1367

IBM Cloud Object Storage System

(powered by Cleversafe)
and its Applications
Tony Pearson, IBM
Master Inventor and Senior Engineer

© 2016 IBM Corporation #ibmedge

Abstract
This session will cover private and
public cloud storage options,
including Flash, Disk and Tape to
address the different types of cloud
storage requirements
#ibmedge
The difference between with block, file
and object stores, and where they are
best used for different workloads will be
explained.
1
This week with Tony Pearson
Day Time Topic
Monday All Flash is Not Created Equal:
2:30pm Tony Pearson Contrasts IBM FlashSystem and SSD
Grand Garden Arena, Lower Level, MGM Grand - Studio A
Wednesday All Flash is Not Created Equal:
11:00am Tony Pearson Contrasts IBM FlashSystem and SSD
Grand Garden Arena, Lower Level, MGM Grand - Studio 2

Tony Pearson Presents

1:15pm IBM Cloud Object Storage System and Its Applications
MGM Grand - Room 114

The Pendulum Swings Back: Tony Pearson Explains Converged and

2:30pm Hyperconverged Environments
MGM Grand - Room 113
Thursday Tony Pearson Presents
09:00am IBM's Cloud Storage Options
MGM Grand - Room 116

#ibmedge 2
 Agenda

What is Object Store?

IBM Cloud Object

Storage System

Applications and
Use Cases
Clients are facing explosive growth in Unstructured Data,
which is exactly why Object Storage is so critical

120
Problem - Traditional and Legacy Storage Designed for
Transactional, Not Unstructured Data
100
Unstructured
data growth of
80 60–80%
per year
creates
* Exabytes

60 Web-scale
storage needs

40 Unstructured Data

20

Structured Data
0
2009 2010 2011 2012 2013 2014 2015 2016 2017
*1 exabyte = 1,000 petabytes =1 million terabytes = 1 billion gigabytes Source: IDC

#ibmedge 4
How is Object Storage Different?
Block and File Storage
• Decide where to put it
– For block, which array/volume/LUN
– For file, which filer/subdirectory
• Remember where it is to get it back
• Read/Write records, append data
• Limits on LUN size, number of files
#ibmedge
Object Storage
• Provide data over to the Object storage
– Get “claim stub” reference locator
• Use or share “claim stub” to access data
HTTP, Openstack Swift, S3
• Get/Put/Delete object in its entirety
• Effectively “unlimited” scalability
5
Object Store for Unstructured data
Hot Data Object Store
High-IOPS and Low-Latency is not designed for
All-Flash and Hybrid Flash/Disk • High IOPS workflows
Information Lifecycle Management (ILM) • Transaction
Structured data / Random-Access Processing
Active logs and traces • Inherent ILM
Virtual Machines and VDI
Single-Tenancy

Static and Stable data

Backups, Files, Archives
Seismic, Research, Telemetry, HPC
Object Store provides a Video, Animation, Body Cams
• Secure Photos, Images, CAD/CAM, GIS
• Reliable Music, Audio
• Scalable Genomic, Medical Images
• Cost Effective Multi-tenancy
Platform For Unstructured data
#ibmedge 6
Volume vs. File vs. Object level access

POSIX NFS, SMB REST

Read record Read record Get, Put
Write record Write record Delete

Access Access
Control Control
List List
SAN Zoning, SCSI File Eventual
(ACL) (ACL)
LUN Masking Reserve Locking Consistency

SAN Volume File LAN Object

or Read block LAN Read record or Get, Put,
LAN Write block Write record WAN Delete

