S014066 scale-ess-orlando-v1705a

1
IBM Systems Technical Events | ibm.com/training/events
© Copyright IBM Corporation 2017. Technical University/Symposia materials may not
be reproduced in whole or in part without the prior written permission of IBM.
IBM Spectrum Scale™
for File and Object Storage
Tony Pearson
Master Inventor and Senior IT Architect
IBM Corporation

2
IBM is ranked #1 in Software
Defined Storage, and the
latest release of IBM
Spectrum Scale provides
universal access to files and
objects across your data
centers.
IBM offers software, pre-built
systems and cloud services
that take advantage of IBM
Spectrum Scale file and
object storage.
This session will cover how it
works, how it scales, and how
it delivers on performance.
Come learn how Spectrum
Scale and Elastic Storage
Server can help you.
Abstract

3
This week with Tony Pearson
Day Time Topic
Monday 10:15am
The Pendulum Swings Back –
Understanding Converged and Hyperconverged Environments
11:30am IBM’s Cloud Storage Options
3:15pm Software Defined Storage -- Why? What? How?
Tuesday 9:00am
Business Continuity –
The seven tiers of business continuity and disaster recovery
3:15pm
Introduction to IBM Cloud Object Storage System and its
Applications (powered by Cleversafe)
4:30pm
New generation of storage tiering: Simpler management,
Lower costs and Increased performance
Thursday 1:45pm IBM Spectrum Scale for File and Object Storage
5:45pm Meet the Experts: IBM Storage

4
The Problem:
Islands of Block, File and Object level data
Volume-level Storage
OS-specific file systems on
block-based devices
Sharing requires file transfers
Provides “Context” for
Analytics of Social and Mobile
transactions
File-level Storage
NAS encourages sharing
across social networks
Desire for file sync-and-
share across desktops and
mobile
HDFS requires transfer
(ingest) from other sources
JFS2
EXT4
NTFS
SMB
HDFS
NFS
Object-level Storage
New Web and Mobile apps
prefer Object-level access
Amazon S3
OpenStack
Swift

5
Why use IBM Spectrum Scale™
Extreme Scalability
Add or Remove nodes and
storage, without disruption or
performance impact to
applications
Universal Access to Data
All servers and clients have access to
data through a variety of file and object
protocols
High Performance
Parallel access with no hot spots
Proven Reliability
Used by over 200 of the top 500 Supercomputers
Survive any node or storage failure with Distributed
RAID and redundant components

6
No single-server performance
and bottleneck scaling limits
- NAS filers can be a bottleneck and
significantly impacts system
performance
No centralized metadata server
- Centralized metadata server can
be a performance bottleneck for
metadata intensive operations
TCP/IP Network
Network
File Server
Client
Nodes
Storage
Metadata
Data Data
Network
data
metadata
data
Centralized
Metadata
Server
) (
) (
IBM Spectrum Scale™ is different than
other clustered/distributed storage solutions

7
IBM Spectrum Scale™ –
Software, Systems or Cloud Services
Software
• Install software on your
own choice of Industry
standard x86, POWER
servers, and z Systems
Pre-built Systems
• Elastic Storage Server
• Storwize V7000 Unified
Cloud Services
• Spectrum Scale can be
deployed on any Cloud,
including IBM SoftLayer
Scale

8
IBM Spectrum Scale – Flexible File and Object Storage
FS1 FS256. . .Exabyte-Scale,
Global Namespace
One big file system or divide into as
many as 256 smaller file/object
systems
Each file system can
be further divided into
fileset containers
Network Shared Disk
(NSD) refers to:
• Flash and Disk
devices
• Servers connected
to these devices
• Protocol between
clients and servers
Metadata can be separated to
its own Pool or intermixed with
data
Files and objects
can be migrated to
Tape, Object store,
or Cloud

