EMC AppSyncUser and Administration Guide

EMC AppSyncUser and Administration Guide
EMC® AppSync™
Version 2.2.3
User and Administration Guide
302-002-782
01
Copyright © 2012-2016 EMC Corporation. All rights reserved. Published in the USA.
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 of merchantability 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, and the EMC logo 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.
For the most up-to-date regulatory document for your product line, go to EMC Online Support (https://support.emc.com).
EMC Corporation
Hopkinton, Massachusetts 01748-9103
1-508-435-1000 In North America 1-866-464-7381
www.EMC.com
2
EMC AppSync 2.2.3 User and Administration Guide
CONTENTS
Chapter 1
Introduction
9
AppSync overview ........................................................................................ 10
Overview of service plans.................................................................10
Role-based management................................................................. 11
Reports show proven protection ...................................................... 11
AppSync architecture.................................................................................... 11
AppSync server ............................................................................... 12
AppSync agent (host plug-in) overview.............................................12
AppSync Console (user interface).....................................................13
REST interface.................................................................................. 14
Differences between AppSync and Replication Manager................................14
Interoperability of AppSync and Replication Manager ......................14
Chapter 2
AppSync Console
15
Console overview.......................................................................................... 16
Performing actions........................................................................... 16
Times shown in the console............................................................. 16
Troubleshooting console problems.................................................. 16
Troubleshooting browser display problems ..................................... 16
Troubleshooting browser refresh issues........................................... 17
Setting preferences....................................................................................... 17
Starting the AppSync console........................................................................17
Chapter 3
Service Plans
19
Service plan overview....................................................................................20
Creating a service plan..................................................................... 23
Summary of Exchange service plans settings.................................................24
Summary of SQL Server service plans settings...............................................26
Summary of Oracle service plan settings....................................................... 29
Summary of file system service plan settings.................................................31
Summary of VMware service plans settings................................................... 33
Chapter 4
AppSync CLI Utility
35
AppSync CLI Utility tutorial............................................................................ 36
Chapter 5
Protect Microsoft Exchange
41
Overview of Exchange support ......................................................................42
Deploying AppSync for Exchange protection: summary of steps.................... 44
Discovering Exchange databases .................................................... 44
Removing an Exchange mailbox server ............................................ 44
Protecting DAG databases in a service plan......................................45
Protect an Exchange database ......................................................................45
Protecting an Exchange database immediately.................................45
Subscribing an Exchange database to a service plan........................46
Unsubscribing a database from a service plan .................................46
Expiring a copy on demand.............................................................. 46
EMC AppSync 2.2.3 User and Administration Guide
3
CONTENTS
Creating a database copy from the Copies page............................... 47
Service plan details.......................................................................................47
Service plan schedule...................................................................... 47
Application discovery.......................................................................48
Application mapping........................................................................49
Pre-copy script ................................................................................ 49
Create copy...................................................................................... 49
Post-copy script............................................................................... 51
Unmount previous copy................................................................... 52
Mount copy......................................................................................52
Validate copy................................................................................... 54
Post-mount script.............................................................................55
Unmount copy..................................................................................55
Mounting Exchange copies............................................................................56
Mount and restore limitations.......................................................... 56
Mounting an Exchange copy on-demand.......................................... 56
Unmounting an Exchange copy ........................................................59
Overview of Exchange copy restore................................................................60
Affected entities during restore........................................................ 60
Restoring from an Exchange copy..................................................... 61
Recovering an Exchange database manually.................................... 62
Restoring a standalone Exchange database on XtremIO .................. 62
Restoring an Exchange database DAG on XtremIO............................ 63
Partial restore.................................................................................. 64
Restoring logs from crash-consistent (APiT) copy..............................65
Restoring a deleted Exchange database........................................... 66
Item level restore ............................................................................ 66
Troubleshooting the EMC AppSync Exchange Interface service ..................... 67
Chapter 6
Protect SQL Server
71
Overview of SQL Server support.....................................................................72
SQL Server prerequisites.................................................................. 72
SQL Server supported configurations............................................... 73
Support for SQL Server on virtual disks............................................ 73
Required permissions and rights......................................................73
Update login credentials for a SQL Server instance...........................74
Support for AlwaysOn Availability Groups......................................................75
SQL Server transaction log backup................................................................ 75
Configure SQL Server transaction log backup................................... 76
Configure log backup scripts............................................................ 78
Run log backup on demand..............................................................79
View log backups for a service plan..................................................80
View SQL database copies............................................................... 80
Log backup expiration......................................................................83
Considerations for working with SQL Server in a cluster.................................84
SQL Server User Databases folder................................................................. 86
Discover SQL Server instances......................................................... 86
Protect a SQL Database................................................................................. 87
Configuring protection for SQL Server database................................88
Unsubscribing a database from a service plan................................. 88
Discovering SQL Server databases .................................................. 88
Creating a database copy from the Copies page............................... 89
Expiring an SQL database copy on demand...................................... 89
Service plan summary and details....................................................90
Mount considerations for SQL Server.............................................................99
4
EMC AppSync 2.2.3 User and Administration Guide
CONTENTS
Mount SQL Server database copy on-demand................................ 100
Unmounting an SQL Server copy ....................................................105
SQL Server database restore overview.........................................................105
Restore considerations for databases in an Availability Group........106
Affected entities during restore...................................................... 106
Restoring a primary database or a secondary database with failover
...................................................................................................... 107
Restoring a secondary database without failover............................107
Restoring a SQL Server copy...........................................................108
Restoring an SQL Server copy on XtremIO.......................................110
SQL Server restore utility (assqlrestore)..........................................114
Repurposing SQL Server database copies....................................................117
Using the Repurpose wizard...........................................................119
Chapter 7
Protect Oracle
121
Overview of Oracle support......................................................................... 122
Oracle permissions........................................................................ 122
Red Hat Cluster Services Integration with AppSync ........................ 122
Oracle Data Guard support............................................................. 123
Veritas Cluster Services integration................................................ 125
HACMP cluster integration..............................................................126
Prerequisites and supported configurations................................... 128
Protecting a database................................................................................. 131
Discovering databases...................................................................131
Subscribe a database to a service plan.......................................... 132
Oracle copies page........................................................................ 132
Service plan summary and details............................................................... 134
Service plan schedule.................................................................... 135
Overriding service plan schedules..................................................135
Application discovery.....................................................................136
Application mapping......................................................................136
Storage preferences....................................................................... 136
Pre-copy script............................................................................... 136
Create copy....................................................................................137
Automatic expiration of copies.......................................................138
Post-copy script............................................................................. 138
Unmount previous copy................................................................. 139
Pre-mount script............................................................................ 139
Mount copies ................................................................................ 140
Overriding mount settings in a service plan.................................... 140
Post mount script...........................................................................141
Unmount copy................................................................................141
Mount an Oracle copy................................................................................. 142
Mounting a copy using the Oracle Mount wizard............................ 142
RMAN cataloging feature ............................................................... 145
Mount on standalone server and prepare scripts for manual recovery
...................................................................................................... 145
Mount on cluster and recover......................................................... 147
Mount/unmount VMAX 3 copies.....................................................147
Restoring an Oracle copy.............................................................................148
Affected entities during restore...................................................... 149
Vdisk restore with affected entities................................................ 150
Restoring a RAC copy..................................................................... 150
Restoring a RAC copy for affected entities.......................................151
Restore a copy from XtremIO.......................................................... 152
EMC AppSync 2.2.3 User and Administration Guide
5
CONTENTS
Repurposing overview................................................................................. 156
Using the Repurpose wizard...........................................................159
Mounting and recovering an Oracle clone of clone of RecoverPoint
Bookmark ..................................................................................... 160
Mounting and recovering an Oracle Snap of Clone RecoverPoint
Bookmark ..................................................................................... 160
Chapter 8
Protect file systems
163
Overview of file system support...................................................................164
Protect NFS file systems on VNX storage.........................................164
Summary of file system service plan settings.............................................. 165
Subscribing a file system to a service plan..................................... 167
Overriding service plan schedules ................................................. 167
Service plan schedule.................................................................... 168
Application discovery.....................................................................168
Application mapping......................................................................168
Pre-copy script phase.....................................................................169
Create copy phase features freeze and thaw callout scripts............169
Post-copy script phase................................................................... 170
Unmount previous copy................................................................. 171
Mount copy....................................................................................171
Post-mount script...........................................................................173
Mounting a copy with the File System Mount wizard.................................... 173
Changing the mount point for an affected file system..................... 174
Unmounting a file system copy ......................................................174
Override mount settings in a service plan.......................................175
Restoring a file system................................................................................ 175
Restore a file system copy manually on XtremIO.............................176
Chapter 9
Protect VMware Datacenters
179
Configuration prerequisites ........................................................................ 180
VMware vStorage VMFS requirements ........................................... 180
Discovering datacenters ............................................................................. 183
List of datacenters .........................................................................183
Adding a VMware vCenter Server ................................................... 184
List of VMware datastores ............................................................. 184
Protect a VMware datastore............................................................184
Considerations when mounting a VMFS copy ..............................................192
Mounting a datastore copy on-demand.......................................... 192
Unmounting a VMware datastore copy .......................................... 194
Restoring a datastore from a copy............................................................... 194
Virtual Machine Operations during restore..................................... 195
Datastore affected entities during restore.......................................195
Restoring a VMware Datastore copy from XtremIO.......................... 196
Restoring a virtual machine from a copy...................................................... 198
Virtual Machine Restore options.....................................................200
File or folder restore with VMFS or NFS datastores....................................... 201
Restoring a file or folder from a virtual disk.................................... 202
Chapter 10
Monitor AppSync
205
RPO concepts and best practices................................................................ 206
Recovery point compliance report.................................................. 206
Exporting an RPO compliance report to CSV....................................206
6
EMC AppSync 2.2.3 User and Administration Guide
CONTENTS
Summary of RPO compliance......................................................... 207
Alerts and associated events.......................................................................207
Acknowledging alerts..................................................................... 207
Sending alerts via email.............................................................................. 208
Configuring server settings for email alerts.....................................208
Specifying email alert recipients.................................................... 209
Repurpose Monitor......................................................................................209
EMC AppSync 2.2.3 User and Administration Guide
7
CONTENTS
8
EMC AppSync 2.2.3 User and Administration Guide
CHAPTER 1
Introduction
This chapter includes the following topics:
l
l
l
AppSync overview ................................................................................................ 10
AppSync architecture............................................................................................ 11
Differences between AppSync and Replication Manager........................................14
Introduction
9
Introduction
AppSync overview
EMC AppSync offers a simple, SLA-driven, self-service approach for protecting, restoring,
and cloning critical Microsoft and Oracle applications and VMware environments. After
defining service plans (such as Gold, Silver, and Bronze), application owners can protect,
restore, and clone production data quickly with item-level granularity by using the
underlying EMC replication technologies. AppSync also provides an application
protection monitoring service that generates alerts when the SLAs are not met.
AppSync supports the following applications and storage arrays:
l
Applications — Oracle, Microsoft SQL Server, Microsoft Exchange, and VMware VMFS
and NFS datastores and File systems.
l
Storage — VMAX, VMAX 3, VNX (Block and File), VNXe, XtremIO, and ViPR Controller
l
Replication Technologies—VNX Advanced Snapshots, VNXe Unified Snapshot, SRDF,
TimeFinder, SnapVX, RecoverPoint, XtremIO Snapshot, and ViPR Snapshot .
Overview of service plans
AppSync protects an application by creating copies of application data.
You indicate to AppSync what you want to protect by subscribing an application object to
a service plan. When the service plan runs, a copy is created. The service plan can also
mount and unmount the copy, validate it, and run user-created scripts. These actions are
called phases of the service plan and may differ between applications.
AppSync includes several application-specific plans that work without change. With the
Subscribe to Plan and Run command, you apply the settings of a service plan to the data
and protect it immediately.
Note
These service plans are provided by default. However, you are not restricted to the
settings in these plans, you can customize these plans and create your own custom
plans.
There are three types of service plans that you can use:
l
Bronze — You can use the Bronze service plan to create local copies of your
applications.
l
Silver —You can use the Silver service plan to create local and remote copies of your
applications.
l
Gold —You can use the Gold service plan to create both local and remote copies of
your applications.
Service plans also set the preferred order of storage technology to use while creating
copies, for example, VNX Snapshot or VMAX clone/snap or RecoverPoint Bookmark.
All service plans can set the period for automatic expiration of copies. When you
subscribe an object to a service plan, it joins any other objects that are already part of the
plan. All objects in the service plan are subject to the workflows and settings that are
defined in the service plan. (The exceptions are the overrides specific to startup or
mounting a copy.)
Service plans also offer a few application-specific copy options. For example, Oracle
service plan has the following copy options:
l
10
Places database in hot-backup mode (Default: enabled )
EMC AppSync 2.2.3 User and Administration Guide
Introduction
l
Copies the Fast Recovery Area (Default: disabled)
l
Indexes and copies BCT (block change tracking) file (Default: disabled)
l
Creates backup control file (Default: disabled)
The default service plans offer tiered levels of protection. If you want to modify a service
plan, change its settings to match the requirements, or create a new service plan.
Role-based management
AppSync supports role-based access to resources and functionality.
You can set up AppSync to have multiple users. Each user can be assigned one or more
roles that correspond to their responsibilities and requirements. You can create users
that are local to AppSync, and optionally add users through an LDAP server which
handles the authorization.
Refer to the EMC AppSync Security Configuration Guide available on the EMC Support
website for specific user roles and their permissions.
Reports show proven protection
AppSync generates reports that tell you whether your data is protected, recoverable, and
compliant with service level agreements.
The reports included with AppSync work without modification. Alerts and reports can be
easily viewed at the top level of the AppSync dashboard. Alerts can be sent in email.
AppSync can export reports to comma-separated value format.
AppSync architecture
AppSync components include the AppSync server, agent (host plug-in software), and user
interfaces (UI or console).
Figure 1 AppSync architecture flow
Role-based management
11
Introduction
AppSync server
The AppSync server software resides on a supported Windows system. It controls the
service plans and stores data about each copy it creates.
The repository is stored in a PostgreSQL database on the AppSync server.
AppSync agent (host plug-in) overview
AppSync installs light-weight agent plug-in software on the production and mount hosts.
AppSync pushes the plug-in software from the AppSync server to the host when you add
the host as a resource. In an environment that prevents the AppSync server from
accessing a host, you can install the agent plug-in manually.
Note
With a push-install, the agent host plug-in remains at the same version as the AppSync
server. If you want to upgrade the agent host plug-in, select Update plug-in with Settings
> Servers.
For UNIX, tar bundles for AIX and Linux are pushed and extracted on the host during host
registration.
Examples of hosts where the plug-in resides are Exchange mailbox servers or Exchange
validation or mount hosts. The agent plug-in is not used for protection of VMware data
stores.
Supported device configurations
For the latest versions, see the EMC AppSync Support Matrix .
Device mapping using:
Table 1 Supported device configurations
AIX
LINUX
Multipath: Multipathing
package is required for
platforms running VNX, VMAX,
or RecoverPoint on VNX in nonvirtualized environments.
Multipath: Multipathing package is required for platforms
running VNX, VMAX, or RecoverPoint on VNX in non-virtualized
environments.
l
Native MPIO
l
PowerPath
Device mapping using:
12
l
Native MPIO
l
PowerPath
l
Veritas DMP
n
Supported where both production and mount host
have DMP installed and DMP devices do not have a
partition.
n
Supported with non-ASM database on VXDMP devices
with or without VXVM Volume manager
Device mapping using:
l
Symbolic links
l
Linux-ASMLib (ASM only)
l
mknod devices
l
UDEV bindings
l
Block devices
l
Dev-mapper devices
l
Block devices with O_DIRECT
EMC AppSync 2.2.3 User and Administration Guide
Introduction
Table 1 Supported device configurations (continued)
AIX
LINUX
File systems:
l
jfs and jfs2
Volume Manager:
l
Native LVM
l
Symbolic links
l
ext2, 3, 4
l
XFS (necessary rpms needed on production as well as
mount host)
l
VxFS (When using VxVM)
Volume Manager:
l
Native LVM
l
VxVM Supported device configuration:
n
Non ASM database on VXVM Volume Group
The EMC AppSync Installation and Configuration Guide provides more information on
installing the AppSync agent.
AppSync Console (user interface)
The AppSync console is web-based. Supported browsers are Chrome, Internet Explorer,
and Firefox. Refer to an AppSync support matrix for supported versions.
There is an issue with running Flash on Windows Server 2012 R2. Adobe Flash is only
installed and enabled when you enable the Desktop Experience. It is not necessary to
enable Flash Player on Windows 2008 R2. For Windows 2008 R2, go to the Adobe website
to download and install the latest version of Flash Player.
For Windows Server 2008 R2 or Windows Server 2012 R2 you need to turn off Internet
Explorer Enhanced Security Configuration. Close any open Internet Explorer windows.
Review these scenarios:
1.
l
If you are running Windows Server 2008 R2, browse to the Security Information
section of Server Summary, and then click Configure IE ESC to open the Internet
Explorer Enhanced Security Configuration dialog.
l
If you are running Windows Server 2012, click Configure this local server to open
the Local Server configuration page. Browse to Properties, next to IE Enhanced
Security Configuration, and then click On to open the Internet Explorer Enhanced
Security Configuration dialog.
2. To edit Internet Explorer Enhanced Security Configuration when members of the local
Administrator group are logged on, browse to Administrators, and then click Off.
3. To edit Internet Explorer Enhanced Security Configuration when all other users are
logged on, browse to Users, and then click Off (Recommended).
4. Click OK to apply your changes.
Additionally, enable the Desktop Experience on Windows Server 2012. Follow these
steps:
1. Open Server Manager and click Add Roles and Features.
2. When the Add Roles and Features Wizard appears, specify the values on the
Installation Type, Server Selection, and Server Roles pages.
3. On the Features page, expand User Interfaces and Infrastructure and select Desktop
Experience.
AppSync Console (user interface)
13
Introduction
4. On the Confirmation page, select Restart the destination server automatically if
required and click Install.
REST interface
AppSync has a REST interface that allows application programmers to access information
controlled by AppSync.
The API is described in the AppSync REST API Reference Guide.
Differences between AppSync and Replication Manager
Although they are related products, there are important differences between AppSync
and Replication Manager.
AppSync is primarily different from Replication Manager in that it uses the concept of
service plans (protection policies) that are designed to meet specific service level
requirements. The service plans are fully customizable and can be applied to the
application with a single click. AppSync is easy to use for applications administrators as
well as storage administrators, because it does not require prior knowledge in data
replication technology.
Interoperability of AppSync and Replication Manager
Replication Manager Server and AppSync Server cannot exist on the same host.
The Replication Manager client and the AppSync agent plugin can be installed on the
same host and can co-exist together.
14
EMC AppSync 2.2.3 User and Administration Guide
CHAPTER 2
AppSync Console
The chapter includes the following topics:
l
l
l
Console overview.................................................................................................. 16
Setting preferences............................................................................................... 17
Starting the AppSync console................................................................................17
AppSync Console
15
AppSync Console
Console overview
The AppSync console is arranged in sections for management, reporting, and
administration.
l
The Dashboard is a customizable view of reports and alerts. The default dashboard
shows recovery point objective (RPO) status of protected applications, service plan
completion status, most recent alerts, and activity in progress.
l
Copy Management lists discovered applications such as Microsoft Exchange and
Microsoft SQL Server, and provides the application-oriented entry point for
protection, mount, restore, service plan subscription, and other operations.
l
The Service Plans tab lets you view and modify service plan settings, view lists of
objects that are subscribed to a service plan, the copies that were made of those
objects, and the events that were generated when the service plan was run.
l
Monitoring displays alerts, recovery point compliance reports, and service plan
completion reports.
l
Settings is for adding servers, VMware vCenter servers, and storage resources to
AppSync. License management, user administration, and server settings are also
found under Settings.
l
Support provides access to how-to videos and the AppSync support page on the EMC
Support website.
User roles control which sections of the console are displayed and which operations are
listed in menus. For example, the console does not display the Copy Management and
Service Plans tabs for a user who has only the Security Administrator role.
Performing actions
To perform an action on an object, select the object and click the action buttons at the
bottom of the page. Use the Shift key to select multiple objects. Use Ctrl to select
noncontiguous objects. To perform an action on multiple objects, click the action button
on the final selection.
Examples of objects are application hosts, storage systems, mailbox servers, databases,
copies and users.
Times shown in the console
Times shown in the AppSync console reflect the local time of the AppSync server, not of
the console.
Troubleshooting console problems
Unexpected behavior by the AppSync console may be related to the browser and its
supporting add-ons and extensions.
Check the latest support matrix for supported browser versions.
Troubleshooting browser display problems
Learn how to resolve data display issues in the AppSync console.
Symptom: AppSync displays stale or corrupted data.
Resolution:
16
EMC AppSync 2.2.3 User and Administration Guide
AppSync Console
1. Clear browser cache. Refer to the browser sepecific documentation for more
information.
2. Clear the Java cache:
a. Select Start > Control Panel (Windows 7) or Start > Control Panel > Programs
(Windows 2008/2012).
b. Double-click the Java icon.
c. Select Settings under Temporary Internet Files.
d. Select Delete Files.
e. Select all available options and select OK.
Troubleshooting browser refresh issues
Symptom: Some contents in the AppSync console are not displayed.
Cause: You refreshed the browser using the browser's Refresh icon or pressing F5 on your
keyboard.
Resolution: To refresh content in the AppSync console, use the Refresh icon on the
console's top right corner.
Setting preferences
You can set the language displayed in the user interface and optimize the console for use
over remote connections.
Procedure
1. Select the language of choice from the list of installed languages.
If you change the language, you must close the browser and launch AppSync again for
the settings to take effect.
2. Check Optimize for Remote Connection to optimize the user interface for use over
remote connections.
For example, certain visual effects may cause slow screen painting when you remotely
access the console. Selecting this option enhances the remote response time but
does not affect management options or functionality.
Starting the AppSync console
You can run the AppSync console on a supported web browser from any system that has
connectivity to the AppSync server.
Use http://appsync_server:8085/appsync to start the console.
If you are running the console on the AppSync server, you can start the console by
clicking on the AppSync shortcut on the installer's desktop. appsync_server must be a
host name, not an IP address.
Troubleshooting browser refresh issues
17
AppSync Console
18
EMC AppSync 2.2.3 User and Administration Guide
CHAPTER 3
Service Plans
This chapter includes the following topics:
l
l
l
l
l
l
Service plan overview............................................................................................20
Summary of Exchange service plans settings.........................................................24
Summary of SQL Server service plans settings.......................................................26
Summary of Oracle service plan settings............................................................... 29
Summary of file system service plan settings.........................................................31
Summary of VMware service plans settings........................................................... 33
Service Plans
19
Service Plans
Service plan overview
Learn about default service plan types, object and copy subscriptions, service plan
settings, schedules, and overrides.
For information on how each array type supports service plans, refer to the section on
prerequisites in the EMC AppSync Installation and Configuration Guide.
AppSync creates and manages copies of application data. A service plan defines the
attributes of these copies. You can subscribe application data objects to a service plan,
then AppSync runs the service plan and creates copies of the data from attributes that
you specified in the plan. Copies that are generated by a service plan are listed in service
plan Copies tab.
There is no limit to the number of objects you can subscribe to a service plan. AppSync
automatically divides up the work for best performance. If you need fine control over
which objects are grouped for mounting, scripting, and validating, consider creating
multiple service plans and distributing objects among the plans. This technique works
when the objects subscribed to a service plan are from the same server. Subscribe no
more than 12 objects to any one service plan when using this method.
Service plan types
AppSync provides the following application-specific tiered plans. There are three types of
service plans:
l
Bronze — You can use the Bronze service plan to create local copies of your
applications.
l
Silver — You can use the Silver service plan to create remote copies of your
applications.
l
Gold — You can use the Gold service plan to create both local and remote copies of
your applications.
Note
Ensure you understand the storage capabilities when selecting a service plan type. Not
all storage technologies support Remote Replication, so Silver or Gold service plans may
not be successful for the application data.
Bronze, Silver and Gold service plans are provided by default, however you can
customize and create your own plans.
The following table describes the service plans and applications supported.
Storage
Replication
type
Bronze
Silver
Gold
RP support Application
support
Repurposing
support
VNX
Advanced
Snapshot
Yes
No
No
Yes
All applications
AppSync supports
Yes
File Snapshot
Yes
Yes
Yes
No
VMware datastores,
File systems, and
Oracle
No
VP Snap
Yes
No
No
Yes
All applications
AppSync supports
Yes
Timefinder Clone Yes
No
No
Yes
All applications
AppSync supports
Yes
VMAX
20
EMC AppSync 2.2.3 User and Administration Guide
Service Plans
Storage
Replication
type
Bronze
Silver
Gold
RP support Application
support
Repurposing
support
SRDF/A a
No
Yes
No
Yes
VMware datastores
and Oracle
Yes
SRDF/S
No
Yes
No
Yes
All applications
AppSync supports
Yes
VNX2eb
Unified
Snapshot
Yes
No
No
Yes
All applications
AppSync supports
No
XtremIO
Snapshot
Yes
No
No
Yes
All applications
AppSync supports
Yes
ViPRc
Snapshot
Yes
No
No
No
VMware datastores, No
Oracle, and UNIX file
systems
VMAX 3
SnapVX Snap
Yes
No
No
No
All applications
AppSync supports
Yes
SnapVX Clone
Yes
No
No
No
All applications
AppSync supports
Yes
SRDF/A d
No
Yes
No
No
VMware datastores
and Oracle
Yes
SRDF/S
No
Yes
No
No
All applications
AppSync supports
Yes
Yes
No
No
Yes
All applications
AppSync supports
Yes
Remote
bookmark
No
Yes
No
Yes
All applications
AppSync supports
Yes
Local and
remote
bookmark
No
No
Yes
Yes
All applications
AppSync supports
No
RecoverPoint Local bookmark
a.
b.
c.
d.
AppSync does not support Windows applications on VMAX 3 and SRDF/A storage. If you create a copy on Windows applications, the
created copies are non-application consistent.
File systems and Oracle NFS are not supported on VNX2e file storage.
AppSync also supports ViPR snapshots as copy technology for applications provisioned using ViPR Controller. AppSync only
supports applications provisioned by ViPR with block virtual pools that are backed only by VMAX/VPLEX (with VMAX2 and XtremIO)/
XtremIO storage systems.
AppSync does not support Windows applications on VMAX 3 and SRDF/A storage. If you create a copy on Windows applications, the
created copies are non-application consistent.
Service plan settings
When you subscribe an object to a service plan, it joins other objects that are already part
of the plan. All objects in the service plan are subject to the workflow and settings that
are defined in the service plan.
Service plans set a storage ordered preference which is the preferred order of storage
technology the service plan uses when creating copies. If AppSync cannot satisfy a
preference, it tries to use the next preference in the storage ordered preference list. You
can adjust the preferences to create service plans that use the replication technology you
want.
The default service plans offer tiered levels of protection. If you must change settings,
modify the service plan.
Service plan overview
21
Service Plans
Any service plan can set the automatic expiration of copies which limits the number of
copies that AppSync keeps, and automatically expires older copies that exceed the
number that is defined for the service plan.
Service plans also offers a few application specific copy options which can be modified.
For example, Oracle service plan has the following copy options:
l
Place a database in hot-backup mode (Default: enabled )
l
Copy the Fast Recovery Area (Default: disabled)
l
Index and copy BCT (block change tracking) file (Default: disabled)
l
Create backup control file (Default: disabled)
To avoid overutilization and depletion of replication storage, when you set up a service
plan, set values in the following fields:
l
RPO in the service plan Startup phase
l
Always keep n Copies in the Create copy phase
Also, monitor the storage system with the storage system user interface on the AppSync
console.
Service plan schedule overrides
You can override a service plan's run schedule settings and specify separate schedules
for individual objects that are subscribed to the plan.
In the Plan Startup phase of the service plan, you select a recurrence type that is based
on which service plan is triggered. This recurrence type is applicable for all application
objects that are subscribed to a service plan. However, you can override the settings and
specify separate settings for selected objects.
You can override only the settings of the Recurrence Type already selected for the plan.
For instance, the chosen recurrence type is to run On selected days... and the settings are
to Run at 12:00 AM on days Fri, Sat. When you override these settings for an object, you
can change only the time and days of the week. You cannot select a different recurrence
type as part of the override.
As the Service Plan Administrator, if you change the generic recurrence settings (such as
the time to run, or minutes after the hour), there is no impact to the settings of the
overrides. If you change the recurrence type itself, then the overrides are no longer valid.
The new recurrence type now applies to all objects until you specify individual settings
that are based on the new recurrence type.
Note
If an application object is subscribed to multiple plans, the plans must not be scheduled
to be running simultaneously.
Service plan events
Events show the progress of an operation. They are generated when a service plan is run,
and when a copy is mounted or restored.
Event information includes:
l
Type (error, warning, or informational)
l
Date and time of the event
l
Description
l
Server
You can view events at:
22
EMC AppSync 2.2.3 User and Administration Guide
Service Plans
l
Service plan Events tab. For example, on the AppSync console, the Events tab in
Service Plans > Microsoft Exchange > Bronze shows you the events that are related to
the Exchange copy under the plan.
l
Application Copies page. For example, go to Copy Management > VMware
Datacenters select a data center, then a datastore to view the Copies page. Select a
copy to view its associated events in the Events page.
l
Events are also displayed at the time they are generated in the Mount and Restore
wizards and when an object is subscribed to a plan and run immediately.
By default only the top level events, which are known as milestone events, are displayed.
A milestone event is generated at the completion of each phase in a service plan cycle.
You can expand a milestone event to show the other events that were generated in the
phase.
Creating a service plan
You can create a new service plan by using an existing plan as a template.
Before you begin
This operation requires the Service Plan Administrator role in AppSync.
Procedure
1. Select Service Plans.
2. Click Create.
3. In the Create New Plan dialog box, select an existing plan to use as a template. Enter
a name and a description for the new service plan.
Note
The new service plan contains the same schedule and other settings as the template,
but there are no objects subscribed to the new service plan.
Running a service plan on demand
Service plans run on a schedule but you can also run a service plan on demand.
Before you begin
This operation requires the Data Administrator role in AppSync.
Procedure
1. Select Service Plans.
2. Select the application to protect.
3. Select a plan from the list and click Run.
This service plan run is applicable to all the application objects currently subscribed
to the plan.