• Device • Mount point • Account

• LUN / Volser or Drive Letter • Container
• Block ID • Path (subdirs) • Object
• File name
#ibmedge 7
Object Store Hierarchy – Account, Container and Object
Account
• Object store can have one or more accounts
• Each account could have separate administrators
to manage ACLs
• Provides Multi-tenancy, Secure isolation
Account 1 Account 2 Account 3
Container (aka “Bucket”)
• Container can have one or more objects
• Unique Container name within the account
B2 D4
A1 Object
C3 • A string of bits, similar to a file
• Unique identifier within the container
• No sub-containers, “flat namespace”, but can
mimic subdirs with slashes a/b/c
Photo123 Doc456 Video789 • Unique URI based on account, container and
object id:
Server-URL/Account2/C3/Doc456

#ibmedge 8
Object Storage is Simpler for Application Development

POSIX – over 60 commands NFS – over 30 commands Object – 5 commands

HEAD
• Read metadata
GET
• Read content and
metadata
PUT
• Write content and
metadata
DELETE
• Remove object or
empty container
SMB – over 80 commands POST
• Update metadata

#ibmedge 9
Where Objects can be Stored
Object-on-Database Photo123 Video789
Relational and NoSQL
Key/value stores Doc456
Object-on-File
Account = File System
Container = Fileset Eventually Consistent
Object = File OpenStack Swift
Metadata = Attributes Amazon S3
Basic HTTP

SAN/LAN

Storage-rich Servers
Multiple copies or
LAN/WAN Erasure Coding

Block or File devices

RAID protection

Proprietary, specialized Software Defined, Commodity,

Storage systems Industry-standard

#ibmedge 10
Data Protection Schemes

2.0X 3.0X

RAID-1 / RAID-10 Triplication 1.7X

K pieces 2 x K slices K pieces 3 x K slices

1.2X 1.5X
Erasure Coding
K pieces K+M = N
RAID-5 RAID-6 slices
K pieces K + 1 slices K pieces K + 2 slices

Tolerate 1 drive failure Tolerate 2 drive failures Tolerate “M” failures

#ibmedge 11
 Agenda

What is Object Store?

IBM Cloud Object

Storage System

Applications and
Use Cases
Storage Positioning – Filling a Gap

IBM was looking to offer easy to

manage, scalable disk-based
Highest Performance

object storage for unstructured

Unified file and data
object storage. • Moderate performance
Optimized for high • Moderate cost
performance, across
flash and disk

Information Lifecycle
Unified file and object
Management (ILM) across tiers
storage on tape

Flash 15K 10K 7200 rpm Tape

Lowest cost
#ibmedge 13
IBM acquires Cleversafe, Inc.
Software Company
• Founded in 2004 in Chicago
• Shipping product since 2008 Notable Milestones
Software Defined
Object Storage
Over 350 Patents Awarded
• Over 100 people dedicated to
IBM Cloud software development
Object Storage Leader
Object 2 years in a row
Storage Runs on Industry Standard x86
System servers Multiple Exabytes in
• Certified platforms to provide Production
enterprise grade experience,
predictable performance and support Acquired by IBM 2015

Object Product Renamed

Store • IBM Cloud Object Storage System

#ibmedge 14
Data Growth at Petabyte (PB) Scale
Traditional
Approach

3 to 5x

PB of data Data
Protection Infrastructure Operations
Proprietary, specialized More than 1 FTE per PB
RAID, Mirrors,
hardware, multiple systems Maintenance outages
Replication, Tape
70%
60% Less Lower
1.7 x Hardware &
Rack space TCO

IBM Cloud
Object Storage
Approach Infrastructure Operations
Data Protection
High Availability & Disaster Recovery Software Defined, Less than 1 FTE per 6 PB
Geo-Distribution & Erasure Coding Commodity Hardware, Single system, Secure
Single System Self-healing
#ibmedge 15
IBM Cloud Object Storage System economics beat
legacy NAS storage and Amazon Web Services (AWS)

IBM Cloud Object vs AWS S3

$10,000,000 Amazon
S3 S3
IBM Cloud Object
IBM Cloud Object vs NAS $8,000,000

$6,000,000
IBM Cloud Object $1,053
dsNet object protected $4,000,000

IBM Cloud Object + NAS $1,613 $2,000,000

gateway
$0
$4,210 480 TB 960 TB 1920 TB 3840 TB
Legacy NAS single copy
Cost: 30 to 60% lower
Legacy NAS DR $8,400
protected

