RecoverPoint Implementation Student Guide
-
Upload
soringheorghe20 -
Category
Documents
-
view
505 -
download
43
Transcript of RecoverPoint Implementation Student Guide
-
8/18/2019 RecoverPoint Implementation Student Guide
1/367
RecoverPoint ImplementationStudent Guide
Education Services July 2013
-
8/18/2019 RecoverPoint Implementation Student Guide
2/367
-
8/18/2019 RecoverPoint Implementation Student Guide
3/367 pyright © 2013 EMC Corporation. All rights reserved
Welcome to RecoverPoint Implementation.
Copyright © 1996, 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011,2012, 2013 EMC Corporation. All RightsReserved. EMC believes the information in this publication is accurate as of its publication date. The information is subject to change
without notice.
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 OFMERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.
Use, copying, and distribution of any EMC software described in this publication requires an applicable software license.
EMC2, EMC, Data Domain, RSA, EMC Centera, EMC ControlCenter, EMC LifeLine, EMC OnCourse, EMC Proven, EMC Snap, EMC SourceOne,EMC Storage Administrator, Acartus, Access Logix, AdvantEdge, AlphaStor, ApplicationXtender, ArchiveXtender, Atmos, Authentica,Authentic Problems, Automated Resource Manager, AutoStart, AutoSwap, AVALONidm, Avamar, Captiva, Catalog Solution, C-Clip, Celerra,Celerra Replicator, Centera, CenterStage, CentraStar, ClaimPack, ClaimsEditor, CLARiiON, ClientPak, Codebook Correlation Technology,Common Information Model, Configuration Intelligence, Configuresoft, Connectrix, CopyCross, CopyPoint, Dantz, DatabaseXtender, DirectMatrix Architecture, DiskXtender, DiskXtender 2000, Document Sciences, Documentum, elnput, E-Lab, EmailXaminer, EmailXtender,Enginuity, eRoom, Event Explorer, FarPoint, FirstPass, FLARE, FormWare, Geosynchrony, Global File Virtualization, Graphic Visualization,Greenplum, HighRoad, HomeBase, InfoMover, Infoscape, Infra, InputAccel, InputAccel Express, Invista, Ionix, ISIS, Max Retriever,MediaStor, MirrorView, Navisphere, NetWorker, nLayers, OnAlert, OpenScale, PixTools, Powerlink, PowerPath, PowerSnap, QuickScan,Rainfinity, RepliCare, RepliStor, ResourcePak, Retrospect, RSA, the RSA logo, SafeLine, SAN Advisor, SAN Copy, SAN Manager, Smarts,
SnapImage, SnapSure, SnapView, SRDF, StorageScope, SupportMate, SymmAPI, SymmEnabler, Symmetrix, Symmetrix DMX, SymmetrixVMAX, TimeFinder, UltraFlex, UltraPoint, UltraScale, Unisphere, VMAX, Vblock, Viewlets, Virtual Matrix, Virtual Matrix Architecture, VirtualProvisioning, VisualSAN, VisualSRM, Voyence, VPLEX, VSAM-Assist, WebXtender, xPression, xPresso, YottaYotta, the EMC logo, and whereinformation lives, are registered trademarks or trademarks of EMC Corporation in the United States and other countries.
All other trademarks used herein are the property of their respective owners.
© Copyright 2013 EMC Corporation. All rights reserved. Published in the USA.
Revision Date: 07/25/2013
Revision Number: MR-7CP-RPSYSDEP.7722.4.0
Course Introduction: RecoverPoint System Deployment
-
8/18/2019 RecoverPoint Implementation Student Guide
4/367 pyright © 2013 EMC Corporation. All rights reserved
This course covers RecoverPoint Installation and Operations.
Course Introduction: RecoverPoint System Deployment
-
8/18/2019 RecoverPoint Implementation Student Guide
5/367 pyright © 2013 EMC Corporation. All rights reserved Course Introduction: RecoverPoint System Deployment
-
8/18/2019 RecoverPoint Implementation Student Guide
6/367 pyright © 2013 EMC Corporation. All rights reserved Course Introduction: RecoverPoint System Deployment
-
8/18/2019 RecoverPoint Implementation Student Guide
7/367 pyright © 2013 EMC Corporation. All rights reserved Course Introduction: RecoverPoint System Deployment
-
8/18/2019 RecoverPoint Implementation Student Guide
8/367 pyright © 2013 EMC Corporation. All rights reserved Course Introduction: RecoverPoint System Deployment
-
8/18/2019 RecoverPoint Implementation Student Guide
9/367 pyright © 2013 EMC Corporation. All rights reserved
This module focuses on the different RecoverPoint configurations and an overview of the
multiple installation and configuration procedures when deploying a RecoverPoint solution.
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
10/367 pyright © 2013 EMC Corporation. All rights reserved
This lesson covers RecoverPoint solutions.
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
11/367 pyright © 2013 EMC Corporation. All rights reserved
EMC RecoverPoint is an out-of-band solution which offers advanced capabilities to deliverenterprise-level data protection, replication, and disaster recovery. This network-basedblock-level solution allows customers to protect data locally for operational and applicationrecovery needs. It also protects the same data remotely for disaster recovery purposes.
RecoverPoint maintains write-order consistency of the replicated data across all volumeswithout impacting the production data flow between an application server and its storage.Write-order consistency guarantees that all data written to storage is delivered in the propersequence so that the data can be re-used by an application after it has been replicated orcopied to a different site.
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
12/367 pyright © 2013 EMC Corporation. All rights reserved
Let’s take a look at how EMC RecoverPoint addresses customers’ challenges.
RecoverPoint ensures data replication, with high availability, and without impact to the production application
server or its data. The replicated data is stored in the same format as the source data, which eliminates the
need for costly encapsulation/de-encapsulation processes that other vendors require.
RecoverPoint is designed for enterprise customers. It includes key features such as bi-directional local and/or
remote replication, WAN optimization, and a policy-driven process. In addition, RecoverPoint ensures
consistent recovery regardless of replication distances, without requiring costly secondary or tertiary copies of
the replicated data.
RecoverPoint is not a one-size-fits-all approach. Customers can specify unique RPO/RTO polices for each of
their applications. In addition, RecoverPoint lets customers leverage under-utilized resources in one data
center to protect the resources in another data center. Providing the same functionality for data protection,
regardless of the host or storage configuration, ensures that the customer’s RecoverPoint investment is
maximized.
Finally, RecoverPoint protects the customer’s existing investment in their server, SAN and storage architectures
by providing replication across heterogeneous server and storage environments. RecoverPoint is also designedto minimize the bandwidth consumed for replication between sites. These sites may be separated by large
physical distances. Reducing the required bandwidth can decrease monthly operating expenses forWAN connections.
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
13/367 pyright © 2013 EMC Corporation. All rights reserved
RecoverPoint provides local data protection and replication for virtual machines. It usescontinuous data capture for local SAN data protection to protect the virtual machines, as wellas the VMware ESX server platform, from data corruption. It also guarantees recoverabilitylocally with no data loss. RecoverPoint supports virtual-to-virtual replication betweenequivalent storage configurations, such as VMFS to VMFS or RDM/P to RDM/P. This supportuses the VNX series or CLARiiON array-based write splitter or an intelligent-fabric writesplitter. Similarly, RecoverPoint supports physical and physical-to-virtual replication, as wellas virtual-to-virtual replication using host-based RecoverPoint write splitters installed in eachguest operating system. Physical RDM (RDM/P) -attached volumes are replicated locallyand/or remotely.
RecoverPoint’s VMware support can be used to enhance existing VMware vMotion andStorage vMotion solutions. RecoverPoint integrates with vCenter Server to allow theRecoverPoint administrator to quickly view virtual machines fully protected by RecoverPoint,and for those not fully protected, to see which of the virtual machine’s LUNs/data stores are
not protected. Additionally, RecoverPoint is integrated with VMware vCenter Site RecoveryManager to simplify and automate disaster recovery in VMware Infrastructure. RecoverPointalso has a vCenter Server plug-in that automates failback for SRM configurations.
