AMS2000 Technical Presentation - Mar11!01!057522

Hitachi Adaptable Modular
Storage 2000 Family

Technical Presentation
20.07.2011
Rafael Noas
Technical Consultant
2008 Hitachi Data Systems
Agenda
Hitachi Data Systems Portfolio
Hitachi Adaptable Modular Storage 2000 family Product Overview
Key Features
Symmetric active/active controller with dynamic load balancing

SAS architecture
Online RAID group expansion
LUN grow/shrink
High Density Expansion Trays
RAID-6
Cache Partitioning
Multiprotocol: FC SAN/iSCSI
Volume Migration Modular software
Hitachi TrueCopy Extended Distance software
GREEN Storage initiative
Improved Security
Hitachi Adaptable Modular Storage 2000 Family

In the Spotlight
The Hitachi Adaptable Modular System

2500 has won the Eco-Responsibility
Award in this years Information Age
Innovation Awards, beating out stiff
competition.
Gartner positioned Hitachi & the Hitachi

AMS2000 family in its Magic Quadrantthis
quadrant reflects highest scores for their ability
to execute and completeness of visionthey
are innovatorswith a clear understanding of
market needs Nov 17, 2008
Hitachi achieves overall best-in-class Storage Performance Council (SPC-1)

benchmark results for its midrange storage system (Hitachi AMS2000). March 24, 2009
Hitachi Data Systems has consolidated

its claim as the top-rated high-end
array vendor by repeating its win...Rich Castagna, Editor
3
The AMS2000 series combination

of enterprise class features
with easy to manage midrange
usability and reduced operational
costs is powerful and worthy of
serious consideration by an IT
organization.
Performance
Hitachi Storage Lineup
Full redundancy for 99.999% availability

Data-in-place upgrades to new controllers
High Performance native SAS backend
Built-in encryption
Full Hyper-V & VMware integration
Common Management
Price/Performance Leaders
AMS2500
Scale Performance &

Capacity Independently!
Virtual Storage Platform
AMS2300
1TB Cache
192 x FC
2048 disks
400,000 IOPS
AMS2100
8GB Cache
4 x FC / iSCSI
120 disks
30,000 IOPS
32GB Cache
16 x FC / iSCSI mix
480 disks
90,000 IOPS
16GB Cache
8 x FC / iSCSI
240 disks
50,000 IOPS
* 60/40 workload with <10% cache hit
Capacity
Adaptable Modular Storage 2000 Family
Throughput
Enterprise design
Host multi-pathing
Hardware load balancing
Dynamic Optimized Performance
Dense storage, SAS and SATA
99.999+% reliability
Adaptable Modular
Storage 2100
4-8GB Cache
4 FC, or 8 FC, or 4 FC and
4 iSCSI ports
Mix Up to 159 SSD, SAS
and SATA-II Disks
Up to 313 TBs
Up to 2048 LUNs
Up to 1024 Hosts
Adaptable Modular
Storage 2300
8-16GB Cache
8 FC, or 16 FC, or 8 FC and 4
iSCSI ports
Mix Up to 240 SSD, SAS and
SATA-II Disks
Up to 472 TBs
Up to 4096 LUNs
Up to 2048 Hosts
Scalability
5
Adaptable Modular
Storage 2500
16-32GB Cache
16 FC, or 8 iSCSI, or 8 FC and
4 iSCSI ports
Mix Up to 480 SSD, SAS and
SATA-II Disks
Up to 945 TBs
Up to 4096 LUNs
Up to 2048 Hosts
Hitachi Adaptable Modular Storage 2100

Hitachi Dynamic Load Balancing controller
Symmetrical active/active design
Host port options:
4 x 8Gb/sec FC
8 x 8Gb/sec FC
4 x 8Gb/sec FC and 4 x 1Gb/sec iSCSI
Cache size: 4GB or 8GB
LUNs maximum number: 2048
Maximum attached hosts through virtual ports: 1024
3Gb/sec SAS links: 16
Hitachi Adaptable
Modular Storage 2100
SATA-II and SAS intermix

Max drives supported: 159 (7200 RPM) 136 (15K RPM)
Drive options:
SAS:
SATA:
SSD:
300GB 15K RPM
1TB 7200 RPM
200GB
450GB 15K RPM
2TB 7200 RPM
600GB 15K RPM
2TB 7200 RPM
Protection
RAID Groups/System: 50
RAID levels: -6, -5, -1+0, -1, -0*
6
* RAID-0 available on SAS drives only
3U trays (max of 7)
2 to 15 disk drives/tray
4U trays (max of 3)
4U controller
4 to 15 disk drives

Hitachi Dynamic Load Balancing Controller
Host port options:
8 x 8Gb/sec FC
16 x 8Gb/sec FC
LUNs maximum number:4096
Hitachi Adaptable

Min/Max drives supported: 4/240
Drive options:
SAS:
SATA:
300GB 15K RPM
1TB 7200 RPM
450GB 15K RPM
2TB 7200 RPM
600GB 15K RPM
2TB 7200 RPM
Protection
RAID levels: -6, -5, -1+0, -1, -0*
7
SSD:
200GB
3U trays (max of 15)

4U trays (max of 4)
4U controller
4 to 15 disk drives

Host port options:
16 x 8Gb/sec FC
8 x 1Gb/sec iSCSI
LUNs maximum number:4096
Hitachi Adaptable

Min/Max Drives Supported: 4/480
Drive options:
SAS:
SATA:
300GB 15K RPM
1TB 7200 RPM
450GB 15K RPM
2TB 7200 RPM
600GB 15K RPM
2TB 7200 RPM
Protection
RAID levels: -6, -5, -1+0, -1, -0*
8
SSD:
200GB

4 to 15 drives/first tray
2 to 15 drives/ additional trays
4U controller
0 disk drives
Hitachi Adaptable Modular Storage 2100 Architecture
Hitachi Adaptable Modular Storage 2300 Architecture:
10
Hitachi Adaptable Modular Storage 2500 Architecture
11
Hitachi Adaptable Modular Storage 2000 Family

Host Port Symmetric Active-Active Feature
The Active-Active Symmetric system allows for a host I/O present on any front-end port to be processed by either
controller. All LUNs are associated with one of the two controllers by the LUN Management Table. If a host request for a
LUN arrives on a port on the current managing controller, then that is called Direct Mode. If that request is for a LUN
managed by the other controller, that is known as Cross Mode. In Cross Mode, the local Intel processor directly sends
the request to the Intel CPU on the other controller for execution over the inter-DCTL communications bus.
12
Hitachi Adaptable Modular Storage 2000 Family Disk Enclosure

8 Links for Maximum Throughput and Reliability
Controller 1
SAS CTL
Expander
15 HDDs
Controller 0
Internal
enclosure
SAS CTL
Expander
2 SAS Wide Cables

8 links x 3Gbit/s
15 HDDs
15 HDDs
15 HDDs
RKAK (SAS and SATA-II Enclosure)