Cost: 80% lower IBM Cloud Object Storage Dedicated

1.7x faster “read” and 9.9x faster “write”
performance than Amazon S3.
--- Frost & Sullivan
#ibmedge 16
IBM Cloud Object Storage System – Components
dsNet® Manager
• Fault management
• Performance monitoring
• Storage configuration
• Reporting
• Provisioning
Accesser® Slicestor®
• Slices data • Storage for slices
• Disperses data • Single or multi-site
• Retrieves data • Capacity based pricing
• Use Load Balancer • Rebuilds slices
across accesser pools
• Stateless
• No cache
• No metadata

Supported on Certified industry standard platforms

Qualified for predictable performance

Software Defined & Faster time to production
Hardware Aware Choice of drive technology
Single Pane Management of HW & SW

#ibmedge 17
IBM Cloud Object Storage System – Terminology
Global Namespace Load
End user Accesser Balancer
HTTP
Application
OpenStack Swift
Server
Amazon S3
Vault Accesser Pool

Object Store organized in Accounts,

Containers and Objects
Storage
IBM Cloud Object Storage System Pool
uses “Vaults”.
Slicestor
Vault = Account / Container

Multiple vaults can share

storage pools dsNet
Manger Site A Site B Site C
#ibmedge 18
Encryption Data-in-Flight
End user
Slicestor
All data-in-flight secured by TLS/SSL

HTTP

Application Accesser dsNet

Server protocol

In all cases, monitoring and

event management secured
by AES in SNMP v3
Rogue devices cannot
dsNet become part of the system
Manger
#ibmedge 19
IBM Cloud Object Storage System – Single vs. Multi-Site

Single Site Two Site / Vault Mirroring

Better performance, Allows customers to leverage existing two-site
when site-tolerance infrastructure. Provides concurrent reads and
not a factor, better writes despite communication disruption
than traditional between locations. Local data better than
RAID-5 / RAID-6 traditional RAID-5 / RAID-6

Geographically Dispersed
Three to Nine Sites
Consider adding IBM SoftLayer or leverage
existing datacenter locations to provide a broader
distribution of data for higher availability, site-
tolerance and scalability

#ibmedge 20
How the IBM Cloud Object Storage System Works

CONTENT
TRANSFORMATION BENEFITS
Data Ingest Data Retrieval

IBM Cloud software The level of

encrypts, slices and resiliency is fully
applies Information customizable
Accesser Accesser
Dispersal Algorithms Software Software resulting in a
otherwise known as massively reliable
erasure coding and efficient way to
policies to the data. store data at scale
Site 1 Site 2 Site 3 Site 1 Site 2 Site 3
as opposed to RAID
and replication
techniques.
Slicestor
Software

Storage Nodes Storage Nodes

Physical Distribution Reliable Retrieval

Slices are distributed to An operator defined subset of

separate disks and industry slices is needed to retrieve
standard x86 hardware data bit perfectly in real time.
across geographic locations.

#ibmedge 21
Writing Data to IBM Cloud Object Storage System
Encrypted, Erasure
Coded Slices Information Dispersal
Original Data
Accesser® Appliance, Application,
12
11
Algorithm (IDA)
VM, Docker Container or Embedded 10
9
Erasure coding is used
8
7
to transform the data
Accesser® into a customizable
$ 7
6
5
4 number of slices (7/12
6 3
5
4 2 in this example)
3 1
Original object is 2
1
encrypted then
cut into pieces

Each slice is written to a

separate storage node. In Slicestor®
this example, the storage Appliances
nodes are geographically
dispersed across 3 sites.
Site 1 Site 2 Site 3
#ibmedge 22
“Zero Touch” Encryption Data-at-Rest
Introducing the All-or-Nothing (AONT) package

#ibmedge 23
IBM Cloud Object Storage System – SmartRead

#ibmedge 24
Scalability

S3 Compatible API Openstack Swift Simple Object API

Compatible API
Scalability – Scale performance
Accesser and/or capacity at any time with no
Software
downtime to operations

Site 1 Site 2 Site 3 Need more Performance?