The service plan runs immediately. The Run Service Plan dialog displays progress as
application storage is discovered and mapped, and application protection begins
according to service plan settings.
4. Click Details to see more events that occurred during each phase.
Creating a service plan
23
Service Plans
Disabling and reenabling a service plan
By default all service plans are enabled. You can disable and reenable a service plan.
Before you begin
This operation requires the Service Plan Administrator role in AppSync.
Procedure
1. Select Service Plans.
2. Select the application.
3. Select the plan and click Disable or Enable.
Deleting a service plan
You can delete a user-created service plan.
Before you begin
l
This operation requires the Service Plan Administrator role in AppSync.
l
You cannot delete a built-in service plan (for example, Bronze, Silver, Gold).
l
You cannot delete a service plan if the plan has subscriptions or if there are valid
copies associated with the plan.
Procedure
1. Select Service Plans.
2. Select the application.
3. Select a user-created plan and click Delete.
4. Click Yes to confirm.
Summary of Exchange service plans settings
The default service plan settings create an application-consistent copy every 24 hours.
Only the replication technology, which is specified by the Copy type in the Create copy
phase, is different from plan to plan.
Table 2 Exchange Service Plan - default settings
Setting
Enabled/Not
enabled
Default settings
Schedule
Plan Startup
Enabled
Automatic schedule
Recurrence type:
Creates a copy
every 24 hours,
with the first run
at midnight
(00:00).
Recovery Point
Objective (RPO):
A copy should
be created every
24 hours. (Alert
is issued if
24
EMC AppSync 2.2.3 User and Administration Guide
Service Plans
Table 2 Exchange Service Plan - default settings (continued)
Setting
Enabled/Not
enabled
Default settings
Schedule
objective is not
met.)
Application
discovery
Enabled
None
Determined by
Plan Startup
phase
Application
mapping
Enabled
None
Starts when
Application
discovery
phase
completes
Pre-copy script
Not enabled
None
Starts when
Application
mapping phase
completes
Create copy
Enabled
l
Starts when Pre-
Copy type is:
n
Bronze
n
Silver
n
Gold
copy script
phase
completes
l
Exchange backup type: Full, Copy or Differential
l
Storage Ordered Preference: Snapshot, Bookmark, and Clone
l
Storage Settings:
n
Include RecoverPoint copies in expiration rotation policy: Check this
option to include RecoverPoint copies when calculating rotations.
Note
If this option is not checked, then RecoverPoint copies accumulate,
and remain until the bookmarks for them "fall off" the RecoverPoint
appliance.
l
Post-copy
script
Not enabled
Event Log Scanning:
n
Fail on -1018 error
n
Fail on -1019 error
n
Fail on -1022 error
n
Fail on Event ID 447
n
Fail on Event ID 448
n
Do not allow databases and logs to reside on the same volume
None
Starts when
Create copy
phase
completes
Summary of Exchange service plans settings
25
Service Plans
Table 2 Exchange Service Plan - default settings (continued)
Setting
Enabled/Not
enabled
Default settings
Schedule
Unmount
previous copy
Not enabled
None
Starts when
Post-copy
script phase
completes
Mount copy
Not enabled
l
Mount on server: Original Host
l
Mount with access: Read-only
l
Mount Path: Default Path
Starts when
Unmount
previous copy
phase
completes
Note
The drive that is specified for mount cannot be a clustered disk.
Validate copy
Post-mount
script
Not enabled
Not enabled
l
Copy metadata files to: Default Path
l
Image access mode: Logged access
l
Check databases and logs in parallel
l
Do not minimize log checking
l
Do not perform throttle checking
None
Starts when
Mount copy
phase
completes
Starts when
Validate copy
phase
completes
Unmount copy
Not enabled
None
Starts when
Post-mount
script phase
completes
Summary of SQL Server service plans settings
Summary of SQL Server service plans settings
Table 3 SQL Server Service Plan - default settings
Setting
Enabled/Not
enabled
Default settings
Schedule
Plan Startup
Enabled
Automatic schedule
Recurrence type:
Creates a copy
every 24 hours,
with the first run
at midnight
(00:00).
Recovery Point
Objective (RPO):
A copy should
26
EMC AppSync 2.2.3 User and Administration Guide
Service Plans
Table 3 SQL Server Service Plan - default settings (continued)
Setting
Enabled/Not
enabled
Default settings
Schedule
be created every
every 24 hours.
(Alert is issued if
objective is not
met).
Application
discovery
Enabled
None
Determined by
Plan Startup
phase
Application
mapping
Enabled
None
Starts when
Application
discovery phase
completes
Pre-copy script
Not enabled
None
Starts when
Application
mapping phase
completes
Create copy
Enabled
l
l
Starts when Precopy script
phase
completes
Copy type is:
n
Bronze
n
Silver
n
Gold
SQL Server Backup Type: Full or Copy
n
Auto Switch to Copy
n
Enable Log Backup
Note
This option must be selected to permit on demand log backup runs.
l
Storage Ordered Preference: Snapshot, Clone, Bookmark
l
Expiration of database copies:
n
Include RecoverPoint copies in expiration rotation policy: Check this
option to include RecoverPoint copies when calculating rotations.
Note
If this option is not checked, then RecoverPoint copies will
accumulate, and will remain until the bookmarks for them "fall off"
the RecoverPoint appliance.
l
Transaction Log backup options:
n
Schedule immediately after database backup = default.
n
Every 15 or 30 minutes or every 1 to 24 hours are other options.
Summary of SQL Server service plans settings
27
Service Plans
Table 3 SQL Server Service Plan - default settings (continued)
Setting
Enabled/Not
enabled
Default settings
Schedule
Note
Scheduled log backups run during times between database
backups.
n
Backup path: Default path
n
Free space on the volume: 5GB
n
Backup group size: 5
n
Truncate logs: Selected
n
Checksum the backup: Unselected
n
Compression: Selected
n
Expiration of log backups: Minimum retention hours 24
Post-copy
script
Not enabled
None
Starts when
Create copy
phase
completes
Unmount
previous copy
Not enabled
None
Starts when
Post-copy script
phase
completes
Mount copy
Not enabled
l
Mount Copy
n
Mount on Server: Original Host
n
Mount with access: Read only
n
Mount Path: Default Path
Starts when
Unmount
previous copy
phase
completes
Note
The drive specified for mount can not be a clustered disk.
n
Copy metadata files to: Default Path
n
Image Access mode: Logged access
Post-mount
script
Not enabled
None
Starts when
Mount copy
phase
completes
Unmount copy
Not enabled
None
Starts when
Post-mount
script phase
completes
Pre-log backup
script
Not enabled
None
Starts after log
backup
Post-log
backup script
Not enabled
None
Starts after log
backup
28
EMC AppSync 2.2.3 User and Administration Guide
Service Plans
Table 3 SQL Server Service Plan - default settings (continued)
Summary of Oracle service plan settings
Use this list of service-plan default settings for Oracle databases in AppSync.
Table 4 Oracle Server Service Plan - default settings
Setting
Enabled/Not
enabled
Default settings
Schedule
Plan Startup
Enabled
Automatic schedule
Recurrence type:
AppSynccreates a
copy every 24
hours, with the
first run at
midnight (00:00).
Recovery Point
Objective (RPO):
Creates a copy
every 24 hours.
Application
discovery
Enabled
None
Determined by
Plan Startup
phase
Application
mapping
Enabled
None
Starts when
Application
discovery phase
completes.
Pre-copy script
Enabled
None
Starts when
Application
mapping phase
completes
Create copy
Enabled
The create copy options on the service plan settings provides various
controls which influence how the Oracle copy is created. Copy types
include:
Starts when
Precopy script
phase completes.
l
Bronze
l
Silver
l
Gold
Create copy setting options include:
l
Place database in hot-backup mode (Default: enabled )
When enabled, the protection puts the database in hot backup and
immediately creates copies of the archive logs. If you disable this
option, the database is not placed in hot backup mode and the copy is
created from the live unquiesced data without any instrumentation of
the database.
l
Copy the Fast Recovery Area. (Default: disabled)
Summary of Oracle service plan settings
29
Service Plans
Table 4 Oracle Server Service Plan - default settings (continued)
Setting
Enabled/Not
enabled
Default settings
Schedule
When enabled, this field tells AppSync to create a copy of the
underlying storage that is used by the FRA when protecting the
database's archive log files.
l
Index and copy the BCT (block change tracking) file. (Default: disabled)
If enabled, AppSync creates an entry in the Oracle block change
tracking file and re-copies the file as part of the protection. This file can
then be leveraged as part of a mount and backup use-case to provide
accelerated incremental backup. This option requires hot backup
mode.
l
Create backup control file for RMAN cataloging. (Default: disabled)
If enabled, AppSync creates a binary backup control file with a request
to catalog the database contents in a remote RMAN catalog. This
option requires hot backup mode.
l
Expiration: Include RecoverPoint copies in expiration rotation policy:
select this option to include RecoverPoint copies when calculating
rotations. If you do not select this option, RecoverPoint copies
accumulate and remain until their bookmarks rotate off the
RecoverPoint appliance.
Note
If RecoverPoint copies are factored in the rotation policy, bookmarks
are created with the ALWAYS_CONSOLIDATE policy. Otherwise,
bookmarks are created with the NEVER_CONSOLIDATE policy. Consult
the RecoverPoint documentation for a definition of these consolidation
policies.
Post-copy
script
Not enabled
None
Starts when
Create copy
phase completes.
Unmount
previous copy
Not enabled
None
Starts when Postcopy phase
completes
Pre-mount
script
Not enabled
None
Starts when
Unmount
previous copy
phase completes
appliance.
Mount and
Recovery
Not enabled
Mount and recovery options:
Starts when Premount script
phase completes.
30
l
Mount on standalone server (RM-equivalent : No recover)
l
Mount on standalone server and create RMAN catalog entry (RMequivalent : Catalog with RMAN )
l
Mount on standalone server and recover database (RM-equivalent:
Recover)
EMC AppSync 2.2.3 User and Administration Guide
Service Plans
Table 4 Oracle Server Service Plan - default settings (continued)
Setting
Enabled/Not
enabled
Default settings
Schedule
l
Mount on standalone server and prepare scripts for manual database
recovery (RM-equivalent: Prepare-only/generate scripts for manual
recovery)
l
Mount on grid cluster and recover as RAC database (RM-equivalent:
Mount as RAC database)
Post-mount
script
Not enabled
None
Starts when
mount phase
completes.
Unmount copy
Not enabled
None
Starts when Postmount script
phase completes.
Summary of file system service plan settings
Use this table to learn default file system settings for service plan phases including
startup, discovery, mapping, pre and post copy scripting, mount/unmount and copy.
Default service plan settings create an application-consistent copy every 24 hours. Only
the replication technology that is specified by the Copy type in the Create copy phase
varies among plans. The following table summarizes the default settings:
Table 5 Default file system Service Plan Settings
Setting
Enabled/No Default settings
t enabled
Schedule
Plan Startup
Enabled
Automatic schedule
Recurrence type: Creates
a copy every 24 hours,
with the first run at
midnight (00:00).
Recovery Point Objective
(RPO): A copy should be
created every 24 hours.
(Alert issued if objective
is not met.)
Application
discovery
Enabled
None
Determined by Plan
Startup phase.
Application
mapping
Enabled
None
Starts when Application
discovery phase
completes.
Pre-copy script
Not enabled
None
Starts when Application
mapping phase
completes.
Summary of file system service plan settings
31
Service Plans
Table 5 Default file system Service Plan Settings (continued)
Setting
Enabled/No Default settings
t enabled
Schedule
Create copy
Enabled
Starts when Pre-copy
script phase completes.
Copy type:
l
Bronze
l
Silver
l
Gold
Also:
32
l
Storage Ordered Preference:
Snapshot, Clone, and
Bookmark.
l
Storage Settings: Include
RecoverPoint copies in
expiration rotation policy—
select this option to include
RecoverPoint copies when
calculating rotations. If you
do not select this option,
RecoverPoint copies
accumulate and remain until
the bookmarks for them "fall
off" the RecoverPoint
appliance.
Post-copy script
Not enabled
None
Starts when Create copy
phase completes.
Unmount
previous copy
Not enabled
None
Starts when Post-copy
script phase completes.
Mount copy
(A pre-mount
script phase is
available for file
system service
plans)
Not enabled
Mount Copy
Starts when Unmount
previous copy phase
completes.
l
Mount on Server: Original
Host
l
Mount with access: Read/
write
l
Mount Path: Default Path
Image
l
Access mode: Logged access
l
Copy to Mount: Local (Only
for Gold Plans)
l
Use Dedicated Storage
Group: Selected by default
Post-mount script Not enabled
None
Starts when Mount copy
phase completes.
Unmount copy
None
Starts when Post-mount
script phase completes.
Not enabled
EMC AppSync 2.2.3 User and Administration Guide
Service Plans
Summary of VMware service plans settings
The default service plan settings create an application-consistent copy every 24 hours.
Only the replication technology, which is specified by the Copy type in the Create copy
phase, is different from plan to plan.
Table 6 VMware Service Plan - default settings
Setting
Enabled/Not
enabled
Default settings
Schedule
Plan Startup
Enabled
Automatic schedule
Recurrence type:
Creates a copy
every 24 hours,
with the first run
at midnight
(00:00).
Recovery Point
Objective (RPO):
A copy should
be created every
24 hours. (Alert
is issued if
objective is not
met.)
Application
discovery
Enabled
None
Determined by
Plan Startup
phase
Application
mapping
Enabled
None
Starts when
Application
discovery phase
completes
Create copy
Enabled
l
l
l
Starts when
Application
mapping phase
completes
Copy type is:
n
Bronze
n
Silver
n
Gold
Copy Consistency:
n
Virtual machine Consistent with a maximum of 4 simultaneous VM
snapshots
n
Ignore VM Snapshots for VMs: link to select individual virtual
machines to ignore while taking VMware snapshots during the
service plan run.
n
Include Virtual Machine Disk: Select this checkbox to protect virtual
machine disks spanning multiple data stores. By default, this option
is not selected.
Storage Ordered Preference: Snapshot, Clone, Bookmark
Summary of VMware service plans settings
33
Service Plans
Table 6 VMware Service Plan - default settings (continued)
Setting
Enabled/Not
enabled
Default settings
l
Schedule
Expiration: Include RecoverPoint copies in expiration rotation policy:
Check this option to include RecoverPoint copies when calculating
rotations.
Note
If this option is not checked, then RecoverPoint copies will accumulate,
and remain until the bookmarks for them "fall off" the RecoverPoint
appliance.
Unmount
previous copy
Not enabled
None
Mount copy
Not enabled
l
No default mount host
l
Mount using new signature
l
For RecoverPoint: mount with logged access, and mount local copy (in
case of local and remote copy plan)
l
For VNX file, mount copy with read-only or read/write access for local or
remote copies
Unmount copy
34
Not enabled
None
EMC AppSync 2.2.3 User and Administration Guide
Starts when
Create copy
phase
completes
Starts when
Unmount
previous copy
phase
completes
Starts when
Mount copy
phase
completes
CHAPTER 4
AppSync CLI Utility
This chapter includes the following topics:
l
AppSync CLI Utility tutorial.................................................................................... 36
AppSync CLI Utility
35
AppSync CLI Utility
AppSync CLI Utility tutorial
The AppSync CLI is a utility that is packaged with AppSync and is used for scripting or
running tasks through a command line interface
The AppSync CLI is installed in the EMC\AppSync\appsync-cli directory. You can
run it on Windows with the file appsync-cli.bat, and in UNIX using appsynccli.sh.
Pre-requisites
l
Java Runtime Environment (JRE) version 7 Update 6 ( jre1.7) and above - must be
installed and available in path.
l
Configured AppSync installation with registered resources
l
Discovered applications on registered hosts
l
Configured service plans
l
For I18N support on Windows, be sure that you set the correct code page before
execution.
Using the CLI
You can run the AppSync CLI on the server where the AppSync installation resides. Also,
you can move the \EMC\Appsync\appsync-cli directory to another location/host.
All actions that are performed, for scripting purposes, return code zero 0 for success and
local system failure code -1 for Windows or 255 for Linux). The syntax for using the
AppSync CLI follows:
appsync-cli.bat -action options=value
You preface the action that you want to perform with a hyphenated -argument. All
options specific to that action are key=value pairs. When using a value that contains
spaces such as a file system or path, you are not required to surround the text in double
quotations. Do not surround a value that ends with a trailing backslash with double
quotes. Java ignores this construct.
The AppSync CLI also has two optional arguments for message handling. At any point,
you can use the argument verbose=true for a more detailed messaging output, and
silent=true to suppress all messages.
The AppSync CLI supports the following actions:
36
l
Login/logout
l
Run a service plan
l
Enable/disable a service plan
l
List all copies that are created for a service plan or application object
l
List all details of an application object
l
Subscribe/unsubscribe an application object to/from a service plan
l
Mount/unmount a copy
l
Expire a copy
l
Run and export AppSync reports
EMC AppSync 2.2.3 User and Administration Guide
AppSync CLI Utility
Using the Help "/?" argument
To evoke a detailed help menu for a command, add the /? argument. This argument
displays all available CLI commands. Because of the complexity and vast number of
arguments, the CLI help uses the following help menu partitions:
l
l
appsync-cli.bat /? Returns information on all CLI-supported actions.
appsync-cli.bat -action /? Returns non-specific application options available
for the selected action.
l
appsync-cli.bat -action app=<value> /? Returns application-specific
options available for the selected action.
l
appsync-cli.bat -mount app=<value> option=<value> /? Returns
mount-specific options for the provided mount option.
Note
When using the help argument on non-English system locales you must enclose the help
argument "/?" in double quotation marks.
CLI actions
Log in/Log out
Before performing CLI actions, authenticate to an AppSync server. The log in command
requires server name, https communication port, AppSync user, and corresponding
password. For example:
appsync-cli.bat -login server=<server> port=8445 user=admin
password=<admin_pass>
After you log in, a file that is named LOCAL_TOKEN is created in the current directory
containing required authentication information. If this file is deleted or the current
session expires, a new session must be created by running the login command once
again.
After you complete actions with the AppSync CLI, ensure you log out. The log out
command not only closes the current session, but also invalidates it. For example:
appsync-cli.bat -logout
Run/Enable/Disable a service plan.
Enable, Disable, and Run a service plan by supplying the application name and service
plan. For example:
appsync-cli.bat -runSP app=sql service_plan=Bronze
appsync-cli.bat -enableSP app=sql service_plan=Bronze
appsync-cli.bat -disableSP app=sql service_plan=Bronze
Subscribe/Unsubscribe
You can subscribe an application object to a service plan with the CLI. Options vary for
each application. Run the help command "/?" for the application that you want to
subscribe/unsubscribe for a complete list of required arguments. For example:
appsync-cli.bat -subscribe app=oracle service_plan=<sp1>
oracle_server=<server> db_name=<db1>
appsync-cli.bat -unsubscribe app=sql service_plan=<sp1>
sql_server=<server> instance_name=<instance> db_name=<db1>
List Copies and Copy Details
AppSync CLI Utility tutorial
37
AppSync CLI Utility
A copy's uuid is required Before you can mount the copy. To get this information, run the
-listCopies command for either a service plan or an application object. Again, the
arguments are application-specific so be sure to use the help command "/?" for more
details.
Use the "age" argument to filter viewable copies on the console by the age of a copy. For
example:
appsync-cli.bat -listCopies app=sql service_plan=Bronze age=month
appsync-cli.bat -listCopies app=sql instance_name=<value>
db_name=<value> age=all
To see more details of an application copy you can run the -copyDetails command
and supply the copy's uuid. For example:
appsync-cli.bat -copyDetails app=<app> copy_ID=<value>
Mount/Unmount
The AppSync CLI supports all mount options that are available through the GUI. Options
vary for each application. Run the help "/?" command for the application you want to
mount to determine mount options. For example:
appsync-cli.bat -mount app=filesystem copy_ID=<value>
mount_host=<value>
appsync-cli.bat -mount app=sql copy_ID=<value> option=recover
recovery_instance=<value> point_in_time=<value>
appsync-cli.bat -mount app=datastore copy_ID=<value>
mount_host=<value> cluster_mount=yes image_access_mode=virtual_roll
appsync-cli.bat -mount app=oracle copy_ID=<value> option=rac
mount_cluster=<value> mount_servers=<server1,server2>
To unmount a copy you need the application name and the copy uuid. For example:
appsync-cli.bat -unmount app=<app> copy_ID=<value>
To unmount the latest or oldest mounted copy specifically for the database\filesystem
\datastore, use the following commands:
l
For Datastores: appsync-cli.bat -unmount app=datastore
datastore=<value> datacenter=<value> vcenter=<value>
option=latestMountedCopy/oldestMountedCopy
l
For SQL application: appsync-cli.bat -unmount app=sql
instance_name=<value> db_name=<value> option=latestMountedCopy/
oldestMountedCopy
l
For Oracle application: appsync-cli.bat -unmount app=oracle
oracle_server=<value> db_name=<value> option=latestMountedCopy/
oldestMountedCopy
l
For file systems: appsync-cli.bat -unmount app=filesystem
fs_server=<value> fs_name=<value> fs_type=<value>
option=latestMountedCopy/oldestMountedCopy
For service plan level, the copy can be of any database that is subscribed to that service
plan To unmount a copy you need the application name and the service_plan. For
example:
The copy can be of any database that is subscribed to that service plan
To unmount a copy you need the application name and the service_plan. For example:
38
EMC AppSync 2.2.3 User and Administration Guide
AppSync CLI Utility
appsync-cli.bat -unmount app=<app> service_plan=<value> option=latestMountedCopy/
oldestMountedCopy
X
Expire a copy
To expire a copy, you need the application name and the copy uuid. Type -listCopies.
For example: appsync-cli.bat -expire app=datastore copy_ID=<value>
Run reports
There are four available reports that you can run and export through the AppSync CLI.
They include:
l
RecoverPoint Objective (rpo)
l
Service Plan Completion (spc)
l
Alert
l
Activity
Run reports in either summary or detailed view using the detailed=true/false
argument. The exception to this rule occurs with an activity report which prints the
activity that is currently running.
All reports are exported to a .csv file in the current directory with unique name from the
report type and local time. For more help, use the help command (/?) for reports. For
example:
appsync-cli.bat -report report_type=rpo detailed=true
AppSync CLI Utility tutorial
39
AppSync CLI Utility
40
EMC AppSync 2.2.3 User and Administration Guide
CHAPTER 5
Protect Microsoft Exchange
This chapter includes the following topics:
l
l
l
l
l
l
l
Overview of Exchange support ..............................................................................42
Deploying AppSync for Exchange protection: summary of steps.............................44
Protect an Exchange database ..............................................................................45
Service plan details...............................................................................................47
Mounting Exchange copies....................................................................................56
Overview of Exchange copy restore........................................................................60
Troubleshooting the EMC AppSync Exchange Interface service ............................. 67
Protect Microsoft Exchange
41
Protect Microsoft Exchange
Overview of Exchange support
Use AppSync to create application-consistent copies of Exchange data.
AppSync support for Microsoft Exchange application includes:
l
Protect and manage Microsoft Exchange in standalone and DAG environments (active
and passive databases).
l
Mount copies to a Windows 2008 or Windows 2012 host for running consistency
check or to back up to long-term storage.
l
Restore from copies to production Exchange databases in the event that production
databases must be brought back to a point-in-time.
l
l
Restore individual mailboxes and mailbox items using Kroll Ontrack®. This action can
also be done using EMC ItemPoint™ for Microsoft Exchange until this product reaches
End of Service Life (EOSL).
Support for databases on physical hosts, RDMs, and virtual disks on virtual hosts.
Note
AppSync only supports RDMs in physical compatibility mode. RDMs in virtual mode are
not supported.
Exchange Server prerequisites
Verify that the Exchange configuration meets supported version requirements for
AppSync, including Windows operating system requirements as well as supported service
packs for Exchange. The AppSync Support Matrix on https://elabnavigator.emc.com/eln/
extendedSupport is the authoritative source of information on supported software and
platforms.
AppSync supports protection and operational recovery of Exchange databases in
standalone and DAG configurations including:
l
Exchange 2010 mailbox servers on Windows Server 2008 SP2 and Windows Server
2008 R2 or later.
l
Exchange 2013 mailbox servers on Windows 2008 R2 SP1 or later.
l
Microsoft Exchange 2010 and 2013 Database Availability Groups (DAGs) including
active and passive copies.
Support for Exchange on virtual disks
You can protect, mount, and restore Exchange databases residing on VMware RDMs in
physical compatibility mode and virtual disks. AppSync supports Full, Copy, and
Differential backup types.
During protection:
42
l
For successful mapping, the Virtual Center must be added to the AppSync server and
discovery must be performed.
l
For successful protection, log files and database files must reside on virtual disks.
There cannot be a combination of physical and virtual storage.
l
Protection of Exchange databases across virtual machines sharing the same
datastore is not supported.
l
Virtual Disk is supported for Exchange ESX 5.0 and above.
l
AppSync versions 1.6.0.1 and above supports circular logging for Exchange
Databases.
EMC AppSync 2.2.3 User and Administration Guide
Protect Microsoft Exchange
AppSync interaction with Microsoft VSS
Microsoft Volume Shadow Copy Service (VSS) is the infrastructure that enables AppSync
to create application-aware copies.
When it creates a copy, AppSync coordinates with VSS and Exchange to create a shadow
copy. The copy is a point-in-time copy of the volumes that contain the data, logs, and
system files for Exchange databases.
AppSync coordinates with VSS and Exchange to quiesce input-output to the databases
when creating the copy, and then resume the flow of data after the copy has been
created. During a restore, AppSync coordinates with VSS and Exchange to recover the
point-in-time shadow copy.
Permissions required by Exchange
Accounts that AppSync uses to work with Exchange require special permissions.
l
On Exchange standalone servers, the account must be a domain user account with
the Databases role.
l
On DAG servers, the account must be a domain user account with the Database and
Database Copies roles.
l
On a mount host, the user account must be a domain user account that is a member
of the local Administrators group.
l
The account must have Log on as a batch job and Log on as a service user rights.
l
The account can have the View-only Organization role. This role is an optional role
applicable only for Microsoft Exchange 2013 if you have public folder mailboxes in
the environment. AppSync uses this role to determine the database containing the
public folder primary hierarchy mailbox.
AppSync Exchange Interface Service Credentials are required the first time that you
access the Exchange server. You are prompted to type two sets of credentials for the
AppSync Exchange Interface Service configuration.
AppSync uses the first set of credentials to install and configure the AppSync Exchange
Interface service on the Exchange production or mount host. The account must have local
administrator privileges. AppSync uses the second set of credentials to run the service. A
user must be a domain user with the following Exchange roles:
l
Database role for standalone server
l
Database and Database Copies roles in DAG environment.
Changes to service plans after upgrade
After an AppSync upgrade, changes to the way service plans operate occur.
Consider the following changes:
l
Storage ordered preferences — After an upgrade, all service plans will have their
storage ordered preferences replaced with the new style preference. After an
upgrade, check the Create Copy instructions for each service plan to confirm that the
storage ordered preference is correct.
l
RecoverPoint copy rotation — After an upgrade, all service plans do not perform
replica rotation for RecoverPoint copies. If you want to enable replica rotation for any
service plan that creates RecoverPoint copies, check the Create Copy instructions for
the service plan and check the checkbox to include RecoverPoint copies for
Expiration.
Overview of Exchange support
43
Protect Microsoft Exchange
Deploying AppSync for Exchange protection: summary of steps
A summary of steps from deployment of AppSync to setting up Exchange protection.
Procedure
1. Install the AppSync server.
2. In the AppSync console, navigate to Settings > Storage Infrastructure and click Add.
This adds the storage system where the Exchange mailbox database resides.
3. Navigate to Settings > Servers and click Add.
This adds the Exchange standalone mailbox servers or DAG member servers as hosts.
4. Navigate to Copy Management > Microsoft Exchange and click a server name from the
list of Exchange standalone and DAG servers.
5. Enter the credentials to configure and run the AppSync Exchange Interface service.
The Exchange databases are discovered.
6. Subscribe an Exchange database for protection by choosing one of the following
options:
l
Protect immediately with Subscribe to Plan and Run, which subscribes the
database to a service plan and runs the protection immediately for the selected
database only. In the case of databases in a DAG, one of the passive databases is
protected by default.
l
Subscribe to Plan, which subscribes the database to a service plan, but does not
run the plan. Protection occurs according to the service plan's schedule.
Discovering Exchange databases
To keep AppSync up-to-date, you should discover databases on the Exchange server
when there is creation, deletion, or renaming of databases.
Before you begin
l
This operation requires the Data Administrator role in AppSync.
l
The AppSync Exchange Interface service must be running.
l
If you are changing a standalone Exchange server that is part of AppSync to be part of
a DAG, you should first remove the standalone server (along with its copies, if any)
from AppSync prior to performing a discovery. During discovery, the erstwhile
standalone server is identified to be part of the DAG.
Procedure
1. Navigate to Copy Management > Microsoft Exchange.
2. Click an Exchange server to display its databases.
3. Click Discover Databases to discover databases for this server.
Removing an Exchange mailbox server
Remove an Exchange mailbox server when there is no longer a need to manage its
protection from the AppSync server.
Before you begin
This operation requires the Resource Administrator role in AppSync.
There should be no copies of the mailbox server that you want to remove.
44
EMC AppSync 2.2.3 User and Administration Guide
Protect Microsoft Exchange
Procedure
1. Select Settings > Servers.
2. Select the server to remove.
3. Select Remove > Remove servers only.
A dialog appears asking for your confirmation.
4. Click OK to confirm your action.
Protecting DAG databases in a service plan
AppSync supports protection of Exchange databases that are part of a Database
Availability Group (DAG).
When a DAG server is subscribed to an AppSync service plan, it is one of the passive
members of the DAG that is selected for protection, by default.
Procedure
l
To protect an active DAG database member, select Active in the Copy to Protect
column from the plan Subscriptions tab.
Protect an Exchange database
Protect an Exchange database by subscribing it to an AppSync service plan.
AppSync uses service plans as its protection mechanism for databases. You subscribe a
database to a service plan and run the service plan immediately, or schedule the service
plan to run at a later time.
l
Choose Subscribe to Plan and Run when you want to protect a selected database
immediately. The service plan is executed for the database alone. In the case of DAG,
one of the passive databases is protected by default.
l
Choose Subscribe to Plan when you want to schedule the protection for later.
