big data – big archives big protection challenges€¦ · an information explosion ... implements...

35
Linear Tape-Open, LTO, LTO Logo, Ultrium and Ultrium Logo are registered trademarks of HP, IBM and Quantum in the US and other countries. Other symbols may be trademarks of other companies. Linear Tape File System is a trademark of the IBM Corp. Big Data – Big Archives Big Protection Challenges LTO Ultrium 5 Technology and Linear Tape File System Bruce Master Sr. Program Manager IBM Corporation

Upload: hoangkiet

Post on 17-Apr-2018

215 views

Category:

Documents


2 download

TRANSCRIPT

Linear Tape-Open, LTO, LTO Logo, Ultrium and Ultrium Logo are registered trademarks of HP, IBM and Quantum in the US and other countries. Other symbols may be trademarks of

other companies. Linear Tape File System is a trademark of the IBM Corp.

Big Data – Big ArchivesBig Protection Challenges

LTO Ultrium 5 Technologyand

Linear Tape File System

Bruce MasterSr. Program ManagerIBM Corporation

2

Agenda

� Storage Issues

� Best Practices – Unveiling the Facts– Addressing performance, data protection and

retention objectives

� Overview - LTO-5 Technology

� Introducing Linear Tape File System (LTFS)

3

Storage Issues……O bjectives

� Data is growing faster than I can manage

� Not all data is alike

� Costs are out of control

� Data is vulnerable to corruption threats

� My backup repository is growing

� Must protect and preserve data assets now and for the future

4

� 2.5 billion RFID tags sold in 2009

� 900 million GPS devices sold annually by 2013

� 76 million smart electric meters in 2009. 200M by 2014

� Text messages generate 400TB of data daily in the U.S.

� MRIs will generate a petabyte of data globally this year

Information is beaming in from everywhere!

BIG DATA !

High Volume !

5

2000 2005 2010 2015

Terabytes

Petabytes

Exabytes

Zettabytes

Gigabytes

� Storage budgets up 1%-5% in 2010

The information explosionmeets budget reality

� Storage requirements growing 20-40% per year

� Backup and Archive requirements growing 40-50% per year

An Information Explosion - Meets Budget Reality

6

Storage Hierarchy Balanced Diet

DRAM CacheSolid-State Drives (SSD)

Phase Change Memory

FC and SAS disksSAN

Tape

SATA disksVirtual Tape

NAS/iSCSI

Build out automated tiered storage

architecture to optimize

performance, protect data and reduce

costs

7

Rec

over

y T

ime

Obj

ectiv

e (g

uide

lines

onl

y)

Sec

onds

-M

inut

esM

inut

es -

Hou

rs.

Hou

rs -

Day

s

Mission CriticalDynamic Data

Active Online Data

Nearline-Arcchive Data

Not All Data Should Be Treated the Same Way

Disk and DB Mirroring

Electronic Vaulting / Replication

Tape Storage

Value of Data / Financial Investment

Time Value of Data Determines RTOAnd Storage Hierarchy Tiers

8

Most Network Data Sits Untouched

90% Never Accessed

Accessed <5 TimesAccessed >5 Times

Accessed Once

90% Never AccessedAccessed OnceAccessed <5 TimesAccessed >5 Times

Source: Government Computer News, July 1, 2008. “Most network data sits untouched” by Joab Jackson

�Three month study of a businesses 22TB disk data access

�Conducted by University of California, Santa Cruz

�90% of the data was COLD - never accessed after being stored on disk

�Another 6.5% of the data was COOL - accessed only once

�U of C recommendation: move data to less expensive and less energy consuming storage units ….. Use Tape!

Access Patterns of Data Stored on Disk.

9

Numerous Threats Can Corrupt and Destroy Data

ACCIDENTAL•Natural Disaster•System Error•Operation Error

INTENTIONAL•Virus •Theft•Hacker •Sabotage•Disgruntled employee

Risks of downtimeLost revenue and market shareLost productivityNon-compliance Loss of reputation and customer trustLoss of the business

• More than a quarter of the companies in a Forrester Research study declared a disaster in the past five years 1

• 76% of companies have experienced a disaster or major business disruptions 1

1 “Building the Business Case for Disaster Recovery Spending,” Forrester Research, April 2008

10

May 2009: Hackers 'destroy' both

production servers and there was no offline backup

Hard Lessons Learned

Jan 2010: University computer infected with a

virus may have exposed the personal information of 3,500

Feb 2009: Hackers broke into government admin

computers accessing 48K names and social security numbers

Mar 2009: Large pharmacy chain pays

$2.25M in fines for improper disposal of

patient data

Feb 2008: Bank lost unencrypted back-

up tapes with sensitive data about 12

million customers

2011: Software error corrupts many records

on primary and replicated disk.

