TSM Symposium 2013 - IBM smarter data protection – which backup target should be used

TSM Symposium 2013: Tivoli Storage Manager: Future Expectations – Vendor Talks

17.-20. September 2013, Hilton Hotel Gendarmenmarkt, Berlin, Germany

Smarter DataProtection

Which Backup Target should be used?

© 2012 IBM Corporation© 2013 IBM Corporation

IBM DataProtection & Retention (DP&R)IBM Smart Backup&Restore SolutionWhich Backup Target should be used?

IBM Data Protection&Retention

Storedata

Josef Weingand +49 171 5526783 - [email protected]

Infos / Find me on: http://sepp4backup.blogspot.de/ https://www.xing.com/profile/Josef_Weingandhttp://www.linkedin.com/pub/josef-weingand/2/788/300http://www.facebook.com/josef.weingand http://de.slideshare.net/JosefWeingandhttps://www.xing.com/net/ibmdataprotection

© 2013 IBM Corporation

Challenge for Backup/Restore

Backups takes too long

Recovery / Restore takes too long

Cost reductionsRessource reductions

More and new Data

Growth Backup

Manage Recover

Data growth


Overall Trends impacting Data Protection & Retention

End-user requirements• More data, data more

valuable • Critical data at ROBO and

individuals• More unstructured file data• Rich Media• Big Data and Analytics

going to change processes by which data is protected

End-user requirements• More data, data more

valuable • Critical data at ROBO and

individuals• More unstructured file data• Rich Media• Big Data and Analytics

going to change processes by which data is protected

IT Organization Constraints (relatively consistent over the years)

• Budget constraints, training budgets slashed

• Fewer people than required• TCO/ROI ever important• Massive complexity from

legacy acquired heterogeneity

IT Organization Constraints (relatively consistent over the years)

• Budget constraints, training budgets slashed

• Fewer people than required• TCO/ROI ever important• Massive complexity from

legacy acquired heterogeneity

Technology Trends• Storage, Server, Compute all cheaper• Increasing availability of Cloud• Shift from block to file, from VTL to NAS• Hardware and Software integration

Technology Trends• Storage, Server, Compute all cheaper• Increasing availability of Cloud• Shift from block to file, from VTL to NAS• Hardware and Software integration

Data Protectio

n & Retentio

n

© 2013 IBM Corporation 5

CPU Performance 8-10x

DRAM Performance 7-9x

Network 100x

Bus Performance 20x

Disk Drive Performance 1-2x

Tape Drive Performance 7-8x

Native Transfer Rate / MB/sec

0

50

100

150

200

250

300

1994 1996 1998 2000 2002 2004 2006 2008 2010 2012

MB

/sec Tape

Disk

Backup-/ Restoretime

0

20

40

60

80

100

120

140

0,1 10 1000 2000 5000 10000 20000 30000 40000 50000

File size / MB

sec Tape

Disk/VTL

History


Challenge of the current and future backup/restor

Hugh Data Growth

But transfer speed was/is not increased in the same manner– Disk drives/system– Network

Bottleneck Backup-Window

Solution: FlashCopy / Snap Backups

– Backups are done by the primary disk – Snaps, FlashCopy, RemoteMirror, FlashCopy Manager

Question:

– Do will still need Backup with TSM? – Is FlashCopy Backup enough?


http://gmailblog.blogspot.com/2011/02/gmail-back-soon-for-everyone.html…. I know what some of you are thinking: how could this happen if we have multiple copies of your data, in multiple data centers? Well, in some rare instances software bugs can affect several copies of the data. That’s what happened here. Some copies of mail were deleted, and we’ve been hard at work over the last 30 hours getting it back for the people affected by this issue.

To protect your information from these unusual bugs, we also back it up to tape. Since the tapes are offline, they’re protected from such software bugs. But restoring data from them also takes longer than transferring your requests to another data center, which is why it’s taken us hours to get the email back instead of milliseconds.

So what caused this problem? We released a storage software update that introduced the unexpected bug, which caused 0.02% of Gmail users to temporarily lose access to their email. When we discovered the problem, we immediately stopped the deployment of the new software and reverted to the old version.

Feb 2011: Google restore gmail from

tape

Hard Lessons Learned

Tape is „last line of defense“!


Trend: SnapBackups

Snap-Backups reduce Backup and Restore/Recovery time

Snap‘s within Storage Systems does not influence Workload on Server/Applications

But: – Take Care about Application integration – consistence!

• Complex Integration• Skill for different layers needed (Application, Disk-System, Backup)

– Snap-Backups does not protect fully • See Amzon Cloud, Google Mail,….

Only the combination of SnapBackups and classic Backup fully protected your data and full fill future requirements!


