View
215
Download
1
Category
Tags:
Preview:
Citation preview
HP Information Lifecycle Management (ILM)
John Sosnowski – Account Executive
Jeff Blackmon – Presales Engineer
Database Growth: Database Growth: Problems & SolutionsProblems & Solutions
2204/10/23 2
AgendaAgenda
The Data ExplosionThe Data ExplosionEffects of Database ExplosionEffects of Database Explosion
End UserEnd UserIT OperationsIT OperationsCase StudiesCase Studies
SolutionsSolutionsQ & AQ & A
3304/10/23 3
The Data ExplosionThe Data Explosion
““Data Explosion”: Data Explosion”: 130,000 Google Hits130,000 Google Hits
““Information Life Cycle Management”: Information Life Cycle Management”: 1.3M 1.3M Google Hits Google Hits
Numerous analyst papersNumerous analyst papers
Archiving vendorsArchiving vendors
4404/10/23 4
Data Explosion: Largest OLTP on Data Explosion: Largest OLTP on Unix DatabasesUnix Databases
Source: Winter Corporation; Top 10 Source: Winter Corporation; Top 10 Survey (www.wintercorp.com)Survey (www.wintercorp.com)
2003: Max: 5.4 TB, Average 2.5 TB2003: Max: 5.4 TB, Average 2.5 TB
2005: Max: 16.4 TB, Average 6.8 TB2005: Max: 16.4 TB, Average 6.8 TB
5504/10/23 5
Drivers for Data ExplosionDrivers for Data Explosion
Increased regulatory Increased regulatory retention requirementsretention requirements
Increased business Increased business intelligence and trend analysisintelligence and trend analysis
Increased detail with-in Increased detail with-in business transactions business transactions
Consolidation of systemsConsolidation of systemsUnderlying growth in Underlying growth in
business volumebusiness volume
6604/10/23 6
Data Explosion: Types of DataData Explosion: Types of Data
Enterprise Information
Business Applications
ERP SystemsERP Systems
Files(un-structured)
Messaging(semi-structured)
(future)Custom AppsCustom Apps ExchangeExchange Lotus NotesLotus Notes Enterprise Document ManagementEnterprise Document Management
Databases(structured)
7704/10/23 7
Data Explosion: Types of DataData Explosion: Types of Data
Enterprise Information
Business Applications
ERP SystemsERP Systems
Files(un-structured)
Messaging(semi-structured)
(future)Custom AppsCustom Apps ExchangeExchange Lotus NotesLotus Notes Enterprise Document ManagementEnterprise Document Management
Databases(structured)
8804/10/23 8
Data has a LifecycleData has a Lifecycle
High
Low
Access / SLA Requirements
Time
Retrieval Activity
Customer inquiryInternal audit
Sales order booked
Change order
Partial shipments
Complete shipment
Customer invoices generated
Sales order closed
Regulatory data retention period ends
Corporate data retention period
9904/10/23 9
Inactive DataInactive Data
Large
Small
DatabaseSize
Time
Retrieval Activity
Customer inquiryInternal audit
Sales order booked
Change order
Partial shipments
Complete shipment
Customer invoices generated
Sales order closed
Inactive Data
Active Data
101004/10/23 10
AgendaAgenda
The Data ExplosionThe Data ExplosionEffects of Database ExplosionEffects of Database Explosion
End User End User IT OperationsIT OperationsCase StudiesCase Studies
SolutionsSolutionsQ & AQ & A
111104/10/23 11
Effects of Database GrowthEffects of Database Growth
End-UserEnd-User
Declining application performanceDeclining application performance
Decreased application availabilityDecreased application availability
Information overloadInformation overloadIs your application presenting irrelevant Is your application presenting irrelevant
information to end-user, causing extra information to end-user, causing extra work?work?