Slicestor
Software
Add more Accesser nodes
Storage
Pool 1
Need more Capacity?
Add more disks to existing Slicestor
nodes, or add more storage pools
Storage
Pool 2 – All pools must have the same
number of nodes
– Difference storage pools can
have different amounts of storage
– All nodes in each storage pool
must have same amount of
Storage
Pool 3
storage

#ibmedge 25
IBM Cloud Object Storage System –
Software, Pre-Built Systems or Cloud Services
Software
• ClevOS – IBM Cloud Object
Storage System software
packaged with Debian Linux OS
• Software-defined, hardware-
aware model for flexibility of x86
platform choice
• Can be deployed on qualified
vendor hardware
Cloud Services
• IBM SoftLayer hardware
infrastructure running IBM Cloud
Pre-built Systems
Object Storage System software
• Fully integrated appliance • Off-premise offering for
models for easy customers that want security and
deployment and support controlled performance
• On-premises object
storage solution
#ibmedge 26
IBM Cloud Object Storage System –
Deployment Options
Cloud Public Consumed
Infrastructure • Standard regional pricing
Public Off-premises • Nearline regional
(IBM SoftLayer) • Nearline geo-
dispersed
Dedicated Dedicated Allocated
(Private) • IBM Managed pricing
Off-premises • Client managed
• Hybrid / Mixed
On-premises
Client Locally managed
infrastructure • Software
Local
(Private)
On-premises • Pre-built systems

(Recommend:
500TB or more)
Infrastructure
#ibmedge 27
Pre-Built Systems for IBM Cloud Object Storage System

dsNet Manager 2105 / 3105 Slicestor 2212

12 drives @ 4, 6 or 8 TB
Health and performance monitoring 48 to 96 TB Nearline HDD
GUI and API access in 2U rack space
Configuration and security
Zero downtime upgrades
Slicestor 2448

Accesser 2100 / 3105 / 4105

16, 32 or 48 drives @ 4, 6 or 8 TB
Slices, Disperses and Retrieves data 64 to 384 TB Nearline HDD
in 4U rack space

#ibmedge 28
IBM Cloud Object Storage – SmartWrite and Rebuild
Optimistically attempts to write all slices in parallel
X ?

Once write threshold is met, returns Entire namespace scanned on ongoing

success response to client basis
• Example: 7/9/12 (7 pieces 12 slices) • Slice integrity check
• Minimum 7 slices required to read back • Missing slice check
• Once 9 slices are written, write is • Slicestors work together to rebuild
considered complete for application missing or corrupted slice(s)

• Best effort to write remaining slices • Rebuilder is “always on” at moderate

• If time-out occurs, unwritten slices dropped rate for I/O rate predictability

#ibmedge 29
Reading Data from IBM Cloud Object Storage System
With erasure coding “k” pieces are turned into “n” slices:

Reads can be performed using any k of the n slices

• This example is a “7 of 12” Information Dispersal Algorithm (IDA)
means only 7 slices are needed to reconstruct the original object

7
6
With this IDA, a read 5
4
can still be executed
with any five storage 3 $
2
nodes being 1
unavailable out of 12.

With 3 sites,
even an entire
Site 1 Site 2 Site 3 site outage (plus
one additional
storage node
outage) can be
tolerated.
#ibmedge 30
The Math Behind Reed-Solomon Erasure Coding

“K” variable of input data: a, b, c, d, e

Here we create “K+M”

equations, adding and
subtracting by different
co-factors

Results in “K+M” slices

that can be geographically
dispersed

We can tolerate losing up

to “M” slices of data, and
still solve for the original
“K” pieces of data.