Controller-1
Expander
Enclosure
15 disks
4-SAS
links
4-SAS
links
SAS
HDD
Slot 0
SAS
HDD
Slot 1
SAS
HDD
Slot 2
SATA
HDD
Slot 3
SAS
HDD
Slot 4
4-SAS
links
13
SATA
HDD
Slot 13
SAS
HDD
Slot 14
4-SAS
links
Expander
Controller-0
SATA
HDD
Slot 5
15 HDDs
240 Disks
1 + 15 enclosures
(SAS and SATA)
0
2 SAS Wide Cables
8 links x 3Gbit/s
15 HDDs
15 HDDs
SAS Engine Overview

The SAS Engine is the set of IOC chips,
SXP chips, and SMX chips. AMS 2100 and
2300 models have 1 SAS Engine per
controller; the AMS 2500 has 2.
The DCTL processor directs disk I/O
requests to the SAS IOP. The IOC will
generate either SAS or SATA-II commands
depending on the target disk.
The SAS Multiplexor chip (SMX) is the
interface from the IOC to the external cables.
It manages 4 SAS links in a single Wide SAS
Cable.
On the AMS 2100/2300 controller, there is
an internal 24-port SAS Expander unit
(SXP). This controls one port from each of
the 15 internal disks. This is the same SXP
used in the external enclosures.
14
External Disk Enclosure 15 Disk
This is the 15-disk shelf which may be freely populated with SSD, SAS and SATA-II disks.
The two SAS Expanders in each enclosure attach each disk port to any of the four SAS links
in that Expander. The SATA-II canisters provide the dual-port logic. Additional enclosures are
daisy-chained from the outbound SAS links port.
15
External Disk Enclosure High Density
This is a representation of the high density shelf for use with 38 SAS disks or 48 SATA disks
(no intermix). The two pairs of SAS Expanders in each enclosure attach each disk port to any of
the four SAS links in that Expander. The SATA-II canisters provide the dual-port logic native to
SAS disks. Additional enclosures are daisy-chained from the outbound SAS links port.
16
Dense Expansion Disk Tray (4U)

Up to 48 high capacity disks
2TB SAS 7200 RPM
2TB SATA 7200 RPM
Up to 38 high performance disks

600GB 15K RPM
450GB 15K RPM
Nearly doubles the density of high

performance SAS storage when
compared to standard disk trays
Intermix high density SAS and
SATA storage in same system
Reduce data center floor space

consumption
17
Dense Expansion Tray (2100 and 2300)

Max System Configurations
High
Density
2100 / 2300
System Tray
Standard
Tray
Tray
SAS / SATA
HDDs per Tray
15
15
38 / 48
Height per Tray
4U
3U
4U
Max w/ only High Density Trays
Max w/ Intermix
Max w/ Intermix
Max w/ only Standard Trays
Max SAS
HDDs *
Max SATA
HDDs *
129
159
136
156
128
138
120
120
167
207
Max w/ Intermix
204
234
Max w/ Intermix
196
216
Max w/ Intermix
11
218
228
15
240
240
AMS2100
AMS2300
* SAS and SATA dense trays can be intermixed in the same system. Max HDD count will depend on the number
of SAS and SATA trays.
18
Dense Expansion Tray (2500)

Max System Configurations
High
Density
HDDS per Tray

Height per Tray
AMS2500
Max w/ Intermix
Max w/ Intermix
Max w/ Intermix
Max w/ Intermix
Max w/ Intermix
Max w/ Intermix
Max w/ Intermix
Max w/ Intermix
Max w/ Intermix
2500
System Tray
Standard
Tray
Tray
SAS / SATA
0
4U
15
3U
38 / 48
4U
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
0
2
6
8
12
14
16
20
24
28
32
10
9
8
7
6
5
4
3
2
1
0
Max SAS
HDDs *
Max SATA
HDDs *
380
372
394
386
408
400
392
414
436
458
480
480
462
474
456
468
450
432
444
456
468
480
* SAS and SATA dense trays can be intermixed in the same system. Max HDD count will depend on the number
of SAS and SATA trays.
19
Dense Expansion Tray

FRU Placement
Top View
Rear View
PS X 4
ENC Board X 4
HDD X 48 (Max)
SAS Cable Holder X 8

Enlarged
Left Side: Input

Right Side: Output
ENC cable
Cable Holder
Lever
20
SAVE Space and Power with smaller Drives

and trays
Cost Effective and scalable
Save up to 60% more space, among the best density in the Industry
High Density + Performance (7.2TB per U rack)
2U trays support 24HDD 2.5 SAS
AMS2100 up to 6, AMS2300 up to 9, AMS2500 up to 20 trays
Sustainable
Save up to half power and cooling compare with 3.5 HDD
300GB and 600GB 2.5 SFF drives
LED:
POWER()
Unit
LED:
Power
(Green)
Unit
LED:
READY
(Green)
LED:
READY()
LED:
LOCATE()
Unit
LED:
LOCATE
(Yellow)
21
2.5 Canister
2.5
Drive
LED:
ACTIVE (Green)
LED:ACTIVE()
Drive LED:
ALARM (Red)
LED:ALARM()
Hardware image for the 2.5 Expansion Unit
LED:
POWER()
Unit
LED:
Power
(Green)
2.5 Canister
2.5
Drive
LED:
ACTIVE (Green)
LED:ACTIVE()
Unit LED: READY (Green)
LED: READY()
Drive LED:
ALARM (Red)
LED:ALARM()
Unit
LED:
LOCATE
(Yellow)
LED:
LOCATE()
ENC LED (POWER, LOCATE, ALARM)
ENC LED ()POWER, LOCATE, ALARM
LINK
(IN) )
LINKLED
LED(IN
Mini
SAS Connector
LINK)
MiniSAS
(UP (UP
LINK)
ENC#0
ENC#1
LINKLED
LED(OUT
LINK
(OUT))
MiniSAS
(DOWN
LINK)LINK)
Mini
SAS Connector
(DOWN
POWER
LOCATE
ALARM
IN
OUT
CONSOLE
IN
POWER
LOCATE
ALARM
CONSOLE
RDY
AC_IN
ALM
PS#0
OUT
RDY
AC_IN
ALM
PS#1
Power
point
22
Comparison between 2.5 Expansion Unit

and DF800 Expansion Unit
#
Item
2.5 Expansion Unit
DF800 Expansion Unit (RKAK)
Unit Size
2U
3U
Connectable Basic Unit
DF800/DF800E
Same as left
# of HDU
24(SAS: 2.5-10Krpm)
15(SAS/SATA/SSD/NL-SAS: 3.5)
# of ENC
Same as left
# of PSU
Same as left
# of Fan(self-powered)
12 units(6units/PSU)
Basic specification
2
3
Module specification
Back-end composition
Path
Input1path (4WideLink)
Output1path (4WideLink)
Same as left
Use device
Expander
PM8005(6G_36Port version)
SXP-24x3G
Flash
8Mbyte
2Mbyte
10
SRAM
Unused (SRAM is mounted)
4Mbyte x 2
11
SEEPROM
32KByte(I2C)
2KByte
12
Power Supply
Power type
12v/5v/3.3v/1.2v/1.0v
Same as left
13
Mini-SAS cable
Path
IN :Key slot3 & key4,6