Protection for databases that are part of the service plan are executed at the
scheduled time.
l
Choose Run from the Service Plan page to run the whole plan immediately. All
databases subscribed to the plan are protected.
Protecting an Exchange database immediately
Click Subscribe to Plan and Run to add a database to an existing service plan and run the
service plan immediately for the selected database alone.
Before you begin
This operation requires the Data Administrator role in AppSync.
Procedure
1. Navigate to Copy Management > Microsoft Exchange
2. Click an Exchange Mailbox Server or DAG from the list to display its databases.
3. From this list, select a database to protect.
When performing an operation on multiple items, be sure to keep the Shift or Ctrl key
depressed.
4. From the Protect list, select the appropriate service plan from Subscribe to Plan and
Run.
Protecting DAG databases in a service plan
45
Protect Microsoft Exchange
In DAG, a passive database is protected by default.
The Subscribe to Plan and Run dialog appears displaying the progress through the
different phases.
Subscribing an Exchange database to a service plan
Select Subscribe to Plan when you want to schedule the protection for later. Protection
for all databases that are part of the service plan are executed at the scheduled time.
Before you begin
This operation requires the Data Administrator role in AppSync.
Procedure
1. Navigate to Copy Management > Microsoft Exchange.
2. Click an Exchange Mailbox Server or DAG from the list to display its databases.
3. From this list, select a database to protect.
Select multiple databases by holding down the Shift or Ctrl keys on your keyboard.
4. From the Protect list, select the appropriate service plan from Subscribe to Plan.
In DAG, a passive database is protected by default. To change the protection type with
another option, specify it from the Subscriptions tab of the service plan.
The plan is added to the Plans column for the database.
Unsubscribing a database from a service plan
You can unsubscribe an individual database from a service plan.
Before you begin
This operation requires the Data Administrator role in AppSync.
Procedure
1. Navigate to Copy Management > Microsoft Exchange
2. Select a mailbox server to display the list of Exchange databases.
3. Select the database to remove from a plan.
l
Select the plan to unsubscribe from Protect > Unsubscribe from Plan.
Only plans to which the database is subscribed to are in the popup list.
l
To unsubscribe from all service plans, select Unsubscribe from Plan > All.
Expiring a copy on demand
Expiring a copy removes it from the AppSync database and can free up storage,
depending on the replication technology and copy state.
Before you begin
This operation requires the Data Administrator role in AppSync.
Expiring a copy that was made with RecoverPoint does not remove the corresponding
bookmark from RecoverPoint itself.
Procedure
1. Select Copy Management > Microsoft Exchange.
46
EMC AppSync 2.2.3 User and Administration Guide
Protect Microsoft Exchange
2. Click an Exchange mailbox server to display its databases.
3. Click an Exchange database to display its copies.
4. Select one or more copies to delete.
You can also perform this action from the Service Plan Copies tab.
5. Select Expire.
Verify that you want to expire the copy you selected and any associated copies listed
and confirm.
Creating a database copy from the Copies page
Create a copy of a database by subscribing it to an AppSync Exchange service plan from
the Copies page.
Before you begin
This operation requires the Data Administrator role in AppSync.
Procedure
1. Navigate to Copy Management > Microsoft Exchange.
2. Click a mailbox server instance.
3. From the list of Exchange databases, click the database to view its copies.
4. From the Create a copy using plan list, select the appropriate service plan.
The service plan runs immediately for the database.
Service plan details
A service plan has the following tabs: Settings, Subscriptions, Copies, Events, and for
SQL Server plans, Log Backups .
The Settings tab shows the name, description, and status (whether enabled or disabled)
of the service plan. Apart from these, the different phases of the plan are also part of this
tab. Click on appropriate tabs to see information regarding Subscriptions, Copies created
by the plan Events generated during the service plan run, and for SQL Servers, Log
Backups created by the plan.
Service plan schedule
The schedule of a service plan is set in the Plan Startup phase.
The Startup Type (scheduled or on demand) determines whether the plan is run
manually, or configured to run on a schedule. Options for scheduling when a service plan
starts are:
l
l
Specify a recovery point objective (RPO)
n
Set an RPO of 30 minutes or 1, 2, 3, 4, 6, 8, 12, or 24 hours
n
Minutes after the hour are set in 5 minute intervals
n
Default RPO is 24 hours
Run every day at certain times
n
Select up to two different times during the day
n
Minutes after the hour is in 5 minute intervals
n
There is no default selected
Creating a database copy from the Copies page
47
Protect Microsoft Exchange
l
l
Run at a certain time on selected days of the week
n
One or more days of the week (up to all seven days) can be selected
n
There is no default day of the week selected. Default time of day is 12:00 AM.
Run at a certain time on selected days of the month
n
Select one or more days of the month (up to all days)
n
Select one time of day. Available times are at 15 minute intervals.
n
Default is the first day of the month
Control replication storage utilization
When you set up a service plan, set values in the following fields so that you avoid
overutilization and depletion of replication storage:
l
RPO value in the Plan Startup phase
l
Always keep n Copies in the Create copy phase
You should also monitor your storage system with the storage system user interface.
Overriding service plan schedules
You can set individual schedules for databases subscribed to a service plan, overriding
the generic recurrence setting.
Before you begin
This operation requires the Service Plan Administrator role in AppSync.
You can override only the settings of the recurrence type already selected for the service
plan.
Procedure
1. Navigate to Service Plans > Microsoft Exchange and select one of the plans from the
list.
2. From the Settings tab, select the Plan Startup phase.
3. From the Plan Startup Defaults pane on the right, note the Recurrence Type selected
for the plan.
A recurrence type can be set only if Scheduled is selected as the Startup Type.
4. Click the Plan Startup Overrides tab.
5. Set individual schedules for selected databases based on your requirement.
For example, if the recurrence type you selected is On specified days of the month,
and the rule setting is to Run at 12:00 AM on the 1st day of every month, you can
override the time and the day for individual databases.
Application discovery
Before creating the copy, AppSync examines the Exchange Mailbox Server to look for
changes such as addition, deletion, renaming, or movement of databases.
There are no user settings associated with this phase and it cannot be disabled.
48
EMC AppSync 2.2.3 User and Administration Guide
Protect Microsoft Exchange
Application mapping
After discovering the application, AppSync maps it to array storage, and protection
services such as RecoverPoint.
There are no user settings associated with this phase and it cannot be disabled.
Pre-copy script
To perform preparatory steps before creating a copy, specify a pre-copy script and
parameters on a service plan's Settings tab.
The pre-copy script runs according to the schedule set in the Plan Startup phase. Valid
script formats are .bat, .exe, and .ps1 (PowerShell scripts). You can optionally enter
credentials to run the script as a specific user. The script runs as Local System by default.
AppSync does not support running of PowerShell scripts directly. You usually must wrap
them in a .bat file. The other option is to make the default "Open" on ps1 files C:
\WINDOWS\system32\WindowsPowerShell\v1.0\powershell.exe. When the
PS script runs, you may get an error and you must set an appropriate execution policy.
To run PowerShell commands from scripts:
1. Specify the full pathname to the PowerShell command file in the .bat file:
powershell -command C:\PshellCommands.ps1 <nul
2. Set the PowerShell execution policy so you can run the script. For example, the first
line in the .bat file should look like the following for an unrestricted policy:
powershell -command set-executionpolicy unrestricted <nul
3. To ensure correct termination of the PowerShell session, add <nul to the end of the
line that calls the PowerShell script. The default location of the script is
%ProgramData%\EMC\AppSync\scripts\ on the application host.
Exact parameters depend on the script. Parameters with spaces must be enclosed in
double quotes.
This phase can be enabled or disabled. This operation requires the Service Plan
Administrator role in AppSync.
Create copy
The Create Copy phase creates a copy based on the replication technology specified in
the service plan.
This phase specifies the type of Exchange copy to make, whether to ignore Exchange
errors in the Application event log, and if database and logs can reside on the same
volume.
Review Overview: Service Plan on page 10 for more service plan copy information.
Exchange backup type
AppSync uses VSS to make a consistent online copy at the volume level.
l
Full creates a copy of the databases in the service plan using VSS, and includes the
database files, transaction logs, and checkpoint files. On successful completion of
the backup, the logs are truncated.
l
Copy creates a copy of the databases in the service plan using VSS, which includes
the database files, transaction logs, and checkpoint files, as it does using the Full
option. However, it does not truncate the logs.
Application mapping
49
Protect Microsoft Exchange
l
Differential copies the entire transaction log volume. A full backup of the selected
database must exist or the backup fails. The transaction logs are not truncated on
completion of the backup.
Automatic expiration of copies
The automatic expiration value in a service plan's Create Copy phase specifies the
maximum desired number of Snap, Clone or Bookmark copies that can exist
simultaneously.
When the "Always keep x copies" value is reached, older copies are expired to free
storage for the next copy in the rotation. Failed copies are not counted. AppSync does not
expire the oldest copy until its replacement has been successfully created. For example,
if the number of copies to keep is 7, AppSync does not expire the oldest copy until the
8th copy is created.
AppSync does not expire copies under the following circumstances:
l
Mounted copies are not expired.
l
A copy that contains the only replica of a database will not be expired.
This setting is independent of the VNX pool policy settings in Unisphere for automatic
deletion of oldest snapshots. The service plan administrator should work with the storage
administrator to ensure that the VNX pool policy settings will enable the support of the
specified number of snapshot copies for the application residing in that pool.
Include RecoverPoint copies in expiration rotation policy: Check this option to include
RecoverPoint copies when calculating rotations.
Note
If this option is not selected, then RecoverPoint copies will accumulate, and will remain
until the bookmarks fall off the RecoverPoint appliance.
Exchange event log errors
Exchange logs certain errors in the Application event log when they occur. These errors
indicate a possible corruption of the data in the .edb or log files. They can cause copy
creation to fail unless you specifically instruct AppSync to ignore them.
AppSync searches the application event log for these errors every time a copy is created.
The first time it runs, AppSync searches the entire log. Subsequent runs search since the
last successful run. If there are no existing copies, then AppSync searches the entire log
when creating the next copy.
In a service plan's Create copy phase, you can configure AppSync to ignore any or all of
these errors.
Table 7 Microsoft Exchange event errors
Error Meaning
-1018 The database tried and failed to verify information about a particular page in the
database.
-1019 Similar to a -1018 error but indicates that the accessed page has returned an invalid page
number (usually all zeros) rather than an invalid checksum.
-1022 Indicates major hardware problems, particularly disk subsystem problems. If the
database engine requests a page from disk but instead receives an error from the I/O
subsystem, a -1022 error results.
50
EMC AppSync 2.2.3 User and Administration Guide
Protect Microsoft Exchange
Table 7 Microsoft Exchange event errors (continued)
Error Meaning
447
Indicates corruption in the logical database structure. This accompanies a message
stating that the information store terminated abnormally.
448
Indicates an inconsistency or corruption in a table in the Microsoft Jet database. This
accompanies a message stating that an information store data inconsistency has been
detected in a table.
Database and log layout
Exchange supports environments in which the database and logs reside on the same
volume when there is more than one copy of the database in a DAG environment. Service
plans can be configured to ignore the restriction that prevents databases and logs from
residing on the same volume.
When creating copies of Exchange databases, it is a best practice to restrict a service
plan from allowing this configuration because having databases and logs on the same
volume limits your restore options. However, you can choose whether service plans with
this configuration should succeed or not.
This option is set in the Create Copy phase of a service plan.
When selecting this option, you are limited to restoring the database and logs together.
Restore overwrites newer log files. To preserve newer log files for use during recovery,
copy them to another volume before restore.
Post-copy script
To perform cleanup or other post-copy steps after creating a copy, specify a post-copy
script and parameters in a service plan's Settings tab.
The script runs on successful completion of the Create copy phase. Valid script formats
are .bat, .exe, and .ps1 (PowerShell scripts). You can optionally enter credentials to run
the script as a specific user. The script runs as Local System by default.
When AppSync creates copies of application items in a service plan, it may break up the
application items and place them in separate groups for protection. This action can be for
performance reasons (for example, VSS for Exchange and SQL) or because items in a
service plan may be protected by different replication technologies. For example, a
service plan may contain some application items that are protected by VNX Snapshots
and some by RecoverPoint bookmarks. As a result, application items in these groups are
protected independently.
When AppSync calls a post-copy script, it passes the copies which were created in the
group by calling the script with -appCopies <APP1> <APP2>, where APP1 and APP2 are the
names of the application items in that grouping.
AppSync does not support running of PowerShell scripts directly. You usually must wrap
them in a .bat file. The other option is to make the default "Open" on ps1 files C:
\WINDOWS\system32\WindowsPowerShell\v1.0\powershell.exe. When the
PS script runs, you may get an error and you must set an appropriate execution policy.
To run PowerShell commands from scripts:
1. Specify the full pathname to the PowerShell command file in the .bat file:
powershell -command C:\PshellCommands.ps1 <nul
Post-copy script
51
Protect Microsoft Exchange
2. Set the PowerShell execution policy so you can run the script. For example, the first
line in the .bat file should look like the following for an unrestricted policy:
powershell -command set-executionpolicy unrestricted <nul
3. To ensure correct termination of the PowerShell session, add <nul to the end of the
line that calls the PowerShell script.
When AppSync runs the post-copy script, it is run for the application items that are part of
a group. If there are multiple groups, the post-copy script runs multiple times. When
AppSync runs the post-copy script, it passes the list of application items in the
replication group as arguments to the script, right after the user arguments. The syntax is:
-applicationCopies <ITEM1> <ITEM2> <ITEM3>
where <ITEMx> is the name of the application item that is being protected.
The default location of the script is %ProgramData%\EMC\AppSync\scripts\ on
the application host.
Exact parameters depend on the script. Parameters with spaces must be enclosed in
double quotes.
This phase can be enabled or disabled. This operation requires the Service Plan
Administrator role in AppSync.
Unmount previous copy
The service plan unmounts a previously mounted copy after creating the new copy. The
exception is a copy that was mounted on-demand as opposed to by the service plan; in
this case the on-demand mounted copy is not unmounted.
There are no user settings associated with this phase and it can be enabled or disabled.
Mount copy
The Mount copy phase mounts the copy. This phase can be enabled or disabled.
The Mount Copy Defaults settings for the mount host value, mount path and mount
access attributes (read-only or read-write) depend on the service plan. Other mount
settings determine where the Exchange metadata files are copied, the type of copy to
mount and the RecoverPoint image access type.
l
Mount on Server
Allows you to choose between Windows hosts you have access to and Original
Server. If you have chosen to validate the copies, only servers that have the Exchange
Management Tools installed are displayed in the drop down. These servers display on
the Microsoft Exchange Protection page as "Utility Host".
l
Mount with access
Choose the type of access the copy should be mounted with - Read/Write or Read
only
l
Mount Path
n
Alternate mount path
The default mount path, when the mount host is the same as the production host,
is SystemDrive:\AppSyncMounts\Production_Server_Name.
path is represented in the console as %SystemDrive%\AppSyncMounts\%
%ProdServerName%%.
To specify the value of a Windows environment variable in the mount path, delimit
the variable name with single percent signs (%). The default path also contains an
52
EMC AppSync 2.2.3 User and Administration Guide
Protect Microsoft Exchange
AppSync variable (ProdServerName) that is delimited with two percent signs (%
%).
The following characters are not valid in the path:
<>:"/|?*
n
Same as original path This is another option for the mount path. You can select
either of the options.
Note
When performing a DAG mount, do not select the mount path as Same as original
path if the mount host also happens to be a DAG node having a copy of the
database that you are mounting.
l
Copy metadata files to
By default, the location to copy VSS metadata files is the default path - SystemDrive:
\AppSyncMounts\Production_Server_Name.
The following characters are not valid in the path:
<>:"/|?*
If you are backing up the database to another media, you must backup these
metadata files as well.
l
Image access options during RecoverPoint mount
RecoverPoint provides a target-side host application the opportunity to write data to
the target-side replication volumes, while still keeping track of source changes.
n
Slow access time, fast image I/O performance (RecoverPoint access mode:
Logged Access)
Use this mount option if the integrity check entails the scanning of large areas of
the replicated volumes. This is the only option available when you mount to the
production host.
n
Fast access time, Fast after roll image I/O performance (RecoverPoint access
mode: Virtual Access with Roll)
Provides nearly instant access to the copy, but also updates the replicated volume
in the background. When the replicated volumes are at the requested point in
time, the RPA transparently switches to direct replica volume access, allowing
heavy processing.
n
Fast access time, Slow image I/O performance (RecoverPoint access mode:
Virtual Access)
Provides nearly instant access to the image; it is not intended for heavy
processing.
n
Desired Service Level Objective (SLO)
Additionally if you are using a VMAX 3 array, a setting called Desired Service Level
Objective (SLO) is available. The option appears in the Mount wizard and it
specifies the required VMAX 3 Service Level Objectives. SLO defines the service
time operating range of a storage group.
l
Copy to mount
Displayed for service plans that create both a local and remote copy. You can select
the type of copy to mount.
l
Additionally if you are using a VMAX 3 array, a setting called Desired Service Level
Objective (SLO) is available. The option appears in the Mount wizard and it specifies
the required VMAX 3 Service Level Objectives. SLO defines the service time operating
range of a storage group.
Mount copy
53
Protect Microsoft Exchange
Mount host overrides in service plan
Select different mount hosts for multiple Exchange servers subscribed to a service plan.
In the Mount copy phase of a service plan, you can specify the host that the copy should
be mounted on along with related mount options. If you have multiple servers as part of a
service plan, you may want to host their copies on different hosts. You can specify
different mount hosts and other options from the Mount Copy Overrides tab of the Mount
copy phase in a service plan.
Overriding mount hosts in a service plan
If there are multiple registered hosts and they are subscribed to the same plan, you can
select a different mount host for each server, overriding the generic mount host settings.
Before you begin
This operation requires the Data Administrator role in AppSync.
Follow these steps when you have multiple hosts subscribed to a plan and you want
different mounts hosts for their database copies.
Procedure
1. Navigate to Service Plans and select one of the plans from the list.
2. From the Settings tab, select Mount copy phase.
In the Mount Copy Defaults tab, the list of servers include all Exchange servers whose
databases are subscribed to this plan. The mount settings display the default
settings.
3. To override the default settings, click Mount Copy Overrides.
4. Select the server whose mount settings you wish to override and click Set Overrides.
5. In the Override Default Mount Settings dialog, select options only for those mount
settings that you wish to override.
For example, if you want to mount a copy to a different path, you would select the
path from the Mount Path list. Fields that do not have a selection retain their default
settings.
6. Select OK to save your changes.
A pencil icon appears in the first column of the server's row whose default mount
settings you changed.
7. To revert back to default settings for a server, select the server and click Use Default
Settings.
Validate copy
Exchange management tools run a consistency check in this phase.
By default, databases and logs are checked sequentially. If the databases are not sharing
the same LUN and the mount host has sufficient resources to support parallel
consistency checks, use the In parallel option. Note that there is a limit of 16 parallel
checks that Exchange can handle.
If the consistency check completes successfully, AppSync instructs Exchange to truncate
the logs so only the changes that are uncommitted to the database remain.
This phase can be enabled or disabled.
54
EMC AppSync 2.2.3 User and Administration Guide
Protect Microsoft Exchange
Advanced options for consistency check
AppSync offers advanced options that change how Exchange consistency checks are
executed. Enabling these features can impact performance.
l
Minimize log checking
Choosing this option speeds up the log checking by instructing the consistency
checking software to check only those logs that are required to recover the database.
Selecting this option improves the performance of the consistency check. If you
disable the option, then consistency check will be performed on all of the database’s
logs.
This command instructs AppSync to check only a subset of the Exchange logs that are
included in the copy. The subset of the logs are actually the logs that are required to
recover the database. If your backup window is small, you may find this option
useful. However, the copy contains logs that have not been checked for consistency.
If you attempt to restore the log volume, you may find that some log files are corrupt
or the log sequence is not complete. Before restoring the log volume, you should
mount the replica and run eseutil /k Enn against the log path.
For maximum protection, clear Minimize log checking. For maximum performance,
select it.
You must also set a working directory, which is where the required log files will be
copied for checking.
The Minimize log checking option is not available when the consistency method is
Differential.
l
Throttle Checking
Consistency checks can be paused to slow down the IOs during the check. You can
specify the number of IOs after which to pause, and the duration of the pause.
Post-mount script
Specify a post-mount script and parameters from the Post-mount script option in the
Settings tab of a service plan.
The script runs on successful completion of the mount copy or mount with recovery
phase. This script is typically used for backup.
From the Server list, select the server on which to run the script. You can optionally run it
on a registered host other than the mount host, and enter credentials to run the script as
a specific user.
The default location of the script is %ProgramData%\EMC\AppSync\scripts\ on
the application host.
Exact parameters depend on your script. Parameters with spaces must be enclosed in
double quotes.
This phase can be enabled or disabled. This operation requires the Service Plan
Administrator role in AppSync.
Unmount copy
The final phase in the service plan unmounts the copy. All the mounted databases are
shut down as part of this phase.
This phase is disabled if the Unmount previous copy phase is enabled. There are no user
settings associated with this phase.
Post-mount script
55
Protect Microsoft Exchange
Mounting Exchange copies
AppSync can mount a copy on-demand, or as part of a plan.
Copies created on a standalone production Exchange server can be mounted to:
l
An alternate host in the same location as the production host.
l
An alternate host in a new location. You specify mount option by adding an alternate
path to the start of the path.
l
The production host in an alternate location.
Copies created in a DAG can be mounted to:
l
An alternate host
l
A server in another DAG
l
Another server in the same DAG
Note
Copies cannot be mounted to the same DAG server on which the copy was created.
Mount and restore limitations
Limitations to mount and restore or Exchange copies appear in the following list:
l
When the root drive letter has mount points on it and they are all included in the
same plan, mounts and restores are likely to fail. For instance, if the log and system
files are on L:\ and the mailbox stores are on L:\SG1DBMP (where SG1DBMP is a
mount point), mounts and restores fail.
l
In Windows 2012 and later environments, when doing a restore, the data on LUNs is
overwritten even if the volume is in use. This action differs from other Windows
platforms in which AppSync displays a warning if the LUN is in use. Since restores
overwrite everything, be sure that there is no other data on that volume and the
volume is not in use.
Mounting an Exchange copy on-demand
You can initiate an on-demand mount of an Exchange copy from a copy or database.
Before you begin
This operation requires the Data Administrator role in AppSync.
Procedure
1. Select Recover > Mount a Copy in the Exchange Databases page.
Alternatively, from the Copies page or Copies tab of the service plan, select a copy
and click Mount.
2. Use the Copies or Service Plan filters to select the appropriate copy to mount.
The Copies list is refreshed based on the filters selected.
3. Select the copy to mount.
For a RecoverPoint copy, you also have the option to select a copy based on a specific
time. Click Select a point in time to select a copy with a specific time stamp. The time
displayed is the console's time. If the console is in a different time zone from the
server, specify the time as per the server's time zone to mount the copy.
56
EMC AppSync 2.2.3 User and Administration Guide
Protect Microsoft Exchange
4. In the Mount Additional Copies page, select one or more additional copies to mount.
The copies listed here are of other databases that were protected at the same time
and on the same host as the copy you selected in the previous step.
5. In the Validate Copy page, select Yes to validate the copies and specify validation
options. See Mount validation options for details.
6. On the Mount Options page, select the mount options.
l
For VMAX 3 arrays (only), you are presented with an SLO drop-down menu. You can
select the desired Service Level Objective (SLO) for the mount copy. If there is a
storage group for the mount host with the desired SLO. AppSync will add the LUN
to the storage group. If this storage group does not exist, AppSync adds the LUN to
any storage group that is masked to the host. If a storage group is configured to
pick target devices, AppSync removes the devices from the storage group at the
time of mount and adds them to the storage group for the mount host. The devices
will be added to the original storage group when the copy is expired. An example
of the Desired SLO menu follows:
Figure 2 For VMAX 3 select Desired SLO
l
Mount on Server
Allows you to choose between Windows hosts you have access to and Original
Server. If you have chosen to validate the copies, only servers that have the
Exchange Management Tools installed are displayed in the drop down. These
servers display on the Microsoft Exchange Protection page as "Utility Host".
l
Mount with access
Choose the type of access the copy should be mounted with - Read/Write or Read
only
l
Mount Path
n
If you specify non-default mount path, the drive specified for mount can not be
a clustered disk.
n
Alternate mount path
The default mount path, when the mount host is the same as the production
host, is SystemDrive:\AppSyncMounts\Production_Server_Name.
Mounting an Exchange copy on-demand
57
Protect Microsoft Exchange
path is represented in the console as %SystemDrive%\AppSyncMounts\%
%ProdServerName%%.
To specify the value of a Windows environment variable in the mount path,
delimit the variable name with single percent signs (%). The default path also
contains an AppSync variable (ProdServerName) that is delimited with two
percent signs (%%).
The following characters are not valid in the path:
<>:"/|?*
n
Same as original path
This is another option for the mount path. You can select either of the options.
Note
When performing a DAG mount, do not select the mount path as Same as
original path if the mount host also happens to be a DAG node having a copy of
the database that you are mounting.
l
Copy metadata files to
By default, the location to copy VSS metadata files is the default path SystemDrive:\AppSyncMounts\Production_Server_Name.
The following characters are not valid in the path:
<>:"/|?*
If you are backing up the database to another media, you must backup these
metadata files as well.
l
Image access options during RecoverPoint mount
RecoverPoint provides a target-side host application the opportunity to write data
to the target-side replication volumes, while still keeping track of source changes.
n
Slow access time, fast image I/O performance (RecoverPoint access mode:
Logged Access)
Use this mount option if the integrity check entails the scanning of large areas
of the replicated volumes. This is the only option available when you mount to
the production host.
n
Fast access time, Fast after roll image I/O performance (RecoverPoint access
mode: Virtual Access with Roll)
Provides nearly instant access to the copy, but also updates the replicated
volume in the background. When the replicated volumes are at the requested
point in time, the RPA transparently switches to direct replica volume access,
allowing heavy processing.
n
Fast access time, Slow image I/O performance (RecoverPoint access mode:
Virtual Access)
Provides nearly instant access to the image; it is not intended for heavy
processing.
7. In the Configure AppSync Exchange Interface Service page, provide the credentials to
configure the service on the mount server.
This page is displayed only if you chose to validate the copy if the service is not
configured.
8. Review the Summary and click Finish to mount the copy.
9. In the Results page, select View Details to see progress of the different phases that
are part of mounting a copy.
58
EMC AppSync 2.2.3 User and Administration Guide
Protect Microsoft Exchange
The last phase completed is displayed at the bottom of the list.
Validation options for a mount copy
Validation for differential backup copies is not supported.
Validate database and logs
When you create a replica of one or more Microsoft Exchange databases, you should
mount the replica and test it for consistency. If you choose to automatically mount the
replica to an alternate host once it has been created, you should run a consistency check
on the replica. The options to validate are:
l
Sequentially — Run tests on one database at a time in order (serial mode). Select this
option if you have several Exchange databases on one LUN.
l
In Parallel — Run tests on several databases simultaneously (parallel mode).
Minimize log checking
By selecting Minimize log checking, AppSync checks a subset of the Exchange logs that
are included in the replica. If your backup window is small, you may find this option
useful. However, the replica may contain logs that have not been checked for
consistency.
For maximum protection, clear Minimize log checking. For maximum performance, select
it.
Working directory — This field allows you to specify the directory to which the relevant log
files will be moved in order to run the check, since a consistency check can only be run
on all logs in a single directory.
Throttle Validation
Select this to throttle the I/Os during a consistency check. This option is for advanced
users and typically should not be selected unless you are working with EMC Support to
resolve an issue related to I/O throughput. Typically, the throttling option is not required.
If you choose to throttle I/Os, you have the following two options.
l
Pause after I/O count of: 100 — This option allows you to choose how many I/Os can
occur between pauses. You can choose any value between 100 and 10,000 I/Os.
l
Duration of pause (in milliseconds): 1000 — You can specify the duration of the
pause in milliseconds. 1000 milliseconds = 1 second. If this option is not available,
the pause will be one second long.)
Unmounting an Exchange copy
When you select a copy to unmount, other copies that were mounted along with the
selected copy will also be unmounted.
Before you begin
l
This operation requires the Data Administrator role in AppSync.
l
If the copy is mounted due to performing an item level restore, verify that all work
within the ItemPoint application is complete prior to unmounting the copy.
You can unmount a copy only from a list of copies made for a database.
Procedure
1. Navigate to the Copies page from the Data Protection or Service Plan pages:
l
Copy Management > Microsoft Exchange > select the Exchange Mailbox Server that
hosts the database, then select the database with the copy to unmount.
Unmounting an Exchange copy
59
Protect Microsoft Exchange
l
Service Plans > Microsoft Exchange> select a service plan, then select the Copies
tab.
2. From the list of copies, select the copy and click Unmount.
The Unmount Confirmation dialog displays copies of other databases that were
mounted along with the selected copy to be unmounted.
3. Click Yes to confirm the unmount of all the copies shown in the dialog.
The Unmount window displays the progress of the unmount operation. All copies
associated with the selected copy will be unmounted.
Overview of Exchange copy restore
Learn about Exchange restore features along with associated storage copy levels.
With AppSync you can restore the following objects:
l
A database with its logs.
l
A database .edb file.
l
Only the logs for a database.
l
An active or passive database (in conjunction with any one of the three points already
mentioned), if the server is a member of a DAG (Database Availability Group).
AppSync restores VNX/VMAX copies at the LUN level and VNXe copies at the LUN group
level. In a RecoverPoint environment, restore is at the consistency group level.
Affected entities during restore
When restoring from a copy, you may be prompted to restore items in addition to the
ones you selected.
An affected entity is data that resides on your production host that unintentionally
becomes part of a replica because of its proximity to the data you intend to protect. You
can prevent affected entity situations by properly planning your data layout based on
replica granularity. The granularity of a replica depends upon the environment.
If there are affected entities in your underlying storage configuration, the Restore Wizard
notifies you of these items. The following scenarios produce affected entities that require
you to acknowledge that additional items will be restored:
l
For RecoverPoint, if the databases are in the same consistency group they become
affected entities when the other database is protected.
l
For VNXe, if the databases are in the same LUN group, they become affected entities
when another database in the group is protected.
l
For VMAX, VNX, VNXe, or XtremIO, if the databases are on the same LUN they become
affected entities when the other database is protected.
l
For VMware virtual disks, since restore involves a datastore, restore of all
applications residing on the same datastore (virtual disks on the same datastore) are
also affected entities.