#ibmedge 31
IBM Cloud Object Storage System – Access Methods
NAS (NFS/SMB) Backup/Archive General Applications
DIRECT API ACCESS
The Accesser Software exposes three REST
APIs for ingest and retrieval. Applications with
knowledge of these APIs can leverage IBM
Cloud Object Storage directly. Hadoop

Data
S3 Compatible API Openstack Swift Simple Object API
Compatible API

Accesser
Software

S3 Compatible API Openstack Swift Simple Object API Site 1 Site 2 Site 3
Compatible API
Slicestor
Software

Accesser
Software

Slicestor
Site 1 Site 2 Site 3 PARTNER BASED
Software
A variety of Certified technology partners can leverage IBM’s
multi tenancy support to satisfy concurrent use cases on a
single IBM Cloud Object Storage instance.

#ibmedge 32
 Agenda

What is Object Store?

IBM Cloud Object

Storage System

Applications and
Use Cases
Market Verticals

Media & Media & Service Financial Heath Care & Government
Entertainmen Entertainment Providers Services & Life Sciences
t Distribution Insurance
Production
Consistently Scale your Secure customer Put medical
Have reliable Manage the data
create engaging storage for your on market offering trust and business progress before essential to serving the
experiences. demand content. without worry. compliance. everything else. good of the public.

Active Archive
Enterprise Storage as a
Content Repository STaaS
Service (STaaS)
Content Genomics
Collaboration Collaboration

Enterprise Collaboration

Backup

#ibmedge 34
Client Reference– Web Based Photo Sharing

Photo and video objects are sent to

Cleversafe via REST based protocols

Users upload photo Metadata is

and video content via captured and
web based application stored

• Scale – 130 petabytes and growing: more than 50 Billion images stored
• Manageability – 3 Administrators manage entire environment
• Security – 50,000+ uploads per minute with zero touch security
• Always-on availability – SLA of 100% download on demand – even during
California to Nevada datacenter move
• Economics – Operating costs reduced by more than 70%
• Key decision makers – Technical team backed by financial cost cutting
mandates

#ibmedge 35
IBM Cloud Object Storage System – Partner Comparison
Vendor Product Protocols Features Use Cases
Avere • NFS Global namespace • Analytics
FXT series • SMB Encryption, Compression, Snapshots, • Media rendering
Clustering
Ctera Networks • NFS, SMB Encryption, Snapshots, Replication, • Backup
CX0 series • AFP, FTP Enterprise File Sync and Share (EFSS) • RO/BO storage
• WebDAV
• iSCSI

IBM Spectrum Scale • NFS, SMB Global namespace & file locking, • Analytics / HPC
• OpenStack Encryption, Snapshots, Compression, • Media rendering
• Amazon S3 Replication • VMs and Databases
• Hadoop • Collaboration
• POSIX • NAS Consolidation
• Backup / Archiving
Nasuni NF series • NFS, SMB Global namespace & file locking, • Collaboration
• FTP, SFTP Encryption, Snapshots, Compression, • File sharing
• HTTPS Dedupe, EFSS • Archiving
• iSCSI

Panzura Global File • NFS Global namespace & file locking, • Collaboration
System • SMB Encryption, Compression, Dedupe • NAS consolidation

#ibmedge 36
IBM Spectrum Protect supports Cloud Object Storage!