OUT:Key slot3 & key2,4
IN:key4,6
OUT:key2,4
SAS cable length
1m, 3m,5m
Same as left
Debug
Connector for serial(RJ-45)
Same as left
14
15
23
Other
Hitachi Adaptable Modular Storage 2000 family - Field

Controller Upgrade Process Easy Upgrades
Upgrade by changing two controller

cards only. Data undisturbed
AMS 2100
AMS 2300
AMS 2500
S CTLR1
M CTLR1
H CTRL1
PS1
PS2
PS1
4U
S CTLR2
PS2
4U
Monterey M
ADD shelf
24
4U
H CTRL2
Note : BATTERY is at the front side in

each model of 2100
- Usable w/S/M/H
- 3 GB/sec capable
- ENC different from
upgrade ENC
PS2
M CTLR2
Monterey S
HDD Chassis
PS1
15 x 3.5 inch SAS/ISATA-II HDDs

ENC 1
ENC 2
PS1
PS2
ENC 1
ENC 2
PS1
PS2
3U

Family Battery Backup Times
Model
Batteries
Small Cache 2GB DIMMS
Large Cache 4GB DIMMS
Standard Configurations
AMS2100
2 Internal
72 hours
48 hours
AMS2300
2 Internal
36 hours
24 hours
AMS2500
4 Internal
48 hours
24 hours
Optional Configurations
25
AMS2500
1 External + 4
Internal
96 hours
48 hours
AMS2500
2 External + 4
Internal
168 hours
96 hours
Hardware Details
Adaptable Modular
Storage 2100
Adaptable Modular
Storage 2300
Adaptable Modular
Storage 2500
Symmetric active/active
controllers
4 or 8 GB cache
1.67GHz Celeron, single core
processor
PCIe internal bus
15 internal disk slots
Dual redundant power
supplies
Dual batteries
controllers
8 or 16 GB cache
1.67GHz Xeon, single core
processor
PCIe internal bus
supplies
Dual batteries
controllers
16 or 32 GB cache
2GHz Xeon, dual core
processor
PCIe internal bus
supplies
Dual batteries
4 x 8Gbps Fibre Channel

(FC)
8 x 8Gbps FC
4 x 8Gbps FC + 4 x 1Gbps
iSCSI

(FC)
16 x 8Gbps FC
iSCSI

(FC)
8 x 1Gbps iSCSI
iSCSI
Max attached host

ports through
virtual ports
1024
2048
2048
Data in place
upgrades
To Adaptable Modular Storage

2300 or 2500
To Adaptable Modular Storage

2500
N/A
Models
(3 models)
Host Interface
Options
(Dedicated FC or iSCSI)
26
Hardware Details
Adaptable Modular
Storage 2100
Adaptable Modular
Storage 2300
Adaptable Modular
Storage 2500
Maximum cache
partitions
16
32
32
Drive Interface
16 Serial Attached SCSI (SAS)

wide links
16 Serial Attached SCSI (SAS)

wide links
32 Serial Attached SCSI

(SAS) wide links
RAID Levels
RAID-1, RAID-1+0, RAID-5, RAID-6 and RAID-0 (SAS drives only)
Max # of RAID
Groups
50
75
100
Max # of LUs
2048
4096
4096
Max LU size
60TB
60TB
60TB
27
Hardware Details
Adaptable Modular
Storage 2100
Adaptable Modular
Storage 2300
Adaptable Modular
Storage 2500
Supported Drives
SAS: 300GB/15k, 450GB/15k, 600GB/15k, 2TB/7200

SATA-II: 1TB/7200, 2TB/7200
SSD: 200GB (SAS interface)
Trays
Controller tray: 4 (min) to 15

(max) HDDs
Expansion tray ( up to 7 ): 2
(min) to 15 (max) HDDs
Dense tray ( up to 3 ): 4 (min)
to 48 (max) HDDs
Max of 159 total HDDs
Controller tray: 4 (min) to 15

(max) HDDs
Expansion trays ( up to 15 ): 2
Dense tray ( up to 4 ): 4 (min)
to 48 (max) HDDs
Controller tray: 0 HDD

First expansion tray: 4 (min)
to 15 (max) HDDs
Up to 31 additional
expansion trays, each with 2
Dense tray ( up to 10 ): 4
Maximum Capacity
313TB
472TB
944TB
Supported
Operating Systems
Microsoft Windows 2000, Windows Server 2003, Windows 2008 ( including Hyper V )
Sun Solaris ( Sparc and x64 )
VMware
RedHat Enterprise Linux
SuSE Linux
RedFlag Linux
Oracle Enterprise Linux
Asianux
MiracleLinux
HP-UX
IBM AIX
MAC OSX
Novell NetWare
HP Tru64 UNIX
HP OpenVMS
28
Full Software and Firmware Offering

Standard Firmware Features
Storage Management Software

Storage Navigator Modular 2
Bundled Storage Functions
Account Authentication
Audit Logging
Dynamic Provisioning
LUN Manager
LUN Grow/LUN Shrink
Online RAID Group Expansion
Cache Residency Manager
System management via GUI and CLI

Capabilities
Provides access control to management functions
Records all system changes
Enables creation of logical, virtual volumes

Manages LUN configuration operations
Dynamically adds or reduces LUN capacity
Dynamically adds HDDs to a RAID group
Loads data for selected LUN into cache
Cache Partition Manager
Customizes cache utilization for applications
Modular Volume Migration
Moves data between RAID groups
SNMP Agent Support Function
Performance Monitor
29
Capabilities
Reports failures and status to a SNMP server
Monitors and reports performance data

Optional Firmware Features
Optional Storage Features
ShadowImage (Clone)
Copy-on-Write (Snapshot)
1 Primary:8 Secondary, 2047 Max , GUI and CLI mgmt

1 Primary:32 Snaps, 2047 Max , GUI and CLI mgmt
TrueCopy (Sync Remote Mirroring)
1 Primary:1 Secondary, GUI and CLI mgmt
TrueCopy Extended Distance

(Asynchronous Remote Mirroring)
1 Primary:1 Secondary, GUI and CLI mgmt
Data Retention Utility (DRU)
Protects LUNs from I/O activity
Device Manager
Configuration GUI and device management for open

systems
Tuning Manager
Performance troubleshooting; deep storage system tuning
Storage Capacity Reporter

Power Savings Service (Spin Down)
30
Capabilities
End-to-end storage capacity reporting

Spins down unused disk drives

Optional Firmware Features
Optional Storage Features
Replication Manager
Simplifies configuration and management of replication

services
Protection Manager
Performs rapid backup and recovery management
Storage Services Manager
Standards-based platform for heterogeneous storage

resource management and active SAN management
Dynamic Link Manger
Path failover and load balancing capabilities
Global Link Manager
Aggregates the management of Hitachi Dynamic Link

Manager instances into a single management station
Virtual Tape Library

Data Protection Suite
Content Archive Platform
31
Capabilities
Provides the benefits of backing up to disk without

changing established backup policies or procedures
Archiving, data migration services and backup/recovery
Digital content archiving