Offline tape storage recovers records.

May 2009: Hackers 'destroy' both production servers and there was no offline backup

2011: Software error corrupts

many records on primary and

replicated storage. Offline tape

storage recovers records.

11

Tape Saves the Day…Provides Offline Protection

Ben Treynor, VP Engineering and Site Reliability Czar for Google Gmail, used the official Gmail blog to explain the situation where some users lost access to their email accounts during a software update that was buggy.

“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 instancessoftware bugs can affect several copies of the data.That’s what happened here. Some copies of mail were deleted...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”.

12

Flood – Brisbane, Australia - January 2011

March 2011: Japan is Devastated

Protect Data - Out of Region

13

Agenda

� Storage Issues

� Best Practices – Unveiling the Facts– Addressing performance, data protection and

retention objectives

� Overview - LTO-5 Technology

� Introducing Linear Tape File System (LTFS)

14

Best Practices in Data Protection

� Have multiple copies or layers of protection: depending upon value of data, keep at least 3 copies, keep in different locations; one out of region – Use disk and tape

� Isolate one copy: at least one copy offline for logical system isolation to avoid intentional or unintentional corruption that can occur with online storage - Use tape; Keep offline

� Have technology diversification: copies on different forms of media to avoid a media or system process disaster - Use disk and tape

� Protect access to data: at rest and in transit – Use encryption & WORM

� Manage backup differently than archive:– Backup multiple, point in time consistent copies for operational and

DR recovery consistent with application specific RTO/RPO - Use disk and tape

– Archive single instanced data for long term retention: combination of disk & tape

*Best Practices Source: Debbie Beech, Sylvatica Consulting / David Hill, Mesabi Group

15

Large Truck Express Line Survives Hurricane Flood

Business Challenge:

• Hurricane Gastone flooded Data Center with 5 ft. of water

• Total loss of hardware, networks, phone systems,

generator, and utility power

• Good news: a tape backup of 100% of the data was

made the night before – stored off site!

Solution:

• Protect assets and business resilience with comprehensive best practice strategy

• Create nightly disk flash copy for fast retrieval and window-less backup to tape

• Backup 100% production data to LTO tape library nightly: tapes moved offsite

• Global Mirror DR site and backup to LTO tape library – Lights out!

• Creates 5 copies of data (2 offline on tape in different remote locations)

Benefits:

•Able to control TCO, access data and protect data with tiered storage strategy

•No production system interruption

•No save window-Set it & forget it

•No production cycles, no operators, lights out operations

•Logical data protection and out of region protection

Dick CrosbySystems Manager, Estes

"You are out of your mind if you think you can live without tape"

Implements Best Practice Data Protection and Retention Systems

16

“The reports of tapes demise are grossly exaggerated, ”

(borrowing a phrase from Mark Twain)

“Tape is cheap, safe and reliable and there is no substitute….the archive data backstop.

It is the data centre's lifebelt and lifeboat. Without it, when the data loss/data corruption storm strikes, you are sunk. It's that simple.”

Chris Mellor “Tales from the Storage Frontier” May 2011

17

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000P

etab

ytes

LTO 5

DAT 320

DAT 160

VS 160

SDLT S4

SDLT II

SDLT I

LTO 4

LTO 3

LTO 2

LTO 1

DLT IV

DDS-4

DDS-3

Dat 72

17

Tape Storage Continues Phenomenal GrowthTape Storage Continues Phenomenal Growth

Source: Santa Clara Consulting Group (SCCG) – Open System Tape Cartridge Shipments

•A total of 6.6M cartridges shipped in Q4 2011, nearly 90% LTO

•8% YtY Tape Growth

•LTO-5 tape continues to be the rising star of the tape market

-Q4 2011 Shipments up 19% quarter on quarter

-LTO-5 shipments up over 93% YtY Quarterly

82.6% of respondents are still using tape as their final destination for backupsPer 2011 Backup Central End User Survey by TruthInIT, Curtis Prestin, CEO

18

� It’s reliable high speed and capacity– Streams very fast and stores high capacity– Read after write verification for reliable writes– Servo tracking to help ensure precision tracking– Better bit error rate than disk! 1x10E17 bits vs. 1x10E15 bits

� It’s Cost-effective and Green– Lowest storage cost for the foreseeable future– Most energy efficient method of storing digital data – Cartridges on a shelf consume no energy

� It’s scalable – Easy to add additional storage (i.e. add cartridges)– Tape provides “infinite capacity” on demand

� It’s removable / transportable / shareable– Off-line and off-sight storage for data protection-archive– Cartridges are easy to ship (XX PetaBytes / Day)

�But tape is difficult to use!– Tape automation has simplified the process– LTFS makes tape easier to use than ever before

Why is LTO Tape on the Storage Hot List?