The RecoverPoint management GUI can be used to monitor virtual machine protection statusas well as physical LUNs. This ability to monitor physical and virtual storage extends toheterogeneous storage.
If you are considering Microsoft Hyper-V, RecoverPoint can fully protect any Hyper-Vvirtualized application as long as its data resides on a SAN. RecoverPoint supports MicrosoftVSS and VDI, Windows Server Core, as well as Cluster Enabler version 4 under Hyper-V.
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
14/367 pyright © 2013 EMC Corporation. All rights reserved
RecoverPoint 4.0 integration with SRM enables users to test or failover SRM protection
groups to any point in time. This is done using the vSphere client and the Virtual Storage
Integrator (VSI) RecoverPoint Management. VSI is free of charge and downloadable from
support.emc.com. Installation, configuration, and usage of VSI RecoverPoint Management isquick, easy, and very similar to all other VSI products.
The process is simple. Simply apply the point in time copy to use with SRM, run the SRM test
or failover procedure, and the test or failover will occur using the point in time specified.
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
15/367 pyright © 2013 EMC Corporation. All rights reserved
RecoverPoint v4.0 and beyond provides the capability to concurrently protect data to
multiple disaster recovery sites using a single management point. For example, data can be
protected locally on one RPA cluster and concurrently copied to up to four remote RPA
clusters (“fan out”), or replicated from up to four remote RPA clusters to one, central RPAcluster (“fan in”). As shown in both examples, there can be up to five RPA clusters in a
RecoverPoint system.
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
16/367 pyright © 2013 EMC Corporation. All rights reserved
The RecoverPoint/SE system supports VNX/CLARiiON storage arrays — when using
VNX/CLARiiON splitters — with hosts running operating systems compatible with
RecoverPoint and the VNX/CLARiiON array.
RecoverPoint/SE Continuous remote replication (CRR) enables block protection as well as afile DR solution. This enables failover and failback of both block and file data from one VNX
to another.
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
17/367 pyright © 2013 EMC Corporation. All rights reserved
RecoverPoint/EX provides all of the data protection of RecoverPoint/CL, with the following
limitations:
• Only the following storage arrays are supported:
• VMAX Series
• VNX Series
• CLARiiON CX3 and CX4 series
• VPLEX Local and Metro storage virtualization platforms
• The following splitters are supported:
• VNX/CLARiiON
• Symmetrix
• VPLEX
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
18/367 pyright © 2013 EMC Corporation. All rights reserved
EMC RecoverPoint/CL provides block-level replication for on-demand protection and
recovery to any point-in-time image. This enables you to implement a single, unified
solution to protect and/or replicate data across heterogeneous storage.
RecoverPoint/CL is the full-featured offering which allows a single cluster to utilize EMC
storage array-based write splitters including the VPLEX splitter.
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
19/367 pyright © 2013 EMC Corporation. All rights reserved
The EMC RecoverPoint family provides cost-effective, local and remote replication dataprotection solutions that include point-in-time data recovery.
RecoverPoint/SE is the entry offering and is ideal for homogeneous VNX/CLARiiON CX
environments. RecoverPoint/SE is limited in function compared to RecoverPoint/CL orRecoverPoint/EX, though does allow for the use of virtual RPAs.
If you require support for arrays other than VNX/CLARiiON, then RecoverPoint/CL orRecoverPoint/EX is a more ideal solution than RecoverPoint/SE. RecoverPoint/CL alsosupports VPLEX Local and Metro.
RecoverPoint/EX provides all of the data protection of RecoverPoint/CL. However, somelimits apply on storage array types. RecoverPoint/EX supports VMAX series, VNX series,CLARiiON CX3 and CX4 series, Celerra unified storage environments, and VPLEX Local andMetro storage virtualization platforms. RecoverPoint/EX supports VNX/CLARiiON-basedsplitters, Symmetrix-based splitters and VPLEX splitters.
Some common features for the RecoverPoint family are:
• Synchronous replication
• Multipath I/O support
• Virtualization hypervisor support for Hyper-V and VMware platforms.
• Integration with VMware vCenter, VMware SRM and VAAI
• Built-in bandwidth reduction
• RPA’s are available as physical servers or virtual appliances
• Two to eight RecoverPoint appliances per RPA cluster for failure resiliency and throughput
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
20/367 pyright © 2013 EMC Corporation. All rights reserved
This lesson covers the RecoverPoint components, the volume allocation requirements and
the general requirements when deploying a RecoverPoint solution.
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
21/367 pyright © 2013 EMC Corporation. All rights reserved
A group of 2-8 physical (or virtual) RPAs that work together to replicate and protect data at the samegeographical location are called an RPA cluster. There can be up to five RPA clusters in a RecoverPointsystem. To scale-up and support a higher throughput rate, you can add up to eight RPAs to your RPAclusters.
The number of RPAs in an RPA cluster must be the same at all RPA clusters in a RecoverPoint system.You can manage all RPAs in an RPA cluster through a floating cluster Management IP address. Inother words, once RecoverPoint is configured, you can manage the entire installation from a single IPaddress.
Each Cluster has its own repository volume. The Replication volumes, a copy or production, belong toa cluster. RecoverPoint Write-Splitters are assigned to the cluster to which they are connected.Clusters are licensed separately; each one must have a license. RecoverPoint clusters can beconnected to each other using Fibre Channel or the WAN connection. A cluster is installed, thenconnected to another during deployment.
A cluster has a high-availability architecture with no single point of failure. New nodes can be addedor removed from the cluster as needed. Consistency Groups are configured within a node.
The maximum number of consistency groups per RPA cluster is 128. The recommended maximumnumber of consistency groups per RPA is 64. In case an RPA fails, consistency groups will fail over toanother RPA.
A single RPA cannot be configured to have more than 64 consistency groups. However, in the eventof RPA failure, groups on one RPA will move to other RPAs in the cluster. In such a case, an RPA cantemporarily hold up to 128 groups, and the data of all the groups will continue to be transferredbetween sites. This state is temporary, however, as an RPA with more than 64 groups may run intohigh loads, and you may not be able to meet your group policies.
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
22/367 pyright © 2013 EMC Corporation. All rights reserved
The Repository volumes contain consistency group metadata, and they are created on
existing storage arrays in the SAN environment. The Repository volumes serve all RPAs and
associated splitters for each cluster at a particular site, and they are created as part of the
installation process.In a CRR configuration, there must be one Repository volume for the RPA cluster on the local
and remote site.
The Repository volume must be presented only to the RPAs.
The minimum size for a repository volume is 3 GB. However, for RecoverPoint/SE, the
following rules apply:
• In RecoverPoint/SE 3.4 and later AND selecting the repository volume manually, the
size should be 2.86 GB on existing RAID groups (not RAID 0 groups); any capacity
beyond 3 GB is not used.• In RecoverPoint/SE 3.4 and later AND selecting the repository volume automatically,
the size should be 5.72 GB on existing RAID groups (not RAID 0 groups); capacity
beyond 5.72 GB is not used.
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
23/367 pyright © 2013 EMC Corporation. All rights reserved
Each copy of a consistency group must contain one or more volumes that are dedicated to
holding system information or point in time history. The type of information contained in the
journal differs according to the journal type.
There are two types of journals:
• “Copy journals”
• “Production journals”
Copy journals are dedicated to holding point-in-time information for each image on the
production storage, as well as bookmarks that mark significant points in time. The copy
journals hold all changes to data, so that the copy storage can be rolled back to a previous
point in time (PIT). There are as many copy journals as there are local and remote copies in a
consistency group.
Production journals are dedicated to store information about the replication process (called
marking information) that is used to make synchronization between the production and copy
volumes, when required, much more efficient.
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
24/367 pyright © 2013 EMC Corporation. All rights reserved
The size of a Journal volume is based on the length of time the user wishes to recover to for a specific
consistency group. For instance, if snapshots for a consistency group must be stored for 24 hours, the
journal volume must be at least as large as the amount of expected snapshot writes for a day.