Architecture - Simplified Installation
Quick and Easy setting at installation:
1. No need to set controller ownership for each LU
2. Set host connection port without regards to controller ownership.
Benefits
Traditional
Adaptable Modular Storage 2000 family
Paths set to match

controller ownership
of LUs
Connect HBAs
to any controller
User required
to set paths
CTL0
LUN: 1
LUN: 2
LUN: 3
CTL1
LUs created
LUs created
LUN: 0
32
User not
required to
set paths
Owning controller of LUN 0
CTL0
LU LUN: 1
LUN: 3
LUN: 0
LUN: 2
CTL1

Architecture - Coordination with path management software
When used with front end load balancing function of path management software,
Hitachi Adaptable Modular Storage 2000 family systems can respond with balanced
performance over multiple data paths.
Benefits
Traditional

Path
Path
Manager Manager
HBA HBA HBA HBA
Path
Path
Manager Manager
HBA HBA
Standard setting
Load balancing
HBA HBA
Path
Path
Manager Manager
HBA HBA
Path
Path
Manager Manager
HBA HBA HBA HBA
HBA HBA
Primary path
Alternate path
User required
to set paths
CTL0
LUN: 0
33
LUN: 1
LUN: 2
CTL1
LUN: 3
User not
required to
set paths
CTL0
LUN: 0
LUN: 1
LUN: 2
LUN: 3
CTL1
Back End Load Balancing

Automatic Internal Load Balancing
Optimal performance is achieved with minimal input from storage administrator.
Load balancing occurs automatically and evens out the utilization rates of
both controllers.
Benefits
Traditional Need to manually re-configure

Load balancing is executed automatically
(5) Change path
(2) Analysis
(3) Planning
Automatically
Changes
(4)Change
setting
(1) Send
information
Performance
monitor
CTL0
CPU usage
70%
LUN: 0
34
LUN: 1
LUN: 2
Load balance
function
CTL1
CPU usage
10%
LUN: 3
CTL0
CTL1
CPU usage:
70% 40%
CPU usage
10% 40%
LUN: 0
LUN: 1
LUN: 2
LUN: 3
Unique
for Midrange
Symmetric
Active- Active
Controllers:
During normal operation
I/O can travel down any
available path to any
front end port using the
native host OS
multipathing capabilities.
This allows I/O to be
balanced over all
available paths without
impacting performance.
Automatic Load
Controller
Load Balancing:
Balancing
Controller :
If the load on the
controllers becomes
unbalanced it will
automatically be
balanced without the
host having to adjust its
its primary
primary
path.
path.
35
SWITCH 1
SWITCH 2
FE CTRL-0
AMS 2500
FE CTRL-1
BE CTRL-0
High Speed
Buses
BE CTRL-1
I/O is redirected to less busy controller
Symmetric ActiveActive and Load

Balancing Controllers:
A Powerful
combination!
With features
traditionally found only in
Enterprise class storage
arrays, the 2000 Family
can dynamically balance
I/O over any available
path through any front
end port on either
controller. All this is
done without having to
load any propriety path
management software.
Point-to-Point SAS Back end architecture

Up to 32 high performance
3GB/sec point-to-point
links for 9600MB/sec
system bandwidth
Each link is available for
concurrent I/O
No loop arbitration
bottlenecks
8 SAS links to every disk
tray for redundancy
Common disk tray for SAS
and SATA drives
Any failed component is
identified by their unique
address
36
SAS Delivers Scalable Performance for OLTP Workloads
100% Random Read, 146GB 15k rpm

Internal Load Balancing Only
8KB Block, 8 Threads per LUN
Transactions / sec
60,000
50,000
40,000
30,000
20,000
10,000
0
Disks In Use
2+2
37
4+4
4+1
2009 Hitachi Data Systems Confidential For Internal Use Only
8+1
SPC-1 Benchmark Results

for Midrange Storage Systems
Hitachi Data Systems
IBM
NetApp
Model
AMS 2100
AMS 2300
AMS 2500
DS5020
DS5300
V7000
FAS3170
Controller Type
Symmetric
Symmetric
Symmetric
Asymmetric
Asymmetric
Asymmetric
Asymmetric
SPC-1 IOPS
31,498.58
42,502.61
89,491.81
26,090.03
62,243.63
56,210.85
60,515.34
Price/Performance
$5.85/IOP
$6.96/IOP
$6.71/IOP
$7.49/IOP
$11.76/IOP
$7.24/IOP
$10.01/IOP
SPC-1 Sustainability
Rate (Throughput
MB/s)
258MB/s
350MB/s
735MB/s
214MB/s
511MB/s
463MB/s
497MB/s
Average Response
Time (50% Load
Point)
2.67ms
2.70ms
2.88ms
3.91ms
2.95ms
4.19ms
4.16ms
Average Response
Time (100% Load
Point)
8.15ms
6.33ms
8.98ms
10.49ms
14.37ms
10.8ms
20.8ms
Leading values shown in RED
38
Simplified Backend Cabling

SAS hard drive connection (system backend)
AMS 2100/2300
CTL0
CTL1
Simple and Easy backend cabling.
Single keyed cable type for all
connections
RKA
RKA
SAS
SAS
SATA
Common HDD Tray for SAS and

SATA drives
RKA
SATA
SAS
RKA
SATA
All SAS Backend paths active

RKA
SATA
RKA
SAS
39
SATA
RKA
SATA
SAS
RKA
SAS
LUN grow/LUN shrink

User adds a new Sub LUN from the unused capacity in an existing
RAID group
Any of the existing LUNs and the sub LUN are unified
LUN grow is an online process
LU0
LU0
LU0
LU1
LU1
Not used
Add LU
Not used
RAID
Group
40
LU0
Sub LU
LU0
Unified
Logical
LU0
LU1
LU1
Sub LU
Not used
Sub
Not used
Physical
LUN grow/LUN shrink

Users can release space from an existing LUN
The released space is no longer provisioned
Released space can be used for new LUNs or to expand an existing LUN
LU0
LU1
LU0
LU0
Not used
LU1
LU1
LU0
LU2
LU1
41
Online RAID Group Expansion

Use with LUN grow to increase LUN and RAID group capacity.
Tunable to allow preference to host I/O or to Expansion. Powerful
controller minimizes disruptions during RAID group expansion
3D+2P
HDD0
HDD1
HDD2
HDD3
HDD4
HDD5
6D+2P
HDD0
42
HDD1
HDD2
HDD3
HDD4
HDD5
HDD6
HDD7
HDD6
HDD7
Online RAID Group Expansion Specifications
LUNs are automatically restriped across expanded RAID groups
LU0
LU1
LU2
LU3
HDD0
HDD1
HDD2
HDD3
HDD4
HDD5
HDD6
HDD7
LU0
LU1
LU2
LU3
HDD0
43
HDD1
HDD2
HDD3
HDD4
HDD5
HDD6
HDD7
Mega-LUNs
Simple to create
Create a LUN, pick your size
LUN can grow up to 60 TB
Can span every disk in the
system
Simple to connect
Select a LUN
Choose a host
No need to choose a preferred
path
No need to choose an owner
controller
Simple to manage
Expand LUNs nondisruptively
Expand Raid Groups nondisruptively
Works with all Hitachi Adaptable
Modular storage software
LUN A
LUN A
LUN A
LUN A
LUN A
LUN A
44
Dynamic Provisioning = Efficient Storage Allocation

