Nathan MuggliESE Program ManagerMicrosoft Corporation

Exchange Storage for Insiders

Todd LuttinenStore Program ManagerMicrosoft Corporation

ARC306

Agenda• Storage Trends• Larger, but not Faster• Data-At-Rest Protection• Power Consumption

• ESE• Exchange 2013 SP1 ESE Enhancements• Update on Windows Server 2012 R2 and Exchange 2013 Sp1

• Future Exchange Storage Enhancements

• Store• Architecture overview• Exchange 2013 SP1 Store Enhancements• Future Exchange Store Enhancements

• Exchange 2013 recap

Storage Trends

LARGER, BUT NOT FASTER

2012 2013 2014 2015 2016 2017

620 700 820 920 1,050 1,200

4,000 5,000

6,000

10,000

12,000

16,000

1,429 2,000

2,286

HDD Capacity and Areal Density Forecast 2012-2017

Larger, but not Faster Drives3.5 HDD Capacity in GB

Areal Density required @ 7 plattersAreal Density GB sq” (PMR)

SMR & HAMR intended to close the Areal Density gap

Perpendicular Magnetic Recording (PMR)

• Perpendicular magnetic recording (PMR) technology is the current HDD technology

• PMR has reached the upper plateau (1.2TB /platter). PMR assist technologies will be necessary to sustain a 15–20% year-over-year production areal density growth rate

Shingle Magnetic Recording (SMR )

• Shingled Magnetic Recording is an Areal Density “assist” technology which extends capacity (30%) of existing PMR head

• Breaks performance characteristics of traditional Databases (ESE/SQL). -80% IOPS based on Jetstress Test .

Heat Assisted Magnetic Recording (HAMR)

• Manufacturing complexities mean HAMR is still a ways out

Shingled Write Overview

head motion

cornerhead

scans

cross

head motion

cornerheadwritehead

progressivewrites

down track

track

In a shingled write, the data tracks are written in a particular direction radially, and are only written once with a write head wider than the track pitch

Conventional Recording uses a track pitch that keeps data separate.

Track Pitch

Shingle Write Recording overlaps tracks, allowing for a narrower track pitch. Each track is only written once until any remaining valid data is moved.

Track Pitch

Storage Trends

DATA-AT-REST PROTECTION

POWER CONSUMPTION

Storage Trends

ESE on Phone

DEVICE POWER CONSUMPTION

http://research.microsoft.com/~ranveer/docs/email-energy.pdf

Storage Trends - Takeaway Larger but not Faster Drives

Exchange 2013 reduces IOPS by +50% compared with Exchange 2010. Supports multiple databases per volume to maximize available IOPS

JBOD still best for COGS (capacity + performance + cost) Continue riding IOPS/Capacity curve up to 10TB (PMR) Drives

Continue riding IOPS/Capacity curve up to 10TB (PMR) Drives SMR Drives not supported with Exchange 2013

Data-at-rest Protection Utilize Bitlocker for Data-at-rest protection with Exchange 2013 Office365 is already utilizing Bitlocker. From http://trustoffice365.com

“All email content is encrypted on disk using BitLocker Advanced Encryption Standard (AES) encryption. Protection covers all disks on mailbox servers and includes mailbox database files, mailbox transaction log files, search content index files, transport database files, transport transaction log files”

Self-Encrypting-Drives (SED) promising but not viable in Enterprise scenarios due to the requirement of Drives being directly attached to ATA channel.

Power Consumption Don’t ignore benefit of power efficient Storage technologies. Helium technology very promising for both Areal Density increase and Power

consumption decrease (-2 watts compared to conventional drive)

What’s newExchange 2013 SP1 ESE Enhancements

Exchange 2013 SP1 ESE EnhancementsTracing improvements• ESE ETW Tracing

New Keywords and Tasks for every component from transactions to i/o xperf -start JetStressTrace -on Microsoft-Exchange-ESE -f Microsoft-Exchange-ESE-Trace.etlxperf -stop JetStressTracexperf Microsoft-Exchange-ESE-Trace.etl

