Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

SRDF Foundations

© 2007 EMC Corporation. All rights reserved.

Welcome to SRDF foundations

The AUDIO portion of this course is supplemental to the material and is not a replacement for the
student notes accompanying this course.
EMC recommends downloading the Student Resource Guide from the Supporting Materials tab, and
reading the notes in their entirety.

Copyright © 2007 EMC Corporation. All rights reserved. These materials may not be copied without EMC's written
consent. Use, copying, and distribution of any EMC software described in this publication requires an applicable software
license.
THE INFORMATION IN THIS PUBLICATION IS PROVIDED “AS IS.” EMC CORPORATION MAKES NO
REPRESENTATIONS OR WARRANTIES OF ANY KIND WITH RESPECT TO THE INFORMATION IN THIS
PUBLICATION, AND SPECIFICALLY DISCLAIMS IMPLIED WARRANTIES OF MERCHANTABILITY OR
FITNESS FOR A PARTICULAR PURPOSE.
Celerra, CLARalert, CLARiiON, Connectrix, Dantz, Documentum, EMC, EMC2, HighRoad, Legato, Navisphere,
PowerPath, ResourcePak, SnapView/IP, SRDF, Symmetrix, TimeFinder, VisualSAN, “where information lives” are
registered trademarks.
Access Logix, AutoAdvice, Automated Resource Manager, AutoSwap, AVALONidm, C-Clip, Celerra Replicator, Centera,
CentraStar, CLARevent, CopyCross, CopyPoint, DatabaseXtender, Direct Matrix, Direct Matrix Architecture, EDM, E-
Lab, EMC Automated Networked Storage, EMC ControlCenter, EMC Developers Program, EMC OnCourse, EMC Proven,
EMC Snap, Enginuity, FarPoint, FLARE, GeoSpan, InfoMover, MirrorView, NetWin, OnAlert, OpenScale, Powerlink,
PowerVolume, RepliCare, SafeLine, SAN Architect, SAN Copy, SAN Manager, SDMS, SnapSure, SnapView,
StorageScope, SupportMate, SymmAPI, SymmEnabler, Symmetrix DMX, Universal Data Tone, VisualSRM are
trademarks of EMC Corporation. All other trademarks used herein are the property of their respective owners.

SRDF Foundations -1
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

Course Objectives
Upon completion of this course, you will be able to:
y Identify the concepts, value, and environmental aspects
of using SRDF
y Identify and differentiate SRDF features
y Describe the architectural components and theory of
operations of SRDF
y Identify and describe SRDF Management Software
offerings
y Identify the business benefits of SRDF

© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 2

The objectives for this course are shown here. Please take a moment to read them.

SRDF Foundations -2
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

Benefits of Symmetrix SRDF

The Value of SRDF

© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 3

In this module, the value of SRDF as a business tool is reviewed.

SRDF Foundations -3
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

Symmetrix Remote Data Facility (SRDF)

y What is SRDF?
– A business tool that allows a client with Symmetrix based data
centers to copy their data between the sites for a number of
purposes

y SRDF offers various levels of Symmetrix based business

continuance and disaster recovery solutions
y SRDF products offer the capability to maintain multiple,
host-independent, mirrored copies of data
y SRDF allows for the Symmetrix systems to be in the
same room, in different buildings within the same
campus, or hundreds to thousands of kilometers apart

© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 4

Symmetrix Remote Data Facility is a business tool that allows a client with Symmetrix based data
centers to copy their data between sites for a number of purposes.

SRDF Foundations -4
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

SRDF - A Remote Data Replication Business Tool

y By maintaining copies of data in different physical
locations, SRDF enables you to perform the following
operations by integrating with your strategies for:
– Disaster restart, Disaster restart testing
– Recovery from planned outages, Remote backup
– Data center migration, Data replication and mobility

y SRDF helps meet your objectives

– Recovery Point Objective
¾ Point in time to which systems and data must be
recovered after an outage
– Recovery Time Objective
¾ Period of time within which systems, applications,
and functions must be recovered after an outage

© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 5

By integration with business solutions for the tasks listed in the slide, SRDF becomes an integral part
of the overall business solution strategy.
SRDF will help you meet your Recovery Point Objective and Recovery Time Objective.
The Recovery Point Objective, RPO, is the point in time to which systems and data must be recovered
after an outage (for example, the end of the previous day’s processing). RPOs are often used as the
basis for the development of replication and backup strategies and as a determinant of the amount of
data that may need to be recreated after the systems or functions have been recovered.
Recovery Time Objective, RTO, is the period of time within which systems, applications, and
functions must be recovered after an outage (for example, one business day). RTOs are often used as
the basis for the development of recovery strategies, and as a determinant as to whether or not to
implement the recovery strategies during a disaster situation. Similar terms include Maximum
Allowable Downtime.

SRDF Foundations -5
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

SRDF Benefits
y Protect against local and regional y Migrate, consolidate or distribute
site disruptions data across storage platforms
– Continuous data availability – Data center consolidations
– Multiple remote recovery sites – Technology refreshes
– Meet regulatory requirements y Enable non-stop operations
y Provide near-instant recovery – Application restart across volumes

Production SRDF links Recovery

Site Site

Source Target

© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 6

The right remote replication solution can limit your exposure to planned and unplanned downtime by
enabling operations at remote sties. Perhaps you need to provide your organization with efficient data
replication to meet corporate or government standards, while still meeting Total Cost of Ownership
requirements. Additionally, you need a flexible solution that changes as your needs change.
No matter what the challenge is, there is one underlying theme: data protection and faster business
restart in the event of a disaster or unplanned outage is critical across the organization.
There are several methods of remote replication available, each with a different profile to help you
manage and conquer your current business challenges, while enabling new processes and procedures
that help you gain a significant competitive advantage - something all businesses strive for in today’s
competitive marketplace.

SRDF Foundations -6
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

SRDF Features and Differences - 1

y The SRDF Product family
SRDF/Star
– Offers solutions for many Multi-point
SRDF/S
different disaster recovery and Synchronous for
replication
option
business continuance zero data exposure

SRDF Family
scenarios SRDF/CG
Consistency
– Consists of three base SRDF/A Group option

solutions Asynchronous for

extended distances SRDF/AR
¾ SRDF/S Automated
Replication
¾ SRDF/A SRDF/DM
option

¾ SRDF/DM Efficient Symmetrix-to- SRDF/CE

Symmetrix data
– Consists of four add-on mobility Cluster
Enabler option
solutions
¾ SRDF/Star
¾ SRDF/CG
¾ SRDF/AR
¾ SRDF/CE
© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 7

The SRDF family of products offers solutions for many different disaster recovery and business continuance scenarios. The
base solutions include SRDF/S, SRDF/A, and SRDF/DM. Add-On solutions include SRDF/Star, SRDF/CG, SRDF/AR,
and SRDF/CE. For simplicity, this course refers to this product simply as "SRDF".
SRDF/Synchronous (SRDF/S) maintains a real-time (synchronous) copy of data at the logical volume level in Symmetrix
or Symmetrix DMX™ systems.
SRDF/Asynchronous (SRDF/A) is another mode of remote replication that allows customers to asynchronously replicate
data while maintaining a dependent write consistent copy of the data at the secondary site at all times. It uses Delta Set
technology.
SRDF/Data Mobility (SRDF/DM) permits operation in SRDF adaptive copy mode only and is designed for data replication
and/or migration between two or more Symmetrix systems.
Optional, Add-on Functionality
SRDF/Star provides advanced multi-site business continuity protection available. It enables concurrent SRDF/S and
SRDF/A operations from the same source volumes with the ability to incrementally establish an SRDF/A session between
the two remote sites in the event of a primary site outage.
SRDF/Consistency Groups (SRDF/CG) is a product offering designed to ensure the consistency of the data remotely
mirrored by the SRDF operations, in the event of a rolling disaster.
SRDF/Automated Replication (SRDF/AR) is an automation solution that uses both SRDF and TimeFinder to provide a
periodic asynchronous replication of a restartable data image.
SRDF/Cluster Enabler (SRDF/CE) for MSCS or VCS provides high availability and automated failover through storage-
based replication and server clustering through SRDF/S and MSCS or VCS.