Use only what you need where you need it when its needed
AMS 2000
HDP Virtual Volumes
Servers
Hitachi Dynamic Provisioning

Pool Volumes
45
Challenges
High cost of storage
Cumbersome provisioning
Expensive optimization
Solution Capabilities
Simplify provisioning
Provision only what is used
Automates performance optimization
Replication Savings
Business Benefits
Reduced storage expense
Reduced operational expense
IT Agility

Capability: Automates performance optimization
Dynamic Provisioning software
effectively combines many applications
I/O patterns and evenly spreads the I/O
activity across available physical
resources
This automatically levels workloads and
avoids particular parity groups
becoming performance bottlenecks
Before Dynamic Provisioning, this had
to be done manually by a storage expert
Servers
HDP Volume
(Virtual LUN)
HDP Pool
RAID
Groups
RAID Group
RAID Group
RAID Group
Physical Drives
46

Reduces Replication Resource Requirements and Overhead
Using ShadowImage to copy Virtual LU-to-Virtual LU requires replicating only the mapped
physical storage which can be significantly less capacity than with a Normal Volume
Virtual LU
ShadowImage
PVOL
Maximum
Savings
Virtual LU
ShadowImage
SVOL
ShadowImage PAIR
The Storage Manager still has the flexibility to using ShadowImage for copying between
Virtual-LU and Normal Volume as needed
Virtual LU
ShadowImage
PVOL
Normal
VOL
ShadowImage
SVOL
Normal
VOL
ShadowImage
PVOL
ShadowImage PAIR
Virtual LU
ShadowImage
SVOL
ShadowImage PAIR
Note: ShadowImage is used in conjunction with the Virtual Volume. Mirroring of the underlying DP Pool with ShadowImage is not supported.
47
Get More From Storage Assets with Zero Page Reclaim

Facility
Hitachi Dynamic Provisioning for the AMS2000 supports the capability to do Zero Page Reclaim to
reduce the DP Pool consumed capacity and improves storage utilization.
Zero Page Reclaim facility examines the volumes of physical capacity and where the firmware determines
that no data other than zeros is found on a Dynamic Provisioning software pool page, the physical storage
is unmapped and 'returned' to the pool's free capacity.
Zero Page Reclaim is intended to be used after initial migration/restore
Migrate from the physical volume to the virtual volume (or restore volume from tape)
Zero Page Reclaim unused pages return physical storage to a dynamically provisioned pool
Note that prior to v8 microcode, the AMS2000 did not zero storage as it was allocated, so for existing volumes less
gain will be seen with ZPR than is seen on a USP.
Physical
Virtual
Virtual
Volume
Volume
Volume
Server Volume
48
Improved Performance Optimization: Dynamic Provisioning on

AMS2000 supports automatic pool rebalancing
Automatic rebalancing after virtual storage pool expansion

- When a new RAID Group is added to a DP Pool, pages are
actively redistributed among all of the RAID Groups in the pool
- Rebalances pool including rebalancing at individual virtual
volume level
- All transparently online with no affect to application I/O
VVOL
1 2 3 4
Pool
1 4
Pool
Vol#
0
Pool
Vol#
1
Pool
Vol#
2
VVOL
1 2 3 4
Pool
1 4
Add Pool Capacity
Pool
Vol#0
3
Pool
Vol#1
Pool
Vol#2
VVOL
1 2 3 4
Pool
Vol#3
Pool
Further simplifies storage

provisioning
Improved performance
optimization
49
1 4
Optimize Pool
Pool
Vol#0
3
Pool
Vol#1
Pool
Vol#2
Pool
Vol#3

Components Overview
Capacity Monitoring
Host
Storage Navigator
Modular 2
(GUI, CLI)
AMS2000
Port
Port
- DP Pool Consumed Capacity Alert

- Over Provisioning Ratio Alert
- DP Pool Trend Information
Port
Capacity Monitoring
Virtual-LU
Virtual-LU
Virtual representation of capacity
presented to Host. Total capacity of
Virtual LUs may exceed the actual
capacity of the DP Pool.
DP Mapping Table
- Mapping information between
Virtual-LU and DP Pool
- Page space information
Cache Memory / Local Memory (CS/DS)
DP Mapping Table
DP Pool
Actual storage capacity that contains
data written by Host/Application.
Comprised of multiple RAID Groups.
DP Mapping Table backup area (=1GB)
DP Pool
RAID Group
RAID Group
Hitachi Dynamic Provisioning provides a host with virtual capacity volumes. Storage is assigned to the volume from a pool of
real storage volumes Just in Time for a write request from the host.
50
Hitachi Dynamic Provisioning Components

Dynamic Provisioning Pool
Dynamic Provisioning Pool (DP-Pool)

Real installed capacity
Cannot be directly referenced from any hosts
DP Pool is composed of 1GB Chunks (32 x
32MB pages) randomly assigned throughout
the available DP Pool volumes
A 1GB Chunk is assigned to a Virtual-LU
DP Pool pages are assigned to Virtual-LUs
Just-in-Time
DP Pool storage will be shared by all
applications whose Virtual-LUs are associated
with the DP Pool
DP Pool itself can not be replicated and it can
not be migrated using the volume migration
function
51
The DP Pool is comprised of

RAID Groups built from physical
HDDs

Virtual Logical Unit
Virtual Logical Unit (Virtual-LU)

Virtual Volume (capacity) that the host
discovers
Must be associated with a DP Pool
Immediately after definition, the
Virtual-LU has no pages assigned
from the DP Pool
Target of Host Reads and Writes
Real storage capacity is assigned to
the Virtual-LU when required for a
write
Is supported as P-VOL or S-VOL for
ShadowImage
52
The DP Pool is comprised of

RAID Groups built from physical
HDDs

Dynamic Provisioning Mapping Table
Dynamic Provisioning Mapping Table
(DP Mapping Table)
Control information for Hitachi
Dynamic Provisioning
Stored in both Cache Memory and
Local Memory
Maps Logical Block Address (LBA) to
DP pages of a Virtual-LU
Provides logical view for Virtual-LU
Mirrored into DP Pool System Area
for real time recovery after complete
power outage
53
Logical View of DP Pool through DP Mapping Table

The DP Pool is comprised of RAID Groups built
from physical HDDs

Improved Storage Management Agility via Monitoring
Capacity Monitoring
54
Monitor the DP Pool utilization