9
IBM Spectrum Scale™ – Supported Topologies
Twin-tailed
SAN
Internal, Direct-Attach
Shared PoolsShare-Nothing Pools
NSD Servers
Access files on direct,
twin-tailed or SAN
attached disk
Can export files to
application nodes
File Placement
Optimization (FPO)
Servers
For AIX, Linux-x86
and Linux on POWER
Access files on direct
attached disk
Exports files to other
FPO servers
Hyperconverged
External Clients
Access data via iSCSI, NAS and
object protocols over IP network
TCP/IP
NSD Clients
For Linux, AIX,
and Windows
Access files via
SAN, TCP/IP or
RDMA
TCP/IP or RDMA network

10
IBM Spectrum Scale™ replaces other POSIX file systems
SAN
Direct-Attach
SAN-level Storage
JFS2
EXT4
NTFS
• Works like OS-specific file systems
• No file transfers required between OS
• Linux on x86, POWER and z Systems
TCP/IP or RDMA Network
Twin-tailed

11
IBM Spectrum Scale™ – More than just a file system!
ROBO
Other NAS
Other
Datacenters
Scale
Active File
Management
(AFM) caches
data to where it is
needed, can be
used to migrate
from other NAS
Hierarchical Storage
Management (HSM) migrates
infrequently accessed files to
tape or object-based cloud,
automatically recalls back when
accessed
Local Read-Only Cache (LROC)
and Highly Available Write Cache
(HAWC) caches the busiest blocks of
files on local flash
Disaster Recovery
(DR) remotely mirrors
data to remote locations
Migrate/Recall
Tape
NSD Client
Information Lifecycle
Management (ILM) moves data
across tiers of flash and disk
Object
Cloud

12
IBM Spectrum Scale™ – Local Read-only Cache
NSD Clients
Local Flash on NSD Clients
Automatically handles the flash cache so data is
transparently available to your application with very
low latency and no code changes
Accelerates I/O performance up to 6x by reducing
the amount of time CPUs wait for data
Improves application performance while keeping all
the manageability benefits of shared storage
Data is never stale
• Cache consistency ensured by standard tokens
• Data is protected by checksum and verified on read
• Write cache on two separate client nodes, or shared
fast storage device
Reduces Network Load
• Decreases the overall load on the IP network,
benefitting performance for others
Local Read-Only Cache
(LROC) and Highly-Available
Write Cache (HAWC) caches
the busiest blocks of files on
local flash

13
IBM Spectrum Scale™ – Policy Management
Migrate/Recall
File Placement
When new files are created, the active
policy assigns it to the right pool – Flash,
15K, 10K or Nearline disk
Files can be marked for having
2 or 3 replicas
Files can be encrypted with
specific keys
Immutability (NENR compliance)
Compression
File Expiration
Delete files automatically
after they are no longer
needed
File Movement (ILM)
Move files between pools
Based on age, size, heat, access
frequently or other criteria
File Movement (HSM)
Migrate files to an external pool
of tapes, object store, or Cloud
Accessed files are automatically
recalled back to internal pool of
flash or disk
Hierarchical Storage
Management (HSM)
Management (ILM)
Tape
Object
Cloud

14
IBM Spectrum Scale™ – Active File Management (AFM)
ROBO
Other
Datacenters
Cloud
Global Namespace
Shows entire file system, on flash, disk and
tape, across all locations
Pre-fetch or Pull on demand
Files can be periodically
pre-fetched in advance, or pulled on
demand when needed
WAN Caching
Files you use most often are cached to
your location for faster access and
availability to avoid WAN delays
NAS Data Migration
Use AFM to cache or migrate data
from other NAS filers
Active File
Management
(AFM)
Other NAS

15
IBM Spectrum Scale™ – Backup and Disaster Recovery
Disaster Recovery
(DR) mirrors data to
remote locations
Backup/Recover
FS1
snap1 snap2
Backup to External Media
• Files can be backed up to
IBM Spectrum Protect, or
third-party backup software
• Snapshotroot parameter
Remote Mirror
• Use Active File Manager
across data center locations
• Specify appropriate RPO
Snapshots
• Up to 256 Snapshots of entire file system, and 256
Snapshots of each file set
• Read-Only, Space-Efficient
• Microsoft VSS Interface
• Writeable File Clones

