IBM Duomenų saugyklų naujienos
XiV Generation3
Storwize V7000 unified
3-tier Easy Tier
IBM Sprendimų diena
Vilnius 2011m. spalio 18d.
© 2011 IBM Corporation
© 2011 IBM Corporation
Short history of disk drives…
System IBM 350 Disk Storage
Unit – 14 September 1956
5MB, 1200 rpm, service time <1 sec.
Total time required to random read all
data with IOsize=512 Bytes ?
5 MB / 512 B ≈ 10.000 IO
(less then 3 hours)
2
© 2011 IBM Corporation
© 2011 IBM Corporation
Modern disk drives
Seagate Cheetah ST3600057FC *)
– 4 plates, 15 Krpm
– Sector size: 512 Bytes
– Total capacity:
• 600 Giga bytes
– Srvice time ≈ 5,6 ms
*) http://www.seagate.com/docs/pdf/pl-PL/datasheet/disc/ds_cheetah_15k_7.pdf
3
© 2011 IBM Corporation
© 2011 IBM Corporation
Modern disk drives
Seagate Cheetah
ST3600057FC *)
Total time required to random read
all data with IOsize=512 Bytes ?
600 GB / 512 B ≈ 1.172 mln IO
(≈ 76 days)
*) http://www.seagate.com/docs/pdf/pl-PL/datasheet/disc/ds_cheetah_15k_7.pdf
4
© 2011 IBM Corporation
© 2011 IBM Corporation
The Drivers of Our Business
Zettabytes
Exabytes
Petabytes
 Information doubling every 18-24 months
 Backup and Archive requirements
growing at 40-50% per year
 Storage budgets up 1%-5% in 2011
Terabytes
The information explosion
meets budget reality
Gigabytes
Megabytes
Kilobytes
2000
Business analytics has
increased the value of
information
5
2005
2010
2015
The Cost of losing data
and searching for it has
skyrocketed
Users expect 100% availability of their applications and their information
© 2011 IBM Corporation
© 2011 IBM Corporation
Smarter Systems are Creating an Information Explosion
 247 Billion e-mails sent daily
 Text messages generate 400TB of data
per day in the US
 MRIs will generate a petabyte of data
in 2010
1980
66
1990
2000
2010
© 2011 IBM Corporation
© 2011 IBM Corporation
Technology and best practices enable storage efficiency
Stop Storing So Much
 IBM Real-time Compression … up to 80% less space
 IBM ProtecTIER … up to 95% smaller disk backups
Move Data to the Right Place
 IBM Easy Tier … 3X more performance with 2% SSD
Store More with What’s On the Floor
 IBM Storage Virtualization … up to 30% more utilization
 IBM Thin Provisioning … up to 35% more utilization
7
© 2011 IBM Corporation
© 2011 IBM Corporation
IBM Storage Portfolio as of 4Q2011
Entry/MidRange
Storage Systems
Enterprise
Storage Systems
• Storwize Family
• DS8000 Family
• V7000
• V7000 Unified
• DS Family
• DS8700
• DS8800
• XIV Family
File Storage
Systems
• Scale-Out NAS
(SONAS)
• N Series
• N7000
• DS5000
• XIV
• N6000
• DS3000
• XIV Gen3
• N3000
Data Protection
And Retention
• TS Family
• TS7000
• TS3000
• TS2000
• TS1000
• ProtecTIER
Deduplication
• Information
Archive
Storage
Management
Software
8
FlashCopy
Manager (FCM)
Tivoli Productivity
Center (TPC)
Tivoli Storage
Manager (TSM)
Efficiency
“Enhancers”
• Real-Time
Compression
• Easy Tier
• SAN Volume
Controller (SVC)
• Active Cloud
Engine
Tivoli Key Lifecycle
Manager
© 2011 IBM Corporation
© 2011 IBM Corporation
SAN Disk + NAS Portfolio 2010
XIV
SAN
DS3000
DS4700
Entry-level
DS5000
DS6800
V7000
XIV
DS8000
Enterprise
Midrange
NAS
Gateway
9
N3000
N6000
Entry-level
Midrange
N7000
Enterprise
SONAS