If the affected entity was protected along with the database selected for restore, AppSync
restores it. Any other database that was not protected but is an affected entity is
overwritten. AppSync calculates affected entities for the consistency groups, LUN groups
or LUNs of the database that is selected for restore. If the affected databases partially
reside on other consistency groups, LUN groups or LUNs, AppSync does not calculate
affected entities on those consistency groups. LUN groups or LUNs.
Affected entities are calculated on the basis of restore granularity. If both data and logs
are selected for restore, then affected entities are calculated for all the consistency
60
EMC AppSync 2.2.3 User and Administration Guide
Protect Microsoft Exchange
groups, LUN groups, or LUNs on which the database resides. If only data or only log
restore is selected, then the affected entities are only calculated for the selected
component's consistency group, LUN Group, or LUN only.
If the database data and log components reside on the same consistency group, LUN
group, or LUN, the option to restore only logs or restore only data is not available. You
have the option only to restore data and logs. The only exception to this scenario is when
you perform a differential copy restore.
Since restore involves a datastore with VMware virtual disks, restore of all applications
residing on the same datastore (virtual disks on the same datastore) are also affected
entities.
Restoring from an Exchange copy
You can perform a restore of an Exchange copy from a copy or a database.
Before you begin
This operation requires the Data Administrator role in AppSync.
Procedure
1. Select Recover > Restore > Databases and logs from the action buttons at the bottom
of the page.
2. Select the copy to restore, and click Next.
3. From the Select Options page, select the appropriate restore options and click Next.
Option
Description
Both data and logs
Available when restoring from full and copy backup.
Data
Available when restoring from full and copy backup.
Logs
Available when restoring from full, copy, and differential
backup.
Recover and mount
databases after restore
Leave this selected. If not selected, you must recover the
database manually.
Allow AppSync to
activate databases
Available for DAG only. If selected, the database copy is
activated on the server prior to restoring it.
Note
Exchange restores only to the active copy and AppSync
restores to the server that created the copy.
The Restore Warnings page is displayed. This page may be displayed if the selected
copy has affected entities.
4. Read the warning messages for the affected databases.
In case of Exchange 2013, AppSync attempts to determine if the selected database or
any of the affected entities contain the public folder primary hierarchy mailbox. If a
match is found, you see an error message that the restore operation cannot continue.
To determine where the public folder primary hierarchy mailbox resides, AppSync
requires your user profile to have an additional Exchange role - the View-only
Organization role. If your account does not have this role, you see a warning message.
Restoring from an Exchange copy
61
Protect Microsoft Exchange
5. Select the checkbox to indicate your agreement to restore other entities along with the
selected copy.
You must manually unmount the databases that will be overwritten.
6. Review the Summary page and click Finish to restore the copy.
7. In the Results page, click View Details to see progress of the different phases that are
part of restoring a copy.
The last phase completed is displayed at the bottom of the list.
Recovering an Exchange database manually
Perform a manual recovery when you have not selected the Recover and mount the
databases after restore option in the Restore wizard.
Before you begin
When you are recovering just a database file, verify that the transaction log files needed
for recovery are present. An unbroken sequence is required. To determine the minimum
required range of logs, run the following command against each database after the
restore and before running recovery: ESEUTIL /mh <database name>. Look for the
Log Required information in the ESEUTIL output.
If the database is the active copy, it must first be unmounted in order to run the ESEUTIL
command successfully.
Procedure
1. Delete the checkpoint file (Enn.chk).
This is optional.
2. Delete the restore.env file (EnnRESTORE.env).
3. Recover the databases manually in soft recovery mode using the ESEUTIL command.
eseutil /r E<nn> /l <logpath> /s <chkpt file path> /d
<database path>
4. Use Exchange Management Console to mount all the restored databases.
Restoring a standalone Exchange database on XtremIO
This procedure is applicable only for XtremIO 3.x. XtremIO 4.0 and later has a formal
restore using the Restore Wizard.
Review and follow these steps to restore an XtremIOSnap copy of a standalone Exchange
database that resides on an XtremIO 3.x array.
The mounted copy is validated before manual restore (Mount with Validation phase) with
this procedure.
Note
Restrictions: Renamed and deleted databases cannot be restored. Restore of data (only)
or logs (only) is not supported.
Procedure
1. Suspend AppSync schedules that affect the copy being restored.
2. To restore the copy to the production host, use the AppSync console to mount the
copy. Ensure that you select the Validation phase.
62
EMC AppSync 2.2.3 User and Administration Guide
Protect Microsoft Exchange
a. Ensure a read/write mount.
b. Mount with default path or any user-defined path
3. Stop the production application, using the Exchange Management Console, Dismount
Production database.
4. Dismount the Exchange database from the Exchange Management Console.
5. Remove drive letters for production database drives.
6. Assign production database drive letters to mounted database (data and log drives
respectively), and then verify data and log paths.
7. In the Exchange Management Console, mount the production database.
8. To remove copies from the AppSync database, click Remove on the copies page of the
AppSync console,
The Remove function is the similar to Expire, but Remove does not delete the copies
from storage.
9. On the AppSync console, move mounted snapshots to the production folder in the
XtremIO array.
10. Test restored contents to confirm successful restore.
11. Remove the production volumes on XtremIO with the AppSync console.
Snapshots of deleted volumes become production volumes.
12. If data and log drives were vDisks, perform VMware Virtual Center re-discovery in
AppSync.
13. To create a fresh copy of restored data, run the service plan.
14. Resume any suspended schedules.
15. Optionally, clean-up or delete production devices from production host and remove
AppSync mount points that are created during the mount-phase.
Restoring an Exchange database DAG on XtremIO
This procedure is applicable only for XtremIO 3.x. XtremIO 4.0 and later has a formal
restore using the Restore button and the Restore Wizard.
These steps show you how to restore an XtremIOSnap copy of an Exchange database with
DAG that resides on an XtremIO 3.x array.
The mounted copy is validated before manual restore (Mount with Validation phase) with
this procedure.
Note
Restrictions: Renamed and deleted databases cannot be restored. Restore of data (only)
or logs (only) is not supported.
Procedure
1. Suspend AppSync schedules that affect the copy being restored.
2. To restore the copy to the production host, use the AppSync console to mount the
copy. Ensure that you select the Validation phase.
a. Ensure a read/write mount.
b. Mount with default path or any user-defined path (DAG member on which the copy
was protected).
Restoring an Exchange database DAG on XtremIO
63
Protect Microsoft Exchange
3. Stop the production application using the Exchange Management Console, suspend
passive copies and then dismount the Production[Active] database.
4. Remove drive letters for production database drives.
5. Assign production database drive letters to mounted database (data and log drives
respectively), and then verify data and log paths.
6. In the Exchange Management Console, mount the production database.
7. To update passive copies, run the following command for all DAG members :
Update-MailboxDatabaseCopy -Identity <Database Name>\<DAG
member> -SourceServer <DAG Member>
8. On the AppSync console, click Remove on the AppSync copies page to remove copies
from the AppSync database.
The Remove function works like Expire, but Remove does not delete the copies from
storage.
9. On the AppSync console, move mounted snapshots to the production folder in the
XtremIO array.
10. Test restored contents to confirm successful restore.
11. Remove the production volumes on XtremIO with the AppSync console.
Snapshots of deleted volumes become production volumes.
12. If data and log drives were vDisks, perform VMware Virtual Center re-discovery with
AppSync.
13. To create a fresh copy of restored data, run the service plan.
14. Resume suspended schedules.
15. Optionally, clean-up or delete production devices from production host and remove
AppSync mount points that are created during the mount-phase.
Partial restore
In a partial restore, you restore data alone or restore data and then restore the logs
separately.
Before you perform a partial restore, ensure that the database layout fulfills some
conditions.
Partial restore considerations
In a RecoverPoint environment, the granularity of restore is at the consistency group level.
When you restore a database from a bookmark, any bookmarks that are newer than the
bookmark being restored are deleted. The corresponding application copies are also
deleted. The following best practices are recommended:
l
The database and logs must reside in different consistency groups.
If you have data and logs for an Exchange database in the same consistency group,
partial restore is not supported.
l
The logs should be restored from a newer Differential backup copy. AppSync does not
support restoring just the logs from a Full or Copy backup in a RecoverPoint
environment.
In a VMAX/VNX environment, the database and logs must reside on different LUNs.
64
EMC AppSync 2.2.3 User and Administration Guide
Protect Microsoft Exchange
Restore data
Restore data from a Full or Copy backup. You can restore data only to preserve the logs
that are on the production host.
In the Restore wizard, restore data from the most recent copy and select the Recover and
mount the databases after restore option.
Restore logs
Restore data from a Full or Copy backup and then restore the logs from a later copy to
make the copy current.
Restoring a copy from the logs is a two-step process. Run the Restore wizard and select a
full backup copy to restore only data. Do not opt to Recover and mount the databases
after restore in this run.
Run the Restore wizard again and select a backup copy (a differential backup in case of
RecoverPoint) to restore only the logs. This time, select the Recover and mount the
databases after restore option. This copy must be later than the backup copy that you
selected during the first run.
Note
If the restore operation includes restoring logs, the restore overwrites any logs that are
created since the copy was created. Therefore, after the restore, the database reflects the
point in time when the copy was created. If you want to preserve logs that are created
since the copy, restore only the databases, preventing AppSync from restoring older logs
over the newer logs. You can also make a copy of the current log files on another volume.
Restoring logs from crash-consistent (APiT) copy
Restore an any point in time (APiT) copy using logs.
Before you begin
This is applicable only in a RecoverPoint environment.
Note
Restoring logs from a crash-consistent copy is not a recommended practice as the
backup is not taken with the Exchange writer. However, the option can be used to
minimize data loss when application consistent copies for that time window are not
available to restore from.
Procedure
1. Restore a database from an application-consistent copy without recovering it.
2. Mount a copy from a newer point in time.
3. Copy the newer log files to the production log volume.
4. Use ESEUTIL /k Enn (Enn is the log prefix for the database) to check the logs, then
recover and mount the database.
Restoring logs from crash-consistent (APiT) copy
65
Protect Microsoft Exchange
Restoring a deleted Exchange database
AppSync can restore a database even if it is deleted from Exchange in standalone and
DAG environments.
Before you begin
l
If you deleted the database files and created an empty database, dismount the
database and delete its files. The database that you are restoring should not have
data and log files at the original location where they were when the empty database
was created. The log file signatures will not match those in the AppSync copy and the
restore will fail.
l
If you completely remove the database and recreate it, the database name and its file
path and names should be exactly the same as those in your AppSync copy. If you do
not recreate the deleted database, AppSync recreates it.
l
In a DAG environment:
n
There should be no active or passive copies of the deleted DAG database.
n
AppSync recreates and restores only the active database copy to the server that
created the AppSync copy. After the database has been restored and recovered,
you can recreate the DAG passive copies.
If you have not selected the Recover and mount the databases after restore option in the
Restore wizard, perform the following manual steps to recover the database.
Procedure
1. Copy the required logs from _restoredLogs directory to the directory where the
current logs reside.
2. If the log file prefix changed, rename the required log files to use the new prefix.
3. Delete the E<nn>restore.env file.
4. Recover the databases manually in soft recovery mode using the ESEUTIL command.
eseutil /r E<nn> /l <logpath> /s <chkpt file path> /d
<database path>
5. Delete the _restoredLogs directory that should be empty after the database is
recovered.
Item level restore
AppSync can restore individual Exchange 2010 and 2013 mailboxes and messages when
™
®
™
EMC ItemPoint for Microsoft Exchange Server is installed.
The following software must be installed and running on the mount host where you want
to perform item level restores:
l
AppSync host plug-in
l
EMC ItemPoint for Microsoft Exchange Server
l
Microsoft Exchange MAPI client
l
Microsoft Outlook 32-bit client
Refer to the ItemPoint documentation on the EMC Support website for further information
on ItemPoint.
66
EMC AppSync 2.2.3 User and Administration Guide
Protect Microsoft Exchange
Restoring individual mailboxes and mailbox items
Use ItemPoint to restore individual Exchange mailboxes and mailbox items.
Before you begin
This operation requires the Data Administrator role in AppSync.
Procedure
1.
Logon to the mount host as the user that will be running ItemPoint. You must be a
local administrator.
2. Select Recover > Restore > Individual items on a Databases or Copies page.
3. From the Select Copy page, select a copy from the list.
If the copy was made with RecoverPoint, you can Add a desired point in time and
select that copy. Specify a date and time (and DAG server if appropriate) when you
select this option. (If you started the wizard from a copy, you do not perform this
step.)
4. If the copy is unmounted you will then select mount options. From the Select Mount
Options page, select the mount host and the mount path (original path or default).
5. From the Item Level Restore Settings page, enter user name and password credentials
to access the item level restore wizard, and select a temporary directory.
The temporary directory is used by ItemPoint to store files used to recover data. These
files may require disk space equal to the size of the .log files. Use the user credentials
discussed in Step 1.
This operation requires the Microsoft Exchange MAPI client and Outlook 32-bit client.
For additional details on the required Exchange roles and permissions for the account
used to access ItemPoint, refer to the EMC ItemPoint for Microsoft Exchange Server
documentation.
6. From the Summary page, confirm your settings and select Finish to launch EMC
ItemPoint for Microsoft Exchange Server.
After you finish
l
After you finish, use ItemPoint to restore mailbox items. For more information about
using EMC ItemPoint for Microsoft Exchange Server, refer to the EMC ItemPoint for
Microsoft Exchange Server documentation.
Note
Do not unmount the copy that is being restored using Item Point until the entire process
is completed within Item Point.
Troubleshooting the EMC AppSync Exchange Interface service
Troubleshooting tips follow:
Is a reboot required?
If you are installing the AppSync host plug-in for the first time, you may need to reboot
your Exchange server after the install completes. In some environments, the install of the
Visual C++ 2010 runtimes require a reboot. In prior releases of AppSync, the install would
reboot automatically without warning. The automatic reboots have been removed with
Troubleshooting the EMC AppSync Exchange Interface service
67
Protect Microsoft Exchange
AppSync 2.0 and above. To verify if a reboot is required, look in Windows\temp
\appsync_host_plugin_setup.log for the following:
>: "C:\Program Files\EMC\AppSync Host
Plug-in\msredist\vcredist-10.0.40219.1-x64.exe" /q
Command.run(): starting stdout monitor...
Command.run(): starting stderr monitor...
Command.complete(): waiting for process to complete.
Command.complete(): process completed with exit
code: 3010
The exit code of 3010 indicates that a reboot is required.
Does the user account have the correct rights?
If the EMC AppSync Exchange Interface service fails to register properly, check the
ExchangeInterfaceInstall.log file in the AppSync host plug-in\logs directory. A
common problem is that the user account for running the service was not granted the
Log on as a batch job permission.
If AppSync fails to discover databases, verify the EMC AppSync Exchange Interface
service user account has been granted the correct Exchange permissions.
How to manually register the service.
1. Grant the user account that will run the EMC AppSync Exchange Interface service Log
on as a batch job and Log on as a service user rights.
2. Open a command prompt and navigate to the directory where the AppSync Host Plugin is installed. The default location is C:\Program Files\EMC\AppSync Host
Plug-in.
3. Run the following command to register the service and the DCOM component:
awExchangeInterface /service /user <"domain\username"> /
password <"password"> /nopriv For example: awExchangeInterface /
service /user mydomain /appsyncexchuser /password
mYp@55W0rd.
4. To configure the password for the DCOM component, run DCOMCNFG.
5. Expand Component Services > Computers > My Computer > DCOM Config.
6. Right click on EMC AppSync Exchange Interface and select Properties.
7. Click on the Identity tab.
8. Select This user and enter the user account and password from step 3.
9. Click OK.
10. Verify that you can start the EMC AppSync Exchange Interface service by running: net
start appsyncexchangeinterface.
11. Use the AppSync console to rediscover the server. Go to Settings > Servers, select the
server, and then click Rediscover.
12. Discover the Exchange mailbox databases. Go to Copy Management > Exchange and
click on the Exchange server. You may have to re-enter the credentials.
What to do if the service is partially registered?
If the rights and permissions are not granted properly to the user account, or if conflicting
software is installed, the EMC AppSync Exchange Interface service will not register
correctly. You may need to do a manual cleanup. Follow these steps:
68
EMC AppSync 2.2.3 User and Administration Guide
Protect Microsoft Exchange
1. Open a command prompt and navigate to the directory where the AppSync host plugin is installed. The default location is C:\Program Files\EMC\AppSync Host
Plug-in.
2. Run the following command to remove the service and delete the DCOM component:
awExchangeInterface /unregserver
3. Using the Services console (service.msc), verify that the EMC AppSync Exchange
Interface service is removed. If it persists, run: sc delete
AppSyncExchangeInterface
4. Using the Component Services console (DCOMCNFG), verify that the EMC AppSync
Exchange Interface DCOM component was removed. Expand Component Services >
Computers > My Computer > DCOM Config.
5. If the component persists, click DCOM Config, then in the center pane, click EMC
AppSync Exchange Interface, and then click Delete.
Troubleshooting the EMC AppSync Exchange Interface service
69
Protect Microsoft Exchange
70
EMC AppSync 2.2.3 User and Administration Guide
CHAPTER 6
Protect SQL Server
This chapter includes the following topics:
l
l
l
l
l
l
l
l
l
Overview of SQL Server support.............................................................................72
Support for AlwaysOn Availability Groups..............................................................75
SQL Server transaction log backup........................................................................ 75
Considerations for working with SQL Server in a cluster.........................................84
SQL Server User Databases folder......................................................................... 86
Protect a SQL Database......................................................................................... 87
Mount considerations for SQL Server.....................................................................99
SQL Server database restore overview.................................................................105
Repurposing SQL Server database copies............................................................117
Protect SQL Server
71
Protect SQL Server
Overview of SQL Server support
Use AppSync to create and manage application-consistent copies of Microsoft SQL Server
databases.
AppSync support for Microsoft SQL applications includes:
l
AlwaysOn Availability Group support.
l
Dynamic discovery of user databases during service plan run.
l
Support for databases on physical hosts, RDMs, and virtual disks on virtual hosts.
Note
AppSync only supports RDMs in physical compatibility mode. There is no support for
RDMs in virtual mode.
l
Protection for standalone and clustered production SQL Server instances.
l
Mount on a standalone server or cluster nodes of alternate cluster or production
cluster as non-clustered resource. Mount with recovery on a non-clustered instance.
Support for Repurposing SQL server database copies.
SQL Server prerequisites
Verify that the SQL Server configuration meets the prerequisites that are listed here. The
AppSync Support Matrix on https://elabnavigator.emc.com/eln/extendedSupport is the
authoritative source of information on supported software and platforms.
72
l
SQL Server database and its transaction logs must be on disks in the same storage
array.
l
The SQL Server database must be online during replication.
l
Full-text catalogs that are associated with a file group are included as part of a replica
of that file group. If the full-text catalogs are not located on supported storage,
protection fails. When using full-text catalogs, ensure that the storage device where
the catalog is located does not include data that is not related to the database.
l
If you want to recover databases from the mounted copy, the mount host must have
an installed SQL Server. It is recommended to use the same version of SQL Server on
the production and mount hosts.
l
In Hyper-V environments, AppSync requires the storage for SQL database and log files
to be on NPIV Fibre devices, iSCSI direct attached devices, or SCSI pass-through
devices. SCSI Command Descriptor Block (CDB) filtering must be turned off in the
parent partition. It is turned on by default. For SQL Server cluster, AppSync requires
storage to be on NPIV Fibre devices or iSCSI direct attached devices.
l
System databases are not supported.
l
SQL Server database snapshots are not discovered.
l
Creating a copy of a database mirror is not supported. Trying to do so results in an
error that the database is not in a valid state.
EMC AppSync 2.2.3 User and Administration Guide
Protect SQL Server
SQL Server supported configurations
AppSync provides support for the SQL configurations listed here.
l
Multiple SQL Server databases can exist on the same volume, or across multiple
volumes. However, it is best practice to not mix databases from more than one SQL
Server instance on a volume.
l
Multiple SQL Server instances can coexist on the same host.
Support for SQL Server on virtual disks
You can protect, mount and restore SQL Server standalone and clustered databases
residing on VMware virtual disks.
During protection:
l
For successful mapping, the Virtual Center must be added to the AppSync server and
discovery must be performed.
l
For successful protection, log files and database files must reside on virtual disks.
There cannot be a combination of physical and virtual storage.
l
Protection of SQL Server databases across virtual machines sharing the same
datastore is not supported.
l
When restoring SQL Server clustered databases, you must add all the owner nodes of
the SQL Server clustered instance to AppSync.
Required permissions and rights
Users require certain permissions and rights to protect databases in a SQL Server
environment. The user account must be configured to use either SQL Server
authentication or Windows authentication.
The Windows user account can either be a member of the local Administrators group or a
non-Administrator account with the restrictions outlined next.
In SQL Server 2012, the default virtual account used in the service startup account of the
database engine does not have the requisite file system permissions for accessing the
mounted or restored database files. Therefore, recovery of SQL databases may fail. To
overcome this, you must change the service startup account for the SQL Server database
engine to use a domain user account with appropriate privileges and permissions.
Setting up permissions for a domain account that does not have local administrator privileges
Additional setup is required if you need to use a domain account that does not have local
administrator privileges.
Procedure
1. Create a Windows domain user (for example, sqluser) and make it part of the Domain
Users group.
2. In SQL Server Management Studio, create a new login, using the newly created
domain account and select Windows authentication.
3. In the General page, select master as the default database.
4. In the Server Roles page, select sysadmin and public.
5. In the User Mapping page, set the database role membership to public.
6. Add the user to each SQL Server instance on which this user needs access:
SQL Server supported configurations
73
Protect SQL Server
a. On the domain controller: On the hosts added to the domain: Start > Programs >
Administrative Tools > Domain Controller Security Policy
On the hosts added to the domain: Start > Programs > Administrative Tools > Local
Security Policy
b. Access security settings and allow login locally (Security Settings > Local Policies >
User Rights Assignment > Allow log on locally)
c. Add the user (the example is sqluser) you created earlier.
7. Log in to the domain controller machine for each host added to that domain that uses
AppSync and set the Security policy.
8. Grant this user read and write permissions on the directory where the AppSync plug-in
is installed (typically C:\Program Files\EMC\AppSync Host Plug-in).
9. Use this user from AppSync when you configure protection or perform other actions
that require access to SQL Server.
10. At the time of restore, if you select the option to back up the transaction logs to a file,
the user must have rights to the target directory.
Setting permissions for a local, non-administrator user
A user account that does not have local administrator privileges needs certain
permissions before it can be used to access SQL Server from AppSync.
Procedure
1. Create a Windows user and make it part of the Users group.
2. In SQL Server Management Studio, create a new login, using the newly created
account. For the authentication type, select Windows authentication.
3. In the Server Roles page, select sysadmin and public.
4. In the User Mapping page, set the database role membership to public.
5. Add the user to each SQL Server instance on which this user needs access:
a. On the host running the plug-in, set the security policy. On the domain controller,
run Start > Programs > Administrative Tools > Local Security Policy.
On the hosts added to the domain: Start > Programs > Tools > Local Security Policy.
b. Access security settings and allow login locally (Security Settings > Local Policies >
User Rights Assignment > Allow log on locally).
c. Add the user (the example is sqluser) you created earlier.
6. Grant this user read and write permissions on the folder where the AppSync plug-in is
installed.
7. If you select the restore option to back up the transaction logs to a file, the user must
have rights to the target directory.
Update login credentials for a SQL Server instance
If the credentials for a SQL Server instance have changed, you need to update them in
AppSync.
Before you begin
This operation requires the Data Administrator role in AppSync. In addition, you should
know the new credentials for the SQL Server instance.
74
EMC AppSync 2.2.3 User and Administration Guide
Protect SQL Server
Procedure
1. Select Copy Management.
2. Select Microsoft SQL Server.
3. Select an instance.
4. Click Connection Settings from the row of buttons below.
5. Enter the SQL Server credentials.
The credentials can be a Windows user or a SQL user with required privileges.
Support for AlwaysOn Availability Groups
The Availability Groups can be part of clustered and non-clustered SQL Server instances
installed on AlwaysOn Failover clusters.
AppSync supports Full or Copy backups of primary databases and Copy backups of
secondary databases. The Auto Switch to Copy option in the SQL Server service plan's
Create copy phase allows you to switch from Full to Copy for secondary databases.
Special considerations when you are using AlwaysOn Availability Groups:
l
To protect secondary databases, they must be read-only. The Readable
Secondary option in the SQL Server Management Studio must be set to Yes;
Read-intent only is not supported.
l
Do not use the original path when mounting an AppSync copy to a node in the same
cluster if that node hosts a copy of the database.
l
It is recommended to protect replicas in the Synchronous-commit mode.
l
The considerations for working with SQL Server in a cluster also apply to Availability
Groups. See Considerations for SQL in a cluster on page 84.
l
Multi-subnets are supported for AlwaysOn Availability Groups as long as none of the
database copies belong to a clustered SQL Server instance.
SQL Server transaction log backup
AppSync 2.1 and above supports SQL Server transaction log backup. Get key
considerations as well as restrictions before implementing your backups.
Every SQL Server database has a transaction log. Write the log backups to EMC storage
systems that are supported by AppSync so you can create copies of the log backup
volume. If you back up logs for databases in a failover cluster environment, use shared
storage or a network share so the log backups are written to the same location.
You can use transaction log backups during recovery of a production database or when
making a copy of a production database. Depending on the database recovery model, the
transaction log can become full. To prevent the accumulation of logs, regularly run
transaction log backups with truncation enabled.
AppSync can backup transaction logs in AlwaysOn Availability Group (AAG)
environments. It can back up primary or secondary database copies. If truncation is
enabled, to initiate truncation, back up either the primary or secondary database
transaction log.
Transaction log backups are supported using only streaming back up; they are not
supported using VSS hardware snapshot technology. You can use AppSync to back up
transaction logs to a file. The file can be written to a local volume or network share using
a UNC path.
Support for AlwaysOn Availability Groups
75
Protect SQL Server
Restrictions
l
To back up a transaction log, the database recovery model must be either “Full” or
“Bulk-logged.” AppSync skips backing up the log for any database with the simple
recovery model.
l
To create any log backups with log truncation, first create at least one full database
backup.
l
To truncate transaction logs, AppSync must have a Full database backup copy.
l
Subscribe a database to only one service plan with log backup enabled.
l
To truncate logs in an AAG environment, subscribe only one copy of a database to a
service plan that is configured for Full database backups and transaction log backups
with log truncation.
l
To back up transaction logs for databases that belong to an availability group, alter
the schedule so that different copies of the database are not backed up at the same
time.
Related topics
l
Configure SQL Server transaction log backup on page 76
l
Run log backup on demand on page 79
l
View log backups for a service plan on page 80
l
View log backup list for a single database on page 82
Configure SQL Server transaction log backup
Learn how to enable transaction log backups for an SQL Server service plan, by selecting
the Enable log backup checkbox on the Create Copy options page of the AppSync
console.
Before you begin
Verify that the user account you select for backups has full control of the directory. This
account is the user account that you entered when discovering databases. Also verify
that the account configured for the SQL Server Database Engine Service of the SQL Server
instance being protected has full control of the backup directory.
After you select this checkbox, the Transaction Log Backup Options dialog box is enabled
where you can customize when and how to run log backups and where to write the log
backup files. Transaction log backups run sequentially.
76
EMC AppSync 2.2.3 User and Administration Guide
Protect SQL Server
Figure 3 Transaction Log Backup Options dialog box
Procedure
1. Use the Schedule field to set log backup runs.
You can select to run the transaction log backup once, immediately after a database
backup is run, or you can select to schedule log backups. You can set log backup
schedules to run every 15 or 30 minutes or every 1 to 24 hours. If you set a service
plan to run on demand, you disable the log backup schedule.
When you schedule log backups to run at a specified interval, the service plan will
have two schedules associated with it: one for database backups and one for log
backups. The log backup is referred to as the alternate schedule. Log backups run
between database backups using the alternate schedule.
2. Edit the Backup path field to set the location where AppSync writes log backup files.
Default path uses the SQL Server instance default backup directory. You can also
enter a path on any volume on the server or the UNC path of a network share.
AppSync creates the directory if it does not exist. It creates a subdirectory using the
name of the SQL Server instance. The log backup file names have the following
format: EMC_AppSync_databasename_timestamp.trn, for example,
EMC_AppSync_AdventureWorks_2014_10_18_15_38_32.trn
3. Use the Free space on volume field to set a value to verify the amount of free space on
the volume before AppSync begins a transaction log backup.
If not enough free space is available, an alert is generated and the log backup fails.
Configure SQL Server transaction log backup
77
Protect SQL Server
4. Use the Backup group size field to control the number of parallel log backups for an
SQL Server instance. The default value is 5, (AppSync runs log backups in groups of
five).
For example, if you subscribe 15 databases from the same SQL Server instance to a
service plan, three log backups will run in parallel. Transaction log backups run
sequentially.
5. Select or clear the Truncate the logs field when you create Full database backups.
This field is checked by default when you select Full backup type, and it is disabled
when you select Copy . To protect secondary databases, truncate logs, select Auto
switch to Copy and Truncate the logs.
6. To perform a checksum on the log backup, select the Checksum the backup field.
7. Set Minimum Retention Hours option to control when transaction log backup files are
deleted.
Transaction log backup expiration is done when no older database backups exist.
AppSync deletes the log backup files and the log backup information contained in the
AppSync database. The default setting is 24 hours which means that AppSync will not
expire any log backup before it is a minimum of 24 hours old. The valid range is 0 to
10,000 hours.
Configure log backup scripts
You can run scripts before and after log backups by enabling the pre- and post- log
backup scripts.
The pre-log backup script runs on the production host. The post-log backup script can run
on the production host or the mount host (if mount is enabled), or you can specify a
server. The server must have the AppSync host plug-in installed.
78
EMC AppSync 2.2.3 User and Administration Guide
Protect SQL Server
Figure 4 Configure log backup scripts
Run log backup on demand
You can schedule and run SQL Server log backups, or you can run log backups on
demand.
Before you begin
To run log backups, make sure you check the service plan’s Enable Log Backup option.