SRDF Foundations -7
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

SRDF Features and Differences - 2

SRDF/Synchronous SRDF/Star
y No data exposure yNo data loss on single point of failure
y Some performance impact yGeographically dispersed protection
y Limited distance yUnlimited distance
y Supports higher-tier yAbility to enable SRDF/A between
applications remote locations
SRDF/Asynchronous
SRDF/Automated Replication
y Seconds of data exposure
yZero or hours of data exposure
y No performance impact
yNo performance impact
y Unlimited distance
yUnlimited distance
y Supports multiple application
tiers yRequires BCVs

© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 8

SRDF/Synchronous
For service-level requirements dictate zero data loss EMC offers synchronous mirroring. However, as with any
synchronous solution, there are characteristics that must be understood: Distance is limited by application
timeouts and speed-of-light issues, and bandwidth must be sized for peak workload at all times.
EMC can also deliver solutions that combine SRDF/S with TimeFinder/Clone or TimeFinder/Mirror Business
Continuance Volumes (BCVs) to create Single-Hop and Multi-Hop environments for specialized needs. These
solutions offer different RPOs and have different requirements for bandwidth, supported distances, etc.
SRDF/Asynchronous
EMC has a solution for service-level requirements that need RPOs in the seconds. So, no matter what your
requirements are, EMC can help deliver the right solution.
SRDF/AR
Through the use of Multi-Hop SRDF/AR, you can configure a solution that offers the protection of no data loss
over any distance in the event of a single-site disruption.
SRDF Star
EMC also offers an advanced mechanism, called SRDF/Star, that combines Concurrent SRDF/A and SRDF/S
with the ability to immediately enable protection between two remote sites in the event of a source-site failure.
This adds continuous protection to the remaining two sites quickly and seamlessly.

SRDF Foundations -8
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

Operational Characteristics

Internal Characteristics of SRDF

© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 9

Having briefly defined SRDF, the internal mechanisms of the tool are reviewed next.

SRDF Foundations -9
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

Symmetrix Remote Data Facility

Data mirroring between physically separate Symmetrix systems
y Foundation for Symmetrix based disaster restart solutions
y Host operating systems, DBMS, file-system independent
y Bi-directional source-to-target(s) architecture
y Architecture uses delta technology (Track Tables)
– only changed data copied

Production A
Mirror A
SRDF

Mirror B Production B

© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 10

SRDF provides comprehensive business continuity and restart capabilities for planned and unplanned
outages. This online, host-independent, mirrored data solution duplicates production site data on one or
more physically separate target Symmetrix systems. Systems can be across the room, across the globe,
or anywhere in between.
What makes this technology a leader in the industry is its use of track tables. While we will not get into
track table details, the point is that we have a unique capability to send only changed information at the
block level, thus allowing us to operate in several modes, such as those listed here, which are
addressed later in this presentation.

SRDF Foundations - 10
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

SRDF Architectural Components – Source and

Target Volumes on DMX System
y Symmetrix Logical Volume types:
– SRDF Source or R1 Volumes - Primary Volume with Read/Write (R/W)
access to local host
– SRDF Target or R2 Volumes - Secondary Volume typically Write Disabled
(WD) to secondary host

y The attached host is unaware of SRDF protection

RW WD
Source Target
M1 M2 M3 M4 M1 M2 M3 M4

y Locally Protected SRDF Source Volume: y Locally Protected SRDF Target Volume:
– Remote mirror “floats” and uses next available – Remote mirror “floats” and uses next available
mirror position mirror position
– Allows for dynamic creation of SRDF volumes – Allows for dynamic creation of SRDF volumes

© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 11

This slide illustrates the representation of the mirror positions for a DMX system when the Source and
Target SRDF Logical Volumes have local protection (RAID-1).
The Target-R2 volume is also represented with 4 mirror positions and has local mirroring
implemented, but only three of the mirror positions are used. The third mirror position (M3) is a
pointer back to the Source Symmetrix Logical Volume, while the first (M1) and second (M2) mirror
positions represent local mirrors. If a BCV is established with the R2 volume, then it occupies the next
available mirror position (M4).
Under normal circumstances, the R1 volume presents a Read-Write (RW) status to the host which
accesses it, and the R2 presents Write-Disabled (WD) to its host.

SRDF Foundations - 11
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

SRDF Modes of Operation

y Modes
– Synchronous Replication
– Semi-Synchronous Replication
– Adaptive Copy Replication Performance vs. Availability
– Asynchronous Replication

y Operational Modes can be changed dynamically using CLI or other

EMC tools
y All methods of replication can co-exist within the Symmetrix array
y You can specify the operational method on a per device basis
y Performance, level of synchronization, and I/O serialization
requirements determine appropriate operational mode

© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 12

The modes of operation are Synchronous Replication, Semi-Synchronous Replication, Adaptive Copy
Replication, and Asynchronous Replication.
Synchronous Replication (SRDF/S) provides real-time mirroring of data between the source
Symmetrix and the target Symmetrix systems.
Semi-Synchronous Replication (SRDF/S) writes data to the source system, completes the I/O, and then
synchronizes the data with the target system.
Adaptive Copy Replication (SRDF/DM) transfers data from the source devices to the remote devices
without waiting for an acknowledgments.
Asynchronous Replication (SRDF/A) places host writes into cycles or "chunks" and then transfers an
entire chunk to the target system. These operational modes only affect those Symmetrix volumes that
are remotely mirrored. I/O operations with local volumes occur as if they are in a non-SRDF
environment as no updates are required to volumes at a different physical location.
There is a preferred mode of operation for each SRDF solution. These operational modes are
selectable based on requirements of performance, distance, and speed of recovery. All methods of
replication can co-exist in a Symmetrix array. SRDF/S and SRDF/DM allows you to specify the
method on a per device basis. SRDF/A is set at the RA Group level, not the volume level. No special
application coding is required and no CPU overhead is incurred when implementing SRDF.

SRDF Foundations - 12
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

Synchronous Mode

SRDF links

Source Target

The write I/O is received from host/server into cache of source

and placed in FIFO queue
The write I/O is transmitted to the cache of the target
Receipt acknowledgment is provided by target back to
cache of source
Ending status is presented to host/server

© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 13

SRDF Semi-Synchronous Mode is used primarily in extended distance environments. In this mode of
operation, data on the remotely mirrored volumes are always synchronized between the source (R1)
volume and the target (R2) volume, prior to initiating the next write operation to these volumes. As
with Synchronous Mode, the data movement is at the block level and it works on a first in, first out
model.
The sequence of operations is:
y The write I/O is received from the host/server into the cache of the source.
y Ending status is presented to host/server and the write I/O is placed in FIFO queue.
y The write I/O is transmitted to the cache of the target. Note: A second write will not be accepted on
a Symmetrix source device until its target device has been synchronized.
y Once the device is synchronized, a receipt acknowledgment is provided by the target back to the
cache of the source.
The target is, at most, one I/O operation behind per source logical volume.
Semi-Synchronous Mode masks the impact of distance in the general case because it allows read
operations while write operations are in transit. Again, SRDF uses a first-in, first-out queue.
It is important to note that Semi-Synchronous mode is not supported on DMX3.