Two customer definable utilization
levels may be set per pool to
generate alerts
Monitor over-provisioning ratio (Total
virtual volume capacity/Total pool
capacity)
Two customer definable over

provisioning ratio thresholds may be
set to generate alerts
Alert options: Blinking LED on
system; Email message; SNMP trap
issued
Error return (WRITE PROTECT) to
the Write command after DP Pool
depletion
Stores the DP Pool Trend information
for the past one year
Transfer the Trend information as
CSV file
Virtual-LU Viewpoint
Over Provisioning
Ratio Alert
Trend
Informatio
n
The DP Pool is comprised of RAID

Groups built from physical HDDs
DP Pool Viewpoint
DP Pool Consumed
Capacity Alert

Example of Trend Information Reported
55

Delivering Performance Improvements
When allocating many Virtual LUs to a Dynamic Provisioning Pool that consists of multiple
storage system groups, the total Host I/O load is evenly distributed across the pools RAID
Groups
This wide striping implementation automatically levels the I/O load and avoids overutilizing a
RAID Group resulting in a performance bottleneck
Without Dynamic Provisioning
LU
LU
LU
LU
LU
With Dynamic Provisioning
Virtual
LU
Virtual
LU
Virtual
LU
Virtual Virtual
LU
LU
Actual capacity pool (DP Pool)

RAID Group#1
RAID Group#2
RAID Group#1
RAID Group #1 becomes

performance bottleneck
56
RAID Group#2
Host I/O load is distributed to

both RAID Group #1 and RAID
Group #2
Hitachi Dynamic Provisioning Operations

Allocating Chunks and Load Balancing
Virtual-LU load balancing
- When receiving new Write data, assign one Chunk (1GB) in the 1st DP Pool RAID Group, then continue writing data to the
(32) 32MB pages that make up Chunk.
- When receiving new Write data for another Virtual-LU, assign the next Chunk in the same 1st DP Pool RAID Group
- When all (32) 32MB pages of the Chunk are allocated, then create a new Chunk in the 2 nd DP Pool RAID Group
Chunk #0 Used Up!!

Chunk #0
Chunk #1
Chunk #2
Virtual-LU #0
DP Pool RAID Group #0
Chunk #3
DP Pool RAID Group

#1
DP Pool
57
Virtual-LU #1
Virtual-LU #2
File System Initial Capacity Requirements

Thin Friendly
Low advantage
No advantage
File-Systems spec for DP Pool capacity utilization.
OS
File System
Meta-Data Writing
(The case of Default parameter)
DP Pool Capacity Consumed
Windows Server2003
NTFS
Writes metadata to first block.
Small (one page)
Windows Server2008
NTFS
Under investigation
Under investigation
Linux
XFS
Writes metadata in Allocation

Group Size intervals.
Depends upon allocation group size. The amount of pool

space consumed will be approximately
[Virtual LU Size]*[32MB/Allocation Group Size]
Ext2 / Ext3
Writes the meta-data in 128MB

intervals.
About 25% of the size of the Virtual-LU.

Note: The default block size for these file systems is 4KB.
This results in 25% of the Virtual-LU acquiring 32MB pages.
If the file system block size is changed to 2KB or less, then
the meta-data is written in 32MB or less intervals, so the DP
Pool capacity consumption becomes 100%
UFS
Writes the metadata in 52MB

increments.
Size of Virtual-LU.
VxFS
Small (one page)
ZFS
Writes metadata in Virtual Storage

Pool Size intervals.
Small (when the Virtual-LU is more than 1GB)
JFS
Writes metadata in 8MB intervals.
Size of Virtual-LU
Note: If you change the Allocation Group Size settings when
you create the file system, the meta-data can be written to a
maximum 64MB intervals. Approximately 50% of the DP
Pool is consumed
JFS2
Small (one page)
VxFS
Small (one page)
Solaris
AIX
58
File System Initial Capacity Requirements

File-Systems spec for DP Pool capacity utilization.
Thin Friendly
Low advantage
No advantage
OS
File System
Meta-Data Writing
(The case of Default parameter)
DP Pool Capacity Consumed
HP-UX
JFS(VxFS)
Writes metadata to first block
Small (one page)
HFS
Writes metadata in 10MB intervals
Size of Virtual-LU.
Windows
Server2003
NTFS
Writes metadata to first block
Small (one page)
Windows
Server2008
NTFS
Under investigation
Under investigation
Linux
Ext2 / Ext3
About 25% of the size of the Virtual-LU.

Note: The default block size for these file systems is 4KB.
This results in 25% of the Virtual-LU acquiring 32MB pages.
If the file system block size is changed to 2KB or less, then
the meta-data is written in 32MB or less intervals, so the DP
Pool capacity consumption becomes 100%.
VMware
Case1: Writes metadata in 2GB intervals

Metadata
2GB
Virtual-LU
Metadata
32MB Page
DP Pool Capacity Consumed: 32MB/2GB = 1.6
59
Case2: Write metadata in 64MB intervals

64MB
Virtual-LU
32MB Page
DP Pool Capacity Consumed: 32MB/64MB = 50
Best practice for creating DP Pool

8D+2P, many RGs in one DP Pool, plural DP Pools
#
Viewpoint
Capacity
utilization
Performance
Items
Best practices
Data disk number for the

RAID Group in the DP
Pool
Recommendation: 2,4,8,16
RG configuration in the
DP Pool
For the site for which high I/O performance is required, you need to configure
the DP Pool with many HDDs and as many RGs as possible
(Ex: 4D+1P, 8D+2P, 16D+2P, etc)

Default Combination = 8D+2P
(By using the Performance Monitor, you can see the Host I/O frequency for
each Virtual-LU and utilization for each RG in the DP Pool.)
3
4
Reliability
RG RAID Level
From performance and reliability viewpoints, the recommendation is 8D+2P.
DP Pool counts
Recommendation: Create plural DP Pools

The backup DP Pool Mapping Table is taken to one of the DP Pools. If there is
a failure in this DP Pool, the backup of DMT is taken to another DP Pool
automatically. Since if there is just one DP Pool this backup process cannot be
executed, creating more than two DP Pools is recommended
60
Best practice for creating Virtual-LU

Multiples of 1GB size
#
1
61
Item
Capacity
Best practice
Recommendation: Set the Virtual-LU on 1GB size boundaries
(If you cannot do this, some DP Pool space will be wasted)
Global sparing with HDD roaming

Improved reliability and
flexibility of recovery point
Ideal for mission-critical data
Mitigate risk of data loss due to
double-disk failures
No need to copy back from

spare HDD after rebuild
Spare becomes part of new RAID
group
Swap bad drive for new spare
Normal Operation
DATA 1
DATA 2
DATA 3
DATA 4
DATA 5
DATA 6
PARITY 1
PARITY 2
62
DATA
3
HOT
SPARE
Rebuild
Operation
Hitachi Cache Partition Manager Feature

Modular storage systems typically have simple single block-size caching
algorithm, resulting in inefficient use of cache and/or I/O
Hitachi Cache Partition Manager feature delivers cache environment that can
be optimized to specific customer application requirements
Less cache required for specific workloads