The Journal Volume has two components
• Snapshot portion (80%) – contains all snapshots replicated plus undo information to rollback
to a specific snapshot
• Target Side Processing portion (20%) – contains all of the undo information while the
volumes are in TSP mode
Best characteristics for Journal volumes
• Mirrored for protection
• Striped and Mirrored on target side for reliability and speed
5 I/Os for each replicated write (Journal volume is not full)
• Flexible Journal volume pool and application tolerance for sizing
Recall that Journal volume is sized by determining the following:
• Data set to be replicated
• RTO/RPO to be achieved – history is required
• Rate of data change
Average I/O write rate must be determined over a long period of time.
Journal volumes should be able to:
• Sustain 2X sequential writes and X sequential reads
• Use striping across a volume set to increase performance
To prevent the possible loss of all journal snapshots during a group resynchronization, the
recommended minimum journal size is 10 GB for a non-distributed consistency group, or 40 GB for a
distributed consistency group.
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
25/367 pyright © 2013 EMC Corporation. All rights reserved
Snapshots include only that data that has changed from the previous snapshot. Upon being
distributed, a snapshot creates a new current image on the remote storage. A snapshot is the
difference between one consistent image of stored data and the next.
Snapshots are taken seconds apart. The application writes to storage; at the same time, thesplitter provides a second copy of the writes to the RPA. In asynchronous replication, the RPA
gathers several writes into a single snapshot. The exact time for closing the snapshot is
determined dynamically depending on replication policies and the journal of the consistency
group. All snapshots are crash-consistent by default.
When replicating synchronously, each write is a snapshot. When the snapshot is distributed
to a copy, it is stored in the journal, making it possible to revert to previous images at any
point in time.
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
26/367 pyright © 2013 EMC Corporation. All rights reserved
RecoverPoint Copies are all of the volumes of a consistency group that are either a source or
a target of replication at a given RPA cluster.
The following types of copies exist in a consistency group:
• The production copy—The production copy consists of all of the volumes that are the
sources of replication of a consistency group, as well as the production journal volumes
• Local copies—A local copy consists of all of the volumes that are the targets of replication,
for a specific consistency group
The following limitations apply to copies:
• A maximum of one production copy and up to four non-production copies can be
configured per consistency group.
• In local replication, there can be only one production copy and one local copy.
• In remote replication, if a local copy exists there can be up to three remote copies. If a
local copy does not exist, there can be up to four remote copies.
• A single cluster can have 2 remote copies maximum
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
27/367 pyright © 2013 EMC Corporation. All rights reserved
The consistency group ensures that updates to the replicas are always consistent and in
correct write order; that is, the replicas can always be used to continue working or to restore
the production source, in case it is damaged. The consistency group monitors all the volumes
added to it to ensure consistency and write-order fidelity. If two data sets are dependent onone another (for instance, a database and a database log), they must be in the same
consistency group.
A consistency group includes these elements:
• Settings and policies: Settings, such as consistency group name, primary RPA,
reservation support; policies, such as compression, bandwidth limits, and maximum
lag, that govern the replication process.
• Replication sets: A consistency group consists of one or more replication sets. Each
replication set consists of a production volume and the replica volumes to which it is
replicating.
• Journals: Receive changes to data. Each copy has a single journal. Changes are
distributed from the replica journal to storage. The replica journals also retain
rollback data for their replica. The production journal does not contain rollback
information. The system marking information is contained in the production journal.
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
28/367 pyright © 2013 EMC Corporation. All rights reserved
The RecoverPoint Appliance (RPA) is the data-protection controller for RecoverPoint. RPA
nodes utilize private LAN and shared RPA volumes for communications using the standard
TCP protocol. No FC/IP converters are needed to replicate across the WAN. The set of RPAs
at each site constitutes an RPA cluster, where each cluster can include from two to eight RPAnodes, as configured during RecoverPoint system installation. The cluster size must be the
same at all sites in an installation. In normal operation, all RPAs in a cluster are active all of
the time. Consequently, if one of the RPAs in a cluster goes down, the RecoverPoint system
supports immediate switchover of the functions of that node to another RPA in the cluster.
With RecoverPoint version 4.0, the RPA is supported on hardware platforms (Gen4 or Gen5)
that each include a single quad port QLE2564 8 Gb HBA which supports native 8 Gb/s Fibre
Channel connectivity.
Previous hardware platforms, Gen 3 and below, are not supported with RecoverPoint version
4.0 and beyond.
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
29/367 pyright © 2013 EMC Corporation. All rights reserved
Although deployed and used in fundamentally the same way as physical RPAs, the
RecoverPoint vRPA is the virtual appliance implementation of the RecoverPoint Appliance. As
with any other virtual machine, the vRPA operates with dynamic virtual resources allocated
to it from the hosting hypervisor.
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
30/367 pyright © 2013 EMC Corporation. All rights reserved
Please note the following special information and best practices when using vRPA. vRPA
cluster nodes should not share the same hypervisor server. This allows for redundancy in
case the host hypervisor server fails. It also allows virtual machine load sharing across the
physical resources of multiple hosts. Where multiple vRPAs must run on the same host,sufficient resources must be made available, especially to allow proper operation during
peak RPA load times.
RPAs and vRPAs cannot be mixed in one cluster. The cluster must be either all physical RPAs
or all vRPAs.
However, replication is supported between any of the following cluster types:
• RPA and RPA
• RPA and vRPA (in either direction)
• vRPA and vRPA
In RecoverPoint/SE installations, the number of RPAs in each of the two clusters must be the
same, regardless of whether they are physical or virtual.
vRPA is currently certified to be deployed only on the VMware vSphere environment.
Although similar in operation and use, distinct differences in installation and deployment
exist between physical and vRPAs. Because of this, the vRPA is not covered in detail in this
course. Please refer to the EMC RecoverPoint vRPA Technical Notes for more information
about vRPA installation, deployment and management within the VMware vSphere and
vCenter environments.
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
31/367 pyright © 2013 EMC Corporation. All rights reserved
There are three types of RecoverPoint licenses: RP/SE, RP/EX, and RP/CL. As with previous
RecoverPoint versions, RP/SE only supports a single VNX and CLARiiON-based splitter for
each cluster. It allows a maximum of two clusters, and each cluster needs its own license.
The license is based on the array and is tied to the array serial number. There are nocapacity-based RP/SE licenses. The license covers the entire array capacity. If a VNX at a
single cluster is being used for both local and remote replication, it will need both a local
replication and a remote replication license.
RP/EX supports all VMAX-based, VNX-based, and VPLEX splitters. Similar to RP/SE licenses,
RP/EX licenses are tied to the array serial number. Each array used in a cluster needs a EX
license for both local and/or remote replication. RP/EX licenses for the VNX-based splitters
are for the entire array. RP/EX licenses for VNX-based splitters are not based on capacity.
RP/EX licenses for VMAX-based and VPLEX splitters are per registered capacity. Registered
capacity means the amount of data on the array that is being protected. Locally registeredcapacity is licensed separately from remotely registered capacity. RecoverPoint/EX licenses
for the VMAX can be purchased either by registered capacity or by raw capacity.
RecoverPoint/CL licenses are based on the replicated capacity. Replicated capacity simply
means the amount of data that is being replicated. RP/CL licenses are not tied to any arrays.
This makes the replication environment very flexible, since adding, changing, and refreshing
arrays does not require any change to the RecoverPoint license. RecoverPoint/Cluster
Enabler (RP/CE) is included with all RP 4.0 offerings.
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
32/367 pyright © 2013 EMC Corporation. All rights reserved
RecoverPoint 4.0 introduces the Unisphere for RecoverPoint user interface, a web based
client that replaces the RecoverPoint Management Application available with previous
versions of RecoverPoint. Users familiar with any one of the EMC Common User Interface
Toolkit-based interfaces will find it easy and intuitive to navigate through the Unisphere forRecoverPoint GUI. This significantly reduces the learning curve for users. When logged in,
users are immediately presented with a topology of their RecoverPoint system, as well as
health and traffic information. All protection and recovery operations are wizard-based with
clearly defined, intuitive steps.