Client nodes

Off-premises:
• IBM SoftLayer

IBM Spectrum • IBM Cloud Object Storage System

Protect Server (using S3-compatible API)
On-premises
#ibmedge 37
IBM Spectrum Storage and IBM Cloud Object Storage
Transparent Cloud Tiering
Object
Highest Performance

Store

Unified file and object IBM SoftLayer

storage. Optimized for OpenStack Swift
high performance, across Amazon Web Services S3
flash, disk and object Swift S3 emulation
store
IBM Cloud Object Storage System
( File, backup and archive interfaces
available through variety of options )

Information Lifecycle Unified file and object

Management (ILM) across tiers storage on tape

Flash 15K 10K 7200 rpm Tape

Lowest cost

#ibmedge 38
Competitive Differentiators

ENCRYPTION IS INHERENT ERASURE CODING IS INLINE

No need for 3rd party Not a post process. Data is

encryption solution or key fully protected upon write
management commit

SOFTWARE PACKAGED WITH OS DISK LIFECYCLE MANAGEMENT

No need to install and Predict drive failures and

manage a separate OS take appropriate action

ALL FEATURES INCLUDED PERPETUAL LICENSING MODE

Base price includes all No need to repurchase
interfaces licenses when refreshing
hardware

#ibmedge 39
Thank You

© 2016 IBM Corporation #ibmedge

IBM Redbooks on IBM Spectrum Scale

IBM Spectrum Implementing IBM IBM Spectrum IBM Spectrum IBM Spectrum
Scale (formerly Spectrum Scale Scale in an Scale – Big Data Scale and ECM
GPFS) OpenStack and Analytics FileNet Content
Environment Solution Manager

www.redbooks.ibm.com
#ibmedge 41
IBM Tucson Executive Briefing Center

• Tucson, Arizona is home for storage

hardware and software design and
development

• IBM Tucson Executive Briefing Center

offers:
• Technology briefings
• Product demonstrations
• Solution workshops

• Take a video tour!

• http://youtu.be/CXrpoCZAazg

#ibmedge
42
 Tony Pearson
9000 S. Rita Road

About the Speaker Master Inventor

Senior Software
Bldg 9032 Floor 1
Tucson, AZ 85744
Engineer
+1 520-799-4309 (Office)
IBM Storage
tpearson@us.ibm.com

Tony Pearson is a Master Inventor and Senior Software Engineer for the IBM Storage product line. Tony joined IBM
Corporation in 1986 in Tucson, Arizona, USA, and has lived there ever since. In his current role, Tony presents briefings on
storage topics covering the entire IBM Storage product line, IBM Spectrum Storage software products, and topics related to
Cloud Computing, Analytics and Cognitive Solutions. He interacts with clients, speaks at conferences and events, and leads
client workshops to help clients with strategic planning for IBM’s integrated set of storage management software, hardware, and
virtualization products.

Tony writes the “Inside System Storage” blog, which is read by hundreds of clients, IBM sales reps and IBM Business Partners
every week. This blog was rated one of the top 10 blogs for the IT storage industry by “Networking World” magazine, and #1
most read IBM blog on IBM’s developerWorks. The blog has been published in series of books, Inside System Storage: Volume
I through V.

Over the past years, Tony has worked in development, marketing and customer care positions for various storage hardware and
software products. Tony has a Bachelor of Science degree in Software Engineering, and a Master of Science degree in
Electrical Engineering, both from the University of Arizona. Tony holds 19 patents for inventions on storage hardware and
software products.

#ibmedge 43
Additional Resources from Tony Pearson

Email:
tpearson@us.ibm.com

Twitter:
twitter.com/az990tony

Blog:
ibm.co/Pearson

Books:
www.lulu.com/spotlight/990_tony

IBM Expert Network on Slideshare:

www.slideshare.net/az990tony

Facebook:
www.facebook.com/tony.pearson.16121

Linkedin:
https://www.linkedin.com/in/az990tony
#ibmedge 44
Please Note: Edge 2016 Disclaimers

• IBM’s statements regarding its plans, directions, and intent are subject to change or
withdrawal without notice and at IBM’s sole discretion.

• Information regarding potential future products is intended to outline our general product
direction and it should not be relied on in making a purchasing decision.

• The information mentioned regarding potential future products is not a commitment, promise,
or legal obligation to deliver any material, code or functionality. Information about potential
future products may not be incorporated into any contract.

• The development, release, and timing of any future features or functionality described for our
products remains at our sole discretion.