16
IBM Spectrum Scale™ -- Graphical User Interface
Provide an easy-to-use Graphical
User Interface for common tasks
Base interface on common IBM
Storage Framework

17
Native Encryption and Secure Erase
Application
IBM Security Key
Lifecycle Manager (SKLM)
TCP/IP or
RDMA
Node-to-Node Encryption
Complies with NIST SP 800-131A
Data-at-Rest Encryption
• Files are encrypted by application node
• Each file assigned random File-key
• Master-key granularity by file or fileset, determined by
policies
• IBM SKLM stores Master-keys, and nodes must have
appropriate SKLM credentials
• Data is encrypted from application node all the way to
NSD (flash or disk) media
• FIPS 140-2 certified
Secure Erase
• Files are cryptographically erased by deleting their
Master-key
• Files that “stay” are re-Mastered to new key

18
SAN
Twin-tailed
Protocol Servers
iSCSI,
NFS v3/v4,
SMB2, SMB3
AIX, Linux, Mac OS,
Windows, VMware,
z/OS, etc.
Feature of IBM Spectrum Scale on
Linux nodes
Share files with clients using NFS,
SMB and Object protocols
All nodes can share the same data
If Protocol Server Node fails,
client connections are moved to
another server
Protocol Server Node(s) need “NSD
Server” License
External clients need no Spectrum
Scale License
Files can be accessed as objects,
objects can be accessed as files!
Clustered Protocol Servers for File and Object access
TCP/IP
OpenStack,
S3 Protocol

19
Spectrum Scale supports OpenStack environments
Glance
• VM images
• Disk images
Cinder
• Volumes
Swift
• Objects
Manila
• File systems
Global Name Space
Volume-on-file Object-on-file
• Create, Delete and Extend volumes
• Take snapshots (FlashCopy) and clones
• Volumes Images, Images Volumes
• Attach and Detach to/from VM instances
• Create and Delete containers in account
• Upload, Download and Delete objects
• List containers or objects in a container
• Display and update metadata
Keystone
• Access control

20
Version 4.21 versus 4.2.2 License Changes
• Native encryption 1
Secure Erase 1
AFM Disaster Recovery
Transparent Cloud Tiering
All of the same
features as
Advanced Edition 4
Unlimited Server,
Client and FPO
sockets (no longer
per-socket charge)
• Information Lifecycle Management (ILM)
• Active File Management (AFM)
• Hadoop Transparency Connector
• Clustered Protocol server 2
POSIX File System
OpenStack drivers
File Placement Optimization (FPO)
Local Read Only (LROC) and High Available Write (HAWC) Cache
Version 4.2.1
Express
edition
Version 4.2.2
Standard
edition
Version 4.2.1
Advanced
edition
Version 4.2.2
Data Management
edition
Socket-based license 3 for Server, Client and FPO Capacity-based
license
1 Not supported with Windows nodes. Metadata, including extended attributes, not encrypted.
2 Available on NSD servers for Linux-x86 and Linux on POWER
3 PVU-license applies to z System nodes
All nodes in a cluster must
use the same edition
4 Supported only on AIX,
Linux-x86 and Linux on
POWER

21
Software
standard x86, POWER,
and even z Systems
mainframes
Pre-built Systems
• Elastic Storage Server with
Erasure Coding
Cloud Services
deployed on any Cloud,
including IBM SoftLayer
Scale

22
Spectrum Scale – Pre-built Systems
5146-GLx models
GL2, GL4, GL6
60-drive 4U drawers
• SSD and
Nearline HDD
5146-GSx models
GS1, GS2, GS4, GS6
24-drive 2U drawers
• All SSD
• SSD and 10K HDD
IBM POWER8
servers
NSD Clients
Windows, Linux
and AIX servers
Twin-tailed
Elastic Storage
Server
“Building block”
TCP/IP or
RDMA
NSD Servers
Windows, Linux
and AIX servers
5146-GLxS models
GL2S, GL4S, GL6S
84-drive 5U drawers
• SSD and Nearline HDD