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
33/367 pyright © 2013 EMC Corporation. All rights reserved
The RecoverPoint CLI can be used to configure and manage an individual RPA or the entire
RecoverPoint Cluster. The 2 user accounts are the Admin account and the boxmgmt account.
The admin account is used to perform replication tasks, such as consistency group creation,
testing an image, and failover. The cluster or site management address can be used via anssh connection. If an RPA’s management address is used, the RPA must be “attached” to the
cluster for a connection to be made.
The boxmgmt account launches the “Installation Manager” menu upon successful login.
This is normally used for each individual RPA. This menu-driven program allows for the
installation, configuration, and diagnostics of an RPA.
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
34/367 pyright © 2013 EMC Corporation. All rights reserved
This lesson covers the RecoverPoint write-splitting options available.
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
35/367 pyright © 2013 EMC Corporation. All rights reserved
This slide shows the types of array-based write splitting used with RecoverPoint. They are:
• CLARiiON CX3/CX4, VNX Block, VMAX arrays and VPLEX.
Note: Only VNX/CLARiiON splitter and VMAX splitter will be covered in details during this
course. If you want to learn more about host-based splitters, fabric-based splitters and VPLEX
splitter refer to EMC Education Services site and search for:
• RecoverPoint VPLEX Splitter Deployment and Management elearning
• RecoverPoint Symmetrix Splitter Deployment and Management elearning
Each elearning has a video demonstration plus associated hands-on labs.
Beginning with RecoverPoint version 4.0, host- and SAN-based splitting are no longer
supported.
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
36/367 pyright © 2013 EMC Corporation. All rights reserved
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.
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.
As shown in this slide, split is through the VNX/CLARiiON splitter which resides on each
storage processor. Host writes are split at the array and sent to the production copy through
the RPA. The original write travels though its normal path to the production LUN.
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
37/367 pyright © 2013 EMC Corporation. All rights reserved
The Symmetrix RecoverPoint write splitter is supported on the VMAX family of arrays with
Enginuity 5876 or later.
The Symmetrix splitter is based on Open Replicator for Symmetrix (ORS). Splitting occurs at
the Symmetrix Fibre Channel Director/Interface Adapter (FA) port. Because of this,RecoverPoint must be zoned to the same FA ports as the replicated hosts.
From RecoverPoint 4.0 and later and Symmetrix Enginuity 5876 Q2 2013 SR and later, it is
permitted to replicate the same device with both RecoverPoint and SRDF, as long as SRDF
and RecoverPoint are replicating to two different target devices, and RecoverPoint is
replicating only to a local copy.
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
38/367 pyright © 2013 EMC Corporation. All rights reserved
RecoverPoint combined with the VPLEX write splitter provides any point-in-time business
continuity recovery to protect VPLEX platform data across a mixture of EMC and non-EMC
arrays. In addition, RecoverPoint can test replicated data on virtual volumes without impact
to production applications.
Before RecoverPoint version 4.0 there were three types of available write splitters: host,
array and intelligent fabric splitters. As of version 4.0, only array-based splitters are available
for RecoverPoint. Particular to VPLEX, the write splitter is built into the VPLEX engine,
making this more of a transparent implementation. Also, this makes RecoverPoint services
uniformly available to EMC and non-EMC arrays alike.
The VPLEX write splitter supports RecoverPoint CRR, CDP and CLR using all arrays that VPLEX
supports and with full VPLEX functionality. The RecoverPoint VPLEX splitter supports
virtualized environments and clustering environments. These environments include VMwareHA, Microsoft MSCS/Failover clustering, Oracle stretched RAC, Oracle OVM and AIX stretched
clusters.
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
39/367 pyright © 2013 EMC Corporation. All rights reserved
All VPLEX splitter directors on a single VPLEX platform will share the same settings. Any
operation on the VPLEX splitter will affect all its directors.
New and replaced directors will be detected and added automatically.
RecoverPoint will support the procedure of replacing directors (i.e. faulty ones) non-
disruptively.
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
40/367 pyright © 2013 EMC Corporation. All rights reserved
This lesson covers the connectivity options, zoning and masking considerations, and cabling
requirements when deploying a RecoverPoint solution.
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
41/367 pyright © 2013 EMC Corporation. All rights reserved
The network connection consists of a management network and a replication network.
• The management network has a static IP address for each appliance and a floating
static IP address for each site. The management network manages the appliances as
well as accesses the replication’s management CLI and GUI.
• The replication network consists of a static IP address for each appliance. It
communicates to the appliances on the opposite site and replicates data between the
sites. An NTP server ensures proper time synchronization with the outside world. It is
located only on the primary site, since the appliances have an internal
synchronization method to ensure all appliances are synchronized.
Note: An NTP server is recommended, but not required, and is typically connected
through the management network.
The RecoverPoint replication solution contains a SAN Discovery feature that enables the
appliances to automatically detect and display information about the LUNs that the
appliances can see on the storage. This information includes: WWN, LUN, size, and storage
type. To determine what LUNs are associated with which replication volume, the WWN, LUN
and size must be determined ahead of time. Volume replication sets are a matching of
replication volumes that replicate between each other. Once the replication volumes have
been determined based on the WWN, LUN and size, they need to be associated into these
replication pairs.
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
42/367 pyright © 2013 EMC Corporation. All rights reserved
In order to size the pipe, you need to know:
• Average incoming I/O for a representative window in MB/s (24hrs/7days/30 days)
• Compression level achievable on the data (often hard to obtain – rule of thumb is 2x
– 6x)
From these numbers, we can compute the minimal bandwidth requirements of the
environment – the multiplication of the compression level by the average incoming data.
Allocating this bandwidth for replication does not give any guarantees on RPO, or the
frequency of high loads. This is due to fact that I/O rate may fluctuate throughout the
representative period. This bandwidth is minimal in the sense that anything below it will not
necessarily suffice to sustain the incoming I/O.
To determine the actual bandwidth needed to meet the end user requirements, one must
understand the user RPO requirements, and analyze the I/O fluctuations.
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
43/367 pyright © 2013 EMC Corporation. All rights reserved
The following are caveats/notes on RecoverPoint compression:
• With RecoverPoint v3.x and above , compression can now be enabled on a
consistency group basis.
• If compression settings are changed, it will result in a loss of the journal volume and
requires a full sweep.
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
44/367 pyright © 2013 EMC Corporation. All rights reserved
The slide shows a typical SAN environment with mirrored fabrics. When mirroring an
environment, start by configuring each physical device in the storage array so that it is
presented across multiple storage ports. These storage ports would then be distributed
evenly across both fabrics. Each host accessing this storage would also have multiple portsdistributed across both fabrics. Volume access and device pathing to these storage arrays
through the servers would be handled by load balancing and path fail over software, such as
PowerPath. PowerPath manages the representation of the multiple instances of the same
devices to the host, and manages the load balancing and path failover necessary to maintain
the high availability of the environment.
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
45/367 pyright © 2013 EMC Corporation. All rights reserved
Physical RPAs connect to the hosts and storage subsystems through a Fibre Channel SAN.
Before deploying RecoverPoint, these subsystems should be in place.
In preparation for RecoverPoint deployment, ensure that there is one RPA Fibre Channel port
available per RPA per FC fabric. This ensures high availability by providing redundancybetween all components at the site.
Note: A dual fabric configuration is required at all sites for RPA, host, and storage
connectivity.
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
46/367 pyright © 2013 EMC Corporation. All rights reserved
RecoverPoint/SE supports two network topologies; dual-fabric and direct attach.
A dual-fabric topology is recommended, and must include the following connections:
• A port on each VNX/CLARiiON storage processor to each fabric.
• Host and storage devices must be connected to the Fabric. In most installations, these
connections already exist.
• Minimum of one RPA FC port connected to each fabric.
