1. Manage Your Risk with Business Continuity and Disaster Recovery
EMC RecoverPoint
Network-Based Intelligent Data Protection
Muhammad Salama
Sr. Technology Consultant

2. Data-Replication Pain Points in Heterogeneous Environments
Application response time
Local site
Remote site
Oracle
Exchange
SQL
Application-consistent recovery
Oracle
Exchange
SQL
Corruption protection
SAN
SAN
SAN
Disaster-recovery testing
Communications cost
infrastructure
SUN
IBM HP
HDS
EMC
SUN
IBM HP
HDS
EMC
Heterogeneous storage

3. Data Replication Technologies…
Host
RepliStor
Network
RecoverPoint
Array
MirrorView/S
MirrorView/A
SAN Copy

4. Remote Recovery: RPO and RTO
Recovery Point Objectives
(RPO)
Recovery Time Objectives
(RTO) 5
a.m. 6 7 8 9 10 11 12 1
p.m. 2 3 4
DECLARE
DISASTER
10 a.m.
RPO Definition: Amount of data lost from failure, measured as the amount of time from a disaster event
RTO Definition: Targeted amount of time to restart a business service after a disaster event

5. Continuous Data Protection
Operational Recovery
SNAPSHOT
Snapshots and Clones
Either space efficient change-based snapshots for backup off-loading
Or, full copies or clones for test and development with minimal production impact
Journal every change made and set the window for the amount of time to record
Creates copies like “time-lapse photographs”
Continuous Data Protection
Change-based “VIDEO CAMERA” with DVR-like roll back capabilities

6. RecoverPoint Continuous Data Protection (CDP)
RecoverPoint splitter drivers
Mirrors server writes to RecoverPoint appliance
Resides on host or in fabric
RecoverPoint appliance
Managed and prioritizes resources
Writes changes to local CDP Journal
Distributes changes to target
Journal
Tracks all data changes to every protected LUN
Utilizes bookmarks for application-aware recovery
Repository for live data updates
Provisioned from existing SAN LUNs
Dynamically compressed, which saves storage
Supports heterogeneous environments
RecoverPoint supports EMC and third-party storage*
RecoverPoint/SE supports CLARiiON arrays only
Application
servers
Database
servers
Messaging
servers
File and Print servers Œ
SAN
Third-
Party
EMC
Disk
systems
Local CDP
Journal
* Support for third-party storage requires intelligent fabric

7. RecoverPoint Local Protection Process—CDP
2a. Host splitter
1. Data is split and sent to the RecoverPoint appliance in one of three ways
3. Writes are acknowledged back from the RecoverPoint appliance
2b. Intelligent-fabric splitter
This slide describes the data flow from the application host to the production volumes, and how the RecoverPoint appliance accesses the data as part of the CDP process.
(Note to Presenter: Click now in Slide Show mode for animation.) An application server issues a write to a LUN that is being protected by RecoverPoint. This write is split, then sent to the RecoverPoint appliance in one of three ways:
<CLICK> The first way is through a host splitter. This host splitter resides in the I/O stack, residing below any file system and volume manager, and just above any multi-path driver (such as EMC PowerPath). The splitter looks at the destination for the write packet. If to a LUN that RecoverPoint protects, the splitter will send a copy of the write packet to the RecoverPoint appliance. It does this by rewriting the target address inside the packet to redirect it to the RecoverPoint appliance’s pseudo-LUN, and reissuing the write down the stack.
<CLICK> The second way is through an intelligent fabric switch-based splitter, such as either the Connectrix AP-7600B or Connectrix ED-48000B with the SAS APIs, or one of the Connectrix MDS-9000 series switches with the Connectrix MDS SSM and the SANTap API. The switch will intercept all writes to LUNs being protected by RecoverPoint, and will send a copy of that write to the RecoverPoint appliance.
<CLICK> The third way is through a CLARiiON array-based splitter, which is supported on CLARiiON CX3 arrays with FLARE 26+ patch code and on CLARiiON arrays with FLARE 28. The array will intercept all writes to LUNs being protected by RecoverPoint, and will send a copies to the RecoverPoint appliance.
In all cases, the original write travels though its normal path to the production LUN. <CLICK> When the copy of the write is received by the RecoverPoint appliance, it is acknowledged back (ACK). This ACK is received by the splitter (either the host or fabric splitter), and held until the ACK is received back from the production LUN. With both ACKs received, the ACK is sent back to the host, and I/O continues normally.
<CLICK> Once the appliance has acknowledged the write, it will move the data into the local Journal volume, along with a time stamp and any application-, event-, or user-generated bookmarks for the write. <CLICK> Once the data is safely in the Journal, it is then distributed to the target replica volumes, with care taken to ensure that write order is preserved during this distribution.
4. The appliance writes data to the Journal volume, along with time stamp and application-specific bookmarks
2c. CLARiiON splitter
/ A
/ C
/ B
r A
r C
r B
Production volumes
Replica volumes
Journal volume
5. Write-order-consistent data is distributed to the replica volumes