• ESE IO Debug Channel

Exchange 2013 JetStressPerformance• Database extension +30% improvement

Tool Enhancements• ESEUTIL verbose output options• Log Dump, Space Report, Table/Page/Node Dump

KeywordsError PerformanceTrace TransactionSpaceBFIOLOGTaskBFRESMGRJETTraceTag

TasksESE_TraceESE_BF_TraceESE_Block_TraceESE_NewPage_TraceESE_ReadPage_TraceESE_PrereadPage_TraceESE_WritePage_TraceESE_EvictPage_TraceESE_TouchPage_TraceESE_LatchPage_TraceESE_DirtyPage_TraceESE_TransactionBegin_TraceESE_TransactionCommit_TraceESE_TransactionRollback_TraceESE_AllocExt_TraceESE_FreeExt_TraceESE_AllocPage_TraceESE_FreePage_TraceESE_IOREQHeapEnqueue_TraceESE_IOREQHeapDequeue_TraceESE_IOCompletion_TraceESE_LogStall_TraceESE_LogFlush_TraceESE_EventLogInfo_TraceESE_EventLogWarn_TraceESE_EventLogError_TraceESE_TimerQueueSchedule_TraceESE_TimerQueueRun_TraceESE_TimerQueueCancel_TraceESE_TimerTaskSchedule_TraceESE_TimerTaskRun_TraceESE_TimerTaskCancel_TraceESE_TaskManagerPost_TraceESE_TaskManagerRun_TraceESE_GPTaskManagerPost_TraceESE_GPTaskManagerRun_TraceESE_ThreadCreate_TraceESE_ThreadStart_TraceESE_VersionPage_TraceESE_VersionCopyPage_TraceESE_CacheResize_TraceESE_CacheLimitResize_TraceESE_CacheScavengeProgress_TraceESE_ApiCall_TraceESE_ResMgrInit_TraceESE_ResMgrTerm_TraceESE_CachePage_TraceESE_MarkPageAsSuperCold_Trace

ESE ETW Tracing in Sp1

Exchange Server Jetstress 2013 Tool• Simulate Exchange disk I/O load on a server• Verify performance and stability of your disk subsystem before putting your Exchange

server into a production environment• Jetstress simulates the Exchange database and log file loads produced by a specific

number of users• Use Performance Monitor, Event Viewer, and ESEUTIL in conjunction with Jetstress to

verify that your disk subsystem meets or exceeds the performance criteria you establish

• Download from http://aka.ms/Jetstress2013• Latest Version 15.0.775.8 Published Sep 11, 2013 • Requirements• .NET Framework 4.5• Exchange 2013 binaries: ESE.DLL, ESEPerf.dll, ESEPerf.ini, ESEPerf.hxx• VC11 runtime

Exchange Server 2013 JetStress Tool NEW UI OPTION TO ALLOW JETSTRESS TO CONTINUE RUNNING EVEN IF A

DISK FAILS

Exchange Server 2013 JetStress Tool TEST SUMMARY (FAILURE) DUE TO LOST FLUSHES AND TEST

LOGS

Windows Server 2012 R2 &Exchange 2013 Sp1Storage Technologies and impact on Exchange

Windows Server 2012 R2 and Exchange 2013 SP1Storage Technologies• Storage Spaces

Tested with Exchange 2013 Sp1

• Storage Tiering Tested with Exchange 2013 Sp1

• ReFS for Data Volumes Tested with Exchange 2013 Sp1 Best Practice: Disable Integrity on ESE

DB and Log Files

• Data Dedupe× Not Supported with live Exchange

2013 Sp1 databases

RoadmapFuture Exchange ESE Enhancements

Database ResiliencyDatabase Divergence Detection

Performance / IOPS reduction

Optics

ESE Developer experience

Future Exchange ESE Enhancements

Performance / IOPS reduction• Search Indexing from Passive Copies• Saves network bandwidth and reduces IOPs