• Performance is based on measurements and projections using standard IBM benchmarks in

a controlled environment. The actual throughput or performance that any user will
experience will vary depending upon many factors, including considerations such as the
amount of multiprogramming in the user’s job stream, the I/O configuration, the storage
configuration, and the workload processed. Therefore, no assurance can be given that an
individual user will achieve results similar to those stated here.

#ibmedge 45
Trademarks and Other Disclaimers
Adobe, the Adobe logo, PostScript, and the PostScript logo are either registered trademarks or trademarks of Adobe Systems Incorporated in the United States, and/or other countries. IT Infrastructure Library is a registered trademark of the Central
Computer and Telecommunications Agency which is now part of the Office of Government Commerce. Intel, Intel logo, Intel Inside, Intel Inside logo, Intel Centrino, Intel Centrino logo, Celeron, Intel Xeon, Intel SpeedStep, Itanium, and Pentium are
trademarks or registered trademarks of Intel Corporation or its subsidiaries in the United States and other countries. Linux is a registered trademark of Linus Torvalds in the United States, other countries, or both. Microsoft, Windows, Windows NT,
and the Windows logo are trademarks of Microsoft Corporation in the United States, other countries, or both. ITIL is a registered trademark, and a registered community trademark of the Office of Government Commerce, and is registered in the U.S.
Patent and Trademark Office. UNIX is a registered trademark of The Open Group in the United States and other countries. Java and all Java-based trademarks and logos are trademarks or registered trademarks of Oracle and/or its affiliates. Cell
Broadband Engine is a trademark of Sony Computer Entertainment, Inc. in the United States, other countries, or both and is used under license therefrom. Linear Tape-Open, LTO, the LTO Logo, Ultrium, and the Ultrium logo are trademarks of HP,
IBM Corp. and Quantum in the U.S. and other countries.

Other product and service names might be trademarks of IBM or other companies. Information is provided "AS IS" without warranty of any kind

The customer examples described are presented as illustrations of how those customers have used IBM products and the results they may have achieved. Actual environmental costs and performance characteristics may vary by customer.

Information concerning non-IBM products was obtained from a supplier of these products, published announcement material, or other publicly available sources and does not constitute an endorsement of such products by IBM. Sources for non-IBM
list prices and performance numbers are taken from publicly available information, including vendor announcements and vendor worldwide homepages. IBM has not tested these products and cannot confirm the accuracy of performance, capability,
or any other claims related to non-IBM products. Questions on the capability of non-IBM products should be addressed to the supplier of those products.

All statements regarding IBM future direction and intent are subject to change or withdrawal without notice, and represent goals and objectives only.

Some information addresses anticipated future capabilities. Such information is not intended as a definitive statement of a commitment to specific levels of performance, function or delivery schedules with respect to any future products. Such
commitments are only made in IBM product announcements. The information is presented here to communicate IBM's current investment and development activities as a good faith effort to help with our customers' future planning.

Performance is based on measurements and projections using standard IBM benchmarks in a controlled environment. The actual throughput or performance that any user will experience will vary depending upon considerations such as the amount
of multiprogramming in the user's job stream, the I/O configuration, the storage configuration, and the workload processed. Therefore, no assurance can be given that an individual user will achieve throughput or performance improvements
equivalent to the ratios stated here.

Prices are suggested U.S. list prices and are subject to change without notice. Starting price may not include a hard drive, operating system or other features. Contact your IBM representative or Business Partner for the most current pricing in your
geography.

Photographs shown may be engineering prototypes. Changes may be incorporated in production models.

© IBM Corporation 2016. All rights reserved. References in this document to IBM products or services do not imply that IBM intends to make them available in every country.

Trademarks of International Business Machines Corporation in the United States, other countries, or both can be found on the
World Wide Web at http://www.ibm.com/legal/copytrade.shtml.

ZSP03490-USEN-00

#ibmedge 46