121204/10/23 12
Effects of Database GrowthEffects of Database Growth
IT OperationsIT Operations
Storage requirementsStorage requirements
Server requirementsServer requirements
Backup and recovery windowsBackup and recovery windows
Maintenance and upgrades windowsMaintenance and upgrades windows
Cloning operationsCloning operations
Database and Application TuningDatabase and Application Tuning
131304/10/23 13
Case StudiesCase Studies
Only studies that have been published: Only studies that have been published: OAUG, OpenWorld, trade press OAUG, OpenWorld, trade press
All Oracle E-Business SuiteAll Oracle E-Business SuiteAll All RIM for DB LiveArchiveRIM for DB LiveArchive from HP/OuterBay from HP/OuterBayIn chronological orderIn chronological order
141404/10/23 14
Applied Materials – Server CapacityApplied Materials – Server Capacity
Data growth of 4GB per weekData growth of 4GB per week14 clones of Production14 clones of ProductionHP HP Live ArchiveLive Archive implementation led to near implementation led to near zero growthzero growth
Storage savings $1 MM/year Storage savings $1 MM/year Significant CPU utilization reductions Significant CPU utilization reductions extending server life by several yearsextending server life by several years
Runtime reductions on key reports by 50%Runtime reductions on key reports by 50%Source: Out of Gas on the Biggest Box Made, Now What? Chei Yeh, Source: Out of Gas on the Biggest Box Made, Now What? Chei Yeh,
Director Core Applications Director Core Applications Applied Materials. Applied Materials. OAUG Fall 2001OAUG Fall 2001
151504/10/23 15
Parsons Brinkerhoff - UpgradeParsons Brinkerhoff - Upgrade
10.7 to 1110.7 to 11i i upgrade of EBS upgrade of EBS Implemented Implemented Live ArchiveLive Archive before upgrade before upgrade Archive 35% of total DBArchive 35% of total DBReduced Upgrade outage from 12 days to 6 daysReduced Upgrade outage from 12 days to 6 daysAllowed 3 clones to be supported on test server Allowed 3 clones to be supported on test server
instead of 2 pre-archiveinstead of 2 pre-archiveUser agreement for archiving was easy to User agreement for archiving was easy to
achieve because of application access to achieve because of application access to archived dataarchived data
Source: “Worldwide Upgrade to Oracle 11Source: “Worldwide Upgrade to Oracle 11ii: A Recipe for Success” OAUG : A Recipe for Success” OAUG EuropeEurope 2002 by Robert P. Dallesandro Parsons-Brinkerhoff2002 by Robert P. Dallesandro Parsons-Brinkerhoff
161604/10/23 16
Apollo Group - UpgradeApollo Group - Upgrade
10.7 to 1110.7 to 11ii upgrade window reduction upgrade window reductionFixed window over 4Fixed window over 4thth July weekend July weekend$50K/day in productivity impact of additional outage $50K/day in productivity impact of additional outage
windowwindowArchived 22% of the 190 GB database pre-archiveArchived 22% of the 190 GB database pre-archiveReduced outage window by 27%Reduced outage window by 27%Avoided additional outageAvoided additional outageLessons learned: Archive before starting upgrade Lessons learned: Archive before starting upgrade
project project
Source: “Oracle 11i Data Upgrade Roadmap – A Lifecycle-Based Source: “Oracle 11i Data Upgrade Roadmap – A Lifecycle-Based Approach”, Approach”, Grant Gasson, Director of Financial Systems Grant Gasson, Director of Financial Systems Apollo Apollo Group, Inc. Group, Inc. Fall 2002 OAUGFall 2002 OAUG
171704/10/23 17
Gevity: Storage CostsGevity: Storage Costs
Reduced budget for storage –Reduced budget for storage – $1.3 savings (2007–2011)$1.3 savings (2007–2011)
Avoid purchase of additional Avoid purchase of additional storage ($664,474) for 2004storage ($664,474) for 2004
Dramatically reduce instances by Dramatically reduce instances by approximately 2TBapproximately 2TB
Saving total storage footprint of Saving total storage footprint of 12TB at implementation12TB at implementation
Additional BenefitsAdditional Benefits Sustainable and predictable Sustainable and predictable
performance/storage costsperformance/storage costs Allows high performance access to Allows high performance access to
clients worldwide via new portal, clients worldwide via new portal, while managing high performance while managing high performance reqsreqs