• Log compression improvements• Reduces WAN costs through IOPs reduction

• Long Value contiguity improvements

Future Exchange ESE Enhancements

Future Exchange ESE Enhancements

Optics• Tracing and

Eventing improvements

Future Exchange ESE EnhancementsESE Developer

experience ISAM PowerShell Example: Export DB Schema• Persistent Dictionary• ISAM Layer• A managed layer with object model which

simplifies writing DB code• Exposed via an assembly

(Microsoft.Exchange.Isam.dll)• Target Customers are Developers and

sophisticated PowerShell IT Pros

• ESE MSDN Bloghttp://blogs.msdn.com/b/jet/

Exchange 2013Information Store

What is “Store” (aka MSExchangeIS)• Process that manages ESE mailbox

database(s)• Responsible for the following functionality:• Logical schema of data in database (tables, rows, columns)• All read/write access to items in mailbox database via RPC operations• Access control to items within mailbox • Management of indices supporting client views over items• Reliable events supporting event-based assistant infrastructure• Maintenance tasks to disconnect and drop mailboxes (SyncWithDS)

Exchange Storage History (1996-1999)

• Exchange 4.0, Exchange 5.0, Exchange 5.5• Supports i386, Alpha, and MIPS architectures• Bundled directory service (ExDS)• Single native store process directly supporting MAPI clients• (1) mailbox EDB and (1) public folder EDB per server• Local RAID storage (18GB drives, volume sizes less than 200GB )• ~10MB mailbox on server (client PST storage)

Exchange Storage History (2000-2006)

• Exchange 2000, Exchange 2003• Supports i386 architecture (virtual address space up to 3GB)• Integration with Windows Active Directory • Single native store process directly supporting MAPI clients• Store support for 3rd party code in-process (events, VSAPI)• Native protocol processes (POP, IMAP, OWA, SMTP) with front-end

proxy• Inter-process communication via shared memory access • Support for up to 20 databases per server (database = EDB + STM)• Integrated full-text search (1st generation)• Shared storage clusters with Storage Array Networks (RAID LUN)• ~100MB mailbox on server (client OST and PST storage)

• Exchange 2007, Exchange 2010• Supports x64 architecture• Integration with Windows Active Directory • Single native store process directly supporting MAPI clients*• Store support for 3rd party code in-process (events, VSAPI)• Managed Event and Time-Based Assistant Infrastructure • Managed protocol processes (POP, IMAP, OWA, EWS, EAS, MOMT*) • Inter-process communication via XSO/MAPI/RPC over TCP/IP• Support for up to 100 ESE databases per server • Integrated full-text search (2nd generation)• Database replication, SAN and local JBOD (1-2 TB drives)• ~10GB mailbox + ~10GB archive (client OST storage for primary)

* MOMT (RPC Client Access) introduced in Exchange 2010 (no MAPI client access direct to store)

Exchange Storage History (2007-2012)

Exchange Storage History (2013+)

• Exchange 2013 and Beyond• Supports x64 architecture• Integration with Windows Active Directory • Managed store worker process per database• Managed Event and Time-Based Assistant Infrastructure • Managed protocol processes (POP, IMAP, OWA, EWS, EAS, MOMT) via Café• Inter-process communication via XSO/MAPI/RPC over LRPC• Support for up to 100* ESE databases per server • Integrated full-text search (3rd generation)• Database replication, local JBOD (4-8TB drives)• ~200GB mailbox + ~200GB archive (OST client storage with sync slider)

* Exchange 2013 CU2 introduced support for up to 100 databases

Exchange 2013 Storage High Level Architecture

RPC Clientprocesses

Store workerRPC Server Processes with MAPI

and AdminRPC endpoints

Local Inter-Process Communication is RPC-based

Databaseand Process

management

Client-Specific Protocols (POP, IMAP, SMTP, HTTP, EAS, etc)

RopParser/RopHandler

Mapi (context/session/mailbox/folder/message)

LogicalDataModel

PhysicalAccess

Managed ESE

ESE.DLL

StoreCommonServices(Mailbox

State, Mailbox Tasks,

Replid Guid Map,

Extended Property

Map)