23
Reduce Costs and Increase Availability
with Spectrum Scale Native RAID
Dedicated
RAID Controllers
Flash / Disk Enclosures
RAID RAID
Spectrum Scale Native RAID distributes data, parity and spare
space uniformly across all drives of enclosure using either:
8+2P Distributed RAID-6
8+3P Distributed Erasure Coding
This distribution reduces the rebuild or disk failure
recovery process overhead compared to
conventional RAID
Critical rebuilds of failed full multi-terabyte
drives can be accomplished in minutes—
rather than hours or even days when using
traditional RAID technology
RAID
Flash / Disk Enclosures

24
Data Protection Schemes
Tolerate 1 drive failure Tolerate 2 drive failures Tolerate “M” failures
RAID-1 / RAID-10
K pieces 2 x K slices
RAID-5
K pieces K + 1 slices
2.0X
1.2X
3.0X
1.5X
1.3XTriplication
K pieces 3 x K slices
RAID-6
K pieces K + 2 slices
Erasure Coding
K pieces K+M = N
slices

25
Distributed Erasure Coding
Distributed RAID6 (8+2) and Erasure Coding (8+3)
No “idle” dedicated spare drives
– All drives contribute evenly to performance
– Especially important when using flash drives
More drives means faster rebuild
– No bottleneck rebuilding to dedicated spare
– Especially important when using large drives
P S
S
S
RAID4 with Dedicated Parity (4+P+S)
Updates require write to parity drive bottleneck
RAID5 with Interspersed Rotating Parity (4+P+S)
RAID6 with Interspersed Rotating Parity (3+P+Q+S)
Avoids bottleneck by spreading out parity updates
Idle dedicated spare drive is bottleneck during rebuild
) (
) (

26
3rd disk failure -
start of critical
rebuild
critical rebuild
finished, continue
normal rebuild
4 Minutes 16 seconds crit rebuild
Rebuild of a Critical Failure in minutes instead of hours and days!
norm rebuildbefore rebuild
1st disk failure
Norm
Rebuild
Distributed Erasure Coding – Array error recovery

27
Twin-tailed Twin-tailed
Aren’t Elastic Storage Servers just “Islands of Storage”? No!
Files translated
in blocks
…
NSD Client does real-time parallel I/O to
all the NSD servers and storage
volumes
All NSD servers export to all
the clients in active-active mode
Elastic Storage wide-stripes the files
across NSD servers in units of
file-system block-size
File-system load spread evenly
across all the servers -- No Hotspots!
Easy to scale file-system capacity
and performance while keeping
the architecture balanced
…
NSD Client
NSD Servers
Global Name Space …

28
IBM Spectrum Scale™ solves “Dropbox problem”
for Enterprise File Sync-and-Share (EFSS)
SAN
Internal,
Direct-Attach
No IT Control:
• Servers and storage
• Security
• Access control
• User provisioning
• Sensitive data
TCP/IP or RFMA Network
Twin-tailed

29
Share-Nothing versus Shared-Disk Deployments
Data
Data
Data Parity
Data
Data
Data
Copy
Copy
Copy
Copy
Copy
Copy
TCP/IP
or RDMA
Need more compute?
Add another node!
Spectrum Scale and Elastic Storage
Server reduce storage to one
RAID-protected copy of the data
Scale compute and storage
capacity separately
Spectrum Scale
can keep 1,2 or 3
replicas of the data
Need more
storage capacity?
Add another
node!
3x versus 1.3x
TCP/IP
or RDMA

30
Hadoop Analytics – HDFS vs IBM Spectrum Scale™
HDFS Save
Results
Discard
Rest
IBM
HDFS Transparency
Connector allows
HDFS-based programs to
process data without
application changes
(100% compatible)
IBM Spectrum Scale
Application data
stored on IBM
Spectrum Scale is
readily available for
analytics
Save
Results
JFS2
NTFS
EXT4
Data Sources
mashup of structured and
unstructured data from a variety
of sources
Actionable Insights
Provides answers to the
Who, What, Where,
When, Why and How
Business Intelligence
& Predictive Analytics
> Competitive Advantages
> New Threats and Fraud
> Changing Needs
and Forecasting
> And More!