Ultrascalable
© 2011 IBM Corporation
© 2011 IBM Corporation
A Consistent Drumbeat of Storage Efficiency Innovation
2010
Storwize V7000
 The most advanced mid-range storage
 Built-in storage efficiency
Easy Tier
 The most powerful and easiest auto
tiering technology in the industry
Real-time Compression Appliance
 Reduce storage space requirements
without degrading performance
DS8800
 Up to 40% better performance
 Up to 36% less energy consumption
10
2011
XIV Gen3
 Up to 69% lower TCO
 Up to 4 times faster
Linear Tape File System
• 2011 Pick Hit award, Broadcast
Engineering
Real-time Compression Appliance
 Support for EMC storage, making even
our competitors’ storage more efficient
TS3500 Tape Library Connector
• Up to 2.7 Exabytes of low cost storage
© 2011 IBM Corporation
© 2011 IBM Corporation
October 2011 Storage Announcements
Announcement letters released: Oct 11, ibm.com live: October 12, press release Oct 19
 IBM® Storwize® V7000 Unified - Integrated support for file storage, higher capacity drives, and more remote
mirroring options (GA Nov 18)
 IBM System Storage® DS8800 - Easy TierTM enhancements, larger drives, and advanced IBM zEnterpriseTM
integration (GA Nov 18)
 IBM XIV® Storage System Gen3 - High capacity 3TB drives, iPad management app (GA Nov 17)
 IBM Scale Out Network Attach Storage (SONAS) R1.3 - Active Cloud EngineTM site-to-site replication, 3 TB
drives, new easy to use GUI (GA Oct 21)
 IBM SAN Volume Controller V6.3 - Stretched Cluster option extends distance up to 300km and Global Mirror adds
low bandwidth option (GA Nov 20)
 IBM Real-time Compression AppliancesTM STN6500 and STN6800 V3.8 - Enhanced operational management,
SMBv2 Windows protocol support (GA Oct 21, Dec 21 for SMBv2)
 IBM System Storage EXP2500 Express® - 900 GB drives, SSD, direct attach to BladeCenter® (GA Oct 28)
 IBM System Storage N Series – OnCommand 5.0, 600 GB drives with encryption (GA Nov 11)
 IBM Tivoli® Storage Manager v6.3 - Hot standby disaster recovery management, more capacity and automated
client agent updates (GA Oct 14)
 IBM Tivoli Storage FlashCopy® Manager v3.1 - Support for Oracle, SAP and IBM DB2® running on HP/UX, file
servers running Microsoft NTFS, and VMware (GA Oct 14)
 IBM Tivoli Storage Manager for Virtual Environments v6.3 - VMware vCenter integration and support for
hardware snapshots (Oct 21)
 IBM Tivoli Storage Productivity Center - Storage tiering reports. TPC Select adds site recovery management
(GA Oct 14)
11
© 2011 IBM Corporation
© 2011 IBM Corporation
Introducing XIV Gen3
Evolution of the Revolution
12
© 2011 IBM Corporation
© 2011 IBM Corporation
Architecture Review
Ground-breaking
Architecture
Large Processing Power (Modules)
Grid Architecture
Innovative and Powerful Caching
Other solutions…
Clustered Controllers
RAID protections
FC
XIV Module
Host
Host
Massive parallelism for IO
Aggressive, parallel pre-fetching
processing
Huge internal bandwidth
Variable
block-size
cache
Infiniband
interconnect
Simpler
cache management
High cache-to-disk
bandwidth
iSCSI
Long RAID rebuilds
Require LUN/Disk Layout
12x 2TB SAS Disk, 24GB Cache &
Autonomic
dataCore
layout
across grid
1Quad
CPU
XIV
Module
XIV
XIV
Module
XIV
Module
XIVModule