DirectoryServices

MAPI.Net/ExRPC

XSO

POP

AD

Dri

ver

MAPI.Net/ExRPC

XSO

EWS

AD

Dri

ver

MAPI.Net/ExRPC

XSO

OWA

AD

Dri

ver

MAPI.Net/ExRPC

XSO

IMAP

AD

Dri

ver

MAPI.Net/ExRPC

XSO

MOMT

AD

Dri

ver

MAPI.Net/ExRPC

XSO

ActiveSync

AD

Dri

ver

Store Worker Process (1 per database)

StoreServiceProcess

MAPI.Net/ExRPC

XSO

MailboxTransport

AD

Dri

ver

MAPI.Net/ExRPC

XSO

Search

AD

Dri

ver

MAPI.Net/ExRPC

XSO

Event and Time-BasedAssistants

AD

Dri

ver

Repl(Active

Manager)

IOPS Reductions: Store Schema Elements

Element E2007 E2010 E2013

Physical Contiguity (ESE)

Poor physical contiguity of leaf pages. Hence many, small size, IOs (1 for each page)

Excellent physical contiguity of leaf pages. So fewer, large size IOs, spanning N pages

Logical Contiguity (Store)

Headers for each folder kept in separate table. So many, small

size, IOs spread over many tables

Folder, Message & Attachment table per mailbox. Message table consists of physical columns and property blobs. High message per page density means fewer large lOs to retrieve many messages for views

Temporal Contiguity (Views)

All views and indexes updated each time a mail is delivered. So many, small size, IOs spread over time

Views and indexes updated only when they are accessed by user. So fewer, large sized, IOs done together.

Replacing Random IO with Sequential IO to squeeze more IOPS from storage

Database Schema ElementsTables optimized for sequential I/O

• Global Tables• Globals – database version, etc • Mailbox – MailboxNumber, Owner Info, Locale, LastLogonTime, etc• DeliveredTo – duplicate delivery information• Events – reliable events for assistants

• Tables partitioned by MailboxNumber• Folder - FolderId, Item Count, Size, PropertyBlob• Message – DocumentId, MessageId, FolderId, PropertyBlob,

OffPagePropertyBlob, MessageClass ordered by DateReceived• Attachment – AttachmentId, Name, Size, CreationTime, etc• PhysicalIndexes (partitioned by LogicalIndex)

Database Schema ElementsIOPS reduction through Message Table Property Storage and Compression

• Blobs used to store collection of MAPI properties• Referred to as On-page and Off-page property blobs• ESE compression optimizes physical storage of blob data• Compression more efficient when input contains more properties

• PropertyBlob• Properties previously stored in Header table in single message table column• Property promotion OffPagePropertyBlob PropertyBlob possible• Blob size limited to eliminate LV tree access for core message properties

• OffPagePropertyBlob• ESE LV Hints push storage of this blob into separate LV tree• Reading LV tree involves large sequential I/O (some fragmentation)

Database ESE cache management• Algorithm will allocate total ESE cache available for all store worker processes based on physical RAM• ~20% of total memory allocated to ESE cache (when running with MaximumActiveDatabases)

• ESE cache allocated to each database (store worker process) based on number of local database copies and MaximumActiveDatabases configuration• Static amount of ESE cache allocated to passive and active database copies

• Passive database allocates 20% of max ESE cache target used for active database• Store worker process will only use max cache target when operating as active

• Max cache target computed at service process startup• Restart service process when adding/removing copies or changing maximum active database

configuration

Exchange IOPS Trend

Exchange 2003

Exchange 2007

Exchange 2010

Exchange 2013

0

0.2

0.4

0.6

0.8

1

DB IOPS/Mailbox

IOPS/Mailbox

+99% * Reduction!