SRDF Foundations - 13
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

Semi-Synchronous Mode

SRDF links

Source Target

The write I/O is received from host/server into cache of source

Ending status is presented to host/server and the write I/O

is placed in FIFO queue
The write I/O is transmitted to cache of target
Receipt acknowledgment provided by target back to
cache of source
© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 14

SRDF Adaptive Copy Mode allows the source (R1) volumes and target (R2) volumes to be a few or many I/Os out of
synchronization. Unlike Synchronous and Semi-Synchronous, it does not use the first in-first out model, target data is only
usable after a full synchronization. This operational mode is not recommended for use when mirroring for disaster
recovery/restart purposes, unless used with TimeFinder. This mode is used primarily for data migrations and data center
moves
There are two types of adaptive copy: Write Pending Mode and Disk Mode. Disk Mode data movement is handled at the
track level. Data from the cache is moved to the R1 volume, then keeps the track-level information as to what data is owed
to the remote side so that information can be subsequently sent one track at a time to the source device. Pending Mode data
accumulates in a local cache. A background process moves the data tasks to the source device and its corresponding target
device. The advantage of this method is that it is typically faster to read data from cache than from disk. The disadvantage
is that cache is temporarily consumed by the data until it moves to disk. Note that both methods allow write tasks to
accumulate on the local side before being copied to the remote side.
The sequence of operations is:
y An I/O write is received from the host/server into the cache of the source
y An ending status is presented to the host/server
y I/O is placed in the SRDF queue
y I/O is destaged and processed using Write Pending Mode or Disk Mode
y The I/O is transmitted to the cache of the target
y A receipt acknowledgment is provided by the target back to the cache of the source

SRDF Foundations - 14
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

Adaptive Copy Mode

SRDF links

Source Target
(and)

The write I/O received from host/server into cache of source

Ending status is presented to host/server
I/O is destaged and processed using Write Pending Mode or Disk Mode
I/O is placed in SRDF queue
I/O is transmitted to cache of the target
Receipt acknowledgment is provided by target back to cache of source
© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 15

SRDF Adaptive Copy Mode allows the source (R1) volumes and target (R2) volumes to be a few or many I/Os out of
synchronization. Unlike Synchronous and Semi-Synchronous, it does not use the first in-first out model, target data is only
usable after a full synchronization. This operational mode is not recommended for use when mirroring for disaster
recovery/restart purposes, unless used with TimeFinder. This mode is used primarily for data migrations and data center
moves
There are two types of adaptive copy: Write Pending Mode and Disk Mode. Disk Mode data movement is handled at the
track level. Data from the cache is moved to the R1 volume, then keeps the track-level information as to what data is owed
to the remote side so that information can be subsequently sent one track at a time to the source device. Pending Mode data
accumulates in a local cache. A background process moves the data tasks to the source device and its corresponding target
device. The advantage of this method is that it is typically faster to read data from cache than from disk. The disadvantage
is that cache is temporarily consumed by the data until it moves to disk. Note that both methods allow write tasks to
accumulate on the local side before being copied to the remote side.
The sequence of operations is:
y An I/O write is received from the host/server into the cache of the source
y An ending status is presented to the host/server
y I/O is placed in the SRDF queue
y I/O is destaged and processed using Write Pending Mode or Disk Mode
y The I/O is transmitted to the cache of the target
y A receipt acknowledgment is provided by the target back to the cache of the source

SRDF Foundations - 15
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

Connectivity Options

SRDF
Connections & Configurations

© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 16

SRDF connectivity options are reviewed in this module.

SRDF Foundations - 16
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

Remote Link Director (RLD)

Remote Remote
Link Link
Director Director
Source Target
Remote Remote
Link Link
Director Director

y Remote Link Directors (RLDs) use a modified ESCON protocol to manage data
transfer to the remote Symmetrix
y Designated as either “RA1” or “RA2”
– Configured in pairs to eliminate single points of failure
y Communications path is either fiber optics or T1/T3, depending on distance
between sites
y Remote Link Directors may be assigned to an RA group
– Represents a logical connection between each Symmetrix and the volumes assigned to the group

© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 17

A Remote Adapter is made up of a Remote Link Adapter (RLA) and Remote Link Director (RLD).
RLAs provide the link connection and fiber optic protocol support between local and remote
Symmetrix units. An RLA is the same hardware component as a serial channel adapter board, but
RLAs use SRDF microcode.
RLDs provide communications control between the local and remote Symmetrix units. An RLD is the
same hardware component as a serial channel director (ESCON), but RLDs use SRDF microcode. This
modification was made to create an RLD for added functionality, but still allows the RLD to be used as
a normal ESCON director (if desired). One feature accomplished with the modifications is the ability
to transmit Open Systems FBA data over the ESCON-based SRDF link.
Volumes are added to SRDF groups, this represents a logical connection between each Symmetrix and
the volumes assigned to the group. One type of SRDF group is an RA group. The Remote Link
Director may be assigned to the RA group. Groups are defined at configuration time.
The ESCON protocol (developed by IBM) was designed for mainframe to peripheral communications.
It is, therefore, typically a one-way communication from mainframe to disk. Because of this design,
some additional overhead is present when implementing SRDF with ESCON because SRDF can be
implemented bi-directionally (sources and targets in both Symmetrix).

SRDF Foundations - 17
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

Remote Fibre Director (RFD)

y Uses the Fibre Channel protocol to facilitate
RFD data transfer between two Symmetrix
systems
Source y Fibre adapter cards designated only as
RFD Remote Adapter Fibre (RAF) rather than
“RA1” or “RA2” within ESCON
– Fibre Channel communication is peer-to-
peer rather than ESCON “master” and
“slave”
– Terms “Source Volume(s)” and “Target
Volume(s)” remain however
y Ability to use a free RAF processor for Fibre
Channel host connection
RFD y Two types of RAF cards are available
– Both ports support multi-mode fiber optic
Target transmission (500 m max. with 50 micron
cable)
RFD – One port supports multi-mode; one port
supports single-mode fiber optic
transmission
© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 18

A Remote Fibre Director (RFD) is an unmodified Fibre Channel Director running SRDF code.
The additional overhead described with the ESCON protocol does not exist with the Fibre Channel
standard. The Fibre Channel peer-to-peer relationship eliminates the potential 20% overhead induced
by an ESCON “peripheral” requesting permission to talk to the ESCON “host” in ESCON-based
SRDF. This increases the flexibility of SRDF in cases where it is desired to have source and target
volumes located at each side of the SRDF link.
SRDF over Fibre Channel uses a Symmetrix RF director type. Remote Fibre Director comes in various
port and processor combinations. The RF director is the same hardware as the FA host interface
director, but is configured to run SRDF emulation code. It has the ability to use a free RAF processors
for Fibre Channel host connections. RF directors use a SCSI 3 layer, running FC-4 specification Point-
to-Point protocol. Connectivity in this environment may be via direct multimode cable up to 500
meters, or single-mode cable up to 10 km between RF directors in two Symmetrix systems. Fibre
Channel switches may also be used for SRDF, with either multimode or single-mode cable segments
between units.

SRDF Foundations - 18
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

GigE Remote Directors and

Multiprotocol Channel Director (MPCD)
y Gigabit Ethernet (GigE) enables direct Symmetrix-
RE
to-IP network attachment
Source y GigE can be used in local or long-distance
RE environments, no need for protocol-conversion
devices
y Gigabit Ethernet adapter cards designated as
Remote Adapter RE
y Communication is peer-to-peer
y Two types, MPCD & GigE card
– MPCD for DMX, two-port and four-port models:
RE combination FICON/GigE or iSCSI/GigE
– MPCD has the ability to use ports for host interface
Target emulation
RE – GigE Remote Directors (RE), two-processor, two-
port, for use in Symmetrix 8000

© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 19

GigE provides Native IP support for any SRDF based product on Symmetrix systems. It is based on
GigE technology that enables direct Symmetrix system-to-IP-network attachment. This increases the
options for Symmetrix system to Symmetrix system connectivity, and allows a Symmetrix system to
connect to existing Ethernet infrastructure and directly access high-speed data transmission conduits
via IP.
Symmetrix DMX series systems provide this connectivity using the Multiprotocol Channel Director.
Unused SRDF ports can be configured with host interface emulations such as iSCSI or FICON, while
the other port is configured for SRDF emulation. Alternatively, both ports can be configured for SRDF
GigE operation.
The Symmetrix 8000 series also provides this connectivity through the GigE remote director.
The MPCD and GigE remote directors provide comparable functionality, with the exception of data
compression, which is a feature of the MPCD only.
Note that GigE ports have higher latency than Fibre Channel ports, which will impact synchronous
performance in campus configurations.
GigE, unlike Fibre Channel or ESCON, can be used in local or long-distance environments, without
the need for protocol-conversion devices. Since the Symmetrix GigE director uses native IP protocol,
traffic can be routed through an IP cloud of various connectivity devices over almost any distance.