Greatly reduced backup, recovery Greatly reduced backup, recovery and clone times and clone times
181804/10/23 18
Agilent Technologies: Storage Agilent Technologies: Storage CostsCosts
Oracle EBS 11Oracle EBS 11ii Single world-wide database Single world-wide databaseGrowth rate 1TB+ per yearGrowth rate 1TB+ per year20 + clones to support business initiatives20 + clones to support business initiativesBenefitsBenefits
Savings for Storage: 2.5% of total 2005 IT budgetSavings for Storage: 2.5% of total 2005 IT budgetHA, Backup & Recovery reduced spend and timeHA, Backup & Recovery reduced spend and timeStable, predictable performanceStable, predictable performanceFree up resources for business initiativesFree up resources for business initiatives
Source: Agilent Case Study: Strategic Application Data Growth Management. By Naresh Shanker Sr. Director ERP Solutions, Agilent Technologies, OpenWorld SF 2004
191904/10/23 19
Agilent: Tuning - TemporaryAgilent: Tuning - Temporary
10,000
15,000
20,000
25,000
30,000
35,000
40,000
400
500
600
700
800
900
1000
1100
1200
1300
Ap
plic
atio
n D
ata
(GB
)
1300
1200
1100
1000
900
800
700
600
500
400
Tota
l B
atc
h R
un
Tim
e (
Hrs
/Mon
th)
Application Data
Batch Run Time
Data Growth Impact
Phase I Tuning
Source: Agilent Case Study: Strategic Application Data Growth Management. By Naresh Shanker Sr. Director ERP Solutions, Agilent Technologies, OpenWorld SF 2004
202004/10/23 20
10,000
15,000
20,000
25,000
30,000
35,000
40,000
2003-06
2003-07
2003-08
2003-09
2003-10
2003-11
2003-12
2004-01
2004-02
2004-03
2004-04
2004-05
2004-06
2004-07
2004-08
2004-09
2004-10
400
500
600
700
800
900
1000
1100
1200
1300
Tota
l B
atc
h R
un
Tim
e (
Hrs
/Mon
th)
Ap
plic
atio
n D
ata
(GB
)
Data Growth Impact
Sustained Predictable Performanc
eArchiving
Phase II Tuning
Agilent: Agilent: Archiving = StabilityArchiving = Stability
Source: Agilent Case Study: Strategic Application Data Growth Management. By Naresh Shanker Sr. Director ERP Solutions, Agilent Technologies, OpenWorld SF 2004
212104/10/23 21
ArvinMeritor – Disaster RecoveryArvinMeritor – Disaster Recovery
60% reduction in database size60% reduction in database size90% reduction in DR window 90% reduction in DR window 1111ii upgrade outage reduction to 2 days upgrade outage reduction to 2 days30% performance improvement30% performance improvementLower database maintenance overheadLower database maintenance overhead
Source: Automotive supplier curbs its runaway production Source: Automotive supplier curbs its runaway production database. Marcel Kuijs, director of IT at ArvinMeritor's European database. Marcel Kuijs, director of IT at ArvinMeritor's European
IT data center. Storage Network World, July 2005IT data center. Storage Network World, July 2005. .
222204/10/23 22
Case Study ThemesCase Study Themes
Many drivers for archiving Many drivers for archiving PerformancePerformanceUpgrade windowUpgrade window
Diverse set of secondary benefitsDiverse set of secondary benefitsCost avoidance, storage, servers, tapesCost avoidance, storage, servers, tapesFree resources for Strategic InitiativesFree resources for Strategic Initiatives
Benefits accrue to diverse parts of the Benefits accrue to diverse parts of the organizationorganizationData centerData centerBusiness userBusiness userApplication Support staffApplication Support staff
232304/10/23 23
AgendaAgenda
The Data ExplosionThe Data ExplosionEffects of Database ExplosionEffects of Database Explosion
End UserEnd UserIT OperationsIT OperationsCase StudiesCase Studies
SolutionsSolutionsQ & AQ & A
242404/10/23 24
Possible SolutionsPossible Solutions
Data Deletion / Data Deletion / Purging Purging
Add Capacity Add Capacity through Hardware through Hardware UpgradesUpgrades
Decentralize or Do Decentralize or Do Not ConsolidateNot Consolidate
Database PartitioningDatabase PartitioningDatabase ArchivingDatabase Archiving
252504/10/23 25
Solutions: Data Deletion/PurgingSolutions: Data Deletion/Purging
Some data is not subject to regulatory Some data is not subject to regulatory requirements and has no ongoing business requirements and has no ongoing business valuevalue
Candidates:Candidates:Concurrent request dataConcurrent request data
MRP runs, Workflows, etc.MRP runs, Workflows, etc.