* Similar usage profile (93% reduction with increased usage profile)

Multiple Database Copies/ JBOD DiskActive User Distribution

• Number of copies/disk less than or equal to number of copies per database

• Use DB copy activation preference to distribute active users across available disks hosting database copies

• Budget for server/disk failures and passive IOPS requirements

• If lag copy used, don’t have to capacity plan for all DB copies active

Server2

DAG

DB1 Active

DB2 Passive

DB3 Passive

DB4 Passive

Server1

Server3

Server4

DB1 Passive

DB2 Active

DB3 Passive

DB4 Passive

DB1 Passive

DB2 Passive

DB3 Active

DB4 Passive

DB1 Passive

DB2 Passive

DB3 Passive

DB4 Active

Database Process Isolation & Management

• Microsoft Exchange Replication service process (msexchangerepl.exe)• Responsible for issuing mount/dismount operations to store• Initiates failovers on failures reported by ESE, Store, Search and Responders

• Store service process/controller (Microsoft.Exchange.Store.Service.exe)• Manages worker process lifetime based on mount/dismount operations received• Logs failure item when store worker process problems detected (exit, hung)• Terminates store worker process in response to “dirty” dismount during failover

• Store worker process (Microsoft.Exchange.Store.Worker.exe)• One process per database, RPC endpoint instance is database GUID• Database started in passive state, responsible for block-mode replication • Fast transition to active when mounted, responsible for processing RPC operations• Transition from passive active increases worker ESE cache size 5X

Elimination of scheduled maintenance

• Recurring maintenance implemented within time-based assistant (TBA) infrastructure as multiple assistants:• StoreMaintenance: lazy index maintenance, isinteg, discretionary Subobject cleanup• StoreUrgentMaintenance: SubObject cleanup (threshold exceeded)• StoreDirectoryServiceMaintenance: disconnected mailbox expiration

• Workload Management monitors CPU and replication health • Task execution throttled/deferred when resource pressure exists• RPC latency monitor disabled (to be replaced by disk health resource monitor in future release )

• Background ESE database scanning rate further throttled• Based on datacenter disk failure analysis, target to complete background database scan within 4

weeks (based on multiple databases on 8 TB disks)

• Periodic tasks to generate mailbox quota notification removed• Quota notifications generated at mailbox logon time no more than once daily

Mailbox Quota Management & Notifications

• Over-quota notification overhead reduced• Storage utilization computed in real-time as CRUD operations performed• At logon time, system evaluates mailbox quota against policy and sends over-quota

notification message once per notification interval (daily)• Quota notifications are NOT sent to inactive mailboxes

• Mailbox size calculation is more accurate measurement of mailbox database storage used• Includes both internal and end-user items/properties• Represents 100% of logical storage used by mailbox in database

• Reported mailbox size will likely increase when moved to Exchange 2013• No increase in database footprint, just the attribution of logical space to each

mailbox• Should plan to increase quota per mailbox by ~30% to accommodate change in

space accounting

Exchange 2013 SP1 Store EnhancementsManagement

Database schema updatesNew mailbox shape limitsEnhancements for space investigationsRetrieve status of mailbox repair tasks

ScalabilitySupports more databases per serverConvert additional RPC operations frequently holding exclusive locks to shared locks Limit concurrent waiters on mailbox lock (MailboxLockMaximumWaitCount = 10)

MonitoringAlerting by HealthSet to consolidate number of alerts per serverDatabaseRepeatedMountsMonitor to detect repeated database failoversDatabaseAvailabilityMonitor DatabaseAvailabilityEscalationProcessingMonitor

MiscelleneousLastLogonTime property moved from source server to target serverMore cases support mailbox signature preserving move (no need to restart Outlook)Performance fix for EmptyFolder (schema change)

Database Schema Upgrades (Exchange Team Blog)• Infrastructure to safely upgrade schema of each database in an

Exchange 2013 DAG deployment

• Requests database schema upgrade for each of the local database copies during CU or Service Pack installation• RequestedDatabaseSchemaVersion set to MaximumSupportedDatabaseSchemaVersion for each database

• Current database schema info can be retrieved for each database using Get-MailboxDatabase (using -Status parameter)• CurrentSchemaVersion: current database schema version• RequestedDatabaseSchemaVersion: equal CurrentSchemaVersion after next database mount

• Supported versions of each database copy can be retrieved using Get-MailboxDatabaseCopyStatus• MinimumSupportedDatabaseSchemaVersion: minimum schema version supported by database copy• MaximumSupportedDatabaseSchemaVersion: maximum schema version supported by database copy