You can run a log backup on demand for an entire service plan, or run a log backup on
demand for a single database instance.
Procedure
1. To run a log backups on demand for an entire service plan, go to Service Plans >
Microsoft SQL Server, select the desired service plan enabling the Run Log Backups
button, and then click Run Log Backups to run the backup for the entire service plan.
2. To run a log backup for a single database go to Copy Management > Microsoft SQL
Server, select an SQL Server instance, select the desired database, and then select
the Log Backups tab. Finally, select the appropriate service plan from the Create log
backup using plan list to run the log backup.
Run log backup on demand
79
Protect SQL Server
View log backups for a service plan
The list of SQL Server log backups can be viewed from the Service Plan Log Backups tab
or from the Database Log Backups tab.
Before you begin
This operation requires the Data Administrator role in AppSync.
The list of copies can be filtered by time of creation, and by service plan. In the Service
Plan Copies tab, you can also filter by instance.
Procedure
1. To view the list of all log backups for a service plan, navigate to Service Plans >
Microsoft SQL Server.
2. Select a service plan.
3. Click the Log Backups tab.
Results
You can now view the log backup list for the service plan. The following table describes
details about the log backup:
Table 8 Service Plan log backup details
Column
Description
Status
l
Green: successful
l
Yellow: some log backups completed with errors when the service plan ran.
l
Red: failed
Instance
SQL Server instance name
Database
SQL Server database name
Name
Name of the log backup copy. The copy is named with the time at which it was
made.
Service Plan Name of the service plan associated with the log backup.
Truncated
Indicates if the transaction log was truncated by the log backup. Yes, if the log was
truncated, otherwise No.
Backup File
The name of the log backup file and its location.
View SQL database copies
View the list of database copies by browsing to Copy Management > Microsoft SQL Server
and selecting a SQL Server, then a database.
Before you begin
This operation requires the Data Administrator role in AppSync.
You can also see details of a copy from the Copies tab of the service plan.
You can filter the list of copies by time of creation, and by service plan. In the Service Plan
Copies tab, you can also filter by instance.
80
EMC AppSync 2.2.3 User and Administration Guide
Protect SQL Server
Table 9 Service Plan Copy details
Column
Description
Status
l
Green: successful
l
Yellow: completed with errors
l
Red: failed
Name
Name of the copy. The copy is named with the time at which it was
made.
Service Plan
Name of the service plan that is associated with the copy. Service
plan field is blank for Repurpose copies.
SQL Server Backup Type
Type of SQL backup: Full or Copy
l
Full protects the database, and the active part of the transaction
log. Copy protects the database and the active part of the
transaction log without affecting the sequence of backups.
l
Secondary databases are read-only and can only be backed up
with the Copy backup type. Auto Switch to Copy is enabled
only when Full is selected as the backup type. However it is
unchecked by default. Checking Auto Switch to Copy tells
AppSync to check if the database role is Secondary, and if so, to
switch the backup type to Copy. If Auto Switch to Copy is not
enabled, backups fail for all secondary databases.
Mount Status
Shows if the copy is mounted. If mounted, the name of the mount
host displays.
Recovery Status
Available values:
l
Not Recovered - when copy is not mounted or it is a file system
mount
l
Successful - when Recovery is successful
l
Failed - when Recovery failed
Availability Group
The Availability Group column lists the availability group the
database belongs to.
Generation
Used for repurposed copies, this column describes how many
generations removed the copy is from the production database.
Source
This column displays the source database or copy from which a copy
was created.
Copy Type
Type of copy can be one of the following:
l
RecoverPoint Continuous Data Protection Bookmark
l
RecoverPoint Continuous Remote Replication Bookmark
l
VNX Snap
l
VNXeSnap
l
VNXe FileSnap
l
VMAX Snap, VMAX Clone
View SQL database copies
81
Protect SQL Server
Table 9 Service Plan Copy details (continued)
Column
Description
For the preceding copy types, the following additional details are
displayed in the Service Plan Copies tab:
l
Instance: The SQL Server instance that hosts the database.
l
Database name: The name of the copy's database.
l
Time: The time at which the database copy was made.
l
Server/cluster: Name of the server or the cluster that hosts the
SQL Server instance.
l
Site: RecoverPoint
Copy type continued:
l
XtremIO snapshot
l
VMAX 3 SnapVXClone, SnapVXSnap
Note
A Repurpose button on this page is enabled. When you select a 1st Generation copy, the
Repurpose wizard launches where you can create 2nd Generation copies.
View log backup list for a single database
You can also view log backups for a single database.
Follow these steps:
Procedure
1. Navigate to Copy Management > Microsoft SQL Server, and then select an SQL Server
instance.
2. Click the User Databases folder.
3. Click on a database in the list and select the Log Backups tab.
Results
You can now view log the log backup list for the database. The following table describes
details about the log backup:
Table 10 Database log backup details: SQL Server instance
Column
Description
Status
l
Green: successful
l
Yellow: some log backups completed with errors when the service plan ran.
l
Red: failed
Name
Name of the log backup copy. The copy is named with the time at which it was
made.
Service Plan Name of the service plan associated with the log backup.
82
EMC AppSync 2.2.3 User and Administration Guide
Protect SQL Server
Table 10 Database log backup details: SQL Server instance (continued)
Column
Description
Truncated
Indicates if the transaction log was truncated by the log backup. Yes, if the log was
truncated, otherwise No.
Backup File
The name of the log backup file and its location.
Log backup expiration
AppSync expires log backups when the service plan runs to create a new log backup.
During expiration, AppSync deletes the log backup file and removes information about
the backup from the AppSync database.
Log backups are always based off the previous Full database backup. However, you do
not have to use AppSync to create the Full database backup. You can use AppSync to
create a Copy database and log backup.
Additionally, AppSync can create Full database backups and log backups with, or without
log truncation. Log backup expiration behavior depends on the type of database backup
you create.
Log backups are eligible for expiration when the following conditions occur:
l
The log backup is older than the service plan Minimum Retention Hours setting.
l
All older database backups are expired. The database backups included in this check
depends on the SQL Server Backup Type.
n
If the log backup service plan has SQL Server Backup Type set to Copy, only
database backups created by that service plan are considered when looking for
older database backups.
n
If the log backup service plan has SQL Server Backup Type set to Full, then Full
database backups created by any service plan are considered.
Example 1: consider the following scenario:
l
Service plan has log backup enabled.
l
Database backup type set to Copy.
l
Rotation set to one.
l
Log backup minimum retention is set to 24 hours.
The service plan has run several times, creating a database backup and several log
backups. The service plan runs again, creating a database backup and expiring the first
database backup. This leaves several log backups with no older database backup. The
service plan runs again, creating a log backup and expiring all of the previous log
backups that are at least 24 hours old.
Example 2: consider the following scenario:
l
You have two service plans.
l
Both have database backup type set to Full.
l
Service plan 1 is scheduled to run a database backup once a week with rotation set
to four.
l
Service plan 2 is scheduled to run daily at 8 PM with a rotation of seven.
l
Service plan 2 has log backup enabled to run every hour and the log backup
minimum retention is set to 24 hours.
Log backup expiration
83
Protect SQL Server
l
Both service plans have been running.
l
Service plan 1 has four database copies and service plan 2 has seven database
copies. Service plan 2 also has many log backups that were run between each of the
seven database copies.
l
Service plan 2 runs again and creates a database copy and then expires its oldest
copy. It runs an hour later to create a log backup and looks for log backups that are
eligible for expiration.
No log backups are eligible because service plan 1 has Full database backups that are
older than all of the log backups. The next time service plan 1 runs, the oldest database
backup will be expired. Log backups will then be eligible for expiration.
Manual expiration of log backups
You can also expire log backups manually.
To expire log backups for several databases:
1. Navigate to Service Plans > Microsoft SQL Server, and click on a service plan.
2. Click the Log Backups tab.
3. Select the log backups that you would like to expire and then click Expire.
4. Click OK on the confirmation dialog. AppSync will delete the log backup file and
remove information about the backup from the AppSync database.
To expire log backups for a single database:
1. Navigate to Copy Management > Microsoft SQL Server and select an SQL Server
instance.
2. Click the User Databases folder.
3. Click on a database in the list and select the Log Backups tab.
4. Select the log backups that you would like to expire and then click Expire.
5. Click OK on the confirmation dialog.
Considerations for working with SQL Server in a cluster
There are special considerations when working with SQL Server in a cluster.
When protecting SQL Server databases in a clustered environment, you must install the
AppSync host plug-in on all of the nodes that are possible owners of a SQL Server
instance. You can use the AppSync console to install the plug-in or manually install the
plug-in on each server. Once the plug-in is installed, use the AppSync console to add the
network name or IP address of the SQL Server clustered instances.
When protecting clustered SQL Server instances:
l
You must add SQL Server virtual server to AppSync after installing the AppSync host
plug-in software on each node.
l
Only single subnets are supported.
l
Mounting AppSync copies:
n
You can mount AppSync copies created on clustered databases to a standalone
server or cluster node.
n
You can mount AppSync copies created on standalone databases to standalone
server or a cluster node.
When mounting a SQL Server copy to a cluster:
84
EMC AppSync 2.2.3 User and Administration Guide
Protect SQL Server
l
Supports mount to either alternate cluster or production cluster as non-clustered
resource.
l
Only supports mount to a cluster node, does not support mount to virtual server or
mount and recover to clustered instance.
l
Mount is supported in the environments of VMAX, VNX, VNXe, XtremIO or
RecoverPoint. The AppSync Installation and Configuration Guide describes the required
storage configuration steps.
l
Select the appropriate mount option that applies for cluster mount based on your
cluster and storage configuration.
l
Manually disable automount. Run diskpart at a command prompt then enter
automount disable at the DISKPART> prompt.
Special considerations for mount to production cluster:
l
Mounting to a production cluster node using the original path is not supported.
l
If the original server is a virtual server, mounting to a production cluster using the
"original server" option is not supported and mounts fail.
l
Statically mounting RecoverPoint SQL copies to production clusters is not supported.
You must use dynamic mounts for this scenario.
l
Performing a RecoverPoint mounted restore while the copy is mounted to a
production cluster is not supported.
l
Mounting RecoverPoint SQL APIT copies to production cluster nodes is not supported.
Example 1 Protect databases owned by clustered instances of SQL Server
In this example, you want to protect databases owned by clustered instances of SQL
Server. In addition, some of those databases belong to AlwaysOn availability groups.
Refer to the following figure for this example:
Figure 5 Cluster information
This cluster has the following configuration:
Considerations for working with SQL Server in a cluster
85
Protect SQL Server
Example 1 Protect databases owned by clustered instances of SQL Server (continued)
l
lrmq093 is a SQL Server virtual server and it hosts databases belonging to AlwaysOn
Availability Groups.
l
lrmq094 is a SQL Server virtual server.
l
lrmq120, lrmq121, and lrmq122 belong to a SQL Server AlwaysOn failover cluster.
l
lrmq120 and lrmq121 are possible owners of the clustered SQL Server instances
owned by lrmq093 and lrmq094.
l
lrmq122 has a standalone instance of SQL Server installed that hosts databases
belonging to AlwaysOn availability groups.
l
lrmq126 is the mount host with a standalone instance of SQL Server installed
To protect the databases belonging to the clustered and standalone instance, follow
these steps:
1. Use the AppSync console to add lrmq120, lrmq121, and lrmq122. AppSync will
install the plug-in on these servers and discover any non-clustered instances. If you
need a mount host, you can add lrmq126 now.
2. With the AppSync console, add the virtual servers for the clustered instances, add
lrmq093 and lrmq094.
SQL Server User Databases folder
The SQL Server User Database folder contains all the user databases for this SQL Server
instance that have been discovered and stored in the AppSync database.
From the Protect button, you can subscribe the folder to a plan. By doing so, all the
databases part of this folder are also protected. Once protected, the Service Plan column
displays the name of the plan.
Clicking on the User Databases folder lists the individual databases part of this SQL
Server instance.
In the Databases page, an entry in the Service Plans column tells you that all the
databases that are part of the folder are protected. Any user databases added to the
instance will also be protected. AppSync will automatically stop protecting any databases
removed from the instance.
Note
If one or more user databases for an SQL Server instance are subscribed to a service plan,
you cannot subscribe the User Databases folder to the same service plan. Conversely, if
the User Databases folder is subscribed to a service plan, you cannot subscribe
individual user database instances to the same service plan.
Discover SQL Server instances
To keep AppSync up-to-date, you should discover SQL Server instances when there is
creation or deletion of instances.
Before you begin
This operation requires the Data Administrator role in AppSync.
86
EMC AppSync 2.2.3 User and Administration Guide
Protect SQL Server
Procedure
1. Navigate to Copy Management > Microsoft SQL Server.
2. From the Discover Instances popup button below:
l
Select On Server and select one of the servers to discover instances from.
l
Select Add Servers to add a new server to AppSync.
Discovering SQL Server databases
AppSync discovers new user databases on demand or automatically on a service plan
run.
When you click the User Databases folder the first time, AppSync discovers databases
and lists them. To manually discover databases again, click Discover Databases in the
Databases page.
On the other hand, when you subscribe the User Databases folder to a plan, databases
are automatically discovered on each run of the plan. All databases that are currently
ONLINE, including those that were added to the SQL instance after the last service plan
run, are automatically protected.
If individual databases are subscribed to a plan instead of the User Databases folder,
AppSync does not automatically discover any new databases that were created after the
last run of the plan. In this case, AppSync rediscovers the database information of all the
databases originally subscribed to the plan and protects the ones that are ONLINE.
Protect a SQL Database
Protect a SQL database by subscribing it to an AppSync service plan.
To optimize performance, AppSync creates copies of a maximum of 35 databases per
instance. If more than 35 databases are subscribed per instance, AppSync breaks them
into groups of 35 and creates copies of the groups sequentially. If more than 35
databases are subscribed to a service plan, and the databases reside on same storage
unit (CG, LUN, DS, and so on), the split into groups with 35 databases does not occur. A
single copy is desirable for a configuration when storage is on the same storage unit.
This number (35) is a server setting and can be modified, if required. Contact EMC
Support to do so.
You can protect objects in different ways from different places in AppSync:
l
Choose Subscribe to Plan and Run when you want to protect a selected database
immediately. The service plan is run for the database alone.
l
Choose Subscribe to Plan when you want to schedule the protection for later.
Protection for databases that are part of the service plan are run at the scheduled
time.
l
Choose an appropriate service plan from Create copy using plan in the database
Copies page.
l
Choose Run from the SQL Server Service Plans page to run the whole plan
immediately.
Protect a SQL Database
87
Protect SQL Server
Configuring protection for SQL Server database
You subscribe a database or the User Databases folder to a service plan and run the
service plan immediately, or schedule the service plan to run at a later time.
Before you begin
This operation requires the Data Administrator role in AppSync.
Procedure
1. Navigate to Copy Management > Microsoft SQL Server.
2. Click a server instance to display its databases.
l
To protect all databases within the User Database folder, select the User Database
folder.
l
To protect an individual database, click User Databases and select a database
from the list.
3. From the Protect popup button below, select the appropriate service plan from:
Option
Description
Subscribe to Plan and To subscribe the database for protection and run the plan
Run
immediately for the selected database(s).
Subscribe to Plan
To subscribe the database for protection. Protection for all
databases that are part of the service plan are executed at
the scheduled time.
Unsubscribing a database from a service plan
When you unsubscribe an individual database from a service plan, all existing database
copies will be retained; only further protection will be removed.
Before you begin
This operation requires the Data Administrator role in AppSync.
Procedure
1. Navigate to Copy Management > Microsoft SQL Server.
2. Click the SQL server instance.
3. Click User Databases.
4. Select the database to unsubscribe from a service plan.
l
Select the plan to unsubscribe from Protect > Unsubscribe from Plan. Only plans to
which the database is subscribed to are in the popup list.
l
To unsubscribe from all service plans, select Unsubscribe from Plan > All.
Discovering SQL Server databases
Use the Discover New Databases command to update the SQL Server databases known
to AppSync.
Before you begin
This operation requires the Data Administrator role in AppSync.
88
EMC AppSync 2.2.3 User and Administration Guide
Protect SQL Server
Procedure
1. Navigate to Copy Management > Microsoft SQL Server.
2. Click a server instance, then the User Databases folder.
3. In the User Databases page, click Discover New Databases from the row of buttons
below.
Discovery can take several minutes to complete depending on the size of the
instance.
Creating a database copy from the Copies page
Create a copy of a database by subscribing it to an AppSync SQL Server service plan from
the Copies page.
Before you begin
This operation requires the Data Administrator role in AppSync.
Procedure
1. Navigate to Copy Management > Microsoft SQL Server.
2. Click a server instance and then click User Databases to display its databases.
3. Click on a folder to display User Databases.
4. From this list, click the database to view its copies.
5. From the Create a copy using plan list, select the appropriate service plan.
The service plan runs immediately for the database.
Expiring an SQL database copy on demand
Expiring a database copy removes it from the AppSync database and can free up storage,
depending on the replication technology and copy state.
Before you begin
This operation requires the Data Administrator role in AppSync.
Expiring a copy that was made with RecoverPoint does not remove the corresponding
bookmark from RecoverPoint itself.
Procedure
1. Select Copy Management > Microsoft SQL Server.
2. Click a SQL Server instance to display its database folders.
3. Click the User Database folder.
4. Click the database whose copies you want to expire.
5. From the Copies page, select one or more copies to expire.
You can also perform this action from the Service Plan's Copies tab.
6. Select Expire from the row of buttons below.
Verify that you want to expire the copy you selected and any associated copies listed
and confirm.
Creating a database copy from the Copies page
89
Protect SQL Server
Service plan summary and details
The service plan Settings tab shows the name, description, schedule, and status of the
service plan. Click the phases for detailed service plan settings and other tabs for
information about subscriptions, lists of copies and events generated by the plan.
Service plan schedule
The schedule of a service plan is set in the Plan Startup phase.
The Startup Type (scheduled or on demand) determines whether the plan is run
manually, or configured to run on a schedule. Options for scheduling when a service plan
starts are:
l
l
l
l
Specify a recovery point objective (RPO)
n
Set an RPO of 30 minutes or 1, 2, 3, 4, 6, 8, 12, or 24 hours
n
Minutes after the hour are set in 5 minute intervals
n
Default RPO is 24 hours
Run every day at certain times
n
Select up to two different times during the day
n
Minutes after the hour is in 5 minute intervals
n
There is no default selected
Run at a certain time on selected days of the week
n
One or more days of the week (up to all seven days) can be selected
n
There is no default day of the week selected. Default time of day is 12:00 AM.
Run at a certain time on selected days of the month
n
Select one or more days of the month (up to all days)
n
Select one time of day. Available times are at 15 minute intervals.
n
Default is the first day of the month
Overriding service plan schedules
You can set individual schedules for databases subscribed to a service plan, overriding
the generic recurrence setting.
Before you begin
This operation requires the Service Plan Administrator role in AppSync.
You can override only the settings of the recurrence type already selected for the service
plan.
Procedure
1. Navigate to Service Plans and select one of the plans from the list.
2. From the Settings tab, select the Plan Startup phase.
You will see the Plan Startup Defaults pane on the right.
3. Note the Recurrence Type selected for the plan.
A recurrence type can be set only if Scheduled is selected as the Startup Type.
4. Select the Start service plan phase.
You will see the Start service plan pane on the right.
90
EMC AppSync 2.2.3 User and Administration Guide
Protect SQL Server
5. Note the Recurrence Type selected for the plan.
A recurrence type can be set only if Automatic is selected in the Startup.
6. Click the Plan Startup Overrides tab.
You can see the list of all databases subscribed to the plan.
7. Select one or more databases and click Override Schedule.
The Override Schedule dialog is displayed.
8. Set the schedule based on your requirement and click OK.
For example, if the default recurrence type is On specified days of the month, and the
rule setting is to Run at 12:00 AM on the 1st day of every month, you can override the
time and the day for individual datastores.
A Pencil icon indicates that default settings have been overidden.
Application discovery
Before creating the User Database folder's copy, AppSync examines the SQL Server
instance to look for changes such as addition, deletion, renaming, or movement of
databases. If individual databases are being protected, AppSync rediscovers information
about the selected database. A database is protected only if it is in the ONLINE state.
There are no user settings associated with this phase and it cannot be disabled.
Application mapping
After discovering the application, AppSync maps it to array storage, and protection
services such as RecoverPoint.
There are no user settings associated with this phase and it cannot be disabled.
Pre-copy script
To perform preparatory steps before creating a copy, specify a pre-copy script and
parameters on a service plan's Settings tab.
The pre-copy script runs according to the schedule set in the Plan Startup phase. Valid
script formats are .bat, .exe, and .ps1 (PowerShell scripts). You can optionally enter
credentials to run the script as a specific user. The script runs as Local System by default.
AppSync does not support running of PowerShell scripts directly. You usually must wrap
them in a .bat file. The other option is to make the default "Open" on ps1 files C:
\WINDOWS\system32\WindowsPowerShell\v1.0\powershell.exe. When the
PS script runs, you may get an error and you must set an appropriate execution policy.
To run PowerShell commands from scripts:
1. Specify the full pathname to the PowerShell command file in the .bat file:
powershell -command C:\PshellCommands.ps1 <nul
2. Set the PowerShell execution policy so you can run the script. For example, the first
line in the .bat file should look like the following for an unrestricted policy:
powershell -command set-executionpolicy unrestricted <nul
3. To ensure correct termination of the PowerShell session, add <nul to the end of the
line that calls the PowerShell script. The default location of the script is
%ProgramData%\EMC\AppSync\scripts\ on the application host.
Exact parameters depend on the script. Parameters with spaces must be enclosed in
double quotes.
Service plan summary and details
91
Protect SQL Server
This phase can be enabled or disabled. This operation requires the Service Plan
Administrator role in AppSync.
Create copy
The Create Copy phase creates a copy based on the replication technology specified in
the service plan.
This phase specifies the backup type of SQL Server copy to make. For VNX Snapshot
copies, this phase also sets the period for automatic expiration of the copies.
Review Overview: Service Plan on page 10 for more service plan copy information.
SQL Server backup type
Two main backup types are supported: Full and Copy.
l
Full protects the database, and the active part of the transaction log. This copy type is
typically used when the copy will be considered a backup of the database or when
the copy will be mounted in order to use a third-party product to create a backup of
the database. This type of copy allows you to restore transaction logs to bring the
database forward to a point in time that is newer than the copy, assuming you have
backed up those transaction logs. AppSync uses Microsoft SQL Server’s VDI snapshot
feature to create this type of copy.
n
Auto Switch to Copy is enabled only when Full is selected as the backup type.
However it is unchecked by default. Checking Auto Switch to Copy tells AppSync
to check if the database role is Secondary, and if so, to switch the backup type to
Copy.
Note
If Auto Switch to Copy is not enabled, backups fail for all secondary databases.
l
Copy protects the database and the active part of the transaction log without
affecting the sequence of backups. This provides DBAs with a way to create a copy
without interfering with third-party backup applications that may be creating full
and/or differential backups of the SQL Server databases. AppSync uses Microsoft
SQL Server’s VDI snapshot feature to create this type of copy.
Note
Secondary databases are read-only and can only be backed up with the Copy backup
type.
Automatic expiration of copies
The automatic expiration value in a service plan's Create Copy phase specifies the
maximum desired number of Snap, Clone or Bookmark copies that can exist
simultaneously.
When the "Always keep x copies" value is reached, older copies are expired to free
storage for the next copy in the rotation. Failed copies are not counted. AppSync does not
expire the oldest copy until its replacement has been successfully created. For example,
if the number of copies to keep is 7, AppSync does not expire the oldest copy until the
8th copy is created.
AppSync does not expire copies under the following circumstances:
Mounted copies are not expired.
A copy that contains the only replica of a database will not be expired.
This setting is independent of the VNX pool policy settings in Unisphere for automatic
deletion of oldest snapshots. The service plan administrator should work with the storage
l
l
92
EMC AppSync 2.2.3 User and Administration Guide
Protect SQL Server
administrator to ensure that the VNX pool policy settings will enable the support of the
specified number of snapshot copies for the application residing in that pool.
Include RecoverPoint copies in expiration rotation policy: Check this option to include
RecoverPoint copies when calculating rotations.
Note
If this option is not selected, then RecoverPoint copies will accumulate, and will remain
until the bookmarks fall off the RecoverPoint appliance.
Post-copy script
To perform cleanup or other post-copy steps after creating a copy, specify a post-copy
script and parameters in a service plan's Settings tab.
The script runs on successful completion of the Create copy phase. Valid script formats
are .bat, .exe, and .ps1 (PowerShell scripts). You can optionally enter credentials to run
the script as a specific user. The script runs as Local System by default.
When AppSync creates copies of application items in a service plan, it may break up the
application items and place them in separate groups for protection. This action can be for
performance reasons (for example, VSS for Exchange and SQL) or because items in a
service plan may be protected by different replication technologies. For example, a
service plan may contain some application items that are protected by VNX Snapshots
and some by RecoverPoint bookmarks. As a result, application items in these groups are
protected independently.
When AppSync calls a post-copy script, it passes the copies which were created in the
group by calling the script with -appCopies <APP1> <APP2>, where APP1 and APP2 are the
names of the application items in that grouping.
AppSync does not support running of PowerShell scripts directly. You usually must wrap
them in a .bat file. The other option is to make the default "Open" on ps1 files C:
\WINDOWS\system32\WindowsPowerShell\v1.0\powershell.exe. When the
PS script runs, you may get an error and you must set an appropriate execution policy.
To run PowerShell commands from scripts:
1. Specify the full pathname to the PowerShell command file in the .bat file:
powershell -command C:\PshellCommands.ps1 <nul
2. Set the PowerShell execution policy so you can run the script. For example, the first
line in the .bat file should look like the following for an unrestricted policy:
powershell -command set-executionpolicy unrestricted <nul
3. To ensure correct termination of the PowerShell session, add <nul to the end of the
line that calls the PowerShell script.
When AppSync runs the post-copy script, it is run for the application items that are part of
a group. If there are multiple groups, the post-copy script runs multiple times. When
AppSync runs the post-copy script, it passes the list of application items in the
replication group as arguments to the script, right after the user arguments. The syntax is:
-applicationCopies <ITEM1> <ITEM2> <ITEM3>
where <ITEMx> is the name of the application item that is being protected.
The default location of the script is %ProgramData%\EMC\AppSync\scripts\ on
the application host.
Exact parameters depend on the script. Parameters with spaces must be enclosed in
double quotes.
Service plan summary and details
93
Protect SQL Server
This phase can be enabled or disabled. This operation requires the Service Plan
Administrator role in AppSync.
Unmount previous copy
The service plan unmounts a previously mounted copy after creating the new copy. The
exception is a copy that was mounted on-demand as opposed to by the service plan; in
this case the on-demand mounted copy is not unmounted.
All the recovered databases are shut down as part of this phase. There are no user
settings associated with this phase and it can be enabled or disabled.
Mount copy
The Mount copy phase either mounts the copy or mounts and recovers the copy. This
phase can be enabled or disabled.
In the Mount Copy Defaults settings, you can set values to Mount copy or Mount and
recover copy.
In the Mount copy settings, you set the mount host value, mount path and mount
permissions (read-only or read-write). Other mount settings determine where the SQL
metadata files are copied and the RecoverPoint image access type.
Field
Description
Mount on
Server
The server on which to mount the copy. Only the nodes of the cluster
or standalone hosts are available for selection. SQL virtual machines
are filtered out.
Mount with
access
Type of access the copy should be mounted with.
Mount path
Copy metadata
files to
94
l
The Default Mount Path is %SystemDrive%\AppSyncMounts\
%%ProdServerName%%.
l
To specify the value of a Windows environment variable in the
mount path, delimit the variable name with single percent signs
(%).
l
The default path also contains an AppSync variable
(ProdServerName) which is delimited with 2 percent signs (%%).
l
The following characters are not valid in the path:< > : " / | ? *
l
The mount path could also be Same as Original Path. However,
this option is not available when the mount host is the same as
production host.
l
If you specify a non-default mount path, the drive that is specified
for mount cannot be a clustered disk.
l
The Default Path is the location to copy VDI and VSS metadata
files:%SystemDrive%\AppSyncMounts\%
%ProdServerName%%
l
The following characters are not valid in the path: < > : " / | ? *
l
If you back up the database to another media, back up the
metadata files as well.
EMC AppSync 2.2.3 User and Administration Guide
Protect SQL Server
Field
Image access
mode (during
RecoverPoint
mount)
Description
l
AppSync can integrate with third-party backup software to create
tape backups of SQL Server copies. The target directory that is
specified here must be part of the backup.
l
Logged access: Use this mount option if the integrity check entails
the scanning of large areas of the replicated volumes. Logged
access is the only option available when you mount to the
production host.
l
Virtual access with roll: Provides nearly instant access to the copy,
but also updates the replicated volume in the background. When
the replicated volumes are at the requested point in time, the RPA
transparently switches to direct replica volume access, allowing
heavy processing. With RP VMAX, and RP XtremIO, virtual access
with roll is not supported.
l
Virtual access: Provides nearly instant access to the image. Virtual
access is not intended for heavy processing. Virtual access with RP
VMAX and RP XtremIO is not supported.
Service Level
For VMAX 3 arrays only, a setting called Desired Service Level
Objective (SLO) Objective (SLO) appears in the Mount wizard and specifies the
required VMAX 3 Service Level Objectives. SLO defines the service
time operating range of a storage group.
Use Dedicated
Storage Group
l
Applicable only for physical hosts or virtual machines with direct
iSCSI as part of cluster.
l
Checked by default, enabling this option allows AppSync to
enforce a dedicated VMAX , VNX, or XtremIO storage group for a
mount. (A dedicated VMAX or VNX storage group contains the
selected mount host only.) For XtremIO, this option applies to an
XtremIO initiator group that only contains an initiator for the mount
host. The mount fails if you are mounting to a node of a cluster
that is in a storage group that is shared with the other nodes.
Note
Use this option to mount the copy to a node for copy validation or
backup to tape. In this scenario, you need two storage groups.
One storage group is dedicated to the passive node being used as
a mount host and the other storage group is for the remainder of
the nodes in the cluster. Both storage groups contain the shared
storage for the cluster.
l
If unchecked, AppSync does not enforce the use of a dedicated
storage group for a mount.
Note
Uncheck this option for manually adding the target devices as
clustered storage and presenting them to clustered SQL Server
instances for data repurposing and data mining.