Temporary and Interface TablesTemporary and Interface TablesSome Challenges! Introduces risk and exposure to meet legal and compliance
requirements No recourse, once the data is gone it is gone forever No validation for full data integrate
262604/10/23 26
Solutions: Add Capacity Solutions: Add Capacity Through Hardware UpgradesThrough Hardware Upgrades
StorageStorage
ServerServerScale upScale up
Scale out (RAC)Scale out (RAC)
NetworkNetwork
Some Challenges! Doesn’t solve the root cause, “Hardware Acquisition
Treadmill” One CIO observes: “Each time I buy storage, it
requires more software, backup, array and bandwidth”
272704/10/23 27
Solutions: Decentralize or Solutions: Decentralize or Do Not ConsolidateDo Not Consolidate
Sometimes managing two medium-Sometimes managing two medium-sized databases is easier than sized databases is easier than managing one large databasemanaging one large database
Some Challenges! Gives up all the advantages of consolidation
− IT Operations
− Consolidating reporting, etc…
282804/10/23 28
Solutions: Database PartitioningSolutions: Database Partitioning
Oracle feature for managing large Oracle feature for managing large tablestables
Improves performance Improves performance ifif queries are queries are able to use partition keysable to use partition keys
““Life cycle” of data may leave active Life cycle” of data may leave active data in same partition as inactive data in same partition as inactive datadataSome Challenges!
Does not address core data growth issue
Provides no help in compliance, data retention
Application must be tuned for partitioning
292904/10/23 29
Solutions: Database ArchivingSolutions: Database Archiving
Remove data from OLTP Databases Remove data from OLTP Databases while retaining ability to accesswhile retaining ability to accessDatabase Backups/SnapshotsDatabase Backups/SnapshotsExports, CSV filesExports, CSV filesXML Files or other self-describing formatXML Files or other self-describing formatRelocation to secondary databaseRelocation to secondary database
Access via reporting tools or business analyticsAccess via reporting tools or business analyticsAccess via native applicationAccess via native application
Maintain full data integrityMaintain full data integrity
Database Archiving: Key Database Archiving: Key ComponentsComponents
High
Low
Access / SLA Requirements
Time
Retrieval Activity
Customer inquiryInternal audit
Sales order booked
Change order
Partial shipments
Complete shipment
Customer invoices generated
Sales order closed
Regulatory data retention period ends
Corporate data retention period
Deploy Data By Cost and Class Of Service
.XML.XSD
SourceDatabaseActive Data
Inactive Data
Compliance Data
Business Applications
HomeGrown
323204/10/23 32
QAQA
Dev2Dev2Dev2
Database to Database Archiving
Database Subsetting
Database to File Archiving
ArchiveDatabase
Moderate Access
.XML.XSD
Infrequent Access
Database
ARCHIVE PLATFORM
User Access
Recovery Services
Information Movement
Configuration
Business Transactions
Policy Management
Analysis
Database Archiving: Key ComponentsDatabase Archiving: Key Components
333304/10/23 33
AgendaAgenda
The Data ExplosionThe Data ExplosionEffects of Database ExplosionEffects of Database Explosion
End UserEnd UserIT OperationsIT Operations
SolutionsSolutionsQ & AQ & A
343404/10/23 34
ConclusionConclusion
Plan for archiving!Plan for archiving!
Understand the impact of doing Understand the impact of doing nothingnothing
Make it part of your Fusion strategyMake it part of your Fusion strategy
04/10/23 35
Plan for Archiving: Next StepsPlan for Archiving: Next StepsData Collection Scripts (DCS) Offer Data Collection Scripts (DCS) Offer
Database measurement at a point in timeSnapshot
Integral part of the Discovery processNon-intrusive to actual data
Provides data/module distributionOracle ApplicationsPeopleSoft
Provides mappings to pre-packaged modulesUnderstand growth trends & areas of growth
04/10/23 36
Next Steps – How to Access DCSNext Steps – How to Access DCS
No-Cost Data Collection Script AnalysisNo-Cost Data Collection Script AnalysisCustomer receives SQL via emailScript is run against production databaseEmail results to John Sosnowski
(john.sosnowski@hp.com)HP analyzes results and provides feedback &
recommendations within 2 weeksProvides growth data & basis for ROI
- Storage optimization and projected performance increases
DCS Script& Survey
to Customer
DCSProcessing
(send to ‘DCS’ alias)REPOS
DCS Spreadsheet
Customer runs DCS, completes survey &Returns via e-mail
.out survey
Non-intrusive Database SQL scripts capture table structure and storage information
Summary: The DCS Process
3939
Thank you!Thank you!
Jeff Blackmon
Jeffrey.blackmon@hp.com
John SosnowskiJohn.sosnowski@hp.com
www.hp.com/go/rim4dbwww.solutionbeacon.com
Real Solutions for the Real World.®
Questions and AnswersQuestions and Answers
Recommended