Embed Size (px)
Citation preview
IBM Transparent Data Migration Facility(TDMF) for z/OSVersion 5 Release 9
Reference Guide
IBM
SC27-8963-03
Note
Before using this information and the product it supports, read the information in the Notices section.
Edition notice
This edition applies to TDMF version 5.9 and to all subsequent releases and modifications until otherwise indicated innew editions.© Copyright International Business Machines Corporation 2010, 2019.US Government Users Restricted Rights – Use, duplication or disclosure restricted by GSA ADP Schedule Contract withIBM Corp.© Teracloud S.A. 2019, 2020.
Contents
About this guide................................................................................................... xiWho should use this guide.......................................................................................................................... xiPublications and related information......................................................................................................... xiOrdering IBM publications.......................................................................................................................... xiFeedback.....................................................................................................................................................xii
If you have a technical problem........................................................................................................... xii
What's new.........................................................................................................xiii
Chapter 1. Product overview..................................................................................1Non-disruptive data migration.....................................................................................................................1Data migration and migration tool - definitions and characteristics.......................................................... 1TDMF Overview............................................................................................................................................ 2
Benefits offered...................................................................................................................................... 2Master/agent(s) relationship..................................................................................................................3Master system responsibilities.............................................................................................................. 4Remote master system responsibilities................................................................................................ 4Agent system responsibilities................................................................................................................ 4
Phases of migration..................................................................................................................................... 5Phases of migration without the switch back feature........................................................................... 5Phases of migration with the switch back feature.............................................................................. 14
What's new.................................................................................................................................................17
Chapter 2. Installing and customizing.................................................................. 19Pre-installation requirements and checklist.............................................................................................19
Operating system requirements.......................................................................................................... 19Pre-installation checklist..................................................................................................................... 19
Security...................................................................................................................................................... 20System authorization facility................................................................................................................20
Migration sessions with a mixture of maintenance levels........................................................................21Installing.................................................................................................................................................... 21Postinstallation tasks.................................................................................................................................21Postinstallation tailoring............................................................................................................................23z/VM Agent for TDMF z/OS........................................................................................................................ 25
Configuring the z/VM Agent for TDMF z/OS.........................................................................................25Preparing to run TDMF...............................................................................................................................26TDMF system defaults............................................................................................................................... 27
System defaults....................................................................................................................................27Updating system defaults with the GTDSOPTN batch job.................................................................. 30
Verifying that required TDMF agents are running..................................................................................... 30Running TDMF............................................................................................................................................31
Terminating TDMF................................................................................................................................ 32Recovering TDMF devices.................................................................................................................... 32How to check the fence status.............................................................................................................33How to check the DDTSIO pointer.......................................................................................................33
Chapter 3. Using control statements.................................................................... 35TDMF control statements.......................................................................................................................... 35
Required control statements............................................................................................................... 35Optional control statements................................................................................................................ 35
iii
Ordering input of control statements.................................................................................................. 36SESSION control statement...................................................................................................................... 37
SESSION parameters........................................................................................................................... 37SESSION options..................................................................................................................................38
REMOTE control statement....................................................................................................................... 44REMOTE format.................................................................................................................................... 45REMOTE parameters............................................................................................................................ 45REMOTE options...................................................................................................................................45
GROUP control statement......................................................................................................................... 46GROUP format...................................................................................................................................... 46GROUP parameters.............................................................................................................................. 46
Volume options.......................................................................................................................................... 47Full speed copy impact........................................................................................................................ 48
MIGRATE control statement..................................................................................................................... 48MIGRATE format.................................................................................................................................. 49MIGRATE parameters.......................................................................................................................... 49
REPLICATE control statement...................................................................................................................49REPLICATE format............................................................................................................................... 50REPLICATE parameters....................................................................................................................... 50
Common options for the SESSION, GROUP, MIGRATE, and REPLICATE statements.............................50FENcesource/NOFENcesource............................................................................................................ 51UNFENcetarget/NOUNFENcetarget.................................................................................................... 51NOCONFirm/CONFirm..........................................................................................................................52NOPRompt/PRompt............................................................................................................................. 52NOCOMPARE/COMPARE...................................................................................................................... 52NOAUTOOPerations/AUTOOPerations, NOAUTOOPS/AUTOOPS ...................................................... 52NOTERMGRoup/TERMGRoup.............................................................................................................. 52NOFASTcopy/FASTCOPY......................................................................................................................52NOPUrge/PUrge....................................................................................................................................53NOALLOWmirrorchange/ALLOWmirrorchange....................................................................................53NOEXTVtoc/EXTVtoc NOXVToc/XVToc............................................................................................... 53SYNCgoal.............................................................................................................................................. 53NOCOMPRESSion/COMPRESSion........................................................................................................54NOOVA/OVA..........................................................................................................................................54NOPPIT/PPIT or NOPERPetual/PERPetual......................................................................................... 54RESETCHAngedflag..............................................................................................................................54NOPAV/PAV...........................................................................................................................................54OFFLINETarget/ONLINETarget............................................................................................................54OFFLINESource/ONLINESource.......................................................................................................... 54
Chapter 4. Improving performance...................................................................... 55Auto-operations interface......................................................................................................................... 55Dynamic volume pacing.............................................................................................................................55
I/O pacing............................................................................................................................................. 56Real-storage pacing............................................................................................................................. 56Reverse pacing..................................................................................................................................... 56When pacing is not in effect.................................................................................................................57User-specified pacing values............................................................................................................... 57Pacing with channel extenders............................................................................................................ 57
SMF recording............................................................................................................................................ 59Copy functions........................................................................................................................................... 59
Point-in-time migration........................................................................................................................59Point-in-time migration and databases...............................................................................................60
Application I/O performance.....................................................................................................................60Synchronization goals................................................................................................................................60Channel utilization..................................................................................................................................... 61Performance impact by phase...................................................................................................................61
iv
Placement of the master system.............................................................................................................. 62Placement of the communications data set............................................................................................. 62RAID subsystems and rank contention.....................................................................................................63Storage subsystem performance.............................................................................................................. 63Suspend or continue a migration.............................................................................................................. 64The TDMF hang monitor............................................................................................................................ 64Dynamic suspension of a migration.......................................................................................................... 64
Chapter 5. Planning............................................................................................. 67Storage requirements................................................................................................................................ 67DASD space requirements.........................................................................................................................69Activating new hardware I/O configurations............................................................................................ 69CA products................................................................................................................................................69
CA ACF2................................................................................................................................................69CA-ASTEX performance....................................................................................................................... 69CA-OPS/MVS event management and automation............................................................................. 69CA-Scheduler job management...........................................................................................................70CA-1 tape management....................................................................................................................... 70CA-HYPERBUF VSAM buffer optimization........................................................................................... 70CA-7 job management......................................................................................................................... 70CA-11 job management....................................................................................................................... 71CA-MIM resource sharing.................................................................................................................... 71Moving multi-image manager control data sets..................................................................................72
Cache fast write and sort programs.......................................................................................................... 72Class names and esoteric names..............................................................................................................72Coexistence with other TDMF products.................................................................................................... 72Cross memory services..............................................................................................................................73Device states: Before and after a swap migration.................................................................................... 73Data facility products.................................................................................................................................73
Control data sets.................................................................................................................................. 73DFSMS...................................................................................................................................................73
Daylight saving time...................................................................................................................................74DFSort........................................................................................................................................................ 74Duplexing functions................................................................................................................................... 75EMC DASD.................................................................................................................................................. 75
Application program interfaces........................................................................................................... 75EMC symmetrix control facility............................................................................................................ 75Consistency groups.............................................................................................................................. 76EMC configuration guidelines.............................................................................................................. 76Symmetrix manager............................................................................................................................. 76Migrating from SRDF to PPRC/HDS TrueCopy volumes...................................................................... 76SRDF session types allowed................................................................................................................ 77Coexistance with EMC’s z/OS migrator................................................................................................78
Enhanced catalog sharing (ECS)................................................................................................................78Extended functions....................................................................................................................................78File allocation managers............................................................................................................................78Global resource serialization.....................................................................................................................79Host software component......................................................................................................................... 79HCD and Input/Output definition files...................................................................................................... 79HFS and zFS files....................................................................................................................................... 80ICKDSF....................................................................................................................................................... 80
Dynamic ICKDSF REFVTOC/EXTVTOC.................................................................................................80Catalogs or VSAM data sets................................................................................................................. 80How dynamic ICKDSF REFVTOC or EXTVTOC functions affect a migration.......................................80Dynamic ICKDSF REFVTOC function................................................................................................... 81Intent of dynamic ICKDSF EXTVTOC function.................................................................................... 82HSM at APAR level OA21574............................................................................................................... 83
v
IPL volumes............................................................................................................................................... 83Invalid count fields.................................................................................................................................... 83JES considerations.................................................................................................................................... 84
Recommendations for spool or checkpoint volume migration...........................................................84JES2 considerations.............................................................................................................................84JES3 considerations.............................................................................................................................85
LISTDATA information............................................................................................................................... 85Migration of control data sets....................................................................................................................85Migration of stand-alone dump data sets................................................................................................. 86Migration of RAID and non-RAID devices................................................................................................. 86Migration of unlike device types................................................................................................................86Migration of volumes formatted for VM use..............................................................................................87Model 204 considerations......................................................................................................................... 87z/VM Agent considerations........................................................................................................................88MVS running under VM.............................................................................................................................. 88
COMMDS data set considerations....................................................................................................... 89Other control data set considerations................................................................................................. 89Parallel access volumes capable device considerations.................................................................... 89Using devices previously formatted for VM.........................................................................................89VM volume tolerance............................................................................................................................89
Operating parameters of TDMF................................................................................................................. 90Number of TDMF sessions................................................................................................................... 90Number of groups per TDMF session.................................................................................................. 90Number of volumes per TDMF session................................................................................................90
Page and swap data sets........................................................................................................................... 90Parallel sysplex considerations.................................................................................................................91
Migrations outside a single sysplex..................................................................................................... 91Peer-to-peer remote copy support........................................................................................................... 91
PPRC session types allowed................................................................................................................ 91PPRC/GDPS environments with HyperSwap enabled.........................................................................92Migrations between a PPRC device and a SRDF device...................................................................... 92Migrations between a PPRC device and a TrueCopy device............................................................... 92
Real-time monitors.................................................................................................................................... 92Shared device definition............................................................................................................................ 93Enterprise storage server support.............................................................................................................93
Messaging.............................................................................................................................................93Migration from ESS to Non-ESS........................................................................................................... 94Migration from non-ESS to ESS............................................................................................................94FlashCopy support............................................................................................................................... 95
Define an I/O device as static, installation static, or dynamic................................................................. 96Storage space requirements for virtual arrays .........................................................................................97TDMF ownership of target volume............................................................................................................ 97TDMF and volume table of contents......................................................................................................... 97Soft Fence.................................................................................................................................................. 97
DEVSERV command to display Soft Fence status............................................................................... 98SPID Fence.................................................................................................................................................98Suppressing warning messages in TDMF..................................................................................................98
Syntax example.................................................................................................................................... 98XRC support............................................................................................................................................... 99
Extent-efficient and space-efficient volumes..................................................................................... 99XRC remote system data movers.........................................................................................................99XRC in a GDPS/XRC (z/OS global mirror) environment..................................................................... 100
Chapter 6. Using advanced functions..................................................................101Overview.................................................................................................................................................. 101TDMF offline volume access....................................................................................................................101
OVA purpose.......................................................................................................................................101
vi
Starting an OVA session..................................................................................................................... 102Functions of the TDMF interface program.........................................................................................103TDMF interface program processing................................................................................................. 103Using the TDMF interface program....................................................................................................105
Delayed offline volume access................................................................................................................106Purpose of Delayed OVA.................................................................................................................... 106
Perpetual point-in-time...........................................................................................................................108Setting up for PPIT............................................................................................................................. 108During an active PPIT session........................................................................................................... 109
TDMF TCP/IP support..............................................................................................................................110Terminology........................................................................................................................................111TCP/IP Password Option....................................................................................................................111TCP/IP migration requirements and considerations.........................................................................111
TDMF version TCP/IP compatibility........................................................................................................ 113
Chapter 7. Using batch utilities.......................................................................... 115Batch utilities overview........................................................................................................................... 115GTDBMON................................................................................................................................................115
Operational considerations................................................................................................................116Direct TDMF session control..............................................................................................................116GTDBMON - operator command processing.....................................................................................116GTDBMON - SYSIN control statement processing........................................................................... 119GTDBMON messages......................................................................................................................... 121GTDBMON - printed output............................................................................................................... 121
GTDCLIP...................................................................................................................................................121Using GTDCLIP after point-in-time copies........................................................................................ 122Using GTDCLIP after migrations........................................................................................................ 122GTDCLIP - return codes..................................................................................................................... 123
GTDCLUP..................................................................................................................................................123GTDCLUP: scanning the entire system..............................................................................................123GTDCLUP - repairing a single device................................................................................................. 126GTDCLUP - program abend codes..................................................................................................... 126GTDCLUP - program messages..........................................................................................................127
GTDEXTV..................................................................................................................................................127GTDEXTV - program abend codes..................................................................................................... 127GTDEXTV program messages............................................................................................................ 128
GTDMUCB................................................................................................................................................ 128Operation............................................................................................................................................128GTDMUCB - return codes...................................................................................................................129
GTDQDSK................................................................................................................................................. 129GTDQDSK - printed output.................................................................................................................130GTDQDSK - user abends.................................................................................................................... 130
GTDSCAN................................................................................................................................................. 130Run parameter................................................................................................................................... 130Operation............................................................................................................................................131GTDSCAN sample output...................................................................................................................132GTDSCAN - messages, return codes and abend codes.................................................................... 132
GTDTERA..................................................................................................................................................132Sample JCL for GTDTERA.................................................................................................................. 132GTDTERA - user abends.....................................................................................................................133GTDTERA sample output................................................................................................................... 133
GTDVTOC................................................................................................................................................. 135GTDVTOC - printed output.................................................................................................................135GTDVTOC - return codes and abend codes.......................................................................................137
GZDRECON.............................................................................................................................................. 137Syntax................................................................................................................................................. 138Options............................................................................................................................................... 138
vii
Usage..................................................................................................................................................139Examples............................................................................................................................................ 139Processing.......................................................................................................................................... 140Notes.................................................................................................................................................. 140
Chapter 8. Viewing migration information.......................................................... 141Overview.................................................................................................................................................. 141Option 0 - Change or submit replication or migration jobs.................................................................... 142Option 1 - Current sessions: monitor progress...................................................................................... 143
Filter volume pair selection............................................................................................................... 144Option 2- Current sessions: user interaction and status........................................................................145
User interaction and status - information......................................................................................... 146User interaction and status - migration options............................................................................... 147User interaction and status - help panel........................................................................................... 148Start a volume.................................................................................................................................... 149Acknowledge mirror change on swap............................................................................................... 150Set volume synchronize goal............................................................................................................. 150Suspend a volume..............................................................................................................................151Continue a volume............................................................................................................................. 151Terminate a volume........................................................................................................................... 151
Option 3 - Current sessions: display messages......................................................................................152Option 4 - Current sessions: display offline volume access (OVA)........................................................ 153Option 5 - Display performance data of current and past sessions.......................................................154
System information............................................................................................................................154Volume information............................................................................................................................155Volume pacing....................................................................................................................................157Source volume performance data..................................................................................................... 158TCP/IP performance data.................................................................................................................. 159Average device response times.........................................................................................................160
Option 6 - Past sessions: display summary............................................................................................160Option 7 - Past sessions: display details................................................................................................ 161Option 8 - Past sessions: display communication data set history....................................................... 163Option 9 - Display installation options and environment.......................................................................163Option 10 - Display/modify installation security environment.............................................................. 164
Adding authorization keys................................................................................................................. 165Deleting keys......................................................................................................................................165Transactional terabyte license (TTL), terabytes purchased, terabytes moved............................... 166
Option 11 - Display/modify user's TSO monitor options........................................................................166Option 12 - Build replication or migration jobs...................................................................................... 166
Enter/update/review job cards..........................................................................................................167Building a Master system batch job...................................................................................................167Building an Agent system batch job.................................................................................................. 171Building a communications data set (COMMDS)...............................................................................172
Option U - TDMF support utilities............................................................................................................173Option U.0 – Display memory............................................................................................................173Option U.1 – View internal details for active sessions......................................................................174Option U.2 – Active session – display/alter tracing bit settings.......................................................174Option U.3 – Active session – display trace table.............................................................................175Option U.4 – Previous session – display trace table.........................................................................175Option U.5 – Module version levels in defined load library.............................................................. 176Option U.6 – Current/past session module version levels from COMMDS...................................... 177Option U.7 – System change summary............................................................................................. 177Option U.8 – Unit control block display.............................................................................................177Option U.9 – Communication data set control blocks...................................................................... 178Option U.10 – Communication data set navigator............................................................................179Option U.11 – Communication data set volume refresh bit maps................................................... 180Option U.12 – Communication data set merged system messages................................................ 180
viii
Option U.13 – Detected source volume I/O errors........................................................................... 181Option U.14 – TDMFSECURE............................................................................................................. 182
Option S - SMF reporting......................................................................................................................... 183Viewing the bytes migrated report.................................................................................................... 183Bytes migrated report descriptions...................................................................................................186
Option H - TDMF help and message detail facility................................................................................. 187Option H.1 – Display TDMF message details.................................................................................... 188Option H.2 – Introduction to TDMF................................................................................................... 188Option H.4 – Monitor line commands................................................................................................189
Chapter 9. Migrating GDPS/PPRC HyperSwap volumes.......................................191Overview of quiescing and swapping a volume...................................................................................... 191Validation and initialization..................................................................................................................... 191HyperSwap groups.................................................................................................................................. 192GDPS controlling systems....................................................................................................................... 192Suspending the HyperSwap environment.............................................................................................. 193Resuming the HyperSwap environment................................................................................................. 193HyperSwap while TDMF is active............................................................................................................ 194Using the TDMF extended MCS console for HyperSwap commands.....................................................194
Chapter 10. Migrating basic HyperSwap volumes............................................... 195Overview.................................................................................................................................................. 195Validation and initialization..................................................................................................................... 195Blocking HyperSwap processing.............................................................................................................195Resuming the HyperSwap environment................................................................................................. 196HyperSwap while TDMF is active............................................................................................................ 196
Chapter 11. Migrating EMC autoswap volumes....................................................197Overview.................................................................................................................................................. 197Validation and initialization..................................................................................................................... 197Suspending the EMC environment.......................................................................................................... 198Resuming the autoswap environment.................................................................................................... 198System automation (SA) policy............................................................................................................... 198
Appendix A. DASD space requirements and serial numbers................................ 199DASD space requirements...................................................................................................................... 199Determining DASD subsystem serial number.........................................................................................199
Appendix B. Determine the CPU serial number................................................... 201
Appendix C. Authorization return codes............................................................. 203Authority checking...................................................................................................................................203Reading and writing the TDMF security record.......................................................................................204Reading, updating, and writing TDMF – express feature information................................................... 205History file recording............................................................................................................................... 206
Appendix D. Messages for automated operations................................................209
Appendix E. Configurable REXX execs................................................................213Sample REXX exec...................................................................................................................................213
Appendix F. Session examples........................................................................... 215Creating the communications data set................................................................................................... 215
Formula for determining data set size...............................................................................................215Performing a swap migration.................................................................................................................. 215
Single System Swap Migration Session.............................................................................................215
ix
Multiple system swap migration session.......................................................................................... 216HyperSwap volume migration session.............................................................................................. 217Swap migration session with rename................................................................................................217Swap migration session with prompt................................................................................................ 218Swap migration session with unidentified systems tolerance......................................................... 219
Running a point-in-time session............................................................................................................. 219Migrate session with OVA.................................................................................................................. 221PIT session with PPIT........................................................................................................................ 222
Running delayed OVA.............................................................................................................................. 223Performing a SCAN ONLY........................................................................................................................ 223Setting up a migration using TCP/IP....................................................................................................... 224
REMOTE master JCL.......................................................................................................................... 224LOCAL master JCL..............................................................................................................................226Remote agent JCL.............................................................................................................................. 227Local agent JCL.................................................................................................................................. 227
Appendix G. How to read syntax diagrams..........................................................229Reading syntax diagrams........................................................................................................................ 229
Appendix H. Operating the z/VM Agent for TDMF................................................ 231Starting the z/VM Agent for TDMF...........................................................................................................231Querying operational status.................................................................................................................... 231Using a secondary user to communicate with the z/VM Agent for TDMF..............................................232Starting a migration session....................................................................................................................232Disconnecting from an active migration session.................................................................................... 233Terminating the z/VM Agent for TDMF....................................................................................................233Retrieving logging information................................................................................................................ 233Displaying status information................................................................................................................. 234Displaying active migration sessions...................................................................................................... 234Displaying maintenance level of z/VM Agent for TDMF software.......................................................... 235Toggling debug messages....................................................................................................................... 236Changing configuration settings..............................................................................................................236Remotely operating the z/VM Agent for TDMF....................................................................................... 236
Sample JCL to send a command over FTP........................................................................................ 237Sample JCL to issue a command over NJE....................................................................................... 237
Notices..............................................................................................................239Trademarks..............................................................................................................................................240
Index................................................................................................................ 241
x
About this guide
This guide provides implementation and usage information for TDMF. It also provides implementationscenarios.
Who should use this guideThis guide is intended for personnel involved in planning and system administrators who are responsiblefor data migration.
Publications and related information
To view a PDF file, you need Adobe Reader. You can download it at no charge from the Adobe website(get.adobe.com/reader/) .
PublicationsYou can order or download individual publications that have a form number.
Table 1. TDMF 5.9.0 product publications
Title Form number
IBM Transparent Data Migration Facility (TDMF) forz/OS (TDMF) Reference Guide
SC27-8963-03
IBM Transparent Data Migration Facility (TDMF) forz/OS (TDMF) Messages and Codes
SC27-8964-02
IBM Transparent Data Migration Facility (TDMF) forz/OS zDMF and TDMF for z/OS Operating SystemSupport Matrices
SC27-4619-02
IBM Transparent Data Migration Facility (TDMF) forz/OS (TDMF) z/OS Program Directory
GI13-3014-04
z/VM Agent for IBM Transparent Data MigrationFacility (TDMF) for z/OS (TDMF) z/OS ProgramDirectory
GI13-4179-03
Ordering IBM publicationsThe IBM Publications Center is a worldwide central repository for IBM product publications andmarketing material.
Procedure
• The IBM Publications Center website (www.ibm.com/shop/publications/order/) offers customizedsearch functions to help you find the publications that you need. Some publications are available foryou to view or download at no charge. You can also order publications. The IBM Publications Centerwebsite displays prices in your local currency.
© Copyright IBM Corp. 2010, 2019 xi
How to send your comments to IBMWe appreciate your input on this publication. Feel free to comment on the clarity, accuracy, andcompleteness of the information or provide any other feedback that you have.
Procedure
• Use one of the following methods to send your comments:
1. Send an email to [email protected]. Send an email from the Contact z/OS® webpage (http://www.ibm.com/systems/z/os/zos/
webqs.html)3. Mail the comments to the following address:
IBM CorporationAttention: MHVRCFS Reader CommentsDepartment H6MA, Building 7072455 South RoadPoughkeepsie, NY 12601-5400US
4. Fax the comments to us, as follows:
From the United States and Canada: 1+845+432-9405From all other countries: Your international access code +1+845+432-9405
Include the following information:
– Your name and address.– Your email address.– Your telephone or fax number.– The publication title and order number.– The topic and page number that is related to your comment.– The text of your comment.
When you send comments to IBM®, you grant IBM a nonexclusive right to use or distribute thecomments in any way appropriate without incurring any obligation to you.
IBM or any other organizations use the personal information that you supply to contact you only aboutthe issues that you submit.
If you have a technical problem
Procedure
Do not use the feedback methods that are listed for sending comments. Instead, take one of the followingactions:• Contact your IBM service representative.• Call IBM technical support.• Visit the IBM Support Portal at z/OS support page (http://www.ibm.com/systems/z/support/)
xii IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
What's new
IBM Transparent Data Migration Facility (TDMF) for z/OS introduces the following new functions andtechnical updates that were added to this release.
Version 5.9The following lists provides the current technical changes and enhancements to the IBM TransparentData Migration Facility (TDMF) for z/OS (TDMF) for z/OS. Changed and new information is indicated by avertical bar (|) to the left of the change.
Table 2. Summary of changes
Function Description
Improved Reporting Session summary information is provided to make it easier to evaluatemigration actions.
z/OS 2.4 support This version of TDMF now supports z/OS 2.4.
Storage efficiencies A new IOBUFFERPOOL option reduces virtual and fixed storage requirementsduring large migrations.
Version 5.8The following lists provides the current technical changes and enhancements to the IBM TransparentData Migration Facility (TDMF) for z/OS (TDMF) for z/OS. Changed and new information is indicated by avertical bar (|) to the left of the change.
Table 3. Summary of changes
Function Description
NOPRompt/PRompt Updated the default description. For information, see “NOPRompt/PRompt”on page 52.
GTDBMON operatorcommands
The GTDBMON operator commands table was updated to include PROMPToption information. For information, see “GTDBMON - operator commandprocessing” on page 116
Operating systemssupported
This version of TDMF supports the following operating systems:
• z/OS 1.13, 2.1, 2.2 and 2.3• z/VM 6.3, 6.4 and 7.1
Querying operationalstatus: "diagnostic" isreplaced with "diagnose"
In the following paragraph, "diagnostic" is replaced with "diagnose:"
"The z/VM Agent for TDMF uses the TDMF diagnose code that is assignedthrough the DIAG parameter in the TDMFVM configuration file. To verify thatthis diagnose code is enabled, enter the q diag 130 CP command (assuming130 is the hexadecimal number of the diagnose code that is assigned by theconfiguration file):"
TDMF supportsCascaded FlashCopy®
During initialization, TDMF processes the FlashCopy relationship table. Ifeither the TDMF SRC or TGT is the target of a FlashCopy relationship, themigration process is terminated. For information, see “FlashCopy support”on page 95.
© Copyright IBM Corp. 2010, 2019 xiii
Table 3. Summary of changes (continued)
Function Description
APAR OA53265 Starting with APAR OA53265 (applies to Version 5.6 and Version 5.7 and isincluded in the Version 5.8 base), TDMF attempts to restore FlashCopybefore a Soft Fence process.
Version 5.7The following lists provides the current technical changes and enhancements to the Transparent DataMigration Facility (TDMF) for z/OS. Changed and new information is indicated by a vertical bar (|) to theleft of the change.
Table 4. Summary of changes
Function Description
z/VM Agent for TDMFz/OS
The z/VM Agent for TDMF z/OS is implemented as a VM server virtualmachine that can run multiple concurrent migration sessions. See “z/VMAgent for TDMF z/OS” on page 25 for details.
Extent-efficient andspace-efficient volumes
Free space is managed depending on whether the volumes are extent-efficient or space-efficient. See “Extent-efficient and space-efficientvolumes” on page 99 for details.
Soft Fence Added FlashCopy re-enablement action following device soft fencing. See“Soft Fence” on page 97 for details.
SPID Fence If a TDMF termination routine is not invoked because of system error (suchas Force and Wait state), then SPID Fence might be left enabled. See “SPIDFence” on page 98 for details.
Editorial updates Editorial updates were made throughout the documentation.
xiv IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Chapter 1. Product overviewTDMF can be used for migrating data over virtually unlimited distances and the data migration is non-disruptive. Information is provided that describes the process TDMF uses to migrate data.
Non-disruptive data migrationAs data availability demands increase, there is a need to also control costs. The following paragraphsdescribe how z/OS-based TDMF supports the requirements for non-disruptive data migration.
As data storage capacity requirements grow rapidly, data availability demands increase. At the sametime, there is a need to control costs. Data center management faces a dilemma; the introduction of newstorage technology is traditionally disruptive. This dilemma conflicts with the need to ensure a maximumavailability of the data.
What is needed is a tool that allows the customer to non-disruptively relocate or migrate data within thedata center, in periods of full production and demand.
Here, the issues of data movement or migration are discussed. Decades ago, IBM Corporation’s technicaladvisory group (GUIDE) published the requirements and capabilities of efficiently managing andmaintaining storage in a modern large data center. In the interim, IBM Transparent Data Migration Facility(TDMF) for z/OS and third-party vendors created a standard set of tools to allow data to be copied,archived, and restored automatically. Through time, improvements in the areas of performance, usability,and data availability evolved.
The problem remains that the abilities of existing data migration tools do not keep pace with therequirements of today’s data centers. A site’s storage administrator must be able to support continuousaround-the-clock data availability. Although other vendors developed migration techniques, theirimplementations are based on the vendors’ hardware capabilities.
Data migration and migration tool - definitions and characteristicsYou can use the following list of basic definitions and characteristics to familiarize yourself with datamigration and the migration tool.
• Data migration is the copying of data from one device (the source) to another device (the target) andredirecting the I/O to the new device. A request can be for Point-in-Time replication; in this case,volume redirection of I/O operations to the new device does not take place.
• Data replication is the copying of data from one device (the source) to another device (the target) withno I/O redirection: that is, the source device remains online during the copy.
• A migration or replication is the logical relationship between a source and target device.• The user initiates and controls all migrations/replications. The user identifies "from" (source) volumes
and "to" (target) volumes.• Multiple volume migrations and replications can be established during any one session.• The migration tool is dynamically started and stopped.• Applications remain unaware that migration is underway. The data is continuously and fully accessible
for read and write activity.• After replication and synchronization are complete, the takeover of the target device is non-disruptive.
For Point-in-Time migration, the pairing is broken, leaving the target device in an offline state, and thedata within congruent to a specific point in time.
• The tool supports multiple system shared data environments.• The tool ensures complete physical data integrity.
© Copyright IBM Corp. 2010, 2019 1
• The use of the tool is not restricted to any control unit model type or device type. Except as noted, alldevices in the data center can participate in a migration session as required.
• A second Point-in-Time copy can be created on the system simultaneously with the creation of a normaltarget DASD volume or device. This action can be done by specifying two target volume serial numbersin the TDMF REPLICATE control statement.
• Remote target volume. A volume that is used as the target for a Point-in-Time copy migration, which isnot directly accessible by the session’s Master System. TDMF completes the replication by transmittingvolume data to a remote Migration session, through TCP/IP.
• No user can be allocated to a target volume during migration.• A source volume cannot contain an active local page data set or swap data set.• The source and target volumes must be of the same track geometry.
The indicated characteristics represent the ideal of an open and non-disruptive migration facility.Considering the state of the industry today, IBM Corporation believes that TDMF takes the lead in meetingthese requirements.
TDMF OverviewTDMF is user-initiated and controlled, and allows for full system sharing throughout the data center.TDMF provides full access to the data at any point during a migration operation. TDMF supports dynamictakeover on the part of the target device, an important consideration in integrity and completeness ofdata migration. TDMF is completely model- and vendor-independent within the supported architecture.New storage subsystem technologies can be brought into the data center with an absolute minimum ofdisruption.
Benefits offeredTDMF offers full target and source flexibility, installs dynamically, and supports multiple system datasharing environments.
TDMF offers the following benefits:
• The ability to introduce new storage subsystem technologies without significant disruption of service.• Protection of the value of existing storage equipment because all devices at the site can participate in a
migration, except as previously noted.• Designation of any subsystem as either source or target.• Lower continuing costs of operation by maintaining a multiple-vendor, non-model-specific environment.• Conducting Parallel to ESCON or FICON migrations, an important consideration when implementing
RAID storage technologies.
Note: TDMF supports High-Performance FICON for the System z (zHPF) data transfer protocol.• Batch window reduction.
See Figure 1 on page 3 to learn about the principles of host-based data migration as carried out byTDMF.
2 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Stockholm, Sweden Dearborn, MI U.S.
DASD 1 DASD 2 DASD 3
LocalCopy
LocalDASD 1 DASD 2
Remote
RemoteCopy
TDMFTDMF
TDMF
TDMF
TCP/IP Link
f2c0
2159
RemoteCopy
ESCON/FICONDirector2
ESCON/FICONDirector2
Remote Masterz/OS X
Remote Agent 4z/OS Y
Remote Agent 5z/OS YLocal Agent 2
z/OS WLocal Agent 1z/OS V
Local Masterz/OS A
Figure 1. TDMF for z/OS host-based data migration
• TDMF offers full target and source flexibility. All Count Key Data/Extended (CKD/E) capable control unitsin the data center can participate in migration sessions. Functions include Hyper-Volumes and Flexi-volumes.
• TDMF installs dynamically; no IPL is required.• TDMF sessions are parameter-driven.• TDMF asynchronously copies data to target volumes to minimize performance impact.• The migration is invisible to applications. Applications continue to access and update the source
volume. TDMF asynchronously reflects updates onto the target device.• TDMF supports multiple system data sharing environments.• TDMF ensures physical data integrity with ongoing internal heartbeat monitoring, error detection, and
recovery capability.• TDMF ensures that the target device dynamically takes over in a swap migration and disconnects from
the source when the synchronization point is reached.• TDMF supports multiple concurrent migration sessions, each containing its own Communications data
set (COMMDS), and parameter inputs. Only agent systems access the COMMDS, and they have noparameter input.
Master/agent(s) relationshipTDMF is initiated as an MVS batch job or Started Task (STC) on the Master system and all Agent systems.The MVS Job Control Language (JCL) for a TDMF session identifies the parameter input and thecommunications data set (COMMDS). The TDMF control statements identify the source volume(s), the
Chapter 1. Product overview 3
target volume(s), the Master system, the attached Agent system(s), and all other overrides and options.The control statements for a Remote Master job define only the systems participating at the remote site;the remote target volume serial numbers are provided by the Local Master job. The COMMDS allows allsystems that are attached to the source and target volumes to communicate and monitor the health ofthe migration in progress. The COMMDS is also used as an event log and a repository for messages,diagnostic and performance information.
The COMMDS may NOT be allocated upon a volume being migrated within that specific session.
In a session with no remote target volumes, there is only one Master system. Multiple Agent system(s)may be involved in a session either locally or remotely.
Note: TDMF will use the Query Host Access to Volume, SPID fence and Soft Fence hardware features toenhance data integrity protection for devices that support those functions.
Important: Possible Data Integrity Exposure: All systems accessing migration volumes must be identifiedto the Master system. TDMF includes various controls and checks that ensure that the user does notassign or direct conflicting migrations to the same devices, or attempt migrations to non-existent devices,or attempt to use the same Communications Data Set for two simultaneous or overlapping migrationsessions.
Master system responsibilitiesThe Master system is responsible for initializing, starting and controlling each session for all participatingsystems, in addition to the following list of actions.
• Initialize the TDMF Master system environment and the COMMDS.• Start and control each session for all participating systems.• Monitor source volume user I/O activity to detect updates.• Monitor target volume user I/O activity to prevent updates.• Copy data from the source volume to the target volume.• Process detected source volume updates from all systems.• Initiate refresh operations to the target volume to reflect the update activity on the source volume.• Check the internal health of the Master environment and the health of all Agent systems.
Important: Master jobs must be run with a high dispatching priority, especially if they are running inCPU bound LPARs.
Remote master system responsibilitiesThe remote master system starts and controls each session for all participating systems, and monitorsremote target volume access in addition to the following actions.
• Initialize the TDMF Remote Master system environment and the remote COMMDS.• Start and control each session for all participating systems, as directed by the Local Master system.• Monitor remote target volume access to prevent access.• Copy data from the local source volume to the remote target volume.• Check the internal health of the Remote Master environment and the health of all Remote Agent
systems.
Both Local and Remote Master systems react to commands entered by the user through the TSO andbatch monitors.
Agent system responsibilitiesThe Agent system is responsible for initializing the TDMF Agent environment, and checking compatibilitywith the TDMF release in addition to the following actions.
• Initialize the TDMF Agent environment and establish communications to the Master system, by usingthe COMMDS.
4 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
• Check compatibility with the TDMF release and maintenance level of the Master system.• Acknowledge and process migration requests from the Master system.• Monitor source volume user I/O activity and detect updates.• Monitor target volume user I/O activity to prevent updates.• Notify the Master system of source volume update activity through the COMMDS.• Check the internal health of the Agent environments and the health of the Master system.
Important: Agent jobs must be run with a high dispatching priority, especially if they are running in CPUbound LPARs.
Phases of migrationWhen you use TDMF, there are several phases of migration. These phases can be slightly differentdepending on how you set the Switch Back feature. Each phase is described in detail in the followingparagraphs.
The TDMF migration process includes more phases that apply when the APPROVAL option of the SwitchBack feature is selected. For more information, see “Phases of migration with the switch back feature” onpage 14.
If you choose to migrate volumes without the use of the Switch Back feature, the migration phases thatfollow apply.
Phases of migration without the switch back featureInformation is provided that describes the TDMF migration process flow without the Switch Back feature.
The following Figure 2 on page 6; summarizes the TDMF migration process flow.
Chapter 1. Product overview 5
Figure 2. Phases of a TDMF session without switch back
The SYSTEM INITIALIZATION phaseOnly after successful initialization of all systems in a TDMF session does migration proceedasynchronously. If any violation occurs during system initialization on any system defined in the session,no migrations are completed.
Successful system initialization is the result of all participating systems performing error-free validationfor all volumes within a session. If all systems in the session are not started within a 60-minute interval,then the session will not complete system initialization. If a system is started that is not defined as a partof an active session, TDMF will terminate the Master job on the Master system and all started andpertinent Agent jobs (jobs on other systems that are using the same COMMDS as the Master).
If the security option has been selected to provide use of the System Authorization Facility (SAF), and anyvolume involved in the migration session fails SAF, the migration session will fail system initialization.
SAF requirements:
• Swap type migrations require ALTER authority on the source and target volumes.
6 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
• Point-in-Time migrations require READ authority for the source volume and UPDATE authority for thetarget volume.
Volumes in a session may be terminated via the TDMF TSO Monitor or Batch Monitor on the Mastersystem prior to the successful system initialization of all Agent systems.
In addition, if the History option has been selected to automatically record information about themigration session, the actual recording requires UPDATE authority for the data set specified in the Historyoption entry. For more information regarding these functions, refer to “System authorization facility” onpage 20.
The Master system initiates and controls all migrations. The Master initiates each phase and all Agentsmust acknowledge this in order to proceed. If any system detects a violation, that specific migrationterminates. Depending on the state of the current migration, it may then be necessary to perform backoutprocessing.
It is useful to break out a migration into major phases, as explained in the following paragraphs.
The INITIALIZATION phaseAll participating systems confirm the validity of the source and target volumes. Volume acknowledgment,confirmation, selection, and initialization occur before the ACTIVATION phase. Requests requiring volumeacknowledgment, volume confirmation, or specifying a limit to the number of concurrently activevolumes, can delay group and volume selection.Volume acknowledgment
Any volume or group of volumes that require an acknowledgment are not eligible for volumeconfirmation and subsequent volume selection until an acknowledgment is received from the TDMFTSO Monitor, a Batch Monitor, or the MVS Write-to-Operator/Write-to-Operator with Reply (WTO/WTOR).
This acknowledgment is required for volumes being migrated whose disaster recovery or mirrortechniques are not compatible, unless the ALLOWmirrorchange (NOACKnowledge) option is specified.TDMF recognizes volumes that use Peer-to-Peer Remote Copy (PPRC), or Extended Remote Copy(XRC) of all vendors’ DASD, TrueCopy from Hitachi, and Symmetrix Remote Data Facility - with orwithout Consistency Groups from EMC.
The general rule is that the source and target volumes’ type of disaster recovery or mirroring, eithersynchronous or asynchronous, must be compatible. Compatibility also extends to include the vendorscommand interface structure and parameters. If they are not compatible, the ALLOWmirrorchangeoption must be specified and an acknowledgment received.
Volume Confirmation
Any volume or group of volumes that require confirmation are not eligible for volume or groupselection until a confirmation is received from the TDMF TSO Monitor, a Batch Monitor or the MVSWrite-to-Operator/Write-to-Operator with Reply (WTO/WTOR). The order of confirmation determinesthe order of volume selection. In other words, if volumes ABC123 and SYSDBR require confirmation,volume selection for those volumes do not occur until confirmation is received.
Volumes or groups that do not require confirmation are immediately available for volume or groupselection.
Volume Selection
By default all volume pairs that are defined in a session are automatically selected during theInitialization phase. However, volume selection is affected when certain user options are specified.Those options that affect volume selection are volume confirmation, number of concurrent volumes,active in copy, group options and allow mirroring changes. These options are discussed later.
Volume Initialization
Initialization of all volume level control blocks and page fixing of all real storage frames necessary fora volume migration. The source and target volumes are SPID fenced preventing any system fromvarying either device online while their migration is in progress. See Figure 3 on page 8.
Chapter 1. Product overview 7
Volume Soft Fence
The Soft Fence status of target volumes are reset if necessary.
Masterz/OS A
Agent 1z/OS V
f2c0
2161
Figure 3. Initialization Phase
The ACTIVATION phaseThe I/O monitor is activated for the source and target device, at which time all systems in the sessionattempt to allocate each source and target volume, preventing them from being inadvertently variedoffline.
For an illustration of the ACTIVATION phase, see Figure 4 on page 9.
8 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
ISPF Monitor
Masterz/OS A
Agent 1z/OS V
f2c0
2160
Figure 4. ACTIVATION phase
The COPY phaseNext the Master system begins a COPY volume task to copy data from the source volume to the targetvolume. There is an independent COPY volume task for each source volume in the migration session.
During the course of the COPY volume phase, if any of the participating systems detect source volumeupdates, the Master system will collect the updated information to be processed in the copy REFRESHphase.
The REFRESH phaseWhen the COPY volume task completes one pass of the source volume, the Master initiates the copyREFRESH task. During this phase, the target volume receives the updates made to the source volume.Multiple refresh phases will occur until TDMF determines that synchronization of the target volume can beachieved, at which time the Master system will signal a quiesce of the source volume.
The QUIESCE phaseAt this point, the Master system tells all systems to stop all I/O activity to the source volume. When anAgent system receives a quiesce request, the Agent system sends to the Master system the final group ofdetected updates. This is necessary for the Master system to complete synchronization.
For an illustration of the QUIESCE phase, see Figure 5 on page 10.
Chapter 1. Product overview 9
Masterz/OS A
Agent 1z/OS V
f2c0
2162
Figure 5. QUIESCE phase
The SYNCHRONIZE phaseWhen all the systems comply with the quiesce request, copy SYNCHRONIZE begins. At the conclusion ofthe SYNCHRONIZE phase, the Master system disables the I/O monitor and starts the volume REDIRECTphase.
For an illustration of the SYNCHRONIZE phase, see Figure 6 on page 11.
10 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Masterz/OS A
Agent 1z/OS V
f2c0
2165
ISPF Monitor
Figure 6. SYNCHRONIZE phase
The Volume I/O REDIRECT phaseTDMF causes I/O activity for the source volume to be permanently redirected to the target device.
The Master system requests that all systems do a redirect, and confirm that the redirect is successful.Upon successful completion of the Agent system redirects, the Master system rewrites the volume labelsto change the source and target devices serial numbers and completes redirect processing. See Figure 7on page 12.
Original source volumes might be put into the Soft Fence state to prevent any updates to those volumesthat might cause data integrity issues.
Note: In a Point-In-Time migration session, this phase does not take place.
Chapter 1. Product overview 11
f2c0
2168
ISPF Monitor
Masterz/OS A
Agent 1z/OS V
Figure 7. REDIRECT phase
The RESUME phaseImmediately after successful I/O redirect processing, the Master system completes RESUME processingand initiates the RESUME request for the Agent systems, so that user I/O can continue to the volume onits new device. After all systems process the RESUME request, the original (source) device is markedoffline and the session enters TERMINATE phase.
When the volumes in a Point-in-Time copy are resumed, the status of the source device is unchanged andthe online status is reset for the target device(s) instead.
For an illustration of the RESUME phase, see Figure 8 on page 13.
12 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
ISPF Monitor
Masterz/OS A
Agent 1z/OS V
Source
f2c0
2164
Figure 8. RESUME phase
The TERMINATE phaseWhen a volume completes a migration, that volume’s fixed storage is freed for possible reuse within thecurrent session. All dynamic allocations are also removed. The SPID (vary online) fence is removed fromthe source and target volumes.
See Figure 9 on page 14 for an illustration of the TERMINATE phase.
Chapter 1. Product overview 13
Masterz/OS A
Agent 1z/OS V
f2c0
2166
Figure 9. TERMINATE phase
Phases of migration with the switch back featureUse the Switch Back feature to complete a swap migration and alternatively, switch back to the originalconfiguration, copying only the tracks that were changed since the swap migration.
See Figure 10 on page 15 for a summary of the TDMF migration process flow with the APPROVAL option.
14 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Figure 10. Phases of a TDMF session with switch back
The Switch Back function is started by the APPROVAL option. This option will change how a swap type ofmigration works after the redirection or swap and the resumption of I/O. When the Switch Back option isselected, the TDMF session stays active and enters a mode that is called "Approval mode". Duringapproval mode, all I/O to the original target volume (now the new source volume) is monitored and thechanges are recorded. Updating the original source volume (now the new target volume) with the changesdepend upon the type of REFRESH selected.
The session stays in Approval Mode until you either approve or disapprove the new configuration:
• If you disapprove the new configuration, TDMF synchronizes the original source volume to the originaltarget volume, swap the volumes, and terminate.
• If you approve the new configuration, TDMF does the necessary clean-up and ends the session.
The use of the Switch Back feature places the following limitations on the functions of the target volumeduring the approval mode.
Chapter 1. Product overview 15
1. If you are migrating to a larger volume, the VTOC on the target volume must not be refreshed by callingICKDSF until the new configuration is approved (the ICKDSF session option is not allowed formigrations that use the Switch Back option). So, while the target volume is in approval mode, youcannot use the additional cylinders. If you use the additional cylinders on the target volume, youcannot switch back.
2. Any advanced or new function capability of the target volume beyond the source volume is notavailable during the approval mode. The use of functions on the target volume that are not supportedby the source volume would prevent the ability to switch back.
3. Depending upon the mode of refresh that you select, the data on the original source volume ismodified.
The migration goes through the same phases as discussed in the previous section until the REDIRECTphase.
When the REDIRECT phase is reached, TDMF leaves the I/O Monitor in place and redirects I/O to thetarget volume and resume I/O operations. The TDMF session remains active in Approval Mode until youapprove or disapprove the new configuration. During the Approval Mode, all I/O to the target volume ismonitored and the changes are recorded.
Note: When completing a Large Consistency Group migration with REPLICATE (Point in Time) andinitiating an IPL to cut over to the new volumes, you cannot use the Switch Back function because of theIPL. To use the Switch Back function, you must use the MIGRATE (swap) option for your migrations.
The migration goes to the TERMINATE phase when you APPROVE the new configuration. Refer to “TheTERMINATE phase” on page 13.)
If you do NOT APPROVE the new configuration, the migration goes through the phases again to revert tothe original configuration on the source volume.
The REFRESH 2 phaseUpdates to the source volume will be made based upon the REFRESH parameter used with the APPROVEoption. In all cases, the number of volumes active in REFRESH will be limited to the number of volumesspecified with the CONCURRENT option. If you disapprove the new configuration, the refresh of thesource volume will begin, regardless of the REFRESH parameter used with APPROVE option, and thenumber of volumes active in the REFRESH phase will be limited to the number of volumes specified withthe CONCURRENT option.
If an error occurs that requires termination while in REFRESH 2 phase, the migration goes to Terminationand the volume and/or session is terminated. The devices will be boxed to allow you to determine whatthe next steps are without the possibility of further corrupting the data.
The QUIESCE 2 phaseThe QUIESCE 2 phase happens for both approve and disapprove of the new configuration. If youdisapprove the new configuration, the QUIESCE 2 phase will be entered when all volumes reachsynchronization. If you approve the new configuration, the QUIESCE 2 phase will be entered only if thevolume table of contents (VTOC) of the target needs to be refreshed.
The I/O monitor will be deactivated for the target volume once it has been quiesced on that system.
If an error occurs that requires termination while the volume is being quiesced, the migration goes inRESUME 2 phase to resume the I/O activity to the target volume before terminating. The devices will beboxed to allow you to determine what the next steps are without the possibility of further corrupting thedata.
The SYNCHRONIZE 2 phaseWhen all the systems comply with the quiesce request, copy SYNCHRONIZE begins. If necessary theoriginal source volume will be taken out of the Soft Fence state. If an error occurs that requirestermination while the source volume is being synchronized, the migration goes in RESUME 2 phase toresume the I/O activity to the target volume before terminating. The devices will be boxed to allow you todetermine what the next steps are without the possibility of further corrupting the data.
16 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
The REDIRECT 2 phaseOnce the synchronization of the source volume has been completed, TDMF completes the swap of thetarget volume back to the source volume. The target volume might be put into the Soft Fence state.
If one or more Agent systems have completed the swap and the swap fails on another system, the Mastersystem will terminate. The devices will be boxed to allow you to determine what the next steps arewithout the possibility of further corrupting the data.
The RESUME 2 phaseThe RESUME 2 phase is entered after all Agent systems and the Master system have completed the swap.If the new configuration was approved, the target volume VTOC gets refreshed at this time.
Note: This phase is called Resume 2 to make the distinction with the Resume phase when TDMF does notrun in Approval mode. However, there is only one Resume phase in all migrations.
What's newIBM Transparent Data Migration Facility (TDMF) for z/OS introduces the following new functions andtechnical updates that were added to this release.
Version 5.9The following lists provides the current technical changes and enhancements to the IBM TransparentData Migration Facility (TDMF) for z/OS (TDMF) for z/OS. Changed and new information is indicated by avertical bar (|) to the left of the change.
Table 5. Summary of changes
Function Description
Improved Reporting Session summary information is provided to make it easier to evaluatemigration actions.
z/OS 2.4 support This version of TDMF now supports z/OS 2.4.
Storage efficiencies A new IOBUFFERPOOL option reduces virtual and fixed storage requirementsduring large migrations.
Version 5.8The following lists provides the current technical changes and enhancements to the IBM TransparentData Migration Facility (TDMF) for z/OS (TDMF) for z/OS. Changed and new information is indicated by avertical bar (|) to the left of the change.
Table 6. Summary of changes
Function Description
NOPRompt/PRompt Updated the default description. For information, see “NOPRompt/PRompt”on page 52.
GTDBMON operatorcommands
The GTDBMON operator commands table was updated to include PROMPToption information. For information, see “GTDBMON - operator commandprocessing” on page 116
Operating systemssupported
This version of TDMF supports the following operating systems:
• z/OS 1.13, 2.1, 2.2 and 2.3• z/VM 6.3, 6.4 and 7.1
Chapter 1. Product overview 17
Table 6. Summary of changes (continued)
Function Description
Querying operationalstatus: "diagnostic" isreplaced with "diagnose"
In the following paragraph, "diagnostic" is replaced with "diagnose:"
"The z/VM Agent for TDMF uses the TDMF diagnose code that is assignedthrough the DIAG parameter in the TDMFVM configuration file. To verify thatthis diagnose code is enabled, enter the q diag 130 CP command (assuming130 is the hexadecimal number of the diagnose code that is assigned by theconfiguration file):"
TDMF supportsCascaded FlashCopy
During initialization, TDMF processes the FlashCopy relationship table. Ifeither the TDMF SRC or TGT is the target of a FlashCopy relationship, themigration process is terminated. For information, see “FlashCopy support”on page 95.
APAR OA53265 Starting with APAR OA53265 (applies to Version 5.6 and Version 5.7 and isincluded in the Version 5.8 base), TDMF attempts to restore FlashCopybefore a Soft Fence process.
Version 5.7The following lists provides the current technical changes and enhancements to the Transparent DataMigration Facility (TDMF) for z/OS. Changed and new information is indicated by a vertical bar (|) to theleft of the change.
Table 7. Summary of changes
Function Description
z/VM Agent for TDMFz/OS
The z/VM Agent for TDMF z/OS is implemented as a VM server virtualmachine that can run multiple concurrent migration sessions. See “z/VMAgent for TDMF z/OS” on page 25 for details.
Extent-efficient andspace-efficient volumes
Free space is managed depending on whether the volumes are extent-efficient or space-efficient. See “Extent-efficient and space-efficientvolumes” on page 99 for details.
Soft Fence Added FlashCopy re-enablement action following device soft fencing. See“Soft Fence” on page 97 for details.
SPID Fence If a TDMF termination routine is not invoked because of system error (suchas Force and Wait state), then SPID Fence might be left enabled. See “SPIDFence” on page 98 for details.
Editorial updates Editorial updates were made throughout the documentation.
18 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Chapter 2. Installing and customizingUse the following information to prepare to install, customize, and verify the installation of TDMF.
Pre-installation requirements and checklistReview the information in this section before you install TDMF.
Operating system requirementsThis version of TDMF supports the following operating systems:
• z/OS 1.13, 2.1, 2.2, 2.3, and 2.4• z/VM 6.3, 6.4 and 7.1
Pre-installation checklistUse this checklist before installation.
Table 8. TDMF pre-installation checklist
Task Description
1 Storage Subsystems support Extended Count-Key-Data (ECKD) ChannelCommand Words (CCWs)?
Note: Older subsystems (pre-3990) might support a subset of the ECKDCCW, which is not sufficient for use with TDMF for z/OS.
Yes [ ] No [ ]
2 MVS running under VM as a guest machine?
If yes, refer to “MVS running under VM” on page 88.
Yes [ ] No [ ]
3 Is this a JES3 environment?
If yes, refer to “JES3 considerations” on page 85.
Yes [ ] No [ ]
4 Have MIM or GRS parameters been set to support TDMF for z/OS?
See:
• “CA-MIM resource sharing” on page 71;• “Coexistance with EMC’s z/OS migrator” on page 78.
Yes [ ] No [ ]
5 Is CA-ASTEX in use?
See “CA-ASTEX performance” on page 69.
Yes [ ] No [ ]
6 Do the Source and Target volumes have the same volume serial numberacross all systems?
While the device address of a volume might be different between LPARs,the volume serial number must be the same.
Yes [ ] No [ ]
7 TDMF agent running on z/VM?
If yes, refer to “z/VM Agent for TDMF z/OS” on page 25.
Yes [ ] No [ ]
© Copyright IBM Corp. 2010, 2019 19
SecurityIf the installation has a security package such as RACF or ACF2 on the MVS system on which TDMF isinstalled, it is necessary to make the appropriate modifications to the security package in order for TDMFto run properly: specifically, profiles and/or command tables should be checked.
Limiting access to the TDMF authorized library to prevent unauthorized use of the TDMF system can beaccomplished through security packages. The user associated with the GTDSOPTN batch job must haveUPDATE authority for the library pointed to by the GTDKEY DD statement. The Master migration job willalso need UPDATE authority for the GTDKEY file if the authorization keys in use specify volume orterabyte limits. To update authorization keys via Option 10 of the TDMF TSO Monitor, the user must haveUPDATE authority for the SGTDLLIB library, which must have been allocated as GTDKEY in theGTDSOPTN job.
Important: The GTDKEY data sets must be defined as PDS type, not LIBRARY. Failure to do this willgenerate an error when trying to apply the product key.
If the History option is selected, UPDATE authority is required for the data set specified in order for TDMFto update that data set. When viewing the history file (and any COMMDS) via the TDMF TSO Monitor, theuser must have READ authority.
System authorization facilityYou can use the System Authorization Facility (SAF) to override authorization default settings for eachlisted command. For example, swap migrations and Point-in-Time migrations might require differentauthorization settings for source and target volumes.
For those installations that need to use the System Authorization Facility (SAF), this option, VOLUMESECURITY = YES, can be selected by using the GTDSOPTN batch job in SGTDSAMP. For moreinformation on the TDMF system defaults, see “Migration sessions with a mixture of maintenance levels”on page 21, and the table “TDMF system defaults” on page 27.
For a Swap migration, ALTER authority or the equivalent must be in effect for the source and targetvolumes. For a Point-in-Time migration, the source volume must have READ authority and the targetvolume must have UPDATE authority. Error messages are created for all volumes not meeting theserequirements in a session.
TDMF checks for two different types of classes:
• CLASS=DATASET for the COMMDS history data set and the TDMF load library that is defined on theGTDKEY DD statement
• CLASS=DASDVOL for volumes that are allowed in a pairing.
If the authorization keys in use specify volume or terabyte limits, the SGTDLLIB (GTDKEY DD statement)must have UPDATE authority to include the user ID submitting the jobs.
For more information on SAF, see the following guides.
• Security Server (RACF) Security Administrator’s Guide (SC28-1915),• CA-ACF2 Administrator Guide (MVS), or• CA-Top Secret User Guide (MVS).
Important: If any volume is determined by the security mechanism not to be eligible for migration, theentire session ends.
Add the TDMF commands that are listed in the following table, to the CA-ACF2 Restricted Commands List.For more information, see CA-ACF2 Systems Programmer Guide.
Table 9. Commands
GTDID GTDSTAT
20 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Table 9. Commands (continued)
GTDVPST GTDMKEY
GTDSUMM GTDMSGM
GTDMSGS GTDVERS
GTDASSO GTDTRC
GTDHMSG GTDTRD
GTDMSTA GTDTRP
GTDPERF GTDLVL
GTDWTCH GTDMEM
GTDBICK
Migration sessions with a mixture of maintenance levelsWhenever a new TDMF release or PTF changes the way in which information is passed between Masterand Agent systems, the Communications Data set for this PTF is said to be at a different level from that ofthe previous release or PTF. At session initialization, the TDMF Master system sets an indicator in theCommunications Data set to inform the Agent systems how the data set is formatted. If an Agent systemis running a version of TDMF that expects a higher or lower Communications Data set level than indicatedby the Master, the session fails to initialize.
For this reason, the Master and Agent systems must all be at the base level of the same version, or at thesame PTF level, across all participating LPARs.
Installing
Postinstallation tasksComplete the following postinstallation checklist to ensure that you completed the needed steps tosuccessfully install TDMF.
Table 10. TDMF postinstallation checklist
Task Description
1 Plans to migrate IPL, Spool, or Checkpoint volumes?
See “HCD and Input/Output definition files” on page 79, “IPL volumes”on page 83, and “Recommendations for spool or checkpoint volumemigration” on page 84.
Yes __ No __
2 Moving from smaller to larger devices?
See “ICKDSF” on page 80 , “Migration of RAID and non-RAID devices”on page 86, “Migration of unlike device types” on page 86.
Yes __ No __
Chapter 2. Installing and customizing 21
Table 10. TDMF postinstallation checklist (continued)
Task Description
3 Does hlq.IBM.HGTD590.SGTDLLIB have read/write authority?
See “Security” on page 20.
Yes __ No __
4 Will you start the System Authorization Facility (SAF) with TDMF forz/OS?
See “System authorization facility” on page 20.
Yes __ No __
5 Determine the number of LPARs involved for each TDMF for z/OSsession.
See “Number of TDMF sessions” on page 90, “Number of groups perTDMF session” on page 90, “Number of volumes per TDMF session” onpage 90.
6 Determine Master System.
See “Placement of the master system” on page 62.
7 Define all Agent Systems.
8 Is there a File Allocation Manager in use?
See “Placement of the communications data set” on page 62' and “Fileallocation managers” on page 78.
Yes __ No __
9 Is Caching turned off at the volume level before you start a TDMF forz/OS Session?
See “Storage subsystem performance” on page 63, “Cache fast writeand sort programs” on page 72, and “Model 204 considerations” onpage 87.
Yes __ No __
10 Target volumes under DFSMS control?
See “Data facility products” on page 73, “DFSMS” on page 73.
Yes __ No __
11 Source volumes not participating in a duplexing function?
See “Duplexing functions” on page 75, “Peer-to-peer remote copysupport” on page 91, “XRC support” on page 99.
Yes __ No __
12 Source volumes not participating in extended functions?
See “Extended functions” on page 78.
Yes __ No __
13 Source volumes contain active Local page or Swap data set?
See “Page and swap data sets” on page 90.
Yes __ No __
14 Is Model 204 installed?
See “Model 204 considerations” on page 87.
Yes __ No __
15 EMC storage subsystems installed?
See “EMC DASD” on page 75.
Yes __ No __
22 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Table 10. TDMF postinstallation checklist (continued)
Task Description
16 RAID storage subsystems installed?
See “RAID subsystems and rank contention” on page 63 and“Migration of RAID and non-RAID devices” on page 86.
Yes __ No __
17 ISPF Keylist function enabled?
See “Postinstallation tailoring” on page 23, Item “7” on page 24.
Yes __ No __
18 Enable TDMF for z/OS TSO Monitor function by using ISPF Option 6 (TSOcommands)?
See “Postinstallation tailoring” on page 23, Item “7” on page 24.
Yes __ No __
19 Is the TDMF agent running on z/VM?
See “Postinstallation tailoring” on page 23, Item “8” on page 24.
Yes __ No __
Postinstallation tailoringAfter you install TDMF, use the following steps to help you set up and get started with TDMF.
About this task
1. Make the SGTDLLIB library an authorized program library.
If the hlq.IBM.HGTD590.SGTDLLIB is to be placed in the system linklist (LNKLSTxx), the STEPLIB DDcard can be removed from the JCL when you run TDMF. If hlq.IBM.HGTD590.SGTDLLIB is to be placedin PROGxx or IEAAPFxx of SYS1.PARMLIB, then the STEPLIB DD statement is necessary.
Tip: If you want to dynamically add any library to the APF list while the system is running, use the MVSSETPROG command. If you want the SGTDLLIB library to be authorized automatically after an IPL, youmust update the PROGxx member in PARMLIB.
2. Allocate the Communications data set.
The Communications data set (COMMDS) is used to pass information between systems thatparticipate in a TDMF session. This data set contains the status and messages that are related to aspecific session. The COMMDS also serves as the input file to the TDMF TSO monitor.
Sample JCL to allocate the COMMDS is in hlq.IBM.HGTD590.SGTDSAMP(GTDALCM). This data setmust be physically on a cylinder boundary with contiguous space. The data set must be on a devicethat supports CKD/E. The COMMDS cannot be on a volume that is involved in the same session. Referto “Point-in-time migration and databases” on page 60.
TDMF periodically sends a RESERVE macro for the COMMDS to serialize communication between theMaster and Agent systems. Refer to “CA-MIM resource sharing” on page 71 or “Coexistance withEMC’s z/OS migrator” on page 78 for details.
The size (number of required cylinders) of the COMMDS can be estimated by using the GTDSPACEREXX command found in the SGTDELIB library. From ISPF Option 6, issue GTDSPACE for help with theGTDSPACE command. If the SGTDELIB library is not concatenated with SYSEXEC or SYSPROC, use theEXECUTE form of the TSO command, for example: ex'hlq.IBM.HGTD590.SGTDELIB(GTDSPACE)'.
Alternatively, when TDMF is run with the SCAN option, TDMF issues message GTD1084I that displaysthe COMMDS space requirements for the set of volumes in the session.
3. Allocate the COMMDS history file. In this optional step member GTDAHIST in SGTDSAMP creates thisfile. Retain the data set name for use in the GTDSOPTN batch job (“System defaults” on page 27).
Chapter 2. Installing and customizing 23
4. Modify member GTDTDMF in data set hlq.IBM.HGTD590.SGTDLLIB, which was unloaded in theinstallation step. Enter the correct high-level qualifier.
If the SGTDELIB library name is to be changed, then modifications must be made to memberGTDTDMF within that library (see Appendix E, “Configurable REXX execs,” on page 213 for anexample). The TDMF TSO Monitor can create two more members in the user’s ISPF Profile data set(userid.ISPF.ISPPROF). They are as follows.GTDEDIT
This member is normally present, but not required. It contains the information necessary for theEDIT windows that the TDMF TSO Monitor presents. This member is created the first time that theuser edits a member that references Chapter 8, “Viewing migration information,” on page 141, oruses the windows within “Transactional terabyte license (TTL), terabytes purchased, terabytesmoved” on page 166.
GTDPROFThis member is normally present, but not required. It contains information that the TDMF TSOMonitor saves for use between sessions. This member is created for the first-time navigation byusing the Monitor windows. And it is done where information might be needed for subsequentsessions.
Note: If the TSO profile prefix is set to a specific user ID, the following statement in memberGTDTDMF of hlq.IBM.HGTD590.SGTDLLIB must appear as secty ="‘hlq.IBM.HGTD590.SGTDLLIB’" otherwise, the monitor fails to start properly. If the TSO profileis set to NOPREFIX, either single quotation marks or double quotation marks followed by singlequotation marks can be used.
5. Ensure that GTDLKEY ran to authorize the product as described in the PGMDIR. If you are runningTDMF for the first time, request authorization keys by contacting the technical support team [email protected].
6. Be prepared to provide the following information:
• CPU serial numbers and CPU model numbers. See Appendix B, “Determine the CPU serial number,”on page 201.
• DASD subsystem serial numbers. See “Determining DASD subsystem serial number” on page 199.• Site ID or number• Company name• Location of site• Name• Phone number
7. To start the monitor feature of TDMF, enter the following command from Option 6 of ISPF: EXEC‘hlq.IBM.HGTD590.SGTDELIB(GTDTDMF)’. If the keylist option is not enabled, the TDMF TSOMonitor issues a message that the keylist is not active.
Tip: To enable the keylist function, select Option 0 of ISPF. At the top of the screen, ‘FunctionKeys’ is displayed. Place the cursor on the "F" and press Enter. If keylists are enabled, the last entryshows an "*0". If keylists are not enabled, the number 10 is displayed.
8. To use a TDMF Agent on z/VM, configure the TDMF server virtual machine after you install the z/VMAgent for TDMF z/OS software.
a. Copy the file GTDVPROF EXEC from the product disk to the virtual machine's A-disk as PROFILEEXEC.
b. Copy the GTDVMCNF CONFIG from the product disk to the virtual machine's A-disk as TDMFVMCONFIG and update. See “Configuring the z/VM Agent for TDMF z/OS” on page 25.
24 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
z/VM Agent for TDMF z/OSThe z/VM Agent for TDMF z/OS is implemented as a VM server virtual machine that can run multipleconcurrent migration sessions.
Each migration session uses a single dedicated COMMDS that is stored on an MVS formatted volume(containing a valid VTOC) which is shared among all host systems that participate in a migration.
When you allocate the COMMDS, special considerations must be taken when MVS is running under VM orwhen a TDMF agent is run on z/VM. Also see “MVS running under VM” on page 88 for details.
Under z/VM, the TDMF Agent uses real Reserve/Release CCWs to serialize access to the SYSCOM dataset. To provide a consistent serialization mechanism, the z/OS Master and Agents must also use Reserve/Release. Therefore, the TDMFRESV resource for the SYSCOM data set that is used in a migration with VMagents must not be converted to global ENQs. For further information, refer to “Global resourceserialization” on page 79.
If the TDMF Master on z/OS does not use Reserve/Release CCWs, the TDMF Server virtual machine onz/VM issues message GTD1286S on its virtual console:GTD1286S Communications Dataset serialization error.
Note: Message GTD1286S is not visible to the TDMF master on z/OS or the TDMF monitor running onz/OS, as the TDMF Agent on z/VM cannot update the COMMDS because of the serialization problem.However, the TDMF Server virtual machine’s console log can be obtained with the GETCONS command.
The COMMDS must be a allocated on a volume which is defined as SHARED in the z/VM systemconfiguration. It is recommended that you place the COMMDS on a volume which is online and free on thez/VM system. The volume is automatically attached to the TDMF Server virtual machine when a session isstarted by using the RDEV or VOLSER parameters of the SESSION command. The volume is automaticallydetached after the last migration session that is using it completes. Because real Reserve/Release is usedto serialize access to the COMMDS, it is also highly recommended that you do not put any other data setson the COMMDS volume.
Note: If you place the COMMDS on a fullpack minidisk defined for the TDMF Agent for VM Server Virtualmachine, the underlying real device must be defined as SHARED in the z/VM system configuration and theV suffix must be used on the MDISK statement in the user directory in order for VM to issue real Reserve/Release CCWs.
Configuring the z/VM Agent for TDMF z/OSConfigure the z/VM Agent for TDMF by modifying a configuration file that is associated to the DD nameCONFIG.
About this task
The default PROFILE EXEC assigns the CMS file TDMFVM CONFIG to the DD name CONFIG. The defaultPROFILE EXEC is distributed as GTDVPROF EXEC and copied to the TDMF/VM Server virtual machine's A-disk during the post-installation step.
The format of a configuration statement is: parameter=value.
When you are configuring the TDMF Agent for z/VM, the most important parameters which you usuallymust modify are the SMFID and DIAG parameters:SMFID
A 1-4 character pseudo-SMFID that is required to be assigned to a z/VM Agent for TDMF.
The pseudo-SMFID is used on the AGENTS session parameter on the TDMF master job on z/OS toidentify this z/VM Agent for TDMF.
DIAGSpecifies the decimal number of the CP DIAGNOSE code to be used for the z/VM Agent for TDMF. Thedefault is 304 (x ’130’).
Chapter 2. Installing and customizing 25
CMD
Defines the CP command name to be used to issue commands to the TDMF/VM I/O monitor. Thedefault is SMON.
DUMP_USERIn case an abend occurred during execution of the z/VM Agent for TDMF, an LE (languageenvironment) dump and a VM dump are transferred to the specified user's virtual machine. Thedefault is the system defined userid for dumps.
LOG_USERSpecifies the default userid to which the console log is sent in case the GETCONS request was issuedlocally or from a remote system which did not transfer sufficient sender information.
Note: The TDMF Server virtual machine tries to send the console log to the requestor (issuer of theGETCONS command) if the command was issued over an NJE connection (via RSCS).
LOG_NODESpecifies the default remote NJE node id which together with the LOG_USER parameter defines thedestination to which the console log is sent in case the GETCONS request was issued locally or from aremote system which did not transfer sufficient sender information.
Sample configuration file (TDMFVM CONFIG)
* assign VM1 as the agent identifier for this z/VM systemSMFID=VM1 * the z/VM Agent for TDMF uses CP diagnose x'130' (d'304)DIAG=304
Note: Refer to Appendix H, “Operating the z/VM Agent for TDMF,” on page 231 for information onoperating the z/VM Agent for TDMF.
Preparing to run TDMFConsider the following information as you prepare to run TDMF.
Member GTDSOPTN is used to install authorization keys and set the TDMF system defaults. Authorizationkeys are provided by email and the exact statements that are provided must be used in the GTDSOPTNjob.
The TDMF 5.7.0 load library (hlq.IBM.HGTD590.SGTDLLIB) can be updated with the GTDSOPTN batchjob. The GTDSOPTN batch job reinitializes the security record with PARM=NEW or update the securityrecord by using PARM=UPDATE.
Review and select the TDMF System Defaults listed in the tables, “TDMF system defaults” on page 27.This applies to those installations that use TDMF for demonstrations or services.
Member GTDMASTJ in SGTDSAMP contains JCL to run on the Master system. Change the data set namesto the correct names for STEPLIB, GTDKEY, and system communications (SYSCOM or COMMDS) data set.
The first control statement in the SYSIN input file is the SESSION statement (see “Ordering input ofcontrol statements” on page 36). The SESSION statement provides the Master and Agent systemidentifiers for the migration, and might set session-wide defaults for a range of migration options. Thesupplied Master system ID must match the SMF identifier of the system on which the Master job isrunning. Agent jobs must be started on all of the systems that are specified in the Agent parameter list.