SRDF Foundations - 19
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

SRDF Connections

Uni-Directional Symmetrix A R1 R2 Symmetrix B

Source R1 R2 Target

Bi-Directional
Symmetrix A R1 R2 Symmetrix B
Source Target R1 R2 Source Target

Dual Symmetrix A R1 R2 Symmetrix B

Configuration
Source Source R1 R2 Target Target

R1 R1
Target Target Source Source
R1 R1

© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 20

SRDF offers three types of configuration: Uni-Directional, Bidirectional, and Dual Configuration. Uni-
directional is a one way mirror relationship; Bi-directional is a two-way mirror relationship; and Dual
Configuration are two uni-directional configurations.
Bi-Directional Configuration would be supported in Campus Mode due to the overhead associated
with RA1/RA2.
Additionally, these configurations can be implemented with Escon Directors, Remote Fibre Directors,
or GigE Directors.

SRDF Foundations - 20
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

SRDF Flexible Configuration Options

Symmetrix A R1
R2 Symmetrix C
Multiple Source Source Source
Symmetrix to a R1
R2 Target Target
Single Target
Symmetrix R2
Symmetrix B R1 Target Target
R2
Source Source R1

R2 Symmetrix B
Single Source Symmetrix A R1
R2 Target Target
Symmetrix to Source Source R1
Multiple Target
Symmetrix R1
Source Source R2 Symmetrix C
R1
R2 Target Target

© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 21

SRDF connections can be configured in many different ways. The current EMC Support Matrix should
be referenced for information on the latest supported configurations.

SRDF Foundations - 21
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

SRDF Flexible Configuration, Switched SRDF

R2
Symmetrix 2
RF
Target for Symmetrix 1

SAN
R1
SAN Symmetrix 3
RF Switch
Switch RF R2 Target for Symmetrix 4
R1

Symmetrix 1
Source to Symmetrix 2
R1 Symmetrix 4
Source to Symmetrix 4
RF Source to Symmetrix 3
R2
Target for Symmetrix 1
y RF Directors no longer Source or Target
y Primary/Secondary R1/R2 relationship determines
mirroring direction

© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 22

Switched SRDF enables Symmetrix systems running SRDF to be connected through any type of SAN
of switches and simultaneously benefit from EMC’s Enterprise Storage Network.
At the same time, switched SRDF enables the Symmetrix to be configured in a “fan out” or “fan in”
configuration. Although the SRDF volume pairing is fixed; that is, the source and target volumes are
not changeable, the customer can share the bandwidth of Fibre Channel across fewer connections. This
is possible due to the implicit multiplexing that occurs in the Fibre Channel protocol.
Users benefit in that fast, optics-based infrastructures can now be used in campus and metropolitan
areas. Configurability is improved, while information access is made that much faster and easier.
More applications can now benefit from SRDF’s disaster recovery, business continuance, improved
risk management, and data replication characteristics.

SRDF Foundations - 22
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

SRDF/Consistency Groups (SRDF/CG)

Maintaining Data Consistency
Mainframe, Open System, Windows, Enterprise

Consistency
Group
UNIX

Consistency
Group
Mainframe

Consistency
y Enabling technology for disaster Group
restart solutions Windows

y Data consistency preserved

during rolling disaster
© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 23

SRDF Consistency Groups allow customers to define logical volume groups, which can be associated
with a given workload. These groups of SRDF logical volumes are automatically suspended in case of
SRDF communications failures. The remote SRDF logical volumes are consistent, even if these
logical volumes span multiple Symmetrix systems, such as when a large database has its tables on one
Symmetrix and its log files on another.
Features include the ability to do an “explicit trip” of the affected consistency group. An explicit trip
means that SRDF transfers to the R2 side are programmatically stopped – leaving the R2 volumes in a
“consistent” state as of the last transfer of information. As such, the R2 volumes can be made re-
startable in a test bed or recovery procedure. Meanwhile, production data can continue to be written to
the R1 side. The R2 side is updated later.

SRDF Foundations - 23
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

Understanding Consistency
y Dependent Write I/O Concept
– An application concept where the start of one write is dependent on
the completion of a previous write
¾ A logical dependency, not a time dependency
¾ Inherent in all Database Management Systems (DBMS)
™ Page (data) write is dependent write I/O based on a successful log write
¾ Applications can also use this technology
¾ Power failures create a dependent write consistent image
¾ Restart transforms dependent write consistent to transactional
consistent

y Creating Dependent Write Consistency

– Database technologies
¾ Hot Backup Mode, Suspend/Resume
– Enterprise TimeFinder Consistent Split
– Enterprise SRDF Consistency Groups
© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 24

Dependent write is a write I/O operation which depends on a previous I/O write to be completed before
its execution. All logging database management systems use this concept to maintain integrity. This
is required for the protection against local power outages, loss of local channel connectivity, or storage
devices. There is a logical dependency between I/Os built into database management systems, certain
applications, middleware tools such as MQ Series, and operating systems.

SRDF Foundations - 24
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

Rolling Disasters

HOST Data
R1(F) ahead 4
R1(E) of Log
3 R1(D)
R1(C)
R2(F)
3
R2(E)
DBMS R2(D)
R2(C)
R2(B)
1
R2(A)
R1(B)
2 R1(A)
1. Rolling disaster begins
A = Logs
C = DBMS Data 2. Log write
E = Application Data 3. Dependent data write
4. Inconsistent data

© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 25

It is important to understand rolling disasters in order to understand how EMC technology prevents
them. On this diagram, there is a host, two source Symmetrix systems, and one target Symmetrix. A
Database Management System is running on the host. The logs of the DBMS are on the bottom source
Symmetrix and the table space of the DBMS is on the top source Symmetrix.
The SRDF link from the bottom source Symmetrix and the target Symmetrix breaks. This is the
beginning of the rolling disaster. However, the table space updates are propagated to the target
Symmetrix, even though log updates are not.
Now the rolling disaster completes. When the restart of the DBMS occurs on the target side, it restarts
fine. There are no errors, but data is corrupt. There is information in the table spaces that is not
reflected in the log. This is known as ‘data ahead of log’. If the table spaces and logs were reversed, it
would be referred to as a ‘log ahead of data’. The only way to detect that the data is corrupt is to be
extremely familiar with the transactions and know what was processing at the exact time the rolling
disaster occurred. This is almost impossible in most environments.