Module
Host
Performance tuning
Higher
cacheonly
hit-ratios
Module
services
its own disks
Consistent performance through
XIV
Module
XIV
XIV
Module
XIV
Module
XIVModule
Module
Internal
workload changes (peak or average)
Connectivity
Active/Active IO parallel access
Disk Hot-Spots
Zero tuning or manual intervention
Islands of storage
13
© 2011 IBM Corporation
© 2011 IBM Corporation
Architecture Review
Ground-breaking
Architecture
Scalable and sustained high performance
Unprecedented Resiliency
FC
Other solutions…
Manual intervention
Performance tuning
Minimum operations impact upon
disk failure (<1%) or module (<7%)
Host
iSCSI
Redesign after changes
Capacity-only scalability
XIV Module
Host
XIV
Module
XIV
XIV
Module
XIV
Module
XIVModule
Module
Host
XIV
Module
XIV
XIV
Module
XIV
Module
XIVModule
Module
Continual data movement
High impact after any failure
XIV
Module
XIV
XIV
Module
XIV
Module
XIVModule
Module
Automatic rebalancing even after a
system component failure or during
rebuild
Internal
Automatic data rebalancing after
Connectivity
new disk/module additions
XIV utilization always balanced
regardless of add/delete/resize LUNs
Linear scalability
on capacity & performance
14
© 2011 IBM Corporation
© 2011 IBM Corporation
Architecture Review: XIV vs Traditional Architecture Disk Utilization
We go
this…
Tofrom
this…
 The result?
Traditional Arrays IOPS profile: Unbalanced Disk L ayout
X IV IOPS / Balanced Disk L ayout
IOPS
IOPS
Disks
Disks
15
© 2011 IBM Corporation
© 2011 IBM Corporation
IBM XIV Gen3 Technology Highlights

20X more internal bandwidth
Using Infiniband

Over 2X more external bandwidth
With 8 Gbps FC ports and over 3x more iSCSI ports (6-22)

New motherboards and processors
2x disk bandwidth, 60 cores, 120 hyper-threads per rack.

50% more cache capacity
Up to 360GB/system, (24GB per module).

SSD ready
Optional cache upgrade of up to 7.5TB

16
3TB disks
50% capacity more, up to 240TB
© 2011 IBM Corporation
© 2011 IBM Corporation
Kicking Off a Second Century of IBM Storage Innovation
Gen3 – Evolution of the Revolution
Gen2 Profile
High performance
Lower cost/TB
Lower entry point
27 TB usable
Gen3 Profile
Ultra-High Performance
 Up to 4X increase
 SSD cache upgradeable
Higher entry point capacity
 55 TB usable
 20% less power consumption
 20% less heat output
 33% less noise reduction
17
© 2011 IBM Corporation
© 2011 IBM Corporation
Architecture Review: Traditional Storage Distribution
Exchange
Oracle
RAID 5
300 GB 15K
RAID 5
1 TB
RAID 5 300 GB
VMware
RAID 5 300 GB
RAID 5
146 GB
RAID 5
Hot
300 GB
Spares
RAID 1/0
146 GB
ERP
RAID 5
146 GB
RAID 5
146 GB
 Scalability is limited
and not linear
 Hot-Spots and
 Capacity is added but
Performance issues
performance is
require analysis,
reduced
design and tuning
 To improve
performance, a
redesign/relayout is
required
 Lots of work to keep
and maintain this
array
18
© 2011 IBM Corporation
© 2011 IBM Corporation
Architecture Review: XIV Storage Distribution
Ground-breaking
Architecture

Data is spread
across all drives

Even Disk
utilization

NO manual
intervention

No Tuning

No Hot-Spots
Oracle
VMware
ERP
19
© 2011 IBM Corporation
© 2011 IBM Corporation
Feature Summary
Powerful
Features