Service plan summary and details
95
Protect SQL Server
In the Mount and recover copy settings, you specify the recovery instance, the type of
recovery, and the database naming details. Other settings are similar to the Mount copy
settings such as mount path and image access type.
Field
Description
Recovery
Instance
The SQL Server instance to be used for recovery. If the connection
settings are not set or are invalid for the instance, the SQL Server
Connection Settings dialog appears. Click Connection Settings to reset
the credentials.
Note
Clustered SQL Server instances are filtered out of this view.
If you are using a VMAX 3 array, a setting called Desired Service Level
Objective (SLO) is available. The option appears in the Mount wizard
and it specifies the required VMAX 3 Service Level Objectives. SLO
defines the service time operating range of a storage group
Recovery Type
Available options are: Recovery (default), No Recovery, and Standby
Database
renaming
This drop down includes:
Naming Suffix
Mount path
Copy metadata
files to
96
l
Use original database names (default if alternate instance): This is
not available for selection if the Recovery Instance is the
production instance.
l
Use original database names with suffix: This is the default if
Recovery Instance is the production instance.
Only displayed when Original database names with Suffix is selected
in the Database renaming dropdown. The default value is AppSync.
l
The default mount path, when the mount host is the same as the
production host, is %SystemDrive%\AppSyncMounts\%
%ProdServerName%%.
l
To specify the value of a Windows environment variable in the
mount path, delimit the variable name with single percent signs
(%).
l
The default path also contains an AppSync variable
(ProdServerName) which is delimited with two percent signs (%%).
l
The following characters are not valid in the path: < > : " / | ? *
l
The mount path could also be Same as Original Path. You can
select either of the options.
l
If you specify a non-default mount path, the drive specified for
mount cannot be a clustered disk.
l
By default, the location to copy VSS metadata files is the same as
the mount path.
l
If the mount path is Same as Original Path, then this defaults to
%SystemDrive%\AppSyncMounts\%%ProdServerName%%.
l
The following characters are not valid in the path: < > : " / | ? *
l
If you are backing up the database to another media, you must
backup these metadata files as well.
EMC AppSync 2.2.3 User and Administration Guide
Protect SQL Server
Field
Image access
mode (during
RecoverPoint
mount)
Use Dedicated
Storage Group
Description
l
AppSync can integrate with third-party backup software to create
tape backups of SQL Server copies. The target directory specified
here must be part of the backup.
l
Logged Access: Use this mount option if the integrity check entails
the scanning of large areas of the replicated volumes. This is the
only option available when you mount to the production host.
l
Virtual Access with Roll: Provides nearly instant access to the
copy, but also updates the replicated volume in the background.
When the replicated volumes are at the requested point in time,
the RPA transparently switches to direct replica volume access,
allowing heavy processing.
l
Virtual Access: Provides nearly instant access to the image; it is
not intended for heavy processing.
l
Applicable only for physical hosts or virtual machines with direct
iSCSI part of cluster.
l
Checked by default, enabling this option allows AppSync to
enforce a dedicated VMAX, VNX, or XtremIO storage group. For
XtremIO, this option applies to an XtremIO initiator group that only
contains an initiator for the mount host. The storage group
contains the selected mount host only for a mount and the mount
will fail if you are mounting to a node of a cluster that is in a
storage group shared with the other nodes.
Note
Use this option to mount the copy to a node for copy validation or
backup to tape. In this scenario, you will need two storage groups.
One storage group is dedicated to the passive node being used as
a mount host and the other storage group is for the remainder of
the nodes in the cluster. Both storage groups contain the shared
storage for the cluster.
l
If unchecked, AppSync does not enforce the use of a dedicated
storage group for a mount and the mount will proceed. Host
initiators can only belong in one initiator group in XtremIO, so use
this option to ensure that you mount to a mount host that is the
only host in the initiator group.
Note
Uncheck this option for manually adding the target devices as
clustered storage and presenting them to clustered SQL Server
instances for data repurposing and data mining.
Overriding mount settings in a service plan
If multiple registered SQL Servers are subscribed to the same plan, you can select
different mount and recover settings for each SQL Server, overriding the generic settings.
Before you begin
This operation requires the Data Administrator role in AppSync.
Service plan summary and details
97
Protect SQL Server
Procedure
1. Navigate to Service Plans > Microsoft SQL Server and click one of the plans from the
list.
2. From the Settings tab, select the Mount copy phase.
3. On the right pane, select the Mount Copy Overrides tab.
The list of servers include all SQL servers whose databases are subscribed to this
plan.
Based on whether Mount copy or Mount and recover copy is selected, the default
settings display for all the Servers.
4. Select the Server whose settings you want to override and click Set Overrides.
The Override Default Mount Settings dialog is displayed.
5. Select options only for those mount settings that you wish to override.
Fields that do not have a selection retain their default settings.
6. Click OK.
A pencil icon appears in the first column of the Server's row whose default mount
settings you changed.
7. To revert back to default settings for a server, click Use Default Settings.
Post-mount script
Specify a post-mount script and parameters from the Post-mount script option in the
Settings tab of a service plan.
The script runs on successful completion of the mount copy or mount with recovery
phase. This script is typically used for backup.
From the Server list, select the server on which to run the script. You can optionally run it
on a registered host other than the mount host, and enter credentials to run the script as
a specific user.
The default location of the script is %ProgramData%\EMC\AppSync\scripts\ on
the application host.
Exact parameters depend on your script. Parameters with spaces must be enclosed in
double quotes.
This phase can be enabled or disabled. This operation requires the Service Plan
Administrator role in AppSync.
Unmount copy
The final phase in the service plan unmounts the copy. This phase is disabled if the
Unmount previous copy phase is enabled. There are no user settings associated with this
phase.
If you have chosen to Mount and recover copy in the Mount copy phase, all the mounted
databases are shut down as part of this phase.
Custom shutdown script prior to unmount
Prior to unmount, if you wish to perform a customized shut down of the databases, you
can place a script at the following location: %ProgramData%\EMC\AppSync\script.
The script name must be in this format:
<ServicePlanName>_<host_ProductionInstanceName OR
ProductionInstanceName>_ ShutdownSQL.bat where:
98
EMC AppSync 2.2.3 User and Administration Guide
Protect SQL Server
l
ServicePlanName is the name of the service plan that the database is subscribed
to
l
host_ProductionInstanceName OR ProductionInstanceName:
n
In host_ProductionInstanceName , you can replace host by another
name, the ProductionInstanceName is needed irrespective of whether there
are different SQL instances or not.
n
Use ProductionInstanceName in case of default production instance which
is equal to the host name.
Note
n
It is recommended that you run the script as a Windows user. To run the script as
a SQL Server user in SQL Server 2012 environment, the Local System user must
have the sysadmin role.
n
Using the _ as a separator in the script file name is mandatory.
In the absence of a customized script, AppSync will perform a shut down of the
databases prior to unmount.
Mount considerations for SQL Server
Get mount host requirements including rules for mount and production host versions and
virtual machine mount host support.
The mount host requires the same versions of the AppSync agent plug-in, SQL Server,
and HBA drivers as the production host. Mount hosts must have an SQL Server installed if
you want to recover databases from the mounted copy. If database recovery is not
performed, then SQL Server is not required on the mount host.
Mount and production host versions
l
If you are mounting to the node of Windows failover cluster, please see the section
"Microsoft cluster server mounts" in the EMC AppSync Installation and Configuration
Guide.
l
If the major version of the SQL Server instance on the production mount host is later
than that of the mount host, recovery will fail for all databases belonging to that
instance.
l
If the major version of the SQL Server instance on the production mount host is earlier
than that of the mount host, recovery will succeed only if the recovery type is either
RECOVERY or NORECOVERY. Recovery will fail if recovery type is STANDBY.
l
If the major version of the SQL Server instance on the production mount host is same
as that of the mount host, but the minor version is earlier, recovery will fail for all
databases belonging to that instance.
l
If the major version of the SQL Server instance on the production mount host is same
as that of the mount host, but the minor version is later, recovery will succeed only if
the recovery type is either RECOVERY or NORECOVERY. Recovery will fail if recovery
type is STANDBY.
Virtual disk support
If the mount host is a virtual machine, the Virtual Center must be registered with
AppSync. This is needed to mount RDMs.
For virtual disks:
Mount considerations for SQL Server
99
Protect SQL Server
l
Production mount is not supported if the ESX host version is prior to 5.0.
l
Non-persistent virtual disks are not supported.
l
For datastore and virtual disk mounts on ESXi 5.x and RecoverPoint 4.1.7.7
environments, disable hardware acceleration to ensure successful virtual access type
mounts. For more details, refer VMware Knowledge Base article 2006858.
For Hyper-V SCSI pass-through, the mount host cannot be a Hyper-V host it has to be a
physical host or VM with NPIV or iSCSI direct attached.
Mount SQL Server database copy on-demand
You can initiate an on-demand mount of a database copy from a copy or a database.
Before you begin
This operation requires the Data Administrator role in AppSync.
Procedure
1. In the Databases page, select Recover > Mount a Copy.
From the Copies page, select a copy and click Mount.
The SQL Server Mount wizard launches.
2. Use the Database, Copies or Service Plan filters to select the appropriate copy to
mount.
The copies list is refreshed based on the filters selected.
3. Select the copy to mount.
For a RecoverPoint copy, you also have the option to select a bookmark based on a
specific time. However, there should be a copy available in AppSync prior to the time
you select Mount.
Click Select a point in time to select a copy with a specific time stamp. The time
shown here is the console's time. If the console is in a different time zone from the
RecoverPoint Appliance (RPA), specify the time as per the server's time zone to mount
the copy.
For VMAX 3 arrays, you are presented with an SLO drop-down menu. You can select
the desired Service Level Objective (SLO) for the mount copy. If there is a storage
group for the mount host with the desired SLO. AppSync will add the LUN to the
storage group. If this storage group does not exist, AppSync adds the LUN to any
storage group that is masked to the host. If a storage group is configured to pick
target devices, AppSync removes the devices from the storage group at the time of
mount and adds them to the storage group for the mount host. The devices will be
added to the original storage group when the copy is expired.
4. In the Mount Additional Copies page, select one or more additional copies to mount.
The copies listed here are of other databases that were protected at the same time
and on the same SQL Server as the copy you selected in the previous step.
5. On the Select Mount Options page, select Mount copy or Mount and recover copy.
6. In the Summary page, review the choices you made in the previous pages and click
Finish to mount the copy.
7. In the Results page, select View Details to see progress of the different phases that
are part of mounting a copy.
The last phase completed is displayed at the bottom of the list.
100
EMC AppSync 2.2.3 User and Administration Guide
Protect SQL Server
SQL Server Mount Copy options
Review SQL server mount copy fields and descriptions.
Field
Description
Mount on
Server
The server on which to mount the copy. Only the nodes of the cluster
or standalone hosts are available for selection. SQL virtual machines
are filtered out.
Mount with
access
Type of access the copy should be mounted with.
Mount path
Copy metadata
files to
Image access
mode (during
RecoverPoint
mount)
l
The Default Mount Path is %SystemDrive%\AppSyncMounts\
%%ProdServerName%%.
l
To specify the value of a Windows environment variable in the
mount path, delimit the variable name with single percent signs
(%).
l
The default path also contains an AppSync variable
(ProdServerName) which is delimited with 2 percent signs (%%).
l
The following characters are not valid in the path:< > : " / | ? *
l
The mount path could also be Same as Original Path. However,
this option is not available when the mount host is the same as
production host.
l
If you specify a non-default mount path, the drive that is specified
for mount cannot be a clustered disk.
l
The Default Path is the location to copy VDI and VSS metadata
files:%SystemDrive%\AppSyncMounts\%
%ProdServerName%%
l
The following characters are not valid in the path: < > : " / | ? *
l
If you back up the database to another media, back up the
metadata files as well.
l
AppSync can integrate with third-party backup software to create
tape backups of SQL Server copies. The target directory that is
specified here must be part of the backup.
l
Logged access: Use this mount option if the integrity check entails
the scanning of large areas of the replicated volumes. Logged
access is the only option available when you mount to the
production host.
l
Virtual access with roll: Provides nearly instant access to the copy,
but also updates the replicated volume in the background. When
the replicated volumes are at the requested point in time, the RPA
transparently switches to direct replica volume access, allowing
heavy processing. With RP VMAX, and RP XtremIO, virtual access
with roll is not supported.
l
Virtual access: Provides nearly instant access to the image. Virtual
access is not intended for heavy processing. Virtual access with RP
VMAX and RP XtremIO is not supported.
Mount SQL Server database copy on-demand
101
Protect SQL Server
Field
Description
Service Level
For VMAX 3 arrays only, a setting called Desired Service Level
Objective (SLO) Objective (SLO) appears in the Mount wizard and specifies the
required VMAX 3 Service Level Objectives. SLO defines the service
time operating range of a storage group.
Use Dedicated
Storage Group
l
Applicable only for physical hosts or virtual machines with direct
iSCSI as part of cluster.
l
Checked by default, enabling this option allows AppSync to
enforce a dedicated VMAX , VNX, or XtremIO storage group for a
mount. (A dedicated VMAX or VNX storage group contains the
selected mount host only.) For XtremIO, this option applies to an
XtremIO initiator group that only contains an initiator for the mount
host. The mount fails if you are mounting to a node of a cluster
that is in a storage group that is shared with the other nodes.
Note
Use this option to mount the copy to a node for copy validation or
backup to tape. In this scenario, you need two storage groups.
One storage group is dedicated to the passive node being used as
a mount host and the other storage group is for the remainder of
the nodes in the cluster. Both storage groups contain the shared
storage for the cluster.
l
If unchecked, AppSync does not enforce the use of a dedicated
storage group for a mount.
Note
Uncheck this option for manually adding the target devices as
clustered storage and presenting them to clustered SQL Server
instances for data repurposing and data mining.
SQL Server Mount and Recover copy options
SQL Server mount and recover copy options are explained in the following table:
Field
Description
Recovery
Instance
The SQL Server instance to be used for recovery. If the connection
settings are not set or are invalid for the instance, the SQL Server
Connection Settings dialog appears. Click Connection Settings to reset
the credentials.
Note
Clustered SQL Server instances are filtered out of this view.
If you are using a VMAX 3 array, a setting called Desired Service Level
Objective (SLO) is available. The option appears in the Mount wizard
and it specifies the required VMAX 3 Service Level Objectives. SLO
defines the service time operating range of a storage group
Recovery Type
102
Available options are: Recovery (default), No Recovery, and Standby
EMC AppSync 2.2.3 User and Administration Guide
Protect SQL Server
Field
Description
Database
renaming
This drop down includes:
Naming Suffix
Mount path
Copy metadata
files to
Image access
mode (during
RecoverPoint
mount)
Use Dedicated
Storage Group
l
Use original database names (default if alternate instance): This is
not available for selection if the Recovery Instance is the
production instance.
l
Use original database names with suffix: This is the default if
Recovery Instance is the production instance.
Only displayed when Original database names with Suffix is selected
in the Database renaming dropdown. The default value is AppSync.
l
The default mount path, when the mount host is the same as the
production host, is %SystemDrive%\AppSyncMounts\%
%ProdServerName%%.
l
To specify the value of a Windows environment variable in the
mount path, delimit the variable name with single percent signs
(%).
l
The default path also contains an AppSync variable
(ProdServerName) which is delimited with two percent signs (%%).
l
The following characters are not valid in the path: < > : " / | ? *
l
The mount path could also be Same as Original Path. You can
select either of the options.
l
If you specify a non-default mount path, the drive specified for
mount cannot be a clustered disk.
l
By default, the location to copy VSS metadata files is the same as
the mount path.
l
If the mount path is Same as Original Path, then this defaults to
%SystemDrive%\AppSyncMounts\%%ProdServerName%%.
l
The following characters are not valid in the path: < > : " / | ? *
l
If you are backing up the database to another media, you must
backup these metadata files as well.
l
AppSync can integrate with third-party backup software to create
tape backups of SQL Server copies. The target directory specified
here must be part of the backup.
l
Logged Access: Use this mount option if the integrity check entails
the scanning of large areas of the replicated volumes. This is the
only option available when you mount to the production host.
l
Virtual Access with Roll: Provides nearly instant access to the
copy, but also updates the replicated volume in the background.
When the replicated volumes are at the requested point in time,
the RPA transparently switches to direct replica volume access,
allowing heavy processing.
l
Virtual Access: Provides nearly instant access to the image; it is
not intended for heavy processing.
l
Applicable only for physical hosts or virtual machines with direct
iSCSI part of cluster.
Mount SQL Server database copy on-demand
103
Protect SQL Server
Field
Description
l
Checked by default, enabling this option allows AppSync to
enforce a dedicated VMAX, VNX, or XtremIO storage group. For
XtremIO, this option applies to an XtremIO initiator group that only
contains an initiator for the mount host. The storage group
contains the selected mount host only for a mount and the mount
will fail if you are mounting to a node of a cluster that is in a
storage group shared with the other nodes.
Note
Use this option to mount the copy to a node for copy validation or
backup to tape. In this scenario, you will need two storage groups.
One storage group is dedicated to the passive node being used as
a mount host and the other storage group is for the remainder of
the nodes in the cluster. Both storage groups contain the shared
storage for the cluster.
l
If unchecked, AppSync does not enforce the use of a dedicated
storage group for a mount and the mount will proceed. Host
initiators can only belong in one initiator group in XtremIO, so use
this option to ensure that you mount to a mount host that is the
only host in the initiator group.
Note
Uncheck this option for manually adding the target devices as
clustered storage and presenting them to clustered SQL Server
instances for data repurposing and data mining.
Supported mount recovery modes
The following mount recovery types are available when you are recovering a SQL
database copy.
Recovery
Type
Description
Recovery
Instructs the restore operation to roll back any uncommitted transactions.
After the recovery process, the database is ready for use.
No Recovery Instructs the restore operation not to roll back any uncommitted
transactions. When in No Recovery mode, the database is unusable. This
option is useful when the Database Administrator needs to restore one or
more transaction log backups. Database is attached to the instance
selected for recovery and is left in the "Restoring" state.
Standby
104
Restores files and opens the database in read-only mode. Subsequently,
the Database Administrator can manually apply additional transaction log
backups.
EMC AppSync 2.2.3 User and Administration Guide
Protect SQL Server
Recovery
Type
Description
Note
If you are restoring a database from an older version of SQL Server onto a
newer SQL Server version, do not use standby mode. If you use standby,
the upgrade to the newer version cannot happen and that will result in a
failure of the operation.
Unmounting an SQL Server copy
When you select an SQL Server copy to unmount, other copies that were mounted along
with the selected copy will also be unmounted.
Before you begin
This operation requires the Data Administrator role in AppSync.
You can unmount a copy only from a list of copies made for a database.
Procedure
1. Navigate to the Copies page from the Copy Management or Service Plan pages:
l
Copy Management > Microsoft SQL Server > select the server which hosts the
filesystem you want to unmount, then select the database instance with the copy
to unmount.
l
Service Plans > Microsoft SQL Server> select a service plan, then select the Copies
tab.
2. From the list of copies, select the copy and click Unmount from the button in the lower
part of the page.
The Unmount Confirmation dialog displays all the copies of other databases that were
mounted along with the selected copy to be unmounted.
3. Click Yes to confirm the unmount of all the copies shown in the dialog.
The Unmount page displays the progress of the unmount operation. All copies
associated with the selected copy will be unmounted.
SQL Server database restore overview
Review and consider the following sections regarding SQL Server database restore
options.
These include:
l
Restore considerations for databases in an Availability Group
l
Affected entities during restore
l
Restoring a primary database or a secondary database with failover
l
Restoring a secondary database without failover
l
How AppSync manages damaged SQL databases
l
Restoring an SQL Server copy
l
Restoring an SQL Server copy on XtremIO
Unmounting an SQL Server copy
105
Protect SQL Server
l
SQL restore utility (assqlrestore)
Restore considerations for databases in an Availability Group
AppSync restores copies of primary and secondary databases. Consider the following
when restoring a database in an Availability Group.
l
Restore is at the LUN level and must be restored back to the source LUN that was
used to create the AppSync copy.
l
AppSync suspends data movement as part of the restore process.
l
A database cannot be restored if it is part of an Availability Group. AppSync removes
the database from the Availability Group as part of the restore process.
l
AppSync does not put the database back in the Availability Group. For more
information on restoring databases in an Availability Group, see "Restoring a primary
database or a secondary database with failover" and "Restoring a secondary
database without failover".
Affected entities during restore
When restoring from a copy, you may be prompted to restore items in addition to the
ones you selected.
An affected entity is data that resides on your production host that unintentionally
becomes part of a replica because of its proximity to the data you intend to protect. You
can prevent affected entity situations by properly planning your data layout based on
replica granularity. The granularity of a replica depends upon the environment.
If there are affected entities in your underlying storage configuration, the Restore Wizard
notifies you of these items. The following scenarios produce affected entities that require
you to acknowledge that additional items will be restored:
l
For RecoverPoint, if the databases are in the same consistency group they become
affected entities when the other database is protected.
l
For VNXe, if the databases are in the same LUN group they become affected entities
when the other database is protected.
l
For VNX/VMAX, VNX, VNXe, or XtremIO, if the databases are on the same LUN they
become affected entities when the other database is protected.
l
For VMware virtual disks, since restore involves a datastore, restore of all
applications residing on the same datastore (virtual disks on the same datastore) are
also affected entities.
If the affected entity was protected along with the database that is selected for restore, it
will be restored by AppSync. Any other database that was not protected but is an affected
entity will be overwritten.
AppSync calculates affected entities for the consistency groups or LUN groups of the
database that is selected for restore. If the affected databases in turn partially reside on
other consistency groups or LUNs groups, AppSync does not calculate affected entities
on those consistency groups or LUN groups.
Depending upon the type of affected entity, the affected databases are detached by
AppSync or you must manually detach them from the SQL Server instance.
Affected entities are calculated only for the SQL Server instances where the credentials
are configured. AppSync does a fresh database discovery for all these instances before
calculating the affected entities.
106
EMC AppSync 2.2.3 User and Administration Guide
Protect SQL Server
Restoring a primary database or a secondary database with failover
Once you click the Finish button in the SQL Server Restore wizard, AppSync performs the
following actions:
1. If you had selected the Failover the Availability Group if the current role is Secondary
checkbox, AppSync verifies the health of the databases in the Availability Group that
are not being restored. If they are not healthy, AppSync cannot perform the failover
and the restore operation fails. You must retry the restore operation without selecting
the checkbox.
2. If you had chosen to backup the transaction log, AppSync backs up the transaction
log.
3. AppSync suspends data movement for all replicas of the selected database before
removing all replicas of the selected database from the Availability Group.
4. If the database being restored is secondary, AppSync initiates the failover.
5. AppSync restores the LUNs of the selected database.
6. Finally, AppSync recovers the database and leaves it in the Recovery state that you
selected in the SQL Server Restore wizard.
After AppSync completes the restore, you must perform the following steps.
Procedure
1. Restore any log backups and recover the primary database.
2. Add the database back into the Availability Group.
3. If the primary database was rolled forward so it is at the same time as the secondary
database, re-join the secondary copies to the Availability Group.
4. If the primary database was not rolled forward:
a. Delete any secondary copies of the restored database.
b. Reseed and re-join the secondary database replicas to the availability group.
Note
After AppSync removes the primary database copy, the copy is in the recovered state
if it is healthy. If you restored a secondary copy with failover, the primary role will have
moved to another SQL Server instance. You must delete the original primary database
and reseed it.
Restoring a secondary database without failover
Once you click the Finish button in the SQL Server Restore wizard, AppSync performs the
following actions:
1. If you had chosen to backup the transaction log, AppSync backs up the transaction
log.
2. AppSync suspends data movement for the selected secondary database replica.
Replication continues to work for other replicas of the database.
3. AppSync removes the selected secondary database replica from the Availability
Group.
4. AppSync restores the LUNs of the selected database.
Restoring a primary database or a secondary database with failover
107
Protect SQL Server
5. Finally, AppSync recovers the database and leaves it in the Recovery state that you
selected in the SQL Server Restore wizard.
After AppSync completes the restore, you must perform the following steps.
Procedure
1. Restore any log backups and leave the secondary database in a "NO RECOVERY" state.
2. Join the secondary database back into the Availability Group.
How AppSync manages damaged SQL databases
Damaged databases may have data files missing or damaged with their log files intact.
AppSync can take tail log backups for damaged databases. A damaged database must
not contain bulk-logged changes and it must not be in OFFLINE state.
If the production database is damaged and you select the Database is damaged
checkbox during restore, AppSync backs up the tail log of the damaged database before
proceeding with restore. If the damaged database is in RECOVERY_PENDING or SUSPECT
state, AppSync first tries to detach the database by setting the EMERGENCY mode on it. If
AppSync fails to set EMERGENCY mode on the database, it drops the database and then
proceeds with the restore. Once the restore is successful, you can recover the database
manually using the tail log backup.
Refer also to Restore procedures with XtremIO on page 196.
Restoring a SQL Server copy
You can perform a restore of an SQL Server copy from the Server's Copies page, service
plan's Copies page or from the Databases page.
Before you begin
This operation requires the Data Administrator role in AppSync.
Procedure
1. In the Databases page, select Recover > Restore .
In the Copies page, select a copy and click Restore.
The SQL Server Restore wizard launches.
2. Select the copy to restore.
Use the Time or Service Plan filters to select the appropriate copy to restore. The
copies list is refreshed based on the filters selected.
3. Click Next.
If the selected copy has affected entities, the Restore Warnings page is displayed.
4. Read the warning messages for the affected databases. Select the checkbox to
indicate your agreement to restore other entities along with the selected copy.
5. In the Backup Transaction Logs step, select Yes to backup logs prior to restore.
a. In the Back up to box, enter the location where the logs will be backed up. The files
will bear the name of the database.
b. Select the Add a file extension to the backup file name checkbox and specify an
extension for the backed up files. The default extension is .trn.
c. Select the Add a prefix to the backup file names checkbox and specify a prefix for
the backed up files. The default prefix is AppSync.
108
EMC AppSync 2.2.3 User and Administration Guide
Protect SQL Server
d. Select the Database is damaged checkbox to backup tail log files.
e. Select the Truncate the transaction logs checkbox as required.
This checkbox is not available for selection if you selected the Database is
damaged checkbox.
f. Select the Overwrite existing backup files checkbox as required.
This checkbox is not available for selection if you selected the Database is
damaged checkbox.
6. Click Next for the Restore Options.
The Restore Options page is displayed.
7. Select the appropriate recovery or restore options.
a. To recover the database, choose from one of these options:
l
Leave database ready to use: This option is not available if you have chosen to
backup transaction logs.
l
Leave databases non-operational
Note
Select this option when restoring a secondary database without failover. This
leaves the secondary database in the restoring state so transaction logs can be
restored allowing the database to rejoin the Availability Group.
l
Leave databases in read-only mode: If you select this option, specify the
location in the
Standby file location box where the standby files must be stored. The default
path is C:\temp.
b. To force a restore by overwriting the existing database, select the Overwrite the
existing databases checkbox.
This option is not required in the normal circumstances.
This option is not available if you have chosen to backup transaction logs.
c. In case of Availability Group, select the Failover the Availability Group if the
current role is Secondary checkbox to initiate a failover before restore if you are
restoring to a secondary database.
A warning message is displayed that the database will be removed from the
Availability Group and that you must rejoin it after AppSync restores the volume.
d. Select the I have read and understand the warning above, and want to continue
with the restore checkbox to acknowledge the message.
8. Click Next.
The Summary page is displayed.
9. Review the Summary page and click Finish to perform the restore.
10. In the Results page, click View Details to see progress of the different phases that are
part of restoring a copy.
The last phase completed is displayed at the bottom of the list.
11. After AppSync restores a database in an Availability Group, perform additional steps
as needed:
Restoring a SQL Server copy
109
Protect SQL Server
l
Restore a primary database or a secondary database with failover on page 107.
l
Restore a secondary database without failover on page 107.
l
Manually restore an SQL Server copy on XtremIO on page 110
Restoring an SQL Server copy on XtremIO
Version 2.2.2 and higher of AppSync supports automated restore of copies on XtremIO
4.0 and higher. This topic shows you how to use AppSync versions 2.1 - 2.2.1 and an
XtremIO version lower than 4.0 to restore an SQL Server database to a point in time with a
semi-manual restore process. Restore of SQL Server transaction log backups are not
supported.
Before you begin
For automated restore, refer to Restoring an SQL Server copy on page 108.
For semi-manual restore ensure the following pre-requisites, then perform the procedure:
l
SQL Server databases and XtremIO storage running on the AppSync configuration.
l
An AppSync service plan that replicates the database.
Procedure
1. On the AppSync console, select Service Plans > Microsoft SQL Server , and then click
the wanted service plan, for example, Bronze.
The Copies page displays existing copies.
2. Mount the copies that you want to restore. When prompted to Mount Additional
Copies, select all database copies that reside on the disk drives you are replacing.
The Select Mount Option dialog box opens.
Figure 6 Select Mount Option
3. In the Select Mount Option dialog box, select Mount copy.
4. Since AppSync does not allow you to recover to the production instance, select
Original Server from the Mount on Server list box.
5. Select Read-write from the Mount with access list box.
110
EMC AppSync 2.2.3 User and Administration Guide
Protect SQL Server
6. Select Default Path from the Mount Path list box (because you mount to the original
server).
7. On the production host, stop any application that uses the databases that you are
restoring.
8. Using the SQL Server Management Studio, take the databases offline, and then
detach them.
Remember that you replace the production disk drives with the AppSync snapshots,
so detach any databases that are using the same devices.
9. After you detach the affected databases, use Windows Server Manager to delete the
drive letter or mount point for the disks that you are restoring. Record the disk number
and drive letter assignments in case you need to undo the restore steps.
After the AppSync mount completes, the drives for the restore copies are located as
mount points in C:\AppSyncMounts\<SERVERNAME>.
10. Start Windows Server Manager to find the mounted devices, and then assign drive
letters or mount points to replace the ones that you removed in Step 8.
11. Use SQL Server Management Studio to attach the databases that you want to restore.
Point to the mounted AppSync copies for the databases. If the database has 17 or
more files, use CREATE DATABASE database_name FOR ATTACH.
Figure 7 Attach Databases
At this point, all the restored databases should be online, and you then can restart
any database application that you previously stopped. You can perform the following
steps without affecting SQL Server.