All systems that have access to the source and target volumes must be included. TDMF does not monitorI/O from a system that is not specified on the SESSION statement. Both source and target volumes mustbe online to the Master system. Also, either both online or both offline on any Agent system. Although, fora Point-in-Time migration, the source volume can be online and either or both of the target or duplexvolumes can be offline to an Agent system.
26 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Member GTDAGENT in SGTDSAMP contains the JCL to run on agent systems. Change the data set namesto the correct names for STEPLIB, GTDKEY, and SYSCOM data set. This JCL must be submitted from allAgent systems, which are specified in the Master system JCL, otherwise, the session does not start.
Important: The TIME=NOLIMIT (or the equivalent TIME=1440) allows the job to use unlimited processortime and disables SMF job wait detection for the job. Refer to members GTDMASTJ and GTDAGENT inSGTDSAMP for examples. For more information on the TIME parameter, see the MVS JCL Referencemanual.
TDMF system defaultsThe following table describes the system defaults, which can be set by running the GTDSOPTN batch job.
Some of the settings might not apply if certain features are not enabled by the authorization keys in use.
A sample GTDSOPTN batch job is supplied in member GTDSOPTN.
The GTDSOPTN job is also used to install the authorization keys. When you initially request yourauthorization keys, contact technical support. You will receive an email that contains the relevantauthorization keys and key installation instructions.
Important: The GTDKEY data sets must be defined as PDS type, not LIBRARY. Failure to do thisgenerates an error when you try to apply the product key.
System defaults set in previous versions or releases of TDMF can be displayed by using the TDMF TSOMonitor and selecting “Average device response times” on page 160.
Note: Data Sets used for Versions 1 and 2 are not compatible with Versions 4 and 5.
System defaultsUse this list of system parameters with their options and default values to help you configure TDMF.
Table 11. TDMF system defaults and options
Parameter Function Options Default
PARM=
(on the EXECPGM= statement)
NEW specifies that this is the first time theGTDSOPTN batch job is being run. Selection ofthis value rewrites the entire key record.
UPDATE specifies that an entry is being updated;for details, refer to “Updating system defaultswith the GTDSOPTN batch job” on page 30.
NEW UPDATE No default
TDMF VERSION Must be the first non-comment entry. Specifieswhat version of TDMF is being run.
1, 2, 4, 5 Must bespecified
KEYxx= Authorization keys, provided by IBM, in order forTDMF to run.
Additionally, starting in column 38, description ofwhat key is related to, for example, CPU, might bespecified (up to 20 characters).
The 16-digit key that isprovided by IBM.
Note: When you initiallyrequest yourauthorization keys byusing the TDMFTechnical Support page,IBM sends you an emailwith the relevantauthorization keys, andthe key installationinstructions.
Must bespecified
Chapter 2. Installing and customizing 27
Table 11. TDMF system defaults and options (continued)
Parameter Function Options Default
CORPORATION = Customer Corporation name. Used in Monitordisplay and sysout listings. Up to 64 charactersallowed.
Customer name Corporation namemust bespecified
SITE NAME = Local customer site name. Up to 64 charactersallowed.
Site Name Site namemust bespecified
SITE NUMBER = The 5-digit customer number. Used in Monitordisplay and sysout listing.
Site Number Sitenumbermust bespecified
SYSCOM HISTORYDATASET NAME =
Record all COMMDS usage in all TDMF sessions.These entries can be accessed by using the TSOMonitor (see “Average device response times” onpage 160 for details). Must have UPDATEauthority.
To remove a data set entry, specify DELETE.
Data Set name
Blank
No historylogging
VOLUMESECURITY=
If a security package is installed, it is possible tohave volume level security by using (SAF) calls.See “System authorization facility” on page 20.
YES NO No SAFcalls
WTO MESSAGESFOR AUTOMATEDOPERATIONSREQUIRED=
If an automated operations package is in use andyou want TDMF use it, TDMF displays specificmessages that are related to start, confirmation,prompts for synchronization and termination onthe system console. See Chapter 4, “Improvingperformance,” on page 55 for more detail.
YES NO No WTO orWTORissued toconsole
AUTOMATICICKDSF =
If "YES" is specified, the ICKDSF program is calledwhen necessary to synchronize a migratedvolume's VTOC with its target device. This settingis ignored for APPROVAL mode migrations.Dynamic ICKDSF running is not supported forAPPROVAL mode migrations.
YES NO NO
CHECK TARGETEMPTY =
If "YES" is specified, target volumes is examinedduring session initialization and the migration jobfails if any contains user data.
YES NO NO
ACTIVE IN COPY=
If "YES" is specified, only volumes in the copy orfirst refresh phases is considered as "active". Thisaffects only migrations in which volume activationis delayed by the setting of a maximum concurrentvolumes limit.
YES NO NO
UNIDENTIFIEDSYSTEMS ACTION=
Determines what action is taken by TDMF when asource volume is found to which logical paths areestablished by a system that is not defined in theSESSION control statement. For moreinformation, see “Updating system defaults withthe GTDSOPTN batch job” on page 30.
IGNORE WARN ERRORTERMINATE
TERMINATE
28 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Table 11. TDMF system defaults and options (continued)
Parameter Function Options Default
OVAREGISTRATIONINTERVAL =
For every OVA volume, provides a time interval (inminutes) after completion of the Point-In-Timemigration during which the 15-minute OVAinactivity timeout does not take effect.
1 to 4-digit number
(max is 1440)
0
WTO AUTOOPERATION MVSROUTCDE =
Write to Operator MVS route codes. Values are 1 -28. Multiple values can be specified and must beseparated by a comma and ended with a ‘)’.
(x, x, x) WTO routecodes 2, 4,6 and 11
SMF RECORD ID=
User specified SMF record to be created for SMFrecording purposes. Valid values are 128 - 255, or0 for no SMF recording (see “Pacing with channelextenders” on page 57 for details.)
128-255
0
0 (No SMFrecording)
DISPLAY TIMEAS =
TDMF defaults to the GMT time display on sysout.It can be set to local time. For information onsetting displays to local time, see “Transactionalterabyte license (TTL), terabytes purchased,terabytes moved” on page 166.
GMT LOCAL GMT
VOLUME PACINGREQUIRED =
If pacing is selected, TDMF dynamicallydetermines the impact to real storage on theMaster system and I/O operations on all sourcevolumes you select in the session. If there are anyimpacts, attempts will be made to decrease theimpact upon the resource dynamically and thenreturn to normal levels after the resourcebottleneck is alleviated. See Chapter 4,“Improving performance,” on page 55.
YES
NO
NO
ALLOW INVALIDCOUNT FIELDS =
If "YES" is specified, the presence of non-standard record count fields do not causetermination of a volume migration.
YES NO NO
STARTUPCONFIRMATIONREQUIRED =
A volume confirmation screen with warningmessages (if applicable) for review and approvalbefore a volume migration proceeds.
Note: If WTO MESSAGES FOR AUTOMATEDOPERATIONS REQUIRED = is set to yes, a Write-to-Operator with Response (WTOR) is sent to thesystem console in addition to the TDMF TSOMonitor.
YES NO No volumeconfirmation
TERMINATE ALLVOLUMES INGROUP ON ERROR=
If an error on any volume occurs within a specific"group" of volumes, the entire group stops.
YES NO No grouptermination
REVERSE PACING=
If "YES" is specified, volumes that are migratedwith the pacing option begin the copy phase byusing single-track buffers. The buffer sizeincreases only if volume activity and performancemeet TDMF criteria.
YES NO NO
Chapter 2. Installing and customizing 29
Updating system defaults with the GTDSOPTN batch jobThe TDMF 5.7.0 load library (hlq.IBM.HGTD590.SGTDLLIB) can be updated with the GTDSOPTN batchjob. The GTDSOPTN batch job updates the security record by using PARM=UPDATE.
About this task
You must be aware of the guidelines that follow when you use the PARM=UPDATE process:
• You must always include the version statement, such as TDMF VERSION 5.• You must NEVER include the key type statement. TDMF REGULAR LICENSE is an example. You cannot
change the license type with this process.• You must NEVER include the FEATURES statement; changing the features invalidates the keys.• If some keys are being changed (adding or deleting a key), all the keys must be included, else only the
keys that are supplied are noted in the security record. This ability is also available with the TSO MonitorOption 10 panel (see “Average device response times” on page 160).
• Only the system default parameter that requires changing needs to be included. If defaults are omitted,the previous settings of the key record remain.
• The effective date of the key (if there is one) is not considered with the PARM=UPDATE process.
Verifying that required TDMF agents are runningTo maintain data integrity, it is essential that a TDMF Agent job is running on all systems that have accessto the migration volumes. The UNIDENTIFIED SYSTEMS ACTION option can be used to verify that thecorrect number of Agents is running.
About this task
An unidentified system is one without a TDMF session and established path to the device. ForTERMINATE, any active parallel system with Master or Agent has migration volume access. An Agentsystem must be active on the system. Instead, one can be specified in the IGNORESYStems option.Otherwise, the session will be aborted.
At session initialization on the Master system, TDMF obtains the path status information from storagecontroller devices having feature 3990-6 compatibility or Query Host Access to Volume. Migrations ofvolumes attached to this controller are terminated in the following conditions:
• Unsupporting storage controller for a Read Subsystem Data command option• Unidentified paths from the Master and all Agent systems• Requested TERMINATE action
The IGNORESCUs(scu-serial) SESSION control statement option is required to migrate volumes awayfrom this storage controller while UNIDENTifiedsystems(T) is in effect.
TDMF checks the data returned from the storage controller to ensure that Agents are active on all thesystems that have established lpaths to the device.
Those processors not running an Agent are listed under message GTD2382I. When an UNIDENT valueother than IGNORE is specified all path data is read and the list of processors is complete as of the timethe query was made. When UNIDENT(IGNORE) is specified, only the path data that will fit in TDMF’sdefault 900 K I/O buffer are read and the list of processors may be unreliable. Message GTD1163I willappear when TDMF does not read all logical path data. Query Host Access to Volume (QHA) data willalways fit in the buffer therefore the list will be accurate when QHA is supported on both the source andtarget devices. To determine if a device supports QHA or display the QHA data for a device, issue theconsole command DEVSERV QDASD,uuuu,QHA
The action taken depends upon the value set for the UNIDENTIFIED SYSTEMS ACTION setting in theGTDSOPTN job, unless this is overridden in the SESSION control statement options list.
30 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Important: Be aware that some storage controllers do not provide path status data, and the support maybe vendor-dependent. To check whether the data is available, run ICKDSF with the ANALYZE NOSCANNODRIVE option. If logical path status data is available, a logical path status table will be printed. Thelpath status data identifies the CPU Serial and Model number of the systems connected to that volume.TDMF cannot determine whether the volume is online to those systems.
Storage controllers maintain system host information for volumes that are varied offline to an LPAR, andeven for LPARs that are not currently active. To allow the use of the TERMINATE parameter on theUNIDENTifiedsystems option when the hardware returns spurious data, the IGNORESYStems optioncan provide TDMF with up to 48 specific CPU serial numbers that might be present in the controllers' pathdata without an active Agent system having a matching CPU ID, while still requiring an active Master orAgent system for every other active host. If the specified action is TERMINATE, the path data will bechecked again, after volume synchronization.
Following path data validation on the TDMF Master system, but only if the requested action isTERMINATE, the Master or Agent job running on one system in each parallel sysplex that comprises theTDMF session obtains the list of the sysplex's active systems. The session will terminate unless there is aTDMF Agent job associated with each active system, or the CPU ID has been provided in theIGNORESYStems option list.
Running TDMFYou can run a single TDMF session or multiple TDMF sessions to submit jobs.
About this task
You can submit TDMF batch jobs, or STCs, in any order. However, you must submit the Master batch jobfirst. If you submit an Agent batch job before the Master batch job, the Agent job waits until the Masterstarts.
Note: Each TDMF session must use a unique COMMDS for reasons of history logging, audit trails,diagnostics, and messages.
You have a 60-minute window after you submit any job to submit any other batch jobs. The session stopsafter 60 minutes if the Master or any attached Agent job is not properly initiated for any reason. Thisincludes any Master or Agent systems that initiated successfully.
Important: If this situation occurs, the entire session must be restarted.
If multiple sessions are to be run, each session must have a unique COMMDS. As previously stated anactive COMMDS volume might not participate in the same TDMF session, but might be migrated in aseparate TDMF session. However, the volume that contains the SGTDLLIB might participate in a session.
Warning messages display during a migration session for the following reasons:
• Migrating a UCB below the 16 MB line to a UCB above the 16 MB line.• Migrating a three-digit UCB to a four-digit UCB.• If the alternate cylinder count does not match on the source and target volumes (infers that an ICKDSF
REFVTOC job needs to be run if Dynamic ICKDSF is not selected).• If migrating an ESCON channel attached volume to a parallel channel attached volume.
Note: The nonzero step completion code that is associated with these warning messages can besuppressed by specifying the message numbers in the NOWARNING session option.
Important: If multiple TDMF sessions are active, the MVS systems must have an active Global ResourceSerialization (GRS) facility unless all Masters run on the same LPAR.
Chapter 2. Installing and customizing 31
Terminating TDMFThrough the TDMF TSO Monitor or Batch Monitor, you can terminate a specific volume pairing, volumegroups, or all volumes within a session dynamically.
About this task
Using the TDMF TSO Monitor or Batch Monitor, it is possible to terminate a specific volume pairing,volume groups, or all volumes within a session dynamically. When termination of a volume pairing isrequested, processing by TDMF can take up to a minute. This delay occurs unless the Master and all theAgent systems are connected with XCF or TCP/IP. See Chapter 8, “Viewing migration information,” onpage 141 for more details.
Note: If a TDMF session (Master or Agent system) must be terminated for whatever reason, the use ofTDMF TSO Monitor is not advised.
Under extreme conditions, the MVS CANCEL command can be used. If the Master system fails leaving anAgent system active, allow the 15-minute interval to expire so that the Agent system can shut downautomatically. It is, however, possible to cancel the Agent system job.
If a TDMF session still appears to hang after an MVS CANCEL command is submitted, resubmit DISPLAYGRS,C. Or use other MVS commands to determine whether canceling a different job might break thedeadlock.
When a TDMF job is canceled, the program attempts to signal the other systems in the session toterminate active volumes and to clean up resources. This processing might run into the same situationthat caused the original hang to occur.
Canceling for a second time causes TDMF to abandon any attempt at an orderly termination, but can alsopermanently break whatever lockout condition caused the hang. During the termination from a secondcancel or a FORCE command, the TDMF RTM Resource Manager program receives control. Normally, onlythe vital clean-up activities are undertaken for a resource manager routine; which means that UCBs areunquiesced and the TDMF I/O Manager is unhooked. If a failed Agent system completed a swap, but theMaster had not, the two devices are varied offline and boxed on the Agent system. In this case, theapplications that use the source volume on that system end and the devices can then be varied onlineagain. Obviously, if the failure occurred when a z/OS system volume was only partially swapped, it mightbe necessary to IPL again.
If a FORCE or second CANCEL command is submitted, the other participating jobs will shut down after 15minutes, or can be canceled individually.
Limited cleanup takes place after multiple CANCEL or FORCE commands. Therefore, the GTDCLUPprogram must run to recheck all the UCBs, re-enable PAV and FlashCopy, and free the common storagethat was acquired by the failing TDMF session. GTDCLUP attempts to write control block, trace, andmessage information to the Communications data set. Thus, it must be run before the data set isunloaded and sent to the support center and on every system where errors occurred.
For instructions for running the GTDCLUP program, see “GTDCLUP” on page 123.
Recovering TDMF devicesIf a TDMF session was canceled and the source or target volumes were left with an invalid DDTSIOpointer or in a quiesced state, you can recover these volumes without an IPL.
Procedure
1. Resubmit the original JCL for the session with the PARM=RECOVERMASTER or PARM=RECOVERAGENTspecified, which causes TDMF to attempt to correct the UCBs for those volumes.
To use this recovery technique, the following conditions must be true:
• The original COMMDS must be available.• No VARY ONLINE | OFFLINE commands were issued to the volumes.• The original JCL must be available.
32 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
2. If the devices cannot be recovered by using the above technique, the GTDCLUP Batch Utility can beused. Additionally, if TDMF failed to pagefree or the release common area virtual storage (SQA/ECSA/HVCOMMON), the GTDCLUP utility can also be used to recover that virtual storage. For details onGTDCLUP, see “GTDCLIP - return codes” on page 123.
3. In case TDMF has left a device soft fenced or SPID fenced, the fence state can be cleared by usingICKDSF. If the device was originally FlashCopy enabled, FlashCopy must be re-enabled after the fenceis cleared. Either run GTDCLUP with PARM=FLASHCOPY or enter the z/OS console command DEVSERVDIAG,uuuu,nn,ALLOWFC (where uuuu is the device number and nn is the number of devices thatstart with uuuu that are to be reset.
4. If the device is offline, it cannot be brought online without clearing the fence state first. This processcan be an issue when IPLing systems that expect access to volumes involved in a migration. If theaffected volume is needed for successful IPL, either wait until the migration is complete or end themigration.
5. The following sample JCL can be used to clear the fence state of a device. Replace the devn by thedevice number. Serial is the last 5 digits of the 12-digit subsystem image serial number, which can beobtained by entering DEVSERV QDASD command; choose SOFT or SPID for the fence to clear):
//STEP EXEC PGM=ICKDSF,PARM='NOREPLYU'//SYSPRINT DD SYSOUT=*//SYSIN DD *CONTROL -UNIT(devn) -[SPID | SOFT]-CLEARFENCE -SERIAL(serial) -SCOPE(DEV)
How to check the fence statusUse the following z/OS console command to display the fence status of a device:
Procedure
DEVSERV PATH,devn
How to check the DDTSIO pointerUse the following steps to check the DDTSIO pointer.
1. Start the TDMF TSO Monitor on the system where TDMF was not properly ended.2. Select option U, then Select Option 8 - Unit Control Block (UCBs) Display.3. In the command line, enter LISTUCB xxxx
where xxxx is the device address of any suspect volume.4. From the command output, find the memory dump of the DDT section of the UCB. It might be
necessary to scroll forward.
Valid DDT exampleUse the following as an example of DDT. If the DDT "eye catcher" at offset X'0' displays DDTR in theEBCDIC conversion, then the volume was not properly ended.
00FC1BE4 C4C4E340 00000000 EF740000 00200000 *DDT ............*DDT C9C5C3D3 01013448 0100EA40 0100CD78 *IECL....... ....* 00000000 01017008 00FD0D68 93940A78 *............lm..* 01019728 00000000 017DB210 C4C4D9C4 *..p.........DDRD* 8100D160 40404040 01015210 *a.J. .... *
To recover from this situation, refer to “Recovering TDMF devices” on page 32.
Chapter 2. Installing and customizing 33
34 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Chapter 3. Using control statementsTDMF control statements can be either required or optional. These statements must be input in arequired order. The information that is covered here includes definitions of each control statement inaddition to examples.
TDMF control statementsTDMF includes control statements such as SESSION, REMOTE, GROUP, MIGRATE, and REPLICATE.Syntax information is provided when you use these control statements.
The five TDMF control statement record types are used with the following syntax rules:
• Each control statement can continue over multiple input records.• The control statement ends at end of file, or when a new control statement is recognized.• The first 71 character positions of each input record are assumed to contain control statement
information. However, if an asterisk is found in the record only the character positions to its left (if any)are parsed.
• Individual fields in a statement, and items within bracketed lists, must be delimited by spaces orcommas.
• Specify only the options you need on the control statements. However, if a particular option is notspecified, the defaults that are specified in the GTDSOPTN batch job take precedence.
Required control statementsRequired control statements are included in local TDMF jobs.
• SESSION
For information about the SESSION statement, refer to “Ordering input of control statements” on page36.
• MIGRATE or REPLICATE
For information about the MIGRATE control statement, refer to “MIGRATE control statement” on page48. For information about the REPLICATE control statement, refer to “REPLICATE control statement”on page 49.
Note: In a remote TDMF session, only the SESSION statement can be used.
Optional control statementsOptional control statements provide information for entering TDMF commands and syntax parameters.
Use the following links for more information on optional control statements.
• REMOTE
For information about the REMOTE control statement, see “REMOTE control statement” on page 44.• GROUP
For information about the GROUP control statement, see “GROUP control statement” on page 46.
© Copyright IBM Corp. 2010, 2019 35
Ordering input of control statementsYou must specify TDMF control statements in the following required order.
Procedure
1. The SESSION control statement must appear first. This statement can include a session name and theSMF identifiers of the Master and Agent systems. It can also include the Communications data setname, options applicable to the session itself, and default settings for volume migration. If thecommunications data set name is omitted, the name is extracted from the session’s JCL
2. In a TCP/IP migration session, where the target devices are controlled by another TDMF Mastersystem at a remote location, the REMOTE control statement is required. This statement mustimmediately follow the SESSION control statement. The REMOTE control statement defines themigration session IP Port number and the remote Master system IP host name. For more information,see “REMOTE control statement” on page 44.
3. A GROUP control statement can precede a number of MIGRATE or REPLICATE statements. Thevolumes in these subsequent statements are synchronized together. Options that are specified on thiscontrol statement apply to the volumes in the group, and are discarded when a new GROUP controlstatement is encountered. Do not specify the grouped volumes with a mixture of MIGRATE andREPLICATE control statements. A maximum of 32 GROUP statements can be supplied in a session.
4. A MIGRATE control statement specifies a volume swap migration. If it is to be treated as a single(ungrouped) volume, the control statement must precede any GROUP statements in the input stream.The statement provides the source and target volume for the migration, and can supply a new volumeserial number to be written to the source volume after the swap. The MIGRATE statement alsoprovides options that apply to this volume migration alone. No MIGRATE statements can be specifiedin a session that contains a REMOTE statement.
5. A REPLICATE control statement specifies a volume Point-in-Time migration. If it is to be treated as asingle (ungrouped) volume, the REPLICATE control statement must precede any GROUP statements inthe input stream. If a remote system is provided on the control statement, the second or only targetvolume is assumed to be at the remote system. If two target volumes are supplied, but no remotesystem exists, a duplex local volume copy is made. Options that are specified on this controlstatement apply to the single volume migration only. Any options that are specified on the precedingGROUP and SESSION statements and the installation defaults are overwritten.
Important: For volumes that are not part of a group, the MIGRATE and REPLICATE control statementsmust precede the GROUP control statement.
36 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
SESSION control statementThe SESSION control statement must be the first specification for the session as the following syntaxdiagram indicates.
About this task
SESSIONname
Master(sysid)
Agents(
,
sysid
)
NOPSWD
NOPASSword
PASSword (remote_password) SYSCOM (dataset_name)
NETView (netview_name)
CONGroup (name)
CONSOLE(CONSTDMF)
CONSole (console_name)
OPTions (|Common Options| |Session Options|)
SESSION parametersYou must use the following rules when you specify SESSION Control Record parameters.
Use the following rules to specify SESSION Control Record parameters.
• NAME. Specify an optional session name. If supplied, the session name must not exceed 8 characters inlength.
• Master (sysid). The Master system identifier. The Master system ID must be supplied, and mustmatch the SMF ID of the system on which the Master job is running.
• Agents (sysid). The agent's identifiers. Up to 63 Agent systems can be specified.• NOPASSword/PASSword. The password to be used to connect to the remote TDMF Master session. The
TCP/IP connection password can be specified only in a remote Master session. The password can be 1 -8 alpha-numeric characters.
• SYSCOM (data set name). Specify an optional Communications data set name. If theCommunications data set name is provided, it must match the data set specified on the SYSCOM DDstatement. However, if the Communications data set is a member of a generation data group, you mustprovide the root name of the group, rather than the actual generation name.
• NetView (name). This parameter is required for GDPS/PPRC HyperSwap managed volumes. It is thename of the NetView started task on the Master system. This name is required for HyperSwap volumeswap migrations and is used in the z/OS Modify commands that are generated and run by TDMF.
• CONSole (console_name). The name of an authorized NetView console that is reserved for use withTDMF. The TDMF extended MCS console is activated on the Master system during initialization of asession migrating GDPS/PPRC HyperSwap managed volumes or EMC volumes that are a part of an EMCConsistency Group and have the EMC AutoSwap function active. If omitted, the default console name isCONSTDMF. The console must have authority to run NetView and GDPS commands or the appropriate
Chapter 3. Using control statements 37
EMC commands to control EMC AutoSwap. Since only one TDMF extended MCS console can be active ina sysplex at any time, it is possible to specify the same console name for every LPAR.
• CONGroup (name). The name of the Consistency Group started task on the Master system. This nameis required for EMC AutoSwap volume swap migrations and is used in the z/OS Modify commands thatare generated and run by TDMF.
SESSION optionsSession options can be used to override the installation defaults. What follows is a list of the options thatcan be specified on the SESSION control statement.
Session options can be used to override the installation defaults (implemented with the GTDSOPTN job.See “Migration sessions with a mixture of maintenance levels” on page 21. They can also be used toprovide volume migration defaults with session-wide applicability.
Note: Most of the SESSION options defaults can be modified in the GTDSOPTN job.
The following syntax diagram displays all options that can be specified on the SESSION controlstatement.
38 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
NOCONFirm
CONFirm
NOPRompt
PRompt
NOCOMPARE
COMPARE
NOPAV
PAV
NOAUTOOPerations
NOAUTOOPS
AUTOOPS
AUTOOPerations
NOTERMGRoup
TERMGRoup
NOFASTcopy
FASTcopy
NOPUrge
PUrge
NOCOMPRESSion
COMPRESSion
NOOVA
OVA
NOPPIT
NOPERPetual
PERPetual
PPIT
NOALLOWmirrorchange
ALLOWmirrorchange
ACKnowledge
NOACKnowledge
RESETCHangedflag
NOEXTVtoc
NOEXVToc
EXTVtoc nnNOFORce
FORce
CONDitional
UNCONDitional
XVToc
SYNCgoal 005 - 999
OFFLINETarget
ONLINETarget
OFFLINESource
ONLINESource
FENcesource
NOFENcesource
UNFENcetarget
NOUNFENcetarget
IOBUFFERPOOL THRESHOLD(512)
THRESHOLD(nnnnn)
NOIOBUFFERPOOL
UNIDENTifiedsystems
• Description: Unidentified systems.• Function: Sets a severity for detection of systems that have logical paths to a subsystem or volume but
that have not been defined as Agents.• Options:
Chapter 3. Using control statements 39
Warning
Error
Ignore
Terminate• Default: Terminate
ALLOWINValidcounts/NOALLOWINValidcounts
• Description: Allow invalid count records• Function: Allows a volume copy to continue after an input record is encountered that has a non-
standard count field.• Default: A count field with a CCHH that does not match its physical disk location will cause volume
termination.
ICKdsf/NOICKdsf
• Description: Setting for whether or not to automatically invoke ICKdsf for a volume table of contents(VTOC) and index correction.
• Function: Causes ICKdsf to be loaded and invoked after a successful swap migration if the sourcevolume has a different number of cylinders to the target device. The ALTTRKS option also causesICKdsf to be invoked if the number of primary cylinders is the same, but there is a difference in thenumber of alternate tracks. ICKdsf cannot be specified for APPROVAL mode migrations.
• Options: ALTTRKS• Default: NOICKdsf (no automatic invocation of ICKdsf will occur and a warning message will be issued
when a VTOC/device mismatch exists).
NOPAcing/PAcing
• Description: Volume Pacing• Function: Sets normal or reverse volume pacing to limit the frequency of source device I/O operations
and real storage usage, in accordance with other system activity. In normal pacing, a single readoperation starts at 15 tracks. If I/O contention is detected, a delay is inserted between I/Os to reduceTDMF device utilization.
If a real storage shortage is detected, the size of the real storage I/O buffer is reduced, which alsoeffectively reduces the number of tracks that are written per I/O.
With reverse pacing, a single read is initially limited to one track. The real storage I/O buffer size isincreased if there is no I/O contention or real storage shortage.
• Default: No volume pacing.
RELABel(xx)
• Description: Relabel source volume after swap• Function: Provides a two alphanumeric character prefix for each new volume serial number. The new
volser is derived by appending the source device number.
Note: This function is not supported in an environment where JES3 manages the devices.• Default: The source volume will be relabeled with the target volume serial.
NOCHECKTarget, NOCHKTarget/CHECKTarget, CHKTarget
• Description: Check target volume is empty• Function: Causes initialization to fail if a target volume contains user data.• Default: No check is initiated.
40 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
TIME
• Description: Time display .• Function: Determines whether system TOD clock values will be converted to local time before being
displayed.• Options:
GMT
LOCAL• Default: No local time zone conversion.
NOCFW
• Description: No Cache Fast Write• Function: Prevents Cache Fast Write from being used for Communications Data Set I/O.• Default: Cache Fast Write will be utilized for VRNQ I/O if the session contains a large number of Agent
systems or volumes and the SYSCOM device supports it.
NOXCF
• Description: Prevents XCF communications from being utilized.• Function: Prevents the Master system from establishing an XCF group for communications with Agent
systems in a parallel sysplex environment.• Default: XCF will be used to speed communications during volume initialization and termination.
SINGLEgroup
• Description: Single group• Function: Treats all volumes in the session as a single group, with a group id of SINGLE.• Default: Volumes are processed individually, or grouped using the GROUP statement.
NOWARNING
• Description. Suppress warning messages.• Function. Enables the suppression of certain TDMF warning messages. Using this option keyword
causes the system to disregard the warning message about the migration, and resets the completioncode to zero.
Note: Before deciding to degrade a warning condition, refer to the message description or the TSOMonitor message help. The message must specifically state that it might be neutralized by NOWARN orthe session fails.
• Options. These messages can be suppressed by specifying the message numeric characters followingthe NOWARNing (nnnn):
– GTD1185W– GTD1293W– GTD1364W– GTD1370W– GTD1372W– GTD1384W– GTD1385W– GTD1387W– GTD1444W– GTD1445W– GTD1446W
Chapter 3. Using control statements 41
– GTD1485W– GTD1574W– GTD1657W– GTD1670W– GTD1692W– GTD1733W– GTD1824W– GTD1955W– GTD1997W– GTD2246W– GTD2455W– GTD2464W– GTD2465W– GTD2667W– GTD2668W– GTD2876W– GTD3690W– GTD3769W– GTD3771W
• Default. Any warning messages result in a condition code of at least 4.
CONCurrent, CNCRNT
• Description: Concurrent volume limit.• Function: Provides a maximum number of concurrently migrating volumesand, optionally, an indication
that this limit should only apply to "active" volumes.
The ACTIVE option refers to volume in their copy or first refresh phases, after the first refresh phase thelimit on the number of volumes concurrently active in refresh is removed. Therefore, many volumes inthe session could be in their second or higher refresh phase simultaneously. This could potentially leadto high I/O activity.
• Options:
(nn)
ACTIVE• Default: All volumes can be initialized and be migrated concurrently. If a number is provided without theACTIVE option, an initialized volume will be counted as part of the concurrent limit until its migrationterminates.
STORCLas
• Description: DF/SMS storage class.• Function: Provides a storage class for allocation of the temporary data sets required for ICKDSF
processing.• Options: A valid DF/SMS storage class must be provided if the option is specified.• Default: Temporary data sets are allocated using an esoteric unit type that has VIO capability.
IGNORESYStems
• Description: Ignore systems' path data.
42 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
• Function: Allows specification of UNIDENT(TERM) even when a storage controller is returning invalidlogical path data.
• Options: Maximum of 48 ten-digit system host identifiers - the fifth to 14th characters of the CPU ID asdisplayed in the GTD2382I message.
• Default: None
HSWAPDisable
• Description: HyperSwap Disable command.• Function: Causes the HYPERSW DISABLE command to be issued before HYPERSW OFF• Default: HYPERSW DISABLE not issued.
IGNOREGdps
• Description: Ignore inactive state of some volumes in a GDPS/PPRC system• Function: Allows GDPS/PPRC HyperSwap managed volumes to be migrated without the need for a
TDMF Agent on every active GDPS/PPRC system.• Options: Maximum of 32 eight-character system names.• Default: TDMF must be started on every active GDPS/PPRC system.
PORT(nnnnn)
• Description: Improves responsiveness where XCF is not applicable.• Function: Allows the Master system to break into the Agent systems' timer interval waits to request a
function.• Options: Port number to be used by the Agents for the TCP/IP connection.• Default: XCF is used within a sysplex to improve responsiveness; systems outside the sysplex wait for
the entire timer interval.
IGNORESCUs
• Description: Ignores unidentified system of a migrating volume.• Function: Allows the specification of UNIDENT(TERM) even when a storage controller is returning
invalid logical path data.• Options: Maximum of 8 ten-character SCU-SERIAL numbers, as they appear in messages GTD2244E,
GTD2245E, GTD2247E and GTD2249E, or in response to the "devserv qd" operator command.• Default: None
SPIDFENCE/NOSPIDFENCE
• Description: SPID fencing action for source and target volumes.• Function: Specifies the SPID fencing action to be taken on source and target volumes during the
migration.• Default: Source and target volumes will be put into the SPID fence state (if supported by operating
system and storage controller).
APPROVAL/NOAPPROVAL
• Description: Option to switch back to the original configuration, or terminate the migration with the newconfiguration.
• Function: Specifying APPROVAL changes the operation of TDMF migrations to enter the Approval Modeafter the swap and resume of I/O operations to the target volume. The session remains active andmonitors and tracks all changes to the target volume. You have the option of approving the migration orto switch back to the original configuration. If you approve the migration, the session will go throughnormal termination. If instead you decide to switch back, TDMF will bring the original source volume tosynchronization with the target and swap the volumes back to the original configuration.
Chapter 3. Using control statements 43
The use of the APPROVAL option requires a prompt to approve or disapprove the new configuration, theresponse will be accepted from the interface dictated by the AUTOOPS option.
The NOFENCETARGET or FENCETARGET option can be used to specify the Soft Fence action to beperformed when a migration is disapproved (switched back).
Note: The APPROVAL option is valid for migrations of MIGRATE type only. The VTOC of the Targetvolume must not be refreshed via ICKDSF before the new configuration is approved. The ICKDSFsession option is not allowed for migrations using the Switch Back feature.
• Options: FENCETARGET and NOFENCETARGET• Default: Target volumes will be put back into the Soft Fence state (if supported by operating system and
storage controller) if it is decided to switch back.
IOBUFFERPOOL/NOIOBUFFERPOOL
• Description: Enable or disable dynamic cylinder and track I/O buffer pool management.• Function: IOBUFFERPOOL enables buffer pool management to dynamically assign cylinder and track
I/O buffers only when needed rather than statically allocating cylinder and track I/O buffers for eachvolume pair defined in the session.
NOIOBUFFERPOOL specifies that cylinder and track I/O buffers be statically allocated to each volumepair defined in the session as in older versions of TDMF.
• Options for IOBUFFERPOOL:
– THRESHOLD(1-65535)
The THRESHOLD value defines the minimum number of volume pairs in a session needed to triggerthe use of shared I/O buffer pool management. The default value is 512.
For example, with the value 512, I/O buffer pool management takes affect with 512 or more volumepairs defined to the session; with anything less than 512 volume pairs, the session uses static I/Obuffer pool allocation.
The valid range for THRESHOLD is 1-65535.• Default: IOBUFFERPOOL
REMOTE control statementThe REMOTE control statement must immediately follow the SESSION control statement. It provides aremote system name for subsequent GROUP and REPLICATE control statements. It also provides aTCP/IP network address, port information for connection to the remote system, and an optionalconnection password.
44 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
REMOTE formatThe format of the REMOTE control statement is as follows.
REMOTE remote_name
HOSTname ip_host_name
ADDRess ip_address
PORT port_number
NOPASSword
PSWD
PASSword remote_password
PSWD
OPTions options
REMOTE parametersUse the following definitions when you specify parameters for the REMOTE control record.
• remote_name. The name is required and must not exceed 8 characters in length. It does not need tocorrespond to any characteristic of the remote system. It is merely used as a label to connectsubsequent GROUP or REPLICATE control statements, or both, to the TCP/IP information on thisstatement.
• ip_host_name. The host name of the z/OS system where the remote MASTER system is running. Thehost name is not case-sensitive when resolved by DNS.
• IP address. The address is an IPv6 or IPv4 format address. IPv6 addresses are specified in ClasslessInter-Domain Routing (CIDR) Notation, which is eight, one to four-digit hexadecimal numbers (1 toFFFF), separated by colons. TDMF accepts any standard IPv6 address shorthand, such as double-colon.IPv4 addresses are specified in dotted decimal notation, which is four, one to three-digit numbers (0 -255), separated by periods.
Note: Use the host name parameter instead of the ADDRess parameter.• port_number. The port number is a decimal number not greater than 65534 and must correspond to
the port number that is specified in the remote Master system’s run parameter.• remote_password: The remote password is a one- to eight-character alphanumeric string that is
compared with the provided password on the remote system’s SESSION control statement.• options. See “REMOTE options” on page 45.
REMOTE optionsUse REMOTE options to override a limited number of TDMF defaults. You can specify only these optionson the REMOTE control statement, and are as follows.
SNDbuf( buffer_size_KB ) RCVbuf( buffer_size_KB )
The purpose of the SNDBUF and RCVBUF options is to provide flexibility in TCP/IP buffer storageallocation. For example, in a storage constrained system, you can specify as little as one track's worth ofdata - OPT(SNDB(56)). And, in a system with plenty of available storage and at the end of a link with ahigh latency time, it might be beneficial to request a much larger receive buffer - OPT(RCVB(512)).
The SNDbuf option is only used by TDMF to increase the buffer size from the installation default. If thisdefault is the usual 65,535 bytes, a specification of anything less than 64 KB does not cause TDMF tochange the buffer size.
Chapter 3. Using control statements 45
Table 12. Remote options
Parameter Description Function Options Default
SNDBuf Minimum size of thedata portion of theTCP/IP sends buffer(in KB).
Speed up the datatransfer.
Numeric valuefrom 1 -2048.
112 (approximatecapacity of two 3390tracks).
RCVBuf Minimum size of thedata portion of theTCP/IP receivesbuffer (in KB). Thisoption takes effectat the Remote TDMFMaster system.
Speed up the datatransfer.
Numeric valuefrom 1 -2048.
112 (approximatecapacity of two 3390tracks)
GROUP control statementA GROUP control statement can precede a number of MIGRATE or REPLICATE statements. The GROUPstatement enables the volumes in subsequent statements to be synchronized together.
Use the following guidelines when you specify a GROUP control statement:
• The options that you specify on this control statement apply to the volumes in the group, and arediscarded when a new GROUP control statement is encountered.
• Grouped volumes must not be specified with a mixture of MIGRATE and REPLICATE control statements.• A maximum of 32 GROUP statements might be supplied in a session.
Important: For volumes that are not part of a group, the MIGRATE and REPLICATE control statementsmust precede the GROUP control statement:
GROUP formatThe following is an example of the format of the GROUP control statement.
GROUP group_name
REMote (remote_name)
OPTions (|Common Options| |Volume Options|)
GROUP parametersGuidelines for supplying migration options for MIGRATE or REPLICATE requests that follow a GROUPstatement.
The GROUP statement provides a group name, and can also supply migration options for MIGRATE orREPLICATE requests that follow a GROUP statement. Follow these guidelines.
• The group name (group_name) and statement options remain in force until a new GROUP statement isencountered. Options can be overridden by the MIGRATE or REPLICATE statements for volumes in thegroup.
• A GROUP statement cannot be used when the SINGLEGROUP option is specified in the SESSION controlstatement.
• If a remote_name is used, it must match the remote_name used in a previous REMOTE statement.Only one REMOTE control statement can be supplied for a session. Every volume must refer to theremote system name, either directly in the REPLICATE control statement or in the preceding GROUPstatement.
46 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Volume optionsVolume options can be specified on the GROUP, REPLICATE or MIGRATE statements.
The options available on the GROUP statement, in addition to those options that can be specified on theSESSION control statement as described in “SESSION options” on page 38, are as follows:
NOPAcing
MAXTRacks (15)
5
3
1
PACing
(Fullspeed)
ABANDONPprc
ABANDONSrdf
ABANDONTruecopy
ABANDONXrc
ABANDONMirror
Note: For convenience, any option that is valid for the individual MIGRATE or REPLICATE controlstatements for grouped volumes can be specified instead on the associated GROUP statement.
The following table provides parameter details for the volume options.
Table 13. Volume options
Parameter Function Options Default
MAXTRacks Maximum tracks per write: Restrictthe maximum duration of each I/Orequest independently of the sourcevolume pacing mechanism.
1, 3, 5 or 15 A read or write of up to 15tracks is allowed.
PACing Sets volume pacing, either normal orreverse as described in the SESSIONstatement pacing option. If theFullspeed copy option isrequested, either the source or targetdevice(s) will be driven at 100% of itscapacity. Without this option, thevolume reads and writes will not beperformed concurrently, halving thereal storage requirement andreducing the impact of the migrationon the volumes and associatedsubsystems.
NormalFullspeed
No volume pacing
Chapter 3. Using control statements 47
Table 13. Volume options (continued)
Parameter Function Options Default
ABANDONPprc
ABANDONSrdf
ABANDONTruecopy
ABANDONXrc
ABANDONMirror
Allows for the migration from devicesthat are mirroring to non-mirroringdevices.
None Do not allow migration ofmirroring devices to non-mirroring devices.
Note: These options are also valid on the individual MIGRATE and REPLICATE control statements.
Full speed copy impactIf the Full Speed Copy option is requested, either the source or target device will be driven at 100% of itscapacity. Without this option, the volume reads and writes will be performed alternately, reducing to halfthe real storage requirement, and reducing the impact of the migration on the users of the source volumeand the source and target DASD subsystems.
The Full Speed Copy feature can be used to shorten the duration of a volume migration. You request theFull Speed Copy by using the fullspeed subparameter of the volume pacing option.
Using Full Speed Copy causes the volume cylinder read and write requests to be overlapped, therebyreducing the time taken by the volume's copy phase, and reducing the number of updated tracks thatmust be refreshed. This process significantly decreases the refresh and synchronization times requiredfor the volume. In practice, however, the additional load on the source and target subsystems willadversely affect the I/O response times for both the migration itself and all other users of the sourcevolume, as well as users of other volumes in the source and target subsystems.
Without considering this probable performance degradation, the duration of the volume copy phase couldbe reduced by up to 50%. The actual rate of copy speed increase will be less; for example, if 1,000cylinders are being copied and the average source and target volume response times are 120 msecs and180 msecs respectively, the copy phase will take five minutes, with no overlap (1000 x (0.120+0.180)seconds).
Running the same migration with Full Speed Copy might result in average source and target responsetimes of 150 msecs and 240 msecs; this would reduce the copy phase duration to about four minutes(0.120 + 1000 x 0.240 seconds). In this example, the I/O of the source volume is faster than the I/O ofthe target, so the TDMF does not need to wait for any read operation to complete, except the first readoperation.
MIGRATE control statementA MIGRATE control statement specifies a volume swap migration. Use these guidelines when you specifya MIGRATE control statement.
• If it is to be treated as a single (ungrouped) volume, the control statement must precede any GROUPstatements in the input stream.
• The MIGRATE statement provides the source and target volume for the migration. Also, it provides anew volume serial number that is written to the source volume after the swap, and options that applyonly to this volume migration.
• No MIGRATE statements can be specified in a session that contains a REMOTE statement.
48 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
MIGRATE formatThe following is a syntax diagram of the format of the MIGRATE control statement.
MIGRATE source_volume target_volume
TARget (target_volume)
TGT (target_volume)
new_volser
RELABel (new_volser)
NEWVSN (new_volser)
OPTions( Common Options Volume Options )
MIGRATE parametersUse these guidelines when you specify the MIGRATE control statement.
• Each MIGRATE statement specifies a swap migration volume pair, with an optional new volume serialnumber to which the source volume will be clipped after the swap.
• The source and target volume serial numbers must be supplied.• If the new volume serial number is not specified, one of the following occurs:
– A new volume serial number is generated that uses the RELABEL prefix that is supplied on theSESSION statement and the source device number.
– The source volume is relabeled with the original target volume serial number.• If a new volume serial number of fewer than 6 characters is required, it must be specified as a
parameter of either the NEWvsn(serial) or RELABel(serial) keywords.
REPLICATE control statementA REPLICATE control statement specifies a volume Point-in-Time migration.
If it is to be treated as a single (ungrouped) volume, the control statement must precede any GROUPstatements in the input stream. If a remote system is provided on the control statement, the second oronly target volume is assumed to be located at the remote system. If two target volumes are supplied,but there is no remote system, a duplex local volume copy will be made. Options specified on this controlstatement are applicable only to this single volume migration and override any options specified on thepreceding GROUP and SESSION statements as well as the installation default options.
Note: The APPROVAL option of the SESSION control statement cannot be used for migrations ofREPLICATE type.
Chapter 3. Using control statements 49
REPLICATE formatUse the following as an example of the format of the REPLICATE control statement.
REPLICATE source_volume local_target
remote_target
remote_target
remote_target
TARGET( local_target
duplex_target
remote_target
)
TGT( local_target
duplex_target
remote_target
)
REMote( remote_name )
OPTions( Common Options Volume Options )
REPLICATE parametersFollow these guidelines to specify the REPLICATE control statement.
• Each REPLICATE statement defines a Point-in-Time volume migration. The source volume and at leastone target volume must be supplied. If a REMOTE statement was supplied and the preceding GROUPstatement did not include the REMOTE parameter, then each REPLICATE statement must refer to theremote system name that uses the REMOTE parameter.
• If two target volumes are specified, the first is a LOCAL target volume and the second is either a volumeat a remote location ("REMOTE" specified for volume or group) or a duplex local target. If one targetvolume serial number is specified, the "REMOTE" specification is used to determine whether it is a localor remote target volume.
• Options that might be supplied on the REPLICATE statement.
– “Common options for the SESSION, GROUP, MIGRATE, and REPLICATE statements” on page 50– “Volume options” on page 47
• Any options that are specified in a control statement, which do not apply to the function, are ignored.
Common options for the SESSION, GROUP, MIGRATE, and REPLICATEstatements
Use the following list to guide you when you specify common options on the SESSION, GROUP, MIGRATE,and REPLICATE control statements.
Note: Some of the options that are listed might not be applicable to all control statements. For example,the NOOVA, OVA, NOPERPetual, PERPetual, ONLINETarget, and OFFLINETarget options are onlyapplicable to REPLICATE control statements and functions. For an accurate listing of options that mightbe used for each individual control statement (SESSION, GROUP, MIGRATE, and REPLICATE), refer to theMASTER member, provided in JCL format in SGTDSAMP. Specify any option that applies to all the volumesin the session and that is valid for the SESSION control statement, on the associated SESSION statement.
The following syntax diagram displays the common options for the SESSION, GROUP, MIGRATE, andREPLICATE control statements.
50 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
NOCONFirm
CONFirm
NOPRompt
PRompt
NOCOMPARE
COMPARE
NOPAV
PAV
NOAUTOOPerations
NOAUTOOPS
AUTOOPS
AUTOOPerations
NOTERMGRoup
TERMGRoup
NOFASTcopy
FASTcopy
NOPUrge
PUrge
NOCOMPRESSion
COMPRESSion
NOOVA
OVA
NOPPIT
NOPERPetual
PERPetual
PPIT
NOALLOWmirrorchange
ALLOWmirrorchange
ACKnowledge
NOACKnowledge
RESETCHangedflag
NOEXTVtoc
NOEXVToc
EXTVtoc nnNOFORce
FORce
CONDitional
UNCONDitional
XVToc
SYNCgoal 005 - 999
OFFLINETarget
ONLINETarget
OFFLINESource
ONLINESource
FENcesource
NOFENcesource
UNFENcetarget
NOUNFENcetarget
FENcesource/NOFENcesource• Description: Soft fencing action for source volumes• Function: Specifies the soft fencing action to be taken on source volumes after a successful MIGRATE.• Default: Source volumes will be put into the Soft Fence state (if supported by operating system and
storage controller).
UNFENcetarget/NOUNFENcetarget• Description: Soft fencing action for target volumes
Chapter 3. Using control statements 51
• Function: Specifies the soft fencing action to be taken on target volumes prior to starting a migration.UNFENtarget removes the target volumes from the soft fence state prior to the beginning of a migration.NOUNFENcetarget does not remove the target volume from the soft fence state prior to the beginning ofa migration.
Note: If NOUNFENcetarget is specified and the device is soft fenced, the source volume’s migration isterminated by TDMF.
• Default: Target volumes are taken out of the Soft Fence state (UNFENcetarget).
NOCONFirm/CONFirm• Description: Confirmation at volume/group initialization• Function: Requests confirmation using the TSO monitor or system operator reply before initializing a
volume or group.• Default: No confirmation required.
NOPRompt/PRompt• Description: Synchronization prompt switch• Function: Requests confirmation using the TSO monitor or system operator reply before synchronizing a
volume or group.• Default: No prompt is issued (NOPRompt).
NOCOMPARE/COMPARE• Description: Volume comparison• Function: Compares the source and target volumes after synchronization. This is not intended for
production use.• Default: Volumes are not compared (NOCOMPARE).
NOAUTOOPerations/AUTOOPerations, NOAUTOOPS/AUTOOPS• Description: Automated operations switch• Function: Issues console messages and accept responses to prompts from the operator. Provide non-
displayed WTO messages for automated operations scheduling.• Default: No operator messages issued.
NOTERMGRoup/TERMGRoup• Description: Terminate a volume group on error• Function: Terminates an entire group if the migration fails for one of its volumes.• Default: The group continues to completion after termination of individual group volumes
(NOTERMGRoup).
NOFASTcopy/FASTCOPY• Description: Enable/disable FASTCOPY• Function: Restricts the cylinders being copied from the source volume to those allocated for data sets.
Notes:
– For z/OS volumes, the NOFASTCOPY statement is not available when an Extent-Efficient or Space-Efficient volume is a migration target volume.
– For z/VM volumes, the FASTCOPY option is not available when migrating a VM formatted volume. VMvolumes are not required to have a VTOC, so the information required for the FASTCOPY option isunavailable. If a VM target volume is Extent-Efficient or Space-Efficient, the volume keeps the Extent-
52 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Efficient or Space-Efficient status, but its free space is removed from Controller management. Thefree space is available only to that VM target volume.
– The two restrictions above are in effect at the volume pair level, so volumes within the scope ofrelated SESSION or GROUP control statements that do not require these restrictions are processednormally.
• Default: For non-Extent-Efficient and non-Space-Efficient volumes, the entire volume, includingunallocated cylinders, is copied (NOFASTCOPY) by default. Otherwise, the defaults are FASTCOPY forz/OS and NOFASTCOPY for z/VM.
NOPUrge/PUrge• Description: Setting for whether to allow purging of the target volume• Function: Erases data on one or more target volumes from those cylinders into which source volume
data is not copied. This option is relevant when a "small to large" migration or when FASTCOPY isspecified.
• Default: NOPUrge (the target volume is erased unless it is an RVA/SVA type)
NOALLOWmirrorchange/ALLOWmirrorchange• Description: Setting for whether or not to allow mirroring change• Function: Allows or disallows migrations between devices that employ different mirroring techniques
(PPRC to SRDF, SRDF to PPRC)• Options:
NOACKnowledge
ACKnowledge• Default: NOALLOWmirrorchange (initialization will fail for a PPRC primary device mirrored by an SRDF
device)
NOEXTVtoc/EXTVtoc NOXVToc/XVToc• Description: Resizing the VTOC on the target device.• Function: Provides required VTOC size for a swap migration to a larger device when the ICKDSF option
will be performed.
The FORCE parameter requires TDMF to attempt to move extents for data sets that are not currently inuse if these are blocking the specified VTOC expansion.
The UNCONDITION parameter causes TDMF to terminate the volume migration if the requested VTOCexpansion is not possible because of the presence of data set extents.
• Options: (nn) is the number of tracks the VTOC will be sized to. If (0) specified, only REFVTOC will bedone. See “ICKDSF” on page 80 for more details.
• Default: The VTOC will be resized based on volume utilization and average data set size (TDMFalgorithm). If the requested VTOC expansion cannot be achieved, TDMF will invoke ICKDSF to extendthe VTOC up to the start of the first data set extent following the VTOC on the volume, or will merelyperform a REFVTOC if there is an allocated dataset adjacent to the VTOC.
SYNCgoal• Description: Parameter for modifying the source device quiesce period• Function: Sets the maximum number of seconds that the source device is quiesced before volume
synchronization occurs.• Options: 005-999• Default: 005 (synchronization occurs during a five second quiesce period)
Chapter 3. Using control statements 53
NOCOMPRESSion/COMPRESSion• Description: Setting to enable or disable TCP/IP data compression• Function: Utilizes software data compression to reduce the amount of physical data transmitted in a
TCP/IP migration.• Default: NOCOMPRESSion (data compression is not enabled)
NOOVA/OVA• Description: Enables or disables Offline Volume Access• Function: Allows OVA registration on one or more point-in-time target volumes in the session.
Note: These options are only applicable to REPLICATE control statements and functions.• Default: NOOVA (Offline Volume Access enabled)
NOPPIT/PPIT or NOPERPetual/PERPetual• Description: Enables/disables Perpetual Point-in-Time copy• Function: Continuously performs new refresh and synchronization cycles on PIT volume(s) in the
session.
Note: These options are only applicable to REPLICATE control statements and functions.• Default: NOPPIT (Perpetual Point-in-Time copy disabled)
RESETCHAngedflag• Description: Resets DS1DSCHA at PIT synchronization.• Function: Causes all the data set changed flags to be reset on the source volume, assuming they will be
reset on the target by an OVA backup job.• Default: Not applicable (the source VTOC is not reset)
NOPAV/PAV• Description: Allow PAV and HyperPAV activity on source volumes.• Function: Controls whether PAV and/or HyperPAV alias devices are allowed to be bound to the source
volume during copy and refresh phases.• Default: NOPAV, which disables PAV and HyperPAV activity during a volume migration.
OFFLINETarget/ONLINETarget• Description: PIT target volume status at completion.• Function: ONLINETarget causes the PIT target volume to remain online at termination.
Note: Both ONLINETarget and OFFLINETarget are only applicable to REPLICATE control statementsand functions.
• Default: PIT target volumes are taken offline after a successful migration.
OFFLINESource/ONLINESource• Description: Setting for whether source devices will be offline or online after swap migrations.• Function: ONLINESource causes the original source device to remain online at completion.• Default: OFFLINESource (source devices are taken offline after a successful swap migration)
54 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Chapter 4. Improving performanceTDMF provides several different ways to measure performance. For example, the Auto-OperationsInterface helps you control TDMF sessions. Dynamic Pacing and I/O pacing can be used to detect adverseeffects after which Real-Storage pacing can be used to dynamically decrease the read/write of tracks in asingle I/O operation.
Auto-operations interfaceAn auto-operations interface is provided in TDMF. The interface is a user-selected option, either as thesystem default, as an option within a migration session, or on a volume level. The intent of this interface isfor the control of TDMF sessions. You can enable the option by using the automated operations package,as a user or operator.
If the auto-operations interface is selected, certain TDMF messages are routed to the MVS systemconsole and the TDMF TSO monitor. Action messages can be responded to from either function. TDMFnotes and displays where the response originated. All messages that can output to the MVS systemconsole for auto-operations are referenced in documentation. See Appendix D, “Messages for automatedoperations,” on page 209.
The messages that are output to the MVS system console include Action (GTDnnnnA), Informational(GTDnnnnI), and Error (GTDnnnE) types.
Action messages can be canceled by responding with the word CANCEL. A volume or group can stop aftera cancel request from the MVS system console or TDMF TSO monitor. They can also stop after a volumepairing failure. In these cases, a GTDnnnnE message displays along with a brief explanation.
Note: If the automated operations option is in effect, the message numbers that appear in the generatedWTO/WTORs have the prefix "GTD" if GTDMAIN is specified.
Dynamic volume pacingTDMF dynamically monitors adverse effects on real storage and source volume usage. When an effect isdetected, TDMF can decrease resource use until degradation is alleviated.
Use of dynamic volume pacing can cause a volume migration to run more slowly due to adjustmentsmade dynamically to real storage or TDMF internal I/O operations. If time is limited for a volumemigration, then do not use dynamic volume pacing for that volume, and do not select the Full Speed Copyoption.
The Full Speed Copy feature can be used to shorten the duration of a volume migration. You can choosethe Full Speed Copy feature through the Full Speed subparameter of the volume pacing option. Full SpeedCopy feature causes the volume cylinder read and write requests to be overlapped. The overlappingreduces the time for the copy phase of the volume. The feature also reduces the number of updatedtracks that must be refreshed.
Note: This process significantly decreases the refresh and synchronization times for the volume.However, the additional load on the source and target subsystems can adversely affect the I/O responsetimes. This effect can occur for both the migration, other users of the source volume, and users of othervolumes in the source and target subsystems.
Without considering this probable performance degradation, the duration of the volume copy phase canbe reduced by up to 50%. The actual rate of copy speed increase is sometimes less. For example, 1,000cylinders are being copied and the average source and target volume response times are 120 msecs and180 msecs. The duration of the copy phase is 5 minutes, with no overlap (1000 x (0.120+0.180)seconds). This migration, if run with Full Speed Copy, can result in average source and target responsetimes of 150 msecs and 240 msecs. This result reduces the copy phase duration to about 4 minutes
© Copyright IBM Corp. 2010, 2019 55
(0.120 + 1000 x 0.240 seconds). In this example, the I/O of the source volume is faster than the I/O ofthe target. So TDMF does not need to wait for any read operation to complete, except the first readoperation.
I/O pacingVolume pacing occurs for I/O contention between each customer application I/O operation to the sourcevolume on all systems. This process occurs for TDMF I/O operations to the source volume during the copyand refresh phases.
Device service time is a percentage of I/O elapsed time and a fraction of the TDMF measurement interval.TDMF assumes that it is adversely affecting application performance if the source volume device busyapproaches 95 percent. (If customer I/O is less than 5%, this assumption is not made.) As a result, TDMFinserts a delay that is equal to the average device service time between each of its I/O operations. If thesum of customer and TDMF I/O operations falls below 75%, TDMF removes the delay that is insertedbetween each of its I/O operations.
If an adverse real storage impact exists within the TDMF Master system, TDMF does not increase its I/Ooperations until the real storage shortage is alleviated.
Note: When Full Speed Copy is specified as a pacing option for a volume or group, TDMF can exceed thepacing threshold (94%) that is set for the source volume. This result occurs regardless of the level ofcustomer I/O activity on the volume. Note the following considerations for Full Speed Copy:
• Any pacing delay between TDMF I/O's is removed only if no customer activity occurred on the volume inthe preceding 30-second interval.
• A pacing delay is inserted between TDMF I/O's only if the customer I/O can generate a device busypercentage. In competition with the migration, the percentage to be generated is in excess of 5%.
Applications in which frequent, short duration I/O requests occur can be seriously delayed by a FullSpeed Copy migration. Therefore, be cautious with this option for the migration of database volumes. Toreduce the risk of performance problems with a Full Speed Copy migration, use one of the followingoptions:
• Reverse pacing• Set a maximum number of tracks per I/O request
For more information, see “Reverse pacing” on page 56 and “User-specified pacing values” on page57.
The impact of TDMF I/O operations is ignored on the target. If performance is an issue on the targetsubsystem, use the User Specified Pacing option.
Real-storage pacingAfter all I/O pacing decisions, pacing for real storage comes into effect. If real storage frames used by themaster system exceed 12.5%, TDMF dynamically decreases the read/write of tracks in a single I/Ooperation. If the used frames go below 6.25%, TDMF dynamically increases the read/write of tracks.
Any value between the limits of 75-95% of I/O operations and 6.25-12.5% of real storage frames do notincur a change in the pacing of TDMF. TDMF does not attempt to increase pacing (for RSM resources) to asource volume that is constrained by application I/O operations.
Note: The amount of storage that is fixed at each pacing increment is doubled with Full Speed Copy.
Reverse pacingFor volumes with heavy channel usage, the Reverse Pacing option can be specified. This option is validon the Session control statement. Optionally, it can be specified as GTDSOPTN for default. Reverse pacingmeans that TDMF starts the volume copy process at one track per I/O operation rather than 15 tracks perI/O.
Note: TDMF dynamically adjusts the number of tracks that are read in a single I/O depending upon usage.
56 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
When pacing is not in effectPacing is not active during the QUIESCE, SYNCHRONIZATON, and RESUME phases. However, normal realstorage and I/O operation requirements continue. Therefore, the phases are completed as quickly aspossible.
The number of real storage frames and tracks for read/write per I/O are shown in the following table.
Table 14. Real storage and I/O pacing operations
Tracks for read/write per I/O
Storage requirement 15 5 3 1
900 K (225 frames) X
300 K (75 frames) X
180 K (45 frames) X
60 K (15 frames) X
Information related to the number of pacing decisions made by TDMF (increase/ decrease I/O operationsand real storage frames) is documented in “Option 5 - Display performance data of current and pastsessions” on page 154.
User-specified pacing valuesUsers can specify to TDMF the maximum number of tracks per I/O operation. This number can bespecified as 1, 3, 5 or 15 in the MAXTRACKS option of the GROUP or MIGRATE control statement. Ifpacing or reverse pacing are specified, and MAXTRACKS specifies 5, I/O starts at 5 tracks and pacesdown. In the case of reverse pacing, I/O starts at 1 track and paces up. In either case, I/O operations donot exceed 5 tracks. If pacing is not specified for a volume or group and a MAXTRACKS value of 5 issupplied, the resulting I/O for the volume is 5 tracks per operation.
Pacing with channel extendersTDMF checks for the presence of channel extenders and the size of the buffers.
Before you begin
When TDMF detects a channel extender, the read/write tracks are dynamically adjusted in a single I/Ooperation based on the buffer size. The buffer size in the channel extender equipment prevents bufferoverflow. This action is forced pacing. TDMF does not increase the tracks that are read in a single I/Ooperation beyond the buffer size. With the pacing option, the tracks are scaled down based on updateactivity on the source volume.
In the following example, TDMF displays information about the source device, which is attached to achannel extender.
Note: In this case, the maximum buffer size is 320 KB paced to read 5 tracks per I/O operation.
Figure 11. Pacing with source attached to channel extender
Chapter 4. Improving performance 57
VOLUME INFORMATION---------------------------------------------------------------------------Source Target New ---- Current ---- Cylinders ---- Number ---- VSN VSN VSN Phase Wait Num Curr %%% Req Compl WaitVDMS9A VDMS9B Copy 0030 02226 00134 06 00000 00000 0000Source Control Unit = 3990-CC Device = 3390-06 Cylinders = 02226Target Control Unit = 3990-CC Device = 3390-06 Cylinders = 02226 SourceRemote Device Information Device Largest Buffer Allocation Unit 16 KB Smallest Buffer Allocation Unit 16 KB Smallest Maximum Buffer Size 320 KB Maximum track reads without I/O split 05Volume Pacing Active using 05 Tracks per Operation. --- T R A C K S --- 15 5 3 1Percent of Copy Phase I/O using: 0 100 0 0Percent of Refresh Phase I/O using: 0 100 0 0
In the next example, the target volume is attached to a channel extender. The maximum buffer size is256 KB, which forces pacing to operate at 3 tracks per I/O operation. Although the displayed informationindicates that 4 tracks can be written to the target without buffer overflow, TDMF uses the next lowestvalue of pacing available.
Figure 12. Pacing with target attached to channel extender
VOLUME INFORMATION--------------------------------------------------------------------------- Source Target New ---- Current ---- Cylinders ---- Number ---- VSN VSN VSN Phase Wait Num Curr %%% Req Compl Wait VDMS9A VDMS9B Copy 0030 02226 00023 02 00000 00000 0000 Source Control Unit = 3990-CC Device = 3390-06 Cylinders = 02226 Target Control Unit = 3990-CC Device = 3390-06 Cylinders = 02226 Target Remote Device Information Device Largest Buffer Allocation Unit 16 KB Smallest Buffer Allocation Unit 16 KB Smallest Maximum Buffer Size 256 KB Maximum track writes without I/O split 04 Volume Pacing Active using 03 Tracks per Operation. --- T R A C K S --- 15 5 3 1Percent of Copy Phase I/O using: 0 100 0 0Percent of Refresh Phase I/O using: 0 100 0 0
In this last example, both the source and target volumes are attached to channel extenders.
Note: The buffer size on the source device is 960 KB but the target device buffer size is 256 KB. TDMFalways selects the smallest buffer size to avoid a buffer overflow situation.
Figure 13. Pacing with source and target attached to channel extender
58 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
VOLUME INFORMATION--------------------------------------------------------------------------- Source Target New ---- Current ---- Cylinders ---- Number ---- VSN VSN VSN Phase Wait Num Curr %%% Req Compl Wait VDMS9A VDMS9B Copy 0030 02226 00245 11 00000 00000 0000 Source Control Unit = 3990-CC Device = 3390-06 Cylinders = 02226 Target Control Unit = 3990-CC Device = 3390-06 Cylinders = 02226 Source TargetRemote Device Information Device Device Largest Buffer Allocation Unit 16 KB 16 KB Smallest Buffer Allocation Unit 16 KB 16 KB Smallest Maximum Buffer Size 960 KB 256 KB Maximum track reads without I/O split 15 Maximum track writes without I/O split 04Volume Pacing Active using 03 Tracks per Operation. --- T R A C K S --- 15 5 3 1Percent of Copy Phase I/O using: 0 100 0 0Percent of Refresh Phase I/O using: 0 100 0 0
SMF recordingDuring a TDMF session, no SMF records are written unless so specified in the system options. If SMFrecording is chosen, a user specified SMF record number is generated for each volume migration. If SMFrecording is not specified, only basic EXCP counts information is generated (SMF record type 4, 14, 15,30, 32 40, and 64). Valid SMF record values are 128 - 255. The SMF record is not written until volumetermination and is written to the master system only.
For details about the generated SMF record, refer to the SGTDSAMP library member TDMFSMF(assembler) and GTDFSASD (SAS). SGTDSAMP member GTDSASE contains sample JCL.
Copy functionsThere are two types of volume copy functions, Swap migrations and Point-in-Time migrations. You controlhow a migration completes by using the MIGRATE control statement or the REPLICATE controlstatement.
By default, synchronization starts without user intervention as soon as the synchronization goal can bemet. To cause the volume or group to wait for a user response, the session, group, or volume PROMPToption must be specified. This option forces a prompt from the master system to the TDMF TSO monitor(or through the MVS system console, if the auto-operations option is selected). A user action is requiredthrough Option 2 of the TDMF for z/OS TSO monitor, from the MVS system console, or from theinstallation's automated operations package.
After successful completion of a swap migration, the source device is varied offline and all I/O for thesource volume is redirected to the target device. At the end of a PIT migration, no serial numbers aremodified. The target devices are varied offline and there is no I/O redirection.
Point-in-time migrationA point-in-time migration is considered a backup facility only. No swap function occurs between thesource and target volumes.
After point-in-time migration, the target devices are varied offline.
Important: Use point-in-time copies on a different operating system from where they were created. MVSdoes not allow duplicate volume serial numbers, and MVS does not allow duplicate VSAM files due tocatalog issues. Refer to Chapter 6, “Using advanced functions,” on page 101 for information about offlinevolume access (OVA) and how this function can be used to access the point-in-time copy.
Note: TDMF allows an agent system to have target volumes (UCBs) offline in a point-in-time session.
Chapter 4. Improving performance 59
Point-in-time migration and databasesWhen point-in-time is used with databases, ensure that all volumes are included in the same session, ifpossible. If a database exceeds the maximum number of volumes per session, a second session can beused to migrate the remaining database volumes. For operational reasons, it is advantageous to run themaster jobs for all the sessions on the same MVS system. All volumes must be grouped within a sessionthat has a different GROUP name. The prompt option necessary to ensure logical data integrity.
Before you respond to the prompt in a database point-in-time session, ensure that the application thatupdates the database logically completed all transactions. Also, ensure that subsequent transactions arequiesced (or queued) for the brief time that is required for the synchronization and termination phases.The result is complete data integrity correct logical transactions (for example, no broken chains).
Before a point-in-time migrated volume is used on another system, relabel the target volume if it containsVSAM or extended function data sets. Relabel the target volume with ICKDSF to the original source name.The data set name of the VSAM volume data set (VVDS), containing information about these data sets,has a fixed format that includes the volume serial number.
The following example shows how to relabel a volume with the original source volume serial number. Ifthe target volume is larger than the source volume, see “ICKDSF” on page 80.
This example is for a volume that is offline:
//STEP1 EXEC PGM=ICKDSF//SYSPRINT DD SYSOUT=*//SYSIN DD * REFORMAT UNIT(xxxx) VERIFY(tgtvsn) VOLID(srcvsn)/*
If the target volume serial number is to be used and the target volume is larger than the source volume,the following steps must be taken:
1. Vary the target volume online to the system2. Run an ICKDSF REFORMAT REFVTOC job
Complete the steps in order. For more information, see “ICKDSF” on page 80. If the source and targetdevice sizes are the same, step 2 is not necessary.
Application I/O performanceApplication I/O performance can be affected during the Copy and Refresh phases.
The performance effects of Copy and Refresh phases on application I/O operations is no greater thanduring a concurrent backup of the source volume. However, the effect can be lessened by selectingvolume pacing or reverse pacing.
Synchronization goalsThe synchronization goal parameter can be set for each volume pairing. The parameter sets the maximumtime (in seconds) that the synchronization phase is allowed to run.
A high I/O rate with many writes can prevent TDMF from meeting the synchronization goal for a volume.The result is multiple refresh phases during the migration. If this result is acceptable, then a change of thesynchronization goal parameter is not necessary. If the prompt is not responded to, TDMF refreshescontinuously.
Before the synchronization phase, a quiesce is sent to the source volume. For a multi-system session, themaster system sends a request to the agents to quiesce all I/O to the source volume (from the agentside). Currently, the last updates are collected and applied to the target volume (synchronization). At theend of synchronization, the master system starts the volume redirect (swap) phase.
60 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Note: During a point-in-time replication session, the swap phase is bypassed.
When all systems verify the redirect, the master system initiates the resume phase so that user I/O cancontinue to the new source volume. The elapsed time between the last quiesce phase and the resumephase is several seconds plus the actual synchronization time. The elapsed time is less than the specifiedsynchronization goal).
The synchronization goal default is 5 seconds. Synchronization does not occur unless the calculatedsynchronization time is less than the goal. If the synchronization goal is increased, then the queue orquiesce time of the customer I/O is greater. The value 999 means that synchronization is to occur as soonas possible without regard to time expenditure. The time that is expended can be a significant dependentupon the write activity of the source volume. Therefore, use the value discriminately. The value can bechanged dynamically for a volume or group that uses the TDMF TSO monitor (Z or ZG command). Formore information, see “Option 2- Current sessions: user interaction and status” on page 145.
Channel utilizationEven though TDMF can adversely affect normal customer I/O operations, TDMF completes I/O operationsfor one cylinder at a time. Therefore, you can plan for these types of sessions.
As described in “Application I/O performance” on page 60, TDMF can adversely affect normal customerI/O operations. However, TDMF completes I/O operations for one cylinder at a time. Thus an increase inchannel utilization can be seen. Therefore, plan for these types of sessions. Check the channel utilizationthrough a real-time monitor (such as RMF or Omegamon) before you start a session. If possible, completethese types of sessions during a low activity period.