Enterprise-class, Robust Features
Unique snapshot technology
Simple, dynamic QoS
Powerful Synchronous and asynchronous mirroring
Streamlined data migration
Strong host software support
Native thin provisioning with thick-to-thin migration
Comprehensive management of events,
statistics and performance
Advanced LDAP-based authentication
Secure role-based access control
20
© 2011 IBM Corporation
© 2011 IBM Corporation
Radical Simplicity
Extreme
Ease-of-Use
 Breakthrough GUI
 Exceptional ease of use
 Powerful management capabilities
 Easy, rapid provisioning
 Minimal administration
 Minimal training required
21
© 2011 IBM Corporation
© 2011 IBM Corporation
Enterprise-Ready, Smarter by Design
Gen2
Gen3
Ethernet Interconnect
Infiniband Interconnect
Capacity: 24-161TB
Max Memory: 240GB
Switches
Capacity: 54-161TB
Max Memory: 360GB
Max FC ports: 24 x 4Gb/s
Max FC ports: 24 x 8Gb/s
Max iSCSI ports: 6 x 1Gb/s
Max iSCSI ports: 22 x 1Gb/s
iSCSI ports with 6 modules: 0
Interface Modules
Disk Type: SATA
SSD-ready (*)
iSCSI ports with 6 modules: 10
Disk Type: SAS, SSD (*)
Data Modules
UPS units
(*) Projected
availability: 1H12
22
© 2011 IBM Corporation
© 2011 IBM Corporation
Performance Review
Evolution of the Revolution
23
© 2011 IBM Corporation
© 2011 IBM Corporation
Gen3 Performance Summary (con’t)
Overall system improvements result in
– higher performance and bandwidth
– Lower latency
– Significant improvement in OLTP environments
 Up to 3x performance on read/write hit loads:
 Up to 4x performance on sequential loads:
 Up to 3x lower latency
>500,000 IOPS (read, hits)
>10 GB/s (read)
170 uSec single-io latency
 RAW single threaded writes, 4x increase
 RAW single threaded reads, 3x increase
24
© 2011 IBM Corporation
© 2011 IBM Corporation
IBM X IV Family Delivers Outstanding Performance Across Applications
Microsoft Exchange Mailboxes
Oracle Data Warehouse (IOPS)
 2 hour performance test
 Requires latency under 20 ms
 ESRP-Storage test
 Oracle DHW Workload
20,000
XIV
60,000
XIV Gen3
55,000
XIV
115,000
XIV Gen3
SAS Business Analytics Workload
Microsoft HyperV Simulation (IOPS)
 Analytics reports created
 200GB dataset
 60% write activity
Simulated via Swingbench load generator
70
XIV
207
XIV Gen3
13,605
XIV
37,856
XIV Gen3
Performance is based on measurements and projections using standard IBM benchmarks in a controlled environment. The actual throughput 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 improvements equivalent to the performance ratios stated here.
25
© 2011 IBM Corporation
© 2011 IBM Corporation
IBM XIV Family Delivers Outstanding Raw Performance
Transaction Processing (IOPS)
Sequential Reads (MB/sec)
 Mixed read / write workload
 System Bandwidth
33,000
XIV
66,000
XIV Gen3
3,034
XIV
File and Print Services (IOPS)
Sequential Writes (MB/sec)
 Mixed file size workload
 XIV Gen3 also had
50% lower latency
 System Bandwidth
42,000
XIV
125,000
XIV Gen3
1,542
XIV
10,300
XIV Gen3
6,788
XIV Gen3
Performance is based on measurements and projections using standard IBM benchmarks in a controlled environment. The actual throughput 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 improvements equivalent to the performance ratios stated here.