• Update-DatabaseSchema can be used to manually request upgrade• Schema upgrade at mount time when RequestedDatabaseSchemaVersion > CurrentSchemaVersion• Mailbox schema upgrade at first logon after database schema upgrade

Example: Database Schema Upgrade

Server2

DAG

DB3 Passive

Server1

Server3

Server4

DB3 Passive

DB4 Passive

DB1 Passive

DB1 Passive

DB2 Passive

1. Remove active databases from S1

2. Upgrade S1, max supported version increased from 0.121 to 0.126 (schema upgrade request for each local DB copy fails because other servers only support 0.121)3. Remove active databases from S2

4. Upgrade S2 , max supported version increased from 0.121 to 0.126 (schema upgrade request for each local DB copy fails because other servers only support 0.121)5. Remove active databases from S3

6. Upgrade S3 , max supported version increased from 0.121 to 0.126 (schema upgrade request for each local DB copy fails because other servers only support 0.121)7. Remove active databases from S4

8. Upgrade S4 , max supported version increased from 0.121 to 0.126 (schema upgrade request for each local DB copy succeeds because all servers support 0.126)

DB1 Active

DB1 Passive

DB1 Passive

0.121 capable0.126

capable0.121

capable0.126

capable0.121

capable0.126

capable0.121

capable0.126

capable

DB2 Passive

DB2 Passive

DB2 Active

DB1 Passive

DB2 Passive

DB3 Passive

DB3 Active

DB3 Passive

DB4 Active

DB4 Passive

DB4 Passive

DB4 Passive

9. Move DB4 to S4 (DB4 schema upgrade succeeds during mount operation)

DB4 Active

DB1, DB2, DB3 schema upgrade on next mount

Mailbox Shape Limits• Enforce practical limits that prevent runaway

usage• Preserve QoS for well behaved mailboxes

• Limits in Exchange Online (online documentation)• MailboxMessagesPerFolderCountWarningQuota : 900000• MailboxMessagesPerFolderCountReceiveQuota : 1000000• DumpsterMessagesPerFolderCountWarningQuota : 2750000• DumpsterMessagesPerFolderCountReceiveQuota : 3000000• FolderHierarchyChildrenCountWarningQuota : 9000• FolderHierarchyChildrenCountReceiveQuota : 10000• FolderHierarchyDepthWarningQuota : 250• FolderHierarchyDepthReceiveQuota : 300

• Limits in on-premises product default to Unlimited• Explicitly set using set-Mailbox cmdlet• Displayed using get-MailboxStatistics cmdlet

# of msgs per

folder# of msgs per dumpster

folder# of folder children

anywhere in hierarchyDepth of folder hierarchy

MailboxStatistics: Physical vs. Logical Space• Previously required ESEUTIL /MS on offline

DB copy• Get-MailboxStatistics extended to display

physical table sizes of each mailbox• Get-MailboxStatistics with identity parameter will provide current stats• Get-MailboxStatistics with database parameter will provide cached

stats

[PS] D:\>Get-MailboxStatistics <MailboxId>| FL *size

TotalDeletedItemSize : 3.545 GB (3,805,959,899 bytes)TotalItemSize : 45.73 GB (49,100,346,075 bytes)

MessageTableTotalSize : 18.02 GB (19,344,031,744 bytes)MessageTableAvailableSize : 31.69 MB (33,226,752 bytes)AttachmentTableTotalSize : 8.071 GB (8,665,759,744 bytes)AttachmentTableAvailableSize : 2.906 MB (3,047,424 bytes)OtherTablesTotalSize : 73.47 MB (77,037,568 bytes)OtherTablesAvailableSize : 736 KB (753,664 bytes)

Logical Mailbox Size (TotalItemSize + TotalDeletedItemSize)

Physical Mailbox Size (MessageTableTotalSize + AttachmentTableTotalSize + OtherTablesTotalSize)