12. On the AppSync console, return to the Database Copies page that is described in Step
1, and then click Remove for each mounted copy.
This action removes the copy from the AppSync database, without removing the
snapshots on the production mount host, or on the XtremIO storage array. You receive
a warning message.
Note
If the service plan contains multiple databases, the storage for all databases in the
service plan remains on the XtremIO storage array. You can remove them at your
discretion.
Restoring an SQL Server copy on XtremIO
111
Protect SQL Server
Figure 8 Confirm the following removal selections
13. Move the mounted snapshots from the /AppSyncSnapshots folder to the folder
where the production LUN that is being replaced resides. Be sure to move the
snapshots that are mounted, that is, the snapshots that are currently masked to the
ESX initiator group (if production host is virtual machine). If the production host is a
physical host, move the snapshots to the folder where the production LUNs are
located.
The production LUN and the snapshot appear in the same folder.
Figure 9 LUN Mapping Configuration
After the move, notice that the restored devices are a snapshot and the production
devices are volumes (LUN).
Figure 10 Snapshot and LUN example
14. Perform this step only if the production host is a virtual
machine.
112
EMC AppSync 2.2.3 User and Administration Guide
Protect SQL Server
a. Unmap the original production volumes from the production host.
b. Rescan the ESX host.
c. Remove the RDMs from the production virtual machine configuration.
15. Perform this step only If the production host is a physical
machine.
a. Unmap the original production volumes from the production host.
b. Perform a device rescan on the host to remove the old production devices.
16. Remove the production volumes.
This action causes the restored snapshots to become volumes.
Figure 11 Restored snapshots become volumes
17. Rename the snapshots to the production volume names.
Figure 12 Rename snapshots to production volume names
In the AppSync XtremIO console, notice that the other snapshots in the /
AppSyncSnapshots folder for the production XtremIO volumes that you removed in
the previous step have also become volumes.
Figure 13 Snapshots become LUNs
18. Run the database service plan to create a snapshot of the restored databases.
Restoring an SQL Server copy on XtremIO
113
Protect SQL Server
SQL Server restore utility (assqlrestore)
AppSync includes a SQL Server restore utility called assqlrestore.This section
describes it's function and uses.
The assqlrestore utility lets you restore individual SQL Server databases from a tape
backup or mounted copy without reverse-syncing the target device over the source
device. It can restore a database, filegroup, or file. The utility can restore to the original
database or to a new database. SQL Server VDI metadata that was created as part of the
replication activity is required to restore a database using assqlrestore.
assqlrestore is a command line interface that you run from a command prompt
window on the AppSync client. It is installed on the client as part of the AppSync
installation.
Restoring an individual database from a mounted copy is especially useful when you
need to recover only one database and do not want to overwrite an entire device which
occurs with a normal AppSync restore.
Assqlrestore command syntax with examples
This topics lists the command syntax for the assqlrestore commend followed by
examples of the commands.
Command syntax
The following table lists the command syntax for the assqlrestore command.
Table 11 assqlrestore Command Syntax
Option
Description
Required
-s
SQL Server name including instance name (host\instance).
-f
Metadata filename and location.
-d
Database name.
Connection Types (-E
or -U)
-E
User used for Windows Authentication (specify username)
-U
SQL Server login ID.
-P
Clear text password (used with -E and -U options).
-p
Encrypted password (used with -E and -U options).
Optional
114
-r
Recover option - RECOVERY, NORECOVERY (default), or STANDBY.
-u
Undo filename, required for STANDBY
-m
Move file. Option has two parameters: logical_file_name and
operating_system_file_name. Pathnames must exist. Repeat
option for each file, including log files or full text catalog files. If you are
restoring to a new database name, use the -m option so you do not
EMC AppSync 2.2.3 User and Administration Guide
Protect SQL Server
Table 11 assqlrestore Command Syntax (continued)
Option
Description
overwrite the original files. For example: -m logicalfilename S:
\existingdir\newfilename.mdf
-fg
Filegroup to restore. Repeat option for each filegroup.
-lf
Logical file to restore. Repeat option for each logical file.
-e
Displays encrypted password when unencrypted password is specified
as an argument. Not used with other parameters.
-v
Verbose mode.
-q
Quiet mode. Will not ask questions.
-l <log_dir>
Creates log files in the specified directory.
-h
Help.
Example 2 Command syntax examples
Command options are case-sensitive. Refer to the "SQL Server books online" for a
description of the T-SQL
Note
Command parameters have changed from the Replication Manager utility
(rmsqlrestore)
l
Using Windows authentication, restore without applying logs.
assqlrestore exe -E Administrator -P password -s sql1\instance1 d custinfo
-f "C:\AppSyncMounts\sql1\APPSYNC_VDI_INSTANCE1_ custinfo.bin" -r
RECOVERY
l
Restore to a new database name and move files using a SQL login and encrypted
password:
assqlrestore -s sql1\instance1
-d custinfoTest
-f "C:\AppSyncMounts\sql1\APPSYNC_VDI_INSTANCE1_ custinfoTest.bin"
-r RECOVERY
-m custinfo_Data S:\custinfoTest.mdf
-m custinfo_Log T:\custinfoTest.ldf
-U sa -p 1EMC_4roJdyU5;x
n
To get the encrypted password:
assqlrestore -e <unencrypted_password>
SQL Server restore utility (assqlrestore)
115
Protect SQL Server
Restoring an SQL Server database with assqlrestore
The basic steps to restore a database are provided here. You may need additional steps
but use these as a framework.
Before you begin
Log in to the SQL Server system as a user with Administrator rights, then back up the SQL
Server transaction log.
Procedure
1. Take the target SQL Server database offline.
2. Restore the database files (.ldf, .ndf, and .mdf) from tape, or copy them from a
mounted replica. You can copy them over the original files or to a new location.
3. Open a command prompt window and cd to: C:\Program Files\EMC\AppSync
\Host Plug-in\
4. Run the assqlrestore command.
Refer to the Assqlrestore command syntax with examples section for sample
commands. The basic command syntax is:
assqlrestore -s <SQLservername> -d <databasename> -f
<metadatafile> -r <recovery_type>
5. If required, apply transaction logs and recover the database.
Restoring a file or filegroup with the SQL Server restore utility
Learn how to restore a file or filegroup with the SQL Server assqlrestore utility.
Before you begin
Be sure you understand how restore of files and filegroups behave in SQL Server before
proceeding.
Note
You cannot use the assqlrestore utility to restore a SQL Server filegroup if the
filegroup name contains non-ASCII characters.
Log in to the SQL Server system as a user with Administrator rights, then back up the SQL
Server transaction log. For file or filegoroup restore, the database must be online.
Procedure
1. Open a command prompt window and cd to: C:\Program Files\EMC\AppSync
\Host Plug-in\
2. Run the assqlrestore command.
a. When assqlrestore displays the restore command that it is about to run, verify
with Y if it is correct.
b. When assqlrestore prompts, restore the files you are
recovering, enter Y to continue.
116
EMC AppSync 2.2.3 User and Administration Guide
Protect SQL Server
To restore two files, for example, run:
assqlrestore -s <SQLservername> -d <databasename>
-f <metadatafile> -lf <logical_filename1>
-lf <logical_filename2> -r norecovery
To restore two filegroups, run:
assqlrestore
-s <SQLservername>
-d <databasename>
-f <metadatafile>
-lf <logical_filename1>
-fg <logical_filegroupname1>
-fg <logical_filegroupname2>
-r norecovery
Do not use the quiet mode for a file or filegroup restore. You can use -lf and -fg in the
same restore command.
Repurposing SQL Server database copies
This topic explains how to use the AppSync repurposing feature for database and
Bookmark copies. ,
Review and consider the following information regarding repurposing features:
l
Repurposing creates a multi-level tree of copies of the database.
l
AppSync identifies copies that are created from a repurpose action by a "generation
removed" number from the production source data, for example, Gen 1, Gen 2, and
so on.
l
There is no practical limit to the number of generations, but support is limited to two
generations that are removed from the production data.
l
A first generation(1st Gen) copy creates a copy that can be used as source for other
copies. This action creates a copy-of-copy.
l
Repurpose copies are meant to be mounted for extended periods of time for various
purposes.
l
After use, repurpose copies are either discarded or refreshed.
l
Repurpose copies do not figure into RPO calculations.
l
You can run 1st Gen copies now or schedule them to run at a future time.
The console displays all scheduled repurpose actions. Select any of them, and then
select the Delete button. The scheduled activity is removed.
If you enable hot-backup, the 1st Gen copy of the database creates the application
consistent copy. It Includes application discovery, mapping, and database freeze/thaw.
AppSync does not set database freeze/thaw as the default for 1st Gen repurpose copies.
Second generation (2nd Gen) copies are created as copies-of-copies, using the 1st Gen
copy as the source. They do not include application discovery, mapping, and database
freeze/thaw. As a result, 2nd Gen copies are created much faster than 1st Gen copies.
Note
Repurposing copies are not supported for Oracle databases on VNX File storage, or NFS
file systems.
Repurposing SQL Server database copies
117
Protect SQL Server
Repurpose refresh
Refresh means to discard the current copy (expire), and recreate the copy contents using
its parent.
118
l
1st Gen and 2nd Gen copies can be refreshed.
l
Refreshing a 1st Gen copy creates an application consistent copy with a new time.
l
2nd Gen copies are not modified if you refresh 1st Gen copy.
l
Refresh of 2nd Gen copy resynchronizes that 2nd Gen copy with the 1st Gen parent.
(Use for throwing away changes of 2nd Gen copy and starting over.)
l
2nd Gen timestamp is the same as the 1st Gen copy.
l
With refresh, it is possible to have first and 2nd Gen copies with different copy time
stamps.
l
If you specified Unmount Previous Copy when creating the 2nd Gen copy, AppSync
unmounts the copy before creating the new copy.
l
Refresh of a 1st Gen copy on XtremIO : If the number of LUNs in the source database
changes, then create only new snapshots. For example, AppSync uses the refresh
provided by the XtremIO array from source LUNs to create the 1st Gen snapshot copy.
l
Refresh of 2nd Gen copy on XtremIO: If you refresh the source 1st Gen copy and the
number of LUNs in the source database changed, then create only new snapshots.
EMC AppSync 2.2.3 User and Administration Guide
Protect SQL Server
Otherwise, use the refresh provided by the XtremIO array from 1st Gen snapshots to
2nd Gen snapshot copies.
l
After the refresh, AppSync remounts the copy if the copy was mounted.
l
2nd Gen copy is recreated from 1st Gen parent. If 1st Gen parent no longer exists,
then the refresh fails.
l
1st Gen copy is recreated from the application.
l
The current copy that you want to refresh must be unmounted (if it is mounted). If you
selected Mount Copy in the Repurpose wizard, AppSync mounts the copy again.
You can refresh a repurposed copy at any time. To start the refresh:
1. From the Applications tab of the AppSync console, select the repurposed copy that
you want to refresh.
2. Click Refresh.
SQL Server repurpose caveats
AppSync supports SQL Server repurposing in standalone environments and failover
cluster environments with or without AlwaysOn Availability Groups.
Repurposing SQL Server copies supports multiple database selection for repurposing
copies. AppSync does not support multiple array repurposing or dynamic database
subscription. The 1st Gen copy can be taken only for a single database or a set of
databases belonging to the same host within the same array. The refresh works only for
those databases that were previously subscribed. AppSync does not subscribe the whole
user database folder to the SQL repurposing service plan.
The system displays repurposed copies of an SQL Server database hierarchically. The
system displays the 1st Gen (parent) copy , and then it displays 2nd Gen copies as
children of the 1st Gen copy.
Note
If you are using XtremIO 3.x as the 1st Gen copy source, and then upgrade to XtremIO 4.x,
and try to create a 2nd Gen copy, refresh the 1st Gen copy or the operation fails.
Using the Repurpose wizard
Use the Repurpose wizard to schedule or immediately create 1st Gen or 2nd Gen copies
as needed.
Before you begin
You need AppSync administrative privileges to Repurpose the database instance.
To display the list of available applications.
Procedure
1. Log in to the AppSync console and go to Copy Management.
Option
Description
To Repurpose an SQL copy:
Select SQL Server
To Repurpose an Oracle copy:
Select Oracle
A list of available databases for the application choice loads.
2. Click the database instance you want to Repurpose, and then select Repurpose from
the drop-down list in the lower left of the console screen.
Using the Repurpose wizard
119
Protect SQL Server
This action launches the Repurpose wizard, and leads you to the Intentions page
where you can tell AppSync which action you want to perform:
l
Create First Generation copy
l
Create First Generation copy and a Second Generation copy
3. Select the desired copy type, and select Local or Remote in the Site drop-down to
continue creating the copy.
4. Click Next to launch the Settings screen.
From the Settings screen, you can define the specific options for 1st and 2nd gen
copies. Specifically you can
l
Define labels for each copy to help identify the copy purpose.
l
Select application-specific copy options for 1st gen copy only.
l
Choose appropriate copy type (wizard fails if incorrect type is chosen).
5. Select the desired options for the copy, and then click Next.
The schedule page of the wizard appears allowing you to identify when the 1st
generation copy should be created. Create the copy now or schedule the copy for a
convenient time.
6. Select Run now or Run later followed by typing run- time date and time, and then
press Next to complete the wizard.
After you finish
The Repurpose Monitor
The Repurpose Monitor allows you to view all currently running repurpose activities, and
monitor their progress. The Repurpose Monitor shows the item being repurposed (source)
and the label of the item being created or refreshed along with the application type. Refer
to The Repurpose Monitor on page 209.
View or cancel scheduled repurpose copies
You can view or cancel any scheduled first generation repurpose copy.
Procedure
1. From the Appsync console, navigate to Application > Copy Management, and then
select the appropriate database.
2. Select Repurpose > View Scheduled Repurpose Copy to view all scheduled repurpose
actions.
3. To delete an action, select one or more desired actions, and click Delete to remove the
action.
120
EMC AppSync 2.2.3 User and Administration Guide
CHAPTER 7
Protect Oracle
This chapter includes the following topics:
l
l
l
l
l
l
Overview of Oracle support................................................................................. 122
Protecting a database......................................................................................... 131
Service plan summary and details....................................................................... 134
Mount an Oracle copy......................................................................................... 142
Restoring an Oracle copy.....................................................................................148
Repurposing overview......................................................................................... 156
Protect Oracle
121
Protect Oracle
Overview of Oracle support
Use AppSync to create and manage application consistent (using hot backup mode) and
crash consistent (without hot backup mode) copies of Oracle® databases. The copies
can be used for mount (with/without recovery) and restore.
The AppSync Support Matrix on https://elabnavigator.emc.com/eln/extendedSupport is
the authoritative source of information on supported software and platforms..
AppSync supports:
l
Oracle - (Standalone and Oracle Real Application Cluster) and on Linux and AIX.
l
Oracle installations on physical hosts as well as virtual machines (with pRDMs and
Vdisks) - There is no support for RDMs in virtual mode.
l
Oracle databases residing on NFS file systems with VNX storage.
l
Oracle databases residing on ASM and file systems.
l
RMAN cataloging of databases to a remote catalog.
l
Repurposing of Oracle database copies. Refer to the repurposing overview topic
earlier in this document for details.
Oracle permissions
These permissions are required for AppSync to work with Oracle.
l
Root or sudo access to Oracle production server and mount server.
l
When connecting to Oracle databases, AppSync uses a bequeath connection and
always connects as SYSDBA.
l
When connecting to Oracle ASM, AppSync uses a bequeath connection and always
connects as SYSASM.
Red Hat Cluster Services Integration with AppSync
AppSync can work with standalone Oracle databases that are configured to failover from
node to node in an RHCS (Red Hat Cluster Services) environment.
Overview
During a replication process, if the node you used to create a service plan is not
accessible, AppSync runs the replication on another node in the cluster. If the node you
used to create the original copy is not accessible, AppSync does not rely on the Virtual IP
of the Oracle service group. Therefore, ensure that you register all nodes in the RHCS
cluster in the AppSync server for database replication.
From a restore perspective, AppSync can only restore to the node where the copy was
originally created, therefore the original node must be active, otherwise the restore
process fails.
Requirements
Review the following requirements to use a standalone database that fails over as part of
an RHCS cluster:
122
l
The AppSync host plug-in must be installed on all nodes of the cluster.
l
The IP resource must be configured in the Oracle service group for the clustered
database.
EMC AppSync 2.2.3 User and Administration Guide
Protect Oracle
l
If a failover occurs while running a replication or restore process, the operation fails.
Node failover should occur before running the service plan, before the start of a
replication, or start of a restore.
l
The Oratab file should have an entry for all possible SIDs that can run on the
specified node (passive and active instances).
l
The package sg3_utils, which contains utilities for accessing devices that use SCSI
command sets, must be installed on all nodes.
Mount considerations
l
The mount host must not be part of the RHCS cluster.
l
The mount host run the same Oracle version as the copy host.
l
The AppSync host plug-in must be installed on the mount host.
l
The package sg3_utils, which contains utilities for accessing devices that use SCSI
command sets, must be installed on the mount host.
Restore considerations
l
AppSync can only restore to the node where the copy was originally created, therefore
the original node must be active. Otherwise, the restore process fails, and corrupts
the database. The console provides a detailed warning message before the restart of
the restore.
l
To perform a restore in an RHCS environment, follow these steps:
1. Perform the restore.
2. Start the mount instance.
3. Perform a manual recovery.
4. Shut down abort database.
5. Enable the Service Group.
Oracle Data Guard support
AppSync supports an Oracle Data Guard configuration for a primary (source database)
and a physical standby (target database) which is open in active or passive/non-active
mode.
There are three types of standby databases:
l
Physical standby
l
Logical standby
l
Snapshot standby
All three configurations can be opened in one of the following modes:
l
Active standby mode—Standby database in read-only or read/write mode
l
Passive/non-active standby mode—Standby database in mounted mode
AppSync currently only supports Data Guard physical standby configuration in active or
non-active mode.
When a physical standby database is open in active mode, the standby database can be
opened in read-only mode while logs are applied. This action allows you to query the
database for information while Data Guard applies logs.
Snapshot and logical standby configurations also allow the database to be open in read/
write mode. A passive/non-active setting means that the database can start in mounted
mode and logs can be applied in the background.
Oracle Data Guard support
123
Protect Oracle
Physical standby
In a physical standby environment, archive logs are applied when they are received. A
physical standby has a 1:1 mapping of the file and storage layout from primary to
standby. A physical standby database can be open in both read-only or mounted mode
which means it can be either an active or passive/non-active configuration.
The following diagram displays a typical primary/standby (source/target) Data Guard
configuration:
Figure 14 Physical standby environment
Copy Management
On the AppSync console, go to the Copy Mangement > Oracle page. A Data Guard
relationship column now displays. If you have an existing Data Guard relationship, you
can view two databases that are part of a Data Guard configuration. One database is the
primary database and one is the physical standby (non-active)database.
Review the following copy management considerations for Data Guard:
l
To protect a primary Data Guard database (source database), create a copy like any
other standalone database. You can take a hot backup copy.
l
For protection of an active standby Data Guard database (Target Database):
Protection in hot backup mode of an active standby database is not allowed because
the standby database is in read-only mode. Also, the standby database contains upto-date archive logs and is an exact copy of the primary and does not require archive
logs to be copied for recovery. You can however take a non-hot backup copy of a
Standby database.
l
For protection of a passive/non-active standby Data Guard database (target
database): A passive/non-active standby database operates the same way as an
active standby database. Hot backup copy of the database is not allowed. The
difference here is that the copy is created from the mounted database without
opening the database in read-only or read/write mode.
l
Creating a copy of a mounted database only succeeds for a passive/non-active Data
Guard standby database in mounted state. Standalone Oracle databases that are
mounted cannot be protected. They appear as offline on the database protection
page of the console.
Mount and restore (recover)
Review the following mount and restore considerations for Data Guard:
l
124
For a primary database (Source database): Mount and restore operate the same way
with a Primary Data Guard database as any Oracle Standalone database. If you use
EMC AppSync 2.2.3 User and Administration Guide
Protect Oracle
RAC to configure the Primary database then the RAC mount/restore rules for AppSync
apply.
l
For an active standby database (target database): Mount and restore operate the
same for an active standby Data Guard database as any other Oracle standalone
database. If the standby database is configured using RAC then the RAC mount/
restore rules for AppSync apply.
l
For a passive/non-active standby database (target database): Mount and restore
operate the same for a passive/non-active standby Data Guard database as any other
Oracle standalone database. If the standby database is configured using RAC, then
the RAC mount/restore rules for AppSync apply.
Note
If you mount and restore either a primary or standby database, the database appears on
the console as a standalone Oracle database. No Data Guard configuration persists.
Repurposing (copy or a copy) Data Guard databases
For general repurposing information, refer to the AppSync user documentation.
Review the following repurpose considerations for Data Guard:
l
Repurposing a primary database (source database): Repurposing operates the same
for a primary Data Guard database like any Oracle standalone database.
l
Repurposing an active standby database (target database): Repurposing operates the
same for an active standby Data Guard database as any Oracle standalone database.
You cannot hot backup a standby database for a repurposed copy.
l
Repurposing a passive/non-active standby database (target database): repurposing
operates the same for a passive/non-active standby Data Guard database as any
Oracle standalone database. You cannot hot backup a standby database for a
repurposed copy.
Restore Data Guard databases
Restore for a primary database (source database): Restore for a primary Data Guard
database operates the same way for any Oracle standalone database. Manually recover
the database and then resynchronize the primary and standby databases after the
AppSync restore process completes.
Veritas Cluster Services integration
AppSync can work with standalone Oracle databases that are configured to failover from
node to node in a VCS (Veritas Cluster Services) environment.
Introduction
During a replication process, if the node that was used to create the service plan is not
accessible, AppSync runs the replication on another node in the cluster. AppSync does
not rely on the Virtual IP of the Oracle service group. Therefore, register all nodes in the
VCS cluster to the AppSync server before you replicate the database.
From a restore perspective, AppSync can only restore to the node where the copy was
originally created. The original node must be active, otherwise the restore process fails.
Requirements
The following are the requirements for using a standalone database that fails over as part
of a VCS cluster:
l
Install the AppSync host plug-in on all nodes of the cluster.
l
Configure the IP resource in the Oracle service group for a clustered database.
Veritas Cluster Services integration
125
Protect Oracle
l
If a failover occurs while running a replication or restore process, then the operation
fails. Node failover occurs before running a service plan, before the start of a
replication, or a restore.
l
The Oratab file should have an entry for all possible SIDs that can run on the
specified node (passive and active instances).
l
Ensure tnsnames.ora files on all nodes contain entries of all standalone
instances, including the virtual IP address of the Oracle service group (per Symantec
documentation).
l
The following files should be accessible to all nodes on the cluster where the
database runs:
l
n
Database init/spfile
n
Password file
Install package sg3_utils, which contain utilities to access devices that use SCSI
command sets, on all nodes.
Mount considerations
l
The mount host must not be part of the VCS cluster.
l
The mount host requires installation of VxVM Storage Foundations minimum 6.1.
l
The package sg3_utils, which contain utilities for accessing devices that use SCSI
command sets, must be installed on the mount host.
Restore considerations
AppSync can only restore to the node where the copy was originally created, therefore the
original node must be active. To perform a restore in a VCS environment, follow these
steps:
1. Freeze the Oracle service group: >hagrp -freeze <service_group_name>
2. Perform the restore.
3. Start the instance.
4. Perform a manual recovery.
5. Open the database.
6. Unfreeze the Oracle service group: >hagrp -unfreeze
<service_group_name>.
Note
AppSync can only restore to the node where the copy was originally created, therefore the
original node must be active. Otherwise, the restore process fails, and leaves the
database in a corrupt state. The console provides a detailed message warning you of this
scenario before the restart of the restore.
HACMP cluster integration
AppSync can work with standalone Oracle databases that are configured to failover from
node to node in an IBM® HACMP cluster environment.
Introduction
AppSync protects the database on the node where the current state is active before the
Service Plan run. AppSync does not rely on the Virtual IP of the Oracle service group.
Therefore, all nodes in the HACMP cluster should be registered in the AppSync server to
126
EMC AppSync 2.2.3 User and Administration Guide
Protect Oracle
replicate the database if the node that was used to create the original copy is not
accessible.
AppSync restores to the cluster node where the copy was originally created. The restore
process fails if the node is not active during the restore process.
Prerequisites for HACMP environment to work with AppSync
The following are the requirements for protecting a standalone database that fails over as
part of a HACMP cluster:
l
The AppSync host plug-in must be installed on all nodes of the cluster.
l
The IP resource must be configured in the Oracle service group for the clustered
database.
l
If a failover occurs while running a replication or restore process, the operation fails.
Node failover should occur before running the service plan, or at the start of a restore.
l
The Oratab file should have an entry for all possible SIDs that can run on the
specified node (passive and active instances).
l
The following files should be accessible to all nodes on the cluster where the
database runs:
n
Database init/spfile
n
Password file
Mount considerations
l
The mount host must not be part of the HACMP cluster.
l
The AppSync host plug-in must be installed on the mount host.
Restore considerations
AppSync can only restore to the node where the copy was originally created, therefore the
original node must be active. To perform a restore in an HACMP environment, follow
these steps:
1. Freeze the Oracle service group: >hagrp -freeze <service_group_name>
2. Perform the restore.
3. Start the instance.
4. Perform a manual recovery.
5. Open the database.
6. Unfreeze the Oracle service group: >hagrp -unfreeze
<service_group_name>.
Note
AppSync can only restore to the node where the copy was originally created, therefore the
original node must be active, otherwise the restore process fails, and leaves the
database in a corrupted state. The console provides a detailed message warning you of
this scenario before the restart of a restore.
Post restore procedure in an HACMP environment
Learn how to perform manual steps with a restore in an HACMP environment after a
restore.
After restore, a file system mounts to the production host in non-concurrent mode.
Remove the file system from the resource group, make it a concurrent volume group, and
then add it back to the resource group.
HACMP cluster integration
127
Protect Oracle
Perform these steps on an active node:
Procedure
1. Unmount file system.
2. Execute Varyoffvg
3. Execute Varyonvg with -c option (to make it concurrent)
4. Run importvg on the passive node.
Verification:
The lspv command should show vg as concurrent on both nodes as follows:
node 2
hdiskpower8
hdiskpower9
00c2bfb0f1ee76ca
00c2bfb0f1f434e3
oradata concurrent
oralogs concurrent
node 1
hdiskpower18
hdiskpower19
00c2bfb0f1ee76ca
00c2bfb0f1f434e3
oradata concurrent
oralogs concurrent
5. Add file system back to resource group.
6. Verify and synchronize configuration.
Prerequisites and supported configurations
Learn about prerequisites and supported configurations for Oracle with AppSync.
Included is information about supported device configurations, Oracle on file systems,
logical volume managers and ASM-based storage, RecoverPoint consistency group-based
storage, Linux and AIX-based configurations including sudo user, and support for
virtualization setups.
AppSync can create application-consistent (using Oracle hot backup) and crashconsistent (without hot backup) copies. For AppSync to create app-consistent copies of
Oracle databases, the data files and archive logs must not share the file system, volume
group, ASM disk group, RP consistency group, or data store. If the Oracle configuration is
such that the data files and archive logs share any of these groupings, then AppSync can
create crash-consistent copies for such databases.
When using VNX, ensure all consistency groups are VNX consistency groups. When using
ViPR controller, ensure all consistency groups are ViPR consistency group. Additionally,
the archive log files must be on a different CG from the rest of the database files.
Oracle on file system-based storage configurations
Some examples of Oracle configurations for which AppSync can offer both app-consistent
as well as crash-consistent copies follow:
128
l
Single database: database files on, for example, /data; archive log files on, for
example, /archive.
l
Multiple databases sharing single archive log location: for example, Database 1 on /
db1, Database 2 on /db2, archive logs on /arch.
l
Multiple databases sharing data location and archive log locations: for example,
Database 1, 2, 3 files on /data, database 1, 2, 3 archive log locations on /
archive.
l
Affected databases scenario: Two filesystems on one volume group with two more
filesystems on another volume group, such that one Oracle database has data on fs1
in vg1 and logs on fs1 on vg2 and second Oracle database has data on fs2 on vg1
and logs on fs2 on vg2.
EMC AppSync 2.2.3 User and Administration Guide
Protect Oracle
Note
AppSync does not support the following configuration: one oracle database has data files
on fs1 in vg1 and logs on fs1 on vg2, and a second Oracle database has data files on fs2
on vg2 and logs on fs2 on vg1.
Oracle on logical volume managers-based storage configurations (LVM/VxVM)
l
Single database: Database files on a volume in, for example, datavg, and then
archive log files in a volume on, for example, archvg.
l
Multiple databases sharing single archive log location: Database 1 files on a volume
in, for example, data1vg, and Database 2 files on a volume in, for example,
data2vg, and then archive logs in a volume on, for example, archvg.
l
Multiple databases sharing data location and archive log locations: Databases 1, 2, 3
files in a volume on, for example, datavg, and then Database 1, 2, 3 archive log
locations in a volume on, for example archvg.
Oracle on ASM-based storage configurations
l
Single database: Database files on, for example, diskgroup +data, then archive log
files on, for example, diskgroup +arch.
l
Multiple databases sharing a single archive log location: Database 1 files on, for
example, diskgroup +data1, and database 2 files on, for example, diskgroup
+data2, then archive logs on, for example, diskgroup +fra.
l
Multiple databases sharing a single archive log location: Database 1 files on, for
example, diskgroup +data1, and database 2 files on, for example, diskgroup
+data2, then archive logs on, for example, diskgroup +fra.
Oracle on RecoverPoint consistency group-based storage
l
Single database: Database files on LUNs in RP consistency group, for example,
DATACG and archive log files in RP consistency group, for example, ARCHCG.
l
Multiple databases sharing single archive log location: Database 1 files on LUNs in
RP consistency group, for example, DATA1CG, then database 2 files on LUNs in RP
consistency group DATA2CG and then archive log files in RP consistency group, for
example, ARCHCG.
l
Multiple databases sharing data location and archive log locations: Database 1, 2, 3
files on LUNs in RP consistency group, for example, DATACG, then database 1, 2, 3
archive logs on LUNs in RP consistency group, for example, ARCHCG.