SRDF Foundations - 25
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

Rolling Disaster Protection: Enterprise SRDF/

Consistency Groups 4 5
Suspend
R1/R2
E-ConGroup Relationship
HOST: MF or OS
Definition DBMS 7
R1(F)
(A,C,E) 1 Restartable
MF: ConGroup Task R1(E) Copy
OS: Solutions Enabler
6 R1(D)
with ConGroup R1(C)
definition R2(F)
R2(E)
DBMS
R2(D)
R2(C)
MF: ISO R2(B)
2
OS: PowerPath R2(A)
R1(B)
3 R1(A) 1. ConGroup protection
2. Rolling disaster begins
A = Logs 3. Log write
C = DBMS Data 4. ConGroup “trip”
E = Application Data 5. Suspend R1/R2
6. Dependent data write
7. Dependent write consistent
© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 26

To prevent corruption of data, EMC created Consistency Groups to use with SRDF. ConGroups is a
started task on the mainframe and is packaged in EMC Consistency Group for z/OS. On Open
Systems, ConGroups work through the SYMCLI/SYMAPI and an RDF process daemon or SYMCLI /
SYMAPI and PowerPath. A ConGroup contains a set of volumes, created by the customer, and needs
consistency maintained. In this example, our host could be Mainframe or Open System and includes
volumes “A”, “C” and “E”. PowerPath is used on Open Systems, and EMC Consistency Group for
z/OS is used on the mainframe.
With the ConGroup defined, our rolling disaster begins with the loss of the SRDF links from bottom
source Symmetrix to the target Symmetrix. A sense code is sent back to the host stating that the data
from volume “A” could not be propagated to its target side. The ConGroup started task on the
mainframe, or the ConGroup definition in Open Systems, detects the sense code and works with IOS
Mainframe or PowerPath for Open Systems to hold the I/O. The DBMS, or application, is not aware
the I/O was held and created a Dependent Write Consistent copy, or DBMS restartable copy of the data
on the target side. We have simulated a local power failure at the target side at the point of the
beginning of the rolling disaster. After complete failure at the source side, the target side can be
restarted and the DBMS can be restarted. This provides transactional consistency.

SRDF Foundations - 26
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

SRDF/Asynchronous (SRDF/A)
y High Performance Remote Replication for
the DMX
y Asynchronous remote mirroring
– Recoverable point-in-time copy
– No impact to production applications
– No distance limitations
y Operational savings through reduced
bandwidth
– Delta Sets vs. ordered writes
y Supports Mainframe and Open Systems
y Complements existing SRDF solutions
– Meet a wide range of RPO and RTO
service level requirements
y Mixed SRDF, SRDF-DM and SRDF/A
– Share links and directors Highest Performance Asynchronous Replication
In The Industry

© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 27

SRDF/Asynchronous (SRDF/A) is a remote mirroring solution for the Symmetrix DMX family. Its
unique architecture delivers a remote mirroring solution that has no impact on production applications
and no distance limitations. This unique architecture enables significant operational savings through
reduced bandwidth requirements; you can size your bandwidth to the average peak workload, versus
the absolute peak workload.
SRDF/A is a single solution supporting both Mainframe and Open Systems attach. It also
complements SRDF solutions to meet mixed service level requirements. In fact, it can also share the
same communication links as SRDF.
The unique architecture uses cache-based “Delta Sets”, which are detailed next.

SRDF Foundations - 27
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

Terms & Concepts

y Delta Set
– Collection of incoming writes maintained in cache
¾ Capture (Source, aka the N cycle)
¾ Transmit (Source, aka the N-1cycle)
¾ Receive (Target, aka the N-1cycle)
¾ Apply (Target, aka the N-2 cycle)

y Write Folding
– Consolidates the writes in cache so that the Capture Delta Set can
be sent across the SRDF/A link only once
– This reduces overhead of transmission, increasing band width

© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 28

A Delta Set is the innovative architecture SRDF/A is built upon. A Delta Set is simply a collection of
writes that are part of an SRDF/A process.
There are four types of Delta Sets which are described in more detail later in this training. The Delta
Set type names are: Capture, Transmit, Receive, and Apply. In some of our literature, you may see the
term “cycles,” as in cycle N, cycle N-1, etc., as shown on the slide.
Applications tend to write data to the same location on a device. When this occurs, it is called locality
of reference phenomenon. A process called Write Folding identifies the locality of reference
phenomenon. Write Folding consolidates the writes in cache so data that is updated multiple times in
the same Capture Delta Set is sent across the SRDF/A link only once. This reduces overhead of the
transmission and increases the bandwidth.

SRDF Foundations - 28
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

SRDF/Asynchronous Operation
R1 R2
Capture
Transmit CAPTURE RECEIVE
Repeat Collects Receives writes
Receive application write from Transmit
Apply I/O Delta Set

TRANSMIT APPLY
Sends final set of Once Receive is
writes to target complete, data is
applied to disk

Source Target
SRDF/A performs “Write Folding” - only sends Transmits
of the very last writes from the Capture Delta Set
© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 29

Let’s take a very high level look at SRDF/A’s innovative architecture. SRDF/A uses “Delta Sets” to
maintain a group of writes over a short period of time. Delta Sets are the enabler of all the efficiencies
that SRDF/A can deliver. There are four types of Delta Sets to manage the data flow process.
SRDF/A’s data flow can be summarized in simple steps.
Source side Delta Sets:
y Capture: Captures, in cache, all incoming writes to the source volumes involved in the SRDF/A
group. Upon completion of the set, the Capture Delta Set is “write folded” and promoted to a
Transmit Delta Set. A new, separate Capture Delta Set is then created to maintain the next Delta
Set of writes.
y Transmit : Transfers its contents (only the last set of writes) from the source to the target system.
Target side Delta Sets:
y Receive: on the target system and receives the data being transferred by the source-side Transmit
Delta Set. Once received in its entirety, it is promoted to the Apply Delta Set.
y Apply: Applies the Delta Set’s writes to the target volume to create the consistent, recoverable
remote copy. This finishes the Delta Set cycle.
A consistent, recoverable copy is maintained at the remote location with each application of a Delta
Set. Once the initial Delta Set cycle completes, it is simply repeated, Delta Set after Delta Set,
providing a continuous checkpoint of Delta Sets. SRDF/A provides a solution for service level
requirements that need data on the R2 side within the seconds to minutes range.

SRDF Foundations - 29
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

SRDF/A Multi Session Consistency

y Ensure consistency across multiple SRDF/A
sessions
– Hosts are writing to the capture cycle on R1
R1 R2
– SRDF/A is transferring the transmit cycle from R1 to R2
– R2 is receiving data in the receive cycle and restoring CAP REC
apply cycle
– When R1 finishes sending transmit cycle, both sides
mark it as complete TRAN App
– When R2 finishes restoring apply cycle it notifies R1

R1 R2 Source Target
R1 R2
CAP REC MSC
or CAP REC
SCF (mf)
TRAN App

TRAN App

Source Target Host

Source Target
© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 30

Achieving data consistency across multiple SRDF/A groups simply requires that the cycle process
described on this slide be coordinated among the participating Symmetrix. SRDF/A Multi Session
Consistency coordinates the cycle of Capture, Transmit, Receive, and Apply across the multiple
Symmetrix R1 source and R2 target relationships.
When the R1 Symmetrix finishes sending the inactive cycle to the R2 Symmetrix, the R1 Symmetrix
also sends an internal message to the R2 side Symmetrix that the transfer is complete. In a similar
fashion, when the R2 side Symmetrix finishes restore of the active cycle, it sends a message to the R1
Symmetrix that the restore is complete. The host receives status messages that indicate cycle
processing readiness; the host then coordinates the execution of the cycles by sending a cycle switch
command to all R1 side Symmetrix.
The single host coordination point drives the cycle process in all participating Symmetrix systems. In a
mainframe, this is provided by ResourcePak Base for z/OS executing Symmetrix Control Facility,
SCF. In an open systems environment, cycle switching is provided by Solutions Enabler, executing
MSC within a RDF process daemon.

SRDF Foundations - 30
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

SRDF/Asynchronous Considerations
y Requires additional cache to support Delta Sets
– Two Delta Sets maintained at each location
¾ ~.75 GB additional cache per 1 TB of data being remotely mirrored per
DMX as a minimum, assuming bandwidth sized to peak write load
¾ Much more cache could be required if bandwidth sized closer to average
write I/O load on busy DMXs

y Bandwidth must be sized relative to cache for maximum

performance/minimum Delta Set size
y Delta Set Extension Feature
– Provides additional Delta set space in a DMX disk

© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 31

Some key considerations for SRDF/Asynchronous include:

y Each DMX system will also require additional cache to support the cache-resident Delta Sets. A
general rule of thumb is ~.75 GB of additional cache per 1 TB of SRDF/A protected data.
y The balancing of cache to bandwidth for maximum performance is also critical.
The Delta Set Extension Feature:
y Provides additional Delta set space in a DMX disk, the disk space is allocated for through software,
a portion of SAVE devices is used for this purpose.