IBM DP Blueprint

Application System

Application Data

Virtualized Servers AppOS

AppOS

ESX Server

Application

SAN / LAN

FCM - TSM

Unified GUI

TSM

CompressionDeDup

Disk VTLPBBA

Disk + Tape + VTL / PBBA

Integration: Applications- consistence Snap-Backups

Security: offline-Media, Chance of media and technology

Cost: Tiering, Compression, DeDup, Tape > 80%

Scalability :

Efficiency : Unified GUI für Server / Infrastructure / Storage / Backup

Compliance: WORM und Encryption

IBM: Fit For Purpose

Application DataSnapshot

Backup

Snapshot Versions

Snapshot Restore

Snapshot Backup

Snapshot Versions

Snapshot Restore


IBM Smarter DataProtection&RetentionIBM Smarter DataProtection&Retention

• Only the combination of different Backup-Method guarantees Security, Performance and EfficiencySecurity, Performance and Efficiency!

„One fits all“ can not full fill all Requirements

Backup-Targets and Backup-Architectures needs to be aliened for Data class and SLAs

Protect data against logical corruption, physical destruction, unauthorized access

– Maintain multiple copies, geographically separated– Use different types of media

• Our goal is 100% security of your Data!


Protecting the business: Backup & Restore: Majority of clients use both Disk & Tape as part of a tiered architecture

11

Enterprise Strategy Group: - Most Enterprises utilize a combination of Physical Tape & Disk for Back-up & Restore

– More large enterprise users have Tape & Disc installations than Midmarket

– Tape-only clients: One-third more midmarket users than large enterprise

Source: ESG Survey of 510 Users, April 2010

% of users

0

10

20

30

40

50

60

70

Midmarket

Lg Enterprise

Disk OnlyTape Only Disk + Tape

Gartner – Confirms that most clients use disk & tape for backup and restore

Gartner: Which Best Describes Your Primary Backup Methodology?

22%

25%

26%

11%

16%

1. Backup to tape (D2T)

2. Backup to a VTL, then data moved to tape (D2D2T)

3. Backup to disk (not a VTL), then data moved to tape (D2D2T)

4. Backup to a VTL - all data stays on the VTL until purged (D2D)

5. Backup to disk (not a VTL) - all data stays on the disk until purged (D2D)

1 in 4 use only disk


Pros and cons of different Storage Tiers/Technologies

Disk+Fast, random access+parallel access-expansive, power, cooling, floor space

=use it for– Buffering, Small files, improve restore

Disk (VTL) with DeDup / Compression+saves capacity+may reduce capacity-decreased performance/throughput

=use it for– Data with high redundancy, data which needs to be stored for a longer time on disk




VTL+tape connection for LAN-free+Compression / DeDuplication+replication with reduced bandwidth requirements-additional cost vs disk

=use it for– Slow LAN-free backups…and some as above – Replication

Tape+cheap+fast single stream, good scalability -bad access time

=use it for– (LAN-free) backup for large databases / files

Pros and cons of different Storage Tiers/Technologies cont.


How to combine different storage tiers with TSM

Define your goals, SLAs, etc.

Classify your data accordingly our goals– Eg. Large Database, Databases, Mail, file systems, clients

Create blueprint for your different data class– Use and combine different storage tiers in order to get most benefit


Large Databases

Flashcopy T1Flashcopy T2

T0 CopyT0 Copy

DB2

Tape backup– LAN-free

For (large) databases

For backup streams faster than 50-80 MB/sec– Cheaper and faster than disk/VTL

Consider FlashCopy Manager


tt

General Data/ Universal using TSM Pool and Migration to Tape

DiskPool

dedicated

Migration

Use TSM Disk Pool for “caching” the data – Some time later migrate data to Tape

Size Disk Pool for at least 1-2 days of daily backup data– May increase the disk pool size to 5-7 days

Take care about performance – most important single stream performance

– Use dedicated disk system for TSM• Does not use disk system which is shared with productive

data/workload- Performance and data protection issues

– NL SAS/SATA may not full fill the performance requirements– DeDuplication appliances will not full fill performance

requirements

– Recommendation: use SAS/FC Disk for primary disk pool with FC/SAN connection

– Rule of Thumb“ • SAS HDDs: Raid5 8+1 -> ~260 MB/sec• NL SAS HDDs: Raid6 8+2 -> ~85 MB/sec


General Data/ Universal using TSM Pool and Migration to Tape

Further considerations:

Use DeDup for data with long retention on disk– Do not use DeDup for data which are migrated to tape after just a few days

Prefer TSM copy pool instead of VTL/Appliances replication– Copy pool data are accessible without any intervention – VTL/Appliance replication data must be “promoted”

For High Availability mirror the disk pool

TSM active TSM standby

Side A Side B


Small Files / Restore optimized

Kopie

Disk

Disk Backup– Copy Pool on Tape

For (small) Files to improve restore

Consider DeDuplication– TSM or ProtecTier

Save license cost with ProtecTier

600TB 400TB

150TB

ProtecTIER

1 PB

4X dedupe400TB

TSM Primary Storage Pool

Move 600TB onto ProtecTIER

Customer benefits from less TSM license requirement


Data DeDuplication (DDD)

Goal: cheaper Disk-Backup

When do use DDD– Data with high redundancy – Data which needs to be stored for a long time on disk– Replication over WAN