26
© 2011 IBM Corporation
© 2011 IBM Corporation
Transparent XIV SSD Caching *
 Upgrade option: SSD Caching
– SSD serves as a cache layer
between RAM and SAS drives
– Up to 7.5TB of fast cache
– Simple, efficient
– No separate disk tier to manage
 Designed to offer significant
performance improvement for all
usage profiles
System Module
RAM
SSD as
CACHE
Non-SSD
Drives
(*) Projected availability: 1H12
27
© 2011 IBM Corporation
© 2011 IBM Corporation
XIV Gen3 Enhancements
 50% more capacity – Integrates high capacity 3TB drives while
maintaining consistently high application performance
 Price Performance
– Exchange ESRP* results demonstrate 2 times better disk efficiency than
EMC, HP, or HDS
– New SPC-2 results (released later in October) will show exceptional
throughput and the best price performance
 IBM XIV Mobile Dashboard – iPad app that monitors XIV from wherever
you are - It doesn’t get any easier
* Microsoft Exchange Solution Reviewed Program (ESRP). http://technet.microsoft.com/en-us/exchange/ff182054
28
Performance is based on measurements and projections using benchmarks in a controlled environment. The actual throughput 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 improvements equivalent to the performance ratios stated here. Disk efficiency is (#mailboxes * IOPS per
© 2011 IBM Corporation
mailbox / number of disks) * mailbox capacity, using vendor postings as of October 1, 2011
© 2011 IBM Corporation
3 Tier EasyTier & DS8800
 Easy Tier automatically balances data across 3 tiers so there is no
tradeoff between cost and performance
 Higher scalability supports up to 45% more drives in only 33% more
floor space
 Large capacity 3 TB drives almost triples overall capacity and supports
data with moderate or varying performance requirements
 System z functional synergy simplifies volume management and
enhances DB2 performance
SSD Pools
29
Enterprise Pools
IBM Confidential until announcement
Nearline Pools
© 2011 IBM Corporation
© 2011 IBM Corporation
The IBM Storwize® V 7000 has
delivered unprecedented power,
flexibility, ease-of-use and value to
mid-range storage
30
© 2011 IBM Corporation
© 2011 IBM Corporation
Now, only one year later,
we’ re delivering the next
major step forward in the
evolution of IBM Storage
31
© 2011 IBM Corporation
© 2011 IBM Corporation
Storwize V7000 Significant Milestones – 3Q11
 Over 4,500 systems purchased world-wide since Nov 2010 GA
– Over 10,000 total enclosures deployed WW
– Over 100
PB of capacity
 Over 2,500 happy customers world-wide
– Presence in every major country and industry
– Delivering five nines data availability
– Over 1,000 net new storage customers
 Over 40 WIN flashes published
– Competitive wins across every geo
– Click here to find additional Storwize Win Flashes on w3

39 official references in customer database
- 10 added in September alone
 One of the fastest product ramps up in IBM storage history
32
© 2011 IBM Corporation
© 2011 IBM Corporation
One Year Later: Technology Integration Continues
Storwize V7000 Delivers Unified Storage
Leveraging the best of IBM storage technologies in a midrange disk system
Snapshots
Innovative GUI
Easy Tier
SAN
Virtualization
IBM Active
Cloud
Engine™
33
RAID
Clustered
Systems
File
Support
© 2011 IBM Corporation
© 2011 IBM Corporation
We’ re bringing to market a powerful,
flexible new midrange unified storage
platform.
One that is actually… Unified.