SRDF Foundations - 31
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

Campus Implementations
y The campus solution enables units to be up to 66 km apart
– ESCON Direct Attach
¾ Distance of up to 3 km
¾ Throughput up to 14 MB/s
– Direct Fibre Channel Attach (Point-to-Point)
¾ Distance of up to 10 km
¾ Throughput up to 55 MB/s
– Switched Fibre Channel Attach
¾ Distance of up to 10 km
– GigE
¾ Enables direct Symmetrix-to-IP network attachment
¾ Allows Symmetrix to access existing Ethernet infrastructure

Symmetrix Symmetrix

ISL

Fibre
Fibre Channel
Channel Fibre
Fibre Channel
Channel

R2 target R1 source Connectrix Connectrix

volumes volume R2 target R1 source
volume volumes

© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 32

There are several SRDF Campus implementations available that enable units to be up to 66 km apart:
y ESCON Direct Attach: Provides connectivity between two Symmetrix unit ports at a distance of up
to 3 km for each cable segment. Throughput for each connection is typically rated at up to 14
megabytes per second (MB/s) maximum, and in most cases is relative to block size of the host
write I/O operation.
y Direct Fibre Channel Attach (Point-to-Point): Provides connection via direct cable segment
attached between two Symmetrix unit ports at a distance of up to 10 km for each cable segment
(single mode hardware), or 500m for each cable segment (multimode hardware, default).
Throughput for each connection is typically rated at up to 55 MB/s maximum, and in most cases is
relative to the block size of the host write I/O operation.
y Switched Fibre Channel Attach: Provides connections between two Symmetrix unit ports by way
of Fibre Channel switch ports, at a distance of up to 10 km for each cable segment, or 500m for
each cable segment (multimode hardware, default).
y GigE Attach:Provides native IP support for any SRDF-based product on Symmetrix systems,
which is based on Gigabit Ethernet (GigE) technology. This enables direct Symmetrix-to-IP
network attachment. This increases the options for Symmetrix-to-Symmetrix connectivity, and
allows a Symmetrix system to connect to an existing Ethernet infrastructure, and to directly access
high-speed data transmission conduits via IP.

SRDF Foundations - 32
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

Metropolitan Area Network Implementations

y The MAN solution enables units to be up to 200 km apart
y Three Optical Link Configurations
– ESCON with Repeaters
– ESCON via DWDM
– Switched Fibre Channel via DWDM
Connectrix Connectrix

DWDM
Symmetrix Dark Fiber Symmetrix

Fiber Fiber
Fiber Fiber
Nortel Nortel
Optera Optera
Metro 5200 Metro 5200
R1 source R2 target
volume volume

© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 33

The use of a Metropolitan Area Network (MAN) is determined by the ability of the customer or carrier
to install/provision “dark fiber” cable that allows the customer to use the entire bandwidth on that
fiber.
MANs are useful for campuses with multiple data centers, ISPs that have several sites along a fiber
optic network right-of-way, and XSPs that need to reach multiple customers in the area. This means
that the maximum bandwidth is available and that the delay or latency is low and controllable. In the
future, optical technology will expand to reach beyond the 200 km distance.

SRDF Foundations - 33
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

Extended Distance Implementations

y The Extended Distance, wide area network solution, enables units
to be at trans-oceanic or trans-continental distances
y Communication options include:
– T1/T3, E1/E3
– Synchronous Optical Networks (SONET)
– ATM-OC3
– IP
IP Network

Router
Symmetrix Router
Symmetrix

100 Mb
Ethernet 100 Mb
Ethernet

SRDF SRDF
w/ FarPoint w/ FarPoint
R1 source CNT USD or INR R2 target
CNT USD or INR
volume 9801 Chl Ext. volume
9801 Chl Ext.

© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 34

The Extended Distance Wide Area Network (WAN) solution enables units to be at trans-oceanic or
trans-continental distances.
Typically, ATM, T3/E3,T1/E1, or Internet Protocol (IP) are offered by the carrier.
Communication options include: T1/T3, E1/E3; Synchronous Optical Networks (SONET); ATM-OC3;
and IP.

SRDF Foundations - 34
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

Ideal SRDF Business Continuance Solution

y Determined by:
– Recovery Point Objective
– Recovery Time Objective
– How well your applications tolerates network latency

© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 35

When picking the ideal business continuance solution, the selection is determined by the distances
separating your data centers, the Recovery Point Objective and Recovery Time Objective, and how
well your applications tolerate network latency. Shorter distances reduce network latency, which
allows using synchronous disk replication and data center mirroring. Business continuance solutions
can be divided into three general categories, based on the distances between the local and remote data
center (as illustrated in this slide):
y Campus solution is a limited subset of metro connectivity that uses fiber-optic cabling to transmit
data over short distances using Symmetrix and Connectrix direct-attach capabilities. Typically the
distance is smaller then 66 km
y Metropolitan Area Network (MAN) provides SRDF connectivity for distances typically less than
100 km, but up to 200 km or more. This environment is characterized by extremely low error rates
and high bandwidth of fiber configurations, so error recovery and compression are less significant
in choosing SRDF connectivity options.
y Extended Distance Wide Area Network (WAN) provides SRDF connectivity over long distances
using telecommunications networks, such as IP, SONET, or ATM. WANs are differentiated from
MANs as being non-lossless and limited bandwidth environments, where error-recovery, data
buffering, and compression capabilities of connectivity.

SRDF Foundations - 35
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

SRDF Data Mobility (SRDF-DM)

y Economical ‘Data Replication’ solution
not intended to be used for Disaster
Recovery operations
y SRDF Data Mobility is just SRDF
running in Adaptive Copy Disk Mode or
Adaptive Copy Write Pending Mode
– Adaptive Copy Disk Mode is
recommended for optimal
performance.
y The Primary Mode (Synchronous or
Semi-Synchronous) is still set on the
Symmetrix Logical Volume, but the
user cannot disable the Secondary
Mode SRDF
y Skew limit cannot be set to below 100 links
tracks per Symmetrix Logical Volume
y Minimum skew limit is set when the
Source Target
microcode is installed

© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 36

SRDF Data Mobility is an economical ‘data replication’ solution and is not intended for use as a
disaster recovery solution. Unlike host level replication technologies, it does not require additional
servers or network capability to perform unlimited distance data replication.
SRDF-DM uses the same user controls as SRDF, and a differences is that: SRDF-DM uses the skew
parameter; the skew parameter defines the number of tracks the source volume can lead the target
volume. This limit cannot be set by the user to a value lower than 100 for SRDF-DM. Another
difference is that the user cannot disable the Secondary mode by setting a skew limit of zero.

SRDF Foundations - 36
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

Concurrent SRDF
y One R1 can be paired with two R2 devices, concurrently
Secondary

M1 M2 M3 M4

Synchronous
Source Target “A”
M1 M2 M3 M4
Target “B”
Adaptive Copy
M1 M2 M3 M4

Primary

Secondary
© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 37

Concurrent SRDF allows two remote SRDF mirrors of a single R1 device. For example, the use of one
remote copy for disaster recovery, and another for decision support or backup.
The example here shows a concurrent SRDF configuration in which the primary volume is
communicating with one secondary volume in synchronous mode. Concurrently, the same primary
volume is communicating with its other secondary volume in one of the adaptive copy modes
(adaptive copy write-pending mode or adaptive copy disk mode).
Each Remote Link Director is assigned to an RA Group. With ESCON, only one RA group per RLD is
allowed, but with Fibre Channel SRDF, the RA Groups can be defined to the same RLD.
Any combination of synchronous/semi-synchronous and adaptive copy is allowed with the following
exceptions: One volume operating in synchronous mode and the other operating in semi-synchronous
mode; and both volumes operating in asynchronous mode.
A write I/O from the host at the primary device side cannot be returned as completed until both remote
Symmetrix systems signal the local Symmetrix that the SRDF I/O is in cache at the remote side.