8. RecoverPoint—Recovery Point Consolidation in Journal
JOURNAL RECOVERY POINT TIMELINE
Two days
ago
Continuous Recovery Points
Daily Recovery Points
Weekly Recovery Points
Monthly Recovery Points
Now
Monthly
Weekly
Daily
Continuous

9. RecoverPoint Continuous Remote Replication (CRR)Œ
SAN
SAN or
WAN
SAN
Third-
Party
EMC
EMC
Third-
Party
RecoverPoint splitter drivers
Intercepts server writes (block level)
Resides on host, CLARiiON, or in fabric*
Mirrors write to RecoverPoint appliance
RecoverPoint appliance
Bandwidth reduction with data compression for WAN transfer
Distributes changes to remote site
Supports instant access to protected data
Journal
Tracks all data changes to every protected LUN
Utilizes bookmarks for application-aware recovery
Dynamically compressed, which saves storage
Provides advanced functionality
Policy-based bandwidth management
5–10 times bandwidth reduction
Replicate over IP or Fibre Channel
Supports heterogeneous environments
RecoverPoint supports EMC and third-party storage*
RecoverPoint/SE supports CLARiiON arrays only
* Support for third-party storage requires intelligent fabric

10. RecoverPoint Concurrent Local and Remote (CLR) Data Protection
SAN
SAN/WAN
PRODUCTION SITE
DISASTER RECOVERY SITE
Cluster Passive Node
Cluster Active Node
Replication
Data Flow
RecoverPoint appliances
RecoverPoint
Replication Services
Tape Backup Manager
Standby Disaster Recovery Server
Tape Library
Local
Journal
Storage Groups
and Logs
Remote
Replicated Storage Groups and Logs
Performance architecture
Out-of-band design leveraging intelligent host and fabric interfaces*
Supports CLARiiON write splitting on CX3 and CX4 arrays
Designed to work in enterprise-class environments
Replication across heterogeneous storage*
Leverage existing storage investments
Co-exists with local CDP
True bi-directional, any-to-any replication
Replicate between arrays at same or different site*
True CDP data protection for applications
All writes stored in Journal with application bookmarks for recovery
Supports Microsoft Volume Shadowcopy Service (VSS) and VDI APIs
Concurrent local and remote data protection
Create local and remote copies of the same LUNs
Recover both locally or remotely to different point-in-time images
No impact to production or the other replica during recovery
Unified management interface
Remotely configure, monitor, manage CDP/CRR
Programming CLI for intelligent scripting
* RecoverPoint/SE does not support intelligent fabric, and only supports a single CLARiiON array at each side/site

11. RecoverPoint Stretched CDP over Stretch Fibre
PRIMARY SITE
REMOTE BUNKER SITE
Application
servers
Database
servers
File and print servers
Application
servers
Database
servers
File and print servers
Management
IP WAN
SAN
RecoverPoint
SAN
Stretch
Fibre
Supports synchronous CDP replication to secondary site (such as a bunker site)
Distance subject to specific configurations; refer to the EMC Support Matrix for guidance
Enables survival of primary site outage
Provides synchronous replication across distance
Remote applications can access any point in time image as read/write without impacting production
All RecoverPoint resources installed at remote site
CDP target storage, journal
RecoverPoint appliances
RecoverPoint repository volume
Standby/disaster recovery servers
Dual fabric that is extended to remote site
Servers and splitters must be at both sites
Reversing replication by promoting CDP replica to production is not supported
Production
Volumes
CDP Replica
Volumes
Stretched CDP
Failover
CDP Journal
Primary
Journal 11