If channel utilization is 30-35% busy before a migration, then TDMF drives the channel to over 80%utilization with little to no impact to the applications. However, if channel utilization is 85% busy before amigration, an application impact results. Generally, one volume per channel path migrates at a time. Theutilization of two volumes per channel path saturates the channel path to 100% busy.
Important: Specify the Full Speed Copy option for a volume migration cautiously. When the Full SpeedCopy option is selected, TDMF attempts to drive each source or target volume at 100% usage. Thisincreased activity causes heightened path usage for channels, switches, and the internal DASDsubsystem data paths. High device usage can degrade performance for other volumes in the source andtarget subsystems. TDMF volume pacing cannot detect or rectify this condition.
If caching is active on a source volume during a migration, normal device response time increases. TDMFreads the data directly from the volume and bypasses the loading of cache. TDMF reads a cylinder of dataper I/O request, unless limited by volume pacing. The application I/O then appears to be queued. Cachenormally contains customer data and TDMF resists interfering with customer application I/O responsetime by polluting the cache. If DASD Fast Write (DFW) is on for the target volume, the device responsetime decreases. TDMF writes a cylinder of data per I/O request. TDMF takes advantage of DFW, ifavailable.
If the pacing option is Full Speed Copy, the source volume DASD subsystem uses sequential accesscaching during the volume copy phase. In addition, if supported by the subsystem, cache track pre-staging completes during the refresh and synchronization phases.
Performance impact by phaseThe master and agent systems poll for processing at specific times. These times are determined by theactual phase that the volume pairing is in.
The phases of a volume migration and their associated time intervals are listed in the following table.
Chapter 4. Improving performance 61
Table 15. Performance impact by phase
Phase Time interval
Copy 30 secs
Refresh pass #1 30 secs
Refresh pass #2… n Interval is halved until the sync goal can be met
Group Option/Monitor Prompt 30 secs
Quiesce 1 sec
Synchronize 1 sec
Compare 1 sec
Swap/Point-in-Time 1 sec
Resume 1 sec
Termination 30 secs
Phases allow TDMF to be responsive with a minimum of CPU time expenditure.
The CPU time expenditure in running TDMF is less than 3% on average for the master system. Theaverage is dependent upon the number of volumes within a session and the write activity to the sourcevolumes. The CPU time expenditure of an agent system is almost immeasurable.
Take, for example, a master job that takes 44 min 22 secs to migrate 16 volumes. The TCB time is 63.5secs and the SRB time is 2.92 secs. The CPU time expenditure for the session is equal to 2.49% ((63.5 +2.92) / 2662).
When multiple volumes are running in a TDMF session, not all volumes are in the same phase at the sametime. Volumes can have different activity and capacities. The number of channels available to eachcontrol unit (CU) is also significant. Therefore, a four-volume session can run while volumes 1-3 are in thecopy, refresh, synchronization phases and volume 4 is completed.
Placement of the master systemThe master system must be placed on the system that has the most updates or on the system where theowning application is running.
If multiple TDMF master sessions are running on multiple operating systems, the MVS system must havea global facility like global resource serialization (GRS) or multi-image manager (MIM). This requirementprevents inadvertent reuse of volumes in a multi-session environment. If a GRS type facility is notavailable in the complex, then all master sessions must run on the same operating system.
Placement of the communications data setEach TDMF session must use a unique communication data set (COMMDS) for reasons of history logging,audit trails, diagnostics, and messages. The COMMDS must be placed on a volume with low activity andthis volume must not be defined to TDMF as a migration volume.
The COMMDS must not be placed on a volume that is subject to physical device reserves by othersoftware subsystems. For this reason, a volume that contains other control data sets (for instance, CA-MIM, StorageTek) is not suitable.
TDMF periodically initiates a reserve macro for the device where the COMMDS exists. Other applicationsthat use the device are affected if the reserve is not converted to a global ENQ (enqueue) by GRS or CA-MIM. The resource major name that is used by TDMF is TDMFRESV. The minor name is equal to the dataset name.
62 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
The appropriate adjustments to MIM or GRS ensure that the reserve macro results in a physical devicereserve or is converted to a global resource ENQ. See “CA-MIM resource sharing” on page 71 and“Global resource serialization” on page 79.
The COMMDS can be defined as a generation data group (GDG), which can ease the tracking of these datasets. In this case, each new COMMDS generation would be defined before the master and agent jobs weresubmitted. Thus, reference is allowed by relative generation number zero in the JCL.
Specify the "History" data set through the GTDSOPTN job when GDGs are used. Reviewing migrationstatus from the individual data sets can then begin. On the Past Sessions: Display Communication DataSet History window of the TDMF TSO Monitor. Individual generations can be selected for viewing.
Some benefits are obtained in a large migration session (many volumes and many agent systems) if theCOMMDS is allocated on a PAV capable volume. TDMF completes each interval read and write as twoconcurrent I/O requests. One I/O request uses Cache Fast Write.
Note: Ensure that every COMMDS file is retained or archived until a successful migration is confirmed.
Important: The reuse of a COMMDS causes the loss of previous session status and message information.In the event of an error, the COMMDS is the primary tool that is used for problem determination or systemrecovery.
RAID subsystems and rank contentionRAID subsystems consist of one or more physical volumes with a possible volume for parity. Thesubsystem is referred to as a rank or array.
Data can be striped across a physical rank to create multiple logical volumes or multiple logical devicesthat can use the same physical device. The number of internal device paths to each rank or array can vary.However, only one actuator exists per physical device. If more than one volume per rank or array is beingread or written to during a TDMF migration, the performance of that volume migration can be adverselyimpacted.
Ensure that a RAID subsystem read and write session involves only one volume per rank. Otherwise,severe performance problems can result.
Review the mapping of logical to physical volumes on a RAID subsystem to avoid overloading a physicalvolume within the array or rank.
Storage subsystem performanceMultiple operating systems can be attached to any storage subsystem and its volumes can be shared.Running a TDMF session with the volumes of one operating system does not adversely affect applicationson other operating systems that are attached to the subsystem.
TDMF bypasses the loading of cache so that customer application I/O is not affected during an activemigration session.
Not all storage subsystems accept cache commands for turning off cache or bypassing cache loading. Itdepends on the vendor architecture and implementation of a DASD storage subsystem. Cache is normallyshared within a DASD subsystem. As a pool, cache is accessed by all connected operating systems andcannot be fenced (partitioned) or allocated per volume. Therefore, a TDMF session on one operatingsystem can affect the application I/O on an operating system that is uninvolved with the session. Select apacing option or limit the active concurrent volumes in a session to minimize the effects.
Chapter 4. Improving performance 63
Suspend or continue a migrationIf a problem arises with a channel extender or real storage, you can dynamically suspend the volume orgroup migration until the network problem or storage issue is resolved.
Note: For real storage usage issues, consider dynamic volume pacing. For more information, see“Dynamic volume pacing” on page 55.
A suspend command that is submitted during the volume migration copy phase initiates before the nextread I/O operation. However, if the volume pairing is in the refresh phase, suspension begins on the nextread I/O operation while updates occur on the source volume. If no updates to the source volume exist,the command is ignored until the next write I/O operation to the target volume. When the suspensionbegins, the storage that is used for TDMF I/O operations is released.
TDMF continues to monitor the source volumes for updates until migrations can resume. When migrationsare resumed with the pacing option, the real storage requirements for TDMF I/O operations arereacquired at the previous TDMF level.
The TDMF hang monitorThe TDMF hang monitor detects a failure in TDMF processing, starts the appropriate error routine, andstops the migration.
A "hang" is detected if the master or agent system did not complete volume processing within a minute.Volume processing includes a timed wait of 1 - 30 seconds, a COMMDS read operation, processing ofvolume status changes, and messages and acknowledgements to the COMMDS.
A session stops when a "hang" is detected, while at least one volume in the session is quiesced by TDMF.
Note: For a quiesced volume such as a synchronized source and target, TDMF processing one volume persecond; the hang monitor interval remains at 1 minute.
A wait for completion of a volume or a COMMDS I/O operation is not considered a "hang". MessageGTD9472I, GTD9473I, or GTD9474I displays to inform the user in this regard. The "heartbeat" is updatedafter the processing of each volume in the session.
Dynamic suspension of a migrationTDMF sets a timeout limit on most volume I/O requests. When the MVS I/O timing facility detects that theoperation exceeded this limit, the operation is stopped and a failure code is displayed. TDMF includes adynamic suspension facility, which prevents further I/O volume pair operations, until the problem iscorrected and the user resumes volume copy.
The missing interrupt handler interval (MIHI) that is in effect for the specific device determines thetimeout value. The following timeout values can be determined for a device.
• MIHI less than 170 seconds, timeout is 1.5 x MIHI.• MIHI less than or equal to 4 minutes, timeout is 255 seconds.• MIHI greater than 4 minutes, I/O timing is suppressed.
The MVS operator command DIOS,MIH,DEV=/xxxx displays the MIHI of a device. The MVS operatorcommand SETIOS DEV=/xxxx,TIME=mm:ss alters the MIHI.
I/O timing is enabled for most COMMDS I/O operations with a timeout interval that is derived from theMIHI of the device. If a timeout occurs, processing continues as though no messages are pending. If atleast one device is quiesced or if the I/O operation is of a type that must complete, COMMDS I/O timing issuppressed.
For agent systems updates that are completed on each source volume (refresh notification), more controlblocks and logic exist to reduce the risk of migration failure.
64 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
If, during an I/O operation, TDMF detects an I/O timeout condition, TDMF dynamically places that volumemigration in a suspended state. Possible messages are GTD3555W, GTD3555W, GTD3556I, andGTD3557I. Messages GTD3562W and GTD3563W indicate when dynamic suspension is caused by thefailure of an I/O request due to an invalid device status.
You must submit the continue command from the TDMF TSO Monitor or through the batch utilityGTDBMON to restart the volume pairing.
When the continue command is submitted, TDMF resubmits the I/O operation that timed out. If atimeout condition occurs again, the dynamic suspension mechanism is reenabled. If a volume can neverbe successfully continued due to repetitive I/O timeouts, it can be stopped through the TDMF TSOMonitor or the batch monitor.
Chapter 4. Improving performance 65
66 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Chapter 5. PlanningSpecific program products or functions must be taken into consideration when you plan or run a TDMFsession. Subsequent user actions are also documented here.
Storage requirementsThis table lists storage requirements for the basic ECSA, CSA, and Extended Private for TDMF.
Table 16. TDMF storage requirements
Location Subpool
AboveorBelow16 MbLine Size
MasterAgentOr Both
Page FixedWhen
Page FreedWhen TDMF Usage
HVCOMMON6 Above 16K to 20K Both Systeminitialize
Systemterminate
Main control block
ECSA 241 Above 24K Both Systeminitialize
TSO Monitor
ECSA 228 Above 16K to 392K Both Systeminitialize
Systemterminate
Main control record
HVCOMMON6 Above 4K Both Systeminitialize
Volumeterminate
WORK control block- per volume pair
HVCOMMON6 Above 4K Both Systeminitialize
Volumeterminate
VMSG control block- per volume pair
HVCOMMON6 Above 1K to 12K Both Systeminitialize
Systemterminate
DDTV control block
CSA 228 Below 256 bytes Both Systeminitialize
Systemterminate
DDTR control block
HIGHVIRTUAL4 Above nK Master Systeminitialize
Systemterminate
MSVE record -nK1*# of Agents
HVCOMMON6 Above 24K- 28K Both Systeminitialize
Volumeterminate
VRBM control block- per volume pair
HVCOMMON6 Above 16K Both Volumeactivate
Volumeterminate
PAV2 alias DDTEblock - per aliasdevice
ECSA 228 Above 4K Both Systeminitialize
Systemterminate
Communicationswith EMC API
ESQA Above (#vols*524) +554
Both Volumeinitialize
Volumeterminate
SRB/IOSB storage
Private4 Above 4K(min)64K(max)
Master Systeminitialize
Systemterminate
CommunicationsNumber ofsystems / 2*4K(rounded up)
© Copyright IBM Corp. 2010, 2019 67
Table 16. TDMF storage requirements (continued)
Location Subpool
AboveorBelow16 MbLine Size
MasterAgentOr Both
Page FixedWhen
Page FreedWhen TDMF Usage
Private4 Above 4K to 20K Agent Systeminitialize
Systemterminate
CommunicationsCCW
Private4 Above 12K Both Systeminitialize
Systemterminate
Common I/O areas
HIGHVIRTUAL4 Above 4K(min)8MB(max)
Master Systeminitialize
Systemterminate
VRNQ I/O area -Number of systems* Number of volumepairs * 4K
HIGHVIRTUAL4 Above nK Master Volumeterminate
Volumeterminate
VRBM control block-per volume pair
HIGHVVIRTUAL5 Above 900K Master Volumeinitialize
Volumeterminate
Cylinder I/O - pervolume pair
HIGHVIRTUAL4 Above 60K Master Volumeinitialize
Volumeterminate
Track I/O area - pervolume pair
Private4 Above 32K Master Volumeinitialize
Volumeterminate
Copy task CCW area- per volume pair
Private4 Above 64K Both Systeminitialize
Systemterminate
Trace Table
• Row 3, TDMF main control record. 16 K-392 K, formula is 8+3*#vols/4 KB (rounded up to 4 K multiple).• 1 Row 8, TDMF MSVE record. nK, formula is 3*#vols/8 KB (rounded up to 4 K multiple).• 2 Row 9, TDMF VRBM control block, 24 K-148 K, formula is 20+K KB. K is 8 for 3390/3, 20 for 3390/9, 68 for
3390/27 and 128 for 3390/27• 3 Row 17.TDMF VRBM control block, nK, formula is 5*K KB. K is 8 for 3390/3, 20 for 3390/9, 68 for 3390/27
and 128 for 3390/27.• 4 Rows 8, 16-19, 21. When possible, TDMF obtains storage above the 2 GB bar (64-bit real) for these control
blocks.• 5 Row 18. If Full Speed Copy is selected, two 900 K Buffers are allocated.• 6 HVCOMMON refers 64-bit common storage (above the bar).
If the Compare option or Full Speed Copy is requested, an extra 900 K buffer for each volume’s migrationis allocated. Therefore, if 16 3390-3 volumes are being migrated with two systems that are involved, andthe Offline Volume Access option is not selected, the storage requirements for the Master and Agentsystem are listed below.
Master system
• Fixed common area storage - 612 K HVCOMMON, 9 K ESQA, 256-bytes CSA• Pageable common area storage - 28 K HVCOMMON• Fixed extended private area storage: 16,600 K
Agent system
• Fixed common area storage - 617 K HVCOMMON, 5 K ESQA, 256-bytes CSA• Pageable common area storage - 28 k HVCOMMON
68 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
• Fixed extended private area storage - 656 K
.
DASD space requirementsRefer to the following topic for all DASD space requirements.
See Appendix A, “DASD space requirements and serial numbers,” on page 199.
Activating new hardware I/O configurationsTDMF prevents operating system corruption of UCB LookUp Tables (ULUT) from swap migrations. Thiscapability is available when the MVS ACTIVATE command is submitted to dynamically change thehardware I/O of the LPAR.
About this taskTo prevent corruption of the internal IOS control blocks, TDMF submits the proper ENQ serialization.
Note: Dynamic I/O configuration activation must be delayed until all swap migrations complete.
CA productsUse the following for information on Computer Associates Products, which includes migrating volumes,event management and automation, job management, resource sharing and moving multi-image managercontrol data sets.
CA ACF2Use the following information when you migrate volumes that contain ACF2 RULES, LOGONIDS, andINFOSTG volumes with other volumes.
Give careful consideration when you migrate volumes that contain ACF2 RULES, LOGONIDS, andINFOSTG volumes with other volumes, especially if the ICKDSF TDMF parm option is requested. Volumesgoing through SWAP processing might start SAF authorization that in turn might require access to theseACF2 volumes. A hang might occur if these control volumes are also going through SWAP. To eliminateany form of hang, switch these data sets to their backup volumes before you migrate the primaries. If thisis not possible, then migrate these volumes individually.
CA-ASTEX performanceInformation is included here on maintenance fixes required for ASTEX's compatibility with TDMF.
Important: ASTEX V2.7 has maintenance fixes that are required for compatibility with TDMF.
ASTEX V2.8 dynamically detects swap migrations. User interaction is unnecessary. It supports OfflineVolume Access in TDMF V2R2.0 and higher. For more information, contact Computer Associates ASTEXTechnical Support.
CA-OPS/MVS event management and automationCA-OPS/MVS intercepts WTO/WTORs that are submitted by licensed programs, including TDMF, when theauto-ops feature is requested.
To migrate a volume that contains the OPSLOG data set, consider the following actions.
• Place the volume in a single volume session, if possible.• Turn off the auto-ops feature for all volumes in the session by specifying OPTIONS(NOAUTOOPS) on
the SESSION, GROUP, MIGRATE, or REPLICATE control statement).
For more information, see the CA-OPS/MVS Administrator Guide.
Chapter 5. Planning 69
CA-Scheduler job managementThe supported releases of CA-Scheduler support the use of four-digit UCBs or UCBs above the 16MB line.
When you migrate volumes that are related to CA-Scheduler, the started task (STC) must be inactiveduring the session. When the migration is completed, the STC can be reactivated. The commands toaccomplish these tasks are as follows.
P CASCHD - stops CA-SchedulerS CASCHD - starts CA-Scheduler
CA-1 tape managementWhen you migrate volumes that contain the CA-1 tape management catalog (TMC) and audit files, ensurethat CA-1 is inactive. When the migration is completed, TMSINIT must be run. CA-1 Release 5.2 genlevel9901 supports UCBs above the 16MB line.
Alternatively, a swap migration with the prompt option can be used. This method signals when thevolume is ready to synchronize at which time CA-1 is made INACTIVE. When CA-1 is INACTIVE, respondto the TDMF prompt. When the swap is completed, TMSINIT runs to reactivate CA-1. The TMC and theaudit files are briefly unavailable while CA-1 is INACTIVE.
For more information, see the CA-1 Administrator and Operator Guide.
CA-HYPERBUF VSAM buffer optimizationYou can improve processing performance using CA-HYPERBUF on z/OS VSAM applications.
Users of the CA-HYPERBUF product can either exclude TDMF from CA-HYPERBUF processing orimplement constraint list changes.
• If you are running CA-HYPERBUF V4.3.0, install PTFs QO15832, QO21795, and QO27604. At this level,CA-HYPERBUF looks for the statement
//CAHBEXCL DD DUMMY
The statement causes CA-HYPERBUF to bypass all of its subsequent activities.• If you are running earlier levels of CA-HYPERBUF, modifications to the constraint list are required.
MODE=EXCHANGE
PROGRAM=ARC+
PROGRAM=IEH+
PROGRAM=EDG+
PROGRAM=ADR+
PROGRAM=ICK+
PROGRAM=TDMF+
For more information, see CA-HYPERBUF documentation or contact CA-HYPERBUF technical support.
CA-7 job managementIf a volume contains the communications data set for use with CA-7, and ICOM is selected for a swapmigration, both the CA-7 and ICOM started task must shut down. When the migration is completed, thesetasks can be restarted.
Alternatively, a Swap migration with the prompt option can be used. This method signals when thevolume is ready to synchronize at which time the application shuts down. After the application has shutdown, respond to the TDMF prompt. When the swap has completed, both the CA-7 and IndependentCommunications Manager (ICOM) tasks are restarted. This operation minimizes the time that CA-7 andthe ICOM are down.
For more information, see the CA-7 Systems Programmer Guide.
70 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
CA-11 job managementIf a volume that contains the job run history file (JEHF) is to be migrated, then the JEHF I/O must be shutdown before the migration. When the migration is completed, the task can be restarted.
Alternatively, a swap migration with the prompt option can be used. This method determines when thevolume is ready to synchronize. Then the application shuts down. When the application shuts down,respond to the TDMF prompt. When the swap is completed, the JEHF I/O task can restart. This operationwould reduce the time that the JEHF is unavailable.
For more information, see the CA-11 Systems Programmer Guide.
CA-MIM resource sharingTDMF periodically issues a reserve macro (TDMFRESV) to the volume where the Communications data set(COMMDS) is resident. Dependent on how certain parameters are set within Multi-image Manager (MIM),this could cause the reserve to be converted by MIM to a global enqueue.
Important: It is recommended that the physical device reserve mechanism be used to ensureserialization of the Communications Data Set. In situations when the volumes suitable for theCommunications Data Set are shared across sysplexes and must not be physically reserved, theTDMFRESV resource can be converted by CA-MIM and propagated to the Agent systems using GDIF.
The following procedure helps determine what actions need to be taken to avoid this reserve conversion.This information has been supplied courtesy of CA-MIM Technical Support.
1. Go into the MIMINIT deck and search for PROCESS=
The options available are SELECT or ALLSYSTEMS. The differences between these two options aredescribed as follows:
• PROCESS=SELECT tells MIM to only manage and convert the enqueues and reserves that areindicated in the MIMQNAMES list and GDIEXMPT list.
• PROCESS=ALLSYSTEMS directs MIM to manage all enqueues and reserves that have aSCOPE=SYSTEMS unless the qname is specified in the MIMQNAMES list with GDIF=NO, or the dataset is specified in the GDIEXMPT list via a LOCAL statement.
2. To prevent MIM from converting the RESERVE to a global enqueue for the COMMDS data set, do one ofthe following:
• If running with GDIINIT PROCESS=SELECT:
Do not define the TDMFRESV qname to MIM via the MIMQNAME member.• If running with GDIINIT PROCESS=ALLSYSTEMS:
Define the following to the MIMQNAME member
TDMFRESV GDIF=NO
If assistance is required to code the MIM parameters, please contact CA-MIM Technical Support. Multi-image Manager (MIM) release 4.1 and above supports four (4) digit UCBs. Support for UCBs above the 16-MB line for disk and tape is available at maintenance level 9608. Support for CTC UCBs above the 16-MBline is available at maintenance level 9706.
For those customers using GRS, refer to “Coexistance with EMC’s z/OS migrator” on page 78.
Chapter 5. Planning 71
Moving multi-image manager control data setsIf you are doing a Swap migration for a volume that contains one or more Multi-image Manager (MIM)control data sets, deallocate the data sets before the migration, and reallocate them after the migration.See the CA-Multi-image Manager Statements and Commands manual.
Cache fast write and sort programsEnsure that cache fast write (CFW) is off for source volumes that are involved in swap migration, ifpossible.
A swap migration can be stopped with the GTD2363I message, which informs that Cache Fast Write(CFW) activity is detected on the source volume. This scenario occurs if the target volume is not part ofthe same DASD subsystem as the source volume. The accompanying GTD2596I message identifies theresponsible job name. Most often, a sort program causes the CFW activity. However, other applicationscan use CFW.
To circumvent this situation, a number of options are available:
• Turn off the CFW feature at the DASD subsystem level. Not all subsystems have the facility to do thisaction. For more information, see the hardware manufacturer's documentation.
• For sort work data sets allocated on work volumes in a DFSMS environment, the volume can be set to astatus of DISNEW (Disable new allocations). For more information, see “DFSMS” on page 73.
• Some applications that use CFW have a facility for turning off the option.
Note: Point-In-Time migrations are not stopped if CFW activity is detected. All CFW data is committedto the source volume by TDMF before it is copied to the target.
Class names and esoteric namesTDMF is not aware of DFSMS class names or system esoteric names that are defined within an operatingsystem.
DFSMS class name definitions are based on volume serial number. Therefore, no action is required after aswap migration since the volume serial number and class name are retained. A point-in-time session doesnot require an action if the target volume is relabeled to the serial number of the original source volumeon a different system and the volume serial number is defined to a class name.
Esoteric names are defined by device address. Therefore, the user must ensure that the esoteric namesare updated to include any device addresses that can be involved in either a swap migration or point-in-time migration.
Coexistence with other TDMF productsThe following use case scenarios apply to all versions of zDMF.
TDMF cannot be used to migrate volumes that an active zDMF migration session is using. Before you useTDMF to migrate volumes that are involved in a zDMF data set migration, zDMF must be shut down.
Attempts to use TDMF to migrate a source volume for a zDMF data set migration, causes error messageGTD2869E to display, and stops the volume migration. Attempts to use TDMF to migrate to a targetvolume for a zDMF data set migration causes error message GTD1117S to display, and stops the volumemigration.
72 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Cross memory servicesWhen TDMF is first run on an MVS system, a system linkage index (LX) is retained for the use ofsubsequent TDMF sessions.
The MVS name/token facility is used and remains valid until the next system IPL. The program call linkagethat TDMF constructs by using this LX enables the TDMF TSO Monitor to instantly notify the Master jobwhen an action is required for a volume or group. The default value for the NSYSLX variable (number ofavailable system linkage indexes) in the IEASYSxx PARMLIB member is 165. The use of one of these LXsby TDMF does not cause a problem.
Device states: Before and after a swap migrationThe states of a device are related back to the subsystem it is connected to. The states in question arecaching (CFW) and DASD fast write (DFW). A device cannot have CFW or DFW active at the volume levelwhile not active at the subsystem level.
TDMF notes the status of CFW and DFW for each source and target volume for a migration, but does notmodify the states of CFW or DFW to any subsystem. If CFW and DFW are on for the source (but not for thetarget) subsystem and devices when volume migration starts, the new source volume (original target)indicates that CFW and DFW are not on at the end of the swap migration.
The status of CFW and DFW can be determined through the TDMF TSO Monitor Feature through “Option 3- Current sessions: display messages” on page 152 .
By default, TDMF will set the Soft Fence on the original source device in a swap migration to preventinadvertent use of the device after the swap is complete. If a device soft fenced by TDMF had been FlashCopy enabled, it will have Flash Copy disabled after the swap. See “Recovering TDMF devices” on page 32for more information.
Data facility productsUse the following information when you work with control data sets or DFSMS.
Control data setsThe control data set of a data facility product (DFP) must not be active at the time of a swap migration.
Before any control data set (ACDS, BCDS, MCDS, OCDS) of the data facility products (HSM, SMS, RMM) ismigrated with TDMF, ensure that the data set is not active. If the DFP product has no means of releasingan active control data set (by using an alternative), then use the synchronization prompt option. This stepminimizes the downtime of the DFP application.
DFSMSIf you use DFSMS, submit the DISNEW (disable new allocation) command for all DFSMS managed targetvolumes of the session. You can complete this action through the ISMF windows before a volumemigration.
Note: DFSMS is an "umbrella" for products such as HSM, DSS, RMM, and DFP. Except for DFP, all havecontrol data sets. Control data sets are referred to as xCDS because the control data set can be ACDS,BCDS, OCDS, or MCDS. For more information, see “Migration of control data sets” on page 85.
Except for a few non-disruptive inquiries, any I/O operation to the target volume by anything other thanTDMF is rejected. A message displays with IOSB completion code 4A (the I/O is prevented). The messagecan be IOS000I or another MVS message that indicates an error through the application or MVS systemcomponent.
The use of the DCOLLECT function against a target volume fails. The format of DFSMS error messages canbe seen in the examples that follow.
Chapter 5. Planning 73
Table 17. Examples of DFSMS xCDS migration error messages
IDC21807I FAILURE DURING LSPACE SERVICE (SVC 78)PROCESSING - RETURN CODEWAS 4
IDC11813I ERRORS WERE DETECTED FOR VOLSER volser. ERRORBYTE WAS X'A0'
IDC21804I FAILURE DURING VVDS/VTOC ACCESS SERVICE PROCESSING - RETURNCODE WAS X'08'
IDC11808I ERRORS WERE DETECTED FOR data.set.name ON volser ERROR BYTE WASX'28'
IEE498I SWITCH LSPACE ERROR. RETURN CODE = 004 REASON CODE = 059 DEVICENUMBER = dev msgtxt
Note: After the session, the DISNEW command must be disabled. Any other commands, such as theQUIENEW command, submitted against the original source volumes must also be disabled. A failure to doso can cause errors such as the ones noted previously.
Important: DFSMS operates on a volume serial number. Therefore, the status of the volumes does notchange during a swap migration. The volume remains the same as before the session started. If theoriginal source volume is renamed, it can be unrecognized by DFSMS. Thus, the original target volumeserial can remain in a pool, while the new volume serial is not included in a pool.
Daylight saving timeAn active TDMF session operates normally during a local time change (a switch to/from "summer time" or"Daylight Saving Time"). Time intervals are not calculated by using local time. The TOD clock is used forall time stamps. The local time option can be specified for the session. Also, the local time option can beset as default when the security key is generated. For both options, the local time offset that is usedthroughout the session is the one in effect at session initialization.
For example, a session is started at 01.30 local time, one half hour before the switch to Daylight SavingTime (typically a Sunday late in March, for the northern hemisphere). For a 1-hour session with the localtime option in effect, the termination messages are time-stamped at 02:30. The time stamp is not03:30, which would be the correct local time.
If the SMF recording option was selected, the records are always written with the correct local time.TDMF generated SMF records are unlikely to cause any confusion when TDMF is run. During the switchfrom Daylight Saving Time (when local time 02:00-02:59:59 is repeated), some users stop all activityduring this extra hour. Be aware of the generation of these records.
The TSO Monitor Option 9 - Display Installation Options and Environment, displays the chosen SMFrecord number if this option was selected.
DFSortYou can use JCL to temporarily turn off CFW. Use either the DFSPARM or SYSIN DD statement to do so.These statements are only valid on 3990 controllers.
Examples:
//DFSPARM DD *DEBUG NOCFW
or
//SYSIN DD *DEBUG NOCFW
74 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
If the DFSPARM DD statement is used, the SYSIN DD statement is not necessary unless the job requireslink editing. For other examples of turning off CFW, see “Cache fast write and sort programs” on page 72.
For more information, see the DFSort Application Programming Guide (SC33-4035) and the IBM3990/9390 Operations and Recovery Guide (GA32-0253).
Note: Point-in-time migrations are not stopped if CFW activity is detected. All CFW data is committed tothe source volume by TDMF before it is copied to the target.
Duplexing functionsTDMF stops migration if volumes are detected that are involved in a primary/secondary duplexingfunction (such as Dual Copy). Such volumes can be found by standard architectural techniques. Duplexingfunctions that are started midway through a migration are included. The operation ensures secondarycopies are not jeopardized in a Disaster Recovery situation.
Before you begin
To ensure that secondary copies remain intact, suspend these session types so they are not active on anyvolume that is targeted for a TDMF session.
EMC DASDUsing EMC DASD might require the installation of EMC Application Interfaces, obtained from EMC.
Application program interfacesUsing EMC DASD may require the installation of EMC Application Interfaces, obtained from EMC, if thesource or target volumes use the Symmetrix Remote Data Facility (SRDF), with or without the EMCConsistency Group software installed or if any EMC SAR function is used.
If the EMC DASD uses standard Peer-to-Peer Remote Copy (PPRC) functions in lieu of SRDF, then the EMCapplication program interface code is not required.
EMC API code can provide three different functions:
• EMCSAI. The low-level Application Program Interface (API) that provides the Symmetrix Control Facility(SCF) interface
• CGRPQDEV. A low-level Application Program Interface (API) for EMC Consistency Group support inCGRPQDEV
• EMCSARQ. A low-level Application Program Interface (API) for EMC SAR support in SARQDEV
Important: TDMF detects if the EMC Symmetrix Control Facility (SCF) Application Program Interface(API) is installed in the operating system on all master and agent systems. If it is installed, the EMCSymmetrix Control Facility (SCF) Application Program Interface (API) must be either active, or it musthave at least been started since the last IPL, although it may not be currently active. If it is not active,and has not been started since the last IPL, TDMF will terminate with a severe error. Without access tothe interface, TDMF cannot detect the existence of the Symmetrix Remote Data Facility (SRDF) devicesor the use of an EMC Consistency Group.
EMC symmetrix control facilityCommunications to EMC DASD is accomplished via the EMC Symmetrix Control Facility, and EMCsubsystem names. To determine the need for additional SCF$xxxx DD statements, issue the MVScommand display (D SSI MVS) to see the subsystem names. The normal default EMC subsystemname for the Symmetrix Control Facility (SCF) is EMC. The normal default EMC subsystem name for theEMC Remote Data Facility is EMC2.
Thus, it is not necessary to have SCF$xxxx DD statements in any TDMF job if there is not more than oneSymmetrix Control Facility (SCF) started procedure. If you are only running one version of the SCF
Chapter 5. Planning 75
subsystem, it defaults internally to EMC if you do not specify a value with the DD statement. It is requiredwhen running multiple copies of SCF in the SAME LPAR, and you need a different SCF$xxxx DD statementso other EMC applications, such as TimeFinder, etc. can match up with the SCF using this same DDstatement.
EMC provides the SCF$xxxx DD statement function for the ability to test different levels of EMCsubsystem code within the same LPAR at the same time.
Consistency groupsTDMF, in conjunction with the proper microcode and EMC Consistency Group software, provides supportfor the swap migration of source volumes that reside on EMC DASD devices utilizing the SymmetrixRemote Data Facility (SRDF), and optionally can exist as part of an EMC Consistency Group.
EMC Symmetrix Remote Data Facility (SRDF), with or without Consistency Group support, ensures that acustomer’s disaster recovery environment is not inadvertently destroyed or changed. TDMF is designed tobe consistent with that type of data protection. At session initialization, TDMF ensures that thecharacteristics of the source and target volumes involved in a swap migration appear to be the same onall systems in the TDMF session. Throughout the migration, TDMF ensures that the volume characteristicsdo not change. If TDMF detects inconsistencies in the SRDF session, then the TDMF session isterminated.
When using source or target volumes controlled by EMC Symmetrix Manager, TDMF supports ConsistencyGroups at level 6.0.0 (FMID SCGP600).
Important: TDMF also detects EMC Symmetrix/Dynamic Address Switch (S/DAS) support, also known asAutoSwap, but does not currently support this EMC function. If the use of the AutoSwap function isdetected, TDMF terminates with a severe error. To migrate AutoSwap volumes, see Chapter 11,“Migrating EMC autoswap volumes,” on page 197.
EMC configuration guidelinesWhen defining EMC devices as members of a Consistency Group, all systems in the shared DASD complexmust activate the EMC CONGroup started task, and the parameters must agree.
EMC DASD devices should be defined via the device-number parameter for the CONGroup configuration,not by volume serial number. An EMC CONGroup configuration should not be changed while a TDMFsession is active. If any changes occur that may inadvertently modify the status of an EMC DASD volumethat is a part of a Consistency Group, TDMF will dynamically terminate the volume migration.
Symmetrix managerIf the source or target volumes are controlled by the EMC Symmetrix Manager, those volumes must beremoved from the Symmetrix Manager’s control, or the Symmetrix Manager must be shut down during amigration session. Failure to do so will cause TDMF to terminate the migration(s) with an error message(GTD2362) indicating an invalid operation code when the Symmetrix Manager issues either a DiagnosticWrite (73) or Diagnostic Control (F3) operation code. Message GTD2596I will follow the error messageindicating where the invalid operation code came from.
See the EMC ResourcePak documentation from EMC for details on removing devices from and addingdevices to EMC Symmetrix Manager control.
Note: Target volumes may not be assigned to EMC Symmetrix if TDMF For Platinum is in use.
Migrating from SRDF to PPRC/HDS TrueCopy volumesTDMF will allow the migration of a device involved in an SRDF session to a volume that is involved in aPPRC (synchronous) session, or a HDS volume active with either a synchronous or asynchronous
76 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
TrueCopy session. In order to allow swaps from these SRDF volumes, the common optionALLOWmirrorchange must be specified.
About this task
For more information on migrating between SRDF and PPRC volumes, see “Migrations outside a singlesysplex” on page 91.
TDMF will require volume confirmation for each volume requesting the ALLOWmirrorchange option.Confirmation can be supplied using the TDMF TSO Monitor, Option 2, A command, or the MVSconsole if the auto-operations interface has been enabled. This confirmation must be supplied beforesubsequent ‘normal’ volume or group confirmation or volume or group selection will take place.
The ALLOWmirrorchange option is not required to be specified in the same way for all volumes in a group.
Following is a syntax example for the ALLOWmirrorchange/NOALLOWmirrorchange option keyword,which is part of the Common Options. For information about this option, refer to “Common options for theSESSION, GROUP, MIGRATE, and REPLICATE statements” on page 50.
+-NOALLOWmirrorchange------------------+
>--+------------------+-------------------+---+-----------------+--->
+-(ACKnowledge)-----+ +-RESETCHAngedflag-+
+-ALLOWmirrorchange--------------------+
+-(NOACKnowledge)---+
SRDF session types allowedTDMF supports sessions where a source or target volume might be involved in an SRDF session.
The following types of active SRDF sessions are available at initialization of a TDMF session:
• Point-In-Time migrations.• Swap migrations, where the source and target volumes are involved in an active SRDF session, and the
source volume is not a member of a Consistency Group.• Swap migrations, where the source and target volumes are involved in an active SRDF session and both
are members of the same Consistency Group on all systems in the TDMF session.• Swap migrations, where the source volume is not involved in an active SRDF session, and the target
volume is involved in an SRDF session that might or might not be a member of a Consistency Group.
Table 18 on page 77 indicates availability when you are migrating from a source volume by using onemirroring technique to a target volume that is using the same or different mirroring technique, and whenthe use of the ALLOWmirrorchange option is required.
Table 18. Source and target availability
Target Device
Non-mirrored TrueCopy PPRC Primary SRDF Primary SRDF and ConsistencyGroup
XRC Primary
Source Device
Non- mirrored Y Y Y Y Y Y
TrueCopy Y
Available ifABANDONTrueCopyoption is used
Y Y
Available ifALLOWmirrorchangeoption is used
Y
Available ifALLOWmirrorchangeoption is used
Y
Available ifALLOWmirrorchangeoption is used
Available ifALLOWmirrorchangeoption is used
PPRC Primary Y
Available ifABANDONPprc optionand ALLOW is used
Y
Available ifALLOWmirrorchange optionis used
Y Y
Available ifALLOWmirrorchangeoption is used
Y
Available ifALLOWmirrorchangeoption is used
Y
Available ifALLOWmirrorchangeoption is used
SRDF Primary Y
Available ifABANDONSrdf option isused
Y
Available ifALLOWmirrorchange optionis used
Y
Available ifALLOWmirrorchangeoption is used
Y Y Y
Available ifALLOWmirrorchangeoption is used
Chapter 5. Planning 77
Table 18. Source and target availability (continued)
Target Device
Non-mirrored TrueCopy PPRC Primary SRDF Primary SRDF and ConsistencyGroup
XRC Primary
Source Device
SRDF andConsistencyGroup
Y
Available ifABANDONSrdf option isused
N Y
The migration IS allowedwhen ABANDONSrdf isspecified and both thesource and target areEMC devices
N Y
Source and targetvolumes must bemembers of the sameConsistency group
N
XRC Primary Y
Available ifABANDONXrc option isused
Y
Available ifALLOWmirrorchange optionis used
Y
Available ifALLOWmirrorchangeoption is used
Y
Available ifALLOWmirrorchangeoption is used
Y
Available ifALLOWmirrorchangeoption is used
Y
Coexistance with EMC’s z/OS migratorEMC’s z/OS Migrator should not be used to migrate data to or from any volumes being used as sourcevolumes in active TDMF migration sessions. TDMF cannot detect the presence of z/OS Migrator andcannot determine its intentions, any attempts to use z/OS Migrator and TDMF to migrate the same datacould lead to unpredictable and unintended results.
Enhanced catalog sharing (ECS)When using TDMF to perform a migration of a volume containing a catalog managed by ECS, it is possiblethat the quiesce is not immediately processed.
The message, IEC378I catname REMOVED FROM ECS DUE TO DDR SWAP, is one of the symptomsrelated to this problem. It is also possible to note discrepancies between catalog entries in the catalogfrom a sharing system.
Extended functionsIf data striping and VSAM data compression is supported on the source subsystem, it is the user’sresponsibility to ensure that these extended functions are supported on the target subsystem as well.
It is the responsibility of the user to ensure that the following extended function is not active on a sourcevolume:
• Concurrent Copy session
The LISTSESS operation will report if a Concurrent Copy session exists for a volume. The MODIFYcommand can be issued from a system that has the XRC software installed, and has the ANTAS000address space active. The command syntax for this command is:
F ANTAS000,LISTSESS addr
For more information about this command and the ANTAS000 address space, refer to the DFSMS/MVSRemote Copy Administrator’s Guide and Reference (SC35-0169).
TDMF will terminate a swap migration if a concurrent copy session is active on the source volume atquiesce (volume synchronization) time.
File allocation managersThe TDMF communications data set (COMMDS) must be allocated in contiguous cylinders. If a productthat actively manages file allocation (for example, ProSMS or SAMS: Vantage) is installed, be aware thatthe possibility that the message GTD1080S might be created, causing the migration session to end. Thismessage states that the COMMDS space is not contiguous. The COMMDS needs to be reallocated outsideof the control of the file allocation manager.
78 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Global resource serializationInformation to ensure that the reserves that TDMF issues are not converted to global enqueues by GRS.
TDMF periodically places a hardware reserve on the volume where the communications data set(COMMDS) resides.
Important: It is recommended that the TDMF reserves are not converted by a resource-serializationproduct such as GRS or CA-MIM. Avoiding this conversion is important if systems that participate in avolume migration are not part of the complex that is covered by the resource serialization product. If aquestion exists about the integrity of the serialization environment, TDMF fails to initialize.
To ensure that the reserves that TDMF issues are not converted to global enqueues by GRS, use thefollowing parameter in GRSRNLxx:
RNLDEF RNL(EXCL) TYPE(GENERIC) QNAME(TDMFRESV)
Similar definitions might be required when you use another resource serialization product like CA-MIM.
Host software componentNote the following for installations using StorageTek’s Host Software Component (HSC) for LibraryStorage Module (LSM).
In those installations using StorageTek’s Host Software Component (HSC) for Library Storage Module(LSM), the following should be noted.
TDMF will successfully migrate the HSC’s Primary or Shadow Control Data Sets (CDS). However, if a CDSis subsequently disabled and re-enabled without recycling (P HSC and S HSC) the HSC Address Space ofall Systems that share it, the HSC will be unable to successfully allocate the CDS being enabled and willgenerate a dynamic allocation error 228 or 22C.
Important: IBM recommends that only one Control Data Set be moved at a time in order to avoid a deadlyembrace.
For information about other Control Data Sets, see “DFSMS” on page 73.
HCD and Input/Output definition filesThe specified values of the hardware configuration definition (HCD) files must be updated for migratingvolumes containing Input/Output Definition Facility (IODF) files before the next initial program load (IPL).
• Load Parameter Definition – You can change this value by using the Operator Control (OPRCTL) frame,or the System Control (SYSCTL) frame.
• Load Profile(s) – For CMOS mainframes.• JES3 Initialization Stream Checker – Using the Create JES3 INISH Stream Checker Data window.
It is your responsibility to ensure that the new device address, or addresses, be communicated to theappropriate areas before the next IPL. Failure to do so might result in a Wait State during IPL. TDMFissues a warning message when an IODF or load-parms volume is migrated.
For more information, see the z/OS HCD User's Guide (SC33-7988).
Chapter 5. Planning 79
HFS and zFS filesTDMF supports the migration of volumes that contain Hierarchical File System (HFS) files or zSeries FileSystem (zFS) files.
Be aware that any systems that are not running a TDMF Master or Agent (that is, any UNIX, Windows/NT,or other systems attached via an Open Systems Interface) should not be actively sharing volumes duringthe swap process. It is your responsibility to reflect any changes to such systems.
ICKDSFDo not run ICKDSF against any volumes involved in an active session. TDMF will dynamically detect theCCW operation codes from this program and terminate that volume’s migration. If an ICKDSF function isnecessary, terminate the migration for that specific volume via the TDMF TSO Monitor, and then restartthe volume migration using the TDMF TSO Monitor or the Batch Monitor.
Dynamic ICKDSF REFVTOC/EXTVTOCAfter a migration, TDMF can call ICKDSF to dynamically reformat or expand a volume's VTOC.
Use this function when the source volume's characteristics do not match the target device, and ifrequested by an installation or migration session option. See “Migration sessions with a mixture ofmaintenance levels” on page 21 for more detailed information. Add DD statement DSFPRINT to themigration session (Master) JCL if the automatic ICKDSF option is requested.
Note: The source volumes that call the ICKDSF program are required to be mounted as PRIVATE.
If SYSVTOC is not specified in the GRS SYSTEMS Exclusion RNL, add it to the Reserve conversion RNL.
Catalogs or VSAM data setsWhen migrating a volume that contains either a catalog or VSAM data sets, and the Auto-ICKDSF option isselected, TDMF takes steps to avoid the possibility of a deadly embrace situation if the SYSZVVDS andSYSVTOC reserve requests for this volume are not treated the same.
However, the recommendation is that SYSVTOC, SYSZVVDS, and SYSIGGV2 either be put in the exclusionlist or conversion list as in the following examples (or be treated identically by CA-MIM, if applicable).
For more information, the following references might be used.
• IBM Informational APAR ll10752 (Item 7 under Catalog Enqueue and Dequeue Resources)• z/OS MVS Planning: Global Resource Serialization (SA22-7600), topic: RNL Candidates, Table 4 -
SYSTEMS Exclusion RNL Recommendations.
When acquiring the SYSZVVDS resource for a volume results in a physical device reserve, and the auto-ICKDSF option is in effect for a volume, TDMF takes steps to avoid a possible lockout at quiesce timebetween a catalog address space on an Agent system and the TDMF Master job (resource namesSYSZBNDX/volser and SYSZVVDS/volser). For example, the volume is not resumed on Agent systemsuntil ICKDSF is complete on the Master system, and TDMF starts "ENQ lockout" and "start pending"monitors before starting ICKDSF. Another precaution that is taken by TDMF in this situation is to suppressthe PAV option on Agent systems.
How dynamic ICKDSF REFVTOC or EXTVTOC functions affect a migrationICKDSF is invoked on the Master system for each applicable volume immediately after the successfulswap on all systems, if the VTOC can be reserved on the target device.
Note: During this process, the volume is available to other jobs on the Master system, but is unavailableto the Agent system(s) due to a volume reserve. However, if the SYSVTOC reserve was converted to a
80 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
global ENQ, the volume will be immediately available on the Agent systems—except for access to theVTOC.
Dynamic ICKDSF processing will not be performed in Approval Mode. Manual ICKDSF processing mightbe required after a migration was approved in order to enable full device functionality.
By default, ICKDSF message output is discarded. If a //DSFPRINT DD statement is supplied, TDMF willgroup the ICKDSF output messages for each volume and write them out. The //DSFPRINT DD statementshould not refer to an existing data set.
The Master system will continue to wait until the last copy task has completed before terminating thevolume migration or session.
The order of events is as follows:
1. At quiesce time, the SYSVTOC ("SYSVTOC/volser") resource is acquired on the Master system.2. The volume is swapped.3. ICKDSF is invoked on the Master system, with the source volume reserved.4. The volume is resumed on all systems while ICKDSF is processing.5. The volume is terminated on all Agent systems.6. The volume is terminated on the Master system.
The TDMF Master system will not end until all ICKDSF tasks have completed, although the TDMF TSOMonitor will not depict it as an active session once the last volume has terminated. In a multi-systemenvironment, the interval between Items 3 and 6 is a few seconds. The larger the VTOC, the more time itwill take for ICKDSF to complete.
It is recommended that a TDMF session that will invoke dynamic ICKDSF have a minimum region size of96M and no more than 16 volumes running concurrently.
Dynamic ICKDSF REFVTOC functionThe dynamic ICKDSF REFVTOC function pertains to migrations only where the number of alternatecylinders in the source volume VTOC does not match that of the target device. If no mismatch exists, noaction is taken. To enable ICKDSF for correcting alternate track counters of volumes when no mismatchexists in the number of primary cylinders, the option must be specified on the SESSION control statementwith the optional ALTTRKS parameter.
If a REFVTOC is not processed to tailor the VTOC and its index to the target device, any programsdependent on the alternate track information being correct may fail after the swap, including olderversions of DF/DSS (for full volume backups) and DCOLLECT. Later releases of these programs andrelated z/OS components have become more tolerant of incorrect "device characteristics" information inthe VTOC to such a degree that the ICKDSF Release 17 INIT and REFORMAT functions no longer place thedevice's actual alternate track configuration data into the VTOC: the DS4NOATK and DS4DEVAC fields arealways set to zero (this is fully documented in IBM APAR II11078). As a consequence of this: if ICKDSFRelease 17 is installed on the Master system, TDMF will only invoke REFVTOC processing to correctalternate track information in the VTOC when these fields are not already zero.
REFVTOC processing of a volume with an indexed VTOC may change the location and size of the indexdata set. In this situation, ICKDSF will indicate to the z/OS allocation component that the description ofthe VTOC index in the Volume Information Block (VIB) in virtual storage must be refreshed. In order tocause the in-storage description of the index data set to be reinitialized on each of the other z/OS systemsthat share the volume, it is normally required to vary the volume offline and online again - unless all thesharing systems are in the same parallel sysplex. TDMF avoids this possible disruption by processing theVIB on each of the Agent systems while ICKDSF processing is active on the Master system. What cannotbe prevented is a certain amount of disruption caused by the SYSVTOC reserve being held for the sourcevolume throughout the synchronization, swap and REFVTOC phases.
If a volume has an index data set that has been disabled for any reason, ICKDSF will consider the volumeto be non-indexed:
• The index data set will be deleted by normal ICKDSF processing.
Chapter 5. Planning 81
• The Volume Table of Contents (VTOC) will be rebuilt in OSVTOC format.
To prevent this from occurring, TDMF will invoke ICKDSF for a "BUILDIX IXVTOC" function before the'REFORMAT REFVTOC'. This 'BUILDIX' is only initiated if a VTOC index data set exists on the volume, andif the volume is under SMS control or the index was disabled by a DADSM function.
Intent of dynamic ICKDSF EXTVTOC functionThe ICKDSF EXTVTOC option will only be invoked in the case of migrating from a smaller device to a largerdevice or if a volume had been previously migrated and no REFVTOC had been performed. Only indexedvolumes are considered for EXTVTOC processing. Non-indexed volumes, including volumes with adamaged VTOC index will only be processed using REFVTOC. If neither EXTVTOC(nn) or NOEXTVTOCoption is in effect for a volume, the minimum size of the new VTOC is the greater of the current VTOC sizeon the source and target. The VTOC may be extended further depending on the number of data sets onthe volume:
• If the volume is less than half full, the VTOC will be extended to contain the current numbers of datasets multiplied by the ratio of target to source device size, plus 25%.
• If the volume is more than half full, the VTOC will be expanded to handle the situation where the targetvolume is full of data sets with the same average size.
When the EXTVTOC(nn) option is specified for a volume, TDMF attempts to expand the VTOC to therequested size:
• If the number of tracks specified in the option is not greater than the current size of the source VTOC, orif NOEXTVTOC was specified, TDMF will only invoke ICKDSF to reformat the Volume Table Of Contents(REFVTOC).
• If the index needs to be extended, but the VTOC does not, and the level of the ICKDSF programinstalled on the Master system will not extend the index data set as part of the REFVTOC function,TDMF will attempt to extend the VTOCIX by one track.
• Where there are data sets on a volume using the tracks into which the VTOC will expand, the FORCEparameter on the EXTVTOC option will cause TDMF to attempt to move these data set extents to someother place on the volume, thus allowing the VTOC expansion to take place.
• The determination of whether a data set can be moved to expand the VTOC is made by testing thestandard SYSDSN ENQ to determine if the data set is in use. Data sets opened with a NODSI optioneither via dynamic allocation or with the SCHED parmlib member could be moved while in use.
• When EXTVTOC is specified with the UNCONDITIONAL parameter for a volume, the migration will beterminated as soon as TDMF detects that the presence of a data set extent close to the VTOC wouldprevent the requested VTOC expansion. Before using this parameter, the user should attempt toprevent new allocations on the volume, and run the GTDEXTV program to determine whether a data setis already allocated on tracks that the VTOC should expand to.
Note: The FORCE and UNCONDITIONAL parameters can be used independently or together.
Use of the FORCE parameter with EXTVTOCThe EXTVTOC FORCE parameter instructs TDMF to attempt to move data set extents away from the spaceinto which the VTOC must expand. To do this, TDMF constructs a map of the volume at the end of theCOPY phase, then modifies the map whenever the VTOC itself is updated. This allows TDMF to determinewhich data sets are allocated in the space adjacent to the VTOC. The VTOC map and associated tables arebuilt in 64-bit virtual storage, if possible. It is recommended that the user specify either REGION=0M orMEMLIMIT=NOLIMIT on the job statement for the TDMF Master job, if the FORCE parameter is required.
To move a data set extent away from the space adjacent to the VTOC, the Master system issues an ENQmacro for the data set or cluster that owns the extent. If the ENQ is unsuccessful, for example, becausethe data set is currently allocated to an application, the data set is considered unmoveable, and the VTOCwill not be expanded into the space it occupies.
Important: It is critical for data integrity that the SYSDSN/dsname ENQ be propagated to all the systemsthat are sharing the volume. TDMF tests the availability of the SYSDSN/data set_name ENQ resource todetermine whether a data set is currently in use by an application. If there are data sets on a volume that
82 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
might be allocated to an application in the installation, without the ENQ request having been propagatedto the z/OS system on which the TDMF Master job is running, the FORCE parameter must not be used.
If there is any doubt about the data sets on a particular volume, the GTDEXTV program can be used toidentify the data sets that TDMF will attempt to move. Other immovable data sets are the volume's VVDS(SYS1.VVDS.V<volser>), and any that have the immovable data set organization attribute.
Data set extents that are moved by TDMF in response to the FORCE parameter are moved to space on thesame volume that is allocated by TDMF before the volume is swapped, which means there must besufficient free space on the source volume before it is migrated to the larger target device. Space that waspreviously occupied by these extents is not made available for reallocation. When TDMF must move anextent that is larger than the space needed for the VTOC expansion, this additional space will becomeavailable when ICKDSF rebuilds the VTOC index, as part of the REFORMAT EXTVTOC function, after thevolume has been swapped.
Note: Because TDMF uses the VTOC index to allocate space on the volume, the FORCE parameter will berejected if it is specified for a volume that does not have a valid VTOC index. The FORCE parameter willalso be rejected unless the TDMF Master system is running on a system at the z/OS V1R6 level or later.
HSM at APAR level OA21574For TDMF option ICKDSF, the Catalog Address Space can appear to suspend without the processing ofMODIFY CATALOG operator commands. This situation can occur during a quiesce of selected catalogs outof ECS (Enhanced Catalog Sharing). The solution is the application of IBM APAR OA16358 with therelevant PTFs.
About this task
The Catalog's Modify task and a Service task are together held up with a BCS enq through SYSIGGV2 andthe ECS Lock. This timing issue occurs during a time of quiesce where catalogs are being removed fromECS. Since the Modify task is involved, attempts to use modify commands, F CATALOG,ABEND and FCATALOG,RESTART do not take affect and the only remedy is to Force catalog address space out.
D GRS,C will only show contention building behind SYSIGGV2 for the catalog in question as the ECS lockis not GRS managed. An F CATALOG,LIST may show a service task waiting on this lock. The case athand occurred during a DDR Swap but this issue can occur for any system quiesce.
To resolve this, apply IBM APAR OA16358 along with the relevant PTFs.
IPL volumesIt is best to migrate active IPL volumes during a low activity period. Test this function in your testenvironment before you run a migration in a production environment. It is your responsibility to ensurethat one or more new device addresses are communicated to the appropriate areas before the next IPL.Failure to do so can result in an invalid IPL.
Invalid count fieldsThe Count-Key-Data (CKD) and Extended-Count-Key-Data (ECKD) channel command sets writedata out to a volume in a CKD format.
About this task
The count field is in the format of CCHHR, where CC is cylinder, HH is head (or track) and R is the record.TDMF checks the CCHH portion of the count field. If the cylinder and track address in any source recordcount field is not equal to its physical location on the volume, then TDMF will normally stop copying data,scan the remainder of the source volume (volume analysis) and terminate the volume pairing withmessage GTD3550E – Invalid count field.
Chapter 5. Planning 83
An invalid count field can occur in a free space area of the volume. In that case, select the FASTCOPYoption, which may prevent this error from recurring or run program IEBDG or IEBGENER in order topopulate a data set (FASTCOPY will copy unallocated tracks if they share a cylinder with a data set). Eitherof these programs will re-write the count field, then delete the temporary data set. If the free space areais large, it is recommended to allocate a data set large enough to span the entire area. If there is data inthe area, a re-write of the data set will be required.
If the error occurs in the middle of a data set, it could potentially be a "broken" data set. This does notmean that the data is inaccessible. Frequently, the data can still be accessed. If it is impractical to repaira data set on the source volume that contains invalid count fields, the user can request that TDMF migratethe volume, including the non-standard count fields, after issuing a warning message (see the optionspecifications that are valid only on the SESSION control statement). Whether the migration wasterminated or completed with a warning message, the disk locations of all instances of records with aninvalid count field can be reviewed using the TDMF TSO Monitor (Option U.13 – Detected Source VolumeI/O Errors).
If records with invalid count fields are known, or suspected, to exist on the source volume, it isrecommended that program GTDSCAN be run against the volume in order to determine where the invalidcount fields are.
Note: Bad count fields are checked when the data is written to the target volume. Therefore, for areplication involving TCP/IP, the SESSION statement on the Remote system must specifyOPT(ALLOWINV). For a replication with both a local and remote target volume, the option would need tobe specified for both sessions.
JES considerationsTDMF can migrate JES2/3 checkpoint and spool volumes. However, you must run the appropriate JEScommand to drain the spool to limit the amount of activity for that volume.
Recommendations for spool or checkpoint volume migrationYou can prevent logging function delays for the migration of spool or checkpoint volumes through thefollowing recommended practices.
TDMF will delay the run of any function, such as allocation/deallocation, that might require access toSYSLOG until there is no longer any volume in a quiesced state.
However, the drawback is that when migrating JES spool or checkpoint volumes, TDMF may encounter anunexpected logging function that requires JES2/JES3 to be able to access a volume that TDMF hasquiesced. In this situation, the session will hang until it has been canceled. For this reason, it is stronglyrecommended that only inactive or drained checkpoint and spool volumes are migrated. If this isimpractical, then under no circumstance must there be more than one checkpoint or spool volumemigrated at one time.
It is recommended you try this function in a test environment before migrating spool volumes in aproduction environment.
JES2 considerationsWhen planning to migrate a volume that contains the JES2 Checkpoint data set, it is important to checkthe HOLD parameter on the MASDEF statement in the JES2 parameters. The HOLD parameter specifies thelength of time that the checkpoint is reserved and unavailable to other requesters.
If the HOLD parameter is not specified, the default is HOLD=99999999, which means that JES2 will neverrelease the reserve. This causes the volume migration to "hang" during refresh processing. TDMF cannotsuccessfully quiesce the I/O when any task has a reserve on the source volume. In order to avoid thissituation, it is recommended that the JES2 command $T MASDEF,HOLD=100 (one second) be issueduntil the migration is complete.
For more information, refer to the applicable OS/390 JES2 Initialization and Tuning Guide or theapplicable z/OS JES2 Initialization and Tuning Guide.
84 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
JES3 considerationsIn order to ensure that JES3 system defined volumes will migrate (not required for a Point-In-Timemigration) in a TDMF (or P/DAS) environment, APARs OW23271, OW28455, and OW28457 must beapplied. These APARs provides JES3 DDR support for P/DAS and therefore, will allow the swapping ofvolumes.
Important: All systems sharing devices where JES3 manages the devices must be involved in the TDMFsession running. This ensures that all JES3 internal tables are properly updated. Failure to do so willcause unpredictable results.
It is recommended that the user check the UCB for the following bit prior to copying volumes in a JES3environment.
UCBJ3DV - device is defined to JES3.
If the bit is off, TDMF will migrate the volume(s) with no errors. If the bit is on, TDMF will make theappropriate calls to JES3 to notify JES3 of the volume redirection needed.
Important: Prior to migrating volumes in a JES3 environment, all target volumes that are JES3 systemdefined must be varied offline to JES3 and online to MVS. The following commands will perform thesefunctions: *V addr mainproc OFF <-JES3 command*V addr,ONLINE <-MVS command
If any of the original source volumes are to be reused after a Swap migration, the following commandmust be issued: *V addr RECOVER mainproc
If any of the original target volumes are to be reused after a Point-In-Time migration, the followingcommand must be issued: *V addr RECOVER mainproc
The following JES3 command will display the status of a volume before or after a migration:*INQUIRY,S,V=volser
If the XTYPE parameter is used with the SETNAME and DEVICE statements, the source and targetvolumes must have the same XTYPE name. If the XTYPE names are not the same, TDMF will terminatethe migration session with message GTD1228S.
Note: JES3 does not support the optional New Volume Serial (NEWVSN) for the original source volumemigration.
If a system is not a participant in a migration, prior to that system joining the complex, the original sourcevolume(s) must have the following command issued to them: *V addr RECOVER mainproc
Failure to issue this command will result in JES3 possibly corrupting the information related to thosevolumes that had been migrated.
LISTDATA informationWhen using the LISTDATA command via IDCAMS to verify the status of the subsystem or volumes (forexample, CFW on or off), there are some commands that query the physical controller and others thatobtain the information from memory.
The following command queries the physical controller: LISTDATA DSTATUS SUBSYSTEM
The command below gets its information from memory. Therefore, if CFW has been turned off at thedevice level, it may not show up in this report. The commands that do physically query the controller are:LISTDATA COUNTS DEVICE and LISTDATA STATUS
Migration of control data setsTDMF can migrate a volume that contains Control Data Set (CDS) for any product.
Important: Migrate only one CDS per product at a time. This practice ensures that no lockouts or deadlyembraces occur.
Chapter 5. Planning 85
Products or subsystems that might be affected by this activity.
• CA-MIM. See “CA-MIM resource sharing” on page 71• StorageTek’s Host Software Component. See “Coexistance with EMC’s z/OS migrator” on page 78• Couple Data Sets related to a Parallel Sysplex environment. See “Extended address volumes (EAV)” on
page 90• JES checkpoint and spool volumes. See “Recommendations for spool or checkpoint volume migration”
on page 84.
Refer to the following for more detailed information on other products or subsystems that are affected.
Migration of stand-alone dump data setsIf the installation has generated the stand-alone dump program to use dump data sets on direct accessdevices, the program locates the data sets using device number rather than volume serial number.
For this reason, you must ensure that the stand-alone dump program and the data sets are recreated, byusing the AMDSADMP macro and the AMDSADDD REXX or IPCS SADMP dump data set utilities, if any ofthe dump data sets are migrated. TDMF issues a warning message whenever a volume containing whatappears to be a stand-alone dump data set is migrated.
Migration of RAID and non-RAID devicesTDMF will migrate volumes that are RAID or non-RAID devices.
However, there are technology differences between non-RAID, RAID-S, RAID-3 and RAID-1 devices.Traditional non-RAID devices (33x0) normally have an alternate cylinder for assignment at the track levelfor those times when a track went "bad". Some RAID-S and RAID-3 devices also have this ability or areconfigured to look like the ability is present. RAID-1 devices normally do not have an alternate cylinder.This presents a compatibility issue when moving between the different technologies. To avoid thissituation, use the ALTTRKS parameter on the Dynamic ICKDSF REFVTOC option.
Migration of unlike device typesIt is now possible to move larger devices to smaller devices, in a Swap migration only.
To move larger devices to smaller devices in a Swap migration, the following must be true:
• Swap migration only• Dynamic ICKDSF option selected• The source volume must have a valid indexed VTOC• "excess" cylinders must be shown as available for allocation in VPSMs at initialization and throughout
the migration until it is time for synchronization.
TDMF copies the number of cylinders from the source volume equal to that of the target volume. Databeyond the size of the target volume is not copied.
It is possible to reverse the original swap migration if a volume is not moved to a larger target device andICKDSF REFORMAT is not initiated.
Note: The source and target volumes must be of the same track geometry. TDMF does not support themigration of 3380 device types (track size 47476) to 3390 device types (track size 56664).
When migrating from a smaller to larger device, the VTOC information might need to be updated to reflectthe extra cylinders for later use. For more information, see “ICKDSF” on page 80.
86 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Migration of volumes formatted for VM useYou can now use TDMF on MVS to copy and relabel VM volumes.
TDMF cannot monitor updates to the source volume from VM while the migration session is in progress,nor can VM (CP and CMS) control blocks be manipulated to achieve a non-disruptive migration.
To migrate a VM formatted volume, it must be varied offline from all attached VM systems. In the instanceof system, spool, or page volumes that are designated for migration, the associated VM system orsystems must be shut down and succeeded by an IPL later.
Note: When you migrate a VM formatted volume the FASTCOPY option is not available, as VM volumes arenot required to have a VTOC. For this reason, the information that is required for the FASTCOPY option isunavailable.
Important: The data integrity of a migration is not ensured if a VM system is not detached from thesource volume during a TDMF session.
Notes:
• If CFW is off at the subsystem level or if caching is off at the device level when Model 204 is started, thisaction should not be necessary. For more information, see the Model 204 Command Reference Manualand the Model 204 Messages Manual.
• Point-In-Time Migrations are not ended if CFW activity is detected.
Model 204 considerationsModel 204 takes advantage of Cache Fast Write (CFW).
Model 204 takes advantage of Cache Fast Write (CFW) for files CCASERV and CCATEMP if the DASDsubsystem supports CFW, caching is active at the volume level, and you changed the default Model 204CACHE parameter from X’00’.
If the update activity is enough for CFW to be called during a migration, TDMF stops the volume with errormessage GTD2363E – GTDIMON error due to the detection of a Cache Fast Writeoperation.
However, it is possible that during the life of a migration, the level of update activity does not warrant I/Ooperations to be run against the previously mentioned files. Subsequent to the swap of the volume, an I/Ooperation might be started to the files that use CFW, which causes the following Model 204 errormessages to be created.
2163 – TROUBLE WITH DISK I/O ON FILE% C POST CODE = X’%X’ CSW X’%X’
2164 – TROUBLE WITH DISK I/O ON POST CODE = X’%X’ CSW X’%X’
These error messages are created because the Subsystem Identifier (SSID) for the target volume (newsource) is different from the original source volume.
Set the Model 204 CACHE parameter to X’00’ before a TDMF migration. You must run a recycle of theModel 204 address space that might have a performance impact on the address space.
Notes:
• If CFW is off at the subsystem level or if caching is off at the device level when Model 204 is started, thisaction is not necessary. For more information, see the Model 204 Command Reference Manual and theModel 204 Messages Manual.
• Point-In-Time migrations are not stopped if CFW activity is detected.
Chapter 5. Planning 87
z/VM Agent considerationsIn z/VM® environments, two associated types of volume labeling are applicable to virtual guest andmultiple volume migrations.
Virtual guest definition
A successful migration of a source volume to a target volume results in assigning the original volume labelto the new real device number.
If volumes are assigned to guest virtual machines by a DEDICATE statement that specifies a real deviceaddress or by an MDISK statement that uses the DEVNO parameter, the directory entry needs to beupdated to use the new real device address.
The following example shows what can occur if real device addresses are used to assign volumes to guestmachines.
• Guest system LINX01 has the following DEDICATE statement in its directory entry:
DEDICATE 1201 5EC0
• Volume LXDAT1 is mounted on 5EC0:
#cp q v 1201DASD 1201 3390 LXDAT1 R/W 60102 CYL ON DASD 5EC0 SUBCHANNEL = 0008
• After volume LXDAT1 has been successfully migrated to FRDAT1 ( real device 7802), a QUERY VIRTUALDASD command displays the following information:
q v 1201DASD 1201 3390 LXDAT1 R/W 60102 CYL ON DASD 7802 SUBCHANNEL = 000E
However, as the directory entry still assigns real device number 5EC0, the guest machine would try tore-use the old source volume (now labeled as FRDAT1) when it logs on to z/VM the next time.
To avoid the above situation, it is recommended to assign volumes to guest machines by using volumelabels instead of real device addresses. This can also be enforced by specifying the ENFORCE_BY_VOLIDparameter in the z/VM system configuration file.
Note: If it is not possible to specify the volume label in the guest machine’s directory entry, you must takemeasures to update the directory entry before the guest systems log on to z/VM the next time.
Duplicate volume labels
While it is possible to have multiple disks with identical volume labels online in a z/VM environment, avolume label must be unique on z/OS. Therefore, you must ensure that the correct volume to be migratedis online to the z/OS Master and Agents.
If multiple z/VM volumes with the same volume label need migrating, you must repeat the z/OS VARYOFFLINE|ONLINE process to migrate each volume individually.
MVS running under VMTDMF provides the following list of detailed information when you run TDMF with VM and MVSenvironments. This information includes data set considerations, parallel access volumes, devices thatwere previously formatted for VM, and VM volume tolerance.
88 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
COMMDS data set considerationsThe following considerations are provided for assistance on reading the mini-disk statement when theCOMMDS is allocated on a device that is defined to VM as a mini-disk.
MDISK vdev type 000 END volser MVS
Do not place a "V" as a suffix to the preceding statement, unless the volume is shared between more thanone MVS Guest virtual machines, and one or more native MVS systems. Refer to "Restrictions for Reserve/Release" in the applicable VM/ESA Planning and Administration Guide, and "Sharing DASD among MultipleVirtual Machines by Using Virtual Reserve/Release" in the applicable z/VM Planning and AdministrationGuide.
The real device statement regardless of disk definition should appear as follows:
RDEVICE rdev type DASD SHARED YES
This defines the device as shared DASD.
Other control data set considerationsVolumes containing control data sets that are shared between MVS and VM, such as control data sets thatare used for StorageTek’s LSM Host Software Component (HSC) or MIM control data sets, should not bemigrated using TDMF.
ML-XXXXXX has no way to detect the VM usage. Furthermore, if such volumes are moved (either with orwithout TDMF) with VM down, there is still an exposure, as VM links to these volumes by DEVICEADDRESS rather than Volume Serial Number.
Parallel access volumes capable device considerationsIn environments where PAV (Parallel Access Volumes) support is being utilized, TDMF requires that theOPTION MAINTCCW must be specified (or OPTION DEVMAINT, which includes all the functions ofDEVINFO and MAINTCCW) for TDMF to successfully run.
If you do not specify the OPTIONS MAINTCCW, you will receive the message HCPDDP1017I A channelcommand was rejected because you do not have MAINTCCW Authorization on the VMGuest MVS system console.
Using devices previously formatted for VMFor target volumes that have been previously used in a VM environment, it is recommended that anICKDSF batch job be run with the INIT, VALIDATE and NOCHECK parameters. This function will perform amedial initialization of the device. The home address and record zero for each track are read, validatedand rewritten. The data remaining on a track is erased.
About this task
Note: This function is not applicable for 3390-9 or larger device types.
For more information related to this subject, refer to the applicable Device Support Utilities User’s Guideand Reference.
VM volume toleranceVM volumes can now be copied and relabeled using TDMF in an MVS environment. However, TDMFcannot monitor updates to the source volume while the migration is in progress, nor can the VM (CP andCMS) control block be manipulated to achieve a non-disruptive Swap migration.
To migrate a VM formatted volume, it must be varied offline from all attached VM systems. In the case ofa system, spool, or page volume, the associated VM system(s) must be shut down and later IPL’d.