SRDF Foundations - 37
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

SRDF/STAR Symmetrix Triangular Asynchronous

Replication
y One source, two targets
– One target in Synchronous mode
¾ Short distance, zero data lag
– One target in Asynchronous mode
¾ Longer distance, variable data lag, no performance impact

y SRDF/STAR: Remote Re-synchronization

– Continued protection upon source failure

Incremental Resync
SRDF/S SRDF/S
(sync) to
SRDF / A
SRDF/A
(async)

© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 38

Building on the foundation of a near and a far site using SRDF/S and SRDF/A respectively, the
remaining problem is that the far and near sites contain different data (the near site is a zero-data-loss
situation, and the far site is seconds to minutes behind).
To resolve these problems, a new product, SRDF/Star, has been created. SRDF/Star is a solution that
provides a third copy of System Data in case one site was lost and not recoverable. SRDF/Star
completes the connection to the two non-primary sites (the bunker and the long-distance site) with an
SRDF/A link. If the primary site is lost to a catastrophic event, business may resume/continue using
data from the bunker or long-distance site. SRDF/CG Consistency Group and SRDF/A Multi Session
Consistency are used to integrate the SRDF/Star configuration.
SRDF Family replicated data can be used from whichever remaining Symmetrix (near or far site) is the
most current. Realizing that determining which site has the better data for a restart is very difficult,
EMC provided a utility to aid in the determination of the status of data in the two remaining sites. This
allows you to make an incremental synchronization decision of which direction to synchronize in prior
to a restart. This link between the near and far with SRDF/Star maintains continuous protection for the
remaining sites.

SRDF Foundations - 38
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

Dynamic SRDF
y Enables user to dynamically define
relationships between R1 and R2 volumes
Symmetrix A
y Dynamic SRDF groups provides flexibility
to tailor SRDF configuration for changing R2
application requirements RF

Symmetrix B

R1 SAN
SAN
RF RF R2
Switch
Switch
R1 1 2
1 2 R2
1

Symmetrix C
3
R1
1. Grey R1 established with R2 in Symmetrix B RF 3
2. Grey R1 relationship with R2 in Symmetrix B broken R2
3. Grey R1 established with R2 in Symmetrix C
© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 39

Prior to Dynamic SRDF, the R1 and R2 pairings were static and defined in the BIN during SDRF
configuration. Any changes to SRDF device pairing required a new BIN file to be defined and loaded
into the Source and Target Symmetrix.
Dynamic SRDF, available since 5x68 code, provides the capability to change device pairings as
needed, without requiring a BIN file configuration change to be performed by EMC. R1 to R2 pairing
will be dynamic and can be changed by the user, much like Standard to BCV relationships.
In this slide, an R1 can have a remote mirror in Symmetrix B and then later be defined to mirror the
data to Symmetrix C instead.
Another scenario, after breaking the R2 in Symmetrix B from its original R1, the R2 can be reused as
the target for an R1 in Symmetrix C.
In conjunction with Symmetrix Manager, this provides the user with maximum flexibility to make ad-
hoc changes to the remote mirroring protection for a volume.
In DMX a user can create Dynamic SRDF Groups and associate dynamic SRDF pairs to the group.
Dynamic SRDF groups provide flexibility within the SRDF environment to change multiple remote
mirroring connections for dynamic devices.

SRDF Foundations - 39
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

SRDF Integration with TimeFinder

Integrating SRDF & TimeFinder Operations

with SRDF/Automated Replication

© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 40

In this module, the process to integrate SRDF with TimeFinder is reviewed.

SRDF Foundations - 40
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

SRDF/Automated Replication (SRDF/AR)

SRDF/S & TimeFinder/Mirror
y Masks performance impact
of distance ECC/OE
ECC/OE Metro East Metro West

Server not
y Minimizes network costs required at
intermediate
node

y Allows business restart site

to be at any distance away Synchronous
from source
Adaptive Copy
San Diego

Extended Distance No performance Impact Rapid Business Resumption

Zero Data Loss Minimize communications cost

© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 41

SRDF/AR pauses transaction updates to a database, spanning multiple volumes or systems, while a
point-in-time copy of the database is being made. This ensures that the database contents aren’t altered
during the copy operation so that the copy is recoverable and re-bootable. SRDF/AR provides a
solution for service level requirements that need data on the R2 side within the minutes to hours range.
SRDF/AR and the automated Multi-Hop command eliminates the multiple, separate SRDF and
TimeFinder commands previously required to implement these capabilities, when establishing BCVs
for remote replication. The same command can be used for Consistent Split and Multi-Hop. In
particular, the symreplicate command benefits Multi-Hop users greatly because it now makes a Multi-
Hop configuration very easy to implement.

SRDF Foundations - 41
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

SRDF/Automated Replication
y The continuous movement of dependent write consistent
data to a remote site in an asynchronous mode
– Standard to BCV/R1 => R2 to BCV Single Hop
– R1 => R2 to BCV/R1 => R2 to BCV Multi-Hop

y Uses Enterprise TimeFinder Consistent Split and SRDF

Adaptive Copy Mode
y Parameters available to manipulate cycles
– Mainframe
¾ EMC Symmetrix Control Facility
– UNIX/Windows
¾ SYMCLI

© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 42

SRDF/AR enables customers to deploy a long distance remote mirroring solution to optimize their
business critical information processes and resources in a cost-effective way.

SRDF Foundations - 42
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

SRDF/Automated Replication Single Hop Configuration

Source Site (A) Target Site (B)

and/or and/or
“Production Site” “Restart Site”

Remote Disk Mirroring

SRDF Links IP, ATM, T3
adaptive copy

STD R1/BCV R2 BCV

1. Consistent split on source
STD
2. SRDF mirroring resumed R2
R1/BCV BCV
3. Incremental establish both
4. BCV split on target
Begin Established

© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 43

This environment is comprised of a source (production) site and target (restart) site. The attached
hosts can be mainframe, UNIX, and/or Windows.
The source site shows the standard volumes established with the BCV/R1 devices, which naturally
suspends the BCV/R1 to R2 synchronization. The target site shows that the R2s are split from their
BCVs.
The standard volumes at the source site represent the dependent write consistent copy.
The standard devices depict that logical consistency is maintained by utilizing TimeFinder Consistent
Split.
The dependent write consistent copy could exist at any of these points throughout the cycle.

SRDF Foundations - 43
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

SRDF/Automated Replication Multi-Hop Configuration

Host: MF & OS
MF: SCF
OS: SYMAPI/ Source Site (A)
SYMCLI “Production Site”
Target Site (B) MF;UNIX;NT
DBMS “Recovery Site”
MF: IOS
OS: PowerPath
ConGroup
Definition

R1 R1
Bunker Site

Restartable
Adaptive Copy

DBMS

Copy
R2 B1 R2 BCV
Synchronous SRDF R2 BCV
R2 B1 Extended Distance

BCV Consistent Split

© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 44

This slide represents a multi-hop configuration, which includes a source site, bunker site, and target
site. In a multi-hop configuration, there is synchronous propagation from the source site to the bunker
site, and adaptive copy propagation from the bunker site to the target site. There is usually an SRDF
Consistency Group defined to maintain consistency, in the event of a rolling disaster between the
source and bunker sites.
The function of the bunker site is to transform synchronous propagation to an adaptive copy
propagation of dependent write consistent copies. The distance from the bunker site to the target site is
usually a long distance and uses the adaptive copy mode of SRDF. The propagation between the
bunker site and target site was once a manual process and done without maintaining consistency.