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
47/367 pyright © 2013 EMC Corporation. All rights reserved
A direct attach, “switchless” topology is intended for iSCSI environments. In iSCSI
environments, you should ensure all RPAs are physically connected to each other through
their HBA FC ports. In iSCSI configurations, there can be a maximum of two RPAs per RPA
cluster since two of the four existing RPA HBA FC ports are already connected directly to thestorage (for high availability). If more than two RPAs are required for the cluster, connect all
of the RPAs in the cluster through FC switches and zone them together.
Summary of the connections:
• Each RPA to one port on each VNX/CLARiiON storage processor.
• Two Fibre Channel ports connecting the two RPAs.
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
48/367 pyright © 2013 EMC Corporation. All rights reserved
Within the supported hypervisor, it’s recommended to create four virtual networks. One for
the LAN connectivity; one for the WAN; and two for the iSCSI (one per VM adapter). Also, it’s
recommended that the hypervisor have at least two physical network adapters to segment
the traffic between the LAN/WAN and iSCSI. For High Availability purposes, consider havingtwo physical adapters for the iSCSI network as shown in the picture.
The physical network will need to be segmented in three; one for the LAN/WAN
connectivity, and one for each iSCSI connection.
Please note that vRPAs do not use the VMware vmkernel iSCSI ports. vRPAs run iSCSI over
standard IP network ports.
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
49/367 pyright © 2013 EMC Corporation. All rights reserved
Zoning instructions vary according to the type of HBA ports installed in your RPAs and the splitter type to
be implemented. Here is information regarding RecoverPoint version 4 and for historical reference,
version 3.5. To determine which HBA ports are installed in your system, use one of the following methods:
1. Check your invoice:•Item numbers RP-HW-1U-GEN4/GN4n/GEN5/GN5n correspond to QLE25xx HBA.
•Item number RP-HW-1U corresponds to QLA23xx HBA.
2. Check the HBA card bracket at the back of the RPA:
•QLE25xx HBA has “8 Gb” stamped on the card bracket.
•QLE24xx HBA has “4 Gb” stamped on the card bracket.
•QLA23xx HBA has nothing written on the card bracket.
Here are some general guidelines and best practices when zoning RPAs:
Do not share VNX/CLARiiON ports between MirrorView™ and RecoverPoint
RPA HBA ports as both initiators and targets. Best practice is to use the RPA HBA ports as both initiators
and targets. In this configuration, maximum performance and redundancy is achieved with optimal use ofresources. Initiator-target separation is not supported in the following cases:
- When mixing different splitter types in the same site
- When using remote replication over Fibre Channel
- When using distributed consistency groups
Zone members. In general, EMC recommends single-initiator zoning, that is, zones with one initiator and
one or more targets. However, RecoverPoint Gen3 and Gen4 RPA ports can act as both initiators and
targets. In some configurations, it is required to add all RPA ports in a fabric to the same zone. The best
practice is always to zone Gen3 and Gen4 RPA ports in the same zone. In some configurations, it is
required.
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
50/367 pyright © 2013 EMC Corporation. All rights reserved
Server-to-storage zones. In most installations, these zones already exist as part of the SAN
implementation. Confirm that these zones exist. EMC requires single-initiator zoning.
RPA-to-storage zones. The zoning guidelines are different for RecoverPoint and
RecoverPoint/SE.
• RecoverPoint. All RPA ports in a fabric must be zoned with all storage ports that are
available for RecoverPoint replication. The best practice is always to use only one
zone per fabric between Gen3 and Gen4 RPAs and storage. This applies to all splitters.
• RecoverPoint/SE. On each fabric, there should be a zone for each RPA, which
includes:
One HBA port of that RPA
One port for each VNX/CLARiiON storage processor (SP A and SP B)
Zoning for distributed consistency groups. Distributed consistency groups require that all
RPA Fibre Channel ports in the fabric can communicate with each other. All RPA ports in a
fabric must, therefore, be in one zone.
RPA WWNs can be recognized in the SAN by their 5001248 prefix. You can use the RPA
Installation Manager option Diagnostics > Fibre Channel diagnostics > View Fibre Channel
details to list the WWNs of each RPA.
Note: Refer to splitter e-learning modules for specifics on zoning.
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
51/367 pyright © 2013 EMC Corporation. All rights reserved
Customers can replicate data between a production site and a remote site over Fibre
Channel. Customers can leverage existing Fibre Channel infrastructure to improve application
efficiency, throughput and performance. In this configuration, the production site RPAs must
be zoned together with the remote RPAs. Listed are a few considerations when replicating toa remote site using FC:
• Uses FC network links between sites for CRR, instead of Ethernet - WAN
• Either FC or IP replication (not both)
• If users choose Fibre, they can then choose to allow control connections (non-data)
between sites over IP
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
52/367 pyright © 2013 EMC Corporation. All rights reserved
Listed are supported fabric environments for implementing asynchronous replication over
Fibre channel.
• RecoverPoint 3.5 and above requires 64-bit HBAs – QLE25XX
• Site layout must be for CRR
• ISL fabrics
• Extended fabrics
Support for either ISL-connected fabrics or extended fabrics are permitted. Please consult
with the EMC TechBooks for examples of stretch fabrics.
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
53/367 pyright © 2013 EMC Corporation. All rights reserved
Listed here are some basic restrictions and constraints. Please take a moment to read them.
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
54/367 pyright © 2013 EMC Corporation. All rights reserved
Replication over Fibre Channel is enabled during the installation process – if replication over
Fibre channel is selected, WAN replication will not co-exist.
To configure, create one zone per fabric that includes all of the site1 RPA ports together with
all of the site2 RPA ports. The fabric in the environment will consist of switches which areconnected by ISL or by extended fabrics. Eth0 (WAN interface) is still used for redundancy
purposes and is called the interface. GUI/CLI will reflect a FC setup with all
relevant information including: fabric statistics, event logs, and schema.
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
55/367 pyright © 2013 EMC Corporation. All rights reserved
Before the RPAs are installed and consistency groups are defined, sufficient volumes must be made
available on the SAN-attached storage at each RecoverPoint cluster.
Repository As part of the installation process, a repository volume must be created at each cluster.
The repository volume should reside in a storage environment that guarantees high availability, andmust be accessible to all RPAs. The repository volume holds configuration data that is created in the
installation and replication process. Repository volumes should not be visible to the source or target
hosts.
Journal Volumes that are to be used as Journals for each Consistency Group must be masked to the
RPA. Only volumes that are masked will be available as choices as Journal volumes. Journal volumes
should be made visible to the RPAs only.
Replication Volumes that are to be used in the creation of Replication Sets must be masked to the
RPAs or vRPAs. These volumes are also masked to the hosts. If a volume(s) are to be used as a copy
or production does not change the masking requirements.
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
56/367 pyright © 2013 EMC Corporation. All rights reserved
The slide shows a typical vRPA configuration in a standard SAN environment. Here, vRPA
nodes connect into the FC fabric using iSCSI through the hypervisor host servers. The host
servers themselves are each connected to storage through Fibre Channel HBAs or thru iSCSI
connections.Please refer to the EMC RecoverPoint vRPA Technical Notes for more information.
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
57/367 pyright © 2013 EMC Corporation. All rights reserved
Here is a comparison of physical and virtual RPA connection and other basic features.
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
58/367 pyright © 2013 EMC Corporation. All rights reserved
This module covered the different RecoverPoint configurations and an overview of the
multiple installation and configuration procedures when deploying a RecoverPoint solution.
Module 1: RecoverPoint Overview
-
8/18/2019 RecoverPoint Implementation Student Guide
59/367 pyright © 2013 EMC Corporation. All rights reserved
This module focuses on implementing a RecoverPoint/SE solution.
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
60/367 pyright © 2013 EMC Corporation. All rights reserved
This lesson covers an overview of RecoverPoint/SE Implementation and its requirements.
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
61/367 pyright © 2013 EMC Corporation. All rights reserved
A RecoverPoint/SE solution provides data protection and disaster recovery using VNX arrays.