Important: For volume migrations, updates to the source volume made during the migration by usersother than the z/OS guest system(s), might not be copied to the target. These users will continue toaccess the obsolete source volume after the swap, and users that link to the source volume by its original
Chapter 5. Planning 89
device address, and not its volume serial number, will need to change their profiles. For Point-In-Timemigrations, updates to the source volume made during the migration by users other than the z/OS guestsystems, might not be copied to the offline target.
Note: The FASTCOPY option is not available for use when migrating a VM formatted volume, as VMvolumes are not required to have a VTOC; thus, the information required for the FASTCOPY option isunavailable.
Operating parameters of TDMFTDMF operates under a certain set of parameters. For example, the number of TDMF sessions, thenumber of groups per TDMF session, and the number of volumes per TDMF session.
Number of TDMF sessionsTDMF can support any number of sessions.
The number of sessions is dictated by the amount of available real storage that the customer hasavailable to use for these sessions. See “Storage requirements” on page 67.
The maximum number of Agent systems/jobs is 63 in a single session.
Number of groups per TDMF sessionA group must consist of at least two volume pairings and cannot exceed the number of volumes per TDMFsession. It is recommended that the SINGLE group option be used only if there is one active session. Ifthere are multiple active sessions, no two concurrently active sessions should contain identically namedvolume groups.
Number of volumes per TDMF sessionThe maximum number of volume pairings per TDMF session is limited by the system resources, inparticular real storage frames. Fixed storage frame requirements rise as the number of volume pairingsincreases.
Review the storage requirements that are listed in Table 16 on page 67.
Extended address volumes (EAV)TDMF is capable of migrating data sets on Extended Address Volumes, which are supported on z/OS 1.10.Extended Address Volumes have a maximum number of cylinders per volume increased from 65520cylinders (or 55 GB) to 1182000 cylinders (1 TB of addressable storage).
However, EAV volumes do not currently support the following data set types in the extended addressingareas:
• Page data sets;• VSAM data sets with IMBED or KEYRANGE attributes. These cannot be created but they might have
been inherited from prior physical migrations or copies.
Page and swap data setsTDMF will migrate a source volume containing an active PLPA or common page data set, or both.
A volume that contains an active page data set must be the only volume in the migration session.
It is recommended that the TDMF Master job is run on the system that "owns" the page volume, simplybecause a volume is quiesced for a shorter time on the Master than on an Agent system. However, if thereare multiple page data sets for multiple systems on a volume it can be migrated in a multi-system sessionand any of these systems can be designated as the Master.
It is recommended that page volumes be migrated during a low activity period of operation.
90 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
TDMF will not migrate a volume containing an active Local page data set or swap data set, while volumescontaining only inactive page or swap data sets may be moved in a multi-volume session.
Note: This function is not available in a JES3 system managed device environment. When cloning asystem or during a data center relocation/consolidation, create these data sets outside the session.
Parallel sysplex considerationsTDMF will migrate Sysplex Couple Data Sets (CDS).
Dependent upon how often XCF expects each system within a Sysplex to update the CDS (default: 15seconds), it is possible that, during the Quiesce and Synchronization phases of a volume migrationcontaining a CDS, a system update to the CDS will not be met. This condition could result in the SystemFailure Manager (SFM) partitioning the "failing" system out of the Sysplex, causing a wait state 0A2 on thepartitioned system.
It is recommended that you move the alternate CDS first, then switch the active CDS to the alternate, andthen move the primary CDS.
After moving a CDS, you may get an incorrect message display although the data sets perform properly.The device number of the CDS displayed in message IXC357I in response to the D(isplay)XCF,COUPLE MVS operator command may need to be updated.
You must reinitialize the message by issuing the MVS command SETXCF COUPLE,ACOUPLE= to point toanother couple data set and again to bring back the primary. For more information regarding thiscommand, refer to the MVS System Commands manual.
In a SYSPLEX environment, TDMF will establish an XCF group called "TDMF", with each Master and Agentjob as a member. This group's message traffic is very sparse and the message length is typically only four(4) bytes, so it can safely be left in the DEFAULT transport class.
Migrations outside a single sysplexWhen the participating systems in a TDMF migration session are not all in the same parallel sysplex, youcan exploit TCP/IP communications between the Master and Agent systems - alongside or instead of thestandard XCF signalling mechanism. To activate this, a port number must be reserved for use by TDMF onall the participating systems. This port number is coded as an option on the SESSION control statementand propagated to the Agent systems.
TCP/IP messaging between the Master and Agent systems will not be performed if the session has aREMOTE control statement or is a remote session type.
Peer-to-peer remote copy supportIBM Peer-to-peer remote copy (PPRC) ensures that a user's disaster recovery environment is notinadvertently destroyed or changed. TDMF is designed to be consistent with this type of data protection.
Standard IBM Peer-to-peer remote copy, a synchronous form of disaster recovery, is also supported bothfunctionally and via its command structure as that provided by both EMC and HDS.
At session initialization, TDMF ensures that the characteristics of the source and target volumes involvedin a swap migration appear to be the same on all systems in the TDMF session. Throughout the migration,TDMF ensures that these characteristics do not change. If TDMF detects inconsistencies in the PPRCsession, then the TDMF session is terminated.
PPRC session types allowedTDMF supports sessions where a source or target volume might be involved in a PPRC session. Thefollowing types of active PPRC sessions are allowed at initialization of a TDMF session.
• Point-In-Time migrations• Swap migrations where the source and target volumes are involved in an active PPRC session.
Chapter 5. Planning 91
• Swap migrations where the source volume is not involved in an active PPRC session and the targetvolume is involved in an active PPRC session.
When a TDMF session starts, the following conditions apply.
• The status of the target volume cannot be changed.• A PPRC session might be stopped on the source volume if the target volume is involved in an active
PPRC session.• A PPRC session might be started on the source volume if the target volume is involved in an active PPRC
session.
For an understanding of when migration can be used with various mirroring techniques, refer to “SRDFsession types allowed” on page 77.
PPRC/GDPS environments with HyperSwap enabledMigration of HyperSwap controlled volumes is described in detail in the following topic.
See Chapter 9, “Migrating GDPS/PPRC HyperSwap volumes,” on page 191.
Migrations between a PPRC device and a SRDF deviceTDMF will allow the migration of volumes in a PPRC pair to volumes in a SRDF pair (EMC SymmetrixRemote Data Facility). A migration from an SRDF volume to a PPRC primary device is also permitted.
TDMF requires that the ALLOWMIRRORCHANGE session option or volume option be specified.Acknowledgement of the device mirroring architecture mismatch will be requested before the volume willbe selected for migration if the NOACKNOWLEDGE parameter was not also specified. For more details,see “Migrating from SRDF to PPRC/HDS TrueCopy volumes” on page 76.
Migrations between a PPRC device and a TrueCopy deviceTDMF allows the migration of volumes in a PPRC pair to volumes in an HDS TrueCopy pair.
A migration from an HDS TrueCopy volume to a PPRC primary device is also permitted. TDMF requiresthat the ALLOWMIRRORCHANGE session option or volume option be specified. Acknowledgment of thedevice mirroring architecture mismatch is requested before the volume is selected for migration if theNOACKNOWLEDGE parameter was not also specified.
For more details, see “Migrating from SRDF to PPRC/HDS TrueCopy volumes” on page 76.
Real-time monitorsVarious real-time monitors can sometimes cause migration-related error messages for which seeking avendor-supplied fix is advisable.
After a migration, TDMF will notify the operating system that a swap has occurred through the EventNotification Facility (ENF).
Real time monitors such as Omegamon, Omegamon II, MainView, The Monitor for MVS™ (TMON/MVS),CMF, or Allocation Control Center (ACC) might not "listen" for these events. As a result, various errormessages may appear after a TDMF migration.
Contact the appropriate vendor to determine if there is a fix for this issue.
92 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Shared device definitionEnsure that the device is defined for sharing before adding a Communications Data Set (COMMDS) formulti-system sessions; otherwise, TDMF initialization fails as indicated by message GTD1386I.
TDMF periodically places a hardware reserve on the volume that contains the Communications Data Set(COMMDS). If the COMMDS for a multi-system session is not placed on a device that is defined as shared,TDMF sends message GTD1386I and initialization fails.
For more information, see Chapter 5, “Planning,” on page 67 and “Coexistance with EMC’s z/OS migrator”on page 78
Enterprise storage server supportTDMF supports the IBM Enterprise Storage Server® (ESS) Storage Controller with the Parallel AccessVolume (PAV) feature.
Support for the PAV feature is in exploitation mode. This means that during the life of a TDMF migrationthe PAV feature is dynamically disabled and dynamically re-enabled when the migration completes forthat volume only.
Before migrating data to, or establishing data on, an IBM Storage Controller (ESS) device, customersshould ensure that they understand these I/O functions.
The ESS subsystem can be defined with "Transparent", "Toleration", or "Exploitation" mode.
Transparent mode
• IODEVICE macro instruction with UNIT=3390
If the preceding is true, then the PAV feature is not enabled. TDMF functions with no issues.
Toleration mode
• CNTLUNIT macro instruction with UNIT=2105• IODEVICE macro instruction with UNIT=3390B
If the preceding is true, TDMF migrates this volume as if it were in exploitation mode.• IODEVICE macro instruction with UNIT=3390A
If the preceding is true, then the device is defined as an alias PAV volume. TDMF will not migratethese volumes.
Exploitation mode
• CNTLUNIT macro instruction with UNIT=2105• IODEVICE macro instruction with UNIT=3390B
If the preceding is true and the proper levels of MVS maintenance is installed, TDMF will migratethis volume as if it were in exploitation mode, and the device is eligible as a base PAV volume. TDMFsupports both standard and dynamic PAV(s).
MessagingTDMF has more messages in support of the Enterprise Storage Server control unit. These messages arecreated during the initialization phase of the volume migration and the volume migration continues on tosubsequent migration phases.
Chapter 5. Planning 93
Migration from ESS to Non-ESSWhen you migrate ESS volumes to non-ESS from TDMF, certain conditions can trigger a warning messagethat are further explained by the following details.
TDMF sends message GTD1670W during initialization, indicating possible errors might occur toapplication I/O operations. Following is a list of conditions that produce this message.
• Swap migration• Source volume resident on the ESS in Exploitation mode• Target volume resident on a subsystem that is non-ESS (3990) control unit.
Message GTD1670W is sent to warn of a migration from a device that supports non-3990 features to adevice that does not. The message is only a warning and the migration is allowed to commence. Duringthe volume initialization phase, TDMF modifies the information in the source volume’s DCE (DASD ClassExtension) to temporarily mask the advanced features of its storage controller. MVSs IOS (input/outputSupervisor) dynamically builds channel programs to detect and use these features by a customer orsystem application channel program. Channel programs constructed subsequent to the DCE modificationuses only those features that are also supported on the target volume. Upon termination of the volumemigration, the information that is dynamically modified in the DCE is restored.
This process is not always successful based on timing conditions although it is felt that the alternative ofbeing able to run only the swap migration of a volume that is not available from a Storage Controllercontrol unit to a 3990 control unit was unacceptable. A channel program that is built before themodification of the DCE or by an I/O driver by using a private copy of the DCE, might be rejected by theTDMF I/O Monitor, which dynamically ends the volume migration.
Migration from non-ESS to ESSWhen you migrate non-ESS volumes to ESS from TDMF, certain conditions can trigger a warning messagethat are further explained by the following details.
TDMF sends message GTD2668W stating that PAV exploitation will not be available for this volume untilthe next IPL. The conditions that produce that message are listed here.
• Swap migration• Source volume resident on a non-ESS subsystem• Source volume resident on an ESS subsystem that is defined as Static PAV and PAV is not currently
active• Source volume resident on an ESS subsystem that is defined as Dynamic PAV and PAV is not currently
active• Target volume resident on an ESS that is in Exploitation mode• Source Unit Control Block (UCB) is not defined as a 3390 B.
TDMF does not send message GTD2668W if the source UCB is defined as 3390 B, and PAV access isimmediately available upon completion of the migration.
TDMF does send the message GTD2667W stating that the target device does not support PAV functionsand PAV access is not available for this volume if and of the statements in this list are true.
• Swap migration• Source volume resident on an ESS subsystem• Target volume resident on a non-ESS subsystem• Swap migration completes normally.
The TDMF I/O monitor ends a migration if unsupported CCWs are detected on the source volume. Inaddition, the I/O monitor can end a Swap migration if I/O operations are detected on the source volumethat are not supported on the target device.
The following table outlines the expected outcome when you move volumes with different attributesrelated to PAV.
94 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Table 19. Expected PAV outcomes based on volume attributes
From Device To Device Result
non-PAV (defined as 3390) PAV (defined as 3390B) PAV unavailable until next IPL
non-PAV (defined as 3390B) PAV (defined as 3390B) PAV available after swapcompletes
PAV (defined as 3390B) PAV (defined as 3390B) PAV available after swapcompletes
PAV static PAV dynamic PAV dynamic
PAV dynamic PAV static PAV static
PAV (defined as 3390B) non-PAV (defined as 3390B) PAV disabled
When migrating volumes from a non-PAV capable subsystem to a PAV capable subsystem, PAV will not beavailable if the source devices are not defined in advance as 3390B type devices; an IPL is required. Thisrequirement is because a PAV UCB is larger than a non-PAV UCB. When you add a PAV capable subsystemthe non-PAV capable IODEVICE macros must be changed to 3390 B.
FlashCopy supportTDMF does not currently support migration of data, which can be modified by the FlashCopy feature. Toprevent this from occurring during a volume migration, the FlashCopy feature is temporarily disabled forthe source and target volumes.
TDMF now recognizes Cascaded FlashCopy. During initialization, TDMF processes the FlashCopyrelationship table. If either the TDMF SRC or TGT is the target of a FlashCopy relationship, the migration isterminated.
Note: During a volume migration that involves FlashCopy (after TDMF has disabled FlashCopy), anycustomer jobs using FlashCopy will proceed as though the feature was not supported by the sourcedevice. More specifically, if using DF/DSS data set copy, the following actions occur based on the optionspecified:
• FASTREPLICATION(PREFERRED), replication proceeds• FASTREPLICATION(REQUIRED), replication will fail
Message ANTF0355E FLASHCOPY TARGET IN USE BY OTHER COPY OPERATION displays while themigration is in progress and attempting to establish a TDMF migration device as a FlashCopy targetdevice. An attempt to enable FlashCopy using the console command DEVSERV DIAG,uuuu,ALLOWFC orany other program that issues the DCTL CCW (F3) will end the volume's migration. The followingmessages then display:
• GTD2330E The GTDVOL program is terminating this migration at the request of GTDIMON• GTD2362I This GTDIMON error was due to the detection of an invalid CCW opcode (F3)
If a TDMF source volume is Soft Fenced, see “FENcesource/NOFENcesource” on page 51 for information.The FlashCopy status is disabled when the migration process completes. See “Recovering TDMF devices”on page 32 for information.
Starting with APAR OA53265 (5.6 and 5.7 and included in 5.8 base), TDMF will attempt to restoreFlashCopy before a Soft Fence process. If message GTD2795W displays, which indicates a failure, followthe above procedure to recover one or more devices.
Chapter 5. Planning 95
Define an I/O device as static, installation static, or dynamicWith MVS/ESA SP V4, the ability to define I/O devices as static, installation static, or dynamic wasintroduced through HCD.
A Static device is a device that cannot be dynamically added, deleted, or modified in the softwareconfiguration definition. Therefore, the device is not available for use until the next IPL of MVS.
An Installation Static device is a device that can be dynamically added to the software definition butcannot be modified or deleted dynamically. Or, the device can be dynamically added, deleted, andmodified in the hardware configuration definition.
A Dynamic device is a device whose device definition can be dynamically added, deleted, and modified inthe software and hardware definition.
You cannot swap a device with different device attributes. This means that a device with an attribute ofDynamic cannot be swapped to a device with an attribute of Installation Static, or vice versa. Attempts todo so result in message "IGF513I - DEVICE devnum INVALID FOR SWAP - DYNAMIC/STATICINCOMPATIBLE" being issued.
When a device attribute is changed from Installation Static to Dynamic, its UCB address changes.Programs within the system that save UCB addresses encounter problems if an attempt to use that UCBthat was valid when the device was Installation Static.
The following table describes the type of swap migrations that are allowed by the MVS operating system:
Table 20. Dynamic, static, and installation static matrix
TARGET VOLUME
Source Volume Dynamic Static Installation Static
Dynamic YES NO NO
Static NO YES YES
Installation Static NO YES YES
The type of device configuration is determined by the Unit Information Modules (UIM) of the operatingsystem with the Dynamic=No or Dynamic=Yes parameter in HCD. DASD devices are allowed to beDynamic or Installation Static based on the Dynamic parameter. Therefore, TDMF is not concerned withStatic devices.
Take the following steps to determine the device configuration of a volume.
1. Open an IPCS session, and set the default to ACTIVE.2. Go to option 6 and issue LISTUCB addr for the specific device.3. Locate the NXUCB data field in the display of the control block (UCBOB + 8).
If the NXUCB data field is zero, the device has a configuration of Dynamic. If the NXUCB data fieldcontains an address, the device has a configuration of Static or Installation Static. Scroll to the bottom ofthe display to see if the device is Static, Installation Static or Dynamic.
If a TDMF Swap migration is attempted with different device configurations, TDMF issues the message"GTD1230S - Both volumes of a Swap migration must be either Dynamic orInstallation Static".
For more information, see the applicable OS/390® HCD Planning, or the applicable MVS/ESA HCDPlanning topics "Defining Whether an I/O Device Will be Dynamic" and "Redefining the DynamicParameter for an I/O Device".
96 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Storage space requirements for virtual arraysCertain storage space requirements should be considered for copying volumes to SVA (shared virtualarray) and RVA (RAMAC virtual array) type arrays.
At the start of the copy phase, the storage subsystem is instructed to release all data associated with thetarget volume, with the exception of track zero. It should be noted that TDMF always constructs apseudo-VTOC on track zero of the target volume.
When a track is read from the source device containing no user records, no I/O is performed to theequivalent track on the virtual target volume. During the refresh and synchronization phases, deletion of adata set and its allocated space on a virtual source volume will cause TDMF to instruct the storagesubsystem to also release the space on the virtual target volume.
Users who maintain "permanent" SnapShot data sets on their virtual DASD will need to take these datasets into account when calculating the space required for a copy to another virtual device array.
TDMF ownership of target volumeTDMF "owns" the target volume during the life of the migration session to preserve physical data integrity.Except for a few non-disruptive inquiries, any I/O operation to the target volume by anything other thanTDMF is rejected. Also, a message is created with IOSB completion code X’4A’ (the I/O was prevented).The message might be IOS000I or some other MVS message that indicates an error by the application orMVS system component
TDMF and volume table of contentsSelecting the dynamic ICKDSF option causes TDMF to reformat the Volume Table of Contents (VTOC) bydynamically starting the ICKDSF REFVTOC or EXTVTOC function when the target volume devicecharacteristics do not match that of the VTOC.
For more information, see “ICKDSF” on page 80.
When the copy phase starts for a migration volume, TDMF builds a non-standard VTOC on track zero ofthe target device. If the migration is terminated by the user, or because of an error, the target volumeremains online, but with this unusable VTOC. The purpose of this pseudo-VTOC is to prevent allocationsto the volume that might allow unauthorized access to data successfully copied from the source volume.Attempts to allocate on this volume results in message IEC603I, which warns of problems in the VTOC onthe volume. This is not an error.
Soft FenceDS8870 disk subsystems can provide a Soft Fence function. The purpose of this function is to preventunintended access to a device in a logical subsystem (LSS). Most I/O operations (like reading and writingdata) to a device are prevented by the disk subsystem if the Soft Fence state is set for the device (thedevice is soft fenced).
TDMF supports this function by resetting a possible Soft Fence state for target devices at the beginning ofa migration session and by setting the Soft Fence state for migration source devices as soon as they havebeen swapped with their associated target devices. Supporting source devices are put into the Soft Fencestate by default in order to prevent any unintended and undetected access to original source volumesafter they have been migrated as this may cause data integrity issues.
Important: If the device was previously FlashCopy enabled and has been put into the Soft Fence state,FlashCopy will be disabled. If the device is to be re-used in a GDPS PPRC environment, then FlashCopywill need to be re-enabled before use. Once Soft Fence has been disabled with ICKDSF, FlashCopy can bere-enabled through the following actions:
Chapter 5. Planning 97
1. GTDCLUP PARM=FLASHCOPY2. Console command DS DIAG,uuuu,nn,ALLOWFC
DEVSERV command to display Soft Fence statusAn example of the DEVSERV command.
IEE459I 07.08.18 DEVSERV QDASD 285 UNIT VOLSER SCUTYPE DEVTYPE CYL SSID SCU-SERIAL DEV-SERIAL EFC07802 JZ7802 2107961 2107900 30051 0078 0175-W2201 0175-W2201 SOF**** 1 DEVICE(S) MET THE SELECTION CRITERIA**** 1 DEVICE(S) FAILED EXTENDED FUNCTION CHECKING
The value SOF indicates that the device is in a soft fenced state.
DS PATH,devn
IEE459I 14.22.46 DEVSERV PATHS 447 UNIT DTYPE M CNT VOLSER CHPID=PATH STATUS RTYPE SSID CFW TC DFW PIN DC-STATE CCA DDC CYL CU-TYPE07802,33909 ,F,000, ,63=+ 71=+ 2107 0078 Y YY. YY. N SIMPLEX 02 02 30051 2107 ** FENCED DEVICE D8FFFF00 D0FFFF00 D0FFFF00 D0FFFF00 00000000 00000000************************ SYMBOL DEFINITIONS ************************F = OFFLINE + = PATH AVAILABLE
The DEVSERV PATH command displays "FENCED DEVICE" if the device is soft fenced.
SPID FenceDS8700, DS8800, and DS8870 storage systems provide a SPID Fence function to prevent an offlinedevice from being brought online.
TDMF utilizes this function to prevent Agentless systems from accessing a migration device. Devices areprotected from access by unidentified systems by setting the SPID Fence state for source and targetdevices at the beginning of its migration and removed when the MIGRATE or REPLICATE is completed.Devices are put into the SPID Fence state by default in order to prevent any unintended and undetectedaccess to the source and target volumes as this may cause data integrity issues.
If the TDMF termination routine is not invoked due to a system error (Force, Wait state), then the SPIDFence function might be left enabled. Since TDMF also disables FlashCopy at the start of a migration, thenthe SPID Fence function will also stay disabled. Any cleanup for SPID Fence should also addressFlashCopy, see Soft Fence above or refer to “Recovering TDMF devices” on page 32.
Suppressing warning messages in TDMFThe NOWARNing option allows users to suppress the effect of individual warning message on thecompletion code of the TDMF session. Using this option keyword causes the system to disregard thewarning message(s) about the migration, and resets the completion code to zero.
Syntax exampleThe following provides an example of the syntax used for the NOWARNing option keyword with 1574 asthe option parameter.
This example prevents message 1574 from contributing to the Master job completion code. If GTD1574Wis the only warning or error message in the migration session, specifying OPT(NOWARN(1574)) wouldcause the Master to set completion code to 0 instead of 4 at termination.
|--+-----------------+--|
+-NOWARNing(1574)-+
98 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
XRC supportTDMF allows a volume that is the primary in an XRC session to be specified as a source or target volumefor a Swap or Point-In-time migration. Restrictions on the use of an XRC active source volume in a Swapmigration are listed below.
• TDMF allows a swap migration source volume to be active in an XRC session to a non-mirrored targetdevice, if the XRCVolume option was specified (see “Volume options” on page 47.) With Release 5.2.0,TDMF allows the XRC migration session's Data Mover address space to be active on the same LPAR asthe TDMF Master job or on an Agent LPAR. In this situation, the target device must be XRC capable.Also, TDMF will use the XDELPAIR and XADDPAIR commands to reestablish the XRC session for thevolume (on the target device) after the swap. You need to make sure that the target device iscompatible with the existing XRC session's Utility devices. For more information, see DFSMS AdvancedCopy Services, SC35-0428 manual.
• For users who run the System Data Mover on the remote system, TDMF allows only the migration of XRCvolumes to mirrored devices (for example, PPRC, SRDF, and XRC). The ALLOWmirrorchange optionmust be supplied for the volume if the target device is not being mirrored by XRC. For more information,see “Common options for the SESSION, GROUP, MIGRATE, and REPLICATE statements” on page 50.
When you swap to a target volume that is active in an XRC session, validate that the System Data Moverdetected the change of volume serial number. For example, by using the TSO command XQUERY<target_volume_XRC_session> VOLUME(<source_volume). The XRC device pair remainsmirrored, even if the SDM did not detect the change of volume serial. Obviously, the disaster recoveryprocedures for the source volume serial must be amended to address the XRC secondary device, as inevery migration to a mirrored target device.
TDMF writes to an XRC target volume as fast as it can, which means that care must be taken when youmigrate multiple volumes into the same XRC session. Judicious use of the concurrent volume limit mightbe required to prevent the XRC link from being overloaded by the TDMF volume copy phase.
Extent-efficient and space-efficient volumesWhen a volume is defined to a DS8800 series controller as extent-efficient or space-efficient, thecontroller manages free space.
The NOFASTCopy/FASTCopy Statement
If the volume is participating in a migration as a TDMF target volume, the NOFASTCopy statement is notavailable. The NOFASTCopy statement directs TDMF to copy free space indicated in the source volume'sVTOC to the target volume. Because this free space is managed by the controller, TDMF does not copy it.If NOFASTCopy is specified under this condition, it is suppressed and converted to FASTCopy.
Message GTD1579I is issued, the migration return code is not changed, and the migration proceedsnormally. Even though the free space indicated in the VTOC was not physically copied, it is still logicallyavailable by virtue of the Extent-Efficient or Space-Efficient status maintained by the controller.
z/VM Volumes
When a z/VM volume is a TDMF target volume, and the volume is defined as Extent-Efficient or Space-Efficient, TDMF performs a full-track copy and the free space on the source volume is copied to the target.This removes the copied free space from controller management. The volume and controller remain fullyfunctional, and this free space is usable by the specific z/VM target volume only.
Message GTD1595W is issued and the migration return code is 4.
XRC remote system data moversTDMF versions 5.4.0 and later support system data movers at remote sites. In prior releases, thepresence of TDMF Agent systems at a remote site (at the other end of extended channels) caused excesscontention for the SYSCOM data set. Remote systems, in conjunction with multiple local systems, couldnot cycle the SYSCOM fast enough. This enhancement solved this problem by eliminating the requirementfor a reserve to be issued against the SYSCOM volume.
Chapter 5. Planning 99
An Agent is considered to be a Remote Agent when the Agent accesses the SYSCOM through a channelextender. Therefore, an Agent is considered to be a Remote XRC System Data Mover Agent when thesource volume is active in an XRC session on that Agent and the Agent accesses the SYSCOM through achannel extender. The number of Remote Agents cannot exceed 14.
XRC in a GDPS/XRC (z/OS global mirror) environmentYour interaction is required because TDMF does not run as a TDMF agent system during the migrationsession upon the GDPS/XRC controlling system.
Important: TDMF 5.2.0 does NOT support XRC in a GDPS/XRC or z/OS Global Mirror environment withoutyour interaction.
Your interaction is required because TDMF does not run as a TDMF agent system during the migrationsession upon the GDPS/XRC controlling system. The GDPS/XRC controlling system, also known as the "KSystem" in GDPS terms, is a system residing at the remote location.
A new volume refresh configuration option, VOLREFR, has been added to allow faster processing of theconfiguration, in order to refresh VOLSER information that GDPS maintains for GDPS-managed XRCdevices. This option should be used after swap migrations are performed by TDMF, or after you haveclipped (changed the volume serial number) of GDPS-managed devices.
As TDMF did NOT run upon the GDPS/XRC controlling system (K system), volumes in the TDMF sessionshould be varied offline and then online prior to run of the VOLREFR function. As with all GDPS operations,these vary commands should be carried out from the Controlling system.
100 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Chapter 6. Using advanced functionsTDMF Advanced Functions are a set of specialized features that use the basic functions of TDMF.
OverviewYou can purchase and license each TDMF advanced function.
The following considerations are provided:
• The TDMF Offline Volume Access feature acts as an interface between one or more offline targetvolumes in a Point-In-Time or Perpetual Point-In-Time session and the user-specified program.
• Perpetual Point-In-Time expands upon the Point-In-Time copy feature, which allows multiplesynchronization points to be taken within the same session. This increased ability provides for multiplebackups to be taken at various points within the batch cycle, for example.
• See “During an active PPIT session” on page 109 to learn how to start a DASD migration that uses theexisting TCP/IP connections within a business environment. Because TCP/IP requires no specializedhardware or dedicated telecommunication circuits, TCP/IP support enables eliminating expenses thatare associated with migrating data that uses specialized channel extension hardware and atelecommunications data link.
TDMF offline volume accessThe TDMF Offline Volume Access feature (OVA) can help you process a Point-In-Time (PIT) volume copy.
With OVA, data can be read from one or more offline target volumes, while normal processing continuesto update one or more source volumes. Using OVA, for example, users can copy or back up to tape fullDASD volumes, the databases on them, or both. This procedure creates a PIT-consistent archived copy,without halting updates to those databases for the length of time that is required for backup jobs.
OVA is enabled by specifying the OVA option on the SESSION control statement or on the GROUP orREPLICATE control statements for the volumes within the TDMF job that are to participate in an OVAsession.
The TDMF Interface program (GTDIPGM) is used to register each OVA job to the TDMF session orsessions, creating the PIT copies. After registration is complete, a user-specified program is started byusing GTDIPGM. While this program is active, I/O requests from the OVA job that is directed towards anOVA source volume are redirected to the PIT target volume.
Control statements for GTDIPGM specify volume serial numbers, which are included in, or excluded from,I/O redirection. It is also permissible to specify data set names rather than volume serial numbers for theinclusion or exclusion functions. However, the entire volume upon which such a data set is resident isalways included or excluded from the I/O redirection.
GTDIPGM also intercepts reserve and DEQ requests from the program that is run, and modifies theserequests to act on the PIT target device.
Explanations of the messages can be accessed by using the message help facility within the TDMF TSOMonitor (see “Option H.1 – Display TDMF message details” on page 188) or TDMF Messages and Codes(LC27-8541).
OVA purposeOVA has been designed to allow read access to the Point-In-Time target copy. However, the mechanismdoes permit updates to be made to the target volume.
Note: As soon as the target volume has been changed, it is no longer a valid Point-In-Time volume copy,therefore, extra care must be taken when scheduling update jobs.
© Copyright IBM Corp. 2010, 2019 101
One data set, which may be modified on the target volume using OVA, is the Volume Table of Contents(VTOC). This allows the user to rename data sets, for example, on the target volume so that differentsecurity profiles may be acquired. In practice, however, because data set allocation for OVA is usuallyperformed before I/O redirection is enabled, most application programs will use data sets that arecataloged as resident on the source volume.
Data sets that are created on the source volume after the Point-In-Time copy has completed are notavailable to a program using OVA. An attempt to open such a data set will result in Abend code S213-04.For this reason, it is recommended that the source volumes be made ineligible for new data setallocations before the migration is synchronized.
Starting an OVA sessionA GTDIPGM job would normally be submitted when replication of its OVA Point-In-Time volumes iscomplete. In fact, it is possible to submit the OVA job up to 15 minutes before you submit the associatedTDMF sessions. After an OVA volume is synchronized it becomes available for registration by theGTDIPGM program and remains available until 15 minutes have elapsed without an OVA sessionregistered to any volume in its migration group.
About this task
For situations where job-scheduling considerations might lead to a delay in starting the first OVA job, ormight result in an interval of more than 15 minutes in the OVA job stream with GTDIPGM not active, the"OVA Registration Interval" installation option needs to be specified. Specifying this option ensures theavailability of OVA volumes for registration, from the time the Point-In-Time copy was made, for thenumber of minutes specified. After the registration interval elapses, the TDMF sessions begin normaltermination for a migration volume, or prompt the user for a Perpetual Point-In-Time (PPIT) recycle, if thevolume or group is idle for at least 15 minutes.
The GTDIPGM program is used to register volumes on either the TDMF Master or Agent sessions. Amaximum of 85 GTDIPGM jobs in each TDMF system can be registered to a volume concurrently.
Registration of OVA jobs can be stopped dynamically for a volume or group by using the TDMF TSOMonitor (Option 2 – User Interaction and Status) by using the H command. If the PPIT option is notselected, the OVA volume is terminated as soon as there are no longer OVA jobs that are registered to it. Ifthe PPIT option is selected, the TDMF session issues the "Recycle PPIT" prompt for the volume or group.
To access a Point-In-Time target volume by using OVA, the volume migration must be processed with theOVA option enabled. You can process it this way by selecting the option on the SESSION, GROUP, orREPLICATE control statement.
Using OVA on different subsystemsCare must be used when defining the source and target volumes for use with OVA. Different subsystemssupport different functions. For example, a Point-In-Time copy is started where the source is on an ESSsubsystem (defined with extended functionality) and the target is resident on a non-ESS subsystem.When the OVA job is started, it will fail due to a mismatch of support functions.
Multi-volume data setsIt is recommended that all volumes that are to participate in an OVA session should be in the same group.If it is not possible to do so, then the multiple TDMF sessions should be synchronized together. Customerapplications should not be restarted until all the sessions are in "OVA Waiting" status.
Incremental volume backupsPrograms that attempt to back up only data sets on a volume that have changed since the previous run ofthe backup program test and reset the DS1DSCHA flag for each data set in the Volume Table of Contents.These are called incremental backups and are widely used because they prevent the backup media fromfilling up with identical copies of the same data. However, if the backup program is running as an OVA job,this flag will actually be reset in the VTOC on the target volume and when the backup program is run againon a subsequent Point-In-Time copy the flag will have been copied from the source volume and anotherbackup will be taken, even if the data set had not been changed in the intervening period. In fact,although DS1DSCHA is popularly known as the "data set changed flag", the bit is only set when a data set
102 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
is opened in a way that allows it to be updated; therefore, an incremental backup strategy is notrecommended for data sets that remain open throughout multiple backup intervals.
To permit the changed flag to be used to minimize the amount of data stored by incremental backups runas OVA jobs, the RESETCHAngedflag option (see TDMF Installation and Customization) can be specifiedfor the session, an individual OVA volume or a group of volumes. When the option is in effect, the TDMFsession will reset the DS1DSCHA flag in ALL the data sets on the source volume after the Point-In-Timecopy has been made. This means data sets with DS1DSCHA set will be backed up from the target volumebut those data sets that are not opened for output in the intervening period will not be backed by OVAbackup jobs running on subsequent Point-In-Time target volumes. The disadvantage of this option is thatthe flag is reset for all data sets including those that will not be backed up using OVA. If the option is to bespecified, steps must be taken to ensure that up-to-date backups will be made of these data sets.
Functions of the TDMF interface programThe TDMF Interface program (GTDIPGM) performs the registration and de-registration functions and thenattaches the user-specified program for which I/O is to be redirected to the target volume. The JCL andcontrol statements necessary for that program must be included to control this processing.
Before registration is attempted, GTDIPGM builds two lists of volume serial numbers: inclusion andexclusion lists. The purpose of these lists is as follows.
Exclusion listThe exclusion list is made up of those volumes that might be allocated during the job step but are not tobe considered for I/O redirection.
A volume serial number is placed on the exclusion list for one of these reasons.
1. An EXCLUDE control statement specifies the volume serial number.2. An EXCLUDE control statement specifies a data set name that is cataloged as resident on the volume.3. The JCL includes a new data set allocated on the volume.4. The JCL includes a new data set with disposition MOD, allocated on the volume.
Inclusion listThe inclusion list contains the volume serial numbers for which the program locates the appropriateTDMF session and registers those volumes for I/O redirection. Registration does not occur unless TDMFsessions are found for all volumes on the list. Therefore, the list must contain only the volumes for whichI/O redirection is required.
A volume is placed on the inclusion list for one of the following reasons.
1. An INCLUDE control statement specifies the volume serial number.2. An INCLUDE control statement specifies a data set name that is cataloged on the volume.3. The JCL includes a data set, with disposition of OLD or SHR, allocated on the volume.4. No EXCLUDE control statement exists for the volume or a data set cataloged on it.
If a volume appears on both lists, an error message is printed and the run is ended.
TDMF interface program processingTDMF provides a list of functions for the Interface Program (GTDIPGM) processes. This list includesinitialization, registration, I/O redirection, and de-registration.
InitializationGTDIPGM issues an ENQ for every OLD or SHR data set specified in the JCL, which does not resideentirely on excluded volumes. The resource major names are "AASF_DSN" and "AASF_VDS"; the minorname is the data set name. Shared control is requested if the data set disposition is SHR, otherwiseexclusive control is requested. To ensure OVA data set serialization between jobs on multiple systems,the appropriate system inclusion entries should be added to the GRS or MIM resource name list. Theseentries would closely resemble those for SYSDSN, but the major name must be specified in single quotes
Chapter 6. Using advanced functions 103
due to the underscore character. If there is another OVA job using the same data set(s) and there is ENQcontention, a "waiting for (OVA) data sets" message will be issued.
Note: This contention can only occur if the other job dynamically allocated the data sets; otherwise thenormal allocation "waiting for data set" message will be issued by the initiator.
RegistrationIf a volume is found only in a "terminated" state within a TDMF session and has not been resubmitted in adifferent TDMF session, the 15-minute registration countdown begins.
The program locates all active TDMF sessions by checking every ten seconds. Active or completedmigrations slated for use by OVA for all volumes on the inclusion list must be found within 15 minutes orthe job will terminate. When all selected volumes have been synchronized, a registration request isenqueued for each TDMF session.
Registration reportThe program prints a Registration report in a table listing for each volume after all volumes are registered.
The Registration report contains the following information.
• The volume to which the I/O operations are redirected.• The job name of the responsible TDMF session.• The time of day at which the two volumes were synchronized.
If registration was unsuccessful, explanatory messages are created.
I/O redirectionThe specified program is attached by GTDIPGM. The TDMF sessions are then responsible for redirectingall included source volume I/O operations that are called by the OVA program to operate on the targetvolume PIT copy.
The program intercepts RESERVE, DEQ, and DYNALLOC requests from the selected program andprocesses the following functions:
• RESERVE and DEQ requests to a volume for which I/O redirection is being processed are sent to the PITtarget volume instead.
• A warning message is sent if the selected program successfully allocates a volume, which is not on theinclusion or exclusion lists.
GTDIPGM redirects Reserve macros for the source volume to the target device. ENQ requests for certainsystem resources, where the minor name is the source volume serial number, is dynamically modified toserialize the target volume. The resource major name of ENQ requests for SYSDSN and SYSVSAM, wherethe data sets being serialized are on INCLUDEd volumes, is changed to "AASF_DSN" and "AASF_VDS".
Allocation requests for the target volume, or for data sets on an OVA source volume, which was notINCLUDEd are failed with error code 0220.
Because OVA can be used to make a full volume backup of the PIT target volume, a restore from this PITtape backup is easier if it contains the source volume serial number. As a result, the I/O redirectionmechanism detects the reading of the volume label from the PIT target volume. Then, it substitutes thesource volume serial number into the data buffer in real storage.
De-registrationAfter termination of the selected program within OVA, the TDMF session(s) are notified to de-register OVAfrom each volume. As part of this process, the session(s) returns I/O redirection statistics to GTDIPGM.
I/O redirection reportThe program prints a I/O redirection report that contains a table listing for each volume.
• The volume to which I/O operations were redirected.• The total number of redirected I/O requests.• The number of redirected I/O requests that modified the target volume.
104 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
The number of I/O requests is a count of the I/O operations that were redirected to the PIT target volume.This value is a reflection of the amount of activity that is directed to the PIT target volume, not the actualnumber of modified records, blocks, or tracks. If the number of updates is zero, the PIT copy is intact. Thecount value is an indication that the target volume changed.
If any serial numbers were substituted for the PIT target volume, a message is sent noting this change.
Using the TDMF interface programThe TDMF Interface program (GTDIPGM) is a batch job that identifies included and excluded volumes toGTDIPGM, and provides program JCL.
About this task
Table 21 on page 105 displays the JCL statements necessary for GTDIPGM.
Table 21. GTDIPGM DD statements
DDNAME DESCRIPTION OF STATEMENT
IPGMOUT Messages and I/O redirection statistics output stream.
IPGMIN Control statement input stream.
Table 22 on page 105 explains the control statements necessary for GTDIPGM.
Table 22. GTDIPGM control statements
CONTROLSTATEMENT
DESCRIPTION OF STATEMENT
PROGRAM The name of the program for which I/O redirection is processed.
Note: Following the program name with NOWARN suppresses message GTD8009W, ifthe program called is not certified for OVA use by IBM.
PARM Parameter string that would normally be specified on the EXEC PGM= statement if thespecified program were being run independently. A parameter, which is too long to becompleted on one control statement, can be specified by using multiple PARMstatements. The parameter string is constructed by concatenating the strings –without supplying commas or other delimiters.
INCLUDE A volser or comma-separated list of volsers that are enclosed in parentheses.
A data set name. Either DSN= or DSNAME= is acceptable. GTDIPGM uses the systemcatalog to determine which volume the data set is resident on and registration occursfor those volumes.
A data set name level or mask. The user must supply an INCLUDE statement, such asINCLUDE LEVEL(dataset.prefix) or INCLUDE ENTRIES(generic.entry.name). Then,GTDIPGM calls IDCAMS LISTCAT to extract the individual data set names from thecatalog.
Note: GTDIPGM automatically generates inclusion list entries for those volumes thatare allocated in the JCL (STEPLIB and JOBLIB DD statements excepted). It isimperative that INCLUDE statements be provided for any volume for which I/Oredirection is to be done and that are dynamically allocated by the selected program.
Chapter 6. Using advanced functions 105
Table 22. GTDIPGM control statements (continued)
CONTROLSTATEMENT
DESCRIPTION OF STATEMENT
EXCLUDE A volser or comma-separated list of volsers that are enclosed in parentheses.
A data set name. Either DSN= or DSNAME= are acceptable. GTDIPGM uses the systemcatalog to determine which volume the data set is resident on. This form of theEXCLUDE statement typically is used when the JCL contains a DD statement thatallocates a data set, for which I/O redirection is not required.
A data set name level or mask. The user must supply an INCLUDE statement, such asINCLUDE LEVEL(dataset.prefix) or INCLUDEENTRIES(generic.entry.name). GTDIPGM calls IDCAMS LISTCAT to extract theindividual data set names from the catalog.
Note: You must provide EXCLUDE statements for all the volumes that are allocatedthrough the JCL, and for which I/O redirection is not required, except for data setswith a disposition of "new". EXCLUDE statements are present for any other volumethat might be dynamically allocated during the selected program’s run.
Figure 14 on page 106 is an example of the TDMF REPLICATE control statements that specify that an OVAsession is taking place and the actual OVA JCL (DF/DSS full volume backup).
SESSION MASTER(MVS1) AGENT(MVS8)OPTIONS(PROMPT NOPACING AUTOOPS FASTCOPY OVA)REPLICATE PROD90 VDMT33//DUMP EXEC PGM=GTDIPGM//STEPLIB DD DISP=SHR,DSN=<TDMF library>//SYSPRINT DD SYSOUT=*//IPGMOUT DD SYSOUT=*//IPGMIN DD * PROGRAM ADRDSSU INCLUDE PROD90//OUTPUT DD DISP=(NEW,CATLG),DSN=......//SYSIN DD * DUMP FULL INDYNAM(PROD90) OUTDDNAME(OUTPUT) - ALLDATA(*) ALLEXCP CANCELERROR
Figure 14. REPLICATE control statement and JCL
In the preceding example, the volume that is being backed up is production volume PROD90. During runof the backup job, requests from the ADRDSSU program to read data from PROD90 are redirected to thePoint-In-Time copy on volume VDMT33.
The INCLUDE statement for the PROD90 volume is needed because it will be dynamically allocated afterOVA registration. No EXCLUDE statement is necessary for the OUTPUT file, even if it is allocated onanother DASD volume, because it has a disposition of "NEW".
Delayed offline volume accessYou can delay the use of Offline Volume Access (OVA) to a later time rather than immediately after thePoint-In-Time (PIT) copy completes.
Purpose of Delayed OVAThe intent of Delayed OVA is to provide a mechanism whereby you can run an OVA session after a TDMFsession is complete. A Delayed OVA job is the equivalent of a single system TDMF session that can becontrolled by using the TDMF TSO or Batch Monitors.
The requirements for a Delayed OVA session are the following.
• The Communications data set (COMMDS) from the original session.
106 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
• All PIT target volumes remain offline.• Running the GTDDOVA program.
You can run a Delayed OVA batch job on any system that was a participant of the original session. Whenyou start the Delayed OVA batch job, the TDMF I/O Monitor (GTDIMON) is called. The COMMDS of theoriginal session are scanned and all completed PIT pairings become eligible for the OVA job. OVA neednot be specified in the original session to be eligible for Delayed OVA. The OVA batch job might be runimmediately after the Delayed OVA job is started. You can also use the TDMF TSO Monitor for thesesessions, except the displayed information for the offline target volume is the device address and not thevolume serial number.
Volume and group initialization proceeds as if a regular TDMF session were running. The GTDDOVAprogram reads the Communication data set and generates REPLICATE statements that contain a sourcevolume serial number and its PIT target device address. If the Delayed OVA job is running on what was anAgent system in the original migration, the target volumes must be available to TDMF at the time of themigration. To prevent corruption of the Point-In-Time copy data resident on the offline target volumes, donot vary online to any system in the complex until all the OVA backup jobs are done.
When the Communication data set is scanned, a temporary COMMDS data set is created for the DelayedOVA job. This temporary COMMDS can be pre-allocated with a DDNAME of DOVACOM, if preferred, inwhich case the data set must be large enough for a single system migration.
The Delayed OVA job remains active for 15 minutes or until an OVA job is submitted. You have no limit tothe number of runs of Delayed OVA if the Communications data set and the offline target volumes areavailable.
An example of the screen displays for a Delayed OVA session as seen from the TDMF TSO Monitor follows.
Sessions Monitor Row 1 to 6 of 6Command ===> Scroll ===> CSR------------------------------------------------------------------------TDMF Master V5.4.0 Session Active.ComDataSet : SYS03321.T021016.RA000.SWH10DVA.R0105353 Source Migration Percent Complete --------->Group VolSer Phase ..10...20...30...40...50...60...70...80...90..100 PROD92 Waiting OVA PROD94 Waiting OVAF1=Help F2=Split F3=Exit F5=Rfind F7=Back F8=ForwardF9=Swap F10=Status F12=Cancel
Figure 15. Delayed OVA session display
The example shows a Delayed OVA session. The COMMDS has a temporary name. Any OVA session cannow be started within 15 minutes of starting the Delayed OVA session. OVA in the JCL was not specified inthe original TDMF session; only the Point-In-Time option was selected.
If F10 is pressed, the status of the job is shown.
Session Status Row 1 to 2 of 2Command ===> Scroll ===> CSRSystem: Master TDMF Version: 5.4.0ComDataSet : SYS03321.T021016.RA000.SWH10DVA.R0105353 Sessions 01Number of volumes migrating : 02 Number of concurrent volumes : 02Number of volumes complete : 00 Number of volumes waiting : 00Requested Volume Device Group --- Migration --- - Error Info - Sync Action Serial Number Name Status Type System Message Goal __ PROD92 22FB Waiting OVA DOVA 22F5 __ PROD94 22F8 Waiting OVA DOVA 22F6F1=Help F2=Split F3=Exit F4=Next F5=Rfind F6=OptionsF7=Back F8=Forward F9=Swap F10=Monitor F11=ActionTg F12=Cancel
Figure 16. Delayed OVA session status
Select F6 to display the options for this session. OVA is yes.
Chapter 6. Using advanced functions 107
Session Status Row 1 to 2 of 2Command ===> Scroll ===> CSRSystem: Master TDMF Version: 4.1.0ComDataSet SYS03321.T021016.RA000.SWH10DVA.R0105353 Sessions 01Number of volumes migrating : 02 Number of concurrent volumes : 02Number of volumes complete : 00 Number of volumes waiting : 00Requested Volume Device Group --- Migration --- - Error Info - Sync Action Serial Number Name Status Type System Message Goal __ PROD92 ActCopy = N PACING = N AUTOOPER = N TIME =LOCAL CONFIRM = N PURGE = N TERMGRP = N COMPARE = N OVA = Y PPIT = N FASTCOPY = N nonPPRC = N __ PROD94 ActCopy = N PACING = N AUTOOPER = N TIME =LOCAL CONFIRM = N PURGE = N TERMGRP = N COMPARE = N OVA = Y PPIT = N FASTCOPY = N nonPPRC = N
Figure 17. Delayed OVA session options
As with any OVA session, you can end it by using the session status window and entering "H".Confirmation for the termination is then requested.
Volume Confirmation Screen Row 1 to 2 of 2Command ===> Scroll ===> CSRComDataSet: . . . . : SYS03321.T021016.RA000.SWH10DVA.R0105353Confirmation required to Halt OVA registrations.Volume Serial No. . : PROD92Confirm? (YES/NO): . . ___Command ===> Scroll ===> CSR F1=Help F2=Split F3=Exit F9=Swap F12=Cancel
Figure 18. Halt OVA registrations confirmation window
Perpetual point-in-timePerpetual Point-In-Time (PPIT) expands on the Point-In-Time (PIT) copy feature, which allows multiplesynchronization points (synch-points) to be taken within the same TDMF session. This means that a targetvolume may be updated multiple times picking up only those updates to the source volume since the lastsynchronization.
Setting up for PPITThe PPIT option can be requested on the SESSION, GROUP or REPLICATE control statement. This optionis valid only if allowed by the authorization key. The PROMPT option must also be specified for all PPITvolumes, and if a PPIT volume is a member of a REPLICATE group, all other member volumes must alsohave PPIT specified. If the option is specified as a session or group default it will be ignored for volumesspecified in MIGRATE control statements.
About this task
Options such as OVA and FASTCOPY are allowed. There are no restrictions so long as the optionsspecified are not in conflict with each other and the customer has the relevant authorization keys.
Once the synchronization prompt has been responded to, TDMF will mark the target volume offline andready for use with OVA if so desired. The TDMF session will not terminate once the initial PIT copy iscomplete. Instead a "Waiting PPIT" prompt is issued and TDMF will keep track of all updates to thesource and target volume(s) until such time that the "Waiting PPIT" prompt is responded to. TDMF willthen go into the REFRESH phase, honoring Pacing values and Active in Copy values if specified. OnceTDMF has determined that it can synchronize the volume(s), the synchronization prompt will be issuedagain. There is no limit to the number of synch-points that may be taken within a single session.
Important: If the PPIT migration source volumes exploit the Parallel Access Volume (PAV) feature, thePAV option should be specified on the SESSION, GROUP or REPLICATE control statement. Failure to dothis will result in the PAV feature being disabled (no bound alias devices) for the duration of the"perpetual" migration, with a probable impact on performance for the applications using the volumes. Thetrade off here is that the PAV option causes TDMF to allocate 16 KB of fixed HVCOMMON storage for everybound alias device on every participating system.
108 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
During an active PPIT sessionDuring the Copy and Refresh phases of an active PPIT session, there is nothing displayed on the panelsthat inform the user that a PPIT session is in progress. This information can be determined by selectingOption 2 – Current Session: User Interaction and Status of the TDMF TSO Monitor then selecting PF6 –Options. This will display the options specified for the volume or session. An example follows.
About this task
Session Status Migration OptionsCommand ===> Scroll ===> CSRSystem: Master TDMF Version: 5.4.0ComDataSet SWH10.TEST01.SYSCOM Sessions 01Number of volumes migrating : 02 Number of concurrent volumes : 02Number of volumes complete : 00 Number of volumes waiting : 00Requested Volume Device Group --- Migration --- - Error Info - Sync Action Serial Number Name Status Type System Message Goal __ PROD92 ActCopy = N PACING = S AUTOOPER = N TIME =LOCAL PROD93 CONFIRM = N PURGE = N TERMGRP = Y COMPARE = N OVA = N PPIT = Y FASTCOPY = Y nonPPRC = N __ PROD94 ActCopy = N PACING = S AUTOOPER = N TIME =LOCAL PROD95 CONFIRM = N PURGE = N TERMGRP = Y COMPARE = N OVA = N PPIT = Y FASTCOPY = Y nonPPRC = NCommand ===> Scroll ===> CSR F1=Help F2=Split F3=Exit F4=Next F5=Rfind F6=Options F7=Back F8=Forward F9=Swap F10=Monitor F11=ActionTg F12=Cancel
Note: Once the user responds to the synchronization prompt, TDMF issues a "Recycle PPIT" or "WaitingPPIT" prompt when the target volume(s) have been marked offline.
Recycle/waiting PPIT prompt display
Session Status Recycle Volume Needed
Command ===> Scroll ===> CSR
System: Master TDMF Version: 5.4.0
ComDataSet SWH10.TEST01.SYSCOM Sessions 01
Number of volumes migrating : 02 Number of concurrent volumes : 02
Number of volumes complete : 00 Number of volumes waiting : 00
Requested Volume Device Group --- Migration --- - Error Info - Sync
Action Serial Number Name Status Type System Message Goal
__ +PROD92 22FB PPIT_GRP Recycle PPIT PIT P 005
PROD93 22F5
__ PROD94 22F8 PPIT_GRP Waiting PPIT PIT P 005
PROD95 22F6
Command ===> Scroll ===> CSR
F1=Help F2=Split F3=Exit F4=Next F5=Rfind F6=Options
F7=Back F8=Forward F9=Swap F10=Monitor F11=ActionTg F12=Cancel
Figure 19. PPIT option display
Respond to the Recycle Volume Needed prompt with an R or RG in the case of a group of volumes.
Chapter 6. Using advanced functions 109
Session Status Recycle Volume NeededCommand ===> Scroll ===> CSRSystem: Master TDMF Version: 5.4.0ComDataSet SWH10.TEST01.SYSCOM Sessions 01Number of volumes migrating : 02 Number of concurrent volumes : 02Number of volumes complete : 00 Number of volumes waiting : 00Requested Volume Device Group --- Migration --- - Error Info - Sync Action Serial Number Name Status Type System Message Goal RG +PROD92 22FB PPIT_GRP Recycle PPIT PIT P 005 PROD93 22F5 __ PROD94 22F8 PPIT_GRP Waiting PPIT PIT P 005 PROD95 22F6Command ===> Scroll ===> CSR F1=Help F2=Split F3=Exit F4=Next F5=Rfind F6=Options F7=Back F8=Forward F9=Swap F10=Monitor F11=ActionTg F12=Cancel
Figure 20. Recycle/waiting PPIT prompt with response display
A confirmation screen is then presented for the volume or group.
Group Confirmation Screen Row 1 to 2 of 2Command ===> Scroll ===> CSRComDataSet: . . . . : SWH10.TEST01.SYSCOMConfirmation required to Recycle a PPiT groupGroup . . . . . . . : PPIT_GRPConfirm? (YES/NO) . . ___
Figure 21. Recycle/waiting PPIT prompt confirmation display
Once TDMF has completed processing the Confirmation, the User Interaction and Status panel willdisplay the following information.
Session Status Row 1 to 2 of 2Command ===> Scroll ===> CSRSystem: Master TDMF Version: 5.4.0ComDataSet SWH10.TEST01.SYSCOM Sessions 01Number of volumes migrating : 02 Number of concurrent volumes : 02Number of volumes complete : 00 Number of volumes waiting : 00Requested Volume Device Group --- Migration --- - Error Info - Sync Action Serial Number Name Status Type System Message Goal __ PROD92 22FB PPIT_GRP Refresh 2 PIT P 005 PROD93 22F5 __ PROD94 22F8 PPIT_GRP Refresh 2 PIT P 005 PROD95 22F6Command ===> Scroll ===> CSR F1=Help F2=Split F3=Exit F4=Next F5=Rfind F6=Options F7=Back F8=Forward F9=Swap F10=Monitor F11=ActionTg F12=Cancel
Figure 22. After recycle/waiting PPIT prompt confirmation display
To end a PPIT session, under Requested Action, enter either an E or EG to stop at the Recycle prompt oralternatively, T or TG to terminate the session.
TDMF TCP/IP supportTDMF supports DASD migration by using the existing TCP/IP connections within the businessenvironment.
Because TCP/IP requires no specialized hardware or dedicated telecommunication circuits, this TCP/IPsupport enables the elimination of expenses that are associated with migrating data. TDMF usesspecialized channel extension hardware and a telecommunications DataLink. TDMF enables the migrationof a DASD volume from a source device that is located upon a local MVS system to a new device locatedupon a remote MVS system, by using the current Point-In-Time capabilities that exist within TDMF.
110 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
TerminologyTDMF TCP/IP support terminology includes TCP/IP terms such as local MVS system, remote MVS system,and duplex migration.
Table 23. TDMF TCP/IP support terminology
Term Description
Local MVSsystem
The MVS system that is sending data on a source DASD volume, device or both.
Remote MVSsystem
The MVS system that is receiving data for a new DASD volume, device or both.
Duplex migration The creation of a Point-In-Time volume copy on the Local MVS systemsimultaneously with the creation of a duplicate copy at the Remote MVS System.
TCP/IP Password OptionAn optional TCP/IP password can be specified to provide security by permitting or denying local access toa remote system for TCP/IP functions.
The TCP/IP password is specified as an optional parameter on the REMOTE control statement of the localMaster job and the SESSION control statement of the remote Master job. If required, the password mustbe specified for both the local and remote systems. For information about the SESSION and REMOTEcontrol statements, see “Ordering input of control statements” on page 36 and “APPROVAL/NOAPPROVAL” on page 43.
Important: The password that is specified for the local and remote systems must match for a successfulconnection to be permitted.
Password Example
The following example shows a password that is specified for the local TDMF Master system that requestsa migration to a remote system.
REMOTE rem_sys HOSTNAME(remote-master-systems-host-name.com) PORT(nnnnn) PASSWORD(password)
TCP/IP migration requirements and considerationsThe requirements and considerations for TCP/IP migration include: the operating system, fast copy,configuration, TCP/IP authorization, and port numbers.
Operating system requirementThe TDMF TCP/IP feature requires the operating system to be z/OS, with the installed TCP/IP processingcomponents.
Fast copy recommendationThe Fast Copy option can reduce the amount of transferred data that decrease the duration of themigration and the resulting load on the internet connection. Use the Fast Copy option on the SESSION,GROUP, or REPLICATE control statement when you transfer data by using the internet.
For more information, see “TDMF control statements” on page 35.
Configuration considerationsTo migrate volumes using the TCP/IP feature, there must be two TDMF Master systems, each running onits own MVS system. The Local MVS Master system must be able to monitor and copy data to the remotetarget volumes. The Remote MVS Master system must be able to monitor and copy data to the remotetarget volumes. Remote MVS Agent systems protect the target volumes from unwanted access, so shouldbe specified for all the other MVS systems that share the target devices.
There are two ways to implement a migration using TCP/IP:
Chapter 6. Using advanced functions 111
• TCP/IP to a Remote System only.• TCP/IP to a remote volume and a local target volume.
The local MVS Master system must be able to access the local target volumes.
Important: A TCP/IP migration cannot be combined with any other TDMF session type. TDMF sessionsthat provide TCP/IP facilities must contain only volume pairings involving TCP/IP.
TCP/IP authorization requirementsIn a TCP/IP session, the Remote Master does not require a license key, but the GTDKEY DD statementmust be in the JCL and pointing to the GTDKEY member in a valid library.
For a sample JCL, refer to “REMOTE master JCL” on page 224. The key can either be uninitialized or itmay be a copy of the local system's key.
When connecting to TDMF on the remote Master system, you will see the following panel:
Command ===>TDMF for z/OSThere is a problem with the installation's Security environment.The problem is most likely one of the following conditions:1. This user doesn't have authority to read the security record.2. The security environment hasn't been defined. Run GTDSOPTNIf this system is being used ONLY as a REMOTE MASTER in a TCP/IPsession, you may proceed with limited functionality.Press ENTER to continue, PF3 to terminate.
PORT number requirementA port number must be specified on the EXEC PARM of each remote Master job. The remote connectionserver listens for connections through this port and also the next higher (n+1).
For an example of remote master job, see the example in “REMOTE master JCL” on page 224. The portnumber in the REMOTE control statement in the Local TDMF session must match the port number that isspecified in the Remote Master JCL. Where the size of a TCP/IP migration necessitates multipleconcurrent TDMF sessions, each session consists of a Local Master and Agents and a correspondingRemote Master job. Each of these sessions must use a unique pair of port numbers.
Note: TDMF uses two ports for communications. Only one port number is specified. TDMF then uses thatport number plus the next sequential port number. For example, if PORT=(8888) is specified, then TDMFuses port 8888 and 8889. The initial port number is used to communicate migration status while thesecond port number is used to migrate data.
Important: If the TCP/IP connection is lost, TDMF initially closes all the socket connects, thenreestablishes communications on the second port, or PORT=(nnnn)+1. When the session is active again,using a single port, the TDMF TSO Monitor at the Remote system is NOT able to display up-to-datevolume migration status.
Documenting the TDMF TCP/IP port numbersTo ensure that the chosen PORT number is identified for TDMF IP use, the port number to be used forTDMF IP needs to be documented in the .ETC.SERVICES data set. The .ETC.SERVICES data set is usedto document all port numbers that are used during product installation.
About this task
The TCP/IP port number is specified in the PORT parameter of the REMOTE control statement, asdescribed in “APPROVAL/NOAPPROVAL” on page 43.
Procedure
To document the TDMF TCP/IP port number:1. Add a statement for each of the port numbers that is being set aside for use by TDMF TCP/IP; these
statements need to be in the following format:
112 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
TDMF nnnnn/tcp # TDMF IP Session 1, Port 1TDMF nnnnn+1/tcp # TDMF IP Session 1, Port 2TDMF nnnnn+2/tcp # TDMF IP Session 2, Port 1TDMF nnnnn+3/tcp # TDMF IP Session 2, Port 2
Note: Including this statement in the ETC.SERVICES data set does not reserve the port numbers, butonly documents its use.
2. If an installation of any product restricts port number usage, the mechanism that is used to implementthat restriction must be modified to allow TDMF TCP/IP port access and usage.
3. If PROFILE.TCPIP is used to reserve the TDMF port assignment, do not use the BIND keyword on theTDMF port number to force usage of a specific IP address.
TDMF version TCP/IP compatibilityThe local and remote TDMF versions in a TCP/IP migration must be within the same IP message/responselevel. TDMF message-format incompatibilities prevent TDMF Version 5.7.0 from participating in a TCP/IPmigration with previous releases of TDMF.
The IP message/response level is used as follows in TDMF:
• TDMF V3.1 - V3.5 use level 001• TDMF V3.6 - V4.1 use level 002• TDMF V5.1 - V5.5 use level 003• TDMF V5.6 and later use level 004
TDMF is not TCP/IP compatible between these ranges. If you use the same version of TDMF at both thelocal and remote sites in a TCP/IP migration, compatibility is not an issue.
Chapter 6. Using advanced functions 113
114 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Chapter 7. Using batch utilitiesThere are ten TDMF batch utilities and one subroutine included with the base product.
Batch utilities overviewSummary description of each batch utility and subroutine are included for general information.
Table 24. Summary of utilities and subroutines
“GTDBMON” onpage 115
Specifies a batch monitor utility to control TDMF sessions at the volume or session level.
“GTDCLIP” on page121
Generates the necessary ICKDSF REFORMAT control statements as input to asubsequent ICKDSF run.
“GTDCLUP” on page123
Specifies a batch utility to clean up after a TDMF session that failed to pagefree orrelease common area (SQA/ECSA/HVCOMMON) virtual storage, or to restore theUCBDDT field in migration device UCBs.
“GTDEXTV” on page127
Specifies when files are migrated from smaller to larger devices. GTDEXTV reports onthe likely effect of TDMF and ICKDSF EXTVOC processing.
“GTDMUCB” onpage 128
Specifies an application programming interface (API) to TDMF, which checks TDMFcontrol blocks, to determine whether a device is involved in a TDMF session. Devicesmight be specified by using volume serial, UCB device address, or the 16-bit unsignedbinary device number.
“GTDQDSK” on page129
Specifies a query device configuration utility that scans all attached DASD subsystemsfor online UCBs and presents the information that is related to these UCBs in a formatsimilar to the DEVSERV QDASD MVS command.
“GTDSCAN” on page130
Specifies a batch utility that scans one or more potential source volumes for possibleI/O errors and count field mismatches before a TDMF session being run.
“GTDTERA” on page132
Specifies the GTDTERA utility which scans CPUs, CPU model types, or total onlinestorage on your systems to determine authorization key charges.
“GTDVTOC” on page135
Extracts device information for each online device in the DASD configuration. The reportfrom this batch job enables the user to easily identify those devices whose VolumeTables of Contents (VTOC) do not accurately describe their physical characteristics.
“GZDRECON” onpage 137
Specifies the REXX procedure that is used when you reconstruct TRANSMIT (XMIT) datasets that are packaged for TDMF distribution.
GTDBMONThe Batch Monitor utility is an alternative to the TDMF TSO Monitor Facility method of controlling theprogress of a TDMF migration.
The Batch Monitor is initiated as either a batch job or a started task. In either case, if a control statementinput stream is provided (by using the SYSIN DD statement), the Monitor processes all these statementsand then ends. If no SYSIN DD statement is provided, the Monitor waits for and reacts to operator modifyand stop commands. When the Batch Monitor processes control statement input, all informational anderror messages are written to the SYSPRINT data set; so always provide this DD statement.
© Copyright IBM Corp. 2010, 2019 115
To run the Batch Monitor with control statement input, use this example of the required GTDBMON JCL.
//<---------------- job card ----------------->//BMON EXEC PGM=GTDBMON//STEPLIB DD DISP=SHR,DSN=<program library>//SYSPRINT DD SYSOUT=<sysout class>//SYSUDUMP DD SYSOUT=<sysout class>//SYSIN DD * <control statements>/*
To run without control statements, use this example of a GTDBMON JCL. The JCL might be installed in asystem procedure library (PROCLIB) to enable the Batch Monitor to be started and used interactively bythe system operator. The SYSPRINT DD statement is optional, and if provided, is used to record theoperator commands and their resulting program actions.
//BMON PROC//BMON EXEC PGM=GTDBMON//STEPLIB DD DISP=SHR,DSN=<authorized program library>//SYSPRINT DD SYSOUT=<sysout class>//SYSUDUMP DD SYSOUT=<sysout class>
Operational considerationsThere are several operational considerations when you use the GTDBMON JCL.
The Batch Monitor will recognize volumes and groups, which are being migrated by a TDMF sessionwhose Master system is not running in the same MVS system as the Batch Monitor. However, the querycommand may not accurately report the status of these volumes, and requests for actions to thesevolumes will be rejected.
It is important for predictable operation of the Batch Monitor that the user adhere to two simple namingstandards for migration groups; group names must be unique across all the concurrently active TDMFsessions and group names and volume serial numbers must not match.
Only one copy of the Batch Monitor can be active at any one time. A monitor using control statements isdeemed to be active while the input records are being read and processed. A monitor reacting to operatormodify commands is active when processing any command other than a "query volume status".
Direct TDMF session controlA useful alternative to starting a GTDBMON job and controlling it with Modify commands, is to issue theModify command directly to the TDMF Master job. This function supports the majority of GTDBMONcommands and also permits the dynamic adjustment of the limit of concurrently active volumes. TheQuery, Synchronize and OVA LISTJOBS commands are not supported by the direct TDMF sessioncontrol facility.
For example, F master_job,CONCurrent nnn changes the maximum number of concurrently activemigration volumes to nnn, if this value is acceptable to the TDMF session of which master_job is theMaster.
GTDBMON - operator command processingThe Batch Monitor accepts the following operator commands when entered through a system console.
Note: P is the abbreviation for the MVS stop command, while F is the short form of modify.
Table 25. GTDBMON operator commands
Command Response
P jobname Monitor processing is immediately ended.
116 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Table 25. GTDBMON operator commands (continued)
Command Response
F jobname,Q operand
F jobname,QUERY operand
Monitor displays the status of volumes currently active in TDMF sessions.
The operands are as follows.
• volume | group – Query the status of a volume orgroup.
• copy | copying – Display the volumes in a copyphase.
• wait | waiting – Display volumes that aresuspended or waiting for confirmation.
• inact | inactive – Display volumes that arecomplete, suspended, or waiting for confirmation.
• refresh – Display volumes in refresh orsynchronization phases.
• active – Display volumes that are active.• complete – Display volumes that are complete or
have ended processing.
F jobname,T volser
F jobname,TERM volser
F jobname,TERMINATE volser
Migration of volume "volser" is terminated.
Note:
• If the volume is part of a migration group, and if theoption "terminate group on error" is set for thevolumes in this group, the T command terminates theentire group, exactly as would be the case in the TSOMonitor.
• Once TDMF enters the Approval Mode, theTERMINATE command terminates the migration withthe original target volume as the source volume. TheTERMINATE command has the same effect asAPPROVE[Yes]. After the session ends, a newmigration is required if you want to go back to theoriginal source volume.
F jobname,T grpname
F jobname,TERM grpname
F jobname,TERMINATE grpname
Migration of group "grpname" is terminated.
F jobname,S volser
F jobname,SUSPEND volser
Migration of volume "volser" is suspended.
F jobname,S grpname
F jobname,SUSPEND grpname
Migration of group "grpname" is suspended.
Chapter 7. Using batch utilities 117
Table 25. GTDBMON operator commands (continued)
Command Response
F jobname,A volser
F jobname,ALLOW volser
Migration of volume volser from one manufacturer’s mirroringtechnology to another manufacturer’s technology (for example, PPRC to SRDF)is confirmed.
F jobname,OVA LISTJOBS Displays the included volumes and registration status for all active orpending OVA jobs.
F jobname,F volser
F jobname,CONF volser
F jobname,CONFIRM volser
Migration of volume "volser" is initiated.
Note: If the volume is part of a migration group, the Fjobname,F volser command is rejected. In thiscase, the F jobname,F grpname command isrequired.
F jobname,F grpname
F jobname,CONF grpname
F jobname,CONFIRM grpname
Migration of group "grpname" is initiated.
F jobname,Z volser secs
F jobname,SYNC volser secs
F jobname,SYNCH volser secs
F jobname, SYNCHRONIZE volver secs
Synchronization target of volume "volser" is set to "secs" seconds.
Note: If the volume is part of a migration group, the Fjobname,Z volser secs command is rejected. Inthis case, the F jobname,Z grpname secscommand is required. The time operand, "secs" mustbe a 1- to 16-character numeric string with a value notless than 5 or greater than 999.
F jobname,Z grpname secs
F jobname,SYNC grpname secs
F jobname, SYNCH grpname secs
F jobname, SYNCHRONIZE grpname secs
Synchronization target of group "grpname" is set to "secs" seconds.
F jobname,C volser
F jobname,CONT volser
F jobname,CONTINUE volser
Migration of volume "volser" is continued.
F jobname,C grpname
F jobname,CONT grpname
F jobname,CONTINUE grpname
Migration of group "grpname" is continued.
118 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Table 25. GTDBMON operator commands (continued)
Command Response
F jobname,P volser
F jobname,PROMPT volser
Synchronization of volume "volser" is confirmed.
F jobname,P grpname
F jobname,PROMPT grpname
Synchronization of group "grpname" is confirmed.