And easy

19

Tape Reliability Soars

Both disk and tape have made significant reliability improvements in recent years. For tape, reliability progress has been even better than disk comparing the BER (Bit Error Rate), which is quickly becoming a more popular means of measuring reliability.

Source: Tape Storage Future Directions and the Data ExplosionFred Moore, President, Horison, Inc. 2011

BER is the percentage of bits that have errors relative to the total number of bits received in a data transfer

2020

Costs Comparison Studies

ESG Backup/DR TCO Study: Dedupe VTL vs. LTO-5 Library System1

The Clipper Group Archive TCO Study: SATA Disk System vs. LTO-5 Tape Library System2

•12 Year TCO Archiving Study: costs covering hardware, maintenance, floor space and energy

•Disk storage was 15x Tape TCO

•The cost of energy alone for the average disk-based solution exceeded the entire TCO for the average tape-based solution

1A Comparative TCO Study: VTLs and Physical Tape, By Mark Peters, ESG, Feb. 2011.

2Clipper Notes report “In Search of the Long-Term Archiving Solution -Tape Delivers Significant TCO Advantage over Disk”, The Clipper Group, Dec.23, 2010. This was a general TCO study and did not specifically focus on LTFS or video storage.

•5 Year TCO Backup/DR Study; VTL with 15:1 Deduplication reduction ratio vs. LTO-5 Library

•Costs included hardware, maintenance, floor space, software, people and energy

•Scenarios included various DR methods (i.e. replication, PTAM)

•Dedupe VTL was from about 2-4 times more costly than tape system

3Cartridge price as of internet search Feb 2012.

0

2

4

6

8

10

12

14

16

TCO

DiskTape

15X

LTO-5 Cartridge is about 3 cents per GB uncompressed! 3

21

Archive Capability

Tape Disk

Source: Tape The Digital Curator of the Information Age. By Fred Moore, President, Horison, Inc.

22

Tape and Disk are Complementary for Optimal Performa nce, Archive, Data Protection and TCO

Virtual Tape LibraryTape Library

Application Servers

Backup Server

VTLTape Library

*Source: Fleishman-Hillard Research

Blended Tiered Storage Example - Layers of Protection

“There is no other medium that offers what tape does for archiving.”LTO tape technology continues to evolve with LTO-5, Curtis Preston,

techtarget.com Feb 18, 2010

Storage Manager Survey Results *• 61% of current disk-only users plan to start using tape

ReplicationDR

23

Agenda

� Storage Issues

� Best Practices – Unveiling the Facts– Addressing performance, data protection and

retention objectives

� Overview - LTO-5 Technology

� Introducing Linear Tape File System (LTFS)

24

LTO-5 Tape Can Preserve Large Backups and Archives

• LTO-5 Tape is Huge and Reliable

– 1.5 TB / cartridge native: 3 TB / cartridge (2:1)

– That’s about 30 DVD movies per cartridge

– Automation offerings from 20-1,000s of cartridge slots

– Highly Reliable: Servo Tracking, Read after Write

Verification, 250K MTBF Hours, up to 30 year shelf life

• LTO-5 Drives are Fast

– Up to 140 MB / sec. native

– Up to 280 MB / sec. (2:1 compressed)

– That’s > 1TB of saved data per drive / hr

(2:1 compressed)

LTOTAPE

25

LTO Data Security

� WORM (Write Once Read Many)– LTO 3, 4 and 5 drives and WORM cartridges – Unalterable tape data storage– Can append data to cartridge