49.273 GB

26.158 GB

Ratio of Logical to

Physical size will vary by

mailbox based on content

Mailbox Repair Task Status• New-MailboxRepairRequest used to create repair tasks, new CorruptionTypes added for more

cases• Can scope all mailboxes on database or a specific mailbox• Some CorruptionType values cause resource expensive tasks that can impact end-user experience

• Get-MailboxRepairRequest will retrieve repair task status from store (not persisted between mounts)• Can scope database, mailbox or specific task ID (returned by New-MailboxRepairRequest)• Summary of progress (% completion) of all tasks (corruption types) and sum of corruptions detected/fixed

[PS] D:\data\scripts>Get-MailboxRepairRequest 425a70a1-7b24-47ef-98f9-3131b7f309b4\da6ef632-e5ed-4193-bbfa-3fa4f70afa11

Identity : 425a70a1-7b24-47ef-98f9-3131b7f309b4\da6ef632-e5ed-4193-bbfa-3fa4f70afa11

Mailbox : ac2dcfd6-555b-460e-85bf-1c656367dc2c

Source : OnDemand

Priority : High

DetectOnly : False

JobState : Succeeded

Progress : 100

Tasks : {SearchFolder, FolderView, AggregateCounts}

CreationTime : 3/18/2014 7:08:49 PM

FinishTime : 3/18/2014 7:10:53 PM

LastExecutionTime : 3/18/2014 7:10:53 PM

CorruptionsDetected : 0

ErrorCode :

CorruptionsFixed : 0

TimeInServer : 00:01:53.1760000

Corruptions : {}

Store-Specific Managed Availability Scenarios

Name Trigger/Recovery sequence

Database Availability 16 logon failures in 22 minutes Escalate

Store service not running 2 failures in 12 minutes Restart service Bugcheck Escalate

Database Free space Free disk space drops below 10% Escalate

Store service process repeatedly crashing 3 crashes for store service in 1 hour Escalate

Store worker process repeatedly crashing 3 crashes for store work (across all workers) in 1 hour Escalate

Percent RPC requests 90% of available threads per database for 10 min Database Failover Escalate

70ms RPC latency 70ms RPC Avg latency for 10 min Determine impact scope Escalate

150ms RPC latency 150ms RPC Avg latency for 10 min Determine impact scope Escalate

Mailbox quarantined More than 1 mailbox quarantined on database for 10 minutes Escalate

Assistants service not running 2 failures in 12 minutes Restart service Escalate

Event assistants behind watermarks Assistant watermark age exceeds 1 hour threshold for 4 hours Escalate

Number of active background tasks Count of active background tasks exceeds threshold for 15 min Escalate

Database Repeatedly Mounting 3 database mount attempts in 1 hour Escalate

Future Exchange Store Enhancements• Pulsing Index Creation• Index operations require exclusive mailbox lock• Divide single long operation into multiple shorter operations• Allows other RPC operations to be processed during index creation• Potential to increase latency of RPC operation that caused index

creation

• Automated Mailbox Repair Tasks• Proactively detect and repair mailbox corruptions• Discretionary workload that occurs during non-peak hours• Tasks interrupted when resource monitors show unhealthy state• Only targets mailboxes with logon activity• Mailbox repair task can occur no more than once per interval (month)

• Management support for passive databases

Summary• Exchange mailbox storage has…• Compared with Exchange 2010, reduced DB IOPS by 40-78%*...again!• Optimized for large disks (up to 8TB) and mailboxes (+100GB) • Optimized for large/slow/low-cost disks (SATA/Tier2)• JBOD/RAID'less storage a viable option for the masses• Better isolation leading to higher reliability• Built-in monitoring and recovery to drive higher availability• Improved with every CU based on experiences from Exchange Online!

* Outlook online vs. cache mode scenarios

Please complete session evaluations

1. Go to the Pre-Release Programs Booth

2. Tell us about your Office 365 environment/or on premises plans

3. Get selected to be in a program

4. Try new features first and give us feedback!

Start now at:http://prereleaseprograms-public.sharepoint.com/

Pre-Release Programs TeamBe first in line!

© 2014 Microsoft Corporation. All rights reserved. Microsoft, Windows and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries.The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.