34
© 2011 IBM Corporation
© 2011 IBM Corporation
Storwize V7000
 New Storwize V7000 Unified
– Unified block and file (SAN and NAS) storage
– Upgradable from existing Storwize V7000 systems
– IBM Active Cloud Engine policy-based management
 Greater flexibility for remote mirror
– Replicate between Storwize V7000 and SVC
– Balance remote data currency with network bandwidth cost to
better meet application requirements
 Additional drive options
– 200GB and 400GB SFF SSDs
– 3TB LFF HDD
• 50% greater maximum system capacity
 Enhanced Tivoli Storage FlashCopy Manager
support
– VMware vSphere 4 and vSphere 5
– DB2, Oracle, and SAP on HP-UX
35
IBM Confidential until announcement
© 2011 IBM Corporation
© 2011 IBM Corporation
Greater Flexibility for Remote Mirror
Replication to Virtually Any Storage
Global Mirror or Metro Mirror
 Storwize V7000 remote mirror now
supports replication to/from IBM
SAN Volume Controller systems
 Enables use of any storage
supported by SVC as replication
target for Storwize V7000
– Over 130 different disk systems from
all major vendors
 Provides new options to redeploy
36 existing storage to extend life and
© 2011 IBM Corporation
© 2011 IBM Corporation
Greater Flexibility for Block Data Remote Mirror
Balance Remote Data Currency With Network Bandwidth
Cost
Remote Copy
Primary
Host
I/O
Background
Copy
Regular
FlashCopy
snapshot
Requires less link
bandwidth
Remote Copy
Secondary
Regular
FlashCopy
snapshot
Guarantees a
consistent copy
 Storwize V7000 remote replication delivers very high data
currency at remote site but requires network bandwidth to
support this currency
 New Global Mirror option makes regular copies of data at
primary location and transmits them to secondary location
 User can select frequency of these copies, which determines
network bandwidth required and data currency
37 Provides much more flexibility to meet specific application
© 2011 IBM Corporation
© 2011 IBM Corporation
Storwize V7000 Unified
What this is
 Unified storage system with tightly
integrated management for both block
and file capabilities
 Support for
NFS/CIFS/FTP/HTTPS/SCP file
protocols in addition to existing block
functions (FCP and iSCSI)
 File replication and file level snapshots
for business continuity and disaster
recovery in addition to existing block
functions
V7000 Unified
6U
V7000 Unified
Why It Matters
 Convergence of even more workloads into one Storwize V7000 Unified system for greater storage
efficiency
 Enables consolidation of multiple stand-alone file servers to reduce management effort
 Upgrade path for installed Storwize V7000 systems (planned availability 1H2012)
 Enables deployment of a broad set of applications from within a single storage system, especially in
shared storage environments
38
© 2011 IBM Corporation
© 2011 IBM Corporation
Storwize V7000 Unified Highlights
 Ordered in eConfig as a unified system
 Same control and expansion enclosures as
Storwize V7000
 Two Storwize V7000 File Modules (2073-700)
running Storwize V7000 File Module Software
(5639-VF1)
V7000 Unified
6U
V7000 Unified
 Minimum config 6U footprint comparable with EMC
and NetApp unified midrange offerings
Software inherited from prior offerings plus enhancements
Proven IBM software functionalities
 Storwize V7000 Block capabilities
Unified storage management
 Converged File/block GUI with Storwize V7000
ease-of-use
 Simplified installation based on Storwize USB key
 Coordinated RAS, event logs and error recovery
 ISV whitepapers for file workloads
39
 Storage Efficiency Advanced SW Functions
 Enterprise Block capabilities at a midrange price
 Grows as your needs grow
 File capabilities as part of IBM’s NAS family





NFS/CIFS/FTP/HTTPS/SCP
File replication and cloning, fileset level snapshots
NDMP and/or TSM Backup/Restore
ILM/HSM capabilities
Anti-virus support capabilities
© 2011 IBM Corporation
© 2011 IBM Corporation
What’s Different About Storwize V7000 Unified?
 Based on highly successful Storwize V7000 block
storage system
 Fully integrated user interface for block and file data
simplifies management
– Not a “unified” launcher for two different interfaces
 Simple USB key based initial setup: extends successful
Storwize V7000 approach
 IBM Active Cloud Engine™ technology delivers automated
storage efficiency capabilities
– True policy-based management of files
– With user-defined policies!