SRDF Foundations - 44
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

SRDF Cluster Integration

y SRDF/Cluster Enabler for Microsoft Cluster Server
– Available from EMC
y SRDF/Cluster Enabler for VERITAS Cluster Server
– Available from EMC
y SRDF/Automated Availability Manager for EMC
SRDF
– EMC FullTime AutoStart
y EMC AutoSwap, redirects IO to the R2 devices in the
event that the R1 devices are unreachable
y HP MetroCluster and Continental Cluster
– Available from HP
y IBM HACMP/HAGEO
– Available from IBM - requires user scripting

© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 45

SRDF is increasingly being used in cluster environments. In a cluster, the server and application
providers have usually worked out the issues of failing over from one server to another at the
application level. This means they can restart the work on the surviving node in the cluster. However,
if the failed node contained the disk storage and working data being used by the application, the
problem would be the surviving node could restart the applications, but only with new data.
SRDF’s value in this scenario is in the mirrored data from one Symmetrix to another within the cluster.
Using SRDF, the surviving node also has all the data because it was previously in an SRDF
relationship. Procedures to make the R2 volumes read/write enabled have to be taken. In some cases
this task is automated, as with SRDF/Cluster Enabler for MSCS, the Microsoft Cluster Server.
When the time comes to bring the failed node back online, SRDF provides the ability to fail back in
the other direction and be ready for work very quickly, with up-to-the-second live data.
EMC AutoSwap works in a mainframe Parallel Sysplex environment with SRDF/S and SRDF/CG.
AutoSwap automatically makes the R2 devices R/W and re-direct I/O to the R2 devices in the event
that the R1 devices are unreachable by the local and remote nodes.

SRDF Foundations - 45
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

Management Options

SRDF Management &

Business Justification

© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 46

SRDF management options and business justifications are reviewed in this module.

SRDF Foundations - 46
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

SRDF Management Software - 1

y SRDF Manager within EMC ControlCenter
– Easy, “point-and-click” access
– Excellent for ad hoc SRDF operations

© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 47

SRDF can be accessed via EMC ControlCenter. Through the graphical user interface (GUI) within
EMC ControlCenter software, related devices are grouped together in device groups. SRDF operations
may be performed on all devices in a device group using a single command, and group information is
maintained in the SYMAPI Database.

SRDF Foundations - 47
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

SRDF Management Software - 2

y Symmetrix Management Console for device and replication
management
– Intuitive, browser-based graphical interface for managing Symmetrix
systems
– Manages most SRDF operations
y Mainframe Symmetrix Control Facility for z/OS and SRDF Host
Component for z/OS started tasks.
– Operator Commands for single commands
– ISPF Panel menus
– Batch JCL for automation with multiple commands
y Solutions Enabler SRDF Feature for Open Systems
– Based on SYMAPI
– Consistent command syntax
– Functions implemented using binary commands
– Used for scripting on Open Systems

© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 48

Symmetrix Management Console for device and replication management:

This very intuitive, browser-based graphical interface for managing Symmetrix systems Manages
many different SRDF functions, it is a user friendly EMC solutions.
Open Systems Solutions Enabler:
Allows access through the command line interface (CLI) and through Open Systems scripting.
Mainframe Symmetrix Control Facility for z/OS:
SRDF Host Component for z/OS started tasks allows operator commands for single commands, ISPF
Panel menus, and Batch JCL for automation with multiple commands.
Open Systems Solutions Enabler allows access through the command line interface (CLI) and through
Open Systems scripting.

SRDF Foundations - 48
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

Decision Drivers Considerations

Business Technical
Considerations Considerations
Recovery and
Cost consistency
Functionality
Capacity
Availability

RTO Bandwidth

RPO Performance

PRIMARY DECISION DRIVERS

© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 49

The key to determine which remote replication solution is best for a particular application is a
thorough understanding service levels. Specifically, you need to balance your requirements for
performance, functionality, availability, and economics against capacity needs, bandwidth
requirements, and overall Total Cost of Ownership. Every solution has benefits and risks, and all of
these factors must be balanced in order to decide the best solution for a given problem.

SRDF Foundations - 49
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

Remote Replication Considerations

Source Target

Response
time Infrastructure Comms Application
consistency

y Response time: Application impact of replication

y Infrastructure: Costs to enable solution
y Communications: Monthly line and bandwidth costs
y Application consistency: Application inter-
dependency
© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 50

Today’s demands require customers to invest in more complex data infrastructures. These data
processing environments require a focused attention to such components as application response times,
application consistency, and the costs associated with the infrastructure and communications.

SRDF Foundations - 50
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

Business Continuance Options: Remote Replication

Service Level Limitations &
Solutions
Requirements Requirements
Product / Max. Link
RPO RTO RGO Mode
Solution Distance Bandwidth

No data Multi- Synchronous and

Medium SRDF/STAR Unlimited Highest
loss location Asynchronous

SRDF/S, Synchronous mode

No data Short
Fastest ConGroup & with Continuous ~200Km High
loss distance
AutoSwap Availability

Up to 200Km High
SRDF/AR Synchronous mode
No data Long for synch.
Fast + Consistent Point
loss* distance Multi Hop Unlimited for
in Time Copy
adcopy. Low

Sec / Long SRDF/A Continuous Medium or

Fast unlimited
Min distance with MSC Asynchronous Copy High

Min / Long SRDF/AR Consistent Point in

Fast unlimited Low
Hours distance Single Hop time copy

© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 51

If an outage occurs on Source Symmetrix, when it comes to business continuity and remote mirroring,
the time it takes to get the business running again - Recovery Time Objective (RTO) - is always “fast”,
no matter which EMC replication solution is deployed.
EMC has the right solution for each particular requirement. If the customer can’t tolerate any data
exposure, we have the industry leader for synchronous mirroring - SRDF. However, as with any
synchronous solution, there are characteristics that must be understood – distance is limited by
application time-outs and speed of light issues and bandwidth must be sized for peak workload at all
times.
We can also deliver solutions that combine SRDF with TimeFinder to create single-hop and multi-hop
environments for specialized needs. These solutions offer different RPOs and have different
requirements for bandwidth, supported distances, etc.
SRDF/A provides a solution for service level requirements that need RPOs in the seconds to minutes
range. SRDF/AR provides a solution for service level requirements that need RPOs in the minutes to
hours range.
y RPO = Recovery point Objective, how much data can be lost
y RTO = Recovery Time Objective, How quickly can data be recovered
y RGO = Recovery Geographic Objective

SRDF Foundations - 51
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

SRDF Benefits Review

y Protect against local and regional site disruptions
– Continuous data availability
– Multiple remote recovery sites
– Meet regulatory requirements

y Migrate, consolidate or distribute data across storage

platforms
– Data center consolidations
– Technology refreshes

y Enable non-stop operations

– Application restart across volumes

© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 52

The right remote replication solution can limit the exposure to planned and unplanned downtime,
enabling non-stop operations. Or perhaps you need to provide your organization with efficient data
replication to meet corporate or government standards, while still meeting your Total Cost of
Ownership requirements. Additionally, you need a flexible solution that changes as your needs change.
No matter what the challenge is, there is one underlying theme: data protection and faster business
restart, in the event of a disaster or unplanned outage, is critical across the organization.
There are several methods of remote replication available. Each has a different profile to help you
manage and conquer your current business challenges, while enabling new processes and procedures
that help gain a significant competitive advantage. This is something all businesses strive for in today’s
competitive marketplace.

SRDF Foundations - 52
 Copyright © 2007 EMC Corporation. Do not Copy - All Rights Reserved.

Course Summary
Key points covered in this course:
y Concepts, value, and environmental aspects of using
SRDF
y SRDF features and their differences
y SRDF connectivity features
y SRDF Management Software offerings
y Business benefits of SRDF

© 2007 EMC Corporation. All rights reserved. SRDF Foundations - 53

These are the key points covered in this training. Please take a moment to review them.
This concludes the training. Please proceed to the Course Completion slide to take the assessment.

SRDF Foundations - 53