Better hit rate for the same cache size
Better optimization of I/O throughput for mixed workloads
No need to mirror all writes where this does not benefit the application
Typical 4KB Database

blocks in 16KB Cache
pages
75% Cache Wasted!
63
Cache Partition Manager Feature

Modular storage systems typically have simple single block-size
caching algorithm, resulting in inefficient use of cache and/or I/O
Cache Partition Manager feature delivers cache environment that can
be optimized to specific customer application requirements
Less cache required for specific workloads

Better hit rate for the same cache size
Better optimization of I/O throughput for mixed workloads
No need to mirror all writes where this does not benefit the application
4KB Database blocks
in 4KB Cache pages
16KB File system
blocks in 16KB
Cache pages
512KB Video data
blocks in 64KB
Cache pages
64

Feature Specifications
Item
65
Conditions
Devices and cache supported
Adaptable Modular Storage 2100: 2GB,

4GB/controller
8GB/controller
16GB/controller
Memory segment size

supported
Master partition: Fixed 16KB

Sub partition: 4KB, 8KB, 16KB, 64KB,
256LB, or 512 KB
RAID stripe sizes supported
64KB, 256KB, or 512KB
Number of partitions
Adaptable Modular Storage 2100: 2 to 16


for Services Oriented Storage Solutions
Cache Partition Manager enables cache memory to be sub-divided and
managed for more efficient use:
Each of the divided portions of cache memory is called a partition
LUNs defined in the Adaptable Modular Storage 2000 family systems can
be assigned to a partition
Customer can specify a size of the partition
Optimize the data transmission from/to a host by assigning the most

suitable partition to a LUN according to the application (data) received
from a host
The same Adaptable Modular Storage 2000 family storage system can be
optimized in many different ways depending upon the characteristics of the
application
66
Overview of the Cache Partition Manager

Cache is divided into partitions that can be exclusively used by assigned
LUNs
Adaptable Modular Storage 2100: 2 16 partitions per system

Partition 0/1 are the master partitions (16KB only)
Partition 2~n have selectable-size segments 4, 8, 16, 64, 256, & 512K
Partition sizes are flexible (each partition has a certain minimum)
Controller #1
Controller #0
Dir #0
67
Partition #0
16KB(Fix)
Partition #2
4KB
Partition #3
256KB
Multipurpose
(Default)
For DB
(Small
size
access)
Streamin
g
(Exclusive
Partition)

Cache Partition Manager allows performance tuning and it includes the
following:
Selectable segment size
Customize the cache segment size for a user application
Partitioned cache memory
Decrease negative effect on performance between applications in a 'storage
consolidated' system by dividing the cache memory into multiple partitions
individually used by each application (LUs)
Selectable stripe size
To increase performance by customizing the disk access size
68
Cache Utilization is Key to Application Performance
Without Cache Partitioning Manager
Data in cache memory is

processed uniformly in units of
16KB regardless of lengths of data
sent from a host
Data with a length of 4KB or 8KB
is scattered on the cache memory
16KB
69
4KB data
Cache-hit rate at the time of

response to the read command is
effectively lowered
Cache Utilization is Key to Application Performance
With Cache Partition Manager
The area of the cache memory

is divided into two (for this
example), making the two
portions fit two data lengths
4KB data
When handling a lot of small

Partition 2
data with lengths shorter than
(Sub partition)
16KB, you can raise the cache
hit rate and lower response
16KB
time (read) by specifying data
Partition 0
to be processed via partition 2
(Master partition)
8KB
Contents of the cache memory

are controlled more efficiently
70
Advantage of Selectable Segment Size

Fixed size cache segments (16KB)
16KB
For an application with short

size (ex.4KB) host access
Low cache hit rate!
Adaptable Modular Storage 2000

Selectable Segment Size 8KB
By choosing a smaller segment size the
cache hit rate will increase, and the
performance will increase
71
Advantage of Selectable Segment Size
Fixed size cache segments (16KB)
16KB
Adaptable Modular Storage 2000
Selectable Segment Size 256KB
72
For an application with long

size (ex.128KB) host access
Takes overhead to handle

many segments for each I/O
By choosing larger segment size,

overhead will decrease,
and the performance
will increase.
Advantage of Selectable Stripe Size
Adaptable Modular Storage 2000:

Selectable Stripe Size
High throughput with

concurrent I/Os to HDDs
Good for applications with
sustained I/Os
Stripe Size
256KB
HDDs
1~2 WRITES to HDD
Stripe Size
64KB
1 WRITE to LUN
2~3 WRITES to HDD
73
128KB
128KB
1 WRITE to LUN
Fixed Stripe Size (64KB)
HDDs
Lower overhead because

of fewer HDD I/Os
Good for applications with
transaction I/Os
Cache Residency Manager
LUN
Extent
74
LUN
LUN
LUN
Extent
Cache
Mirrored Cache
Extent
Enhanced
Read/Write
Performance
Cache Residency
Area
Storage Administrator defines a

portion of cache for use by Cache
Residency Manager
Storage Administrator or application
(through API) dynamically maps
extents
All Read/Write activity occurs in
cache
Write data is mirrored and protected
on disk asynchronously.
Unsurpassed Heterogeneous Connectivity

Simplify: Connectivity
Facilitates consolidation with broadest
range of open systems O/S support and
easily accommodates dynamic,
heterogeneous environments, using
native path management and failover
Hitachi Universal Storage Platform V and
VM
Microsoft Windows 2000, Windows
Server 2003, Windows 2008
Hyper-V
Sun Solaris (Sparc and x64 )
VMware
RedHat Enterprise Linux, SuSE Linux,
RedFlag Linux, Oracle Enterprise Linux,
Asianux, MiracleLinux, others
HP-UX
IBM AIX
MAC OSX
Novell NetWare
HP Tru64 UNIX
HP OpenVMS
ADAPTABLE
MODULAR STORAGE
2100
75
Host Storage Domains allow multiple

heterogeneous servers to share a single
port
Host Protocol Options

AMS 2100
Fibre Channel System
(base)
AMS 2500

(max)
iSCSI System
FC and iSCSI
Multiprotocol
FC and iSCSI
Multiprotocol
AMS 2300
(base)
(max)
FC and iSCSI
Multiprotocol
76
FC Port
iSCSI Port
Host Port Intermix

New AMS2000 Controller Designs
AMS2100 (Embedded FC + FC Option)
Optional 2x 8Gb/s FC
User LAN
B
Embedded 2x 8Gb/s FC
F
Mnt LAN
AMS2100 (FC Onboard + iSCSI Option)

Optional 2x 1Gb/s iSCSI
User LAN
Mnt LAN
AMS2300 (FC Onboard + FC Option)

Optional 4x 8Gb/s FC
User LAN
H
Mnt LAN
AMS2300 (FC Onboard + iSCSI Option)

Optional 2x 1Gb/s iSCSI
User LAN
D
F
Mnt LAN
77
10 GbE / 1GbE FCIP with the iSR6250-HDS
Remote replication with AMS2500 and USP

Cost-effective remote BC and DR connections
iSR6250-HDS
AMS with TrueCopy SAN-over-WAN replication

Storage Subsystem Drag; SW + Array revenue
1GbE FCIP
8Gb FC
WAN
iSR6250-HDS
78
Hitachi Volume Migration Modular Software for