 Built-in IBM Tivoli Storage Manager client simplifies backup
to TSM server
– System also supports NDMP for third-party backup clients
40
© 2011 IBM Corporation
© 2011 IBM Corporation
Storwize V7000 Unified Key Points
 Storwize V7000 Unified is a midrange disk system
– Customer set-up and maintained: built on proven Storwize V7000
– Intuitive GUI with wizards: built on proven Storwize V7000
– Support call flow starts in the same manner as Storwize V7000
 Storwize V7000U is the first IBM-developed midrange unified storage
offering
– Presenting both file and block interfaces from the same system
 Storwize V7000U has a similar architecture to EMC Celerra and EMC
VNX
– Unified management, ordering, and support but separate file/block devices
– File “module” hardware/software in front of block storage system
– File “module” only supports one type of storage (Storwize V7000), not
heterogeneous gateway
– Virtualized external storage supported for file but only for NFS access (not
CIFS)
– Unified capability cannot be used with clustered systems
41
Storwize V7000U is a single orderable system with block and file
© 2011 IBM Corporation
© 2011 IBM Corporation
One system, one management
framework, one GUI, one unified
solution. Simple.
Because one > many.
42
© 2011 IBM Corporation
© 2011 IBM Corporation
Storwize V7000 Unified: Designed for Easy Setup
Proven USB Key Approach Simplifies Setup for Unified System
43
© 2011 IBM Corporation
© 2011 IBM Corporation
Storwize V7000 Unified User Interface
File management
integrated into
navigation
Storage pools may be
shared between block and
file workloads
File storage
management completely
integrated
44
© 2011 IBM Corporation
© 2011 IBM Corporation
Storwize V7000 Unified User Interface: System Status
Click on File
Modules to see
status
Storwize V7000
Unified system
monitoring display is
simple extension of
block display
45
© 2011 IBM Corporation
© 2011 IBM Corporation
Storwize V7000 Unified User Interface
Familiar right-click
for file system
options
Visual display of file
system size and
utilization
46
© 2011 IBM Corporation
© 2011 IBM Corporation
Storwize V7000 Unified User Interface
Switch easily
between file and
block performance
display
47
© 2011 IBM Corporation
© 2011 IBM Corporation
What is IBM Active Cloud
E ngine™ ? Think of it as a very
smart, very fast robot – that never
sleeps – keeping your cloud
storage neat, tidy, and running
smoothly…
48
© 2011 IBM Corporation
© 2011 IBM Corporation
IBM Active Cloud Engine™ for Storwize V7000 Unified
 IBM Active Cloud Engine
– Policy-driven engine that helps improve storage efficiency by automatically
• Placing files when they are created on the appropriate storage
• Moving files during their life to the right tier of storage including tape in a TSM
hierarchy
• Deleting expired or unwanted files
• Identifying files for backup or replication to a DR location
– High-performance: can scan large numbers of files very rapidly
 IBM Active Cloud Engine Value
– Lowers storage cost by moving files transparently to the most appropriate tier of
storage
– Controls storage growth by moving older files to tape and deleting unwanted or
expired files
– Enhances administrator productivity by automating file management
– Improves data protection by identifying files for backup or DR
 Statement of Direction
– IBM intends to enhance Active Cloud Engine on Storwize V7000 Unified with the remote
caching capabilities which are available starting with Active Cloud Engine in SONAS V1.3.0,
which will provide multi-site remote caching capability to Storwize V7000 Unified and between
SONAS and Storwize V7000 Unified
49
© 2011 IBM Corporation
© 2011 IBM Corporation
Understanding IBM Active Cloud Engine
Terminology
• A fileset is a logical subtree of the file
system namespace and provides a way to
partition the namespace into smaller, more
manageable units
• Multiple levels of rules (policies) can be
applied because the complete policy rule set