12. Cascaded RecoverPoint Topology Example
PRIMARY SITE
SECONDARY SITE
TERTIARY SITE
Manage-ment
IP WAN
Manage-ment
IP WAN
WAN
SAN
SAN
SAN
Stretch
Fibre
CDP Replica
Volumes
CRR Replica
Volumes
Production
Volumes
Failover
CDP Journal
CDP
Journal
CRR
Journal
Near-Synchronous*
Asynchronous Unlimited*
Local and remote data protection minimizes data loss
* Refer to the EMC Support Matrix for the maximum distance for your configuration

13. Policies for Dynamic Sync/Async Replication

14. Journaling for Application-Aware Recovery
Journal Includes Data Plus Metadata
Time/date
Identifies the time image was saved
Bookmarks:
System-generated group bookmarks
e.g., Volume Shadowcopy Service (VSS) backup
User-generated bookmarks
Other EMC product bookmarks
EMC Replication Manager
System-event-generated bookmarks
Microsoft SQL Server
Microsoft Virtual Device Interface (VDI) operations
Microsoft Exchange
Microsoft VSS

15. RecoverPoint Recovery Process
Request image via management interface
ΠUser selects by bookmark or point in time
RecoverPoint recovers image as read/write
User chooses virtual  or physical target 
Existing Journal data  used to play back changes to appropriate bookmark or point in time
Target-mounted read/write to recovery server 
All new writes saved in Journal 
Existing disaster-recovery protection not impacted
Image changes committed on request
Recovery-image use cases
Failover production to image
Start new application from image
Surgically repair data from image
Fast production resynchronization
Rollback production from changed blocks
Source for backup
Data analysis
Disaster-recovery development and testing (“fire drill”) Œ 
Recovery
request
Servers at
recovery site
RecoverPoint appliance
SAN v
History Journal
Target volumes

16. Understanding Data Consistency
CRM
Order Entry
Applications and data are interrelated (Federated)
All data movement must be stopped/started at the same point in time
To restart applications you must have all the data—not parts of it
Recovery requires dependent-write consistency across all volumes and systems DB DB DB
SCM
Systems share information…how do you get a consistent view?

17. Grouping for a Consistent View
Allows application recovery to be tiered by service level
Multiple volumes per group
Mixed recovery point objectives within same infrastructure
Provides independent replication controls
Recover by group, locally or remotely
Start/stop by group
Enables grouping of optimization
Importance
Resource usage
Recovery point and recovery time objectives OE
CRR
Group 1
CRM
CDP
CRR
Group 2
SCM
CDP
CRR
CRR
Group 3

18. Grouping for Federated Environments
Each tier has different service level agreements
Consistency groups per tier
Operational recovery of tier
Parallel consistency across tiers
Federated environments
Recover to a known point for all applications
Disaster recovery for tier or application
Spans operating systems, applications, storage, and servers
Enables advanced functions
Full environment cloning
Application upgrade testing
Data mining
Consistent production rebuild
1: Linux (Web OE)
Consistency
group
2: Windows (CRM)
Consistency
group
3: UNIX (SCM, Financials…)
Consistency
group

19. Summary Lowest total cost of ownership Integrated CLARiiON support
Reduction and consolidation of server, storage footprint (capital expenditure), and WAN bandwidth (operational expenditure)
Integrated CLARiiON support
CLARiiON-based splitter on CX3 and CX4
Virtual Provisioning support on CLARiiON CX4
Intelligent SAN-switch integration
RecoverPoint supports Connectrix using Brocade and Cisco technology
Any-point-in-time recovery
Rollback of data at the local site and/or remote site
Highest availability, highest performance
Network-based architecture optimized for application availability and performance
Heterogeneous support
RecoverPoint supports spectrum of host, storage, and SAN elements

20. Thank you.