A RecoverPoint 4.0 implementation can contain 1 or 2 clusters, with a single array (CX or VNX
models only ) for each cluster. Physical or virtual RPAs can be used. A single cluster must
contain only one or the other type. A cluster using physical RPAs can be connected to aseparate cluster containing vRPAs.
The vRPA can reduce the total solution cost for VNX customers by up-to 60%. With a
RecoverPoint/SE license, any number of vRPAs can be ordered to scale your environment for
no additional charge.
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
62/367 pyright © 2013 EMC Corporation. All rights reserved
The Virtual RecoverPoint Appliance (vRPA). vRPAs are virtual machines that run the
RecoverPoint software. They access the repository, journal, production, and copy volumes
via the iSCSI protocol, and therefore do not require any FC infrastructure. vRPAs are a great
way to get many of the benefits of RecoverPoint without the need for physical appliances ora SAN infrastructure. One important consideration is that vRPAs are only available for use in
RecoverPoint/SE (RP/SE) configurations. This means that the storage array must be a VNX.
Another important consideration is that since RP 4.0 supports synchronous replication over
IP, if the WAN is sufficiently robust, vRPAs can be used for remote synchronous replication.
vRPAs can replicate any block data, regardless of how the hosts are connected to the VNX.
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
63/367 pyright © 2013 EMC Corporation. All rights reserved
Within the Data Protection window, you are able to manage the different methods available
to protect the storage. The Snapshots option allows for LUN snapshots and File System
Checkpoint management. The Mirrors and Replications icon allows you to create and
manage Mirrors and Consistency Groups for block storage and Replication for File Storage.On the right-hand side, there are many wizards that allow for quicker access to create and
manage data protection.
Unisphere also has plug-ins that allow for direct start-up of many storage-related tasks
including RecoverPoint/SE.
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
64/367 pyright © 2013 EMC Corporation. All rights reserved
This lesson covers the RPA preparation process when deploying a RecoverPoint/SE solution.
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
65/367 pyright © 2013 EMC Corporation. All rights reserved
The RecoverPoint Procedure Generator (RPPG) is a stand-alone Windows application used to
generate precise, user-driven steps for high demand tasks carried out by Customer Support
and field personnel.
In the RecoverPoint Procedure Generator, each procedure is tailored and generated based onuser selectable prompts. This generated procedure gathers the most critical parts of
RecoverPoint product guides and combines sections of documentation along with expert
advice from RecoverPoint SMEs into a sequential set of steps in a single document.
One of the major differences between RPPG and other EMC Procedure Generators is that
there is no expiration of the product or a forced re-install. RPPG will check for application
and content updates each time RPPG is launched and automatically download them (if
allowed by the user).
RPPG includes the ability to submit feedback via a form to a back-end database. We
encourage everyone to submit improvements for the tool, ideas for future procedures, and
corrections/additions to the content via this mechanism.
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
66/367 pyright © 2013 EMC Corporation. All rights reserved
To facilitate a successful deployment of RecoverPoint and RecoverPoint/SE, it is important to
assemble selected information about your installed networks and storage systems. It is also
necessary to ensure that the LUNs needed by RecoverPoint are available on your storage.
EMC Global Services will assist you with these pre-deployment activities:• RecoverPoint connectivity
• Zoning
• Unpacking the RPAs
The Wan IP port is given a Temporary Virtual IP address: 10.77.77.77 by default. This can be
used to connect via SSH and assign a management address to be used during configuration.
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
67/367 pyright © 2013 EMC Corporation. All rights reserved
Using “boxmgmt” user, log in to all RPAs to be deployed within the same global cluster and
set a management IP address for them.
Although this solution is not commonly used, in some scenarios where there is a DHCP
server, Deployment Manager can set the IP address during the installation. In this case, youdo not have to set them manually before the installation begins. vRPAs have a temporary
address set during the installation of the VM, but this method can be used to alter the
address.
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
68/367 pyright © 2013 EMC Corporation. All rights reserved
For the Fibre Channel connections, a dual-fabric topology is recommended. It must include
the following connections:
• A port on each VNX/CLARiiON storage processor to each fabric.
• Each RPA HBA to each fabric
A direct attach, “switchless” topology, intended for iSCSI host-connected environments, must
include the following connections:
• Each RPA to one port on each VNX/CLARiiON storage processor
• Two Fibre Channel ports connecting the two RPAs
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
69/367 pyright © 2013 EMC Corporation. All rights reserved
vRPAs are available to be installed using VMware vSphere server. They are deployed from
OVA files which can be downloaded from support.emc.com. During installation, there are
three options for vRPAs. The minimum vCPU, disk space, and memory configurations for
each of these is shown here. The performance of each of these options will be documentedin the RecoverPoint 4.0 Release Notes. Please see the release notes to match the proper
configuration to the environment’s needs. Please note that the performance numbers in the
Release Notes will be given on a reference hardware, and that performance can vary
depending on the physical hardware on which the vRPAs will run. It is also possible to change
the hardware configuration post-deployment, as long as the resources are not less than
those of the minimal configuration shown in the table.
There is no special license needed for vRPAs. Since vRPAs are only supported with RP/SE, the
RP/SE license for the array is the only license needed.
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
70/367 pyright © 2013 EMC Corporation. All rights reserved
Configuring virtual networks for vRPA:
- In low-traffic deployments, all four network cards may be placed on the same virtual
network (on a single vSwitch).
- Where high performance is desired, separate LAN and WAN traffic from iSCSI traffic. For
even better performance, place each network (LAN, WAN, iSCSI1, and iSCSI2) on a separate
virtual switch.
- For high-availability deployments, in which clients have redundant physical switches, the
best practice is to route each iSCSI card to a different physical switch. This will usually involve
specifying a different virtual network for each of the iSCSI cards.
iSCSI bandwidth:
In VNX, either of two types of iSCSI modules may be available:
- 1Gb port cards, which allow bandwidth of 1Gb ,2Gb, 4Gb
- 10Gb port cards, which allows bandwidth of 10Gb, 20Gb
Since the vRPA relies on virtual networks, the bandwidth that the user exposes to the vRPA
iSCSI vSwitches will determine the performance that the vRPA can reach. The user can
configure vSphere hosts with quad port NICs and present them to the vSwitches.
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
71/367 pyright © 2013 EMC Corporation. All rights reserved
vRPAs can be deployed using the OVA package, which can be downloaded from
support.emc.com
From vCenter Server management console, Select the File menu, then Deploy OVF
Template. The OVF deployment wizard appears. Specify the vRPA OVF package location.
The best practice is to create each vRPA instance individually from an OVF package. This
makes temporary IP address assignment simpler. To allow for redundancy, vRPAs should not
share the same ESX server.
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
72/367 pyright © 2013 EMC Corporation. All rights reserved
The vRPA OVF template details appear, showing the general properties of this OVF template.
To accept, click Next.
The EMC End-User License Agreement appears. If you accept the terms, click Accept and
Next.
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
73/367 pyright © 2013 EMC Corporation. All rights reserved
Enter the name you wish to assign to this vRPA. If you enter the name of an existing vRPA,
you will not be permitted to continue deployment.
Select the desired configuration profile. The vRPA supports dynamic provisioning of these
resources, as long as they do not go below the values selected.
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
74/367 pyright © 2013 EMC Corporation. All rights reserved
Specify the vRPA Storage (datastore) that will be used to host the vRPA virtual machine files.
The best practice is to select a high-performance datastore. Do not use a datastore that you
intend to replicate with RecoverPoint. The datastore should be 80GB minimum.
Specify the vRPA datastore virtual Disk Format. SelectThin Provision.
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
75/367 pyright © 2013 EMC Corporation. All rights reserved
This step in the deployment wizard maps the four required vRPA Ethernet ports (Source
Networks) to the available vSphere virtual networks (Destination Networks). Do not select
the internal VMware iSCSI network (iSCSI Network option).