F jobname,P grpname volcount
F jobname,PROMPT grpname
Confirms serial synchronization of all group volumes in chunks of “volcount” volumes which are synchronized simultaneously. By specifying a “volcount” value of 1, a serial synchronization for all group volumes is performed. Specifying 0 has the same affect as omitting it (that is, simultaneous synchronization of all group volumes).
F jobname,I volser Reinitializes the migration of volume "volser".
F jobname,H volser
F jobname,HALT volser
Halts all new Offline Volume Access (OVA) Registrations of volume "volser".
F jobname,H grpname
F jobname,HALT grpname
Halts all new Offline Volume Access (OVA) Registrations of group "grpname".
F jobname,E volser
F jobname,END volser
Ends the Perpetual Point-In-Time (PPIT) processing of volume "volser".
F jobname,E grpname
F jobname,END grpname
Ends the Perpetual Point-In-Time (PPIT) processing of group "grpname".
F jobname,R volser
F jobname,RECYCLE volser
Recycles a Perpetual Point-In-Time (PPIT) process of volume "volser".
F jobname,R grpname
F jobname,RECYCLE grpname
Recycles a Perpetual Point-In-Time (PPIT) process of group "grpname".
GTDBMON - SYSIN control statement processingThe following control statements are acceptable in the input stream. Commands and operands can startanywhere on the input record, except that the entire statement must be completed by column 71 of theinput record. No continuation of a control statement is allowed. An input record that is blank in the first 71positions and records containing an asterisk in the first character position are treated as comments.Comments can also be appended to control statements, leaving at least one blank character after theoperand.
Note: In the following table, the single character in parentheses is the acceptable abbreviation for thecommand.
Chapter 7. Using batch utilities 119
Table 26. GTDBMON SYSIN control statements
Command ShortForm
Purpose
QUERY (Q) Report on volume status in active TDMF migration sessions.
Operands are:
• volume | group – Query the status of a volume or group.• copy | copying – Display volumes in copy phase.• wait | waiting – Display volumes that are suspended or waiting
for confirmation.• inact | inactive – Display volumes that are complete, suspended,
or waiting for confirmation.• refresh – Display volumes in refresh or synchronization phases.• active – Display volumes that are active.• complete – Display volumes that are complete or have been
terminated.
TERMINATE volser (T) Migration of volume "volser" is terminated.
Note: If the volume is part of a migration group, and if the option"terminate group on error" is set for the volumes in this group, theTERMINATE command will terminate the entire group, exactly as wouldbe the case in the TSO Monitor.
Once TDMF has entered the Approval Mode, the TERMINATE commandwill terminate the migration with the original target volume as the sourcevolume. The TERMINATE command will have the same effect asAPPROVE[Yes]. After the session has terminated, a new migration isrequired if you need to go back to the original source volume.
TERMINATEgrpname
Migration of group "grpname" is terminated.
ALLOW volser (A) Migration of volume "volser" from one manufacturer’s mirroringtechnology to another manufacturers technology (for example, PPRC toSRDF) is confirmed.
CONFIRM volser (F) Migration of volume "volser" is initiated.
Note: If the volume is part of a migration group, the CONFIRM volsercommand will be rejected. In this case the CONFIRM grpname commandis required.
CONFIRM grpname Migration of group "grpname" is initiated.
OVA Report on OVA jobs status.
Operands are:
• Listjobs - display the included volumes and registration status for allactive or pending OVA jobs.
120 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Table 26. GTDBMON SYSIN control statements (continued)
Command ShortForm
Purpose
SYNCHRONIZEvolser secs
(Z) The Synchronization target of volume "volser" is set to "secs" seconds.
The time operand, "secs" must be a one to 16 character numeric stringwith a value not less than five or greater than 999.
Note: If the volume is part of a migration group, the SYNCHRONIZEvolser secs command will be rejected. In this case, the SYNCHRONIZEgrpname secs command is required.
SYNCHRONIZEgrpname secs
The Synchronization target of group "grpname" is set to "secs" seconds.
SUSPEND volser (S) Migration of volume "volser" is suspended.
SUSPEND grpname Migration of group "grpname" is suspended.
CONTINUE volser (C) Migration of volume "volser" is continued.
CONTINUE grpname Migration of group "grpname" is continued.
PROMPT volser (P) Synchronization of volume "volser" is confirmed.
Note: This command will be rejected if the volume is part of a migrationgroup.
PROMPT grpname Synchronization of group "grpname" is confirmed.
REINIT volser (I) Reinitialize the migration of volume "volser".
HALT volser (H) Halt all new Offline Volume Access (OVA) Registrations of volume"volser".
HALT grpname (H) Halt all new Offline Volume Access (OVA) Registrations of group"grpname".
END volser (E) End the Perpetual Point-In-Time (PPIT) processing of volume "volser".
END grpname (E) End the Perpetual Point-In-Time (PPIT) processing of group "grpname".
RECYCLE volser (R) Recycle a Perpetual Point-In-Time (PPIT) process of volume "volser".
RECYCLE grpname (R) Recycle a Perpetual Point-In-Time (PPIT) process of group "grpname".
GTDBMON messagesSee IBM Transparent Data Migration Facility (TDMF) for z/OS Messages and Codes (SC27-8964-00) formessage information.
GTDBMON - printed outputIf a SYSPRINT DD statement is provided, operator messages that are issued by the authorized version ofthe Batch Monitor (driven by operator modify commands) appear in the printed report. When the BatchMonitor is run with control statement input, some of the printed messages are identical to these operatormessages. See IBM Transparent Data Migration Facility (TDMF) for z/OS Messages and Codes(SC27-8964-00) for information.
GTDCLIPUsing the GTDCLIP program you can easily change the serial numbers of Point-In-Time target, or theSwap Migration source devices.
Chapter 7. Using batch utilities 121
Using GTDCLIP after point-in-time copiesThe GTDCLIP program allows the user to easily change the serial numbers of TDMF Point-In-Time Copytarget volumes to match their original source volumes. The program produces "REFORMAT" controlstatements for a later ICKDSF program run. This facility can speed up data center relocation, for example,by automating the volume relabeling process, which is necessary before IPLing with the targetconfiguration.
Before you begin
Important: The user should be aware, however, that after the run of the ICKDSF program with the inputstatements generated by GTDCLIP, there would be duplicate volume serial numbers in the configuration.The facility is intended for use in an environment where access to the source volumes is not possible fromthe IPLing processor(s).
Using GTDCLIP after migrationsThe GTDCLIP program can be used after a migration to generate "INIT" statements for ICKDSF to re-initialize the original source volumes. In this case, the user can provide a "MODEL" DD statement referringto an online volume with the VTOC and index size and locations that ICKDSF will apply. If no "MODEL" DDstatement is supplied, GTDCLIP will initialize each source volume with its previous VTOC and index sizeand location.
About this task
The input file is a single TDMF Communications Data Set generated by a previous TDMF session, specifiedby the SYSCOM DD statement. This session must not have contained a mixture of Swap and Point-In-Time migrations. GTDCLIP reads the TDMF control blocks in this data set and will produce a "REFORMAT"control statement for each successful Point-In-Time volume copy or an "INIT" statement for eachsuccessful volume Swap.
By default, the GTDCLIP program will generate control statements to initialize each swap migrationsource volume as if it were to be uninstalled, erasing customer data from every track. An optional runparameter can be used to prevent this time-consuming process, or to reinitialize the volume as system-managed.
The supported parameters are:
• PARM=NOVAL
Initialize the volume at the minimal level.• PARM=STGR
Initialize the volume as system-managed.
The parameters can also be used in combination, separated by a comma. Additional customization of thegenerated INIT commands can be achieved by directing the output control statements to a data set,where they can be modified before being used as input to ICKDSF.
Note: The TDMF I/O service routine ( GTDCDIO) does not support concatenated input files, hence onlyone Communications Data Set can be processed by GTDCLIP in one job step.
The output file, with a DDname of GTDOUT, is sequential with 80-byte records and contains thegenerated ICKDSF control statements. It is expected that these card images will usually be written to adirect access or VIO data set for input to the ICKDSF program.
The following is an example of the JCL required to run GTDCLIP:
122 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
//<---------------- job card ----------------->//FIXIT EXEC PGM=GTDCLIP//STEPLIB DD DISP=SHR,DSN=<program library>//SYSCOM DD DISP=SHR,DSN=<TDMF Communications Data Set>//GTDOUT DD DISP=(,PASS),UNIT=SYSALLDA,SPACE=(TRK,1)//* Model DD statement only required for INIT (Swap) type COMMDS//MODEL DD DISP=OLD,UNIT=SYSALLDA,VOL=SER=<model volume>//*//* OPTIONAL CLIP STEP FOLLOWS//*//CLIPIT EXEC PGM=ICKDSF,PARM=NOREPLYU,COND=(0,LT)//SYSPRINT DD SYSOUT=*//SYSIN DD DISP=(OLD,DELETE),DSN=*.FIXIT.GTDOUT
Figure 23. GTDCLIP JCL
GTDCLIP - return codesGTDCLIP generates an explanatory message before a non-zero return code is issued. These return codesare explained in IBM Transparent Data Migration Facility (TDMF) for z/OS Messages and Codes(SC27-8964-00).
GTDCLUPThe GTDCLUP program can be used to clean up after a TDMF session that failed to pagefree or releasecommon area (SQA/ECSA/HVCOMMON) virtual storage, or to restore the UCBDDT field in migration deviceUCBs. Under the control of run parameters, the program also cleans up the UCB for a single device thathas an incorrect DDT pointer or is incorrectly set to the "DDR" quiesce level.
GTDCLUP has two primary modes of operation:
1. Scan and repair the entire system.
GTDCLUP scans the entire system for common virtual storage that was not pagefreed or released by aTDMF session and then scans the devices in every installed DASD subsystem and repair devices foundto have an incorrect DDT pointer or incorrectly set to the "DDR" quiesce level. To run GTDCLUP in thismode, see “GTDCLUP: scanning the entire system” on page 123 for more details.
2. Scan and repair a specific device.
GTDCLUP scans and repairs a specific device if the device is found to have an incorrect DDT pointer orincorrectly set to the "DDR" quiesce level. To run GTDCLUP in this mode, see “GTDCLUP - repairing asingle device” on page 126 for more details.
GTDCLUP: scanning the entire systemThe GTDCLUP program can be used to scan the entire system.
Running JCLList of the required JCL to run GTDCLUP to clean up storage and devices from a failed session, and then tovalidate all the DASD UCBs in the system.
//CLUP EXEC PGM=GTDCLUP//STEPLIB DD DISP=SHR,DSN=<TDMF load library>//SYSPRINT DD SYSOUT=*//SYSSNAP DD SYSOUT=*
Figure 24. GTDCLUP JCL - failed session
To clean up a session that did not initialize, or that was at a different version of the product, add thePARM=’FORCE’ option to the run statement:
Chapter 7. Using batch utilities 123
//CLUP EXEC PGM=GTDCLUP,PARM=’FORCE’//STEPLIB DD DISP=SHR,DSN=<TDMF load library>//SYSPRINT DD SYSOUT=*//SYSSNAP DD SYSOUT=*
Figure 25. GTDCLUP JCL - force
To scan for devices that are incorrectly disabled for FlashCopy, add PARM=’FLASHCOPY’ to the runstatement:
//CLUP EXEC PGM=GTDCLUP,PARM=’FLASHCOPY’//STEPLIB DD DISP=SHR,DSN=<TDMF load library>//SYSPRINT DD SYSOUT=*//SYSSNAP DD SYSOUT=*
Figure 26. GTDCLUP JCL - FlashCopy
Control block search and scan entire systemFor more information, see the following list, which includes instructions on scanning and repairing theentire system.
Common area virtual storage scan and repairWhen started without any run parameter, the program uses the VSMLIST service to map ECSA and thensearches every storage range that is allocated in subpool 228, protect key zero, non-fetch protected,storage for a TDMF main control block. For each control block that is found, representing a migrationsession that is/was active in the system, the program initiates the following processing.
1. Version check
The GTDCLUP program will attempt to clean up storage left allocated by an earlier release of TDMF.2. Communications Data Set serialization
The program extracts the data set name and volume serial number from the TDMF control block andissues an exclusive ENQ with SYSTEM scope for the volser/dsname resource. If the data set is stillowned by a migration session, the ENQ will fail, the TDMF block is ignored and the search of allocatedECSA continues.
3. Communications Data Set allocation
The program attempts to dynamically allocate, then OPEN for update, the session’s CommunicationsData Set. If either the allocation or the OPEN fails, the data set will be discarded but processing of theorphaned control blocks will continue.
4. Communications Data Set currency check
The first two records are read from the newly allocated Communications Data Set and the TDMFMSEfor the running system is found, using the relative system number field from the TDMF block. TheTDMF block for this system is then read from the Communications Data Set and its latest heartbeatTOD clock value is compared with the control block found in ECSA. If the heartbeat time on disk islater than the value in storage, the data set is discarded but storage cleanup processing continues.
5. Communications Data Set update
If the Communications Data Set is successfully ENQueued, and the TDMF heartbeat time in virtualstorage is not earlier than the equivalent time in the data set for this system, the session beingcleaned up is assumed to "own" the data set and the program attempts to bring the data set up todate by writing the session’s control blocks to disk, as they are found in storage.
6. Session and volume message processing
The session level messages from the TDMF and TSOM control blocks are formatted and written toSYSPRINT, immediately following the program’s own status messages.
The control blocks for each volume in the session are accessed and the volume messages are printedfor the system. If possible, these control blocks are also written to the SYSCOM data set, so that theycan be displayed by the TSO Monitor (Display Details or Merged Messages option).
124 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
7. Control block SNAP
If a SYSSNAP DD statement was provided, the orphaned control blocks are snapped as they arefound and processed.
8. UCB cleanup
For each volume in the session, the source, target and duplex UCBs are examined and, if necessary,the UCBDDT and UCBLEVEL fields is corrected.
If the associated DDT is invalid, the UCBDDT pointer is changed to point to one of the DDTs in thenucleus (IGGDDT01 or IGGDDTA1).
If the DDT is imbedded in a DDTR control block for the session being cleaned up, the DDT address inthe DDTR are stored back into the UCBDDT field.
If a UCB that had an invalid DDT, or one that was imbedded in an orphaned TDMFDDTR control block,has the DDR quiesce level set, the IOSLEVEL service is used to reset this quiesce level. However, ifthe program cannot determine whether swap processing was started for the volume, the UCB is leftquiesced for processing by the final UCB scan.
9. Storage cleanup
The program will unfix and release the ECSA, HVCOMMON and SQA storage owned by the session.The total number of bytes of storage cleaned up is displayed in a message at termination.
10. DASD UCB scan
As a final check, the program attempts to validate the UCB for every installed Direct Access device.This processing is described in detail in “Scan all installed DASD UCBs” on page 125.
Scan all installed DASD UCBsWhen the common area virtual storage scan completes, the program scans the devices for every installedDASD subsystem.
This processing involves the following validation and repair operations:
1. Dynamic LPA check
When a TDMF I/O Monitor module loaded in dynamic LPA and it is not shared by a migration session,its start address is saved. Any DDT found in the UCB scan with a DDTSIO field that points to thismodule is considered invalid.
2. DDT Address check
When a device’s UCBDDT field points to one of the four DDTs at label IGGDDT01 in the nucleus, it isassumed to be valid.
3. DDT format check
When the DDTNAME field does not contain the characters "DDT" followed by a space, the DDT isassumed to be invalid. The UCBDDT field is corrected to point to one of the nucleus resident DDTs.
4. DDTSIO address check
When the DDTSIO field points to IECDPRFX or KDFSSIOE (Candle Corporation), the DDT is assumed tobe valid. If the DDT is part of a TDMFDDTR control block for a session that is no longer active, theUCBDDT pointer is repaired. If the DDT is not embedded in a TDMFDDTR block, but the DDTSIO fieldpoints to a TDMF I/O Monitor module, the DDTSIO field is changed to point to IECDPRFX.
5. DDR quiesce level check
The quiesce level is not reset for any device in an active migration session. In this case, the programmust be resubmitted with a "FIXUCB" parameter for each device to be repaired.
6. Volume Label check
When there are no active TDMF sessions and an online volume with the DDR quiesce level set in itsUCB, the program generates a DCB and DEB and attempt to read the serial number from the volumelabel. If the device is offline or boxed, or the volume serial number in the label matches the serialnumber in the UCB, the DDR quiesce level is reset.
Chapter 7. Using batch utilities 125
7. Enable Parallel Access Volumes (PAV)
Every base device that previously had bound alias devices, but that currently has none, the programcalls the API routine to attempt to bind any PAV alias UCBs that are permanently associated with thebase.
8. Enable FlashCopy (optional)
Every device that is attached to a storage subsystem that supports FlashCopy, but that does notappear to have FlashCopy enabled, is called by the API routine to re-enable the feature. This optioncauses the program to send a warning message for every device on the controller that is notconfigured for FlashCopy. For this reason, the FlashCopy run parameter must be used only when thereare devices that are left disabled for FlashCopy.
GTDCLUP - repairing a single deviceTo run GTDCLUP to repair a single device, see the instructions in Running JCL.
Running JCLTo run GTDCLUP to repair a single device, the following JCL is required. You must specifythe ’FIXUCB=unit’ parameter on the run statement, where the unit is the device address that needs to berepaired.
The following is an example of a GTDCLUP JCL (repairing a single device)
//CLUP EXEC PGM=GTDCLUP,PARM=’FIXUCB=unit’//STEPLIB DD DISP=SHR,DSN=<TDMF load library>//SYSPRINT DD SYSOUT=*//SYSSNAP DD SYSOUT=*
When run with a running parameter of "FIXUCB=xxxx", the program locates the UCB and DDT for device"xxxx" and attempt to repair it:
1. DDT address check
If the UCBDDT field points to an inaccessible or invalid DDT it is changed to point to one of thestandard nucleus resident DDTs. If the DDT is part of a TDMFDDTR control block for a session that isno longer active, the UCBDDT pointer is replaced with the DDT address from the TDMFDDTR controlblock.
2. DDR quiesce level check
It is possible to reset the quiesce level for a single device, even if there is an active TDMF session, sogreat care must be taken. Before using the IOSLEVEL RESET service, the program attempts to acquireSYSTEMS scope control of the "MIGRTDMF/volser" ENQ resource. This will only protect a device that isactive in a TDMF session (but not currently being monitored) if the Master system for the session isrunning on a system in the same GRS- or MIMplex as the GTDCLUP program.
GTDCLUP - program abend codesGTDCLUP can issue the following program abend codes.
Table 27. GTDCLUP abend codes
U0001 An invalid data area was returned by the VSMLIST macro. In the memory dump forthis abend, register 14 points to the 1 MB work area passed to the VSMLIST service.
U0101 A nonzero return code was received from a DEBCHK (ADD) macro. This return code isused as the abend reason code.
126 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
GTDCLUP - program messagesGTDCLUP messages can be found in IBM Transparent Data Migration Facility (TDMF) for z/OS Messagesand Codes (SC27-8964-00).
GTDEXTVUsers migrating volumes to devices with a greater number of primary cylinders can use the "automaticICKDSF EXTVTOC" facility of TDMF to invoke the REFORMAT function without the necessity of varyingthese volumes offline to sharing systems. The GTDEXTV program can be run before the start of amigration to report on the likely effect of this TDMF and ICKDSF processing.
In particular, the program will report:
• Volume status that would prevent a VTOC reformat attempt by TDMF.• The "optimum" extended VTOC size, as calculated by TDMF.• The data set(s) that must be moved to facilitate the "optimal" VTOC extension.
GTDEXTV should be run on the same z/OS system as intended for the TDMF Master job, because it mightmake a prediction based on the level of ICKDSF installed which would not be valid if the TDMF Masterwas run on another of the participating systems.
The following JCL is required to run the GTDEXTV program:
//step EXEC PGM=GTDEXTV//STEPLIB DD DISP=SHR,DSN=<authorized library>//SYSPRINT DD SYSOUT=<sysout class>//SYSIN DD *< control statements >/*
Figure 27. GTDEXTV JCL
The control statements processed by GTDEXTV have the same format as a TDMF Swap migration.Migration statements contain the following:
• The word "MIGRATE"• A source volume serial number• A target volume serial number
These three items must appear on the same input record.
The program uses the EXCP macro to read the VTOCs of each source and target volume. The sourcevolume's mount status and VTOC indicators are checked for compatibility with the ICKDSF REFORMATfunction.
Where a target device has more primary cylinders than are described by the source volume's VTOC, theoptimum size to which the VTOC should be extended is:
• extracted from the EXTVTOC (tracks) option if present in the input stream, or• set to be the same as the size of the target volumes, or• calculated using the algorithm described in “Coexistance with EMC’s z/OS migrator” on page 78.
If the tracks into which the VTOC would need to extend to achieve this optimum size are currentlyoccupied by data sets, the names of these data sets are reported.
GTDEXTV - program abend codesA failure to OPEN the input control statement or output message files or an unsuccessful EXCP requestcauses abnormal termination of the GTDEXTV program.
Abend codes for GTDEXTV are described in the table below.
Chapter 7. Using batch utilities 127
Table 28. GTDEXTV abend codes
Abend Code Description
U0103 OPEN error. Probably a missing SYSIN or SYSPRINT DD statement.
U0135 I/O Error on a Read Device Characteristics CCW for a migration targetvolume. The Program issues message XTV035S before the Abend.
U0136 I/O Error on a Read Data CCW for the first VTOC record on amigration target volume. The Program issues message XTV036Sbefore the Abend.
U0137 I/O Error on a Read Device Characteristics CCW for a migrationsource volume. The Program issues message XTV037S before theAbend.
U0138 I/O Error on a Read Track CCW for a migration source volume. TheProgram issues message XTV038S before the Abend.
GTDEXTV program messagesAll GTDEXTV messages are documented in IBM Transparent Data Migration Facility (TDMF) for z/OSMessages and Codes (SC27-8964-00).
GTDMUCBThe GTDMUCB application program interface (API) examines TDMF control blocks for active migrationsand returns information related to caller-specified devices. The program can be invoked using the LINKmacro, or can be called as either a linkage-edited subroutine or a dynamically loaded module.
GTDMUCB must be entered and always returns in 31-bit addressing mode.
OperationGTDMUCB is passed either a parameter structure or the address of a single UCB by the calling program. Ifgeneral-purpose register one (R1) is non-zero, it must point to a control structure as described below.Within this structure, the caller can specify up to 65535 devices to be processed by the routine. Eachdevice must be specified using its UCB address, its 16-bit unsigned binary device number or its volumeserial. All devices in the list must be specified in the same way, and a code describing this specificationmust be stored in the third byte of the structure.
The parameter structure has the following format:
Table 29. GTDMUCB parameter structure
OFFSET (BYTES) LEN Field Description
0 2 Number of devices in the list. If this field is zero, the device list isassumed to contain one entry, else it should contain a binary half-word of the number of devices in the list.
2 1 Device specification type. The format of the device information in theattached list. This is a one-byte binary value. Supported values are:
• 0: UCB address. This can be a captured or actual address.• 1: Device number.• 2: Volume serial number.
128 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Table 29. GTDMUCB parameter structure (continued)
OFFSET (BYTES) LEN Field Description
3 1 Function code. This is a one-byte binary value; supported value are:
• 0: Return device status information.• 1: Return extended device status.
4 varies Device list. Each entry in the list comprises a two-byte return code, atwo-byte address space ID and a two-, four- or six-byte fieldcontaining a hexadecimal device number, a 31-bit UCB address or avolume serial number.
GTDMUCB - return codesOn return from GTDMUCB, register 15 contains return codes. Return codes reflect the status of a sourcevolume (copy, refresh, synch ready, OVA ready, PPIT ready, etc.) The return codes vary, depending onwhether function code 0 or function code 1 is being used. Note that if more than one of the followingconditions are met, the resulting return code will be the highest of those applicable. For more informationabout return codes and messages, see IBM Transparent Data Migration Facility (TDMF) for z/OS Messagesand Codes (SC27-8964-00).
Table 30. GTDMUCB - return codes
Function Code Return Code Description
0 01 The device is not involved in a migration.
0 42 A source volume is being migrated.
0 82 The device is the target of a migration.
0 12 The device is offline or the volume serial could not be located on anonline device.
1 32 Activated (copy phase).
1 36 First refresh phase.
1 40 Suspended.
1 44 Subsequent refresh phase.
1 48 Ready to synch.
1 52 Volume swapped.
1. If R1 is zero at entry to GTDMUCB, R0 is assumed to point to a UCB.2. For each device entry whose return code is four or eight, the ASID of the associated TDMF session is
returned in bytes two and three of the entry.
For every device for which a migration is not currently active or pending, but which is found to be"completed" or "terminated" in an active TDMF session, the ASID of the TDMF job for this session isreturned, together with a device return code of zero.
GTDQDSKThe GTDQDSK (Query Device Configuration) utility provides the TDMF user with a convenient method ofextracting configuration data from a device in each DASD subsystem. This information contains
Chapter 7. Using batch utilities 129
equipment serial numbers and is described in "Read Configuration Data" in the manual, IBM 3990/9390Storage Control Reference (GA32-0274).
GTDQDSK is initiated as a batch job. The DF/SMS control blocks describing the attached DASDsubsystems and for each subsystem the associated Unit Control Blocks (UCBs) are examined. Theprogram can not allocate volumes formatted for VM/390 use. When an online unit has a standard VTOCpointer (), the volume is dynamically allocated using the special data set name "FORMAT4.DSCB", openedfor input, and the device and configuration data are read. This data is formatted in a manner similar to theresponse from the DEVSERV MVS command in the SYSPRINT output file. Once a usable volume has beenfound, processing continues with selection of the next subsystem.
The following is an example of the JCL required to run GTDQDSK.
//<---------------- job card ----------------->//QDSK EXEC PGM=GTDQDSK//STEPLIB DD DISP=SHR,DSN=<program library>//SYSPRINT DD SYSOUT=<sysout class>//SYSUDUMP DD SYSOUT=<sysout class>/*
Figure 28. GTDQDSK JCL
GTDQDSK - printed outputMessages may be written to the SYSPRINT file. See IBM Transparent Data Migration Facility (TDMF) forz/OS Messages and Codes (SC27-8964-00).
GTDQDSK - user abendsUser abend codes are documented in IBM Transparent Data Migration Facility (TDMF) for z/OS Messagesand Codes (SC27-8964-00).
GTDSCANThe GTDSCAN program scans one or more volumes, reporting on I/O errors and count field mismatches.The program outputs a report on the suitability of the volume for migration using TDMF.
GTDSCAN must be installed and run as an APF-authorized program. GTDSCAN must not be linkage editedwith the "RENT" or "REFR" attributes.
There are no input files to the program; a run parameter controls which unit(s) will be processed and theprogram then uses dynamic allocation.
The report is written to the output file defined by the SYSPRINT DD statement. Because an I/O errorcondition can trigger an ABEND with the dump option, a SYSUDUMP DD statement should be provided.
The following is an example of the JCL required to run GTDSCAN:
//<---------------- job card ----------------->//FIXIT EXEC PGM=GTDSCAN,PARM=’uuuu’//STEPLIB DD DISP=SHR,DSN=<authorized program library>//SYSPRINT DD SYSOUT=<sysout class>//SYSUDUMP DD SYSOUT=<sysout class>/*
Figure 29. GTDSCAN JCL
Run parameterThe GTDSCAN run parameter is used to specify the device(s) to be scanned ("uuuu" in the exampleabove). It must be supplied and must be a valid device number. Wild card values are allowed by use of anasterisk.
By using the asterisk as a wild card, the low order device range is replaced with zeroes and the high orderdevice range is replaced with ‘FF’: EXEC PGM=GTDSCAN,PARM='xx**' will cause the program to readevery track of every online volume in the device range xx00 to xxFF. For example, PARM=’213C’ directs
130 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
processing to one specific unit, while PARM=’1F**’ causes all the online DASD units in the range 1F00through 1FFF to be scanned.
TCP/IP migrationsThe FASTCOPY parameter is used to scan only cylinders that are allocated on the selected devices forTCP/IP migrations.
PARM='FASTCOPY,xx**'
To use FASTCOPY for sample output, refer to GTDSCAN - FASTCOPY Sample Output, here: “GTDSCANsample output” on page 132.
Also, specifying PARM='TCPIP,xx**' not only bypasses deallocated cylinders, but also counts thenumber of bytes of data that needs to be copied from each volume. The number of empty tracks is alsoreported because they are sent as a bitmap, rather than the 8 bytes of data that an empty track wouldoccupy on the target.
To use TCPIP for sample output, refer to GTDSCAN - TCPIP Sample Output, here: “GTDSCAN sampleoutput” on page 132.
OperationGTDSCAN operates by reading each specified device, a cylinder at a time, and examines each record.
Processing of the volume proceeds as follows:
Table 31. GTDSCAN volume processing summary
Volume Label If the CCHH component of the VTOC pointer in the label is zero, TDMF is treated as aVM volume during migration.
Record zero The record zero on every track must have a count field whose CCHH (bytes zerothrough three) matches its physical location and whose record number (byte four) iszero, key length (byte five) is zero and data length (bytes six and seven) is eight.
All Records The track number (bytes two and three of every count field) must match the physicaltrack number.
First data recordon track zero
For a VM volume it is permissible for the cylinder number (bytes zero and one of thecount field) in the first data record on track zero of a cylinder to be lower than thephysical cylinder on which it resides.
Other datarecords
Except for VM volumes, all count fields’ CCHH must match the physical cylinder andtrack, or select the Tolerate Invalid Count Fields option. For volumes recognized asbeing formatted for CP or CMS use, the cylinder number in the count fields of all thedata records on a cylinder must match either the physical cylinder or the count fieldof record one on track zero of the cylinder.
Chapter 7. Using batch utilities 131
GTDSCAN sample outputTwo examples of GTDSCAN sample output.
04/20/2009 04:11:56 GTD9110I Beginning scan of volume QASE00.04/20/2009 04:15:30 GTD9120I Scan completed on volume QASE00.04/20/2009 04:15:30 GTD9125I Number of cylinders allocated: 23,127.04/20/2009 04:15:30 GTD9110I Beginning scan of volume QASE01.04/20/2009 04:15:32 GTD9120I Scan completed on volume QASE01.04/20/2009 04:15:32 GTD9125I Number of cylinders allocated: 78.04/20/2009 04:15:32 GTD9110I Beginning scan of volume QASE02.04/20/2009 04:18:22 GTD9120I Scan completed on volume QASE02.04/20/2009 04:18:22 GTD9125I Number of cylinders allocated: 13,257.04/20/2009 04:18:22 GTD9110I Beginning scan of volume QASE03.04/20/2009 04:18:24 GTD9120I Scan completed on volume QASE03.04/20/2009 04:18:24 GTD9125I Number of cylinders allocated: 158.04/20/2009 04:18:24 GTD9110I Beginning scan of volume RGPE01.04/20/2009 04:18:24 GTD9120I Scan completed on volume RGPE01.04/20/2009 04:18:24 GTD9125I Number of cylinders allocated: 108.04/20/2009 04:18:24 GTD9109W Device 1605 cannot be scanned; offline or pending.
GTDSCAN - TCPIP sample output
04/20/2009 02:55:55 GTD9110I Beginning scan of volume RGPE01.04/20/2009 02:55:55 GTD9120I Scan completed on volume RGPE01.04/20/2009 02:55:55 GTD9125I Number of cylinders allocated: 108.04/20/2009 02:55:55 GTD9126I Number of empty tracks: 0.04/20/2009 02:55:55 GTD9127I Number of KBs of data: 27,089.
Figure 30. GTDSCAN - FASTCOPY sample output
GTDSCAN - messages, return codes and abend codesMessages written to the SYSPRINT file are documented in IBM Transparent Data Migration Facility (TDMF)for z/OS Messages and Codes (SC27-8964-00).
GTDTERAThe GTDTERA utility identifies TDMF authorization key charges based on CPUs, CPU model types, or totalonline storage on your systems. IBM might request that the GTDTERA utility be run on every system todetermine charges.
Sample JCL for GTDTERAThe following is an example of the JCL required to run the GTDTERA utility.
//<--------------------- job card -------------------->//TERA EXEC PGM=GTDTERA//STEPLIB DD DISP=SHR,DSN=<program library>//SYSPRINT DD SYSOUT=<sysout class>//SYSUDUMP DD SYSOUT=<sysout class>/*
Figure 31. GTDTERA JCL
132 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
GTDTERA - user abendsUser Abend codes are documented in IBM Transparent Data Migration Facility (TDMF) for z/OS Messagesand Codes (SC27-8964-00).
GTDTERA sample outputExamples of GTDTERA sample output.
Table 32. GTDTERA sample output
TDMF for z/OSVersion 5.4.0Operating EnvironmentCPU Serial / Model : 0001232E 2066CPC Node Description : 002066.0A2.IBM.02.00000001232ECPC STSI Description : 2066.0A2.IBM.02.000000000001232EIODF Data Set : Z800.IODF20Machine Hardware Name: Z800Machine Configuration: Logical PartitionMachine MSUs : 00000046LPAR MSUs : 00000046LPAR Name : Z8L1System Name : TDM1Sysplex Name : TDMFPLEXSystem ID ( SMFID ) : TDM1SCP Name : z/OS 01.09.00SCP FMID : HBB7706SCP PID : 5694-A01SCP Product Owner : IBM CORPJES SUBSYSTEM : JES2ETR ID : 09Local Time : 01/29/2009 08:45:49.22GMT Time : 01/29/2009 16:45:49.22Local Offset : -08:00:00Leap Seconds : +015
Chapter 7. Using batch utilities 133
Table 32. GTDTERA sample output (continued)
Serial SCU UCBs UCBs GBSSID MFG No. Type Defined Online OnlineC100 HTC 30396 3990006 47 45 79C200 HTC 30396 3990006 15 14 1190002 AMH 40103 3990006 28 28 3380003 AMH 40103 3990006 24 21 3200004 AMH 40103 3990006 26 26 3330005 AMH 40103 3990006 23 23 2930006 AMH 40103 3990006 28 28 3380007 AMH 40103 3990006 23 23 3220008 AMH 40103 3990006 64 60 1640009 AMH 40103 3990006 24 23 3250058 AMH 30258 3990G03 16 15 941000 IBM 14468 2105F20 15 13 01002 IBM 14468 2105F20 01 01 51003 IBM 14468 2105F20 32 28 1682000 AMH 30355 3990006 64 64 1532001 AMH 30355 3990006 64 63 1672002 AMH 30355 3990006 64 53 1272003 AMH 30355 3990006 64 64 1682004 AMH 30714 3990006 64 64 1642005 AMH 30714 3990006 64 64 1282006 AMH 30714 3990006 64 64 1682007 ***** 30714 3990 64 00 04650 AMH 30345 3990006 64 64 1484651 AMH 30345 3990006 64 64 1524652 AMH 30345 3990006 64 64 1285130 AMH 30363 3990006 64 64 645131 AMH 30363 3990006 64 64 566555 IBM 92361 3990006 04 04 1677C00 AMH 20104 3990006 28 28 1697C01 AMH 20104 3990006 28 28 1687D00 AMH 20104 3990006 28 28 1687D01 AMH 20104 3990006 28 28 1687E00 AMH 20104 3990006 28 28 487E01 AMH 20104 3990006 28 28 687F00 AMH 20104 3990006 28 28 1677F01 AMH 20104 3990006 28 28 07140 IBM 90233 3990006 17 17 638801 AMH 00001 3990006 18 18 33Number of DASD subsystems : 38Number of DASD volumes defined : 2,961Number of DASD volumes online : 2,867Amount of online storage (GB) : 6,2854650 AMH 30345 3990006 64 64 1484651 AMH 30345 3990006 64 64 1524652 AMH 30345 3990006 64 64 1285130 AMH 30363 3990006 64 64 645131 AMH 30363 3990006 64 64 566555 IBM 92361 3990006 04 04 167
Serial SCU UCBs UCBs GBSSID MFG No. Type Defined Online Online7C00 AMH 20104 3990006 28 28 1697C01 AMH 20104 3990006 28 28 1687D00 AMH 20104 3990006 28 28 1687D01 AMH 20104 3990006 28 28 1687E00 AMH 20104 3990006 28 28 487E01 AMH 20104 3990006 28 28 687F00 AMH 20104 3990006 28 28 1677F01 AMH 20104 3990006 28 28 07140 IBM 90233 3990006 17 17 638801 AMH 00001 3990006 18 18 33Number of DASD subsystems : 38Number of DASD volumes defined : 2,961Number of DASD volumes online : 2,867Amount of online storage (GB) : 6,285
134 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
GTDVTOCGTDVTOC first dynamically allocates and opens the VTOC, and then runs three or more I/O operations. Itthen creates the TDMF Device Characteristics Report.
You enter the GTDVTOC command as a batch job. The program uses the Unit Control Block scan service(UCBSCAN) to find the serial number of every online direct-access volume. Using the volume serialnumber, the program dynamically allocates and opens the VTOC. The program then runs three or moreI/O operations.
• The first operation reads the device characteristics.• The second operation checks if the volume is attached to a channel extender.• The third operation reads the first track of the VTOC.
If the VTOC has an index, subsequent I/O operations might continue to read tracks from the VTOC. Theyalso might continue to locate the index, and then to read the index itself, to analyze the VPSM records.
The TDMF Device Characteristics Report program extracts device information for each online device in theDASD configuration, then outputs either a full or exception report. These reports help you to easilyidentify those devices whose Volume Table of Contents (VTOC) does not accurately describe theirphysical characteristics. It also helps identify which devices might be eligible for invocation of dynamicICKDSF in a TDMF session.
Information from the UCB, the device characteristics data and the Format 4 DSCB are formatted on thereport line. Three flags also appear for each device, of which two are considered to indicate an errorcondition on the device.
This example shows the JCL required to run GTDVTOC program.
//<---------------- job card ----------------->//VTOC EXEC PGM=GTDVTOC,PARM=’parm1,parm2’//STEPLIB DD DISP=SHR,DSN=<program library>//SYSPRINT DD SYSOUT=<sysout class>//SYSUDUMP DD SYSOUT=<sysout class>/*
Figure 32. GTDVTOC JCL
The allowable parameters are as follows.
Table 33. GTDVTOC parameters
BRIEF Indicates that only two types devices are reported: devices with a mismatchbetween the VTOC and the physical device characteristics, and devices for whichGTDVTOC encounters an error when the VTOC opens. The default is that the statusof every online volume is reported.
LINECT=nnn Controls how many pages are skipped in the output report. The minimum value forthe number of lines on the page (including headings) is 25 and the maximum is999. The default value for LINECT is 55.
GTDVTOC - printed outputFor each device, information displays in the GTDVTOC output file that is defined by the SYSPRINT DDstatement.
The following information is contained in the GTDVTOC printed output file:
Table 34. GTDVTOC device level output
Unit The four-digit device number. Only devices that are online and not flagged as "changingstatus" (pending offline) are selected.
Volser The volume serial number.
Chapter 7. Using batch utilities 135
Table 34. GTDVTOC device level output (continued)
Primary Cyls -Device
The number of primary cylinders that are returned in the Read DeviceCharacteristics data.
Primary Cyls -VTOC
The number of primary cylinders that are calculated from fields in the Format 4 DSCB. Ifthe DS4DEVAC field (number of alternative cylinders) is valid, the value is subtracted fromDS4DEVSZ (the number of logical cylinders). In older VTOCs, the number of alternativecylinders is calculated by dividing DS4NOATK (the number of available alternative tracks)by 15 and rounding up.
Alt. Tracks -Device
The number of alternative tracks that are returned in the Read DeviceCharacteristics data.
Alt. Tracks - VTOC The number of available alternative tracks from the VTOC, taken from field DS4NOATK.This column contains both values that are separated by a slash if the format for DSCB alsocontains the number of alternative cylinders, and a mismatch exists between DS4NOATKand DS4DEVAC multiplied by 15.
DS4DEVAV set Blank, or a flag that indicates the VTOC contains a valid DS4DEVAC.
Primary CylinderMismatch1
Blank, or a flag to draw attention to the fact that the values in fields, Primary Cyls - Deviceand Primary Cyls - VTOC are different.
Alternative TrackMismatch1
Blank, or a flag that indicates the calculated number of alternative tracks does not matchthe value that is returned in the Read Device Characteristics data.
Note: The numbers of alternative tracks are not compared, and this field is blank if thenumbers of primary cylinders are unequal.
XRC Blank or a flag that indicates "+" for XRC supported or "X" for XRC active on the volume.
PPRC Blank or a flag that indicates "+" for PPRC supported or "X" for PPRC active on the volume.
Channel Extender Blank or a flag that indicates the volume is attached to a channel extender
Other Messages1 Blank, or a message that reports an index error or explaining why no counters or flags arereported for the device.
1 An entry is listed if any of the last three items in the table are non-blank for a device, even if the "BRIEF"parameter was specified.
Messages that might appear in the Other Messages field are as follows.
Table 35. GTDVTOC messages
Volume has non-standard VTOC
The pointer to the Volume Table of Contents in the UCB shows that the VTOC does notstart at the beginning of a track. So, GTDVTOC does not attempt to allocate to the volume.This volume might be a VM volume.
Device has nooperational paths
The online device is unusable.
Device is TDMFtarget
The device is the target of a TDMF volume migration. It would be impossible to OPEN theVTOC, as TDMF does not allow access to the volume.
OPENJ AbendSxxx-rr
The DCB Abend exit was called during OPEN TYPE=J processing for the VTOC or VTOCindex. The system completion and reason codes that are encountered by OPEN areformatted into the message. GTDVTOC will de-allocate the volume and continue with thenext device.
OPENJ timeout More than 5 seconds elapsed during run of the OPEN TYPE=J. The cause might be thatanother system reserved the device, or a hardware/pathing error. GTDVTOC willdeallocate the volume and continue with the next device.
136 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Table 35. GTDVTOC messages (continued)
I/O Error (RDC).Sense - ‘ssss’
The VTOC was opened successfully, but the control unit rejected the Read DeviceCharacteristics request. The contents of the first two sense bytes are displayed in themessage.
I/O Error (Read).Sense - ‘ssss’
The VTOC was opened successfully and the device characteristics were obtained from thecontrol unit. The I/O operation to read a track from either the VTOC or its index failed. Thecontents of the first two sense bytes are displayed in the message.
VTOC format error An invalid DSCB was encountered during the read of the VTOC
Unable to openVTOC index
An OPENJ of the VTOC index whose name was found in the VTOC itself was unsuccessful.
VTOC index formaterror
An invalid VIR record was found in the index, or the number of tracks mapped by the VPSMrecords was not a multiple of 15.
Index size error -nn,nnn Cyls
The number of logical cylinders that are described in the index’s VPSM records (displayedin this message) differs from the value found in the format four DSCB.
XRC Session ID =session name
An XRC session is active for this volume on this system.
XRC Session An XRC session is active for this volume on a different system.
GTDVTOC - return codes and abend codesAll GTDVTOC Return Codes and ABEND Codes are documented in IBM Transparent Data Migration Facility(TDMF) for z/OS Messages and Codes (SC27-8964-00).
GZDRECONThe GZDRECON REXX procedure is used to reconstruct data sets packaged by the GZDTERSE procedure.
The advantages of XMIT over other transmission formats are listed below.
1. Multiple data sets are passed in a single file.2. Errors are detected at the logical data set and file levels before the creation of new data sets or
modifying existing data sets.
Take these steps to restore files with GZDRECON.3. If the data file is received as a self-extracting compressed file on the PC, run the data file on the PC to
create one or more upload files.4. Upload the xxxxxxxx_DATA file, by using BINARY transmission, to a sequential data set. The data set
might be RECFM=F, FS, FB, FBS, V, or VB, with an LRECL up through 32760. The preferred attributesare RECFM=FB, BLOCK=27998, and LRECL=27998.
5. Run a job by using GZDRECON in a TSO BATCH environment to re-create the original data sets. See thefollowing examples. If you have any problems, or encounter error messages that you don'tunderstand, contact the technical support for assistance.
Chapter 7. Using batch utilities 137
SyntaxAn example of a syntax diagram for GZDRECON.
OptionsGZDRECON provides the options listed here.ABOUT
Specifies that information regarding the EXEC is to be displayed. All other parameters will be ignored.This is optional.
HELP|?Specifies that help text for the EXEC will be displayed. All other parameters will be ignored. This is thedefault.
indsnSpecifies the fully qualified input data set name. This is required. If not specified, the help text for theEXEC will be displayed.
outhlq
• outdsn specifies the high level qualifier(s) to be used for output data sets, or the output data setname if the input data set contains a single data set transmission (identified by an OPTIONSSINGLE DSNAME statement in the input data set). This is optional.
If outhlq is not specified, outhlq defaults to the high level qualifier(s) specified by indsn, with theuser's TSO PREFIX replacing the first high level qualifier.
For the case of outdsn not being specified, the outhlq rules will apply.
DATACLASSpecifies the SMS DATACLAS to be used for the allocation of data sets. It is optional.
DSNTYPESpecifies the data set type as PDS or LIBRARY (PDS/E) for partitioned data sets that are to beunloaded. This is optional. If not specified, dsntype defaults to PDS.
UNITSpecifies the unit name to be used for the allocation of data sets. This is optional.
VOLSERSpecifies the volser to be used for the allocation of data sets. This is optional.
MGMTCLASSpecifies the SMS MGMTCLAS to be used for the allocation of data sets. This is optional.
REALLOCATE(NO|YES)Specifies if the target data sets may be deleted and reallocated. Use of the REALLOCATE(YES) optionis only recommended for recovery following an allocation failure. This is optional. The default isREALLOCATE(NO).
STORCLASSpecifies the SMS STORCLAS to be used for the allocation of data sets. This is optional.
WORKUNITSpecifies the unit name to be used for the allocation of work data sets. This is optional.
138 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
UsageBatch invocation of the EXEC is preferred.
The outline for the required JCL is as follows:
//jobname JOB//STEP1 EXEC PGM=IKJEFT01,DYNAMNBR=20//SYSEXEC DD DSNAME=execlibrarydsname,DISP=SHR//SYSTSPRT DD SYSOUT=*//SYSTSIN DD *%GZDRECON input.dsname/*//
ExamplesFive sample job examples.
Example 1
A sample job to receive the files for the COFFEE product, Version 2 Release 8.0 (V2080), transmitted infile MY.COFFEE.V2080.DATA, as MY.COFFEE.V2080.xxxxx would be:
//MYINST JOB//STEP1 EXEC PGM=IKJEFT01,DYNAMNBR=20//SYSEXEC DD DSNAME=MY.LIBRARY,DISP=SHR//SYSTSPRT DD SYSOUT=*//SYSTSIN DD *%GZDRECON MY.COFFEE.V2080.DATA/*//
Example 2
A sample job to receive the files for the COFFEE product, Version 1 Release 4.0 (V1040), transmitted infile MY.COFFEE.V1040.DATA, as MYALT.COFFEE.V1040.xxxxx on volume ALTVOL would be:
//ALTINSTA JOB//STEP1 EXEC PGM=IKJEFT01,DYNAMNBR=20//SYSEXEC DD DSNAME=MY.LIBRARY,DISP=SHR//SYSTSPRT DD SYSOUT=*//SYSTSIN DD *PROFILE PREFIX(MYALT)%GZDRECON MY.COFFEE.V1040.DATA - VOLUME(ALTVOL)/*//
Example 3
Another version of Example 2, but without using the TSO PROFILE PREFIX command:
//ALTINSTB JOB//STEP1 EXEC PGM=IKJEFT01,DYNAMNBR=20//SYSEXEC DD DSNAME=MY.LIBRARY,DISP=SHR//SYSTSPRT DD SYSOUT=*//SYSTSIN DD *%GZDRECON MY.COFFEE.V1040.DATA - MYALT.COFFEE - VOLUME(ALTVOL)/*//
Example 4
A sample job to receive the files for the COFFEE product, Version 1 Release 23.4 (V1234), transmitted infile MY.COFFEE.V1234.DATA, as YOUR.CAFFEINE.V1234.xxxxx would be:
//YOURINST JOB//STEP1 EXEC PGM=IKJEFT01,DYNAMNBR=20//SYSEXEC DD DSNAME=MY.LIBRARY,DISP=SHR//SYSTSPRT DD SYSOUT=*
Chapter 7. Using batch utilities 139
//SYSTSIN DD *%GZDRECON MY.COFFEE.V1234.DATA - YOUR.CAFFEINE/*//
Example 5
A sample job to receive a single patch file for the COFFEE product, transmitted in fileMY.COFFEE.MOCHA.DATA, as YOUR.NEXT.CAFFEINE.FIX would be:
//YOURFIX JOB//STEP1 EXEC PGM=IKJEFT01,DYNAMNBR=20//SYSEXEC DD DSNAME=MY.LIBRARY,DISP=SHR//SYSTSPRT DD SYSOUT=*//SYSTSIN DD *%GZDRECON MY.COFFEE.MOCHA.DATA - YOUR.NEXT.CAFFEINE.FIX/*//
ProcessingThe GZDRECON REXX procedure processing is as follows.
1. Verifies the EXEC to ensure that no conversion error occurred during transmission.2. Allocates the data sets with the recommended space as listed for the products, along with temporary
data sets as required.3. Transfers the file input data to temporary data sets and the appropriate program is run to restore the
files.
NotesList of notes for the GZDRECON procedure.
• No user interaction with the EXEC is expected or required.• Region size for the job step might be estimated as:
2n+4 MB
where n is the size of the distribution file in MB.• The GZDRECON temporary data set size is allocated based on the number of records of the largest
transmitted file that is contained within the stacked transmission file that is specified by indsn.• Input lines for SYSTSIN might be continued on subsequent lines by specifying the TSO continuation
character (normally a hyphen or dash).• When REALLOCATE(YES) is specified, data sets with unexpired expiration dates are not deleted.• For greater speed, GZDRECON might be compiled with the REXX Compiler. In compiling GZDRECON,
the SLINE compiler option must be specified and the compilation completes with RC=4. Ignore all"FANGAO55W Unsupported TRACE options will default to OFF" messages.
• GZDRECON uses AMATERSE or TRSMAIN. A STEPLIB DD statement is required in the TSOBATCH JCL ifneither program is in LNKLST. Further information about TRSMAIN is available at http://techsupport.services.ibm.com/390/trsmain.html.
• AMATERSE and TRSMAIN allocate a temporary work data set that normally does not exceed 1.08 timesthe size of each target data set.
140 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Chapter 8. Viewing migration informationThe TDMF TSO Monitor is installed along with TDMF. The TSO Monitor can be used to view active or pastmigrations.
OverviewThis topic presents overview information for the TSO Monitor including a listing of the TSO Monitor menuoptions.
The TDMF TSO Monitor is a feature that is installed at the same time as the TDMF function. The TDMF TSOMonitor feature consists of REXX execs, which require ISPF Version 3.6 and TSO/E Version 2.4 or later.The TSO Monitor can be used to view active or past migrations.
After the TSO Monitor (EXEC ‘hlq.IBM.HGTD590.SGTDELIB(GTDTDMF)’) opens, it provides completecontrol over TDMF sessions.
Figure 33 on page 141 shows the options available on the TDMF Monitor menu.
Figure 33. TDMF Monitor menu
Note: The keylist function must be enabled in order for the TDMF TSO Monitor to work properly. Failure toenable this function results in the message "TDMF keylist not active" with a tone when you areusing the TSO Monitor. If this message occurs, see “Migration sessions with a mixture of maintenancelevels” on page 21 to correct the situation.
Tip: The ISPF settings for V5.7.0 are saved in a separate member. So if you go back to previous versions,your old settings are active for those versions.
If function key values (PF keys) are displayed on the bottom of a help window, additional informationmight be hidden. To prevent the PF keys from displaying, enter PFSHOW (which toggles) or FKA OFF onthe ISPF command line. As an alternative, press the Resize key (PF4) to expand the display so thatadditional information is visible.
As a result of running the TDMF TSO Monitor, your ISPF profile data set might be updated. See “Migrationsessions with a mixture of maintenance levels” on page 21 for a description of members that might beadded.
© Copyright IBM Corp. 2010, 2019 141
Help windows are available for all actions within the TDMF TSO Monitor. Some windows have multiplehelp panes dependent upon the location of the cursor.
If the TDMF TSO Monitor is run without the proper security settings, or if the security environment wasnot defined, the following window is displayed.
TDMF V e r s i o n 5Copyright (C) Teracloud S.A. 2019-2020 and IBM Corporation 1996-2020 There is a problem with the installation's Security environment. The problem is most likely one of the following conditions: 1. This user doesn't have authority to read the security record. 2. The security environment hasn't been defined. Run GTDSOPTN Press Enter to terminate.
Figure 34. TDMF security warning window
Correct the security settings and rerun the REXX exec to start the monitor.
The following figure is an example of the TDMF Monitor menu.
TDMF Monitor menu Option ===> 0 Change or Submit Replication or Migration Jobs 1 Current Sessions: Monitor Progress 2 Current Sessions: User Interaction and Status 3 Current Sessions: Display Messages 4 Current Sessions: Display Offline Volume Access (OVA) 5 Current/Past Sessions: Display Performance Data 6 Past Sessions: Display Summary 7 Past Sessions: Display Details 8 Past Sessions: Display Communication Data Set History 9 Display Installation Options and Environment 10 Display/Modify Installation Security Environment 11 Display/Modify User's TSO Monitor Options 12 Build Replication or Migration Jobs U Support Utilities H Help and Message Detail Facility Option ===> F1=Help F3=Exit F12=Cancel
Figure 35. TDMF Monitor menu
From this window it is possible to create, submit, monitor, or end volume and session migrations.
Option 0 - Change or submit replication or migration jobsUse TDMF TSO Monitor option 0 to create, change, or submit data migration sessions.
You can use this option to create, change, and submit data migration sessions without leaving the TDMFTSO Monitor. The samples that are provided include jobs to allocate the COMMDS, a Master system batchjob, an Agent sample batch job, and the GTDSOPTN batch job.
The following figure shows the first panel, which is a brief overview of the jobs that are contained withinthis option.
Change/Submit Jobs PanelWhen you press the enter key from this screen a data set containing acollection of job templates used to perform various functions.You may browse, edit, submit and/or change any member as long as itis not being edited by another user. You can select a new member andcopy the desired member into it as you may do in any edit session.The reason this capability is provided is to enable you to accomplishall tasks from within the Monitor. This is not however, a requirement.Press Enter to continueCommand ===> F1=Help F2=Split F3=Exit F9=Swap F12=Cancel
Figure 36. Change/Submit Jobs Panel overview
142 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
The following figure shows the second panel, which displays the sample jobs necessary to create, change,and submit a TDMF session. Review “Preparing to run TDMF” on page 26 for a description of each job andthe actions that must be completed before the jobs can be submitted.
Menu Functions Utilities Help---------------------------------------------------------------------------EDIT FBL00.IBM.HGTD590.GTDSAMP Row 00001 of 00032Command ===> Scroll ===> PAGE Name Prompt Size Created Changed ID. GTDAGENT. GTDCSMP. GTDALTGT. GTDALCM. GTDCFTP. GTDDDDEF. GTDDOVAE. GTDAHIST. GTDINCSI. GTDINSTL. GTDCLAGN. GTDCLMST. GTDRMAGN. GTDRMMST. GTDJCARD. GTDMNUAL. GTDMASTJ F1=Help F2=Split F3=Exit F5=Rfind F7=Up F8=Down F9=SwapF10=Left F11=Right F12=Cancel
Figure 37. Sample batch jobs
Option 1 - Current sessions: monitor progressUse TDMF TSO Monitor option 1 to watch active migration sessions.
You can use this option to watch active migration sessions in progress. A bar chart is displayed with anarrow ( --> ) to indicate how far along a specific volume migration is in the copy phase and then whatphase each volume pairing is in. Each point represents 2% of a volume.
For volumes that take several minutes to complete the copy phase, an estimate of the remaining time isdisplayed to the right of the percentage complete bar. This value is only displayed if the copy phase is notyet at 80% complete and there are more than 30 seconds remaining.
Tip: Press PF10 to transfer the display to Option 2 – Current Sessions: User Interaction and Status.
In the following figure, volume QAS100 completed migration, while volumes QATE03 and QAS101 are inthe copy phase, and volume QAS102 is inactive. To determine why volume AME794 terminated, pressPF10 for status and enter M.
Sessions Monitor Start CONFIRM NeededTDMF Master V5.4.0 Session Active.ComDataSet : HLQ.GTD550.SYSCOM3 Source Migration Percent Complete --------->Group VolSer Phase ..10...20...30...40...50...60...70...80...90..100 PROD9A Waiting OVATDMF Master V5.4.0 Session Active.ComDataSet : HLQ.GTD550.SYSCOM3 Source Migration Percent Complete --------->Group VolSer Phase ..10...20...30...40...50...60...70...80...90..100 QATE03 Copy ---------> 0:10:28 QAS100 Complete QAS101 Copy -----------------------------------------> QAS102 Inactive AME794 Terminated Command ===> Scroll ===> CSR F1=Help F2=Split F3=Exit F5=Rfind F7=Back F8=Forward F9=Swap F10=Status F12=Cancel
Figure 38. Sessions Monitor panel
Chapter 8. Viewing migration information 143
Note: Depending on how the TSO session is configured, the panel can be displayed in normal (80-character) line display or wide-screen (160-character) line display. Users with 160-character displays seesomething similar to the 80-character line display. Detail display is the default.
The SUMMARY option includes an 80-character summary display. Place the cursor on the ComDsn lineand press Enter to toggle between Detail and Summary. The toggle can be different for ALL ComDsndisplay and retains this setting for the life of the TSO Monitor session.
The Summary Display wide screen provides information for all active sessions on two lines. Variouscounts are provided to assist the user in knowing the status of the active TDMF sessions.
RefrDisplays the count of volumes in the refresh phase.
CopyDisplays the count of volumes in the copy phase.
InactDisplays the count of volumes that are inactive.
WaitDisplays the count of volumes that are waiting for a user response.
SuspDisplays the count of volumes that are suspended.
CompDisplays the count of volumes that completed migration.
TermDisplays the count of volumes that were terminated. Examine the message log to determine thereason for termination.
Number of agent systems
Highest estimated time to completion
Number of volume pairs in the session
Note: You can use TSO Monitor option 11 to set your default. Each SYSCOM can have its own SUMMARY/DETAIL property and they stay set while the TSO Monitor is active. (If you exit the TSO Monitorcompletely, the next restart opens all of the displays according to the defaults set through option 11).
In the following figure, the SDM identifier on the last line indicates that the session is using the SystemData Mover facility. It is blank in all other cases.
Session Monitor Row 1 to 5 of 5Command ===> Scroll ===> CSRTDMF Master V5.2.0 Session Active.ComDataSet : HLQ.GTD550.SYSCOM Source Migration Percent Complete --------->Group VolSer Phase ..10...20...30...40...50...60...70...80...90..100 GFG100 Suspended ----------------------> SDM
Figure 39. Session Monitor panel with SDM
Filter volume pair selectionUse the FILTER command to display only volumes with particular characteristics.
You can use filters to reduce the number of volume pairs that are displayed to more easily work with largegroups. When you enter the FILTER command, the following panel is displayed.
144 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Enter filters to be applied to the table display. Source volume: Target volume: Source UCB : Target UCB : Volume status: PF3 to apply filters to the table.
Figure 40. Filter menu
After you complete the filter criteria, press the PF3 key. The next panel displays only volume pairs thatmatch the criteria that you selected. Filtering remains in effect until you return to the main panel or untilall filter fields are blank.
The following table lists the volume pair values that can be specified with the FILTER command.
Table 36. Volume pair filter criteria
Criteria Description
Source Volser The 6-character source volume serial number, orany leading portion of it.
Target Volser The 6-character target volume serial number, orany leading portion of it.
Source UCB The 4-character source volume serial number, orany leading portion of it.
Target UCB The 4-character target volume serial number, orany leading portion of it.
Volume status A 13-character field that matches character bycharacter with the volume status. For example:COPY, UNINIT, TERM, and so on.
You can use the percentage sign (%) as a wildcard to mask one character. For example, A%%300 matchesall values that start with A, have 2 characters, and end with 300. ABC matches all values that start withABC. Do not use an asterisk (*) at the end of the criteria.
Option 2- Current sessions: user interaction and statusUse TDMF TSO Monitor option 2 to end a specific volume migration.
Use this option to end a specific volume migration or an entire group of migrations, respond to a promptfor migrations, change the synchronization goal parameter, confirm migration activation, and suspend orcontinue a volume or group. You can set a limit on the number of concurrently active migration volumes ormodify a limit that was specified on the SESSION control statement. The maximum permitted value forthe number of concurrent volumes is the total number of volume pairs in the session, and the minimum isthe size of the largest volume group, if the active only in copy option is not in effect - or one.
Press the Next key (PF4) to skip from one active TDMF session to another.
You can use the FILTER command to display only volume pairs that meet certain criteria (see “Filtervolume pair selection” on page 144).
The following table describes each field of the panel.
Chapter 8. Viewing migration information 145
Table 37. Option 2 - Current Sessions: User Interaction and Status
Field Description
Sessions The number of active TDMF sessions
Number of volumesMigrating
The number of volumes that are migrating in this session
Number of volumescomplete
The number of volumes that completed the migration process in thissession
Number of concurrentvolumes
The number of concurrent tasks in a single session (see “Orderinginput of control statements” on page 36).
Number of volumes waiting The number of volumes that are waiting for a response (prompt orconfirmation) or initialization. This field cannot be used todynamically change the number of concurrent volume migrations bytyping over the current number displayed.
Requested Action Action that is needed for a volume or group. See the help panel thatfollows the session status (see “User interaction and status - helppanel” on page 148).
Volume Serial The source and target volume serial numbers
Device Number The source and target device number on the master system
Group Name The group name that is provided for a multi-volume group
Migration Status The current phase of volume migration.
Migration Type,Synchronization Type
Migration type is either point-in-time (PIT) or swap. Syncrhonizationtype is either prompt or automatic
Error Information/ SystemMessage
Error or system messages for the volume migration. Errors andsevere error messages are displayed. The system indicates whichLPAR determined the error. If no messages are displayed, then themigration completed with no errors, however warning messagesmight have been issued.
Sync Goal The synchronization goal for the volume or group migration
User interaction and status - informationPress PF4 to display information about other active sessions.
In the following figure, a six-volume session is in progress. Of those six volumes, two are in a groupnamed GROUP1; one volume is not part of a group; three are in a group that is named GROUP2, wherevolume AME794 ended migration on system TDM1 with message GTD3537E.
146 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Sessions Status Start CONFIRM NeededTDMF Master V5.4.0 Session Active.ComDataSet : GTD550.V.SYSCOM Sessions 01Number of volumes migrating . : 06 Number of concurrent volumes 06Number of volumes complete . : 01 Number of volumes waiting 00Requested Volume Device Group --- Migration --- - Error Info - Sync Action Serial Number Name Status Type System Message Goal __ AMEE95 B081 GROUP1 Copy Swap A 007 AME741 B095 __ AMEE92 B082 GROUP1 Copy Swap A 007 AME742 B092 __ SCB083 B083 GROUP2 Copy PIT A 007 AME793 B093 __ AME744 B084 GROUP2 Copy PIT A 007 AMEE94 B094 __ AME794 B087 GROUP2 Terminated PIT A TDM1 GTD3537E 007 AME791 B091 __ PROD9A 879A Waiting OVA PIT P 010 PROD9B 879B
Figure 41. User Interaction and Status panel
You can use option H.1 to display an explanation of message GTD3537E (see “Option H.1 – Display TDMFmessage details” on page 188), or select “Terminate a volume” on page 151 to display the message.
In the following figure, the SDM identifier on the last line indicates that the session is using the SystemData Mover facility and the 000 identifier below it indicates the owning system. It is blank in all othercases.
Session Status Row 1 to 1 of 1Command ===> Scroll ===> CSRSystem: Master TDMF for z/OS Version: 5.2.0ComDataSet HLQ.GTD550.SYSCOM Sessions 01Number of volumes migrating : 001 Number of concurrent volumes 001Number of volumes complete : 000 Number of volumes waiting 000Requested Volume Device Group --- Migration --- - Error Info - Sync Action Serial Number Name Status Type System Message Goal __ GFG100 5565 Suspended Swap P 005 SDM GFG103 5165 000
Figure 42. User interaction and status panel with SDM
User interaction and status - migration optionsPress PF4 to display migration options for other active sessions.
The following figure shows all options for each volume in the active session. These options can be set as asystem default by using the GTDSOPTN batch job. You can also override them with the OPTIONS keywordon the SESSION, GROUP, REPLICATE, or MIGRATE control cards that are used in the MASTER systembatch job.
Chapter 8. Viewing migration information 147
Session Status Migration OptionsSystem: Master TDMF for z/OS Version: 5.4.0ComDataSet GTD550.V.SYSCOM Sessions 01Number of volumes migrating : 006 Number of concurrent volumes 001Number of volumes complete : 000 Number of volumes waiting 001Requested Volume Device Group --- Migration --- - Error Info - Sync Action Serial Number Name Status Type System Message Goal __ VDMF80 ActCopy = N PACING = Y AUTOOPER = N TIME =LOCAL VDMF81 CONFIRM = N PURGE = N TERMGRP = N COMPARE = N OVA = N PPIT = N FASTCOPY = N NonPPRC = N __ VDMF82 ActCopy = N PACING = Y AUTOOPER = N TIME =LOCAL VDMF83 CONFIRM = N PURGE = N TERMGRP = N COMPARE = N OVA = N PPIT = N FASTCOPY = N NonPPRC = N __ AMEE95 ActCopy = N PACING = Y AUTOOPER = N TIME =LOCAL AME741 CONFIRM = N PURGE = N TERMGRP = N COMPARE = N OVA = N PPIT = N FASTCOPY = N NonPPRC = N __ AMEE92 ActCopy = N PACING = Y AUTOOPER = N TIME =LOCAL AME742 CONFIRM = N PURGE = N TERMGRP = N COMPARE = N OVA = N PPIT = N FASTCOPY = N NonPPRC = N __ SCB083 ActCopy = N PACING = Y AUTOOPER = N TIME =LOCAL AME793 CONFIRM = N PURGE = N TERMGRP = N COMPARE = N OVA = N PPIT = N FASTCOPY = N NonPPRC = N __ AME744 ActCopy = N PACING = Y AUTOOPER = N TIME = LOCAL AMEE94 CONFIRM = N PURGE = N TERMGRP = N COMPARE = N OVA = N PPIT = N FASTCOPY = N NonPPRC = N
Figure 43. User Interaction and Status panel – migration options
User interaction and status - help panelA help panel for User Interaction and Status is available. To access this panel, tab to a volume pair underthe heading Requested Action and press PF1.
The following figure shows an example of a portion of the help panel.
IBM Transparent Data Migration Facility (TDMF) for z/OSPF3=End PF7=Page Up PF8=Page Down Following valid on all Systems Command Display Migration Messages : M Display Migration Performance Data : D Display Volume's Offline Volume Access (OVA) Job(s) : X Following valid only on MASTER System Command Acknowledge Mirror Change on Volume Swap : A Continue Volume Migration or its Group : C or CG Confirm PPiT Recycle for Volume or its Group : R or RG Confirm Start Volume Migration or its Group : F or FG Confirm Synchronization for Volume or its Group : P or PG Disallow OVA registrations for Volume or its Group : H or HG Discontinue PPiT migrations at "Recycle" prompt : E or EG Reinitialize Migration of terminated Volume : I Set Goal Synchronization Goal for Volume or its Group : Z or ZG Suspend Volume Migration or its Group : S or SG Terminate Volume Migration or its Group : T or TG PF4 will display the next active session. PF6 is a toggle key for displaying the Session options. PF10 will immediately display the Session Monitor Panel. PF11 is a toggle key for filtering the volumes.
Figure 44. User Interaction and Status – help panel
Enter M on the Requested Action area of a volume pair to display all messages for that specific pairing(see “Terminate a volume” on page 151).
Enter D to display the performance data for that migration (see “System information” on page 154through “Average device response times” on page 160).
The following panels depict other options available to the Master System noted in the panel.
Note: Commands that are entered on this panel for a volume or group are usually presented to therelevant session immediately. However, if there are TDMF sessions active at different maintenance levels,the command might not be processed by the TDMF session for up to 30 seconds.
148 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Start a volumeIf volume confirmation was selected through the system defaults (GTDSOPTN batch job) or by using theOPTIONS keyword on an input control statement, a response must be received by TDMF on either theVolume Confirmation panel or the MVS system console, if auto-operations were selected.
Under Requested Action for the volume pair or volume group, enter either F for volume pair or FG forvolume group and press the Enter key. The following panel is displayed.
Volume Confirmation Screen Row 1 to 1 of 1ComDataSet: . . . . : GTD550.V.SYSCOMConfirmation required to start volume.Volume Serial No. . : SPMS86Confirm? (YES/NO): . . ___Command ===> F1=Help F2=Split F3=Exit F9=Swap F12=Cancel
Figure 45. Start volume display
Synchronize a volumeIf the prompt option was requested for the volume or volume group, the message SYNC VolumeNeeded is displayed in the upper right corner of any active session window of the TDMF TSO Monitor, oron the MVS console if auto-operations is selected. Additionally, an asterisk (*) is displayed next to thevolume pairing or group that requires the response.
Note: In a group of volumes, the message is not displayed until all of the volumes within the group reachthe synchronization ready point.
In the following figure, the SYNC Volume Needed message is displayed and all volumes reached thesynchronization ready point. All of the volumes are in the refresh phase.
Session Status SYNC Volume Needed.System: Master TDMF Version: 5.4.0ComDataSet : GTD550.V.SYSCOM Source Migration Percent Complete --------->Group VolSer Phase ..10...20...30...40...50...60...70...80...90..100 DMC87 Refresh 29 -----------------------------------------------> TD9C2J Refresh 1 -----------------------------------------------> TD8E01 Refresh 34 -----------------------------------------------> SPMS86 Refresh 59 ----------------------------------------------->Command ===> ___________________________________________________________ F1=Help F2=Split F3=Exit F5=Rfind F7=Back F8=Forward F9=Swap F10=Status F12=Cancel
Figure 46. Sync volume needed
Press PF10 or select Option 2 from the primary window to display the following window and respond tothe prompt.
Session Status SYNC Volume Needed. System: Master TDMF Version: 5.4.0 ComDataSet : GTD550.V.SYSCOM Sessions 02 Number of volumes migrating . : 04 Number of concurrent volumes 04 Number of volumes complete . : 00 Number of volumes waiting 00 Requested Volume Device Group --- Migration --- - Error Info - Sync Action Serial Number Name Status Type System Message Goal __ *AME780 B080 GROUP01 Waiting Sync PIT P 007 AMEB81 B090 __ AMEA81 B081 GROUP01 Refresh 2 PIT P 007 AME791 B091 __ AMEB82 B082 GROUP01 Refresh 2 PIT P 007 AME79E B09E __ SCB083 B083 GROUP01 Refresh 2 PIT P 007 AME793 B093 Command ===> Scroll ===> CSR F1=Help F2=Split F3=Exit F4=Next F5=Rfind F6=Options F7=Back F8=Forward F9=Swap F10=Monitor F12=Cancel
Figure 47. Sync volume needed – session status
Chapter 8. Viewing migration information 149
In this example, the volume group reached the synchronization ready point and the prompt was created.Note the asterisk (*) at the first volume pairing of GROUP01. Only one response is necessary for the entiregroup to go into the synchronization phase.