31
Three ways to move cold data out of IBM Spectrum Scale
Migrate/Recall
LTFS
Tape
Management (ILM) moves data
across tiers of flash and disk
Migrate/Recall
Migrate/RecallTransparent
Cloud Tiering
Other NAS
AFM
IBM Cloud Object
Storage System
IBM Bluemix
Amazon S3

32
Positioning – IBM Spectrum Storage vs. IBM Cloud Object
Unified file and object
storage. Optimized for
high performance, across
flash, disk and object
store
Flash
Object
Store
15K
Object storage on disk
( File, backup and archive interfaces
available through variety of options )
IBM Bluemix
OpenStack Swift
Amazon Web Services S3
Swift S3 emulation
Unified file and object
storage on tape
Transparent Cloud Tiering
Management (ILM) across tiers
HighestPerformance
Lowest cost
Tape10K 7200

33
Software
standard x86 or
POWER servers
Pre-built Systems
• Elastic Storage Server with
Erasure Coding
Cloud Services
deployed on any Cloud
Scale

34
IBM Spectrum Scale on any Cloud
Active File
Management
Private VLAN
Scale
Scale
NSD Clients and Servers
can be deployed within a
Private VLAN on any Cloud

35
IBM Software Defined Storage
Data Plane
IBM Software Defined Storage
Control Plane
IBM for Software Defined Storage
Control Protect
IBM ranked #1 in
Software Defined Storage
Accelerate Virtualize Scale Archive
Universal access to data
• Global Namespace across dozens
of data center and remote locations
• POSIX, NAS, HDFS, Object APIs
Proven Reliability
• Introduced as GPFS in 1998
• Over 5,000 production systems
High performance
• Over 400 GB/sec throughput on
single system
Extreme scalability
• Clusters with over 10,000 nodes
• File systems with over 30 PB of data

36
Your feedback about this session is very important to us.
Submit a survey at:
ibmtechu.com

37

38
http://www.ibm.com/training/events
https://www-03.ibm.com/systems/storage/
http://www.redbooks.ibm.com/
https://www-03.ibm.com/systems/storage/spectrum/suite/
https://www-03.ibm.com/systems/storage/spectrum/scale/
Session reference links

39
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

40
About the Speaker
Tony Pearson is a Master Inventor and Senior IT Architect 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
solutions.
Tony writes the “Inside System Storage” blog, which is read by thousands 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 consulting 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 is an inventor or co-inventor of 19 patents in the field of electronic data
storage.
9000 S. Rita Road
Bldg 9032 Floor 1
Tucson, AZ 85744
+1 520-799-4309 (Office)
tpearson@us.ibm.com
Tony Pearson
Master Inventor
Senior IT Architect
IBM Storage

41
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
Additional Resources from Tony Pearson

42
Continue the conversation
view event
highlights
talk to tech
experts
connect with
attendees
read training
articles
IBM Systems
Technical Events
LinkedIn
community
Join today
bit.ly/IBMTechUconnect

43
Continue growing your IBM skills
ibm.com/training
provides a comprehensive
portfolio of skills and career
accelerators that are designed
to meet all your training needs.
If you can’t find the training that is right for you
with our Global Training Providers, we can help.
Contact IBM Training at dpmc@us.ibm.com

44
Trademarks and 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 2017. 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

S014066 scale-ess-orlando-v1705a

Recommended

Recommended

More Related Content

What's hot

What's hot (20)

Similar to S014066 scale-ess-orlando-v1705a

Similar to S014066 scale-ess-orlando-v1705a (20)

More from Tony Pearson

More from Tony Pearson (20)

Recently uploaded

Recently uploaded (20)

S014066 scale-ess-orlando-v1705a