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
76/367 pyright © 2013 EMC Corporation. All rights reserved
This step determines the method of assigning a temporary IP address for management of the
vRPA. This address will be used during the installation of the RecoverPoint cluster. It is only
a temporary address and will not become permanent until the cluster is installed.
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
77/367 pyright © 2013 EMC Corporation. All rights reserved
Specify the desired IP address for the vRPA. This address will only be used during cluster
installation by Deployment Manager.
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
78/367 pyright © 2013 EMC Corporation. All rights reserved
The Ready to Complete screen summarizes all your selections. Click Finish to initiate the
deployment of this instance of the vRPA.
The Deploying vRPA appears, showing the progress of the deployment.
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
79/367
-
8/18/2019 RecoverPoint Implementation Student Guide
80/367 pyright © 2013 EMC Corporation. All rights reserved
This lesson covers deploying the VNX write-splitter for a RecoverPoint/SE solution.
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
81/367 pyright © 2013 EMC Corporation. All rights reserved
The VNX/CLARiiON splitter runs in each storage processor of a VNX, CLARiiON CX3 or CX4
array and splits (“mirrors”) all writes to a VNX/CLARiiON volume, sending one copy to the
original target and the other copy to the RecoverPoint appliance.
All RecoverPoint families support the VNX/CLARiiON splitter. The VNX/CLARiiON splitter issupported on VNX, CLARiiON CX3 and CLARiiON CX4 arrays.
The VNX/CLARiiON splitter is supported on both iSCSI and FC attached volumes and supports
hosts running OSes compatible with the VNX/CLARiiON array.
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
82/367 pyright © 2013 EMC Corporation. All rights reserved
The RecoverPoint Splitter driver is embedded in the operating system for a VNX or CLARiiON
array. Installing the enabler for RecoverPoint can be executed using the Unisphere Service
Manager tool.
To do it, first connect to the array, as shown above.
• Using Unisphere Service Manager tool, select the system to upgrade or install.
• In the System tab, select Software option. The Storage System Installation Software
window will appear.
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
83/367 pyright © 2013 EMC Corporation. All rights reserved
In the installation options, select Custom Install (default) and click Next.
Implementation tasks that are taken for granted as being done here are that the array is
available via an IP address and that security has been enabled on the array.
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
84/367 pyright © 2013 EMC Corporation. All rights reserved
Proceeding with the implementation, we see the USM tool shows the software which is
already installed on the array. Using the browse button, we can locate the splitter code and
splitter enabler which we will install on the array. Note that the splitter code and enabler will
be available for EMC employees on the GS Intranet. You have to download this code and fileto a local management host in order to perform the installation.
• Click Browse and locate the copy of software installed on your management host
Installing the Splitter bundle
Installing the Splitter enabler
• Proceed with file transfer – Click Next to proceed
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
85/367 pyright © 2013 EMC Corporation. All rights reserved
The USM tool performs several base tasks as part of the installation. It checks for connected
servers, collects diagnostics, and checks a variety of rules before installation is allowed. If
there are issues found, it is possible to correct or override the existing rules. Once checking is
complete, the actual install is initiated.
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
86/367 pyright © 2013 EMC Corporation. All rights reserved
Shown is the installation beginning on each SP of the array. Once a single SP has been
updated with the enabler and splitter code, it is rebooted. After a successful reboot, the
other SP is then installed and rebooted.
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
87/367 pyright © 2013 EMC Corporation. All rights reserved
On VNX deployments, the splitter has been added to the VNX OE code. However, it is still
required to install the splitter enabler. This can be accomplished using USM.
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
88/367 pyright © 2013 EMC Corporation. All rights reserved
This lesson covers the installation of RecoverPoint/SE.
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
89/367 pyright © 2013 EMC Corporation. All rights reserved
The RecoverPoint Deployment Manager is a package of graphical user interface (GUI) wizards
that help you perform a number operations from a single management point.
The Deployment Manager can be downloaded from the Downloads section of
support.emc.com.
It is available as a zip file for use from a Windows machine.
To launch the Deployment Manager, open the folder and double-click the executable file:
RecoverPoint_DM.exe. Select the correct RecoverPoint version for the task to be completed.
At the login screen, select Other operations (EMC personnel) and enter the default CLI
RecoverPoint login credentials (user admin, password admin).
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
90/367 pyright © 2013 EMC Corporation. All rights reserved
Deployment Manager 2.0 branches off into two paths. Throughout the solution for the
implementation scenario presented, both paths will need to be accessed. The first path
provides options for installation, upgrade, and other operations for RecoverPoint 4.0. End
users who would like to install RecoverPoint 4.0 on a new hardware appliance or virtualRecoverPoint Appliance (vRPA) should use this option. Users who would like to upgrade to
RecoverPoint 4.0 from version 3.5 or earlier should choose the second option. This option
will launch Deployment Manager 1.1 to allow maintenance of RecoverPoint 3.x clusters.
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
91/367 pyright © 2013 EMC Corporation. All rights reserved
The RecoverPoint/SE Installer Wizard is a graphical user interface (GUI) wizard that helps you
install a new RecoverPoint/SE configuration from a single management point. The Installer is
intended for deployment of RecoverPoint/SE using only VNX series and CLARiiON CX3 and
CX4 series splitters. This wizard is used to install a single cluster of either vRPAs or PhysicalRPAs.
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
92/367 pyright © 2013 EMC Corporation. All rights reserved
The next step in the SE installation wizard outlines the required prerequisites for the
RecoverPoint environment. The conditions are defined for both Physical and Virtual RPAs.
The environment, SAN, IP, connectivity, and Array code versions are defined.
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
93/367 pyright © 2013 EMC Corporation. All rights reserved
The RecoverPoint/SE Installer allows you to select one of two installation modes:
• Create a new installation configuration file – The system creates a new installation
configuration file from the beginning. Starting with a blank configuration, the user must
enter configuration settings for the installation. The system uses these settings tocomplete the installation.
• Create or continue an installation from a saved configuration file – The system loads the
settings from a previously saved installation configuration file.
• If using the file to create a new installation from a saved file, and the file was
already applied to a previous installation, the system will allow you to use the file
and modify the configuration parameters.
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
94/367 pyright © 2013 EMC Corporation. All rights reserved
This step defines the RecoverPoint cluster environment. Entries under the General category
create the Cluster name and the Number of RPAs. The Connectivity category lists the IP
version type and MTU size. The MTU defaults at 1500 and should not be altered unless
instructed to do so. The Environment category lists options such as the DNS server, Timezone, and NTP server.
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
95/367 pyright © 2013 EMC Corporation. All rights reserved
The next step in the wizard defines how to set the IP addresses of uninstalled RPAs. This
address will be used for discovery until the wizard sets the final management address.
Two options are provided :
“I have already set IP addresses for the RPAs”
This option is recommended when one or more of the following is true:
• You have already set the IP addresses.
• You are installing RPAs running releases earlier than 3.5
• Your environment does not include a DHCP server.
“I want Deployment Manager to discover the RPA IP addresses”
This option is only relevant when RPAs are running release 3.5 or later and your
environment includes a DHCP server. When you select this option, the Discover
uninstalled RecoverPoint Appliances (RPAs) section becomes active.
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
96/367 pyright © 2013 EMC Corporation. All rights reserved
This screen allows the user to define the management and WAN networks. The cluster
management address will be used to configure replication once the cluster is created. The
WAN address must be entered for communication even if the cluster will not be connected
to another cluster. The wizard allows for additional routes to be added if the default gatewayis not sufficient.
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
97/367 pyright © 2013 EMC Corporation. All rights reserved
Enter the default RecoverPoint login credentials. The system prompts you to enter your
RecoverPoint login information. For username boxmgmt, the default password is boxmgmt.
Review the results of the RecoverPoint connectivity test. The system communicates and
validates the statuses of the RPAs and confirms that all components are OK, or that errorsare detected. The system presents the following results: Status, RPA, Site name, RPA IP
address, Gen (RPA hardware platform), and Comment (if applicable).