� Tape Drive Encryption– LTO 4 and 5 Tape Drive Hardware Encryption– AES 256 bit encryption data key provided to tape drive– Data is compressed then written to tape cartridge in

encrypted form to maximize capacity and protect sensitive information

– Virtually no impact to drive performance– Helps eliminate need for encryption SW or appliance– Get encryption key management software from tape

vendors – Straight forward implementation process

26

Providence Health & Services Encrypts with LTO Tape

� Six data centers in five states all encrypting off-site media

� Daily backups are between 1 – 8 TB per site� Centralized, automated data protection system

eliminated manual management of backups� Effectively established a disk to disk to tape

strategy � Assured data is protected – LTO-4 addresses

security and compliance requirements

*See white paper: Securing Sensitive Information -- with LTO tape drive encryption. by Silverton Consulting at www.ultrium.com/whitepaper

“…it took only 1 to 2 days to implement encryption.” Mack Kigada, Data Storage Manager, Providence Health and Services

27

Agenda

� Storage Issues

� Best Practices – Unveiling the Facts– Addressing performance, data protection and

retention objectives

� Overview - LTO-5 Technology

� Introducing Linear Tape File System (LTFS)

“ Tape not only provides the best option for protecting most data today,

it is finding new roles to play.”Jon Toigo Mar 2012

28

■ A open software specification that allows simple and new ways of accessing data on tape (LTFS spec doc available at: www.ultrium.com/ltfs )

■ Self-describing tape format to address data archive requirements

■ Implemented on dual-partition LTO-5 tape

■ First partition holds the tape index / metadata

■ Second partition holds the content

■ It presents a tape as an extension of the operating system: appears as another drive letter, icon or folder like a disk or memory stick

What is the Linear Tape File System?

LTO Tape Joins the Ranks of Easy to Use Portability

29

LTFS: What are the potential benefits?LTFS: What are the potential benefits?

•Improved archive storage-“Memory stick like” self describing file system

-Tape can tell you what’s on it now and in the future

-Up to 30 year shelf life

•Easy to use-View contents in OS browser directory tree

-Simple “Drag & Drop” movement of data

•Increased data mobility-portability-Compatibility across OS environments

-No backup/archive software needed to view content

-A single storage media standard

-File, HW, SW and camera agnostic

•Reduced costs and energy consumption-LTO tape is less expensive than other storage formats

-Tape is “green”…a cartridge draws no power!

"I am shocked! This is exactly what we need!"

"LTO-5 technology gives tape-less work-

flow....with tape!"

"Now I can offer an LTO-5 archiving

service to my movie clients."

LTFS: How does it work?

• LTFS utilizes media partitioning (new to LTO Gen 5)

• Tape is logically divided “lengthwise” into two partitions

- Index partition : File system info, index, metadata (37.5 GB)

- Content partition : Contains the files / content bodies (1425 GB)

� When mounting the tape, the Index is copied to the workstation/server memory for fast access and updates

� Periodically the index is backed up to the content partition

BOT

EOT

Content Partition

File File File File

Index PartitionGuard Wraps

Index/Metadata

31

Tape browsing on Linux

Files can be accessed on tape directly from any application

Device Directory

Tape Contents

"We think that LTFS could be one of the most significant developments in

the tape drive space since the introduction of LTO itself.“

George Crump, Analyst, Storage Switzerland, May 2010

How it Works: LTFS in Action with File Browser

See LTFS in Action -1 Min. Movie Clip at: www.ultrium.com/ltfs

Oscar Winners

32

LTFS – Easily Exchange, Archive and Share Files

Easy to Use - Archive - Share

LTO-5 Drive

Share

Share

Share

33

Thought Equity Motion: Video Archiving in the Cloud

•Business Need

• Needed a low cost delivery platform for enterprise scale Video Supply Chain as a Service

• Information growth of ~100 TB per month

• Easy self-serve access required by clients

•Solution

• Linear Tape File System at several global locations, including some client facilities

• Tape Libraries and LTO-5 tape drives

•Benefits

• Opened up new business opportunities

• Enabled more predictable and transparent pricing for clients

• Portable, interoperable, scalable, cost-effective data protection and long-term storage

“LTO 5 and LTFS significantly reduce the ancillary costs around storage. This is a real game-changer!”

Mark Lemmons

CTO, Thought Equity Motion

3434

LTO Ultrium Roadmap to the Future

�Over 4M LTO Tape Drive Shipments�Over 200M LTO Cartridge Shipments

35

Protect Your Data Now and Down the Road

LTO Ultrium technology can provide:

�Cost effective backup�Reliable archive�Disaster recovery �Low energy consumption�Ultimate data protection

LEARN more at TRUSTLTO.com

Your Costs - Energy - Data - Company