To suspend a group and then later to continue the group while the first volume pairing is still in WaitingSync status, enter the SG and CG group line commands on any volume pairing other than the first one. Ifyou enter SG or CG on the first volume pairing while the Waiting Sync status is displayed, the message Aprompt is outstanding is displayed in the upper right corner.
Confirm synchronization (pairing)When you enter either P or PG under Requested Action for the first pairing of that volume group, theGroup Confirmation panel is displayed.
The following figure shows an example of the synchronization confirmation panel.
Note: This confirmation panel is displayed for both point-in-time and swap migrations, if the promptoption is specified in the control records.
Group Confirmation Screen Row 1 to 1 of 1ComDataSet: . . . . : GTD550.V.SYSCOMConfirm Synchronization for Group Migration.Group . . . . . . . : GROUP01Confirm? (YES/NO) . . ___Command ===> F1=Help F2=Split F3=Exit F9=Swap F12=Cancel
Figure 48. Sync volume confirmation display
Acknowledge mirror change on swapIf a volume that is being migrated is the primary in a PPRC/SRDF/TrueCopy session and the migrationtarget is mirrored with a different hardware technique, it is necessary to select the ALLOWmirrorchangeoption for the volume. This option can be specified on the SESSION, GROUP, or MIGRATE controlstatements and usually requires that the user either allow or acknowledge this situation before thevolume can be selected and activated.
This user intervention is not required if the option was supplied with the NOACKnowledge parameter. Ifthe option is specified but is not applicable, it does not cause the volume selection to be delayed.
Enter A for the volume pairing under Requested Action to display the Volume Confirmation panel, asshown in the following figure. Confirmation requires a YES or NO response. Press the PF3 key to exit thispanel without action.
Volume Confirmation Screen Row 1 to 1 of 1ComDataSet: . . . . : GTD550.V.SYSCOMAcknowledge Mirror Change on SwapVolume Serial No. . : SPMS86Confirm? (YES/NO): . . ___Command ===> F1=Help F2=Split F3=Exit F9=Swap F12=Cancel
Figure 49. Acknowledge mirror change on swap
Set volume synchronize goalYou can set a volume synchronization goal.
The TDMF default for volume synchronization is five seconds. You can change this value dynamically byentering either Z for a specific volume or ZG for a volume group under the Request Action and pressingthe Enter key.
The following figure shows an example of the Set Synchronize Goal confirmation panel.
150 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Set Volume Synchronize Goal Row 1 to 1 of 1ComDataSet: . . . . : GTD550.V.SYSCOMConfirmation required to set Sync Goal.Synchronization Volume Serial No. . . : AMEB81Synchronization Time . . . . . . . . . . ___Confirm? (YES/NO): . . . . . . . . . . . ___Command ===> F1=Help F2=Split F3=Exit F9=Swap F12=Cancel
Figure 50. Set Synchronize Goal panel
Suspend a volumeYou can use the SUSPEND command to suspend a volume pairing. Enter S for a specific volume or SG forthe entire group to stop the migration for a volume or volume group. These commands are valid during thecopy and refresh phases of a volume migration. TDMF monitors the source volume for updates until theCONTINUE command is issued.
If the SUSPEND command is issued during the copy phase, the volume migration is suspended at the startof the next I/O operation.
If the suspend command is issued during the refresh phase, and no updates to the source volume areoccurring, the command is queued until the next update I/O operation occurs. At that time, the volumemigration is suspended.
The following figure shows an example of the suspend confirmation panel:
Volume Confirmation Screen Row 1 to 1 of 1ComDataSet: . . . . : GTD550.V.SYSCOMConfirmation required to Suspend a Volume.Volume Serial No. . : TD9C2JConfirm? (YES/NO): . . ___Command ===> F1=Help F2=Split F3=Exit F9=Swap F12=Cancel
Figure 51. Suspend confirmation panel
Continue a volumeIf a volume migration process was suspended, you must issue the Continue command to complete themigration process.
Enter either C for a volume or CG for a volume group to continue migrating the volume or volume groupfrom where it left off. The following figure shows an example.
Volume Confirmation Screen Row 1 to 1 of 1ComDataSet: . . . . : GTD550.V.SYSCOMConfirmation required to continue a volume.Volume Serial No. . : TD9C2JConfirm? (YES/NO): . . ___Command ===> F1=Help F2=Split F3=Exit F9=Swap F12=Cancel
Figure 52. Continue volume panel
Terminate a volumeYou can use the TERMINATE command to end a TDMF volume pairing dynamically.
Under Requested Action, enter T to terminate a volume or TG to terminate a group and press the Enterkey. If the Terminate Group on Error option was selected, a selection of T defaults to TG. The followingfigure shows the confirmation panel.
Chapter 8. Viewing migration information 151
Volume Confirmation Screen Row 1 to 1 of 1ComDataSet: . . . . : GTD550.V.SYSCOMConfirmation required to Terminate a Volume.Volume Serial No. . : SPMS86Confirm? (YES/NO): . . ___Command ===> F1=Help F2=Split F3=Exit F9=Swap F12=Cancel
Figure 53. Confirm termination panel
Note: After TDMF enters Approval Mode, the TERMINATE command ends the migration with the originaltarget volume as the source volume. The TERMINATE command has the same effect as APPROVE[Yes].After the session is terminated, a new migration is required if you need to go back to the original sourcevolume.
Option 3 - Current sessions: display messagesUse TDMF TSO Monitor option 3 to display all messages for a specific session. The messages aredisplayed only for the system where the TDMF TSO Monitor is started.
The following figure shows an example of the Session Messages panel.
Session Messages Row 1 to 38 of 158 Command ===> Scroll ===> CSR ----------------------------------------------------------------------------------- TDMF Master V5.4.0 Session Active. ComDataSet : SWH10.TEST01.SYSCOM This is a Master system. Current Messages for this session are as follows: The current messages for SMFID TDM1 follow. 04/10/2009 02:16:05.142 GTD1732I The I/O Monitor module has been loaded into Dynamic LPA. 04/10/2009 02:16:05.163 GTD1728I Installed software for this system is: TDMF Base Level. 04/10/2009 02:16:05.186 GTD1279I The installation option to call ICKDSF for VTOC and index maintenance is set. 04/10/2009 02:16:05.186 GTD1278I The installation option to count volumes active only in their copy phase is set. 04/10/2009 02:16:05.186 GTD4152I The Synchronization Prompt option has been set. 04/10/2009 02:16:05.186 GTD1449I The Time-of-Day option was specified as LOCAL. 04/10/2009 02:16:05.186 GTD1453I The Pacing System Option was specified as OFF. 04/10/2009 02:16:05.186 GTD1465I The Operator Messages Option was specified as OFF. 04/10/2009 02:16:05.186 GTD1457I Confirmation on Initialization system option was specified as OFF. 04/10/2009 02:16:05.186 GTD1460I Terminate All Volumes in Group on Error Option was specified as ON. 04/10/2009 02:16:05.186 GTD4164I The Perpetual Point in Time option has been requested. 04/10/2009 02:16:05.186 GTD4165I The FASTCOPY option has been requested. 04/10/2009 02:16:05.236 GTD1380I The application program interface (API) for STK is not available.The current messages for SMFID TDM1 follow. 04/10/2009 02:16:05.187 GTD1509I A volume Pacing option was changed to ON by a Migrate card option.F1=Help F2=Split F3=Exit F5=Rfind F7=Back F8=ForwardF9=Swap F12=Cancel
Figure 54. Current Sessions: Display Messages panel
Note: This option supports a 160-character line display. Both wide screen (160) and normal (80) are thesame except for the size of the line.
Tip: You can use the Find and Repeat Find functions to search for a specific volume or message number.
152 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Option 4 - Current sessions: display offline volume access (OVA)Use TDMF TSO Monitor option 4 to display any TDMF offline volume access (OVA) batch jobs that arerunning on the same MVS system as the TDMF TSO Monitor.
If OVA was not specified in the TDMF Master session for a volume, the following panel is displayed.
Session Associated Address Spaces Row 1 to 6 of 6------------------------------------------------------------------------TDMF Version 5.4.0Communication Data Set: GTD550.V.SYSCOM Volume PROD9A migrating to volume PROD9B didn't specify OVA.--------------------------------------------------------------------------Command ===> Scroll ===> CSR F1=Help F2=Split F3=Exit F5=Rfind F7=Back F8=Forward F9=Swap F10=Status F12=Cancel
Figure 55. Current sessions: OVA not selected
If the OVA option was in effect for a volume migration but no OVA batch job was submitted for thevolume, the message within the panel displays as shown in the following figure.
Offline Volume Access (OVA) ** no sessions active **Command ===> Scroll ===> CSR------------------------------------------------------------------------ No OVA jobs are active on this system. TDMF. Master V5.4.0 Session Active. Communications Data Set: SWH10.PROD01.SYSTEMVolume VDMF92 I/O being redirected to VDMF93 This pair has no active Offline Volume Access (OVA).Volume VDMF94 I/O being redirected to VDMF95 This pair has no active Offline Volume Access (OVA).
Figure 56. Current sessions: OVA not active
After the OVA batch job is started, the following information is displayed, as shown in the following figure.
• The source volume I/O that is being redirected to the target volume• Total number of active OVA jobs on each system• Total number of redirected I/O requests on each system• Total number of updates that are redirected on each system• The jobname, program, and I/O count for the active OVA jobs
Session Associated Address Spaces Row 1 to 14 of 14------------------------------------------------------------------------- TDMF Version 5.4.0 Communication Data Set: GTD550.V.SYSCOM Volume PROD9A I/O being redirected to PROD9B (Device Number 879B) Copy valid at 04/30/2009 14:55:20.671 Local Time Total active OVAjobs on system TDM1 = 1 Total active OVAjobs on system TDM2 = 0 Total redirected I/O Requests on TDM1 = 56 Update redirected I/O Requests on TDM1 = 0 Currently active OVAjobs on TDM1. Jobname Program I/O Count GTD550IPG ADRDSSU 28Command ===> Scroll ===> CSR F1=Help F2=Split F3=Exit F5=Rfind F7=Back F8=Forward F9=Swap F10=Status F12=Cancel
Figure 57. Current sessions: display active OVA jobs
Note: This information is displayed while the active OVA jobs are running. After all OVA jobs arecompleted, the panel that is shown in “Terminate a volume” on page 151 is displayed.
Chapter 8. Viewing migration information 153
Option 5 - Display performance data of current and past sessionsUse TDMF TSO Monitor option 5 to view current or past performance data.
To view current or past performance data, select the appropriate COMMDS either by entering the data setname of a specific COMMDS or, if the History data set option was selected, selecting Option 8 from theTDMF Monitor menu window. If a session is active, the window defaults to the COMMDS currently in use.If the COMMDS that is displayed is not the COMMDS that you want, type over the current COMMDS withthe appropriate data set name and press the Enter key.
The information shows the amount of real storage that is used, volume pacing (if selected), the number ofpacing changes due to real storage constraints and I/O contention, device type, read and write statistics,elapsed time of each phase, and all messages that are created by TDMF. The information is presented asSYSTEM information and VOLUME information.
System informationThe System Information section of the Performance Data panel displays system performance data forcurrent and past sessions.
Table 38. Option 5 - Current/past sessions: system information
Type Type of Job, Master, or Agent System
SMFID The SMFID of each LPAR that each job is running on.
ASID The address space ID for each job.
System Number The relative system number.
The Master system is always relative system number 00.
Current Interval The length of time between TDMF polls
Heartbeat A timer (GMT) that increases for every current interval that the Master or Agentsystem is active on that specific system.
Expires In The amount of time a system is allowed to be inactive to other systems.
Fixed StorageFrames
The number of available fixed storage frames within the LPAR at system initializationthat can be dynamically adjusted if pacing is selected.
TDMF FixedStorage Frames
The number of fixed storage frames that TDMF used and is using.
TDMF FixedStorageThresholds
The high and low watermarks of fixed storage that TDMF uses during the session(obtained at system initialization). These values can change from one session toanother.
AverageRESERVE delay(msecs)
The cumulative amount of time that is spent waiting for SYSCOM Volume Reserverequests divided by the number of READ I/O issued.
Average intervalREAD (msecs)
The cumulative amount of time that is spent waiting for SYSCOM READ I/O tocomplete divided by the number of READ I/O issued.
Average volumeloop (msecs)
The cumulative amount of time that is spent processing SRC/TGT volumes betweenSYSCOM READ and SYSCOM WRITE divided by the number of WRITE I/O issued.
Average intervalWRITE (msecs)
The cumulative amount of time that is spent waiting for SYSCOM WRITE I/O tocomplete divided by the number of WRITE I/O issued.
154 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
The following figure shows an example of system information with performance data, using TDMF PTFlevel TD51901. The additional information, between "system" and "volume", is "volume processing looptiming data".
Performance Data Row 1 to 33 of 118Command ===> Scroll ===> CSRComDataSet: . . SWH10.TEST01.SYSCOMVOL Interval: : 0030.00Init Status: : All systems initialized.----------------------------------------------------------------------------------IBM Transparent Data Migration Facility (TDMF) for z/OS Version 5.2.0Communication Data Set: SWH10.TEST01.SYSCOM SYSTEM INFORMATION Type SMFID Asid System Current Number Interval ------ Heartbeat ------ --Expires In-- Master TDM1 0061 00 0030 06/09/2009 07:13:10.508 00898 Secs Agent TDMA 0053 01 0005 06/09/2009 07:13:12.992 00900 Secs Agent TDM2 0055 02 0005 06/09/2009 07:13:12.978 00900 Secs Agent TDMB 0051 03 0005 06/09/2009 07:13:13.007 00900 Secs Agent DRM1 0051 04 0030 06/09/2009 04:13:12.962 00900 Secs Previous Count Current Count TDM1's Fixed Storage Frames 24,052,079 33,406 Session Fixed Storage Frames 336 336 Session Fixed Storage Threshold - low 29,735 29,735 Session Fixed Storage Threshold - high 59,470 59,470 TDMA's Fixed Storage Frames 2,949,165 2,949,157 Session Fixed Storage Frames 37 37 Session Fixed Storage Threshold - low 30,026 30,026 Session Fixed Storage Threshold - high 60,052 60,052 TDM2's Fixed Storage Frames 2,949,165 2,949,157 Session Fixed Storage Frames 37 37 Session Fixed Storage Threshold - low 29,348 29,348 Session Fixed Storage Threshold - high 58,696 58,696 TDMB's Fixed Storage Frames 2,949,165 2,949,157 Session Fixed Storage Frames 37 37 Session Fixed Storage Threshold - low 30,004 30,004 Session Fixed Storage Threshold - high 60,008 60,008 DRM1's Fixed Storage Frames 2,949,165 2,949,157 Session Fixed Storage Frames 37 37 Session Fixed Storage Threshold - low 30,456 30,456 Session Fixed Storage Threshold - high 60,912 60,912 TDM1 Average RESERVE delay (msecs): 1.60 Average interval READ (msecs): 4.89 Average volume loop (msecs): 8.97 Average interval WRITE (msecs): 2.32 TDMA Average RESERVE delay (msecs): 4.55 Average interval READ (msecs): 2.97 Average volume loop (msecs): 3.72 Average interval WRITE (msecs): 2.59 TDM2 Average RESERVE delay (msecs): 6.93 Average interval READ (msecs): 3.13 Average volume loop (msecs): 3.79 Average interval WRITE (msecs): 3.53 TDMB Average RESERVE delay (msecs): 9.99 Average interval READ (msecs): 3.13 Average volume loop (msecs): 4.09 Average interval WRITE (msecs): 5.98 DRM1 Average RESERVE delay (msecs): 3.17 Average interval READ (msecs): 2.82 Average volume loop (msecs): 3.72 Average interval WRITE (msecs): 3.37
Figure 58. Performance Data panel - system information
Volume informationThe Volume Information section of the Performance Data panel displays performance data for theselected volume pairing.
The following table describes the information displayed in the Volume Information section.
Table 39. Option 5 - Current/past sessions: volume information
Field Description
Source VSN The source volume serial number.
Chapter 8. Viewing migration information 155
Table 39. Option 5 - Current/past sessions: volume information (continued)
Field Description
Target VSN The target volume serial number.
New VSN The new volume serial number the original source volume is to berenamed.
Current Phase The phase that the volume pairing is in.
Current Wait The current interval for that phase.
Cylinders Num The total number of cylinders to be processed for a specific phase.
Cylinders Curr The number of cylinders that were processed for that phase.
Number %%% The percentage completed for this volume pairing.
Number Req An indication of the amount of update activity that is detected byTDMF.
Number Compl The number of update requests completed.
Number Wait An indication of contention on the source volume when TDMF isattempting to quiesce the I/O for synchronization.
The following figure shows an example of the Volume Information section:
ComDataSet: . . HLQ.GTD560I.SYSCOM VOL Interval: : 0030.00 Init Status: : All systems initialized. _____________________________________________________________________________
VOLUME INFORMATION _____________________________________________________________________________ Source Target New ---- Current ---- Cylinders ---- Number ---- VSN VSN VSN Phase Wait Num Curr %%% Req Comp WaitJW71A0 JW71A1 Complete 0002 00200384 00200384 00000 00000 000Source Control Unit = 2107-E8 Device = 3390-0E Cylinders = 01182006 Target Control Unit = 2107-E8 Device = 3390-0E Cylinders = 01182006 Volume Pacing Active using 15 Tracks per Operation. --- T R A C K S --- 15 5 3 1 Percent of Copy Phase I/O using : 100 0 0 0 Percent of Refresh Phase I/O using : Number Pacing Changes Due to Real Storage Manager. Raise to : 0 0 0 0 Lower to : 0 0 0 0 Pacing Delay Due to I/O Contention Cumulative Last Interval Pacing I/O Delay 00.000 Secs 0.00 msec Customer Performance on Volume Cumulative Last Interval Total Elapsed Time Available 00:16:39 00:00:30 Total Elapsed Device Response Time 00:00:00 00:00:00 Average Device Response Time 00.001 Secs 00.000 Secs Average No. of I/O Operations per system 1 0 Percent of Device used by Customer 0 0 Percent of Device used by session 52 0
Figure 59. Performance Data panel - volume information
The following figure shows an example of volume information for a session that is using the System DataMover facility.
156 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Performance Data Row 35 to 68 of 90Command ===> Scroll ===> CSRComDataSet: . . HLQ.GTD550.SYSCOMVOL Interval: : 0030.00Init Status: : All systems initialized.------------------------------------------------------------------------VSN VSN VSN Phase Wait Num Curr %%% Req Comp WaitGFG100 GFG103 Suspended 0030 00000208 00000098 47 00000 00000 000Source Control Unit = 2105-E8 Device = 3390-0A Cylinders = 00001113Target Control Unit = 2105-E8 Device = 3390-0A Cylinders = 00001113SDM Owning System Number : 000 CCA : 00000101 x'00000065' Type : 016 x'10' ID : 002 x'02' Volume Threshold : 00001280 Volume Actual : 00000000
Figure 60. Performance Data panel - volume information with SDM
The six SDM lines are displayed only when the session uses SDM.
• The first line is the owning system number.• The second line is the CCA address that is used by SDM.• The third line is the XRC session type.• The fourth line is the XRC session ID.• The fifth line is the XRC volume threshold.• The sixth line is the XRC actual volume count.
Volume pacingThe Volume Pacing section of the Performance Data panel displays information about volume pacing forcurrent and past sessions.
Table 40. Option 5 - Current/past sessions: volume pacing
Field Description
Volume Pacing If volume pacing is active, the level of read operations that is ineffect (15, 5, 3, and 1). If volume pacing is inactive, it is stated.
Percent of TDMF Copy Phase I/O Shows the percentage of I/O operations that used 15 tracks, 5tracks, 3 tracks, and 1 track in a single I/O operation during the copyphase.
Percent of TDMF Refresh PhaseI/O
Shows the percentage of I/O operations that used 15 tracks, 5tracks, 3 tracks, and 1 track in a single I/O operation during therefresh phase.
Number Pacing Changes Due toReal Storage Manager (RSM)
The number of pacing changes made due to RSM constraints.
Note: The amount of storage that is used automatically changes thenumber of tracks read/written in a single I/O operation.
Pacing Delay Due to I/OContention
The amount of delay that is inserted between each TDMF I/Ooperation to provide more device availability. The delay is notedcumulatively and for the last interval.
The following figure shows an example of volume pacing (if selected).
Chapter 8. Viewing migration information 157
Volume Pacing Active using 15 Tracks per Operation. --- T R A C K S --- 15 5 3 1 Percent of Copy Phase I/O using : 100 0 0 0 Percent of Refresh Phase I/O using : Number Pacing Changes Due to Real Storage Manager. Raise to : 0 0 0 0 Lower to : 0 0 0 0 Pacing Delay Due to I/O Contention Cumulative Last Interval Pacing I/O Delay 00.000 Secs 0.00 msec
Figure 61. Performance Data panel - volume pacing
If reverse pacing was selected for the volume pairing, it is indicated at the beginning of the session on thispanel, where the number of tracks per operation is 01. If the level of activity permits, TDMF increases thenumber of tracks that are read per I/O operation.
Source volume performance dataPart of the performance data that is provided by TDMF is customer performance on the source volume.
The data that is displayed is cumulative for the migration and for the last interval (30 seconds). Thefollowing table describes the data that is displayed for source volume customer performance.
Table 41. Option 5 - Current/past sessions: source volume performance data
Field Description
Total Elapsed Device ResponseTime
These values represent the sum of the I/O response times across allmonitored systems.
Because I/O requests from multiple MVS systems can be "active" inthe I/O subsystems, the storage controller, or the device at any onetime, it is possible for the reported device response time to nowexceed the elapsed time.
Percent of Device used byCustomer/TDMF
The sum of the I/O service times across all systems, divided by thelength of the volume interval. The TDMF I/O response times do notinclude an IOS queue component.
It is possible to exceed 100% utilization of the device. If the sum ofthe TDMF and customer initiated I/O response times exceeds themonitoring interval; the individual percentages are scaled to total101.
The following figure shows an example of customer performance, in which the customer is using thevolume 110% of the time (percent of device that is used by customer). TDMF used approximately 1.1seconds to read a cylinder, so it used roughly 10 minutes 30 seconds so far. The percentages are scaledto total 101; the raw cumulative percentages would be 68.5% for TDMF and 114.4% for customer I/O.
158 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Performance Data Row 35 to 68 of 78Command ===> Scroll ===> CSR ComDataSet: . . SWH10.LOCAL2.SYSCOM VOL Interval: : 0030.00 Init Status: : All systems initialized. -------------------------------------------------------------------------- Customer Performance on Volume Cumulative Last Interval Total Elapsed Time Available 00:04:00 00:00:30 Total Elapsed Device Response Time 00:00:00 00:00:00 Average Device Response Time 00.001 Secs 00.000 Secs Average No. of I/O Operations per system 2 0 Percent of Device used by Customer 0 0 Percent of Device used by session 12 0 TCP/IP Performance Data Cumulative Last Interval Total written to target (KB): 1,444,771 0 Effective TCP/IP Data Rate (KB/sec): 24,762 0 Data compression effective (%): 37.3 Average Device Response Time by Operation: Read Write Copy 0.73 msec 0.44 msec Refresh 0.00 msec 0.00 msec Copy Cylinder 10.95 msec 6.60 msec Refresh Cylinder 0.00 msec 0.00 msec Estimated Sync Time 00.000 Secs 00.000 Secs Refresh and Synchronization Information. Refresh Sync. Refresh and Synchronization I/O counts Tracks 00000 00000 Cylinders 00000 00000 Synchronization Goal in Seconds : 005 Estimated Sync time required in seconds : 000 Copy Phase started : 04/10/2009 06:55:55.82 Copy Phase ended : 04/10/2009 06:56:43.75 Copy Elapsed Time : 00 Mins 48.93 Secs Refresh Phase started : 04/10/2009 06:56:55.32
Figure 62. Performance Data panel - customer performance
TCP/IP performance dataWhen the migration type is TCP/IP, TDMF collects additional information about the transfer to the remotesystem at the local session, or from the local system at the remote session.
The following table describes TCP/IP performance data.
Table 42. Option 5 - Current/past sessions: TCP/IP volume performance data
Field Description
Total written to target (KB) The amount of data that is transferred to the remote system for thisvolume. If the COMPRESS option was specified, this valuerepresents the amount of data before compression.
Effective TCP/IP Data Rate (KB/sec)
The amount of uncompressed data (in KB) that was transferred,divided by the number of seconds required by the TCP/IP APIroutine to send the data (on the local system) or receive the data(on the remote system). Because the TCP/IP send and receivebuffers are not infinite, some of this elapsed time includes writingthe data to the target volume or reading the data from the source.
Data compression effective (%) The approximate reduction in the amount of data that is transferredbecause the COMPRESS option was specified.
The following figure shows an example of TCP/IP performance data.
TCP/IP Performance Data Cumulative Last Interval Total written to target (KB): 1,444,771 614,089 Effective TCP/IP Data Rate (KB/sec): 24,762 36,812 Data compression effective (%): 37.3
Figure 63. Performance Data panel - TCP/IP performance data
Chapter 8. Viewing migration information 159
Average device response timesThe Average Device Response Times section of the Performance Data window displays the read andwrite times for the copy, refresh, copy cylinder, refresh cylinder, and estimated sync time operations.
The Average Device Response Times section provides the following information:
• Volume Refresh and Synchronization Information displays the number of tracks and cylinders that arerefreshed and synchronized for that specific migration.
• The Sync Goal in Seconds is the value that was set at job submission or modified through the TDMFMonitor.
• Estimated Sync Time Required in Seconds is the calculated value for the synchronization phase. Thisvalue should be less than the total Estimated Sync Time (read and write) that is displayed in theAverage Device Response Time section.
• The last display is the start and end times for each phase and the total elapsed time for the copy,refresh, and synchronization phases of the volume pairing.
Note: The estimated synchronization time is normally the sum of the predicted synchronization readand write times. However, if the Full Speed Copy option was requested, the read and write operationsoverlap, so the estimate is the greater of the two times, rather than their sum.
The following figure shows an example of average device response times.
Average Device Response Time by Operation: Read Write Copy 00.018 Secs 00.034 Secs Refresh 00.019 Secs 00.035 Secs Copy Cylinder 00.270 Secs 00.510 Secs Refresh Cylinder 00.285 Secs 00.525 Secs Estimated Sync Time 03.230 Secs 05.925 SecsRefresh and Synchronization Information. Refresh Sync. Refresh and Synchronization I/O counts Tracks 00123 00020 Cylinders 00017 00010 Synchronization Goal in Seconds : 005 Estimated Sync time required in seconds : 010
Figure 64. Performance Data window - average device response times
Each volume in the session is listed in the order that is specified on the MIGRATE or REPLICATE controlrecord in the Master batch job. Therefore, the previous five windows are displayed for each volume. Toview performance data for a specific volume within the session, you can use the find command or scrollthrough the windows.
If you want to see information about a different session, change the entry on the ComDataSet line at thetop of the window.
Note: Full Speed Copy Effect - Using the Full Speed Copy option can reduce the elapsed time of the copysubphases of each volume that is being migrated. Real memory is reduced for an extra data buffer so thatread and write I/O operations overlap or take place simultaneously. The impact of using Full Speed Copyis reflected in the average device response times. For example, if the device service times of the read andwrite cylinder operations for the copy and refresh phases were equal (510 milliseconds in the previousexample), if you use the Full Speed Copy option it would decrease the required time for I/O operations byhalf. The migration completes in about half the usual elapsed time
Option 6 - Past sessions: display summaryUse TDMF TSO Monitor option 6 to display previous session summaries. All that is required is theCOMMDS name of the session.
You can also view previous session summaries on the Average Device Response Times section of thePerformance Data panel by moving the cursor to a specific COMMDS and pressing the PF5 key (see“Average device response times” on page 160).
The following figure shows an example of a previous session summary.
160 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Past Session Summary Row 1 to 14 of 51ComDataSet : GTD550.V.SYSCOM------------------------------------------------------------------------TDMF Version 5.4.0Session started 04/11/2009 10:07:47.313Volume Device Group --- Migration --- - Error Info - SyncSerial Number Name Status Type System Message GoalSPMS84 8784 DASD1 Terminated SWAP TDM1 GTD3537E 005SPMS85 8785TD9C2J 8411 DASD1 Complete SWAP 005TD8D0J 8410TD8E01 8421 DASD1 Complete SWAP 005TD8E11 8420SPMS90 8790 DASD2 Terminated PIT TDM1 GTD3537E 005SPMS91 8791 PROD3C E43C DASD2 Terminated PIT TDM2 GTD2268E 005PROD3D E43DSPMS86 8787 Complete SWAP 005SPMS87 8786 PRODB2 87B2 Complete SWAP 005PRODB3 87B3
Figure 65. Past Session Summary panel
In the following figure, the SDM identifier on the last line indicates that the session used the System DataMover facility. It is blank in all other cases.
Past Session Summary Row 1 to 9 of 9 Command ===> Scroll ===> CSR ComDataSet : HLA.GTD550.SYSCOM --------------------------------------------------------------------- IBM Transparent Data Migration Facility (TDMF) for z/OS Version 5.9.0 Session started 06/24/2009 11:31:14.612 Volume Device Group --- Migration --- - Error Info - Sync Serial Number Name Status Type System Message Goal GFG100 5565 Complete SWAP 005 SDM GFG103 5165
Figure 66. Past session summary with SDM
Option 7 - Past sessions: display detailsUse TDMF TSO Monitor option 7 to display all data that is related to the COMMDS, the source and targetvolumes, and all messages that are issued by the Master and Agent systems from previous sessions.
To display a past session, enter the COMMDS name. You can also view previous session summaries onthe Average Device Response Times section of the Performance Data panel by moving the cursor to aspecific COMMDS and pressing the PF5 key (see “Average device response times” on page 160).
The following figure shows an example of past session details.
Chapter 8. Viewing migration information 161
Past Session Display Row 1 to 14 of 96ComDataSet : GTD550.V.SYSCOM------------------------------------------------------------------------TDMF Version 5.4.0 Completed MigrationComDataSet : GTD550.V.SYSCOM Volume serial number TD2398 on 3390 device 2398 Starting Cylinder is x'00000270', 00624 decimal Pin Token = 02A259C8 UCB = 0274E9F0Number of Systems active = 02. Volumes active = 02.MSE for SMFID TDM1, ASID: 0075, Relative System Number 00MSE for SMFID TDM2, ASID: 0073, Relative System Number 01The current messages for SMFID TDM1 follow.04/11/2009 10:07:46.960 GTD1731I The I/O Monitor module was found in Dynamic LPA.04/11/2009 10:07:47.268 GTD1448I The Time-of-Day System Default Option is GMT.04/11/2009 10:07:47.268 GTD1451I The TOD Option has been overridden by the OPTIONS statement and set to LOCAL.04/11/2009 10:07:47.268 GTD1453I The Pacing System Option was specified as OFF.Volume Migration summaries follow: Src=TDE972 Target=TDE973 100% Done MaxCyl=03339 CurCyl=03339Messages for volume TDE972 SMFID TDM1 follow.04/11/2009 16:17:49.615 GTD1478I A volume AutoOper option was changed to ON by a Migrate card option.04/11/2009 16:17:49.615 GTD1469I The Perpetual Point-in-Time option was specified for a volume.04/11/2009 16:17:49.716 GTD1177I The source volume TDE972 is mounted on device E973 on this system.04/11/2009 16:17:49.716 GTD1182I At TDMF initialization, the source volume has caching (CFW) activated.04/11/2009 16:17:49.716 GTD1184I At TDMF initialization, the source volume has dasd fast write allowed.Command ===> Scroll ===> CSR F1=Help F2=Split F3=Exit F5=Rfind F7=Back F8=Forward F9=Swap F12=Cancel
Figure 67. Past Session Details panel
Note: This panel supports a 160-character line display. Both wide screen (160) and normal (80) are thesame except for the size of the line.
The following figure shows an example of a past session that used the System Data Mover facility.
Past Session Display Row 51 to 86 of 293Command ===> Scroll ===> CSRComDataSet : HLQ.GTD550.SYSCOM----------------------------------------------------------------------Volume Migration summaries follow:Src=GFG100 Target=GFG103 100% Done MaxCyl=00208 CurCyl=00208SDM Owning System Number : 000 CCA : 00000101 x'00000065' Type : 016 x'10' ID : 002 x'02' Volume Threshold : 00001280 Volume Actual : 00000000
Figure 68. Past session display with SDM
The six SDM lines are displayed only if the session uses SDM.
• The first line is the owning system number.• The second line is the CCA address that is used by SDM.• The third line is the XRC session type.• The fourth line is the XRC session ID.• The fifth line is the XRC volume threshold.• The sixth line is the XRC actual volume count.
162 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Option 8 - Past sessions: display communication data set historyUse TDMF TSO Monitor option 8 to view the data set history.
If the GTDAHIST batch job is run and a SYSCOM History Data Set Name was specified in the Systemdefaults, TDMF keeps a history log for all COMMDS that were used. The history log displays the date andtime that the COMMDS was used, if it was reused, and if there was an x37 error. The following figureshows an example.
History Data Set Display Row 1 to 4 of 4 TDMF Version 5.4.0 Company Name : IBM Site Number and Name : 50069 M2 Data Center History Data Set Name : 'HLQ.GTD550.LOG' Report Date and Time : 04/13/2009 11:45:15 GTD550.VA.SYSCOM 04/10/2009 10:25:30 GTD550.VB.SYSCOM 04/10/2009 12:19:02 ReUsed GTD550.VC.SYSCOM 04/09/2009 12:47:42 GTD550.VX.SYSCOM 04/13/2009 09:46:56 Command ===> Scroll ===> CSR F1=Help F2=Split F3=Exit F4=Detail F5=Summary F6=Perform F7=Back F8=Forward F9=Swap F10=Updates F11=Trace F12=Cancel
Figure 69. History Data Set Display panel
Use the PF keys to navigate the History Data Set Display panel. For example, to see performancestatistics move the cursor to the COMMDS that you want to select and press PF6. A help panel is availableto aid navigation through this display, as shown in the following figure.
Help for Communication Data Set Display More: - Displays information logged into history file. o Placing cursor on a given data set line and depressing ENTER or PF4 displays that communication data set's messages. o Placing cursor on a given data set line and depressing PF6 displays that communication data set's performance information. o Placing cursor on a given data set line and depressing PF10 displays time and dates of updates to the User's Install Option information. o Placing cursor on a given data set line and depressing PF11 displays that communication data set's trace data. If you use any of the above options to display different information, when you depress PF3 to exit that display you will be returned to the History Data Set Display. F1=Help F3=Exit F5=Exhelp F6=Keyshelp F7=PrvTopic F8=NxtTopic F10=PrvPage F11=NxtPage F12=Cancel
Figure 70. History data set display – help
Option 9 - Display installation options and environmentUse TDMF TSO Monitor option 9 to review the current default options set by the GTDSOPTN batch job.The TDMF TSO Monitor operating system environment is also displayed.
The following figure shows an example of the Installation Options panel.
Chapter 8. Viewing migration information 163
Installation Options Row 1 to 4 of 15Command ===> Scroll ===> CSRCompany : IBM CorporationSite . : TDMF Version 5 Release 9.0Site ID : 05656 Date Initialized . . . : April 7, 2019History Data Set . . . . . : ** History Logging not installed **Security package volume protection . . . : NoWTO/WTOR for automated operations . . . : YesAutomatic ICKDSF REFORMAT on swap . . . : NoCheck target volumes are empty . . . . . : NoVolume counted as active only in copy . : NoUnidentified connected systems . . . . . : WarnOVA Registration Interval (minutes) . . : NoneWTO AutoOps Route Codes . : 2,4,6,11,27SMF Record Type . . . . . . . : 205 Use Startup Confirmations . . : YesUse Local Time on Messages . . : Yes Terminate group on error . . . : YesUse Pacing during Migrations . : Yes Use Reverse Pacing algorithm . : NoAllow invalid count fields . . : No Operating EnvironmentIBM Transparent Data Migration Facility (TDMF) for z/OSVersion 5.9.0SECURITY DSN.: 'HLQ.GTD550.GTDLLIB'USER ID. : FBL00CPU ID. : 0015D77D 2084SCP NAME : SP7.0.9SCP FMID : HBB7740ETR ID. : 00Local Time : 04/13/2019 11:37:02.73GMT Time : 04/13/2019 15:37:02.73Local Offset : -04:00:00Leap Seconds : +000
Figure 71. Installation Options panel
Note: You might need to scroll forward to see all of the information in the Operating Environmentsection. The default installation options do not scroll off the panel.
To display the installation security environment, press the PF10 key.
Option 10 - Display/modify installation security environmentUse TDMF TSO Monitor option 10 to display the security keys that are set by the GTDSOPTN batch job andthe TDMF feature in use.
The TDMF authorization program (GTDAKEY) uses the internal hardware clock to check the date and time.This internal hardware clock operates on Greenwich mean time (GMT). Therefore, GMT is used todetermine the effective date or expiration date of certain types of keys. Leap seconds (the one-secondadjustments that are occasionally applied to keep Coordinated Universal Time aligned with GMT) areignored.
The following table describes the fields in the Installation and Security Environment panel:
Table 43. TDMF installation and security environment
Field Description
Company Company name where TDMF is installed.
Site Location where TDMF is installed.
Site ID Five-digit value that was assigned by IBM Corporation.
Date Initialized Date when GTDSOPTN batch job was run.
Maintenance Expiration Date Date that maintenance for TDMF or a specific feature expires.
Commit Changes YES commits changes. NO does not commit changes.
Return Code from TDMF Security nnnnnnnn is the rc from GTDAKEY. See Appendix C, “Authorizationreturn codes,” on page 203 for TDMF authorization return codes.
164 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Table 43. TDMF installation and security environment (continued)
Field Description
__ Enter A to add a key. Enter D to delete a key
Security Key The 16-digit security key that is assigned by IBM.
Key Description User-supplied description of each key assigned.
Added Date and time (GMT) that the security record was updated.
By Batch job name (GTDSOPTN) or TSO UserID that updated thesecurity record through the TDMF Monitor.
Return codes from the security and authorization programs are listed in Appendix C, “Authorization returncodes,” on page 203.
Adding authorization keysYou can add multiple authorization keys from the Installation Security Environment panel.
About this task
Enter KEYnn as specified by IBM Corporation, otherwise TDMF does not function on the CPU.
Procedure
To add an authorization key, complete the following steps from the Installation Security Environmentpanel.1. Enter A under Requested Action on the bottom row.2. Enter the authorization key under Security Key.3. Enter the key number (01-16) under Key Number.4. Press the Enter key.5. Repeat these steps to add multiple keys.6. When you are done adding authorization keys, enter yes or no to Commit the Changes and press the
PF3 key.A message that the changes were accepted or not accepted is displayed under the Commit Changesline.
Deleting keysYou can delete authorization keys from the Installation Security Environment panel.
About this taskOnly TDMF regular license authorization keys can be deleted from the TDMF Monitor.
Procedure
To delete an authorization key, complete the following steps from the Installation Security Environmentpanel.1. Enter D’ under Requested Action.2. Press the Enter key.3. To commit the changes, type Yes on the appropriate line and press the PF3 key.
Chapter 8. Viewing migration information 165
Transactional terabyte license (TTL), terabytes purchased, terabytes movedThe transactional terabyte license (TTL), terabytes purchased, and terabytes moved are displayed on theInstallation and Security Environment panel.
In the following figure, the TDMF Full Function feature is noted at the top of the panel.
Command ===> ----------------------------------------- Scroll===> CSR
Company : XYZ CorporationSite : XYZ CorporationSite ID : 99999 Date Initialized ................ January 28, 2013Expiration Date................................... February 28,2013Migration Terabytes Purchased .................... 001.0000000Migration Terabytes Performed .................... 000.0000000Return Code from Security ........................ 00000000Features Enabled ................................. MIGRATE REPLICATE PPIT OVA
TDMF for z/OS Version 5.9.0 Base
No Security Key Key Description Added Date/Time By
01 CCA692F2A654E27A CPU AD6E 01/28/2013 GTDSOPTN
Figure 72. Terabyte license panel
Option 11 - Display/modify user's TSO monitor optionsUse TDMF TSO Monitor option 11 to set default options for the TDMF TSO Monitor.
These options include monitor time display, default first panel display, and how volumes are displayedwithin the session. The following figure shows an example of user monitor options.
User's Monitor OptionsCommand ===> Scroll ===> CSR PF3 to exit ---- PF5 to save changesUser's ID SWH10 Display Monitor Messages in Local or GMT Time LOCALUser's First Monitor Function - with cursor on the field, depress PF1for list of valid specifications. . . . . . . . . . . . . . . . . . . . XDisplay Preference (1 or 2, (if 2, define priority below)) . . . . . . 1Waiting Allow 01 Copy 08 Uninitialized 15Waiting Reply 02 Resume 09 Synchronize 16Waiting PPIT 03 Backed out 10 Compare 17Waiting OVA 04 Swap 11 Terminating 18OVA Active 05 Quiesce 12 Terminated 19Refresh 06 Suspended 13 Complete 20Activation 07 Inactive 14Qualified Data Set for savingJCL members -- Optional . . ____________________________________________F1=Help F2=Split F3=Exit F5=Save F7=Back F8=ForwardF9=Swap F12=Cancel
Figure 73. User’s Monitor Options panel
Option 12 - Build replication or migration jobsUse TDMF TSO Monitor option 12 to create and submit TDMF Master and Agent batch jobs.
All TDMF Master and Agent batch jobs can be created and submitted through the TDMF TSO Monitor. JCLgenerated by using the following panels is saved in a user-defined data set in Option 11 User’s MonitorOptions or by default in the sample library SGTDSAMP.
Press the PF11 key to display the installation defaults as set by the GTDSOPTN batch job (seeInstallation Options panel).
The following figure shows the Build/submit Jobs panel.
166 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Command ===> Scroll ===> CSR
PF 4 --> Enter/Update/Review Job Cards
PF 5 --> Build/Submit Master Migration Job
PF 6 --> Build/Submit Agent(s) Migration Job PF 10 --> Build JCL to Create a Communication Data Set
PF 11 --> Display Current User Installation Options
F1=Help F2=Split F3=Exit F4=Define F5=Master F6=Agent F9=Swap F10=ComDaSet F11=Options F12=Cancel
Figure 74. Build/submit Jobs panel
Enter/update/review job cardsPress the PF4 key to open the Enter/Update/Review Job Cards panel.
The following figure shows an example of the Enter/Update/Review Job Cards panel with a sample jobcard.
Enter/Update/Review Job Cards JCL saved Enter Job Control Statements, insure correctness prior to saving PF3 to exit ----- PF5 to save data > //EXAMPLE JOB 'TDMF SESSION', > // CLASS=A,MSGCLASS=Y,NOTIFY=USERID > //*Command ===> Scroll ===> CSR F1=Help F2=Split F3=Exit F5=Save F7=Back F8=Forward F9=Swap F12=Cancel
Figure 75. Enter/Update/Review Job Cards panel
To return to the previous panel, press the PF3 key.
Building a Master system batch jobYou can build a Master system batch job from the Build/submit Jobs panel.
Procedure
1. After a job card is created and saved, press the PF5 key to start the process for creating, updating,reviewing, or submitting a Master system batch job.The following figure shows the Master Job Selection panel.
Master Job Selection PF 4 --> Build a new Migration Session PF 5 --> Enter/Update/Review Saved Session PF 6 --> Final Review/Submit/Save Saved SessionCommand ===> Scroll ===> CSR F1=Help F2=Split F3=Exit F4=Library F5=Session F6=Migrate F7=Review F9=Swap F12=Cancel
Figure 76. Master system job selection2. Press the PF4 key to build a new migration session.
The Master Confirmation panel is displayed.
Chapter 8. Viewing migration information 167
Master Confirmation ScreenYou have requested the building of a New TDMF Job.The old job will be deleted, unless you requested duringits build that it be saved in the GTDSAMP library.Confirm with "YES" to continue building a New TDMF job.Confirm? (YES/NO): . . ___Command ===> Scroll ===> CSR F1=Help F2=Split F3=Exit F9=Swap F12=Cancel
Figure 77. Master system confirmation screen3. From the Master Confirmation panel, confirm the new job.
4. After confirmation is accepted, the Job Build Menu is displayed. In this example, the confirmation isYES.
Job Build Menu Reply was YES PF 4 --> Enter/Update/Review Job Name, Libraries, & SMF IDs PF 5 --> Enter/Update/Review Session Job Options PF 6 --> Enter/Update/Review Migration Information PF 7 --> Final Review/Submit/Save Session JobCommand ===> Scroll ===> CSR F1=Help F2=Split F3=Exit F4=Library F5=Session F6=Migrate F7=Review F9=Swap F12=Cancel
Figure 78. Master system job build menu5. Press the PF4 key to continue. The Define Libraries & SMF IDs panel is displayed. Complete all of
the appropriate fields.
Note: Until you press the PF5 key, the message ** UNSAVED record ** is displayed. After youpress the PF5 key, the message changes to SAVED record.
Define Libraries & SMF IDs PF3 to exit ----- PF5 to save dataMember name for saving . . . EXAMPLE_TDMF Load Library . . . . . HLQ.IBM.HGTD590.GTDLLIBTDMF Security Library . . . HLQ.IBM.HGTD590.GTDLLIBCommunication Data Set . . . GTD550.V.SYSCOM___________________________Master Complex ExecutionAffinity JCL Statement . . . ____________________________________________Master SMF ID . . . . . . . . . . TDM1Agent SMF IDs (optional) . . . . TDM2 > ____ > ____ > ____> ____ > ____ > ____ > ____ > ____ > ____ > ____ > ____ > ____> ____ > ____ > ____ > ____ > ____ > ____ > ____ > ____ > ____> ____ > ____ > ____ > ____ > ____ > ____ > ____ > ____ > ____ ** UNSAVED record **Command ===> F1=Help F2=Split F3=Exit F5=Save F7=Back F8=Forward F9=Swap F12=Cancel
Figure 79. Master system – define libraries and SMF IDs6. Press the PF5 key to save the library. The message SAVED record is displayed.
Define Libraries & SMF IDsCommand ===> Scroll ===> PAGE PF3 to exit ----- PF5 to save dataMember name for saving . . . EXAMPLETDMF Load Library . . . . . HLQ.IBM.HGTD590.GTDLLIBTDMF Security Library . . . HLQ.IBM.HGTD590.GTDLLIBCommunication Data Set . . . GTD550.V.SYSCOMMaster Complex ExecutionAffinity JCL Statement . . . /*JOBPARM SYSAFF=TDM1Master SMF ID . . . . . . . . . . TDM1Agent SMF IDs (optional) . . . . TDM2 > ____ > ____ > ____> ____ > ____ > ____ > ____ > ____ > ____ > ____ > ____ > ____> ____ > ____ > ____ > ____ > ____ > ____ > ____ > ____ > ____> ____ > ____ > ____ > ____ > ____ > ____ > ____ > ____ > ____ SAVED record
Figure 80. Master system – define libraries & SMF IDs - saved
168 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
7. After the TDMF libraries and SMF IDs are defined and saved, press the PF3 key to open the DefineSession Options panel. These session options can override the default system options that wereselected when the GTDSOPTN batch job was run.
Note: The option specifications from the GTDSOPTN batch job are set as the initial defaults. Changesthat are made on the Define Session Options panel are saved in the user ISPF profile and becomethe default values. Where appropriate, the TDMF TSO Monitor adds volume migration options to thegenerated SESSION control statement to simplify the generated MIGRATE statements. Some optionsthat can be specified on the SESSION or MIGRATE control statements cannot be set on these panels.To supply additional options, manually edit the generated control statements (refer to “TDMF controlstatements” on page 35).
Define Session OptionsCommand ===> Scroll ===> PAGE PF3 to exit ----- PF5 to save dataMigration Session time display in local time . . . YMigration Session Pacing Requested . . . . . . . . NReverse Pacing Requested . . . . . . . . . . . . . NConfirmation Required at Initialization . . . . . NTerminate group on error . . . . . . . . . . . . . NMVS operator messages issued . . . . . . . . . . . NTreat all volumes as a single group . . . . . . . _Maximum number of concurrent migrations . . . . . __Check for empty Target VTOC . . . . . . . . . . . NInvoke ICKDSF for REFORMAT after swap . . . . . . NControl Response to unidentified systems . . . . . W ** UNSAVED record **
Figure 81. Master system – define session options8. After you define the session options, press the PF5 key to save the selected options and mark the
record as saved.The following figure shows an example of the saved session options.
Define Session Options Session Options SavedCommand ===> Scroll ===> PAGE PF3 to exit ----- PF5 to save dataMigration Session time display in local time . . . YMigration Session Pacing Requested . . . . . . . . NReverse Pacing Requested . . . . . . . . . . . . . NConfirmation Required at Initialization . . . . . NTerminate group on error . . . . . . . . . . . . . NMVS operator messages issued . . . . . . . . . . . NTreat all volumes as a single group . . . . . . . NMaximum number of concurrent migrations . . . . . __Check for empty Target VTOC . . . . . . . . . . . NInvoke ICKDSF for REFORMAT after swap . . . . . . NControl Response to unidentified systems . . . . . W SAVED record
Figure 82. Master system – define session options - saved9. After you save the session options, press the PF3 key to open the migration definition panel.
From this panel, you can define a volume migration. To define additional volume migrations, pressthe PF6 key.
Define a Migration Row 1 to 1 of 1Command ===> Scroll ===> PAGEPF4 Previous Migration PF5 Save PF6 Next MigrationMigration Number . . . . . . . . . 01Source Volume Target Volume New Source Volume (opt) ______Type of Migration . . . . . . . . . Type of Synchronization . . . . .Volume Purge . . . . . . . . . . . Volume Comparison . . . . . . . .Synchronization Goal . . . . . . 005 Group Name . . . . . . . . ________Volume Pacing . . . . . . . . . . . Automated Operation Messages . . _Volume Confirmation Requested . . . _ Terminate Group on Error . . . . _Allow PPRC to non-mirrored device Copy Only Allocated Data . . . . _Allow Offline Volume Access (OVA) _ Perform Perpetual Point-In-Time _ ** UNSAVED record **Current Time is . . . . . 04/29/2009 01:44:42.13
Figure 83. Master system – define a volume migration
Chapter 8. Viewing migration information 169
10. Enter the necessary information and press the PF5 key when complete.If there are errors in the migration definition, they are noted on the panel. The following figure showsan example.
Define a Migration *** ERROR(s) ***Command ===> Scroll ===> PAGEPF4 Previous Migration PF5 Save PF6 Next MigrationMigration Number . . . . . . . . . 01Source Volume SPMS03 Target Volume SPMS05 New Source Volume (opt) ______Type of Migration . . . . . . . . . _ Type of Synchronization . . . . . _Volume Purge . . . . . . . . . . . _ Volume Comparison . . . . . . . . _Synchronization Goal . . . . . . 005 Group Name . . . . . . . . ________Volume Pacing . . . . . . . . . . . Y Automated Operation Messages . . YVolume Confirmation Requested . . . N Terminate Group on Error . . . . NAllow PPRC to non-mirrored device _ Copy Only Allocated Data . . . . YAllow Offline Volume Access (OVA) _ Perform Perpetual Point-In-Time _ ** UNSAVED record **There are errors in this record, you can notmove to another record. Must fix or PF3"Type of Migration" value in error.Position cursor in the error field.Depress PF1 to get an explanation of valid parameters.Current Time is . . . . . 04/29/2009 01:46:29.46
Figure 84. Master system – define a volume migration - with errors11. Correct the identified errors and press Enter or PF5.
After the errors are corrected, the following panel is displayed.
Define a Migration Migration Data SavedCommand ===> Scroll ===> PAGEPF4 Previous Migration PF5 Save PF6 Next MigrationMigration Number . . . . . . . . . 01Source Volume SPMS03 Target Volume SPMS05 New Source Volume (opt) ______Type of Migration . . . . . . . . . S Type of Synchronization . . . . . _Volume Purge . . . . . . . . . . . _ Volume Comparison . . . . . . . . _Synchronization Goal . . . . . . 005 Group Name . . . . . . . . ________Volume Pacing . . . . . . . . . . . Y Automated Operation Messages . . YVolume Confirmation Requested . . . N Terminate Group on Error . . . . NAllow PPRC to non-mirrored device _ Copy Only Allocated Data . . . . YAllow Offline Volume Access (OVA) _ Perform Perpetual Point-In-Time _ SAVED recordCurrent Time is . . . . . 04/29/2009 01:47:37.23
Figure 85. Master system – define a volume migration - saved12. After all of the volume migrations are defined and saved, press the PF3 key to open the Master Job
Build Menu (see “Building a Master system batch job” on page 167). On this panel, press the PF7 keyto display information about how the JCL is saved. If you want to submit a job, press the PF5 key.
Migration Job Review Row 1 to 4 of 4Command ===> Scroll ===> PAGE PF3 exit ------ PF5 Review/Submit JCL JCL will be saved in 'HLQ.IBM.HGTD590.GTDSAMP', Member : EXAMPLE Current Time is : 04/29/2009 01:48:39.89
Figure 86. Master system migration job review - saved information13. On the Migration Job Review panel, press the PF5 key to review the JCL that was created based on
the information that you entered in the previous panels. The following figure shows an example.
170 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Menu Utilities Compilers HelpBROWSE HLQ.IBM.HGTD590.GTDSAMP(TEMPNAME) - 01.01 Col 00001 000072Command ===> Scroll ===> CSR******************************* Top of Data *******************************//EXAMPLE JOB 'TDMF SESSION',// CLASS=A,MSGCLASS=Y,NOTIFY=GTD550/*JOBPARM SYSAFF=TDM1//TDMF EXEC PGM=GTDMAIN,PARM=MASTER//STEPLIB DD DISP=SHR,DSN=HLQ.IBM.HGTD590.GTDSAMP.GTDLLIB//GTDKEY DD DISP=SHR,DSN=HLQ.IBM.HGTD590.GTDSAMP.GTDLLIB//SYSCOM DD DISP=SHR,DSN=GTD550.V.SYSCOM//SYSPRINT DD SYSOUT=*//SYSSNAP DD SYSOUT=*//SYSIN DD * SESSION Master(TDM1) Options( Time(Local) Nopacing Unident(W) NoAutoOPs NoICKDSF NoCHKtarget NoAllowinvalidcounts NoConfirm NoPrompt NoTermgroup FASTCOPY NoPurge NoAllowMirrorChange Sync(005)) Agents(TDM2)************************************************************************ Additional Session options: NOXCF NOCFW RELABel(xx) ** NOWARNing(nnnn,nnnn) STORCLas(class) ** ** Additional Volume options: PAcing(YES,FULLSPeed) EXTVtoc(nn) PAV ** MAXTRacks(n) RESETCHAngedflag ************************************************************************ MIGRATE SPMS03 SPMS05 Options(Pacing AutoOPs)/****************************** Bottom of Data ******************************
Figure 87. Master system migration job – ISPF display
The Master system batch job was defined and saved. If Agent systems were identified in the Masterbatch job, one or more Agent system batch jobs must be defined.
Building an Agent system batch jobYou can define Agent system batch jobs on the Agent Migration JCL Build panel.
Procedure
1. On the Build/Submit Jobs panel (see Build/submit Jobs panel), press the PF6 key to open the AgentMigration JCL Build panel:
Agent Migration JCL Build Row 1 to 3 of 3 PF3 exit ---- PF5 ContinueAfter building the input needed to generate a TDMF Master Migration,all the data is available to generate the JCL needed for startingAgent jobs. Remember, all systems that have access to the sourcevolume MUST have either a Master or Agent job running. All jobsmust be started with in 15 minutes of each other.Two of the parameters for the Agent jobs are optional. If there areno errors and you wish to continue, the program will prompt you forthese parameters.------------------------------------------------------------------------Current Time is : 04/11/2009 14:08:35.83Command ===> Scroll ===> CSR F1=Help F2=Split F3=Exit F5=Continue F7=Backward F8=Forward F9=Swap F12=Cancel
Figure 88. Agent system migration JCL build2. Press the PF5 key to open the Define Agent Optional Parameters panel to create the Agent system
JCL.
The only required field in the following panel is a member name to save the JCL in the specified library.
Chapter 8. Viewing migration information 171
Define Agent Optional Parameters Row 1 to 5 of 5 PF3 to exit ------- PF5 to Review/Submit JCLMember name for saving . . . EXAMPLESTDMF Load Library. . . . . : HLQ.IBM.HGTD590.GTDLLIBSecurity Library . . . . . : HLQ.IBM.HGTD590.GTDLLIBCommunication Data Set . . . : GTD520.V.SYSCOMAgent Complex ExecutionAffinity JCL Statement . . . ____________________________________________Master SMF ID . . . . . . : TDM1Agent SMF ID . . . . . . . : TDM2------------------------------------------------------------------------Current Time is : 04/11/2009 14:09:01.96JCL will be saved in 'HLQ.IBM.HGTD590.GTDSAMP', Member : EXAMPLESCommand ===> Scroll ===> CSR F1=Help F2=Split F3=Exit F5=Save F7=Back F8=Forward F9=Swap F12=Cancel
Figure 89. Agent system optional parameters panel3. After you complete the Define Agent Optional Parameters panel, press the PF5 key to save and
review the JCL generated. If it is saved, the following panel is displayed so that you can define theother Agent systems.
Menu Utilities Compilers Help---------------------------------------------------------------------------BROWSE HLQ.IBM.HGTD590.GTDSAMP(TEMPNAME) - 01.01 Col 00001 00072****************************** Top of Data ********************************//EXAMPLE JOB 'TDMF SESSION',// CLASS=A,MSGCLASS=Y,NOTIFY=GTD550//*//*//* This AGENT job is for SMF ID: TDM2//*//TDMF EXEC PGM=GTDMAIN,PARM=AGENT//STEPLIB DD DISP=SHR,DSN=HLQ.IBM.HGTD590.GTDLLIB//GTDKEY DD DISP=SHR,DSN=HLQ.IBM.HGTD590.GTDLLIB//SYSCOM DD DISP=SHR,DSN=GTD550.V.SYSCOM//SYSPRINT DD SYSOUT=*//SYSIN DD DUMMY/****************************** Bottom of Data ******************************Command ===> Scroll ===> CSR F1=Help F2=Split F3=Exit F5=Rfind F7=Up F8=Down F9=Swap F10=Left F11=Right F12=Cancel
Figure 90. Agent system migration job – ISPF display4. After you define all Agent systems, press PF3 again to return to the Build/Submit Jobs panel (see
Build/submit Jobs panel).
Building a communications data set (COMMDS)If a Communications Data Set (COMMDS) was not previously defined, you must define it before yousubmit the Master and Agent batch jobs. Failure to do so results in failure of the Master and Agent batchjobs.
Procedure
To create a COMMDS, complete the following steps.1. Press the PF10 key from the Build/Submit Jobs panel to open the Create JCL to Allocate a
Communication Data Set panel.
172 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Create JCL to Allocate a Communication Data Set Row 1 to 5 of 5 PF3 to exit ----- PF5 to generate jclMember name for saving . . . COMMDSNo of Systems in Migration 01No of Volumes in Migration 01Communication Data Set Name GTD550.V.SYSCOMUnit Name . . . . . . . . . SYSDAVolume Serial Number . . . . PROD01Complex Execution AffinityJCL Statement . . . . . . . ____________________________________________------------------------------------------------------------------------Current Time is : 04/11/2009 14:14:24.85JCL will be saved in 'HLQ.IBM.HGTD590.GTDSAMP', Member : COMMDSCommand ===> Scroll ===> CSR F1=Help F2=Split F3=Exit F5=Review F7=Back F8=Forward F9=Swap F12=Cancel
Figure 91. JCL create for COMMDS panel2. Complete the necessary fields and press PF5 to view the JCL that was created for the COMMDS
allocation.
Results
The following figure shows the JCL that was created from the information in the previous example. Youcan save the JCL for later use or submit it from this panel.
Menu Utilities Compilers Help---------------------------------------------------------------------------BROWSE HLQ.IBM.HGTD590.GTDSAMP(TEMPNAME) - 01.01 Col 00001 00072******************************* Top of Data *******************************//EXAMPLE JOB 'TDMF SESSION',// CLASS=A,MSGCLASS=Y,NOTIFY=GTD550//*//* This creates a COMMUNICATION DATA SET//*//TDMF EXEC PGM=IEFBR14//SYSCOM DD DISP=(NEW,CATLG,DELETE),// DSN=GTD520.V.SYSCOM,// DCB=(LRECL=4096,BLKSIZE=4096,RECFM=F),// SPACE=(CYL,004,,CONTIG),// VOL=SER=PROD01,// UNIT=SYSDA//SYSPRINT DD SYSOUT=*//SYSIN DD DUMMY/******************************* Bottom of Data *****************************Command ===> Scroll ===> CSR F1=Help F2=Split F3=Exit F5=Rfind F7=Up F8=Down F9=Swap F10=Left F11=Right F12=Cancel
Figure 92. Allocate JCL COMMDS panel
By using the previous panels, a migration session was defined and is ready to be run.
The remaining option that is listed on the Build/Submit Jobs panel is PF11, which displays the currentuser environment and installation options as shown in Installation Options panel.
Option U - TDMF support utilitiesUse TDMF TSO Monitor option U to access TDMF support utilities.
You can use the functions on the Support Utilities menu to help resolve problems.
Option U.0 – Display memoryUse TDMF TSO Monitor option U.0 to display active memory by specifying an address and length that arerelated to the operating system of the TDMF TSO Monitor.
In the following example, the physical configuration communication area (PCCA) is displayed. It containsinformation that is related to the physical facilities that are associated with each processor in the system.
Chapter 8. Viewing migration information 173
Memory Display Row 1 to 8 of 8Address . . 00f4c008Length . . 8000F4C008 D7C3C3C1 F0F9F0F3 F8F0F0F3 F5F9F9F5 :PCCA090380035995:00F4C018 00008000 00FCF1E8 00F8B000 0C293000 : . .1Y 8. ... :00F4C028 FF000000 FE000000 00585503 00006FC4 :. . ... ?D:00F4C038 3D184240 00001000 00000000 00000000 :... . :00F4C048 00000000 00000000 00000000 00000000 : :00F4C058 00000000 00000000 00000000 00000000 : :00F4C068 00000000 00000000 00000000 00000000 : :00F4C078 00000000 00000000 00000000 00000000 : :
Figure 93. Memory Display panel
Note: This utility supports 64-bit addresses as an input and supports 160-character lines.
Option U.1 – View internal details for active sessionsUse TDMF TSO Monitor option U.1 to display specific data about active TDMF sessions. This informationincludes the location of specific control blocks within TDMF.
The following figure is an example of the internal details display utility.
Internal Details Display Row 1 to 16 of 16--------------------------------------------------------------------------TDMF. Master V5.4.0 Session Active. ComDataSet : GTD550.V.SYSCOM Volume serial number STOR0A on 3390 device DE4F Starting Cylinder is x'000002C6', 00710 decimal Pin Token = 02155730 UCB = 00F1E940Master System. Relative System Number=00, SMFID=TDM1, ASID=0032Number of Systems active = 02 Volumes active = 01TDMF Control Block is at 07F2B000.MSE Control Block is at 07F15000.MSV Control Block is at 07F17000.MSVE Control Block is at 07F1D000.Command ===> Scroll ===> CSR F1=Help F2=Split F3=Exit F5=Rfind F7=Back F8=Forward F9=Swap F12=Cancel
Figure 94. Internal Details Display panel
Note: This utility supports the display of 64-bit addresses.
To display the control block addresses use Option U.0 - Display Memory for review or diagnosis.
Option U.2 – Active session – display/alter tracing bit settingsUse TDMF TSO Monitor option U.2 to display the default trace bit settings for a TDMF session. You can setthese bits to other values, if instructed by technical support personnel.
You must enter the bit values and then press the PF5 key before the settings take effect. After you modifythe settings, press PF3 to exit the panel.
Each entry represents various tracing functions within TDMF for analysis by technical support personnel.To view these trace functions press PF1 from each flag entry. The following figure is an example of theTrace Status panel.
174 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Trace Status Row 1 to 6 of 6ComDataSet : GTD550.V.SYSCOM 01 of 02Environment : Master SystemTracing Flag 1 . . . . . . . . 1 1 1 1 1 1 1 1Tracing Flag 2 . . . . . . . . 1 1 1 1 1 1 1 1Tracing Flag 3 . . . . . . . . 1 1 1 1 1 1 1 1Tracing Flag 4 . . . . . . . . 0 0 0 0 0 0 0 0Diagnostic Flag 1 . . . . . . 0 0 0 0 0 0 0 0Diagnostic Flag 2 . . . . . . 0 0 0 0 0 0 0 0Initialization Flag 1 . . . . 0 0 0 0 0 0 0 0Recovery Flag 1 . . . . . . . 0 0 0 0 0 0 0 0TDMF Display/Update Tracing BitsTDMFVersion 5.4.0User ID: GTD550Date and Time : 04/13/2009 09:42:36.95Command ===> Scroll ===> CSR F1=Help F2=Split F3=Exit F4=Next F5=Update F7=Back F8=Forward F9=Swap
Figure 95. Trace Status panel
Option U.3 – Active session – display trace tableUse TDMF TSO Monitor option U.3 to display a TDMF trace table with the most current entries at thebeginning of the display. Technical support personnel use these trace table entries to identify problems.
The following figure shows an example of the TDMF Trace Entries panel.
TDMF Trace Entries Row 1 to 15 of 4,610ComDataSet : GTD550.V.SYSCOM 01 of 01Environment : Master System Tracing Bits : FF FF FF 00 00 00 00 00---------------------------------------------------------------------------04/12/2009 E6E2C930 8986C316 B2C6276A DE8FE282 *WSI.ifC..F....Sb*09:42:36.953230 0986C000 00000316 00000000 00000000 *.f..............* 1 00000000 00000000 00000000 00000000 *................* 00000000 00000000 00000000 00000000 *................* 23000000 07F555E4 00F78100 07F555E4 *.....5.U.7a..5.U* 00F742E8 00F74314 00EF8A40 07F55000 *.7.Y.7..... .5..* 00FCF3C0 07F4A000 07F48000 0986D000 *..3..4...4...f..* 8986C000 7F70D000 8986C29A 00000000 *if......ifB.....*04/12/2009 E6E2C930 8986C316 B2C6276A C73BC200 *WSI.ifC..F..G.B.*09:42:36.957676 0986C000 00000316 00000000 00000000 *.f..............* 0 00000000 00000000 00000000 00000000 *................* 00000000 00000000 00000000 00000000 *................* 23000000 07F555E4 00F78100 07F555E4 *.....5.U.7a..5.U* 00F742E8 00F74314 00EF89C0 07F55000 *.7.Y.7....i..5..*Command ===> Scroll ===> CSR F1=Help F2=Split F3=Exit F4=Next F5=Rfind F6=Refresh F7=Back F8=Forward F9=Swap
Figure 96. Trace entries – active session
Option U.4 – Previous session – display trace tableUse TDMF TSO Monitor option U.4 to display a trace table for a specific completed session, with the mostcurrent entries at the beginning of the display. Technical support personnel use these trace table entriesto identify problems.
The following figure is an example of the Past Session Trace Entries panel.
Chapter 8. Viewing migration information 175
Past Session Trace Entries Row 1 to 16 of 4,613ComDataSet : GTD550.V.SYSCOM--------------------------------------------------------------------------Trace Entries for System : TDM1Final Trace Bit Settings : F7 FE FE 00 00 00 00 0004/12/2009 C5E5D330 97894100 B726BB1C 79E21900 *EVL.pi.......S..*16:46:38.533729 17894000 00000100 00000000 00000000 *.i .............* 0 00000009 00000000 004A0000 00000000 *................* FFFFFFFF E3F1E9F1 C3F0ED48 20CFC000 *....T1Z1C0......* 00000027 00000644 00000000 00000009 *................* 17487000 17489000 1748F000 15F7A000 *..........0..7..* 15F7D000 17425000 17424000 17895000 *.7........ ..i..* 97894000 17424048 978940B6 1788BF30 *pi ... .pi ..h..*04/12/2009 E4D5D930 9789E670 B726BB1C 79DE9300 *UNR.piW.......l.*16:46:38.533673 1789E000 00000670 00000000 00000000 *.i..............* 0 00000000 00000000 00000000 00000000 *................* 00000000 00000000 00000000 00000000 *................*Command ===> Scroll ===> CSR F1=Help F2=Split F3=Exit F5=Rfind F7=Back F8=Forward F9=Swap
Figure 97. Trace entries – completed session
Option U.5 – Module version levels in defined load libraryUse TDMF TSO Monitor option U.5 to display the assembly date and time of all TDMF load modules in thedefined TDMF load library.
The version level of each load module reflects the PR level that is applied, as shown in the followingfigure.
Module Version Level Display Row 1 to 39 of 133Command ===> Scroll ===> CSR IBM Transparent Data Migration Facility (TDMF) for z/OS Version 5.4.0 Report Date and Time 05/15/2009 11:44:30.48 Load Library Name : 'HLQ.IBM.HGTD590.GTDLLIB' GTDADDT level 31413 assembled on 2009-04-16 at 00.55 PT. GTDAKEY level 31437 assembled on 2009-04-13 at 00.53 PT. GTDAMAP level 30388 assembled on 2009-04-12 at 01.16 PT. GTDASPP level 30338 assembled on 2009-04-12 at 13.36 PT. GTDASSO level 30338 assembled on 2009-04-12 at 01.16 PT. GTDAVOL level 30318 assembled on 2009-04-09 at 00.36 PT. GTDBDDT level 30420 assembled on 2009-04-11 at 07.37 PT. GTDBMON level 30339 assembled on 2009-03-04 at 23.04 PT. GTDBUPM level 30339 assembled on 2009-03-04 at 22.26 PT. GTDBVOL level 30338 assembled on 2009-03-02 at 11.16 PT. GTDCDDT level 30378 assembled on 2009-04-17 at 08.15 PT. GTDCDIO level 30388 assembled on 2009-04-29 at 04.49 PT. GTDCFTP level 30338 assembled on 2009-04-12 at 13.37 PT. GTDCLIP level 30398 assembled on 2009-04-14 at 23.32 PT. GTDCMSG level 30391 assembled on 2009-04-12 at 01.18 PT. GTDCNAV level 30338 assembled on 2009-04-12 at 01.19 PT. GTDCOMF level 30388 assembled on 2009-04-12 at 01.19 PT. GTDCOMM level 30338 assembled on 2009-04-12 at 13.37 PT. GTDCOM1 level 30376 assembled on 2009-03-15 at 02.09 PT. GTDCOM2 level 30426 assembled on 2009-04-14 at 00.55 PT. GTDCOPY level 30318 assembled on 2009-04-09 at 00.35 PT. GTDCRCV level 30338 assembled on 2009-03-01 at 00.32 PT. GTDCSEL level 30318 assembled on 2009-04-09 at 00.34 PT. GTDCVOL level 30338 assembled on 2009-03-01 at 00.32 PT. GTDDDDT level 30378 assembled on 2009-04-17 at 08.15 PT. GTDDOVA level 30335 assembled on 2009-04-29 at 01.42 PT. GTDEDDT level 30378 assembled on 2009-04-17 at 08.15 PT. GTDEMC level 30378 assembled on 2009-04-17 at 09.01 PT. GTDEVOL level 30336 assembled on 2009-04-29 at 01.53 PT. GTDEXTV level 30332 assembled on 2009-04-05 at 07.38 PT.