is evaluated for each file when the policy
engine executes
File System
Filesets
• Storage pools allow you to group multiple
physical storage devices within a single file
system (SAS, NL-SAS, external storage)
Policy Engine
Storage Pools
Pool 1
Pool 2
External
Pool Interface
File Data
External Storage
Management Application
TSM/HSM
50
© 2011 IBM Corporation
© 2011 IBM Corporation
IBM Active Cloud Engine Policies and Rules
 Active Cloud Engine supports two types of policies
 File placement policies
– Used to automatically place newly created files in a specific storage pool
– Files in same fileset may reside on different storage pools
 File management policies
– used to manage files during their lifecycle
– transparently move files to another internal or external storage pool
– changing their replication status
– Deletion
 Examples of policy rule uses
– Initial file placement
– File migration
– Backup file data
 Characteristics that cause a rule or action to be executed
– Last access time, last modification time, archive bit status
– Fileset name, file name, extension, or file size
 Policy rules evaluated in order, from first to last, as they appear in the policy
51
© 2011 IBM Corporation
© 2011 IBM Corporation
File Example Using Active Cloud Engine
File creation
30 days
EasyTier
SSD / SAS
180 days
60 days
NL-SAS
(deduplicate)
SATA
Externally Virtualized HP EVA
(For files accessed by NFS only)
Backup
catalog
Intelligent Backup offers
(deduplicate)
52
Better efficiency: Progressive incremental, dedu
Tighter control: File-level policy control
Faster recovery: Backup catalog
vs NDMP
© 2011 IBM Corporation
© 2011 IBM Corporation
Define Active Cloud Engine ILM Policies in the GUI
This dialog is
modeled on the
Storwize V7000
volume creation
dialog
When to run this
policy
Which files do not
get migrated under
this policy
53
© 2011 IBM Corporation
© 2011 IBM Corporation
Define Active Cloud Engine ILM Policies in the GUI
Display and manually edit policy rules
Advanced users can
display rules created
through the GUI
Users can edit policy rules directly
54
© 2011 IBM Corporation
© 2011 IBM Corporation
Improved Data Protection with SVC Enhanced Stretched Cluster
2-site Stretched Cluster
Server Cluster 1
SVC
Stretched-cluster
Failover
Improve availability, load-balance, and
deliver real-time remote data access
by distributing applications and their
data across multiple sites.
Seamless server / storage failover
when used in conjunction with server
or hypervisor clustering (such as
VMware or PowerVM)
Up to 300km between sites (3x EMC
VPLEX)
Server Cluster 2
Stretched
virtual volume
Up to
300km
Data center 2
Data center 1
4-site Disaster Recovery
Server Cluster 1
Failover
Server Cluster 2
Stretched
virtual volume
Data center 1
Server Cluster 1
Data center 2
Server Cluster 2
Stretched
virtual volume
Metro or Global Mirror
Data center 1
High Availability
Failover
For combined high availability and
disaster recovery needs,
synchronously or asynchronously
mirror data over long distances
between two high-availability stretch
clusters.
Data center 2
High Availability
Disaster Recovery
55
© 2011 IBM Corporation
© 2011 IBM Corporation
IBM Storage Portfolio as of 4Q2011
Entry/MidRange
Storage Systems
Enterprise
Storage Systems
• Storwize Family
• DS8000 Family
• V7000
• V7000 Unified
• DS Family
• DS8700
• DS8800
• XIV Family
File Storage
Systems
• Scale-Out NAS
(SONAS)
• N Series
• N7000
• DS5000
• XIV
• N6000
• DS3000
• XIV Gen3
• N3000
Data Protection
And Retention
• TS Family
• TS7000
• TS3000
• TS2000
• TS1000
• ProtecTIER
Deduplication
• Information
Archive
Storage
Management
Software
56
FlashCopy
Manager (FCM)
Tivoli Productivity
Center (TPC)
Tivoli Storage
Manager (TSM)
Efficiency
“Enhancers”
• Real-Time
Compression
• Easy Tier
• SAN Volume
Controller (SVC)
• Active Cloud
Engine
Tivoli Key Lifecycle
Manager
© 2011 IBM Corporation
© 2011 IBM Corporation
57
© 2011 IBM Corporation
© 2011 IBM Corporation