If errors are detected for an RPA, a comment appears describing the problem. To proceed to
the next step in the wizard, you must resolve all errors. Once errors are resolved, click Retry
to revalidate the communication statuses.
When all statuses appear OK (green check mark), click Next to continue.
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
98/367 pyright © 2013 EMC Corporation. All rights reserved
This screen only appears if you are installing a virtual RPA (vRPA).
Define the RPA cluster iSCSI settings.
Connectivity
For iSCSI 1
IP type — IPv4 or IPv6.
MTU — Default value is 1500 (do not modify unless instructed to do so).
For iSCSI 2
IP type — IPv4 or IPv6.
MTU — Default value is 1500 (do not modify unless instructed to do so).
If relevant, select the I want to configure CHAP credentials for the RPAs iSCSI ports
checkbox and modify the CHAP credentials (username and password).
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
99/367 pyright © 2013 EMC Corporation. All rights reserved
Select how to update the RecoverPoint release running on the RPAs. The RecoverPoint
release running on the RPAs is displayed (the ISO image that comes pre-installed on the
RPAs). To ensure the RPAs are running the target RecoverPoint release, you should update
them now.Update options:
• Update with an ISO image available from EMC Online Support site (recommended)
Select this option to download an ISO image available from EMC Online Support site
(login credentials required). This option will be grayed out if no ISO image can be found on
EMC Online Support site.
• Update with an ISO image available locally
Select this option if you have an ISO image on your local machine, a DVD, a USB device, or an
FTP server.
• Do not update. Continue with current RecoverPoint release on RPAs.
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
100/367 pyright © 2013 EMC Corporation. All rights reserved
The system attempts to log in to each of the RPAs and validate system information. The
results of applying the configuration settings to all RPAs are displayed. Once installation has
completed successfully, click Next.
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
101/367 pyright © 2013 EMC Corporation. All rights reserved
Enter the login credentials for your VNX/CLARiiON storage system used for this cluster.
The system now has the necessary information required to validate the VNX/CLARiiON
settings. The system attempts to log in to the VNX/CLARiiON array and validate the following
information:
The RecoverPoint/SE Installer can communicate with VNX/CLARiiON SP A and SP B.
• The login credentials are correct for SP A and SP B.
• SP A and B are not already attached to RecoverPoint/SE cluster.
• IP settings were applied successfully.
The system communicates and validates the statuses of the RPAs and confirms that all
components are OK or that errors are detected. If errors are detected, you can click the error
to display details, and instructions on how to correct the problem. The array name and serial
number of the array are also displayed. To proceed to the next step in the wizard, you must
resolve all errors. Once errors are resolved, click Refresh to revalidate the communication
statuses. Warnings are displayed, but give the user the option to continue or not.
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
102/367 pyright © 2013 EMC Corporation. All rights reserved
This screen only appears if you are installing a virtual RPA (vRPA). The current cluster iSCSI IP
configurations and RPA iSCSI IP settings are displayed. These were configured during the
installation of the RPA VM. Enter the iSCSI settings for the new RPAs. Interface netmask is
based on the ip type for the LAN and WAN. This was chosen in the "Environment settings"section.
IMPORTANT: It is strongly recommended to put the iSCSI interface on a different subnet
than the LAN and WAN interface, to avoid iSCSI communication errors.
Additional gateways for the iSCSI network can be added at this time.
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
103/367 pyright © 2013 EMC Corporation. All rights reserved
The Deployment Manager wizard allows the user to select from four options for zoning the
RPAs to storage.
• Automatic - The Installer zones the RPAs and the VNX/CLARiiON arrays automatically.
• Semi-automatic - The Installer creates the necessary zoning scripts, but you must
apply them yourself.
• Manual - The Installer indicates which ports should be zoned together, and you are
responsible for actually creating the zones.
• Direct-attach Storage - The Installer verifies that, in this switchless configuration, each
RPA is directly connected to SP A and SP B of the VNX/CLARiiON array.
Automatic zoning can be performed on Cisco fabric switches with an active zoneset, and on
Brocade (Connectrix B-Series) fabric switches running FOS 6.0 and higher with an active zoneconfiguration. A Telnet connection to Cisco and Brocade switches is mandatory.
The Installer automatically creates an RPA-to-storage zone for each RPA at this site, where
each zone includes:
• One HBA port of that RPA
• One port for each VNX/CLARiiON storage processor
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
104/367 pyright © 2013 EMC Corporation. All rights reserved
The Installer retrieves the WWNs for the RPA ports and the SP ports on the VNX. Zone and
mask LUNs using the RPA port WWNs are displayed in the RPAs WWNs table(s), and the
VNX/CLARiiON array port WWNs are displayed in the VNX/CLARiiON Array table.
On each fabric, create a zone for each RPA, which includes:
• One HBA port of that RPA
• One port for each VNX/CLARiiON storage processor
At each site, on each fabric, create an RPA-to-RPA zone containing all local RPA HBAs that are
not used in RPA-to-storage zones (best practice). This zone provides an additional
communication link between RPAs. It can also be used for synchronous replication of
distributed Consistency Groups. If replicating over Fibre Channel, merge the RPA-to-RPA
zones at each site into one.
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
105/367 pyright © 2013 EMC Corporation. All rights reserved
The system performs SAN diagnostics for the site, attempting to validate that the current
configuration is appropriate for the RPA, and that RecoverPoint/SE can successfully operate
in the current SAN environment. The system presents the following results: Severity, RPA,
Message ID, Category, and Description. The system confirms that all components are OK andthat no errors exist. If errors are detected, you can click the error to display details and
instructions on how to correct the problem.
Once errors are resolved, click Refresh to rerun the SAN diagnostics process. It is highly
recommended to resolve all SAN discovery warnings, even though the system allows you to
ignore them and continue the installation.
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
106/367 pyright © 2013 EMC Corporation. All rights reserved
Determine how you want the repository volume to be selected:
Automatic - The Installer automatically creates and configures a repository volume.
Manual - The Installer lists possible volumes, and you can select the one to be configured as
the repository volume. The list is volumes configured on the VNX. Care must be taken not to
choose a volume already in another storage group.
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
107/367 pyright © 2013 EMC Corporation. All rights reserved
After the Repository volume has been configured, the next step is for Deployment Manager
to apply the configuration and create the cluster.
These are the steps Deployment Manager Performs:
Registering RPA WWNs on VNX/CLARiiON array, Creating storage groups, Configuring RPAs,
Creating repository volume and assigning it to the newly created storage group.
From there, the wizard finishes the installation by Attaching the RPAs to the cluster and
waiting for the RPAs to come up after they reboot.
Progress bars indicate that the system is applying settings and completing the installation.
Once the progress bars complete, the system confirms that all components are OK, or that
errors are detected. If errors are detected, you can click the error to display details and
instructions on how to correct the problem. To proceed to the next step in the wizard, you
must resolve all errors. Once errors are resolved, click Retry to reapply settings andrevalidate that the RPA installation is successful.
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
108/367 pyright © 2013 EMC Corporation. All rights reserved
You have successfully completed the RecoverPoint/SE installation. The RecoverPoint/SE
storage group created will be displayed.
IMPORTANT:
The storage group contains all the VNX/CLARiiON LUNs that RecoverPoint/SE replicates. Do
not change the name of the group, or RecoverPoint/SE may fail to function properly. To start
managing the RecoverPoint system, click the Unisphere for RecoverPoint hyperlink or click
Finish and enter the system using your browser.
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
109/367 pyright © 2013 EMC Corporation. All rights reserved
This lesson covers Connecting RecoverPoint clusters.
Module 2: RecoverPoint/SE Installation
-
8/18/2019 RecoverPoint Implementation Student Guide
110/367 pyright © 2013 EMC Corporation. All rights reserved
RecoverPoint clusters can be connected together in order to perform replication to or from
different geographical locations. New clusters can be connected to existing clusters. A new
cluster is a RecoverPoint cluster that has been installed, but not licensed. Connecting RP