Figure 98. Module version level display
176 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Option U.6 – Current/past session module version levels from COMMDSUse TDMF TSO Monitor option U.6 to display the TDMF load module version levels.
This panel displays the TDMF load module version levels based on an active or completedcommunications data set. This panel is similar to the load module version display shown in “Option U.5 –Module version levels in defined load library” on page 176; however, you can have more than one loadlibrary.
In the following figure, the communications data set is listed as inactive. This panel shows the modulesfrom a completed migration session. If a current session is needed, change the data set name for thatsession. The selected communications data set is listed as active.
Module Version Level Display Row 1 to 30 of 75Command ===>ComDataSet : SWH10.TEST01.SYSCOM------------------------------------------------------------------------ IBM Transparent Data Migration Facility (TDMF) for z/OS Version 5.4.0 Report Date and Time : 03/29/2009 03:33:12.72 Comm Data Set (inactive): SWH10.TEST01.SYSCOM GTDTERM level 30415 assembled on 2009-03-03 at 07.23 PT. GTDTRAC level 30338 assembled on 2009-03-12 at 23.28 PT. GTDRTNM level 29991 assembled on 2009-03-17 at 01.31 PT. GTDMSGN level 30423 assembled on 2009-03-17 at 01.22 PT. GTDCMSG level 30391 assembled on 2009-03-12 at 01.18 PT. GTDRTNV level 29991 assembled on 2009-03-17 at 01.31 PT. GTDIMON level 30378 assembled on 2009-03-13 at 01.59 PT. GTDSIO level 30399 assembled on 2009-03-15 at 00.19 PT. GTDVOL level 30364 assembled on 2009-03-12 at 00.55 PT. GTDAVOL level 30318 assembled on 2009-03-09 at 00.36 PT. GTDBVOL level 30338 assembled on 2009-03-02 at 11.16 PT. GTDCVOL level 30338 assembled on 2009-03-01 at 00.32 PT. GTDEVOL level 30336 assembled on 2009-03-29 at 01.53 PT. GTDMVOL level 30339 assembled on 2009-03-12 at 23.24 PT. GTDNVOL level 30339 assembled on 2009-03-01 at 00.52 PT. GTDQVOL level 30356 assembled on 2009-03-01 at 02.25 PT. GTDRCVR level 30336 assembled on 2009-03-01 at 02.38 PT. GTDRVOL level 30427 assembled on 2009-03-14 at 04.41 PT. GTDSVOL level 30339 assembled on 2009-03-12 at 07.30 PT. GTDTVOL level 30339 assembled on 2009-03-01 at 02.00 PT. GTDUVOL level 30339 assembled on 2009-03-01 at 02.02 PT. GTDWVOL level 30388 assembled on 2009-03-29 at 02.55 PT. GTDASPP level 30338 assembled on 2009-03-12 at 13.36 PT.
Figure 99. Module version level display based on COMMDS
Option U.7 – System change summaryUse TDMF TSO Monitor option U.7 to show which program temporary fixes (PTFs) were installed for theTDMF system. The PTFs are identified by a fix number.
In the following example, no PTFs were applied. The TDMF system is at base level.
System Change Summary Display Row 1 to 12 of 12Command ===> Scroll ===> CSR IBM Transparent Data Migration Facility (TDMF) for z/OS Version 5.4.0 System Change Summary Report Date and Time 04/13/2009 11:55:56.62 Base System released June 1, 2009
Figure 100. System change summary display
Option U.8 – Unit control block displayUse TDMF TSO Monitor option U.8 to display all unit control blocks (UCBs) that are used by TDMF.Additionally, the total number of DASD UCBs that can be addressed from an LPAR is listed.
The following figure shows an example of the Unit Control Block Display utility.
Chapter 8. Viewing migration information 177
Unit Control Blocks Display Row 1 to 14 of 14Command ===> Scroll ===> CSR DASD Unit Control Block Display TDMF Version 5.4.0 04/13/2009 14:38:49.79 Volume PROD03 UCB 8503 is being used by TDMF. Volume HOT812 UCB 8512 is being used by TDMF. Volume SPMS95 UCB 8794 is being used by TDMF. Volume SPMS94 UCB 8795 is being used by TDMF. Number of DASD UCBs : 7,997 To display data about a particular UCB Enter on the command line: LISTUCB xxx or xxxx
Figure 101. Unit control blocks display
If you need more information about a specific UCB, enter the LISTUCB command, followed by the deviceaddress on the command line. The following figure shows an example of the information that is displayedfor the selected UCB.
Unit Control Blocks Display Row 1 to 17 of 27 DASD Unit Control Block Display TDMF Version 5.4.0 04/13/2009 04:32:51.55176970A0 0088CC84 22F80000 00000000 00E4C3C2 *.h.d.8.......UCB*UCBAREA 3030200F 00000000 00000500 E3C4D4C6 *............TDMF* (COPY) F9F51000 00000000 00000000 00000000 *95..............*176970D0 00040040 00000000 00000000 0001167D *... ............*UCBPAREA 289C0F11 E00020E0 BE707100 00000000 *................* (COPY) 01080000 00000001 00000000 00F76E00 *.............7..*17697100 00000940 202A0000 00000000 00000000 *... ............*CMXTAREA 00000000 00FCCBEC 17697130 00000000 *................* (COPY) 00000000 00000000 00000000 00000000 *................*17697130 38788074 16C92FE0 00000000 020F9898 *.....I........qq*DCEAREA D9000D0B 0D1A2424 15960000 05EB0000 *R........o......* (COPY) 00FCCEC0 9400F0E0 F8153400 00030000 *....m.0.8.......* 00000000 00000000 *........ *00FCCBEC C4C4E340 00000000 EF740000 00200000 *DDT ............*DDT C9C5C3D3 010111F8 0100C7F0 0100AB28 *IECL...8..G0....* 00000000 01014DA8 00FDAD68 972B0A78 *.......y....p...* 010174C8 00000000 016C9210 C4C4D9C4 *...H......k.DDRD* 8100AF10 40404040 01012FC0 *a... .... * Command ===> Scroll ===> CSR F1=Help F2=Split F3=Exit F5=Rfind F7=Back F8=Forward F9=Swap F12=Cancel
Figure 102. LISTUCB display
Option U.9 – Communication data set control blocksUse TDMF TSO Monitor option U.9 to display TDMF control blocks. You can select specific volumes froman active or completed session. The relative system number can also be specified, so that control blocksfor a specific volume can be viewed from either the Master or Agent system.
In the following example, the TDMF Master System Entry (TDMFMSE) control block is displayed forrelative system number 00 (TDM1), which is the Master system.
178 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Communication Data Set Control Blocks Row 1 to 15 of 15Command ===> Scroll ===> CSRCommunication Data Set . . SWH10.TEST01.SYSCOMControl Block . . . . . . TDMFMSESource Volume . . . . . . PROD92System number . . . . . . 00--------------------------------------------------------------------------DISP HEX DUMP CHAR DATA0000 E2C4D9D7 D4E2C535 E3C4D4F1 F0F00000 :TDMFMSE.TDM100 :0010 BA560411 47C82720 BA560411 3D49CF80 :.....H..........:0020 F7FBFE00 00000000 00000000 94B0171E :7.. m...:0030 FFFF94B6 2E000000 0000000E 4E1C0000 :..m.. .+. :0040 010003D0 00180001 232E2066 BA4512CA :. .} . .........:0050 00007380 00006FCC 116D116D 44455445 : .. ?.._._....:0060 000000A3 00015F90 00015F90 00000BB8 : t .^. .^. ..:0070 00008000 1224F000 1DE78000 0000E406 : . ..0 .X. U.:0080 0000E401 00000000 00080008 00010000 : U. . . . :0090 121D8000 1B0B6000 0000E404 0000E401 :... ..- U. U.:00A0 00000000 00100010 80000000 00000000 : . .. :00B0 0001002E 00000000 01E10011 FFFF8009 : . . .. .....:00C0 48280880 00000000 00000000 00000000 :.... :==========================================================================
Figure 103. Communication data set control blocks – Master system
In the following example, the TDMFMSE control block is displayed for relative system number 01 (TDM2),which is an Agent system.
Communication Data Set Control Blocks Row 1 to 10 of 17Communication Data Set . . GTD550.V.SYSCOMControl Block . . . . . . TDMFMSESource Volume . . . . . . TDE972System number . . . . . . 01--------------------------------------------------------------------------ComDataSet is in use: GTD550.VH.SYSCOMDISP HEX DUMP CHAR DATA0000 E3C4D4C6 D4E2C514 E3C4D4F2 F0F10001 :TDMFMSE.TDM201 .:0010 B6A2AD16 66956000 B6A2AD16 66956000 :¶sY.Ãn- ¶sY.Ãn- :0020 00000001 3CB52A07 00000377 00015F90 : ..§.. .I .¬°:0030 00015F90 00000BB8 00008000 16514000 : .¬° .½ Ø .é :0040 4DDEB000 0000E406 0000E401 00000000 :(ú^ U. U. :0050 00080008 00016000 164FE000 4DDBD000 : . . .- .|\ (û} :0060 0000E404 0000E401 00000000 00160016 : U. U. . .:Command ===> Scroll ===> CSR F1=Help F2=Split F3=Exit F5=Rfind F7=Back F8=Forward F9=Swap F12=Cancel
Figure 104. COMMDS control block display – Agent system
Option U.10 – Communication data set navigatorUse TDMF TSO Monitor option U.10 to step through the communications data set by entering either theTTRZ or CCHR value and pressing the Enter key. This utility works with both active or completed sessions.
The following figure is an example of the Communication Data Set Navigator utility.
Communication Data Set Navigator Row 1 to 11 of 28Communication Data Set . . GTD550.V.SYSCOMTTRZ: . . . . . . . . . . 00000300CCHR: . . . . . . . . . . 0003------------------------------------------------------------------------Communication Data Set is in use: GTD550.VD.SYSCOMDISP HEX DUMP CHAR DATA0000 E3C4D4C6 D4E2E550 E3C4C5F9 F7F2E3C4 :TDMFMSV TDE972TD:0010 C5F9F7F3 40404040 40400F01 29000BB8 :E973 ... .½:0020 0000D0B0 00000000 FFFFFFFF 01FFFFFF : }^ ........:0030 01FFFFFF 01FFFFFF 01FFFFFF FFFFFFFF :................:0040 FFFFFFFF FFFFFFFF 00000401 92551CD6 :........ ..kè.O:0050 000000AC 85CA7127 00000000 00000000 : De-É. :0070 00000000 00000000 00000000 3C531CFC : .ë.Ü:0080 00000000 00000000 000008D4 F8C00000 : .M8{ :Command ===> Scroll ===> CSR F1=Help F2=Split F3=Exit F5=Rfind F7=Back F8=Forward F9=Swap F12=Cancel
Figure 105. COMMDS navigator display
Chapter 8. Viewing migration information 179
Option U.11 – Communication data set volume refresh bit mapsUse TDMF TSO Monitor option U.11 to display communication data set volume information for completedsessions. This utility is not intended for an active COMMDS.
This utility might display the following types of bit maps for each volume in the session, depending on thetype of situations that are encountered within the session.
• The Miscellaneous Refresh bit map shows locations where compare errors might have occurred.• The Cumulative Refresh bit map indicates those cylinders and tracks that were updated during the
migration session by customer application I/O operations. The number of times a specific track orcylinder was updated is not indicated.
• The FASTCOPY Refresh bit map indicates those cylinders and tracks that were identified as containingdata and will be copied if the FASTCOPY option was specified for this session.
• The Target Empty Tracks bit map shows those cylinder and tracks that are empty (no data is allocatedon them).
• ECAM/DDSR Space Release bit map indicates those cylinders and tracks that were released during asession.
In the following example, one volume had no entries for the Miscellaneous Refresh bit map but hadentries in the Cumulative Refresh bit map.
Display Refresh Bit Maps Row 1 to 34 of 18,984ComDataSet: . . HLQ.HGTD590.SYSCOM------------------------------------------------------------------------ IBM Transparent Data Migration Facility (TDMF) for z/OS Version 5.4.0Report Date and Time : 04/14/2009 13:19:11.43Communication Data Set : HLQ.HGTD590.SYSCOM Created 04/06/2009 16:30:59.99Miscellaneous Refresh Bit Map for volume TD4400 All bit map locations werezero for this volume --------------------------------------------------------------------------Cumulative Refresh Bit Map for volume TD4400Cylinder : 0 (x'0000') Tracks : 1Cylinder : 1 (x'0001') Tracks : 0Cylinder : 36 (x'0024') Tracks : 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14Cylinder : 37 (x'0025') Tracks : 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14Cylinder : 38 (x'0026') Tracks : 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14--------------------------------------------------------------------------Target Empty Tracks Bit Map for volume TD4400Cylinder : 35 (x'0023') Tracks : 9, 10, 11, 12, 13, 14Cylinder : 276 (x'0114') Tracks : 11, 12, 13, 14Cylinder : 277 (x'0115') Tracks : 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14Cylinder : 748 (x'02EC') Tracks : 10, 11, 12, 13, 14Cylinder : 749 (x'02ED') Tracks : 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14--------------------------------------------------------------------------ECAM/DDSR Space Released Bit Map for volume TD4400All bit map locations were zero for this volume
Figure 106. COMMDS volume refresh bit map display
Option U.12 – Communication data set merged system messagesUse TDMF TSO Monitor option U.12 to display messages in date time sequence. Press the PF10 key tospecify a date time filter. To cancel the filter, press PF11. This option also displays the external timereference (ETR) status for a complex.
The following figure shows an example of the Past Session Merged Messages panel.
180 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Past Session Merged Messages Row 1 to 5 of 62Communication Data Set : GTD550.V.SYSCOM--------------------------------------------------------------------------IBM Transparent Data Migration Facility (TDMF) for z/OS Version 5.4.0 Completed Migration04/20/2009 10:07:46.960815 SMFID = TDM1 TOD synchronized on ETR ID 09 TDM1731I The I/O Monitor module was found in Dynamic LPA.04/20/2009 10:07:47.268883 SMFID = TDM1 TOD synchronized on ETR ID 09 TDM1448I The Time-of-Day System Default Option is GMT.04/20/2009 10:07:47.268885 SMFID = TDM1 TOD synchronized on ETR ID 09 TDM1451I The TOD Option has been overridden by the OPTIONS statement and set to LOCAL.04/20/2009 10:07:47.268888 SMFID = TDM1 TOD synchronized on ETR ID 09 TDM1453I The Pacing System Option was specified as OFF.04/20/2009 10:07:47.268890 SMFID = TDM1 TOD synchronized on ETR ID 09 TDM1455I Pacing has been changed to ON by the OPTIONS statement in TDMF job stream.04/20/2009 10:07:47.268892 SMFID = TDM1 TOD synchronized on ETR ID 09 TDM1465I The Operator Messages Option was specified as OFF.04/20/2009 10:07:47.268894 SMFID = TDM1 TOD synchronized on ETR ID 09 TDM1466I The Operator Messages Option has been changed to ON.04/20/2009 10:07:47.313190 SMFID = TDM1 TOD synchronized on ETR ID 09 TDM1380I The application program interface (API) for STK is not available.04/20/2009 10:07:47.506132 SMFID = TDM1 Volume Serial Number = TD6005 TDM1177I The source volume TD6005 is mounted on device 3005 on this system.04/20/2009 10:07:47.506136 SMFID = TDM1 Volume Serial Number = TD6005 TDM1182I At TDMF initialization, the source volume has caching (CFW) activated.Command ===> Scroll ===> CSRF1=Help F2=Split F3=Exit F5=Rfind F7=Back F8=ForwardF9=Swap F10=A_Filter F11=D_Filter F12=Cancel
Figure 107. Past session merged messages
Note: The 160-character line display and normal (80-character) are the same except for the size of theline. The wide display does not contain any new information that is not shown in the 80-character linedisplay.
Option U.13 – Detected source volume I/O errorsUse TDMF TSO Monitor option U.13 to display the locations where an error situation exists for eachvolume within a session.
While TDMF is copying a volume, it checks the CCHH value of each track and cylinder that it is reading andwriting. If the CCHH value does not match the location that is specified by the count key data protocol,TDMF stops the volume migration at that point and scans the remainder of the volume for any otherpotential error situations. When this occurs, it might be due to an invalid count field, in which casemessage TDM3550E is issued, or a physical I/O error on the source volume, in which case messageTDM3536E is issued.
In either case, TDMF does not migrate the volume until these errors are corrected (see “HSM at APARlevel OA21574” on page 83 for corrective procedures).
Note: This utility is not active when a COMMDS is in use.
The following figure shows an example of COMMDS source volume I/O errors.
Chapter 8. Viewing migration information 181
Display Source Volume I/O Errors Row 1 to 14 of 2,038ComDataSet: . . GTD550.V.SYSCOM--------------------------------------------------------------------------- IBM Transparent Data Migration Facility (TDMF) for z/OS Version 5.4.0Report Date and Time : 04/15/2009 12:53:08.96Communication Data Set : GTD550.V.SYSCOM Created 04/02/2009 08:24:09.19--------------------------------------------------------------------------Detected I/O Error locations for volume SPMS84 for SMFID TDM1No errors detected on this volume------------------------------------------------------------------------Detected Invalid Count locations for volume SPMS84 for SMFID TDM1Cylinder : 137 (x'00000089') Tracks : 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14Cylinder : 138 (x'0000008A') Tracks : 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14Cylinder : 139 (x'0000008B') Tracks : 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14Cylinder : 140 (x'0000008C') Tracks : 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14Cylinder : 141 (x'0000008D') Tracks : 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14Cylinder : 242 (x'000000F2') Tracks : 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14
Figure 108. COMMDS source volume I/O error display
Option U.14 – TDMFSECURETDMF TSO Monitor option U.14 uses the ICKDSF TRKFMT utility to provide a source disk erasure failure,which is a standard requirement for overwriting.
TRKFMT is part of the Device Support Facility (DSF) mainframe software. It can be run at the specificcylinder or track range, or full volume level on any IBM disk subsystem. It can also be run while all otherdata on the volume remains online and active to the server or host processor (except in full volumemode). This allows security to immediately clean up and recover from accidental write downs withoutimpact to operations. TRKFMT works on all IBM z Systems.
Number of times to overwrite dataThe following list displays the factors that determine what the number of required overwrites depend on.
• The sensitivity of data• If the client business is regulated• If the storage media are reused
The US Department of Defense requires a minimum of three levels of overwriting for storage media. Toget the three levels requirement by using ICKDSF TRKFMT, specify CYCLES(3).
Duration required to erase dataThe duration that is needed to erase data depends on the level of the erasure. Use of ICKDSF INIT withthe VALIDATE option is the quickest. Use of the TRKFMT ERASE command takes more time. The reasonis because data is not immediately written to disk; it requires a de-stage of cache, which "hardens" thedata on the physical device and, in turn, elongates the process. Regardless, it is important that volumelevel erasure is completed with the volumes offline.
Using TDMFSECUREFrom the TDMFSECURE panel, you can submit a JCL that you can use to initialize a volume, erase thetracks, rewrite the data, and then erase all tracks.
Procedure
1. Select the volumes that you want to submit the JCL for, as shown in the following figure.
182 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Y Select volumes that reached Swap complete? N Select volumes that reached Point-in-time complete? Note. Volumes involved with TCP/IP are ignored. F1=HELP F2=SPLIT F3=END F4=RETURN F5=RFIND F6=RCHANGE F7=UP F8=DOWN
Figure 109. Option U.14 - Generate ICKDSF JCL from communication data set2. Press the Enter key.
The following panel is displayed.
Y Create INIT NOVALIDATE statements? Initialize a volume. N Create INIT VALIDATE statements? Perform erase track on all tracks. N Create TRKFMT statements? Rewrite data on each track prior to performing the erase track. F1=HELP F2=SPLIT F3=END F4=RETURN F5=RFIND F6=RCHANGE
Figure 110. Option U.14 - ICKDSF JCL selection3. Select one or more required options and press Enter.
The following panel is displayed. You are prompted to changes your VTOC and INDEX parametersaccording to your environment settings
You have requested statements to initialize the volume.The VTOC and VTOC INDEX locations and size need to be setprior to submitting the JCL.A global change is needed to change the VTOC and INDEX parms.Example: CHA VTOC(ccccc,hh,###)' 'VTOC(00010,00,15)' ALLExample: CHA VTOC(ccccc,hh,###)' 'INDEX(00010,00,15)' ALLTo use ICKDSF's default parameters a global changewill be needed to remove the VTOC and INDEX parms.Example: CHA 'VTOC(ccccc,hh,###)' '' ALLExample: CHA 'INDEX(ccccc,hh,###)' '' ALL F1=HELP F2=SPLIT F3=END F4=RETURN F5=RFIND F6=RCHANGE F7=UP F8=DOWN F9=SWAP F10=LEFT
Figure 111. Option U.14 - ICKDSF JCL selection4. After you complete the appropriate changes, submit the JCL.
Option S - SMF reportingUse TDMF TSO Monitor option S to view information that is included in the bytes migrated report.
SMF reporting provides additional information to help determine the extent of TDMF processing thatoccurred and the amount of data that remains to be moved as part of a migration process.
Viewing the bytes migrated reportWhen TDMF SMF recording is activated, you can generate a report that lists the amount of data that wasmigrated for specified systems and the time frames that are associated with the migration.
About this task
The bytes migrated report is used to determine the extent of TDMF processing that occurred so far or theamount of data that remains to be moved as part of a migration process.
Procedure
To generate the bytes migrated report, complete the following steps:1. On the TDMF Monitor Menu, select Option S.2. Enter the input that is shown in the following figure.
(The panel that is displayed is a scrolling panel.)
Chapter 8. Viewing migration information 183
Figure 112. Bytes migrated report - panel 13. Press the DOWN arrow key to view the entire screen and the UP arrow to return to the top.
The following figure shows an example of the subsequent panel.
Figure 113. Bytes migrated report - panel 24. Use the More indicator on the upper right side of the panel to determine your position.5. Press Enter.
The scrollable display in the following figure displays the amount of data that is migrated for eachmigration group within the parameter specifications. You can use the SORT primary command to sortthe report into the required order. Storage quantities are reported in the most appropriate unit. Forexample, if 13,463,704,766 bytes were migrated, it would be reported as 12.54 gigabytes, whichequals 13,463,704,766 divided by 1024 to the third power (1 gigabyte) and then rounded at thehundredths digit.
184 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Note: If Process is set to Submit on the panel that is shown in Figure 112 on page 184, this report isgenerated in batch. The report will not contain the command line and other interactive elements, suchas the Scroll input field.
Figure 114. Bytes migrated report - panel 3
The definitions in the Bytes Migrated Report are provided for more information. The panel shown inFigure 114 on page 185 has field-level help available for all fields. Place the cursor on the field inquestion, and press PF1 (HELP) to view field details.
6. Scroll to the last line to see the grand total, as shown in the following figure.
Note: The earliest and latest record time stamps are reported on the total line.
Figure 115. Bytes migrated report - panel 4
Chapter 8. Viewing migration information 185
Bytes migrated report descriptionsUse the fields and filters in TDMF TSO Monitor option S to view bytes migrated reports.
Field definitions
The following table provides descriptions of the fields in the bytes migrated report.
Table 44. Bytes migrated report - field definitions
Field Description
Process Enter E (Execute) to generate the report online within the active TSOsession.
Enter S (Submit) to submit a batch job to generate the report.
Report Live SMF Data? Enter Y (Yes) to add the SMF data in the system buffers to the report. Thebuffered data remains in the buffer. Only buffered data from the currentsystem is used.
Enter N (No) to ignore the buffered SMF data in the report.
SMF Dump DS Name Specifies which logs that are processed by IFASMFDP SMF Dump Utility touse as input. These logs are used in addition to the live SMF data if it isalso reported. The data set names are 44 bytes. Standard wildcardsasterisk (*) and percentage sign (%) are supported.
Report Output DSN Specifies the 44-byte data set name for the report. It can be pre-allocatedwith minimum LRECL of 80 and RECFM FB, FBA, VB, or VBA. If thespecified data set does not exist, it is created. It can be a GDG, PDSmember, or PDSE member. If a data set name is not specified, thefollowing defaults are used:
• Process is set to Execute• Temporary Data Set• Process is set to Submit• SYSTSPRT Data Set• Batch Job Card• Specifies a JCL job card that is used if Process is set to Submit.
TDMF SMF Record ID If possible, the interface determines the SMF record ID. The ID is definedby creating sample job GTDSOPTN.
Other TDMF SMF Record IDs Specifies that SMF record numbers are included in the report.
Filter Options
The following table describes the filters that are available for the bytes migrated report.
Table 45. Bytes migrated report - filter options
Filter Description
TDMF SMFIDs (SYSIDs) Filters the output to include SMF records that are generated on TDMFAgents that are running only on the specified SMFIDs.
Groups Filters the output to include only the specified session groups. Standardwildcards asterisk (*) and percentage sign (%) are supported.
Starting Date Specifies the creation date of the earliest reported data.
186 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Table 45. Bytes migrated report - filter options (continued)
Filter Description
Starting Time Specifies the creation time of the earliest reported data The starting dateis required if a time is specified.
Ending Date Specifies the creating date of the most recent reported data.
Ending Time Specifies the creation time of the most recent reported data. The endingdate is required if a time is specified.
Diagnostic Assistance
There are several things that you can do if you need help diagnosing a problem.
If the SMF Reporter interface does not initialize and you suspect that saved data in your ISPF profilemight be the problem, enter command TSO GTDSMF1 RESET on the ISPF command line from anywherewithin the TDMF ISPF Monitor. This deletes all SMF Reporter dialog variables in your ISPF profile.
If you want to trace the SMF Reporter execution, complete the following steps to send a detailed trace tothe SYSTSPRT
1. Enter the TDMF Monitor in the normal manner.2. Select option "S" to enter "SMF Reporting".3. Enter TSO GTDSMF1 TRACE on the TSO command line.4. Response "Trace active" confirms that the trace is active.5. Enter normal input on the window for desired reporting.6. Enter "B" in the "Process" option for batch job submission. This step is required as any other value
does not trace.7. Ensure that job card fields are correct.8. After verifying all panel input is correct, press Enter. The batch job is submitted. When the job
completes, the SYSTSPRT DD contains the trace output.9. To turn off tracing while still in the TDMF Monitor, enter TSO GTDSMF1 TRACE OFF on the TSO
command line.10. Response "Trace inactive" confirms that the trace is inactive.11. Exiting the TDMF Monitor also turns off the trace.
Option H - TDMF help and message detail facilityUse TDMF TSO Monitor option H to display detailed information for TDMF messages and supportinformation.
This selection opens a window that displays detailed information for TDMF messages, supportinformation, an overview of the TDMF TSO Monitor line commands and a brief introduction to TDMF andthe process necessary to start a session. The following figure shows an example of the help windowl.
Help Selection: 1. Display TDMF Message Details 2. Introduction to TDMF 4. Monitor Line Commands
Command ===> F1=Help F2=Split F3=Exit F9=Swap F12=Cancel
Figure 116. Help Selection window
Chapter 8. Viewing migration information 187
Option H.1 – Display TDMF message detailsUse TDMF TSO Monitor option H.1 to display all TDMF messages within the system. These messagesprovide a more detailed explanation than what is displayed in the TDMF TSO Monitor or in the outputlisting.
To search for a message, use the 4-digit value within the message ID. For example, message TDM3771Wwould be found by searching for 3771, as shown in Figure 117 on page 188.
Message Display Row 1 to 24 of 24Command ===> Scroll ===> CSRWithin the message text there may be special string variablesthat will be replaced when the message is generated.______________________________________________________________Message Number: 3771GTD3771WMessage Text:The devices involved contain a different number of primary cylinders.Explanation:At initial contact with the remote system for this TCP/IP volume migration,it has been determined that the target device for this volume at the remotesystem contains a greater number of primary cylinders. In order to enableallocation of data upon these additional cylinders, and also to resolve anymismatch between the number of alternate cylinders/tracks as found in the VTOCand the number on the physical device, the execution of ICKDSF "REFORMAT"with either the "REFVTOC"or "EXTVTOC" parameter is recommended.This warning indication can be suppressed by using the NOWARNING(3771) optionon the SESSION control statement.System Action:TDMF continues normally.User Response:See the TDMF Installation and Reference Guide for more details.
Figure 117. Message Display example
The last character in a TDMF message ID indicates the severity of the message. The following tableprovides descriptions of the message severity levels.
Table 46. TDMF message severity table
SeverityCharacter
Return Code Description
A Action message: WTO/WTOR option selected; requires aresponse from the MVS console or the TDMF TSO monitor.
I 0000 Informational message: All volume migrations successful; noaction required.
W 0004 Warning message: All volume migrations successful; one or morewarning messages were issued; an action might be required.
E 0008 Error message: One or more volume migrations did not completesuccessfully; a corrective action might be required.
S 0012 Severe error: The session did not start or complete successfully;a corrective action must be taken.
TDMF displays the highest return code for a session in the output listing. For example, if a five-volumesession is started and one volume pairing fails; the return code is 0008.
Option H.2 – Introduction to TDMFTDMF TSO Monitor option H.2 provides a brief overview of TDMF and how it works. It also provides aquick reference list of what is necessary to create, run, and monitor a session.
The following example shows the introduction to TDMF.
188 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Table 47. TDMF introduction
TDMF is a software based "volume" level asynchronous disk-to-disk data replication solution that performsPoint-in-Time copies with minimal disruption to active applications. TDMF also performs non-disruptive"swap" migrations without interruption to active applications. In both cases any storage vendor’s subsystemmay be used.
The SOURCE and Target volumes must be the same geometry (3390 to 3390) however the TARGET volumemay be larger or smaller than the original SOURCE volume.
There are multiple phases during the replication / migration process. When a migration is performed, if allphases complete successfully, the TARGET volume becomes the SOURCE volume. This "swap" migration isperformed using an IOS swap mechanism. If a replication is performed, the TARGET volume is marked offlineand the SOURCE volume remains online. The TARGET volume is then available for other uses, including the useof Offline Volume Access (OVA).
The system on which the replication / migration is controlled is referred to as the 'MASTER' system. Anysystem that may access the SOURCE volume during the replication / migration is called the 'AGENT' and bedefined in the MASTER job. There can be 1 to 63 AGENTS in a session. There must be no activity on theTARGET volume from either the MASTER or AGENT systems.
Terminology
Master the job controlling the copy function and the replication / migrationprocess Agent - associated jobs in a shared DASD environment with theMaster
COMMDS the Communication Data Set: used by the Master and Agent(s) tocommunicate with each other
Source the volume contained data to be replicated / migrated
Target the volume receiving the data being replicated / migrated
VolumePair
the Source and Target volumes for a specific replication / migration
Group two or more volume pairs with the same group name to be synchronizedtogether
Session the Master and Agent(s) sharing the COMMDS with the groups/pairsprocessed in a single job
Sample JCL is available in TDMF for z/OS library SGTDSAMP (which may be accessed using Option 0 under theTDMF for z/OS Function Menu), or may be independently created in another library.
For more information about TDMF for z/OS and defining sessions, please refer to the TDMF for z/OS Installationand Reference Guide.
Option H.4 – Monitor line commandsThe TDMF TSO Monitor does not support the standard ISPF commands on most screen displays. TDMFTSO Monitor option H.4 displays information about the line commands that the TDMF TSO Monitor doessupport and how they might be different from standard ISPF commands.
The following figure shows the TDMF Monitor Line commands panel.
Chapter 8. Viewing migration information 189
Menu Utilities Compilers Help---------------------------------------------------------------------------BROWSE HLQ.IBM.HGTD590.GTDMLIB(MONITORC) Line 00000000 Col 001 080****************************** Top of Data ******************************** TDMF Monitor Line Commands Most screens presented by the TDMF monitor will not perform the normal ISPF commands. Detailed below is the line commands that the TDMF Monitor does support and the differences in their action. TDMF Monitor functions that invoke the ISPF "browse" support all ISPF line commands normally. TDMF Monitor functions that build panel display support the following commands as described. FIND Enter FIND followed by your search argument(s). Unlike ISPF, the TDMF FIND doesn't require a ' (hex '7d') if the user wished to specify more than one word. Both of the following FINDs are valid. FIND copy FIND average seek time Both the FIND argument and data are translated to uppercase prior to the comparison. The line containing the match will be the first line displayed. FIND supports no parameters (ie: PREVIOUS). RFIND (set to PF5). Entering RFIND as a line command will give the ISPF message "RFIND not active". The ISPF RFIND command is NOT passed to the TDMF monitor. The monitor simulates this function by entering FIND with no search argument. Depressing PF5 will present a FIND command with no search argument. When used, the first line searched is the 2nd line on the display. Once at the bottom of the data, the next PF5 will cause the search to wrap to the top. PRINT PRINT is a ISPF command and will only print the current physical screen. TDMFPRT (P is the short command) Unlike ISPF, TDMFPRT will print all the lines contained in that panel to ISPF's print data set. Each invocation of TDMFPRT starts on a new page.
Figure 118. TDMF TSO Monitor line commands
190 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Chapter 9. Migrating GDPS/PPRC HyperSwapvolumes
To migrate GDPS/PPRC HyperSwap volumes successfully, read the following information that includesinstructions on validating and initializing on the Master system, detailed information on HyperSwapgroups, and GDPS controlling systems. In addition, see information on suspending and resuming theHyperSwap environment when the migration is complete.
Overview of quiescing and swapping a volumeTo quiesce and swap a volume in a GDPS/HyperSwap environment, the HyperSwap function itself mustfirst be disabled, and the device UCBs must be unprepared. This process resets the UCBHSWAP flag bits.
TDMF automatically runs the necessary HYPERSW commands to disable HyperSwap for the length of timethat is required to quiesce, synchronize, and swap the migration volumes. In a GDPS xDR for MVSenvironment, the HYPERSW commands are run to the foreign systems through the proxy system.
Validation and initializationAll HyperSwap volumes in a migration session must be grouped so that HyperSwap needs to be disabledonly one time - when all the volumes are ready to be quiesced and synchronized. Use these notes tocomplete the validation and initialization necessary to complete this task.
Validation on the Master system proceeds as follows.
• For the first source or target device of a swap migration that is determined to be HyperSwap prepared(the UCBHSWAP flag is set), the session environment itself is validated.
• The Master system must be part of a parallel sysplex and the user must specify a NetView started taskname on the SESSION control statement (this name must be usable in a Modify command for an activetask in the system).
• The Master activates an extended MCS console by using the name CONSTDMF.
Note: If this name does not meet installation console-naming standards, a different name can besupplied on the SESSION control statement.
This console must be authorized to run Netview and GDPS commands. Only one TDMF console can beactive in the sysplex at any time, so initialization fails if another TDMF session is migrating HyperSwapvolumes. The MODIFY Netview,HYPERSW SYS command is run through the TDMF console andinitialization continues if it receives a valid response to the command.
• The HYPERSW SYS NetView command returns the names of all the active systems in the GDPSplex. TheTDMF Master or an Agent system must be active on every system in the command response, unless thesystem name was specified in the IGNOREGDPS session option. The TDMF Master must not be active ona GDPS controlling system. If an active Agent is in the session on a system that is not part of theGDPSplex, it must not have online HyperSwap prepared devices.
Note: TDMF system names are SMF identifiers, while GDPS uses the SYSNAME specified in theIEASYSxx parmlib member. It is not required that these identifiers are equal to each other; TDMF usesboth names, where appropriate, for each participating system. For a GDPS/PPRC configuration where alldevices are not configured to all the production systems, it is possible to initialize the session withoutan Agent on every active GDPS/PPRC system. You can set up this type of initialization by specifying thesystems that do not have access to ANY of the migration source or target volumes in the IGNOREGDPSsession option. For more information, see “IGNOREGdps” on page 43.
© Copyright IBM Corp. 2010, 2019 191
Important: It is the user's responsibility to run TDMF on all systems that can access the migrationvolumes. Do not add a system to the IGNOREGDPS list if none of the volumes can be physically broughtonline to it.
• All HyperSwap volumes in a migration session must be grouped so that HyperSwap needs to bedisabled only one time - when all the volumes are ready to be quiesced and synchronized. The easiestway to achieve this grouping is to specify OPTion(SINGLEGroup) on the SESSION control statement,which causes the volumes to be migrated in a group called SINGLE. Users who have no experience withmigration volume groups need to be aware that the ACTIVE parameter needs to be specified when theyare limiting the number of concurrent volumes. For more information, see “SESSION options” on page38.
If it is necessary to migrate HyperSwap volumes in two phases, for example to avoid migrating both theprimary and secondary checkpoint, or control data set volume for an application simultaneously, it isnot permissible to define two migration groups that contain HyperSwap controlled volumes in the sameTDMF session. You are required to submit two separate sessions.
• The additional validation for HyperSwap controlled volumes at Agent system initialization consists ofdetermining that the Agent system name was returned to the Master system in response to the MODIFYNetview,HYPERSW SYS command.
HyperSwap groupsTDMF provides this detailed information on HyperSwap groups, which includes the number of volumes ina single session and working with the ICKDSF option.
A HyperSwap group has no known size limit, even if you specify the ICKDSF option.
• You can have as many volumes in a single session as you need to move, if the guidelines for systemvolumes are followed.
• The quiesce process takes the same amount of time that the same-sized group of PIT volumes wouldtake. When ICKDSF sessions are active, jobs are disrupted that need to allocate or extend on thesevolumes or while waiting in TDMFCSEL for below the line storage.
• You can have as many ICKDSF invocations as necessary in a single session. TDMF controls the numberthat are "active" at any one time because ICKDSF uses virtual storage "below the line" (16 MB). Forexample, a 128-volume migration with REGION=500 M might have as many as 20 ICKDSF processesthat are running in parallel. However, the volume level ENQ resources that ICKDSF requires are alwaysobtained by TDMF before volumes are even quiesced. They are then held until ICKDSF processing iscomplete for each volume, which might cause allocation delays. The length of time that ICKDSF isactive for a volume depends on both the responsiveness of the hardware and the size of the VTOC andindex. This is because the VTOC is scanned in its entirety and the index is rebuilt.
GDPS controlling systemsBecause of the way that GDPS/PPRC is structured and controlled, any device swap that occurs on theproduction systems must also take place on the controlling systems. To process this swap, a TDMFMaster or Agent system must be specified on the SESSION control statement for, and run on, every activeGDPS system, including the controlling system.
In practice, the Master system, with its high real-storage usage and intensive I/O activity on the migrationvolumes is ill-suited to a GDPS/PPRC Controlling System. Initialization fails if you attempt to run amigration session in this way.
The need to run TDMF on a GDPS Controlling System runs counter to previously accepted methods andcan cause some configuration difficulties. For example, the batch job or started task that is the TDMFAgent on a controlling system cannot allocate and access the Communications data set that is sharedwith the production systems. To allow this data set to be shared in this way, it must be placed on avolume that is accessible by all the systems on both sites. However, the volume must not be GDPS/PPRCHyperSwap managed.
192 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
A single STEPLIB/GTDKEY library for all the participating systems might be placed on the same volume. Itis more likely that the production systems would share a data set that is placed on a normal PPRCvolume. Each controlling system would have access to its own copy on an infrastructure volume. In thelatter case, the controlling system library copies must be kept at the same maintenance level as theproduction system libraries. And, unless the maintenance procedures allow for a full data set copy fromthe production library, the security keys need to be installed in all these libraries at the same time.
Suspending the HyperSwap environmentWhen a volume group that contains HyperSwap volumes is about to be quiesced, TDMF suspends theGDPS/PPRC HyperSwap environment. There are three ways that TDMF suspends the HyperSwapenvironment.
TDMF starts by passing the GDPS commands to NetView with an internal MODIFY command (for example,MODIFY netview_proc,HYPERSW OFF).
There are three different ways for TDMF to suspend the environment.
• Using the HYPERSW OFF command. This method is always chosen by TDMF if the HSWAPDisableoption is not requested. The command disables HyperSwap and causes GDPS to unprepare all theHyperSwap managed devices. The subsequent HYPERSW ON command might take several minutes,since the devices must be "prepared" again and this function also validates the PPRC paths.
• Using both HYPERSW DISABLE and HYPERSW OFF commands. TDMF uses this command sequence ifthe HSWAPDisable option was requested and the maintenance level of GDPS/PPRC does not supportthe "DISABLE/SWAP/ON" sequence. Again, the HYPERSW OFF command requires the subsequentHYPERSW ON to validate PPRC paths.
• Using the HYPERSW DISABLE command only. TDMF bypasses the HYPERSW OFF command ifOPTion(HSWAPDisable) was specified on the SESSION control statement. See “HSWAPDisable” onpage 43. There is support on every participating LPAR for the GDPS/PPRC HS AND TDMF COEXISTENCEENHANCEMENTS. This support is added by APAR PK69445. Because the HYPERSW OFF command isnot run, the subsequent HYPERSW ON command does not need to prepare the GDPS/PPRC HyperSwapmanaged devices. This results in a much faster return to normal operation.
When the expected responses are received, the volume group is set for quiesce processing.
If the HYPERSW OFF command was run, each TDMF system quiesces the volumes in the group as soon asthe UCBHSWAP flags are reset in the source and target UCBs by GDPS Unprepare processing.
If only the HYPERSW DISABLE command was run, the quiesce is processed immediately after thecommand response is received.
Refer to the GDPS/PPRC Installation and Customization Guide for complete information about thecombination of HYPERSW DISABLE and HYPERSW OFF commands.
Resuming the HyperSwap environmentAs each volume synchronization and swap phase completes, the source and target volumes are resumedas normal.
When all the TDMF Agent systems acknowledge the Master system's resume request for every quiescedHyperSwap volume, the MODIFY Netview,HYPERSW ON command is issued and termination is startedfor the group.
On a controlling system, the source device is left online at termination, instead of being marked offline ason the production systems.
The TDMF Master system waits during termination until HyperSwap Prepared status is detected.However, if there is a problem that is causing this processing to be delayed for more than a few minutes,the session terminates with a warning message.
Chapter 9. Migrating GDPS/PPRC HyperSwap volumes 193
Important: TDMF does not detect the situation where HyperSwap is disabled or suspended after thesession initializes. For this reason, the active TDMF session needs to be terminated before the HYPERSWDISABLE or HYPERSW OFF commands are issued manually. If the PROMPT option is in effect for theTDMF session, it is not necessary to terminate the session; however, do not reply to the TDMFsynchronization prompt until you are ready to turn HyperSwap back on.
HyperSwap while TDMF is activeIf a HyperSwap occurs while TDMF is active, the migration or replication session stops every activeHyperSwap controlled volume when the swap is detected.
During termination, TDMF attempts to clean up the devices that were being migrated. This processinginvolves restoring the original contents of the UCBDDT field for each monitored device and calling theAOMSSDS service to restore FlashCopy functions and bind PAV alias devices.
Due to the possible delay in detection of the device swap and the time that is needed for volumetermination, it is not possible to reestablish FlashCopy soon enough for what would become the PPRCsecondary devices in a HyperSwap RESYNCH. For this reason, a HyperSwap RESYNCH of FlashCopycapable devices fails when TDMF is active with the following message for every failing CESTPAIRoperation:
ANTP0100I DUPLEX OPERATION ERROR. MSG FRMT=0, MSG NMBR=F, REAS: 87
Do not run any planned HyperSwaps while TDMF is active. TDMF cleans up devices after an unplannedHyperSwap (equivalent to a HyperSwap SUSPEND).
Additionally, running the GTDCLUP program on every participating system after a migration that failedbecause of an unplanned HyperSwap would be worthwhile.
Note: A service command can be used to reestablish FlashCopy on one or more devices if TDMFencountered an error during clean-up after a HyperSwap. The GTDCLUP program, withPARM=FLASHCOPY, needs to first be run on all the active systems. Contact technical support forassistance if the problem persists.
Using the TDMF extended MCS console for HyperSwap commandsTo use the TDMF extended MCS console for HyperSwap commands, associate the console with a RACFuser ID and an automation function with authority to run the modify netview and GDPS/PPRCHYPERSW commands.
For more information, see GDPS/PPRC Installation and Customization Guide (ZG24-6703), or GDPS/PPRCHyperSwap Manager V3.3 Installation and Customization Guide (ZG24-6746).
194 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Chapter 10. Migrating basic HyperSwap volumesUse this information to learn how to quiesce and swap a volume in a Basic HyperSwap environment. Inaddition, learn how to resume HyperSwap and run HyperSwap while TDMF is active.
OverviewTo quiesce and swap a volume in a Basic HyperSwap environment, the HyperSwap function itself mustfirst be blocked - preventing both planned and unplanned HyperSwaps. When the HyperSwap volumes inthe session are swapped and terminated, the block is removed.
Validation and initializationUse this information to initialize a volume for HyperSwap and then run validation on the Master system.
• A volume is assumed to be managed by Basic HyperSwap if it is prepared (the UCBHSWAP flag is set),and no GDPS/PPRC GEOCB control block is in storage.
• For the first source or target device of a swap migration that is determined to be Basic HyperSwapmanaged, the session environment itself is validated.
– The blocking interface module (IOSHMBLI) must be present in SYS1.LINKLIB.– The blocking interface must be active on all systems in the sysplex (the IOSHXBLK macro must set a
zero return code for the TESTBLOCK function).– Ensure that only one TDMF session migrates HyperSwap volumes.
Validation on the Master system proceeds as follows:
All HyperSwap volumes in a migration session must be grouped so that HyperSwap needs to be blockedonly one time - when all the volumes are ready to be quiesced and synchronized. You can achieve thisresult by specifying OPTion(SINGLEGroup) on the SESSION control statement, which causes thevolumes to be migrated in a group called SINGLE.
Note: The ACTIVE parameter needs to be specified when you are limiting the number of concurrentvolumes. For more information, see Chapter 10, “Migrating basic HyperSwap volumes,” on page 195, and“SESSION options” on page 38.
Blocking HyperSwap processingWhen a volume group that contains Basic HyperSwap volumes is about to be quiesced, TDMF blocksHyperSwap processing, by using the IOSHXBLK macro.
The block request has an application name of "TDMFV5" and applies to every system in the sysplex. ADISPLAY HS,STATUS command displays TDMFV5 as the application name.
IOSHM0303I HyperSwap Status 923Replication Session: <session name>HyperSwap enabledNew member configuration load failed: DisablePlanned swap recovery: DisableUnplanned swap recovery: DisableFreezeAll: NoStop: NoHyperSwap programmatically blocked by application named TDMFV5
© Copyright IBM Corp. 2010, 2019 195
Resuming the HyperSwap environmentAs each volume synchronization and swap phase completes, the source and target volumes are resumedand terminated as normal.
The IOSHXBLK macro is used to unblock Basic HyperSwap processing when all the HyperSwap managedvolumes are stopped. If the unblock request fails on the Master system, it is retried on the Agent systems.For every failure, TDMF issues message GTD2966E, resulting in a nonzero completion code for themigration. If the request continues to fail, the TDMF session ends without unblocking Basic HyperSwap.
To remove the block after the migration session ends, the GTDCLUP program needs to be run withPARM=UNBLOCK. It is also possible to remove the block with the SETHS UNBLOCK command, but do soonly after you determine that the blocking application is TDMFV5.
For information on GTDCLUP, see “GTDCLIP - return codes” on page 123.
HyperSwap while TDMF is activeIf a HyperSwap occurs while TDMF is active, the migration or replication session stops every activeHyperSwap controlled volume when the swap is detected. During termination, TDMF attempts to clean upthe devices that were being migrated.
This processing involves restoring the original contents of the UCBDDT field for each monitored device,and calling the AOMSSDS service to restore FlashCopy functions and bind PAV alias devices.
A delay in detection might occur of the device swap, and extra time is needed for volume termination.Therefore, it is not possible to reestablish FlashCopy soon enough for what becomes the PPRC secondarydevices in a HyperSwap RESYNCH. For this reason, a HyperSwap RESYNCH of FlashCopy capable devicesfails when TDMF is active with the following message for every failing CESTPAIR operation:
ANTP0100I DUPLEX OPERATION ERROR. MSG FRMT=0, MSG NMBR=F, REAS: 87
Note: Do not process any planned HyperSwaps while TDMF is active. TDMF cleans up devices after anunplanned HyperSwap (equivalent to a HyperSwap SUSPEND). Additionally, run the GTDCLUP program onevery participating system after a migration that failed because of an unplanned HyperSwap.
Note: A service command can be used to reestablish FlashCopy on one or more devices if TDMFencounters an error during clean-up after a HyperSwap. First, run the GTDCLUP program, withPARM=FLASHCOPY on all the active systems. Contact the technical support center for assistance if theproblem persists.
196 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Chapter 11. Migrating EMC autoswap volumesUse this information to learn how to swap an EMC volume that is part of an EMC consistency group andactive in an EMC autoswap environment. In addition, information is provided for suspending and resumingthe EMC environment, and instructions for system automation (SA).
OverviewTo quiesce and swap an EMC volume that is not only part of an EMC Consistency Group, but also active inan EMC AutoSwap environment, extra considerations must be taken. You must also use an extraparameter.
Before you begin
With the EMC AutoSwap enhancement to TDMF Version 4, Release 1 (PTF TD41801), and thecorresponding EMC AutoSwap feature available and active, the duration of the AutoSwap outage isreduced. It is reduced to the length of time that is required to quiesce, synchronize, and swap themigration volumes. Basically, no external EMC commands are required: the Disable and Enablecommands are now automatically created by TDMF.
Unlike GDPS/PPRC in an IBM environment, this support allows the swap migration of multiple EMCvolumes that are part of a single EMC Consistency Group. It also allows multiple volumes within multipleEMC Consistency Groups (by name). Use a single TDMF GROUP statement to minimize the time.
Validation and initializationOne step in validating and initializing your EMC Consistency Group is to ensure that all volumes are a partof an EMC Consistency. Also, all EMC AutoSwap volumes in a migration must be grouped so thatAutoSwap is disabled only one time.
Before you begin
• All volumes must be a part of an EMC Consistency Group. Also, the TDMF source and target volumesthat are identified in a MIGRATE statement must be within the same named EMC Consistency Group.The EMC Consistency Group must be active on every LPAR.
Note: The additional validation for HyperSwap controlled volumes on each system, on every LPARwithin the TDMF session, still applies.
• Additionally, EMC AutoSwap volumes cannot be migrated in the same TDMF session that is alsoattempting to migrate other DASD volumes that are active in an IBM GDPS/PPRC HyperSwapenvironment.
• All EMC AutoSwap volumes in a migration session must be grouped so that AutoSwap needs to bedisabled only one time - when all the volumes are ready to be quiesced and synchronized. You canachieve this result by specifying OPTion(SINGLEGroup) on the SESSION control statement, whichcauses the volumes to be migrated in a group called SINGLE.
Note: The ACTIVE parameter needs to be specified when you limit the number of concurrent volumes.For more information, see “SESSION options” on page 38.
Place the SYSCOM on a volume that is accessible by all participating systems. By doing so, the productionsystems share a single copy of the TDMF load module library, while each controlling system has access toits own copy on an infrastructure volume.
© Copyright IBM Corp. 2010, 2019 197
Suspending the EMC environmentUse these instructions to suspend the EMC environment when volumes belong to one or multiple EMCConsistency Groups.
Before you begin
When a volume group that contains AutoSwap volumes is about to be quiesced, a MODIFYCONGroup,DAS TS GRP * DIS command is run on the TDMF Master system.
Note: The asterisk (*) is used when candidate volumes belong to multiple EMC Consistency Groups. Theasterisk can be replaced with the consistency group name when all volumes are part of the same EMCConsistency Group.
Resuming the autoswap environmentAs each volume synchronization and swap phase completes, the source and target volumes are resumedas normal.
Before you begin
When all the TDMF Agent systems acknowledge the Master system's swap request for every quiescedAutoSwap volume, the MODIFY CONGroup,DAS,TS GRP * ENA command is issued. Next, the resumeand termination phases are started for the group.
Note: The asterisk (*) is used when candidate volumes belong to multiple EMC Consistency Groups. Theasterisk might be replaced with the consistency group name if all volumes are part of the same EMCConsistency Group.
System automation (SA) policyTake these steps to use the TDMF extended commands for automation.
To use the TDMF extended MCS console for AutoSwap commands, it is necessary to associate theconsole with a RACF user ID. You must also grant an automation function the authority to run the modifyCONGroup command, and to run EMC AutoSwap disable and enable commands.
198 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Appendix A. DASD space requirements and serialnumbers
Use this information to learn about DASD space requirements, and what is required to determine theserial number of a specific DASD subsystem.
DASD space requirementsFor DASD space requirements information refer to the TDMF Program Directory.
Determining DASD subsystem serial numberTo determine the serial number of a specific DASD subsystem, MVS command authorization is necessary.You can enter this command from the operator console or by using a product like System Display andSearch Facility (SDSF).
You can enter the DEVSERV QDASD command against a volume to determine the serial number of asubsystem. The following output is an example of the command and response:
DS QDASD,0A90,RCD,1IEE459I 14.51.17 DEVSERV QDASD 338UNIT VOLSER SCUTYPE DEVTYPE CYL SSID SCU-SERIAL DEV-SERIAL EF-CHK0A90 SPMS90 3990Q03 3390A2F 2226 0088 XXF1-30896 XXF1-00000 **OK** READ CONFIGURATION DATAD40101004040F3F3 F9F0C1F2C6C1D4C8 C6F1F0F0F0F0F0F0 F0F0F0F0F0F00010D40000004040F3F3 F9F0C1E7C6C1D4C8 C6F1F0F0F0F0F0F0 F0F0F0F0F0F00000D40200004040F3F9 F9F0D8F0F3C1D4C8 C6F1F0F0F0F0F0F9 F0F3F0F8F9F60000F00000014040F3F9 F9F0F0F0F0C1D4C8 C1F9F0F0F0F0F0F9 F0F3F0F8F9F600000000000000000000 0000000000000000 0000000000000000 00000000000000000000000000000000 0000000000000000 0000000000000000 00000000000000000000000000000000 0000000000000000 0000000000000000 00000000000000008000008000000000 0088C09010000000 0000000000000000 0000000000000000**** 1 DEVICE(S) MET THE SELECTION CRITERIA**** 0 DEVICE(S) FAILED EXTENDED FUNCTION CHECKING
The bold and underscored area is the serial number of the subsystem.
For more information, see IBM 3990/9390 Storage Control Reference (GA32-0274), Read CharacteristicsData.
© Copyright IBM Corp. 2010, 2019 199
200 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Appendix B. Determine the CPU serial numberTo determine the serial number of a specific mainframe, authorization to enter MVS commands isnecessary. The command can be entered from the operator console or by using a product like SystemDisplay and Search Facility (SDSF).
Issue the MVS command Display Matrix for CPU. The command syntax is shown below.
D M = CPU
The operating system issues a Store CPU ID (STIDP) instruction, which returns a doubleword (8 bytes, 16digits) of information. The format of this information is shown.
Table 48. STIDP information format
00 an xxxx mmmm 0000
0 1 2 3 4 7 8 11 12 15
VersionCode
Logical PartitionIdentifier (LPID)
CPU Serial Number CPU Model Number zeros
The operating system displays the following information:
RESPONSE=TDM1 IEE174I 09.24.55 DISPLAY M 528 PROCESSOR STATUS ID CPU SERIAL 00 + 15D77D2084 01 + 15D77D2084 02 - 03 - CPC ND = 002084.C24.IBM.02.0000000FD77D CPC SI = 2084.324.IBM.02.00000000000FD77D CPC ID = 00 CPC NAME = RP55A LP NAME = RALTDM1 LP ID = 15 CSS ID = 1 MIF ID = 5
The serial number that is displayed is 15D77D2084. Using the preceding chart the CPU serial numbermaps out to the following information.
• Version Code is not displayed• 15 is the LPID• D77D is the CPU serial number• 2084 is the CPU model type• Zeros are not displayed
The CPC SI field displays the MTM (machine type and model) information.
For more information on the Store CPU ID instruction, see ESA/390 Principles of Operations (SA22-7201).
© Copyright IBM Corp. 2010, 2019 201
202 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Appendix C. Authorization return codesUse the following information for checking authorization, recording, and reading history files and writingthe TDMF security record.
Authority checkingTDMF includes return codes and their meanings from authority checking.
Table 49. TDMF authority checking return codes
Return Code Description Reason
00
(00)
Passed authority checking withlicense key.
Normal return code. Authorization is good.
04
(04)
Passed authority checking with aTrial Express key.
Normal return code. Authorization is good.Limited to Trial Feature.
08
(08)
Failed authority checking. CPU is not defined to TDMF. Verify that CPU isauthorized to run TDMF.
28
(1C)
Failed authority checking. TDMF feature expired. Verify that key is correct,check expiration date of key, or that key wasinstalled within 5 days of issuance.
64
(40)
Program check Verify input statements. Contact technicalsupport.
516
(204)
Error loading GTDUKEY Verify that correct library is pointed to.
520
(208)
Error locating CPU PCCA Rerun job. Contact technical support.
524
(20C)
Error in caller identification Verify input statements. Contact technicalsupport.
528
(210)
Error in date conversion Verify that the Express statement and key arecorrect. Rerun failing batch job or contacttechnical support.
532
(214)
Error with Enterprise key Verify that key is correct or check expirationdate of key.
536
(218)
Error with Basic Express key Verify that key is correct and the expiration dateof key.
554
(220)
Error with Express key Verify that key is correct and the expiration dateof key.
© Copyright IBM Corp. 2010, 2019 203
Table 49. TDMF authority checking return codes (continued)
Return Code Description Reason
558
(224)
Error with Vendor key Verify that key is correct and that DASDsubsystem ID is valid.
550
(21C)
Error with Trial Express key Verify that key is correct or check expirationdate of key.
552
(228)
Error with version definition Verify that the proper TDMF version is specified.
556(22C)
Program error Verify input statements. Contact technicalsupport.
4096(1000)
No valid keys detected Uninitialized keys data set. Run GTDSOPTNbatch job
Reading and writing the TDMF security recordTDMF includes possible return codes and their meanings from the program for reading and writing theTDMF security record.
Important: The GTDKEY data sets must be defined as PDS type, not LIBRARY; otherwise, you receive areturn code of 204.
Table 50. TDMF security record return codes
Return Code Description Reason
00
(00)
Successful Normal return code. Authorization isgood.
200
(C8)
BLDL error Missing GTDKEY DD statement or dataset does not exist or is misspelled.
204
(CC)
BLDL list in error GTDKEYS load module has multiple textrecords or the format is invalid. Contacttechnical support.
208
(D0)
TTR conversion error Verify that JCL is correct. Contacttechnical support.
212
(D4)
Unsuccessful I/O operation (read) The first eight bits of RC are copied fromthe ECB used for the EXCP request. Referto description of IOBECBCC field in theDFSMS/MVS documentation. Contacttechnical support.
216
(D8)
Unsuccessful I/O operation (write) The first eight bits of RC are copied fromthe ECB used for the EXCP request. Referto description of IOBECBCC field in theDFSMS/MVS documentation. Contacttechnical support.
204 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Table 50. TDMF security record return codes (continued)
Return Code Description Reason
220
(DC)
GETMAIN error Unable to allocate required storage.Rerun job.
224
(E0)
RDJFCB error The data set does not exist or ismisspelled
228
(E4)
Bad Buffer address Verify that JCL is correct. Contacttechnical support.
232
(E8)
Invalid Function code Verify that JCL and input statements arecorrect. Contact technical support.
Reading, updating, and writing TDMF – express feature informationTDMF documents all possible return codes and their meanings from the program for reading, updating orwriting TDMF - Express information.
Table 51. TDMF express return codes
Return Code Description Reason
00
(00)
Successful Normal return code. Authorization isgood.
300
(12C)
BLDL error Missing GTDKEY DD statement or dataset does not exist or is misspelled.
304
(130)
BLDL list in error GTDKEYS load module has multiple textrecords, or its format is invalid. Contacttechnical support.
308
(134)
TTR conversion error Verify that JCL is correct. Contacttechnical support.
312
(138)
Unsuccessful I/O operation (read) The first eight bits of RC are copied fromthe ECB used for the EXCP request. Referto the description of IOBECBCC field inthe DFSMS/MVS documentation. Contacttechnical support.
316
(13C)
Unsuccessful I/O operation (write) The first eight bits of RC are copied fromthe ECB used for the EXCP request. Referto description of IOBECBCC field in theDFSMS/MVS documentation. Contacttechnical support.
320
(140)
GETMAIN error Unable to allocate required storage.Rerun job.
324
(144)
RDJFCB error The data set does not exist or misspelled.
Appendix C. Authorization return codes 205
Table 51. TDMF express return codes (continued)
Return Code Description Reason
328
(148)
Express information error Verify that input statements are valid.Rerun job.
History file recordingTDMF documents all possible return codes and their meanings from the history file recording.
Table 52. TDMF history file return codes
Return Code Description Reason
00
(00)
Successful Normal return code. Authorization isgood.
400
(190)
History logging not installed This option was not specified in theGTDSOPTN batch job. If needed, rerunthe job with this option.
404
(194)
RACF authority checking failed Update the authority that is required forhistory data set. See “Postinstallationtasks” on page 21
408
(198)
ESTAE recovery Contact technical support.
412
(19C)
ENQ for History data set failed An attempt to acquire exclusive control ofthe History data set was not successful; ifRC is 04, the program’s 30 second timeelapsed while you were waiting for thedata set to be available. This code is onlyreturned to APF-authorized callers, forothers, the DYNALLOC SVC runs the ENQand code 424 is returned. Contacttechnical support.
416
(1A0)
RDJFCB error for the SYSCOM data set GTDLOGS completed in error. The reasoncode in bits 0-15 is the RC from theRDJFCB macro request. Evaluate the RCfrom RDJFCB and respond.
420
(1A4)
Parsing of data set name failed GTDLOGS completed in error. Contacttechnical support.
424
(1A8)
DYNALLOC error during allocation of dataset
History data set is in use by another jobor TSO user. Evaluate the RC fromDYNALLOC and respond. Contacttechnical support.
428
(1AC)
RDJFCB error for History Logging data set GTDKEY data set does not exist ormisspelled
432
(1B0)
Open of History Logging data set failed Missing GTDKEY DD statement or dataset does not exist or is misspelled.
206 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Table 52. TDMF history file return codes (continued)
Return Code Description Reason
436
(1B4)
DYNALLOC error during
Deallocation of data set
Missing GTDKEY DD statement or dataset does not exist or is misspelled.
440
(1B8)
History Logging data set definedincorrectly
Incorrect allocation. See memberHISTORY in SGTDSAMP.
444
(1BC)
Invalid call Contact technical support.
448
(1C0)
Error writing log entry Contact technical support.
Appendix C. Authorization return codes 207
208 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Appendix D. Messages for automated operationsThe following messages are intended for use with automated operations packages.
Table 53. Messages for automated operations
Message Number Text
GTD2411A Confirmation requested, reply CANCEL, or volume <volser>.
Explanation A reply is required prior to migration of this volume because this option wasselected.
Action Reply with the volume serial number located in the message or CANCEL if themigration of this volume is not to be allowed.
GTD2412A Confirmation requested, reply CANCEL, or group <groupid>.
Explanation A reply is required prior to migration of this group of volumes because this optionwas selected.
Action Reply with the group name located in the message or CANCEL if the migration ofthis group is not to be allowed.
GTD2413A Ready to synchronize, reply CANCEL, or volume <volser>.
Explanation A reply is required prior to synchronization of this volume because this optionwas selected.
Action Reply with the volume serial number located in the message or CANCEL if themigration of this volume is not to start synchronization.
GTD2414A Ready to synchronize, reply CANCEL or group <groupid>.
Explanation A reply is required prior to synchronization of this group of volumes this optionwas selected.
Action Reply with the group name located in the message or CANCEL if the migration ofthis group is not to start synchronization.
GTD2415E Migration is being terminated by request for volume <volser>.
Explanation This volume migration is being terminated due to operator CANCEL reply tomessage GTD2411A, or via a request from the TSO monitor. Migration of thisvolume continues termination.
Action None
GTD2416E Migrations are being terminated by request for group<groupid>.
Explanation This group of volume migrations is being terminated due to operator CANCELreply to message GTD2412A, or via a request from the TSO Monitor. Migration ofthis volume group continues termination.
Action None
GTD2419I Migration initialization process starting for volume<source-volser>.
Explanation This volume migration is being initialized. Migration of this volume continues.
Action None
© Copyright IBM Corp. 2010, 2019 209
Table 53. Messages for automated operations (continued)
Message Number Text
GTD2420I Migration initialization process starting for group<groupid>.
Explanation This group of volume migrations is being initialized. Migration of this group’svolumes continues.
Action None
GTD2421I Swap Migration process completed successfully for volume<source-volser>.
Explanation This volume migration is being completed successfully.
Action None
GTD2422I Swap Migration has completed successfully for group<groupid>.
Explanation This group of volume’s migrations has completed successfully.
Action None
GTD2423I Swap Migration was not completed successfully volume<source-volser>.
Explanation This volume migration was not completed successfully.
Action Error messages issued; review output.
GTD2424I Swap Migration not completed successfully for group<groupid>.
Explanation This group of volume’s migrations has not completed successfully due to errors.
Action Error messages issued; review output.
GTD2425I Point In Time migration completed successfully volume<volser>.
Explanation This volume migration is being completed successfully.
Action None
GTD2426I Point-In-Time migration completed successfully group<groupid>.
Explanation This group of volume’s migrations has completed successfully.
Action None
GTD2427I Point-In-Time migration was not successful for volume<volser>.
Explanation This volume migration was not completed successfully.
Action Error messages issued; review output.
GTD2428I Point-In-Time migration was not successfully for group<groupid>.
Explanation This group of volume’s migration has not completed successfully due to errors.
Action Review output.
210 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Table 53. Messages for automated operations (continued)
Message Number Text
GTD2579A Allow swap to non-PPRC requested, reply CANCEL, or volume<volser>.
Explanation Confirmation is required for the migration of a PPRC primary volume to a device,which does not seem to have a remote mirroring function active.
Action Reply with the volume serial number located in the message or CANCEL if themigration of this volume is not to be allowed.
GTD2580I PPRC to non-PPRC confirmation received from console<consoleid>.
Explanation This volume’s migration confirmation was received from the console indicated.
Action None
GTD2666E Volume <source-volser> in group <groupid> has beenterminated.
Explanation During a migration, a volume has been prematurely terminated due to an error.The group has option terminate group on error off and auto-ops on.
Action Volume may be re-submitted in another group if necessary.
GTD2802I Volume activity is being resumed for volume <source-volser>.
Explanation This Point-In-Time volume migration has completed synchronization and userI/O requests will be allowed to proceed.
Action None
GTD2803I Volume activity is being resumed for group <groupid>.
Explanation This group Point-In-Time migration has completed synchronization and user I/Orequests will be allowed to proceed on the volumes in the group.
Action None
Appendix D. Messages for automated operations 211
212 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Appendix E. Configurable REXX execsReview the following sample when you create a REXX exec.
Sample REXX execThe following is an example of a REXX exec.
/* REXX *//*********************************************************************//* *//* *//* All user variables are defined in this REXX and passed to TDMFMON *//* *//*********************************************************************//* Get the user's high level qualifier */parse upper arg hlqif hlq = "" then hlq = "hlq.GTD550"Upper hlq/* Define the dataset and variables for TDMF LOAD LIBRARY */address ISPEXEC "LIBDEF ISPLLIB DATASET ID('"hlq".GTDDLLIB')"megtb = "'"hlq" .GTDLLIB'"/* Define the dataset where the security key resides. May be defined to reside in the TDMF load library (definition above) or a different library with like attributes. If the security record is contained in the load library, a definition is still required. *//* Define the SECURITY RECORD *//* secty = "'"hlq" .SECURITY'" */ secty = "'"hlq".GTDLLIB'"/* Define the dataset and variables for TDMF MESSAGE LIBRARY */address ISPEXEC "LIBDEF ISPMLIB DATASET ID('"hlq" .GTDMLIP')"intro = "'"hlq".GTDMLIB(INTRO)'"suptf = "'"hlq".GTDMLIB(SUPPORT)'"moncmd = "'"hlq".GRDMLIB(MONITORC)'"/* Define the dataset TDMF PANEL LIBRARY */address ISPEXEC "LIBDEF ISPPLIB DATASET ID('"hlq".GRDPLIB')"/* Define the dataset and variable for TDMF SAMPLE LIBRARY */wrkfl = "'"hlq".GTDSAMP'"/* Define the dataset TDMF TABLE INPUT LIBRARY */address ISPEXEC "LIBDEF ISPTLIB DATASET ID('"hlq".GTDTLIB')"/* Start the main REXX and pass the user's parameters */address TSO "ALTLIB ACTIVATE APPLICATION(EXEC)", "DATASET('"hlq".GRDELIB') UNCOND"address ISPEXEC "SELECT CMD(%TDMFMON "wrkfl" "megtb" "intro" "suptf" "moncmd" "secty" ) NEWAPPL(GTD) PASSLIB"address TSO "ALTLIB DEACTIVATE APPLICATION(EXEC)"address ISPEXEC "LIBDEF ISPMLIB"address ISPEXEC "LIBDEF ISPLLIB"address ISPEXEC "LIBDEF ISPTLIB"address ISPEXEC "LIBDEF ISPPLIB"address ISPEXEC "LIBDEF GTDKEY"
exit 0
© Copyright IBM Corp. 2010, 2019 213
214 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Appendix F. Session examplesReview this list of examples to help you create sessions such as performing a swap migration, creatingthe communications data set, or running a delayed OVA.
Creating the communications data setTo create the Communications Data Set (COMMDS), member GTDALCCM, which is located in SGTDSAMP,is necessary.
Note: The Communication Data Set must be on a cylinder boundary with contiguous space. It must resideon a device that supports CKD/E.TDMF may attempt to exploit the Parallel Access Volume (PAV) and/orCache Fast Write (CFW) features if available for the Communications Data Set device. This processingcomes into effect for Agent systems if the device supports PAV and there are at least 11 volumes in asession. For the Master system, the trigger for PAV/CFW utilization is more complicated and involves boththe number of volumes and the number of Agent systems; therefore, for a large migration session, aperformance improvement can be expected if the Communications Data Set is allocated on a device thatsupports both PAV and CFW.
Formula for determining data set sizeThe size (number of required cylinders) of the COMMDS can be estimated by running the GTDSPACE REXXcommand found in the SGTDELIB library.
From ISPF Option 6, issue the GTDSPACE command for help by using the GTDSPACE command. If theSGTDELIB library is not concatenated with SYSEXEC or SYSPROC, use the EXECUTE form of the TSOcommand. For example:
dsname(GTDSPACE) where dsname is the data set name of the SGTDELIB library.
Performing a swap migrationThe JCL (MASTER or AGENT) necessary to perform a Swap migration is located in SGTDSAMP.
Single System Swap Migration SessionIf only one system is connected to the DASD devices to be migrated, only one job, MASTER, need besubmitted. In the following example, 16 volumes are to participate in a Swap migration.