• Saves bandwidth requirements

However– DDD needs CPU “power”– Performance might be slow– Restore might be much slower

• Example backup of database- with LTO3 -> 166 MB/sec- With DDD Appliance 180 MB/sec

• Restore- With LTO3 -> 160 MB/sec- With DDD Appliance 110 MB/sec

DDD ratio depends on data, retention time, number of versions– DDD ration with TSM 2:1 and 15:1– Good for DDD

• Full Backups, database, OS– Challenging for DDD:

• Encrypted data, compressed data• Databases with many reorgs, like Exchange• Databases with compression, like Oracle, DB2• Logs

Keep one copy on tape


VTL for LAN-free

Use it for “slow” and small LAN-free Backups

Save license cost with ProtecTier

600TB 400TB

150TB

ProtecTIER

1 PB

4X dedupe400TB

TSM Primary Storage Pool

Move 600TB onto ProtecTIER

Customer benefits from less TSM license requirement


Disk/VTL & Tape Performance

Tape - TS1140 VTL - TS7650

Library Mount Zeit Ca. 3 – 8 sec < 1 sec

Drive Mount Zeit 14 sec < 1 sec

Positionierung 40 sec < 1 sec

Access Time ~ 60 sec ~ 1-2 sec

Single Stream native 250 MB/sec ~ 80 - 120 MB/sec

Max. Single Stream compressed ~700 MB/sec ~ 80 - 120 MB/sec

Performance 134 400 MB/sec 2 800 MB/sec

Max Stream 192 512

Single Stream Performance Disk vs Tape

0

1000

2000

3000

4000

5000

6000

1 2 3 4 8 12 16 32 64

Number of parallel streams

0

50

100

150

200

250

300

Tape Single Stream

Disk Single Stream

Disk Throughput

Tape Throughput

Backup-/ Restoretime

0

20

40

60

80

100

120

140

0,1 10 1000 2000 5000 10000 20000 30000 40000 50000

File size in MB

sec Tape

Disk/VTL


Disk

TAPE DISK Disk/VTL

IBM DP&A: Smarter Backup Solution

LAN

SAN

SAN Disk

NAS

Disaster Recovery –

3. Location

CopyCopy

DB2

MoveMove

T0 CopyT0 Copy

Disk

Backup Storage

Data Classes


Example

Data Class Blueprint Option 1 Blueprint Option 2

Structured Data

Large DBs + short SLA FLC Manager + LAN free Tape

Large DBs + medium SLA LAN free + Tape

Large DBs + slow network LAN free + VTL

other DBs VTL Disk Pool + Tape

DBs - normal network Tape (über LAN)

DB - Log VTL Disk Pool

Unstructured Data

Virtual environment TSM for VE + FLC Manager, VTL

OS Image VTL TSM Disk Pool mit Dedup.

Full Image VTL TSM Disk Pool mit Dedup.

Incremental short SLA VTL TSM Disk Pool (active Data Pool, dedup.)

Filer Backup Tape Snap + Tape


Infos / Find me on: http://sepp4backup.blogspot.de/https://www.xing.com/profile/Josef_Weingandhttp://www.linkedin.com/pub/josef-weingand/2/788/300http://www.facebook.com/josef.weingand


Disclaimers

The performance data contained herein was obtained in a controlled environment based on the use of specific data. Actual results that may be obtained in other operating environments may vary significantly. These values do not constitute a guarantee of performance.

Product data is accurate as of initial publication and is subject to change without notice.

No part of this presentation may be reproduced or transmitted in any form without written permission from IBM Corporation.

References in this document to IBM products, programs, or services do not imply that IBM intends to make these available in all countries in which IBM operates. Any reference to an IBM program product in this document is not intended to state or imply that only IBM's program product may be used. Any functionally equivalent program may be used instead.

The information provided in this document has not been submitted to any formal IBM test and is distributed "As Is" basis without any warranty either express or implied. The use of this information or the implementation of any of these techniques is a customer responsibility and depends on the customer's ability to evaluate and integrate them into their operating environment. While each item may have been reviewed by IBM for accuracy in a specific situation, there is no guarantee that the same or similar results will be obtained elsewhere. Customers attempting to adapt these techniques to their own environments do so at their own risk.


Trademarks

The following terms are trademarks or registered trademarks of the IBM Corporation in either the United States, other countries or both.

IBM, S/390, ES/3090, ES/9000, AS/400, RS/6000, MVS/ESA, OS/390, VM/ESA, VSE, TPF, OS/2, OS/400, AIX, DFSMS/MVS, DFSMS/VM, ADSTAR Distributed Storage Manager, DFSMSdfp, DFSMSdss, DFSMShsm, DFSMSrmm, FICON, ESCON, Magstar, Seascape

Other company, product, and service names mentioned may be trademarks or registered trademarks of their respective companies.

Windows NT is a registered trademark of Microsoft Corporation.

TSM Symposium 2013 - IBM smarter data protection – which backup target should be used

Technology

Transcript of TSM Symposium 2013 - IBM smarter data protection – which backup target should be used