Data Lifecycle Management
Example: SAS to SATA
SAS
LU1
Migration
finish
SAS
LU2
Online
Migration
SATA
LU2
79
SATA
LU1
Volume Migration does

not require host resource
Does not require change
of host configuration,
mount, or boot.
Low I/O access LU
migration from SAS HDD
to SATA HDD
Or move heavy IO LU
from 10K RPM to 15K
RPM HDDs
Hitachi Volume Migration Software for

Performance Improvement
Example: Performance Tuning
Use Hitachi Performance Monitor feature

to find performance bottleneck.
High load I/O
LU1
Online
Migration
High load RG
LU2
Low load RG
80
High load I/O
LU2
Migration
finish
Medium load RG
LU1
Medium load RG
Performance imbalance is
addressed
Performance bottleneck is
removed
ShadowImage In-System Replication/Copy-on-Write

Snapshot Software Cloning
Hitachi ShadowImage In-System Replication software

Full Volume Clone/Copy-on-Write (PiT) Snapshot
software
ShadowImage full volume clones per source 8 maximum
ShadowImage command devices 128 maximum
ShadowImage copy requests per AMS2500 - 2047 maximum
Copy-on-Write snapshots per source - 32 maximum

Copy-on-Write command devices 128 maximum
Copy-on-Write PiT snapshots per AMS2500 - 2046 maximum
Reverse resynchronization - supported
Instant snap restore - supported
Auto I/O Switch on double/triple drive failure - supported
81
TrueCopy Remote Replication Software for

Business Continuity
TrueCopy Synchronous
1000
2500
Bi-directional
Differential re-synchronization
Reverse re-synchronization
R/W Source and Read Target Continuous Access
Copy between AMS/WMS and AMS2000 models
TrueCopy Extended Distance (Asynchronous)
2500
82
1000
Automated source and remote synchronization

Automated local and remote RPO snapshots
Consistency Group support
Write order fidelity preservation
Bi-directional
Local or remote site back-up capability
Copy between AMS2000 and AMS1000/500 models
TrueCopy Remote Replication Software for

Business Continuity
TrueCopy Synchronous
Bi-directional
Differential re-synchronization
Reverse re-synchronization
R/W Source and Read Target Continuous
Access
TrueCopy Extended Distance (Asynchronous)
83
Automated source and remote synchronization

Automated local and remote RPO snapshots
Consistency Group support
Write order fidelity preservation
Bi-directional
Local or remote site back-up capability
TrueCopy Extended Distance Software

TrueCopy Update Sequence
Updates
T0
PVOL
Updates
T2
T1
T1
SVOL
T2
Snapshot
T2
T1
T2
T3
Snapshot
T3
Send
Differentials
Send
Differentials
T1
Snapshot
T1
Available 2H CY06
Update Cycle
84
T3
T2
Snapshot
Same
Data
Disaster
Updates
T2
T2
Snapshot
T2
Restore
Time
Power Savings
Lower Total Cost of Ownership
Hitachi Data Systems HDD spin down solutions
Spinning down of
ShadowImage drive groups involved in backup to tape

VTL drive groups involved in backups
Local internal backup copy of data
Drive groups within archive storage

Unused drive groups
Spun-Up RAID Group
Deliver power consumption savings by
Spun-Down RAID Group
Reduce electrical power consumption of idled hard drives

Reduce cooling costs related to heat generated by the hard drives.
85
The Hitachi Green Storage Initiative

Power savings (power down) option available
Higher density drives utilizing less power = fewer BTUs
Saves money, saves the environment
SATA II drives take advantage of power modes

Heads unload whenever HDD is idle for 2 hours
Variable fan speeds

Internal temperature of the subsystem will control the speed of the fans
86
System and Data Security

Role Based Access Control
System management access based on one of three roles (account administrator,
storage administrator or auditor)
Unique login and password for each user
Active Directory integration in conjunction with Command Suite 7.0
System Audit Logging

Records all system changes and all power ups and downs
Sends all changes to Syslog server
Enables IT managers to meet compliance requirements including auditable trails
Data and System Security
Self-encrypting drive option

Communication between management software and systems is SSL/TLS encrypted
Maintenance ports support IPv6 security specs
All models meet Common Criteria EAL2 Certification
Data Retention Utility (optional)

Tamperproof Write Once Read Many (WORM) makes selected data non-erasable and
non-rewriteable
Read access can be limited
87
Questions/Discussion
88
Thank You
89

AMS2000 Technical Presentation - Mar11!01!057522

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

AMS2000 Technical Presentation - Mar11!01!057522

Uploaded by

Copyright:

Available Formats

Hitachi Adaptable Modular

Storage 2000 Family

2008 Hitachi Data Systems

Symmetric active/active controller with dynamic load balancing

Hitachi Adaptable Modular Storage 2000 Family

The Hitachi Adaptable Modular System

Gartner positioned Hitachi & the Hitachi

Hitachi achieves overall best-in-class Storage Performance Council (SPC-1)

Hitachi Data Systems has consolidated

The AMS2000 series combination

Hitachi Storage Lineup

Full redundancy for 99.999% availability

Scale Performance &

Virtual Storage Platform

* 60/40 workload with <10% cache hit

Adaptable Modular Storage 2000 Family

Hitachi Adaptable Modular Storage 2100

SATA-II and SAS intermix

* RAID-0 available on SAS drives only

Hitachi Adaptable Modular Storage 2300

SATA-II and SAS intermix

* RAID-0 available on SAS drives only

3U trays (max of 15)

Hitachi Adaptable Modular Storage 2500

SATA-II and SAS intermix

* RAID-0 available on SAS drives only

3U trays (max of 32)

Hitachi Adaptable Modular Storage 2100 Architecture

Hitachi Adaptable Modular Storage 2300 Architecture:

Hitachi Adaptable Modular Storage 2500 Architecture

Hitachi Adaptable Modular Storage 2000 Family

Hitachi Adaptable Modular Storage 2000 Family Disk Enclosure

2 SAS Wide Cables

RKAK (SAS and SATA-II Enclosure)

SAS Engine Overview

External Disk Enclosure 15 Disk

External Disk Enclosure High Density

Dense Expansion Disk Tray (4U)

Up to 38 high performance disks

Nearly doubles the density of high

Reduce data center floor space

Dense Expansion Tray (2100 and 2300)

HDDs per Tray

Height per Tray

Max w/ only High Density Trays

Max w/ only High Density Trays

Max w/ only Standard Trays

Dense Expansion Tray (2500)

HDDS per Tray

Dense Expansion Tray

SAS Cable Holder X 8

Left Side: Input

SAVE Space and Power with smaller Drives

Hardware image for the 2.5 Expansion Unit

Unit LED: READY (Green)

ENC LED (POWER, LOCATE, ALARM)

ENC LED ()POWER, LOCATE, ALARM

Comparison between 2.5 Expansion Unit

2.5 Expansion Unit

DF800 Expansion Unit (RKAK)

Connectable Basic Unit

Unused (SRAM is mounted)

IN :Key slot3 & key4,6