//STEP1 EXEC PGM=GTDMAIN,PARM=MASTER,TIME=1440//STEPLIB DD DISP=SHR,DSN=HLQ.IBM.HGTD590.GTDLLIB//GTDKEY DD DISP=SHR,DSN=HLQ.IBM.HGTD590.GTDLLIB//SYSCOM DD DISP=SHR,DSN=HLQ.HGTD590.SYSCOM//SYSPRINT DD SYSOUT=*//DSFPRINT DD SYSOUT=*//SYSUDUMP DD SYSOUT=*//SYSSNAP DD SYSOUT=*//SYSIN DD * SESSION SESSION1 Master(TDM1) SYSCOM(HLQ.HGTD590.SYSCOM) OPTIONS(TIME(LOCAL) PACING(NORMAL) NOAUTOOPS NOCONF CHECKTARGET CONCURRENT(04 ACTIVE) FASTCOPY ICKDSF ) MIGRATE SRC000 TGT100 MIGRATE SRC004 TGT104 MIGRATE SRC008 TGT108
© Copyright IBM Corp. 2010, 2019 215
MIGRATE SRC00C TGT10C MIGRATE SRC001 TGT101 MIGRATE SRC005 TGT105 MIGRATE SRC009 TGT109 MIGRATE SRC00D TGT10D MIGRATE SRC002 TGT102 MIGRATE SRC006 TGT106 MIGRATE SRC00A TGT10A MIGRATE SRC003 TGT103 MIGRATE SRC007 TGT107 MIGRATE SRC00B TGT10B MIGRATE SRC00F TGT10F MIGRATE SRC010 TGT110/*//
In the preceding example, 16 volumes have been defined for swap migrations (MIGRATE statements)with the FASTCOPY option specified. The OPTIONS keyword has set the session defaults, which mayoverride what was specified in the system defaults (GTDSOPTN batch job). The following has been set:Local Time display, Pacing, no WTO/WTOR, no volume confirmation, the target volumes must have nodata on them (only VTOC, VTOCIX and VVDS entries allowed), Active in Copy has been set on and themaximum number of volumes to be concurrently migrating is 4, and the Dynamic ICKDSF option has beenset. Only those options desired need to be specified, all others may take the default.
Multiple system swap migration sessionIf multiple systems are connected to the DASD devices to be migrated, then the MASTER JCL requiressubmission as well as AGENT JCL for each participating system. Using the previous example of 16volumes participating in the Swap migration with the same options specified.
//STEP1 EXEC PGM=GTDMAIN,PARM=MASTER,TIME=1440//STEPLIB DD DISP=SHR,DSN=HLQ.IBM.HGTD590.GTDLLIB//GTDKEY DD DISP=SHR,DSN=HLQ.IBM.HGTD590.GTDLLIB//SYSCOM DD DISP=SHR,DSN=HLQ.IBM.HGTD590.SYSCOM//SYSPRINT DD SYSOUT=*//DSFPRINT DD SYSOUT=*//SYSUDUMP DD SYSOUT=*//SYSSNAP DD SYSOUT=*//SYSIN DD * SESSION SESSION1 Master(TDM1) AGENT(TDM2 TDM3 TDM4) SYSCOM(HLQ.IBM.HGTD590.SYSCOM) OPTIONS(TIME(LOCAL) PACING(NORMAL) NOAUTOOPS NOCONF CHECKTARGET CONCURRENT(04 ACTIVE) FASTCOPY ICKDSF ) MIGRATE SRC000 TGT100 MIGRATE SRC004 TGT104 MIGRATE SRC008 TGT108 MIGRATE SRC00C TGT10C MIGRATE SRC001 TGT101 MIGRATE SRC005 TGT105 MIGRATE SRC009 TGT109 MIGRATE SRC00D TGT10D MIGRATE SRC002 TGT102 MIGRATE SRC006 TGT106 MIGRATE SRC00A TGT10A MIGRATE SRC003 TGT103 MIGRATE SRC007 TGT107 MIGRATE SRC00B TGT10B MIGRATE SRC00F TGT10F MIGRATE SRC010 TGT110/*//
216 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
The following JCL is what would be submitted for each of the participating systems: TDM2, TDM3 andTDM4.
//STEP1 EXEC PGM=GTDMAIN,PARM=AGENT,TIME=1440//STEPLIB DD DISP=SHR,DSN=HLQ.IBM.HGTD590.GTDLLIB//GTDKEY DD DISP=SHR,DSN=HLQ.IBM.HGTD590.GTDLLIB//SYSCOM DD DISP=SHR,DSN=HLQ.IBM.HGTD590.SYSCOM//SYSPRINT DD SYSOUT=*//SYSUDUMP DD SYSOUT=*//SYSSNAP DD SYSOUT=*//SYSIN DD DUMMY
HyperSwap volume migration sessionFor a swap migration in a GDPS/ PPRC HyperSwap environment, the Master system will need tocommunicate with a Netview started task for HyperSwap commands and responses. There must be aTDMF Master or Agent system running on every active GDPS LPAR, including the controlling system(s).
This example migrates eight volumes as a single group (a requirement for HyperSwap managed volumes),relabeling each source volume as $$uuuu, where uuuu is the device number. The optional HYPERSWDISABLE Netview command will be issued before HYPERSW OFF.
The started task name for the internally issued MODIFY operator command is CNMPROC and theauthorized Netview console name to be used is CONSTDMF, which is also the TDMF default.
//STEP1 EXEC PGM=GTDMAIN,PARM=MASTER,TIME=1440//STEPLIB DD DISP=SHR,DSN=HLQ.IBM.HGTD590.GTDLLIB//GTDKEY DD DISP=SHR,DSN=HLQ.IBM.HGTD590.GTDLLIB//SYSCOM DD DISP=SHR,DSN=HLQ.IBM.HGTD590.SYSCOM//SYSPRINT DD SYSOUT=*//DSFPRINT DD SYSOUT=*//SYSUDUMP DD SYSOUT=*//SYSSNAP DD SYSOUT=*//SYSIN DD * SESSION MASTER(TDM1) AGENT(TDM2 TDM3 KSY1) NETVIEW(CNMPROC) CONSOLE(CONSTDMF) OPTIONS(TIME(LOCAL) PACING(NORMAL) RELABEL($$) HSWAPDISABLE CHECKTARGET CONCURRENT(03 ACTIVE) FASTCOPY SINGLEGROUP ) MIGRATE SRC000 TGT100 MIGRATE SRC004 TGT104 MIGRATE SRC008 TGT108 MIGRATE SRC00C TGT10C MIGRATE SRC002 TGT102 MIGRATE SRC006 TGT106 MIGRATE SRC00A TGT10A/*//
Swap migration session with renameUsing the previous example of 16 volumes that participate in the Swap migration with the same optionsspecified. In this case, the original source devices are to be relabeled as if the subsystem were to beremoved. By using XXnnnn as the "new."
Note: This function is not supported in an environment where JES3 manages the devices.
//STEP1 EXEC PGM=GTDMAIN,PARM=MASTER,TIME=1440//STEPLIB DD DISP=SHR,DSN=HLQ.IBM.HGTD590.GTDLLIB//GTDKEY DD DISP=SHR,DSN=HLQ.IBM.HGTD590.GTDLLIB//SYSCOM DD DISP=SHR,DSN=HLQ.IBM.HGTD590.SYSCOM//SYSPRINT DD SYSOUT=*//DSFPRINT DD SYSOUT=*//SYSUDUMP DD SYSOUT=*
Appendix F. Session examples 217
//SYSSNAP DD SYSOUT=*//SYSIN DD * SESSION SESSION1 Master(TDM1) AGENT(TDM2 TDM3 TDM4) SYSCOM(HLQ.IBM.HGTD590.SYSCOM) OPTIONS(TIME(LOCAL) PACING(NORMAL) NOAUTOOPS NOCONF CHECKTARGET CONCURRENT(04 ACTIVE) FASTCOPY ICKDSF ) MIGRATE SRC500 HDF001 XX2300 MIGRATE SRC504 HDF002 XX2304 MIGRATE SRC508 HDF003 XX2308 MIGRATE SRC50C HDF004 XX230C MIGRATE SRC501 HDF005 XX2301 MIGRATE SRC505 HDF006 XX2305 MIGRATE SRC509 HDF007 XX2309 MIGRATE SRC50D HDF008 XX230D MIGRATE SRC502 HDF009 XX2302 MIGRATE SRC506 HDF00A XX2306 MIGRATE SRC50A HDF00B XX230A MIGRATE SRC50E HDF00C XX230E MIGRATE SRC503 HDF00D XX2303 MIGRATE SRC507 HDF00E XX2307 MIGRATE SRC50B HDF00F XX230F MIGRATE SRC510 HDF010 XX2310 or MIGRATE SRC500 HDF001 RELABEL(XX2300)/*//
Swap migration session with promptIn this example, only one volume is in the session as the volume may contain a control data set that has ahigh utilization rate and cannot have a dynamic swap take place due to the application saving UCBinformation.
Since it is not practical to stop such an important application (such as resource sharing) for the length oftime it may take to migrate the volume from start to end, alternative ways may be used. Use of theprompt option will allow the application to remain active during the copy phase. Once TDMF determinesthat the Synchronization Goal can be met, a prompt will be issued notifying the user of this. At that time itis possible to stop the application, respond to the prompt at which time TDMF will pick up the last of theupdates to the source volume, write them to the target volume, and then perform the swap. Once theswap is complete, the application may be restarted. In this way, the time the application is unavailable isa very brief period.
//STEP1 EXEC PGM=GTDMAIN,PARM=MASTER,TIME=1440//STEPLIB DD DISP=SHR,DSN=HLQ.IBM.HGTD590.GTDLLIB//GTDKEY DD DISP=SHR,DSN=HLQ.IBM.HGTD590.GTDLLIB//SYSCOM DD DISP=SHR,DSN=HLQ.IBM.HGTD590.SYSCOM//SYSPRINT DD SYSOUT=*//DSFPRINT DD SYSOUT=*//SYSUDUMP DD SYSOUT=*//SYSSNAP DD SYSOUT=*//SYSIN DD * SESSION SESSION1 Master(TDM1) AGENT(TDM2) SYSCOM(HLQ.IBM.HGTD590.SYSCOM) OPTIONS(TIME(LOCAL) PACING(REVERSE) NOAUTOOPS NOCONF CHECKTARGET ICKDSF ) MIGRATE CDS001 NEWCDS OPT(FAST PROMPT)/*//
218 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Note: There is a second system participating in the session, so an Agent session must be running onTDM2 as well.
Swap migration session with unidentified systems toleranceIn this example, one or more volumes may be connected to other systems not included in the session.
While this may not be a cause for concern, TDMF will issue a warning message (GTD2377W) and post areturn code of 04 at session termination. If a different action is preferred, TDMF may be directed to treatthis condition as an error, which terminates the volume migration (RC=08), or as an informationalmessage (RC=00). Please refer to the “TDMF system defaults” on page 27 for more details.
//STEP1 EXEC PGM=GTDMAIN,PARM=MASTER,TIME=1440//STEPLIB DD DSN=HLQ.IBM.HGTD590.GTDLLIB,DISP=SHR//GTDKEY DD DSN=HLQ.IBM.HGTD590.GTDLLIB,DISP=SHR//SYSCOM DD DSN=HLQ.IBM.HGTD590.SYSCOM,DISP=SHR//SYSPRINT DD SYSOUT=*//DSFPRINT DD SYSOUT=*//SYSUDUMP DD SYSOUT=*//SYSSNAP DD SYSOUT=*//SYSIN DD * SESSION SESSION1 MASTER(TDM1) SYSCOM(HLQ.IBM.HGTD590.SYSCOM) OPTIONS(TIME(LOCAL) PACING(NORMAL) NOAUTOOPS NOCONF NOTERMGROUP CHECKTARGET CONCURRENT(04 ACTIVE) ICKDSF FASTCOPY UNIDENT(W) ) MIGRATE SRC000 TGT100 MIGRATE SRC004 TGT104 MIGRATE SRC008 TGT108 MIGRATE SRC00C TGT10C MIGRATE SRC001 TGT101 MIGRATE SRC005 TGT105 MIGRATE SRC009 TGT109 MIGRATE SRC00D TGT10D/*
Running a point-in-time sessionThe JCL (MASTER or AGENT) necessary to perform a Point-In-Time session is located in SGTDSAMP.
Point-In-Time (PIT) sessions may be used to create an offline backup for later use with a Delayed OVAsession and data center relocations or consolidations. Options such as OVA or Perpetual Point-In-Timemay also be selected with PIT sessions.
Comments have been removed in the examples for brevity.
In the following example, 16 volumes are to participate in a PIT migration session.
//STEP1 EXEC PGM=GTDMAIN,PARM=MASTER,TIME=1440//STEPLIB DD DISP=SHR,DSN=HLQ.IBM.HGTD590.GTDLLIB//GTDKEY DD DISP=SHR,DSN=HLQ.IBM.HGTD590.GTDLLIB//SYSCOM DD DISP=SHR,DSN=HLQ.IBM.HGTD590.SYSCOM//SYSPRINT DD SYSOUT=*//DSFPRINT DD SYSOUT=*//SYSUDUMP DD SYSOUT=*//SYSSNAP DD SYSOUT=*//SYSIN DD * SESSION SESSION1 MASTER(TDM1) AGENT(TDM2) SYSCOM(HLQ.IBM.HGTD590.SYSCOM) OPTIONS(TIME(LOCAL) PACING(NORMAL) NOAUTOOPS
Appendix F. Session examples 219
NOCONF CHECKTARGET FASTCOPY CONCURRENT(04 ACTIVE) PROMPT ) REPLICATE SRC000 TGT100 REPLICATE SRC004 TGT104 REPLICATE SRC008 TGT108 REPLICATE SRC00C TGT10C REPLICATE SRC001 TGT101 REPLICATE SRC005 TGT105 REPLICATE SRC009 TGT109 REPLICATE SRC00D TGT10D REPLICATE SRC002 TGT102 REPLICATE SRC006 TGT106 REPLICATE SRC00A TGT10A REPLICATE SRC003 TGT103 REPLICATE SRC007 TGT107 REPLICATE SRC00B TGT10B REPLICATE SRC00F TGT10F REPLICATE SRC010 TGT110/*//
In the preceding example, 16 volumes have been defined for migrations (REPLICATE statements) withthe FASTCOPY option specified. The OPTIONS keyword has set the session defaults, which may overridewhat was specified in the system defaults (GTDSOPTN batch job). The following has been set: Local Timedisplay, Pacing, no WTO/WTOR, no volume confirmation, the target volumes must have no data on them(only VTOC, VTOCIX and VVDS entries allowed), Active in Copy has been set on and the maximum numberof volumes to be concurrently migrating is 4. Only those options desired need to be specified, all othersmay take the default.
The following example depicts a migration session with all volumes treated as a single group.Additionally, the prompt option has been selected as well as the auto-operations, Active in Copy, andFASTCOPY options.
//STEP1 EXEC PGM=GTDMAIN,PARM=MASTER,TIME=1440//STEPLIB DD DISP=SHR,DSN=HLQ.IBM.HGTD590.GTDLLIB//GTDKEY DD DISP=SHR,DSN=HLQ.IBM.HGTD590.GTDLLIB//SYSCOM DD DISP=SHR,DSN=HLQ.IBM.HGTD590.SYSCOM//SYSPRINT DD SYSOUT=*//DSFPRINT DD SYSOUT=*//SYSUDUMP DD SYSOUT=*//SYSSNAP DD SYSOUT=*//SYSIN DD * SESSION SESSION1 MASTER(TDM1) AGENT(TDM2) SYSCOM(HLQ.IBM.HGTD590.SYSCOM) OPTIONS(TIME(LOCAL) PACING(NORMAL) AUTOOPS NOCONF NOTERMGROUP CHECKTARGET CONCURRENT(04 ACTIVE) PROMPT FASTCOPY SINGLE ) REPLICATE SRC000 TGT100 REPLICATE SRC004 TGT104 REPLICATE SRC008 TGT108 REPLICATE SRC00C TGT10C REPLICATE SRC001 TGT101 REPLICATE SRC005 TGT105 REPLICATE SRC009 TGT109 REPLICATE SRC00D TGT10D REPLICATE SRC002 TGT102 REPLICATE SRC006 TGT106 REPLICATE SRC00A TGT10A REPLICATE SRC003 TGT103 REPLICATE SRC007 TGT107 REPLICATE SRC00B TGT10B REPLICATE SRC00F TGT10F REPLICATE SRC010 TGT110
220 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
/*//
When using the auto-ops option, certain TDMF messages are routed to the MVS console for operatoraction or reply. When volumes are grouped, these messages will usually contain the group name ratherthan the volume serial number. It is recommended that the user supply a meaningful group name,especially if there are multiple TDMF Master jobs active and issuing the same messages. For thisparticular example, the messages will contain the group name SINGLE.
All volumes within a group will synchronize at the same time and can be controlled as a group by theTDMF TSO Monitor. As the name infers, the Single group option groups all the volumes within the samesession. Multiple group names may be employed within the same session, but are mutually exclusive withthe single group option. When a user supplied group name is employed for only one volume, it will bediscarded.
In the next example these volumes will be migrated as a group again, but with a user-supplied groupname.
//STEP1 EXEC PGM=GTDMAIN,PARM=MASTER,TIME=1440//STEPLIB DD DISP=SHR,DSN=HLQ.IBM.HGTD590.GTDLLIB//GTDKEY DD DISP=SHR,DSN=HLQ.IBM.HGTD590.GTDLLIB//SYSCOM DD DISP=SHR,DSN=HLQ.IBM.HGTD590.SYSCOM//SYSPRINT DD SYSOUT=*//DSFPRINT DD SYSOUT=*//SYSUDUMP DD SYSOUT=*//SYSSNAP DD SYSOUT=*//SYSIN DD * SESSION SESSION1 Master(TDM1) agent(TDM2) SYSCOM(HLQ.IBM.HGTD590.SYSCOM) OPTIONS(TIME(LOCAL) PACING(NORMAL) AUTOOPS NOCONF NOTERMGROUP CHECKTARGET CONCURRENT(04 ACTIVE) PROMPT ) GROUP GROUP001 OPT(FAST PROMPT) REPLICATE SRC000 TGT100 REPLICATE SRC004 TGT104 REPLICATE SRC008 TGT108 REPLICATE SRC00C TGT10C REPLICATE SRC001 TGT101 REPLICATE SRC005 TGT105 REPLICATE SRC009 TGT109 REPLICATE SRC00D TGT10D REPLICATE SRC002 TGT102 REPLICATE SRC006 TGT106 REPLICATE SRC00A TGT10A REPLICATE SRC003 TGT103 REPLICATE SRC007 TGT107 REPLICATE SRC00B TGT10B REPLICATE SRC00F TGT10F REPLICATE SRC010 TGT110/*//
The group name and options remain in force until a new GROUP statement is encountered.
Migrate session with OVAIn this example, the selected volumes will be participating in an OVA session once the migration sessionhas completed the synchronization phase and has marked the volumes offline.
//STEP1 EXEC PGM=GTDMAIN,PARM=MASTER,TIME=1440//STEPLIB DD DISP=SHR,DSN=HLQ.IBM.HGTD590.GTDLLIB//GTDKEY DD DISP=SHR,DSN=HLQ.IBM.HGTD590.GTDLLIB//SYSCOM DD DISP=SHR,DSN=HLQ.IBM.HGTD590.SYSCOM//SYSPRINT DD SYSOUT=*//DSFPRINT DD SYSOUT=*//SYSUDUMP DD SYSOUT=*
Appendix F. Session examples 221
//SYSSNAP DD SYSOUT=*//SYSIN DD * SESSION SESSION1 MASTER(TDM1) AGENT(TDM2) SYSCOM(HLQ.IBM.HGTD590.SYSCOM) OPTIONS(TIME(LOCAL) PACING(NORMAL) NOAUTOOPS NOCONF CHECKTARGET ) REPLICATE DBS435 XX7856 OPTIONS(PROMPT FAST OVA)/*//
In the following OVA sample, a full DFDSS volume copy has been selected. DBS435 is involved in amigration (it is also listed on the INCLUDE list, for clarity). The offline target pair of DBS435 (at sync time)will be input for the copy to XX1975. XX1975 is listed on the EXCLUDE list because it is the output in theOVA JCL (VOL2 DD statement). If the EXCLUDE statement is not used, the OVA job will assume thatXX1975 is part of another migration session.
//IPGM EXEC PGM=GTDIPGM,REGION=2047M//STEPLIB DD DISP=SHR,DSN=HLQ.IBM.HGTD590.GTDLLIB//IPGMOUT DD SYSOUT=*//SYSUDUMP DD SYSOUT=*//IPGMIN DD * PROGRAM ADRDSSU name INCLUDE DBS435 EXCLUDE XX1975//SYSPRINT DD SYSOUT=*//VOL1 DD DISP=SHR,UNIT=3390,VOL=SER=DBS435//VOL2 DD DISP=SHR,UNIT=3390,VOL=SER=XX1975//SYSUDUMP DD SYSOUT=Y//SYSIN DD * COPY FULL INDDNAME(VOL1) OUTDDNAME(VOL2)/*
//
PIT session with PPITPPIT sessions may be used during a batch cycle to allow for multiple synchronization points for fallbackpurposes. Once the volumes have been synchronized, the target volumes are marked offline. TDMF willcontinue to monitor updates to the source volumes so that at a later time, the target volume may beupdated to a new synch time.
OVA has been selected in this example in order to take periodic backups to tape as well as having "hot"backup volumes available, if desired.
In this example, the selected volumes will be participating in a PPIT session once the migrate session hascompleted the synchronization phase and marked the volumes offline.
//STEP1 EXEC PGM=GTDMAIN,PARM=MASTER,TIME=1440//STEPLIB DD DISP=SHR,DSN=HLQ.IBM.HGTD590.GTDLLIB//GTDKEY DD DISP=SHR,DSN=HLQ.IBM.HGTD590.GTDLLIB//SYSCOM DD DISP=SHR,DSN=HLQ.IBM.HGTD590.SYSCOM//SYSPRINT DD SYSOUT=*//DSFPRINT DD SYSOUT=*//SYSUDUMP DD SYSOUT=*//SYSSNAP DD SYSOUT=*//SYSIN DD * SESSION SESSION1 MASTER(TDM1) AGENT(TDM2) SYSCOM(HLQ.IBM.HGTD590.SYSCOM) OPTIONS(TIME(LOCAL) PACING(NORMAL) NOAUTOOPS NOCONF CHECKTARGET PROMPT SINGLE )
222 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
REPLICATE DBS001 XX1001 OPTIONS(PROMPT FAST OVA PPIT) REPLICATE DBS078 XX1012 OPTIONS(PROMPT FAST OVA PPIT) REPLICATE DBS080 XX10F3 OPTIONS(PROMPT FAST OVA PPIT) REPLICATE DBS105 XX10C4 OPTIONS(PROMPT FAST OVA PPIT)/*//
Running delayed OVAThe JCL for Delayed OVA (DOVA) is located in SGTDSAMP. All that is required for this type of session is aCommunications Data Set(COMMDS), which contains volumes that successfully completed a PIT session.
An example follows:
//STEP1 EXEC PGM=GTDDOVA//STEPLIB DD DISP=SHR,DSN=HLQ.IBM.HGTD590.GTDLLIB//GTDKEY DD DISP=SHR,DSN=HLQ.IBM.HGTD590.GTDLLIB//SYSCOM DD DISP=SHR,DSN=HLQ.IBM.HGTD590.SYSCOM//SYSPRINT DD SYSOUT=*//SYSUDUMP DD SYSOUT=*//SYSSNAP DD SYSOUT=*//
Once this session has initialized, a normal OVA batch job may be submitted. If an OVA batch job is notsubmitted within the standard 15 minute interval, the Delayed OVA job will terminate (normally).
Performing a SCAN ONLYThe SCAN parameter allows the validation of options, volumes, and the migration environment, withoutcopying the volume. During a scan, the TDMF copy task does not start and the target volume remainsunchanged. If the WTO option is specified, you receive automated operations messages (GTDxxxxA). Ifthe Confirmation option is specified, this prompt needs to be satisfied before SCAN starts.
The SCAN feature assists users in preparing for migrations by ensuring that all options that are selectedare valid, and that the source and target volume pairing is valid.
The SCAN ONLY can be specified in the following ways:
• By specifying a //GTDSCAN DD DUMMY Statement in the JCL stream.• By specifying SCAN on the EXEC statement in the JCL stream, such as PARM=MASTER,SCAN
When the SCAN parameter is specified as shown in the following example JCL, TDMF simulates theINITIALIZATION and ACTIVATION phases. As a result of this simulation, you receive the normal TDMFmessages. You also receive the message GTD2722I, Volume termination requested by "SCAN ONLY."
The //GTDSCAN DD DUMMY statement is shown in Figure 119 on page 224.
Appendix F. Session examples 223
JOBCARD//STEP1 EXEC PGM=GTDMAIN,PARM='MASTER,SCAN'//STEPLIB DD DISP=SHR,DSN=HLQ.IBM.HGTD590.GTDLLIB//GTDKEY DD DISP=SHR,DSN=HLQ.IBM.HGTD590.GTDLLIB//SYSCOM DD DISP=SHR,DSN=HLQ.IBM.HGTD590.SYSCOM//SYSPRINT DD SYSOUT=*//*GTDSCAN DD DUMMY//DSFPRINT DD SYSOUT=*//SYSUDUMP DD SYSOUT=*//SYSSNAP DD SYSOUT=*//SYSIN DD * SESSION SESSION1 MASTER(TDM1) SYSCOM(HLQ.IBM.HGTD590.SYSCOM) OPTIONS(PROMPT PURGE COMPARE NOCHECKT FASTCOPY CONC(03 ACTIVE) ) GROUP TESTGRP OPT(NOPU NOCOMPARE) MIGRATE TD239A TD239B/*//
Figure 119. Sample SCAN ONLY JCL
SCAN only to create DS8000 Heat Map Transfer Utility input file creation
TDMF can create an XML or standard z/OS file that contains control cards as DS8000 Heat Map TransferUtility (HMTU) input for initiating a heat map move from the source to the target device. HMTU must beordered and installed separately. It is a Unix based application and can be run under the z/OS BPXBATCHUnix System Services shell. TDMF contains two sample JCL members that demonstrate how to generatethe HMTU control cards (GTDHTMU1) and a sample of how to run the HTMU using BPXBATCH(GTDHTMU2). GDPS TDMF users should use the GDPS provided REXX EXEC to run HTMU.
Setting up a migration using TCP/IPUse this information to help you set up a migration with Remote Master and Local Master JCL, andRemote Agent and Local Agent JCL.
REMOTE master JCLWhen performing TCP/IP migration to a remote system, a port number must be specified for the remotemaster system that is the target of a transparent migration using TCP/IP.
Important: The port number specified in the remote master job must match the port number specified inthe local master job.
224 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
JOBCARD//*
//* +-------------------------------------------------------------------------------------------------+
//* | |//* | THIS MASTER JOB IS USED ON THE REMOTE SYSTEM OF A TCP/IP |//* | CONNECTION. THE PORT NUMBER SPECIFIED ON THE REMOTE MASTER, |//* | MUST MATCH THE PORT NUMBER SPECIFIED ON THE LOCAL MASTER. |//* | THE PORT NUMBER MUST BE IN THE RANGE OF 1-65534. THE SYSCOM |//* | USED FOR THE REMOTE MASTER MUST ALSO BE USED FOR ANY REMOTE |//* | AGENT(S). |//* | |
//* +-------------------------------------------------------------------------------------------------+
//*//RMASTER EXEC PGM=GTDMAIN,PARM='MASTER,PORT=8888'//STEPLIB DD DSN=HLQ.IBM.HGTD590.GTDLLIB,DISP=SHR//GTDKEY DD DSN=HLQ.IBM.HGTD590.GTDLLIB,DISP=SHR//SYSCOM DD DSN=GTD550.REMOTE.SYSCOM,DISP=SHR//SYSPRINT DD SYSOUT=*//DSFPRINT DD SYSOUT=*//SYSUDUMP DD SYSOUT=*//SYSSNAP DD SYSOUT=*//SYSIN DD * SESSION SESSION1 MASTER(TDM2) AGENT(DRM1) OPT(TIME(LOCAL))/*//
Figure 120. Remote master JCL example
Guidelines for creating remote master JCLKeep in mind the following when creating remote master JCL for a TCP/IP migration:
• The PORT=port parameter on the EXEC statement indicates to TDMF that this is a remote session.• MIGRATE control statements cannot be used in the remote Master input stream; the migration volumes
can be defined only on the local session.• The agent system (in this example, DRM1) on the remote session does not perform normal agent
functions (source volume updates monitoring and notification). In the case of a remote master job for aTCP/IP migration, the agent system only protects the target volume from unauthorized access andvaries it offline at termination time.
• Make sure to have the GTDKEY DD statement mentioned in the JCL and pointing to the GTDKEYmember in a valid library.
Appendix F. Session examples 225
LOCAL master JCLBefore a TCP/IP migration is initiated, a LOCAL master job must be submitted that identifies a localmaster system as a sending system that uses TCP/IP.
Important: The port number that is specified in the remote master job must match the port number thatis specified in the IP statement of the local master job.
JOBCARD//*
//* +-------------------------------------------------------------------------------------------------+
//* | |//* | THIS MASTER JOB IS USED ON THE LOCAL SYSTEM OF A TCP/IP |//* | CONNECTION. THE PORT NUMBER SPECIFIED ON THE LOCAL MASTER, |//* | MUST MATCH THE PORT NUMBER SPECIFIED ON THE REMOTE MASTER. |//* | THE PORT NUMBER MUST BE IN THE RANGE OF 1-65534. THE SYSCOM |//* | USED FOR THE LOCAL MASTER MUST ALSO BE USED FOR ANY LOCAL |//* | AGENT(S). |//* | |
//* +-------------------------------------------------------------------------------------------------+
//*//LMASTER EXEC PGM=GTDMAIN,PARM=MASTER//STEPLIB DD DSN=HLQ.IBM.HGTD590.GTDLLIB,DISP=SHR//GTDKEY DD DSN=HLQ.IBM.HGTD590.GTDLLIB,DISP=SHR//SYSCOM DD DSN=GTD550.LOCAL.SYSCOM,DISP=SHR//SYSPRINT DD SYSOUT=*//DSFPRINT DD SYSOUT=*//SYSUDUMP DD SYSOUT=*//SYSSNAP DD SYSOUT=*//SYSIN DD * SESSION SESSION1 MASTER(TDM1) AGENT(BRO1) OPT(FAST UNIDENT(W)) REMOTE TESTING HOSTNAME(prod.lpar7-sys14.image68.ibm.com) PORT(8888) REPLICATE SCR000 TGT100 REMOTE(TESTING)//
Figure 121. Local master JCL example
226 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Remote agent JCLIn the remote agent JCL, the SYSCOM data set name must match the SYSCOM data set name in theremote master JCL.
JOBCARD1JOBCARD2/*JOBPARM SYSAFF=DRM1
//* +-------------------------------------------------------------------------------------------------+
//* | |//* | THIS AGENT JOB IS USED ON THE REMOTE SYSTEM OF A TCP/IP |//* | CONNECTION TO ENSURE DATA INTEGRITY FOR THE REMOTE VOLUME(S). |//* | THE SYSCOM USED FOR THE REMOTE MASTER MUST ALSO BE USED FOR |//* | THE REMOTE AGENT(S). |//* | |
//* +--------------------------------------------------------------------------------------------------+
//*//RAGENT EXEC PGM=GTDMAIN,PARM=AGENT//STEPLIB DD DSN=HLQ.IBM.HGTD590.GTDLLIB,DISP=SHR//GTDKEY DD DSN=HLQ.IBM.HGTD590.GTDLLIB,DISP=SHR//SYSCOM DD DSN=GTD550.REMOTE.SYSCOM,DISP=SHR//SYSPRINT DD SYSOUT=*//SYSUDUMP DD SYSOUT=*//SYSSNAP DD SYSOUT=*
Figure 122. Remote agent JCL example
Local agent JCLIn the local agent JCL, the SYSCOM data set name must match the SYSCOM data set name in the localmaster JCL.
JOBCARD/*JOBPARM SYSAFF=BRO1
//* +-------------------------------------------------------------------------------------------------+
//* | |//* | THIS AGENT JOB IS USED ON THE LOCAL SYSTEM OF A TCP/IP |//* | CONNECTION TO ENSURE DATA INTEGRITY FOR THE LOCAL VOLUME(S). |//* | THE SYSCOM USED FOR THE LOCAL MASTER MUST ALSO BE USED FOR |//* | THE LOCAL AGENT(S). |//* | |
//* +-------------------------------------------------------------------------------------------------+
//*//LAGENT EXEC PGM=GTDMAIN,PARM=AGENT//STEPLIB DD DSN=HLQ.IBM.HGTD590.GTDLLIB,DISP=SHR//GTDKEY DD DSN=HLQ.IBM.HGTD590.GTDLLIB,DISP=SHR//SYSCOM DD DSN=GTD520.LOCAL.SYSCOM,DISP=SHR//SYSPRINT DD SYSOUT=*//SYSUDUMP DD SYSOUT=*//SYSSNAP DD SYSOUT=*
Figure 123. Local agent JCL example
Appendix F. Session examples 227
228 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Appendix G. How to read syntax diagramsReview the following information to help you create and decipher syntax diagrams.
Reading syntax diagramsTo read syntax diagrams, follow one line at a time from the beginning to the end, and code everything youencounter on that line.
The following conventions apply to all syntax diagrams for control cards:
• Read the syntax diagrams from left to right and top to bottom.• Each syntax diagram begins with a double arrowhead (►►) and ends with opposing arrowheads (►◄).• An arrowhead (►) at the end of a line indicates that the syntax continues on the next line. A continuation
line begins with an arrowhead (►).• Commands and keywords are shown in uppercase and lowercase letters. The uppercase portion is the
minimum needed to code the command properly; the lowercase portion is optional. For example,FASTCOPY can be coded in any of the following ways: FAST, FASTC, FASTCO, FASTCOP, or FASTCOPY.
Note: Commands must be entered in uppercase. Lowercase is not recognized.• Some commands and keywords have alternative abbreviations; these appear as part of the stack for
that command or keyword. For example, the alternative abbreviation for TARget is TGT.
TARget
TGT
(target_volume)
• Words in all lowercase letters represent information you supply. For example, target_volume, orseconds.
• You must provide all items enclosed in parentheses, ( ), and you must include the parentheses.• Where you can choose from two or more keywords, the choices are stacked one above the other. If one
choice within the stack lies on the main path, you must choose a keyword. In the following example youmust choose either target_volume, TARGET, or TGT.
target_volume
duplex_target
remote_target
remote_target
TARget
TGT
(
local_target remote_target
duplex_target
)
If one or more keywords are below the main path, they are optional. You can choose one, the other, ornone. In the following example SINGLEgroup is an optional keyword.
SINGLEgroup
If a stack of keywords is below the main path and one keyword is above the main path, the use of thekeyword is optional, and the above item is the default. In the following example, if no keywords arespecified, the default NOPUrge is taken.
© Copyright IBM Corp. 2010, 2019 229
NOPUrge
PUrge
• The repeat symbol appearing above keywords and variables indicates that you can specify thosekeywords and variables more than once. If a comma appears in the repeat symbol, you must separaterepeated keywords or variables with a comma or a blank. For example, after the keyword Agents, youcan enter multiple system identifiers separated by commas.
Agents (
,
sysid )
• Substitution blocks are used to simplify the diagrams. They indicate that blocks of the syntax diagramare located outside of the main diagram. You insert the keywords for that block where the symbolappears, and return to the main diagram to continue with the command. This technique is used forOptions.
230 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Appendix H. Operating the z/VM Agent for TDMFThe TDMF Server virtual machine provides commands to start and manage migration sessions. Thecommands can be entered locally (from within the same z/VM system) or remotely (via NJE or FTP).
Note: In contrast to a TDMF for z/OS Agent job, a migration session on z/VM runs as a thread in the TDMFServer virtual machine running zCMS.
Starting the z/VM Agent for TDMFThe z/VM Agent for TDMF typically runs in the disconnected TDMF Server virtual machine.
About this task
This TDMF Server virtual machine must be active during TDMF volume migrations that include volumesonline to the z/VM system.
You can add the TDMF Server virtual machine to the VM AUTOLOG process that starts server virtualmachines when the z/VM system is IPLed. If the TDMF Server virtual machine is not started, use theXAUTOLOG command to start it. To start the virtual machine for the tdmfvm user ID, enter:
xautolog tdmfvm
Querying operational statusYou can query the operational status of the z/VM Agent for TDMF by entering several commands.
About this task
Some components of the z/VM Agent for TDMF are loaded into the CP system execution space (SXS)when the TDMF Server virtual machine is started. You can verify the status of these components byentering the QUERY CPXLOAD command.
query cpxload epn gtdvsm
Loaded as identifier 14 File: GTDLOAD CNTL Z Loaded by: TDMFVM on 09/04/16 at 07:06:53 CPXLOAD options: NODELAY NOCONTROL TEMPORARY Associated entry points and external symbols: GTDVDIAG is at 51ABD390 GTDVDI is at 51ABD000 module size is 0027B0 GTDVSMIN is at 51AB3400 GTDVSMCL is at 51AB3400 GTDVSM01 is at 51AB3498 GTDVSMAD is at 51AB38A0 GTDVSMDE is at 51AB3A50 GTDVSMDY is at 51AB3BE8 GTDVSMST is at 51AB3C18 GTDVSMSY is at 51AB3CA0 GTDVSMIR is at 51AB3D90 GTDVSM is at 51AB3000 module size is 001588Ready; T=0.01/0.01 05:06:13
The z/VM Agent for TDMF uses the TDMF diagnose code that is assigned through the DIAG parameter inthe TDMFVM configuration file. To verify that this diagnose code is enabled, enter the q diag 130 CPcommand (assuming 130 is the hexadecimal number of the diagnose code that is assigned by theconfiguration file):
q diag 130
Code Status EPName PrivClasses
© Copyright IBM Corp. 2010, 2019 231
0130 Enabled GTDVDIAG <ANY>Ready; T=0.01/0.01 05:11:40
To verify that the TDMF Server virtual machine is started and disconnected, enter the following CPcommand (assuming TDMFVM is the userid that is used for the server virtual machine):
q tdmfvm
TDMFVM - DSCReady; T=0.01/0.01 09:28:46
Using a secondary user to communicate with the z/VM Agent for TDMFA secondary user can directly communicate with the z/VM Agent for TDMF while the TDMF Server virtualmachine runs disconnected by using the CP SEND command.
About this task
To see the currently assigned secondary user ID for the TDMF Server virtual machine, TDMFVM issue thiscommand.
q secuser tdmfvm
SecondaryUserid Userid StatusTDMFVM IBM logged onReady; T=0.01/0.01 09:31:26
You can define a default secondary userid in the VM directory. To dynamically associate a userid as asecondary user after the primary user virtual machine disconnects, enter the SET SECUSER command.For example, to associate userid IBM as a secondary user, enter:
set secuser tdmfvm ibm
A secondary user can verify whether the TDMF Server virtual machine is operational by entering the sendcommand with the tdmfvm and status parameters:
send tdmfvm status
send tdmfvm status Ready; T=0.01/0.01 05:07:05 TDMFVM : GTD6004I command accepted: STATUS TDMFVM : GTD6127I the console log will be sent to IBM at BOULDER TDMFVM : GTD6129I abend dumps will be sent to OPERATNS TDMFVM : GTD6078I no active sessions found TDMFVM : >
Starting a migration sessionAll host systems that participate in a migration session (the TDMF Master and all TDMF Agents sharing theCOMMDS) must connect and confirm readiness within approximately 15 minutes before the actualvolume migration starts.
About this task
Use the SESSION command to connect the z/VM Agent for TDMF to a migration session.
SESSION commds vdev
RDEV= rdev
VOL= volser
Note: Before you connect the z/VM Agent for TDMF to a volume migration session, ensure that allreferenced target volumes are online and free on the z/VM system.
232 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
commdsThe communication data set name (COMMDS).
vdevThe virtual device address of the COMMDS volume (which already must be assigned to the TDMF/VMServer virtual machine).
rdevThe real device address of the COMMDS volume.
Note: The TDMF/VM Server virtual machine tries to automatically assign the device to itself. It alsounassigns it after the session completes. The CP commands ATTACH and DETACH are used to assignand unassign a COMMDS volume.
volserThe volume serial number of the COMMDS volume.
Note: The TDMF/VM Server virtual machine tries to automatically assign the device to itself. It alsounassigns it after the session completes. The CP commands ATTACH and DETACH are used to assignand unassign a COMMDS volume).
Disconnecting from an active migration sessionAn active session can be ended by entering a CANCEL command.
About this taskCANCEL commds
commdsSpecifies the data set name of the COMMDS (communication data set) used by the migration sessionto be stopped.
Terminating the z/VM Agent for TDMFYou can stop the TDMF/VM Agent running in the TDMF/VM Server virtual machine by entering the STOPcommand.
About this taskSTOP
FORCE
STOP terminates the TDMF/VM agent. However, an active migration session prevents termination of theTDMF/VM agent. By specifying FORCE, all active migration sessions are stopped and the TDMF/VM agentis terminated.
Retrieving logging informationThe virtual console of the TDMF/VM server virtual machine is closed. It is then transferred back to theissuer of the GETCONS command, or the user that is specified by the LOG_USER and LOG_NODEconfiguration parameters if the issuer cannot be determined.
About this taskGETCONS
Appendix H. Operating the z/VM Agent for TDMF 233
Note: RSCS is used to transfer the console log over the NJE network in case the command was enteredremotely.
Displaying status informationYou can display various information about the status of the TDMF/VM server virtual machine by using theSTATUS command. This information includes a list of currently active migration sessions.
About this taskSTATUS
ExampleResulting output (common)
GTD6004I command accepted: STATUS GTD6139I z/VM Agent for TDMF loaded at 0xepaddr GTD6127I the console log will be sent to log_user at log_node GTD6129I abend dumps will be sent to dump_user
epaddrIs the load address of the GTDVMAIN MODULE within the TDMF Server virtual machine.
log_userIdentifies the user to whom the console log of the TDMF Server virtual machine is sent in responseto a GETCONS command.
log_nodeIdentifies the NJE node to which the console log of the TDMF Server virtual machine is sent inresponse to a GETCONS command.
dump_userIs the ID of the local VM user to which dumps are transferred in case of an abend of the TDMFServer virtual machine.
Resulting output 1GTD6078I no active sessions found
Resulting output 2
GTD6076I SYSCOM MASTERGTD6077I commds smfid
commds
Is the name of the communications data set that is being used.
smfidIdentifies the SMFID of the TDMF Master z/OS system.
Displaying active migration sessionsYou can display active migration sessions by using the QUERY SESSION command.
About this taskQUERY SESSION
ALL
commds
234 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
commdsThe name of the communications data set that identifies the migration session to be displayed.Omitting this parameter, or specifying ALL, displays information about all currently active migrationsessions.
ExampleResulting output 1
GTD6150I no active session found
Resulting output 2GTD6149I session commds waiting for master
commdsIs the name of the communications data set that is being used.
Resulting output 3GTD6148I commds volsrc(devs) voltgt(devt) xxxxxxxx error
commdsIdentifies the name of the communications data set that is being used for this migration session.
volsrcIs the volume label of the migration source volume.
devsIs the real device number of the migration source volume.
voltgtIs the volume label of the migration target volume.
devtIs the real device number of the migration target volume.
errorIs the latest error message that is issued on behalf of the source and target migration-volumepair.
Displaying maintenance level of z/VM Agent for TDMF softwareYou can display the applied software maintenance for components of the z/VM Agent for TDMF productby using the QUERY LEVEL command.
About this taskQUERY LEVEL
ALL
name
nameSpecifies a valid component name for which the maintenance level is displayed. Omitting thisparameter, or specifying ALL, displays the maintenance level for all components.
ExampleResulting output
GTD6154I name EP=0xepaddr LVL=aparlvl BUILD=yyyy/mm/dd (hh:mm)
name
Identifies the component name.
Appendix H. Operating the z/VM Agent for TDMF 235
epaddr
The entry point address in hexadecimal.
aparlvl
Identifies the applied maintenance level (APAR identifier). Components without maintenanceapplied display the base identifier HGTD590.
yyyy/mm/dd (hh:mm)
The date and time the component was built.
Toggling debug messagesIf problems occur when you run the z/VM Agent for TDMF, you can do one of two things.
About this task
• Start the TDMF Server virtual machine in debug mode by specifying the -d runtime option.• Enter the DEBUG command while the TDMF Server virtual machine is already running.
DEBUG
Changing configuration settingsYou can change several configuration settings while the TDMF Server virtual machine is active by usingthe SET command.
About this taskSET LOG_USER user_id
DUMP_USER user_id
LOG_NODE node_id
*
user_idSpecifies the default target userid that receives console logs requested by the GETCONS command orthe local userid to which abend dumps are transferred.
node_idSpecifies the default NJE node name of the target system that receives console logs requested by theGETCONS command.
Remotely operating the z/VM Agent for TDMFInstead of directly entering commands on the virtual console of the TDMF Server virtual machine (or by asecondary user that uses CP SEND), you can send commands in files to the agent's virtual reader.
About this task
The z/VM Agent for TDMF recognizes a file in its virtual reader as a command file if the file type (obtainedfrom the spool file control block), member name, or any data set name qualifier of a data set sent fromz/OS (obtained from the INMR02 NETDATA control record) equals TDMCMDS or JOB.
You can send files to the z/VM Agent for TDMF virtual reader by using various protocols such as FTP andNJE.
236 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Sample JCL to send a command over FTP
About this task
The following sample JCL transfers a SESSION command to the virtual reader of the virtual machineTDMFVM and assigns a file name of SESSION and a file type of TDMCMDS. It expects the password forTDMFVM to be defined in the IBM.NETRC data set.
Note: In the sample, TDMCMDS must be specified as filetype to identify the file as a command file.
//VMCMD JOB //FTPSTP1 EXEC PGM=FTP,REGION=2048K, // PARM='RALTDMVM.RALEIGH.IBM.COM (TIMEOUT 20' //SYSPRINT DD SYSOUT=H //SYSTCPD DD DISP=SHR,DSN=SYS1.TCPPARMS(TCPDATA) //NETRC DD DISP=SHR,DSN=IBM.NETRC //OUTPUT DD SYSOUT=H //COMMAND DD * SESSION IBM.TDMF.SYSCOM VOL=SYSCM1 //INPUT DD * CD TDMFVM.RDR PUT //DD:COMMAND SESSION.TDMCMDS QUIT /*
Sample JCL to issue a command over NJE
About this task
This sample JCL transfers the GETCONS command to the virtual reader of the TDMF Server virtualmachine.
//VMCMD JOB,// MSGLEVEL=(1,1),// NOTIFY=&SYSUID,MSGCLASS=T//SEND XMIT DEST=RALTDMVM.TDMFVM GETCONS/*
Appendix H. Operating the z/VM Agent for TDMF 237
238 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Notices
This information was developed for products and services offered in the U.S.A.
IBM may not offer the products, services, or features discussed in this document in other countries.Consult your local IBM representative for information on the products and services currently available inyour area. Any reference to an IBM product, program, or service is not intended to state or imply that onlythat IBM product, program, or service may be used. Any functionally equivalent product, program, orservice that does not infringe any IBM intellectual property right may be used instead. However, it is theuser's responsibility to evaluate and verify the operation of any non-IBM product, program, or service.
IBM may have patents or pending patent applications covering subject matter described in thisdocument. The furnishing of this document does not give you any license to these patents. You can sendlicense inquiries, in writing, to:
IBM Director of LicensingIBM CorporationNorth Castle DriveArmonk, NY 10504-1785U.S.A.
For license inquiries regarding double-byte character set (DBCS) information, contact the IBM IntellectualProperty Department in your country or send inquiries, in writing, to:
Intellectual Property LicensingLegal and Intellectual Property LawIBM Japan, Ltd.3-2-12, Roppongi, Minato-ku, Tokyo106-8711 Japan
The following paragraph does not apply to the United Kingdom or any other country where suchprovisions are inconsistent with local law: INTERNATIONAL BUSINESS MACHINES CORPORATIONPROVIDES THIS PUBLICATIONS "AS IS" WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESS ORIMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF NON-INFRINGEMENT,MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Some states do not allow disclaimer ofexpress or implied warranties in certain transactions, therefore, this statement may not apply to you.
This information could include technical inaccuracies or typographical errors. Changes are periodicallymade to the information herein; these changes will be incorporated in new editions of the publication.IBM may make improvements and/or changes in the product(s) and/or the program(s) described in thispublication at any time without notice.
Any references in this information to non-IBM Web sites are provided for convenience only and do not inany manner serve as an endorsement of those Web sites. The materials at those Web sites are not part ofthe materials for this IBM product and use of those Web sites is at your own risk.
IBM may use or distribute any of the information you supply in any way it believes appropriate withoutincurring any obligation to you.
Any performance data contained herein was determined in a controlled environment. Therefore, theresults obtained in other operating environments may vary significantly. Some measurements may havebeen made on development-level systems and there is no guarantee that these measurements will be thesame on generally available systems. Furthermore, some measurement may have been estimatedthrough extrapolation. Actual results may vary. Users of this document should verify the applicable datafor their specific environment.
Information concerning non-IBM products was obtained from the suppliers of those products, theirpublished announcements or other publicly available sources. IBM has not tested those products andcannot confirm the accuracy of performance, compatibility or any other claims related to non-IBM
© Copyright IBM Corp. 2010, 2019 239
products. Questions on the capabilities of non-IBM products should be addressed to the suppliers ofthose products.
All statements regarding IBM's future direction or intent are subject to change or withdrawal withoutnotice, and represent goals and objectives only.
This information is for planning purposes only. The information herein is subject to change before theproducts described become available.
This information contains examples of data and reports used in daily business operations. To illustratethem as completely as possible, the examples include the names of individuals, companies, brands, andproducts. All of these names are fictitious and any similarity to the names and addresses used by anactual business enterprise is entirely coincidental.
TrademarksIBM, the IBM logo, and ibm.com® are trademarks or registered trademarks of International BusinessMachines Corp., registered in many jurisdictions worldwide. Other product and service names might betrademarks of IBM or other companies. A current list of IBM trademarks is available on the Copyright andtrademark information website.
Adobe, the Adobe logo, PostScript, and the PostScript logo are either registered trademarks ortrademarks of Adobe Systems Incorporated in the United States, and/or other countries.
IT Infrastructure Library is a registered trademark of the Central Computer and TelecommunicationsAgency which is now part of the Office of Government Commerce.
Java and all Java-based trademarks and logos are trademarks or registered trademarks of Oracle and/orits affiliates.
240 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
Index
AACDS 72acknowledge mirror on swap 150action messages 188ACTIVE IN COPY 27active only in copy option 141agent system
CPU overhead 61initiation 2, 6number of 90quiesce request 9, 60responsibilities 4run order 30sample job 21, 141termination of 32time intervals 61
Agent systembuild batch jobs 171initiation 4responsibilities 4
agents running, verifying 30ALLOW INVALID COUNT FIELDS 27ALLOWmirrorchange 53, 150AMATERSE 140AOMSSDS service 191APARs
APAR II11078 81APAR OA14558 78APAR OA16358 83APAR PK69445 193
APPROVAL 43authorization keys
adding 165deleting 165GTDSOPTN 21
authorization return codesauthority checking 203history file recording 206TDMF security record 204TDMF/Express feature information 205
auto-operationsinterface 55messages 209
AUTOMATIC ICKDSF 27AUTOOPS, AUTOOPerations 52autoswap 197average device response times 160
BBasic Express key 203Basic HyperSwap 195batch utilities
overview 115Batch utilities
GTDBMON 115
Batch UtilitiesGTCLUP
Repairing a single device 126GTDBMON
command processing 116control statement processing 119messages 121printed output 121
GTDCFTP 121GTDCLIP
return codes 123GTDCLUP
messages 127program abend codes 126Scanning the entire system 123
GTDEXTVmessages 128Program abend codes 127
GTDMUCBoperation 128return codes 129
GTDQDSKmessages 130printed output 130user abends 130
GTDSCANabend codes 132messages 132operation 131return codes 132run parameter 130sample output 132
GTDTERAsample output 133user abends 133
GTDVTOCabend codes 137messages 135printed output 135
GZDRECONoptions 138processing 140
BCDS 72BLDL error 204, 205build replication or migration jobs 166building
a COMMDS 172an Agent system batch job 171
BUILDIX IXVTOC 81bytes migrated report
definitions 186viewing 183
CCA Products, See Computer Associates Productscache fast write
Index 241
cache fast write (continued)and sort programs 72
CCHH 181CCHR 179change or submit replication or migration jobs 141, 142changes due to I/O contention 157channel utilization 61CHECK TARGET EMPTY 27checking target volume for user data 40checking the fence status 33checklist 19class names 72CLIP, See GTDCLIPcommand word
Channel 57, 78, 229COMMDS
purpose 3See also Communications data set
comments, sending xiicommunication data set
generate ICKDSF JCL 182navigator 179
Communication data set 2, 4Communication Data set
allocation 19Communication Data Set
allocation 78, 89hardware reserve 78, 92history file 21purpose 3
Communications data set 121COMPARE 52compare option
errors 180COMPRESSion 54Computer Associates products
CA-11 Job Management 71CA-7 job management 70CA-ASTEX performance 69CA-MIM resource sharing 71CA-OPS/MVS Event Management and Automation 69CA-Scheduler job management 70moving multi-image manager control data sets 72
Computer Associates ProductsCA ACF2 69
concurrent copy 78CONCurrent, CNCRNT 42CONFirm 52confirmation
common option 52volume 7
Consistency Groups 76CONSole (console_name) 37CONSTDMF 191continue a volume 151Control block search and scan entire system 124Control Statements
GROUP 46MIGRATE 48, 49optional 35order of input 36REPLICATE 49required 35SESSION 37
Control Statements (continued)TDMF 35
copy phase I/O 157CORPORATION 27count key data protocol 181count records
allowing invalid 40couple data sets 90CPU
determining serial number 201PCCA 203serial number 21
cumulative refresh 180current sessions
display messages 152display offline volume access (OVA) 153monitor progress 141, 143user interaction and status 145
DDASD
determining 199fast write 61, 72space requirements 69, 199subsystem serial number 199UCBs 177
DASD Fast Write 73Data facility products 72, 73data migration
definition and characteristics 1data set control blocks 178databases
point-in-time migrations 60DCOLLECT 73, 81defaults
for SESSION options 38system 27
Delayed Offline Volume Access (DOVA)purpose 105
deleting keys 165detected source volume I/O errors 181device recovery 32device states
before and after a swap migration 72, 73DEVSERV command 98DFHSM 72DFSMS
class names 72DCOLLECT 73DISNEW command 73QUIENEW command 73volume status 73
DFW 61DISPLAY TIME AS 27display/modify user's TSO monitor options 166Documenting TDMF TCP/IP port numbers 112DS1DSCHA 54DS4DEVAC 81DS4NOATK 81duplex_target 50duplexing functions
Dual Copy 75PPRC 91
242 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
duplexing functions (continued)XRC 75
dynamicdevices 96ICKDSF, how functions will affect a swap migration 80ICKDSF, intent of 80pacing, I/O pacing 157pacing, real storage pacing 157suspension 64
dynamic devices 95dynamic volume pacing
I/O 56pacing with channel extenders 57real storage pacing 56reverse pacing 56user-specified pacing values 57
EECB 204ECSA 124ENF 92, 93Enhanced catalog sharing (ECS) 78ENQ 206enter/update/review job cards 167Enterprise key 203Enterprise storage server
considerations 95support 95, 97
Enterprise Storage Serverconsiderations 92support 92–94, 96
error messages 188esoteric names 72ESTAE recovery 206Event Notification Facility 92, 93Express key 203extended functions
concurrent copy 78data striping 78VSAM data compression 78
extenders, pacing withChannel 57, 78, 229
external time reference (ETR) 180EXTVtoc 53EXTVTOC
FORCE parameter 82nn 82
FFASTCOPY
NOFASTcopy 52refresh 180required with TCP/IP 111
feedback, sending xiiFENcesource/NOFENcesource 51file allocation managers
ProSMS 78SAMS Vantage 78
FIND 189FIXUCB 126FORCE 53
FTP processGTDCFTP 121
Full Speed Copyinpact 48
GGDPS
/PPRC hyperswap 191/PPRC HyperSwap 191/PPRC HyperSwap groups 192/PPRC HyperSwap volumes 191/PPRC suspending HyperSwap environment 193/XRC environment 100Controlling systems 191
GDPSplex 191generate ICKDSF JCL from communication data set 182GEOCB control block 195GETMAIN 204, 205global mirror 100Global resource serialization 30, 62, 79GMT time
setting conversion to local time zone 41GROUP
control statement 46statement SINGLEgroup option 41
group_name 46groups
number of, per session 90GRS, See Global resource serializationGTDCLIP 122GTDCLUP 191, 195GTDEDIT 21GTDEXTV
abend codes 127program messages 128
GTDLOGS 206GTDMUCB 128GTDPROF 21GTDSCAN 83GTDSOPTN 141GTDTERA 132GTDUKEY 203GZDRECON 137
Hhang monitor 64HCD
and IODF files 79help and message detail facility
introduction to TDMF 188help messages, See TDMF Monitor feature help and messagedetail facilityHFS and zFS files 80Host software component 78, 79, 89HSC 78HSM 72HSWAPDisable 43HyperSwap
Basic 195Groups 191resuming the basic environment 195
Index 243
HyperSwap (continued)Resuming the GDPS-PPRC environment 191Suspending the basic environment 195suspending the GDPS-PPRC environment 191Validation and initialization 191
II/O
pacing 56performance 60timeout condition 64
ICKDSFJCL 182REFVTOC 30, 60relabel 60restrictions 78, 80session 191
ICKdsfNOICKdsfNOICKdsf 40
IDCAMScache fast write 72, 73DASD fast write 72DASD Fast Write 73LISTDATA command 85
IEC378I 78IGNOREGdps 43, 191IGNORESCUs 43IGNORESYStems 42INDEX 182informational messages 188INIT NOVALIDATE 182INIT VALIDATE 182initialization
systeminitialization 91
system initialization 6volume initialization 7
Installingpre-installation checklist 19
invalid count fieldcorrection of 83
invalid count recordsallowing 40
IOBUFFERPOOL 44IODF 78IOSHMBLI 195IOSHXBLK 195IPL volumes 83ISPF
keylist 21, 141profile data set members 21, 141
ISPF.ISPPROF 21
JJES2
checkpoint volumes 84considerations 84MASDEF statement 84spool volumes 84
JES3checkpoint volumes 84
JES3 (continued)considerations 78, 85P/DAS support 85spool volumes 84XTYPE parameter 85
job cards, create 167
KK system 100Keylist
enable 21keys
deleting 165
Llabeling
relabeling source volume following swap 40line commands 189LISTDATA information 85LISTUCB 177ll10752, APAR 80LNKLST 140local time
setting conversion to 41LPAR 177
Mmaster system
activation phase 9copy task 9CPU overhead 61I/O pacing 56initiation 3placement 62quiesce phase 9real storage pacing 56redirect phase 10refresh phase 9relationship 3, 21responsibilities 4resume phase 3, 11run order 30sample job 21session status 141SMF recording 59synchronization goal 60synchronize phase 10termination of 32time intervals 61
MAXTRacks 47MCDS 72merged system messages
communication data set 180message severity 188messages, See TDMF monitor feature help and messagedetail facilityMIGRATE control statement 48, 49migration
autoswap overview 197basic HyperSwap volumes 195, 196
244 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
migration (continued)checkpoint volumes 84control data sets 85definition and characteristics 1dynamic devices 95dynamic suspension 64EMC autoswap validation and initialization 197EMC autoswap volumes 197GDPS
/PPRC resuming HyperSwap environment 193controlling system 192
GDPS/PPRC HyperSwap volumes 191, 195GDPS/PPRC HyperSwap Volumes 191HyperSwap while TDMF is active 194installation static devices 95intializing 197IODF 78IPL volumes 83JES3 system managed devices 85master/agent system relationship 3page data sets 1, 90phases 5PPRC volumes 91RAID and non-RAID devices 85, 86SRDF volumes 76stand-alone dump data sets 86static devices 95suspend 64swap data sets 1sysplex couple data sets 90types, point-in-time 59unlike device types 85, 86validating 197VM formatted volumes 85VM volumes 89volumes formatted for VM use 87
migration phasesactivation 9copy volume 9quiesce 9redirect 3, 11refresh 9resume 12synchronize 9system initialization 6, 7
migration sessionauto-operations 55canceling
recovery of devices 32channel utilization 61history file 6rank contention 63terminating 32
MIHI 64MIM, See Multi-Image Managermiscellaneous refresh 180missing interrupt handler interval 64Model 204 considerations 85, 87model number
CPU 21MODIFY Netview,HYPERSW SYS 191module version levels in defined load library 176Monitor
feature, run 21
Monitor (continued)line commands 189
MUCB, See GTDMUCBMulti-Image Manager 62, 89MVS
running under VM 89
NNetView (name) 37new features xiii, 17new_volser 49No valid keys detected 203NOALLOWmirrorchange 53NOAPPROVAL 43NOAUTOOPerations, NOAUTOOPS 52NOCFW 41NOCOMPARE 52NOCOMPRESSion 54NOCONFirm 52NOEXTVTOC 82NOEXTVtoc, NOXVToc 53NOFASTcopy 52NOIOBUFFERPOOL 44NOOVA 51, 54NOPAV 54NOPERPetual 54NOPPIT 54NOPRompt 52NOPUrge 53NOTERMGRoup 52NOWARNING 41NOXCF 41
OOCDS 72Offline Volume Access
control statements 105de-registration 104exclusion list 103functions 101I/O redirection 104I/O redirection report 104inclusion list 103initialization 103JCL statements 105multi-volume data sets 102processing 103purpose 101registration 104Registration Report 104starting a session 102usage 101, 102using on different subsystems 102using the interface program 105
OFFLINESource 54OFFLINETarget 54ONLINESource 54ONLINETarget 54operating system requirements 19Optional Control Statements 35options
Index 245
options (continued)0 1421 14311 16612 1662 145–1482, Pairing (P) 1502, synchronize goal 150auto-operations 55compare 67H 187H.1 188H.2 188H.4 189history file 6multiple group names 60parm statement 27prompt 60REMOTE 45S 183SMF recording 59U 173U.0 173U.1 174U.10 179U.11 180U.12 180U.13 181U.14 182U.2 174U.3 175U.4 175U.5 176U.6 177U.7 177U.8 177U.9 178volume 47volume pacing 27
OptionsCommon for SESSION, GROUP, MIGRATE, andREPLICATE 50
OVAREGISTRATION INTERVAL 27
OW23271, OW28455, and OW28457APARs 78, 80, 81, 83, 85, 191, 193
PP/DAS, See Peer-to-Peer remote copy/dynamic addressswitchingPACing 47pacing with channel extenders 57page and swap data sets 90page data sets
local 1, 90PLPA 90
pairing 150Parallel sysplex considerations 91PAV
alias devices 195PCCA 173Peer to Peer remote copy
support migrating PPRC to a non-PPRC device 92
Peer-to-Peer remote copy/dynamic address switchingJES3 support 85
percent of device used by customer 158performance
application impact 61channel utilization 61considerations 63customer on volume 158statistics storage subsystem 63storage subsystem 63TDMF monitor feature 158
PERPetual 54Perpetual point-in-time
during an active 106setting up for 108
PG 150physical configuration communication area 173PIT
volumes 191See also Point-in-time
PK38350, APAR 191PK38351, APAR 191Point-in-time
databases 60migrations 60
PORT(number) 43postinstallation tailoring 20PPIT 54Pre-installation
considerations 19preparing to run TDMF 21previous session
display trace table 175PRINT 189program temporary fixes 177PRompt 52PTFs 177publications
ordering xiproduct xirelated xi
PUrge 53
Qquiesce phase 9
RRACF 191, 194, 206RAID subsystems
rank contention 63RAMAC virtual array 97RCVBuf 45RDJFCB 204, 205Reading syntax diagrams 229real storage pacing 56real time monitors 92recovering TDMF devices 32REFORMAT EXTVTOC 82REFORMAT REFVTOC 81relabeling source volume after swap 40remote_name 46, 50
246 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
remote_target 50REPLICATE control statement 49Required control statements 35RESETCHAngedflag 54resume phase 12resuming
the autoswap environment 198the HyperSwap environment 191, 195, 196
RESYNCH 191, 195return code
0000, 0004, 0008, 0012 188authorization 203
reverse pacing 157Reverse pacing 27REXX execs
requirements 141RFIND 189RMM 72RSM 157running TDMF 30RVA support 95
SSAF 6SCAN, See GTDSCANSCAN parameter 223Scanning the system 223security 6serial number
CPU 21, 201DASD subsystem 199
SESSION 195session examples 215SESSION options
ALLOWINValidcounts 40APPROVAL 43CONCurrent, CNCRNT 42HSWAPDisable 43ICKdsf 40IGNOREGdps 43IGNORESCUs 43IGNORESYStems 42IOBUFFERPOOL 44NOAPPROVAL 43NOCFW 41NOCHECKTarget 40NOIOBUFFERPOOL 44NOPAcing 40NOWARNING 41NOXCF 41PORT(number) 43RELABel 40SINGLEgroup 41STORCLas 42TIME 41UNIDENTifiedsystems 39
SESSION parametersCONSole (console_name) 37NetView (name) 37
sessionsnumber of 90
SETHS UNBLOCK 195SETIOS 64
severe errors 188shared versus non-shared devices 92, 93shared virtual array 97Shared virtual array, See SVA supportShark support
CCW support 78, 94FlashCopy support 95messaging 93swap from ESS to non-ESS 94swap from non-ESS to ESS 94
single group option 60Singlegroup
option 46parameter 41
SITE NAME 27SITE NUMBER 27SMF RECORD ID 27SMF recording 59SMF reporting 183SMS 72SNDBuf 45Soft Fence 97sort programs 72source volume
DFSMS considerations 73dynamic suspend command 64dynamic suspension 64extended functions considerations 78invalid count field 181JES3 considerations 85Model 204 considerations 85pacing 55performance 61performance data 158PPRC considerations 91requirements 1, 85, 90, 95smaller to larger device migration 60status after migration 59, 60system
authorization facility requirements 20source_volume
migrate 49replicate 50
SPID Fence 98SRDF 75start a volume 149STARTUP CONFIRMATION REQUIRED 27statements, Required control 35static devices 95, 96storage requirements 67STORCLas 42support utilities
active sessiondisplay trace table 175display/alter tracing bit settings 174
communication data set control blocks 178current session
module version levels for COMMDS 177system change summary 177unit communication data set merged system messages180unit communication data set navigator 179unit communication data set volume refresh bit map180
Index 247
suppressing warning messages in TDMF 98suspending
the EMC environment 198the HyperSwap environment 191, 195volume 151
SVA support 95swap data sets 1, 90SYNC Volume Needed 149SYNCgoal 53synchronization goal 60synchronize goal 150SYNCHRONIZE Phase 10syntax
example 98Syntax diagrams
reading 229SYSCOM 206SYSCOM HISTORY DATASET NAME 27SYSIGGV2 83system
authorization facility 20automation (SA) policy 191, 194, 198data mover 141defaults 21failure manager 90information 154initialization 6
Ttarget empty tracks 180target volume
checking for user data 40DFSMS considerations 73dynamic suspension 64extended functions considerations 78JES3 considerations 85Model 204 considerations 85performance 61PPRC considerations 91requirements 1, 21, 85, 95, 97smaller to larger device migration 60status after migration 59, 60system
authorization facility requirements 6TDMF ownership 95
target_volumemigrate 49replicate 50
TCP/IPconnection 43Fast copy requirement 111migrations 131
TDMFadvanced functions 101advanced functions overview 101and volume table of contents 97control statements 35customizing 19delayed offline volume access 106installation 19planning considerations 67TCP/IP support 110
TDMF help and message detail facility 187
TDMF monitor featureauto-operations 55help and message detail facility 188session termination 32support utilities
invalid count field 83TDMF Monitor feature
change or submit migration jobs 141current sessions
user interaction and status 141help and message detail facility 141ISPF keylist 141ISPF profile data set members 141overview 141run 141support utilities 141
TDMF TSO monitor featurehelp and message detail facility
introduction to TDMF 188support utilities
display memory 173view internal details for active migrations 174
TDMF TSO Monitor featurebuild data migration jobs
Agent system batch job 171build communication data set 172
help and message detail facilitydisplay message details 188
help and message detail facility monitor line commands189master system batch job 167menu 141message severity 188options
10 1643 1524 1535 1546 1607 1618 1639 163
overview 141support utilities
detected source volume i/o errors 181invalid count field display 181module version levels in defined load library 176system change summary 177unit communication data set merged systemmessages 180unit communication data set navigator 179unit control block display 177
TDMFMSE 178TDMFoffline volume access 101TDMFPRT 189TDMFSECURE 182Technical advisory group 1TERA (see GTDTERA) 132TERMGRoup 52terminate
a volume 151Terminate
phase 13TDMF 32
248 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
TERMINATE ALL VOLUMES IN GROUP ON ERROR 27TESTBLOCK function 195time display option 41total elapsed device response time 158trademarks 240transactional terabyte license (TTL) 166Trial Express key 203TRKFMT 182TRSMAIN 140TSO Monitor, See TDMF monitor featureTTR conversion error 204TTRZ 179
UUCB (see Unit Control Block(s)) 30UCBDDT field 195UCBHSWAP 191, 195UCBs 177UNCONDITION 53UNIDENTIFIED SYSTEMS ACTION 27unidentified systems option 39Unit Control Block(s)
above the 16MB line 30below the 16-MB line 30four-digit 30JES3 considerations 85LISTUCB command 95three-digit 30
unit control blocksdisplay 177
user datachecking target volume 40
user interaction and status panelhelp 148information 146migration options 147
user-specified pacing values 57utilities, See TDMF Monitor feature support utilities
VValid DDT example 33Vendor key 203verifying TDMF agents 30version definition 203view internal details for active sessions 174virtual machine
considerations 89restrictions 19volume tolerance 89
VM, See virtual machineVOLREFR 100volume
information 155pairing 150performance data on source 158
Volumechecking target for user data 40I/O redirection 11information block (VIB) 81options 47table of contents 95
volume pacingapplication performance 60, 61channel extenders 57I/O pacing 56, 157real storage pacing 56, 157reverse pacing 56setting 40specific pacing values 57when pacing is not in effect 57
volume refreshcommunication data set 180
VOLUME SECURITY 27volumes
acknowledge swap 150number of volumes per session 90synchronize goal 150
VSMLIST service 124VTOC 182, 191
Wwarning messages 188websites xiWTO AUTO OPERATION MVS ROUTCDE 27WTO MESSAGES FOR AUTOMATED OPERATIONS 27
XXCF 43XMIT 137XRC (see Extended Remote Copy) 99XVToc 53
ZZ 150z/VM Agent for TDMF
changing configuration settings 236communicating using secondary user 232configuring TDMF server virtual machine 23displaying active migration sessions 234displaying applied software maintenance forcomponents 235preparing to run 26querying operation status 231remote operation 236retrieving logging information 233starting 231starting migration session 232status information 234terminating the z/VM Agent 233
ZG 150
Index 249
250 IBM Transparent Data Migration Facility (TDMF) for z/OS : TDMF Reference Guide
IBM®
SC27-8963-03