Oracle on datastore-based storage layouts
l
Single database: Database files on vDISKs from data store, for example, DATADS and
archive log files on vDISKs from data store, for example, ARCHDS.
l
Multiple databases sharing single archive log location: Database 1 files on vDISKs
from data store, for example, DATA1DS, then database 2 files on vDISKs from data
store DATA2DS and then archive log files on vDISKs from data store, for example,
ARCHDS.
l
Multiple databases sharing data location and archive log locations: Database 1, 2, 3
files on vDISKs from data store, for example, DATADS, then database 1, 2, 3
archive logs on vDISKs from data store, for example, ARCHDS.
Oracle on VNXe LUN Group-based storage
The following configurations are supported
Prerequisites and supported configurations
129
Protect Oracle
l
Single database: Database files on LUNs in VNXe LUN group, for example, data files
in LUN Group DATALUNGRP and archive log files in LUN group ARCHCG.
l
Multiple databases sharing single archive log location: Database 1 files on LUNs in
VNXe LUN group, for example, DATA1LUNGRP, then database 2 files on LUNs in VNXe
LUN group DATA2LUNGRP, and then archive log files in VNXe LUN group, for example,
ARCHLUNGRP.
l
Multiple databases sharing data location and archive log locations: Database 1, 2, 3
files on LUNs in VNXe LUN group, for example, DATALUNGRP, then database 1, 2, 3
archive logs on LUNs in VNXe LUN group, for example, ARCHLUNGRP.
Supported virtualization configurations
AppSync supports protection, mount, and restore of Oracle databases on vDisks in
standalone and RAC.
AppSync does not support configuration where data and archive logs are on mix of RDM
and VDisks.
Considerations:
l
For Oracle databases on VDisks on VMs on ESX:
n
Disable the VMFS3hardwareaccelated locking flag on the ESX that is hosting
the VMs hosting the Oracle databases on vDisks.
n
If ATS locking is enabled for VMFS3/5 datastore, AppSync datastore mount fails.
l
To run SCSI commands from AppSync, set disk.EnableUUID on the VM.
l
Ensure your VM datastore does not share the same VMFS as your Oracle databases
Note
AppSync does not support the following configurations:
l
ASM database on VXVM volume groups
l
ASM database on VXDMP devices
l
ASM database on raw devices under the control of VXVM
l
Non-ASM database on Native LVM volume group residing on VXDMP devices
Support for Oracle on VMware virtual disks
You can protect, mount and restore Oracle standalone and clustered databases residing
on VMware virtual disks.
Consider the following information when working with Oracle and VMware virtual disks.
l
For successful mapping, add the vCenter to the AppSync server and then perform
discovery before adding the Oracle host. Otherwise you must rediscover the Oracle
host after adding the vCenter.
l
For successful protection, log files and database files must reside on virtual disks.
There cannot be a combination of physical and virtual storage.
l
AppSync does not support:
l
130
n
NFS datastores
n
Protection of Oracle databases across virtual machines sharing the same
datastore
To perform Oracle mount and recovery to a virtualized host, you need VMware
permissions to modify the VMware configuration of the mount VM (create RDM / SCSI
adapter), as well as rescan datastores/VMFS.
EMC AppSync 2.2.3 User and Administration Guide
Protect Oracle
Refer also to Oracle vDisk restore with affected entities on page 150.
Support for VIO vSCSI
Oracle with AIX LPARs can now also use "virtual" connections to the storage.
Overview: Support for VIO (Virtual I/O disk) vSCSI
Previously, AppSync supported Oracle on AIX physical machines and on AIX virtual
machines (LPARs) that use physical or NPIV connections to the array storage. AppSync
now supports Oracle AIX LPARs with virtual connections.
All supported applications and use cases for AIX Hosts using physical or NPIV storage
connections are now also supported on VIO VSCSI devices. Two restrictions apply to this
support:
l
Mounting of replicas must be done to mount hosts using physical or NPIV storage
connections. Mounts cannot be created as virtual disks .
l
The VIO Server must map whole raw disks to the VIO Clients. Do not map logical
volumes from the VIO Server.
In addition AppSync can coexist with AIX Live Partition Mobility. AppSync will continue to
protect and repurpose applications after the migration of a client partition to a new
managed server.
Supported versions
When referring to an AppSync support matrix, AIX Virtual I/O disks are supported as a
valid virtual disk type known as Virtualization Server Solutions.
Protecting a database
To protect a database, subscribe it to an AppSync service plan.
You can protect objects in different ways from different places in AppSync:
l
Select Subscribe to Plan and Run when you want to protect a selected database
immediately. The service plan is executed for that database alone.
l
Select Subscribe to Plan when you want to schedule protection for later. Protection
for databases that are part of a service plan is executed at a scheduled time.
l
Select an appropriate service plan from Create copy using a plan in the database
Copies page.
l
Select Run from the Oracle Service Plans page to run the entire plan immediately.
Discovering databases
To keep AppSync up-to-date, you should discover databases on the Oracle server when
there is creation, deletion, or renaming of databases.
Before you begin
This operation requires the Data Administrator role in AppSync.
Procedure
1. Navigate to Copy Management > Oracle to display the Databases page.
Only databases that are started and are in an open state show up as online on the
databases page. Databases that do not have an entry in the /etc/oratab file as
well as shutdown databases do not appear.
Protecting a database
131
Protect Oracle
2. From the Discover Databases drop-down menu on the bottom left of the screen, click
On Server, and then click on the desired server where the database you want to
discover resides.
Subscribe a database to a service plan
You can subscribe a database to a service plan and run the service plan immediately, or
schedule the service plan to run at a later time.
Before you begin
This operation requires the Data Administrator role in AppSync.
Procedure
1. Navigate to Copy Management > Oracle.
2. Select one or more Oracle databases.
3. From the Protect popup button, select the appropriate service plan, for example:
Table 12 Sevice plan protection options
Option
Description
Subscribe to Plan and
Run
To subscribe the database for protection and run the plan
immediately for any selected database(s).
Subscribe to Plan
To subscribe the database for protection. Protection for all databases
that are part of the service plan is executed at the scheduled time.
Unsubscribe database from a service plan
When you unsubscribe an individual database from a service plan, AppSync retains all
existing database copies; only further protection will be removed.
Before you begin
This operation requires the Data Administrator role in AppSync.
Procedure
1. Navigate to Copy Management > Oracle.
2. Select the database to unsubscribe from a service plan.
l
Select the plan to unsubscribe from: Protect > Unsubscribe from Plan. Only plans
to which the database was subscribed appear in the popup list.
l
To unsubscribe from all service plans, select Unsubscribe from Plan > All.
Oracle copies page
You can see details of a copy from the Copies tab of the Service Plan. The list of copies
can be filtered by time of creation, and by service plan.
Table 13 Copy page fields
132
Column
Description
Status
l
EMC AppSync 2.2.3 User and Administration Guide
Green: successful
Protect Oracle
Table 13 Copy page fields (continued)
Column
Description
l
Yellow: completed with errors
l
Red: failed
Name
Name of the copy. The copy name is the time AppSync created it.
Service Plan
Name of the service plan associated with the copy.
Label
Label assigned to the copy in case of repurposing.
Application
Consistent
l
Yes, if database was successfully put in hot backup mode while creating
the copy.
l
No, if hot backup mode was not selected in the Create Copy phase.
l
No, if hot backup mode was selected and database failed to go into hot
backup mode.
Mount Status
Status of the copy: mounted or not mounted. If mounted, the name of the
mount host displays.
Recovery Status
Was copy recovered post mount or not. Values are:
Copy Type
l
Not Recovered - copy was not mounted or copy was a filesystem mount.
l
Successful - recovery was successful.
l
Failed - recovery failed.
l
CDP Bookmark
l
CRR Bookmark
l
VNX Snap, VNXeSnap
l
VMAX Clone
l
VMAX Snap
l
VMAX V3: SnapVXSnap, SnapVXClone
l
XtremIO Snap
l
ViPRSnap
Generation
First or second generation copy - for repurposing
Source
Production database (for first generation copy) or a copy of a copy ( second
generation) copies.
Site
RecoverPoint site information.
Storage System
Array serial number/name
Viewing database copies
Follow these steps to view an Oracle database copy on the AppSync console.
Before you begin
This operation requires the Data Administrator role in AppSync.
Oracle copies page
133
Protect Oracle
Procedure
1. Navigate to Copy Management > Oracle
2. Click a database to view existing copies of the database.
You can see details of a copy from the Copies tab of the Service Plan. The list of copies
can be filtered by time of creation, and by service plan.
Creating a database copy from the Copies page
Create a copy of a database by subscribing it to an AppSync Oracle service plan from the
Copies page.
Before you begin
This operation requires the Data Administrator role in AppSync.
Procedure
1. Navigate to Copy Management > Oracle
2. Click a database to view existing copies.
3. From the Create a copy using a service plan list, select the appropriate service plan.
The service plan runs immediately for the selected database.
Expiring a database copy on demand
Expiring a database copy removes it from the AppSync database and can free up storage,
depending on the replication technology and copy state.
Before you begin
This operation requires the Data Administrator role in AppSync.
Expiring a copy that was made with RecoverPoint does not remove the corresponding
bookmark from RecoverPoint itself.
Procedure
1. Navigate to Copy Management > Oracle.
2. Click the desired database for copy expiration.
3. From the Copies page, select one or more copies to expire.
You can also perform this action from the service plan's Copies tab.
4. Select Expire from the row of buttons on the lower part of the screen.
5. Verify that you selected the appropriate copy, and any associated copies that are also
listed and confirm.
Service plan summary and details
The service plan Settings tab shows the name, description, schedule, and status of the
service plan.
Click the phases for detailed service plan settings and other tabs for information about
subscriptions, lists of copies and events generated by the plan.
Review Overview: Service Plan on page 10 for more service plan copy information.
134
EMC AppSync 2.2.3 User and Administration Guide
Protect Oracle
Service plan schedule
The schedule of a service plan is set in the Plan Startup phase. The Startup Type
(scheduled or on demand) determines whether the plan is run manually, or configured to
run on a schedule.
Options for scheduling when a service plan starts include:
l
l
l
l
Specify a recovery point objective (RPO).
n
Set an RPO of 30 minutes or 1, 2, 3, 4, 6, 8, 12, or 24 hours
n
Set minutes after the hour in 5 minute intervals.
n
Default RPO is 24 hours.
Runs every day at specific times.
n
Select up to two different times during the day.
n
Select minutes after the hour in 5 minute intervals.
n
There is no default selected.
Run at a certain time on selected days of the week.
n
You can select one or more days of the week (up to seven days).
n
There is no default for day of the week. Default time of day is 12:00 AM.
Runs at a certain time on selected days of the month.
n
Select one or more days of the month (up to all days).
n
Select one time of day. Available times are at 15 minute intervals.
n
Default is the first day of the month.
Overriding service plan schedules
You can set individual schedules for databases subscribed to a service plan by overriding
the generic recurrence setting.
Before you begin
This operation requires the Service Plan Administrator role in AppSync.
You can only override the settings of the recurrence type previously selected for the
service plan.
Procedure
1. Navigate to Service Plans and select one of the plans from the list.
2. From the Settings tab, select the Plan Startup phase.
The Plan Startup Defaults pane appears on the right.
3. Note the Recurrence Type selected for the plan.
A recurrence type can be set only if Scheduled was set as the Startup Type.
4. Select the Start service plan phase.
You will see the Start service plan pane on the right.
5. Note the Recurrence Type selected for the plan.
A recurrence type can be set only if Automatic is selected in the Startup phase.
Service plan schedule
135
Protect Oracle
6. Click the Plan Startup Overrides tab.
You can view the list of all databases subscribed to the plan.
7. Select one or more databases and click Override Schedule.
The Override Schedule dialog appears.
8. Set the schedule based on your requirement and then click OK.
For example, if the default recurrence type is for specified days of the month, and the
rule setting is to Run at 12:00 AM on the 1st day of every month, you can override the
time and the day for individual datastores.
Results
A Pencil icon indicates that default settings have been overidden.
Application discovery
Before creating the database copy, AppSync examines the Oracle database on the host to
look for changes such as addition, removal, and shutdown or for changes in open status.
A database is protected only if it is in the ONLINE state. This means the database(s) must
be in open mode (databases started in nomount mode or in mount mode and offline
databases are not protected).
There are no user settings associated with this phase and it cannot be disabled.
Application mapping
After discovering the application, AppSync maps it to array storage, and protection
services such as RecoverPoint.
There are no user settings associated with this phase and it cannot be disabled.
Storage preferences
Sets the preferred order of storage technology to use while creating copies, for example,
VNX Snapshot or VMAX-Clone/Snap or RecoverPoint Bookmark.
Use the Move Up and Move Down buttons. Copies are made using the first technology
preference when possible. If conditions are such that the first technology can no longer
be used, then any remaining copies will be handled by the next preference instead. For
example, if your first preference was a bookmark but not all the application data in the
service plan could be mapped to RecoverPoint, then AppSync uses Snap instead.
Note
A single service plan can contain a mix of datasets configured on VNX/VMAX block/file
and RecoverPoint. For example, with VNX, if you have a Bronze service plan for Oracle, the
databases subscribed can on a mix of RecoverPoint and VNX/VMAX block objects.
A database mix of VNX and VMAX is not supported. Also to get an RP bookmark copy for a
database, all LUNs in that database should be configured with RecoverPoint protection; if
not Snap copies are created for that database.
Pre-copy script
To perform preparatory steps before creating a copy, specify a pre-copy script and
parameters on a service plan's Settings tab.
The pre-copy script runs according to the schedule set in the Plan Startup phase.
AppSync executes this script once per host per service plan run on the production host.
136
EMC AppSync 2.2.3 User and Administration Guide
Protect Oracle
All script phases are non-blocking, which means that even if they fail, service plan
execution does not terminate and the next phase continues.
This operation requires the Data Administrator role in AppSync.
For a successful script run ensure:
l
The script phase is enabled.
l
The script exists in the specified path. You provide absolute path to script; there is no
default location.
l
You use valid script formats: all executables on UNIX are supported. The script
requires execute permissions for the specified user.
l
The pre-copy script runs per the schedule set in the Plan Startup phase.
l
The script runs as Local System by default for Windows only.
l
The script does not put the database/tablespaces in backup mode.
l
The script does not shut down the database.
Table 14 Pre-copy script console fields
Field in UI
Description
Full path to script The complete path to the script location.
Script parameters Parameters that will be passed to the script during the run.
Run as username User that has execute permissions on the script.
Password
Password of the user.
Create copy
The Create Copy phase creates a copy that is based on the replication technology that is
specified in the service plan.
The Create copy phase specifies the backup type for the Oracle database copy that
AppSyncs creates including: Snap (VNX Snapshot, XtremIO Snapshot, VNX File SnapSure,
TimeFinder VP Snap), Clone (TimeFinder Clone), or bookmark (RecoverPoint CDP/CRR/
CLR) .
This phase also sets the period for automatic expiration of the copies.
There are three types of copy phases depending on which plan you use :
l
Create local copy—For Bronze service plans, creates a local copy on the production
storage system or local site. Copy type includes:
n
Snap—VNX Snapshot, VNX File SnapSure, TimeFinder VP Snap, XtremIO Snapshot,
VNX2e Advanced Snapshot, ViPR Snapshot, VMAX3 SnapVX Snapshot in NoCopy
mode
n
Clone—TimeFinder Clone
n
Bookmark—RecoverPoint RecoverPoint Continuous Data Protection, VMAX3
SnapVX snapshot in Copy mode
– Source VNX volume, target VNX volume—virtual and virtual with roll access
modes are supported.
– VMAX V3 with RecoverPoint is not supported. VMAX V2 with RecoverPoint is
supported in 2.2.2 and later.
Create copy
137
Protect Oracle
l
Create remote copy—For Silver service plans, creates a remote copy on the remote
storage system (across RemoteReplicator or SRDF) on remote site. Copy type
includes:
n
Bookmark—RecoverPoint RecoverPoint Continuous Remote Replication
– Source VNX volume, target VMAX volume—virtual and virtual with roll access
modes are not supported.
– Source VMAX volume, target VNX volume—virtual and virtual with roll access
modes are supported.
l
Create local and remote copy—For Gold service plans, creates both local and remote
copies. Copy type includes:
n
Snap— VNX File SnapSure
n
Bookmark—RecoverPoint CLR
Automatic expiration of copies
The automatic expiration value in a service plan Create Copy phase specifies the
maximum desired number of Snap, Clone or Bookmark that can exist simultaneously.
When the "Always keep x copies" value is reached, older copies are expired to free
storage for the next copy in the rotation. Failed copies are not counted. AppSync does not
expire the oldest copy until its replacement has been successfully created. For example,
if the number of copies to keep before expiration is 7; AppSync does not expire the oldest
copy until the 8th copy is created. AppSync does not expire copies under the following
circumstances:
l
Mounted copies are not expired.
l
A copy that contains the only replica of a database will not be expired.
This setting is independent of any storage policy setting (for example the VNX pool policy
settings in Unisphere for automatic deletion of oldest snapshots.) The service plan
administrator should work with the storage administrator to ensure that the Storage
policy settings will enable the support of the specified number of snap copies for that
application.
Include RecoverPoint copies in expiration rotation policy: Check this option to include
RecoverPoint copies when calculating rotations.
Note
If this option is not selected, then RecoverPoint copies accumulate, and remain until the
bookmarks fall off the RecoverPoint appliance.
Post-copy script
To perform cleanup or other post-copy steps after creating a copy, specify a post-copy
script and parameters in the service plan Settings tab.
The pre-copy script runs as per the schedule set in the Plan Startup phase. You can
execute this phase once per host per service plan run. If this script phase is enabled but
the permissions to run it are improper, or if the script does not exist in the specified path,
the Service Plan run fails with appropriate error.
This process requires the role of AppSync Data Administrator. AppSync executes this
script once per host per service plan run on the production host.
All script phases are non-blocking, which means that even if they fail, service plan
execution does not terminate and the next phase continues.
138
EMC AppSync 2.2.3 User and Administration Guide
Protect Oracle
For a successful script run ensure:
l
The script exists in the specified path. You provide absolute path to script; there is no
default location.
l
You use valid script formats: all executables on UNIX are supported. The script
requires execute permissions for the specified user.
Table 15 Post copy script console fields
Field in UI
Description
Full path to script The complete path to the script location.
Script parameters Parameters passed to the script during the run.
Run as username User that has execute permissions on the script.
Password
Password of the user.
Unmount previous copy
The service plan unmounts a previously mounted copy after creating the new copy.
The exception is a copy that was mounted on-demand instead of mounted by the service
plan; in this case the on-demand mounted copy is not unmounted.
All the recovered databases are shut down as part of this phase. There are no user
settings associated with this phase and it can be enabled or disabled.
Pre-mount script
You can enable this phase if you want to run a script prior to AppSync performing a mount
operation.
This script will be executed once per host per service plan run. If you enable the script
phase but the permissions to run it are improper, or if the script does not exist in the
specified path, the service plan run fails with appropriate error.
Show caution when using several mount hosts in a Service Plan run. (Refer to Overriding
mount settings on a service plan on page 140. You must select Same as mount host in
the Run on host option so that the script runs on all mount hosts.
Table 16 Pre-mount script field descriptions
Field in UI
Description
Full path to script The complete path to the script location
Script parameters Parameters passed to the script during the run
Run as username User with execute permissions on the script
Password
Password of the user
Run on host
Host where the script needs to run. Select Same as mount host if several
mount hosts are involved.
Unmount previous copy
139
Protect Oracle
Mount copies
The Mount copy phase either mounts the copy or mounts and recovers the copy. This
phase can be enabled or disabled.
In Mount Copy Defaults settings, you can set values to Mount copy or Mount and recover
copy.
For Mount copy settings, you can set the mount host value and mount path and the
RecoverPoint image access type.
For Mount and recover copy settings, you specify the recovery instance, the type of
recovery, and the database naming details. Other settings are similar to the Mount copy
settings such as mount path and image access type.
For Mount on standalone server and prepare scripts for Manual Recovery Oracle mount
option, if you enable script phase after the mount operations completes AppSync creates
scripts on the mount host that you run to recover the database. The scripts are two types,
RMAN and SQL. The scripts are created under /tmp/<MOUNTED_SID_NAME>/
RecoveryScripts.
Console field descriptions:
l
Host name: This field is used to specify the host where you want to mount the Oracle
copy.
l
Mount to path: The path on which to mount database files and file systems. For ASM
RAC, this setting is unused/ignored.
l
Service Level Objective (SLO): If you are using a VMAX 3 array, a setting called
Desired Service Level Objective (SLO) is available. The option appears in the Mount
wizard and it specifies the required VMAX 3 Service Level Objectives. SLO defines the
service time operating range of a storage group.
l
Database name: This field represents the format of the mounted database name. To
specify the original database name use the token %DB% . For example: To use the
original name that is prefixed by TEST, use TEST%DB%.
l
SID name: This field represents the format of the mounted instance name. To specify
the original instance name use the token %SID%. For example: To use the original
name that is prefixed by TEST, use TEST%SID%.
l
ASM Diskgroup: This field represents the format of the ASM disk group. To specify the
original disk group name use the token %DG%. For example: To use the original name
that is prefixed by TEST, use TEST%DG%.
l
Custom initialization parameters: This field is a multi-line field which allows you to
specify settings which override any original database setting on the mounted
database copy. This field is useful for editing options such as memory settings.
Overriding mount settings in a service plan
If multiple registered databases are subscribed to the same plan, you can select different
mount settings for each database, overriding the generic settings. Recovery settings
cannot be overridden.
Before you begin
This operation requires the Data Administrator role in AppSync.
Procedure
1. Navigate to Service Plans > Oracle and click one of the service plans from the list.
140
EMC AppSync 2.2.3 User and Administration Guide
Protect Oracle
2. From the Settings tab, select Mount copy phase.
3. On the right pane, select the Mount Copy Overrides tab.
The list of servers includes all Oracle hosts whose databases are subscribed to this
plan.
4. Select the server for settings override, and then click Set Overrides.
The Override Default Mount Settings dialog displays.
5. Select options for the mount settings that you want to override.
Fields that do not have a selection; they retain their default settings.
6. Click OK.
A pencil icon appears in the first column of the row of the server with your changed
settings.
7. To revert back to default settings for a server, click Use Default Settings.
Post mount script
You can enable this phase if you want to run a script after AppSync performs a mount
operation.
This script will be executed once per host per service plan run. If you enable the script
phase but the permissions to run it are improper, or if the script does not exist in the
specified path, the service plan run fails with appropriate error.
Show caution when using several mount hosts in a Service Plan run. (Refer to the
Overriding mount settings on a service plan section. You must select Same as mount host
in the Run on host option so that the script runs on all mount hosts.
Table 17 Post-mount script field descriptions
Field in UI
Description
Full path to script The complete path to the script location
Script parameters Parameters passed to the script during the run
Run as username User with execute permissions on the script
Run on host
Host where the script needs to run. Select Same as mount host if several
mount hosts are involved.
Unmount copy
The final phase in the service plan unmounts the copy.
This phase is disabled if the Unmount previous copy phase is enabled. There are no user
settings associated with this phase.
If you have chosen to mount with recovery options (standalone, RMAN, or cluster mount)
in the Mount copy phase, all the mounted databases are shut down as part of this phase.
Post mount script
141
Protect Oracle
Mount an Oracle copy
Before performing an oracle mount on a standalone server, you need to understand the
AppSync console Mount fields and their meanings.
Mount operations
Table 18 Console field descriptions
Field
Description
Mount on Server
The server on which to mount the copy.
Mount path
The Default Mount Path is /appsync. The mount path could also be Same
as Original Path. However, this option is not available when the mount host
= production host. You can also change Default Mount Path, for
example, /EMC instead of /AppSync.
Image access
mode (during
RecoverPoint
mount)
l
Logged access: Use this mount option if the integrity check entails the
scanning of large areas of the replicated volumes. This is the only option
available when you mount to the production host. Virtual access with RPVMAX, is not supported.
l
Virtual access with roll: Provides nearly instant access to the copy, but
also updates the replicated volume in the background. When the
replicated volumes are at the requested point in time, the RPA
transparently switches to direct replica volume access, allowing heavy
processing. With RP-VMAX, and RP-XtremIO, virtual access with roll is not
supported.
l
Virtual access: provides nearly instant access to the image; it is not
intended for heavy processing. With RP-VMAX, and RP-XtremIO, virtual
access is not supported.
You can mount a copy created on any multipathing device production host, and mount it
on any multipathing device mount host. This means you can create a copy on Block/
PowerPath/MPIO devices and mount it on a mount host with any of these combinations.
For DMP, make sure you install DMP on both production and mount hosts.
Addition server information
l
With AppSync 2.2.1 and above, you can configure a temporary location per UNIX host
from the AppSync console in the Servers page.
n
AppSync uses the set temporary location during Oracle mount operations for
storing information that previously resided in /tmp/<SID>/.
n
/tmp/ is the default temporary location unless you specify otherwise.
Mounting a copy using the Oracle Mount wizard
From the AppSync console, you can perform a mount of a copy using the Oracle Mount
wizard .
Before you begin
This operation requires the Data Administrator role in AppSync.
142
EMC AppSync 2.2.3 User and Administration Guide
Protect Oracle
Procedure
1. On the Databases page, select Recover > Mount a Copy.
A list of the Oracle database instances appears.
2. From the Copies page, select a copy and click Mount
The Oracle Mount wizard launches.
3. Use the Copies or Service Plan or Type filters to select the copy to mount.
The copies list is refreshed based on the filters selected.
4. Select the wanted copy to mount. (For a RecoverPoint copy, you also have the option
to select a bookmark that is based on a specific time, however, ensure that there is a
copy available in AppSync.) Select the copy, and click Mount to launch the Mount
Copy of Oracle wizard.
5. Click Select a point in time to select a copy with a specific timestamp. The time that is
shown here is the console's time. If the console is in a different time zone from the
RecoverPoint Appliance (RPA), specify the time in the server's time zone to mount the
copy.
6. From the Mount Options page, Mount operation drop-down list, select one of the
following options: Mount on standalone server, Mount on standalone server and
create RMAN catalog entry, Mount on standalone server and recover, Mount on
standalone server and prepare scripts for manual recovery, or Mount on grid cluster
and recover as RAC database.
If you select Mount on standalone server and recover, Mount on standalone server
and prepare scripts for manual recovery, or Mount on grid cluster and recover as RAC
database , with read/write open mode for recovery, the Create TempTable Space
option is enabled. This option is used to create the TEMP TableSpace on the recoverymounted database copy. After you select Create TEMP TableSpace, AppSync shows
two other options:
a. Number of TEMPFILES': Number of files to be added to TEMP TableSpace, each of
size specified in 'Size of each file' option
b. The size_clause specifies a number of bytes, kilobytes (K), megabytes (M),
gigabytes (G), terabytes (T), petabytes (P), or exabytes (E) . The size_clause
allows you to establish amounts of disk or memory space, for example 10M. The
size of the TempTable Space equals the Temp table file that is multiplied by
the size of each file . For example, if the Temp table file count = 2 and the size of
each file = 10M, the TempTable Space Size = 20M.
AppSync generates the name of the TempTable Space in the form of
<DBNAME>_TEMP. This newly created TableSpace is set as the default TEMP
TableSpace of the mounted database instance. During unmount, AppSync drops the
created TEMP TableSpace.
Mounting a copy using the Oracle Mount wizard
143
Protect Oracle
Note
l
With manual recovery mount, scripts are prepared to both create
('Step-5_createTempTableSpace.sql') and drop
('Step-6_dropTempTableSpace.txt') TEMP TableSpace. You should drop the
created TEMP TableSpace manually before unmounting a copy with AppSync.
l
If AppSync fails to drop the TEMP Tablespace during unmount, and if a restore
operation is performed using this copy, the tablespace is restored.
l
If you attempt to restore a RecoverPoint copy, the TEMP TableSpace, if created
during mount with recovery, is also restored to production. You should drop the
TEMP TableSpace manually from the mounted database copy, and then attempt a
restore.
7. Review the default Mount and Recovery settings and make changes if wanted.
For VMAX 3 arrays, you are presented with an SLO drop-down list. You can select the
Service Level Objective (SLO) for the mount copy. If there is a storage group for the
mount host with the wanted SLO. AppSync adds the LUN to the storage group. If this
storage group does not exist, AppSync adds the LUN to any storage group that is
masked to the host. If a storage group is configured to pick target devices, AppSync
removes the devices from the storage group at the time of mount and adds them to
the storage group for the mount host. The devices are added to the original storage
group when the copy is expired. An example of the SLO menu follows:
Figure 15 For VMAX3™ select SLO
8. Click Next to display the Summary page.
9. Review the mount settings and click Finish to complete the mount.
10. In the Results page, you can view the progress of the different phases that are part of
mounting a copy.
144
EMC AppSync 2.2.3 User and Administration Guide
Protect Oracle
The last phase that is completed displays at the bottom of the list.
RMAN cataloging feature
This section includes prerequisites and restrictions for creating RMAN catalog entry, and
copying BCT file.
Mount Operation: Mount on standalone server and create RMAN catalog entry
Table 19 Console field descriptions
Field in UI
Description
RMAN user
Catalog owner
RMAN password
Catalog owner's password
RMAN connect string
The TNS alias used to connect to remote RMAN catalog
TNS_ADMIN
Path of the tnsnames.ora file where the TNS alias is specified. (Default
Path : $ORACLE_HOME\network\admin\)
ORACLE_HOME
ORACLE_HOME path for the Oracle binaries. Default: Same as production
host
ASM Diskgroup Name Specify prefix or suffix to rename diskgroups on mount host or %DG% (if
production ASM diskgroup name is to be used during mount). Default:
APS%DG%
Skip Data Files
Skip cataloging of database data files. Default: Not selected.
Notes on prerequisites
l
RMAN catalog database must exist and be accessible on the same network as the
mount host.
l
The tnsnames.ora file on the mount host must contain a TNS alias that points to
the RMAN catalog database where EMC AppSync should catalog the copy.
l
The catalog and catalog owner must be created prior to mounting a copy to be
cataloged.
l
Production database must be registered in the RMAN catalog before mounting the
copy.
l
The Oracle version running the RMAN catalog database must be equal to or greater
than the highest Oracle version of all production databases registered to that catalog.
l
Copies mounted with RMAN integration cannot be renamed using the database
rename option. This also implies that only one copy per database can be mounted on
a mount host for RMAN cataloging, and Mount to Original Host is not possible.
l
Copies mounted with Read-only access cannot be cataloged using RMAN.
l
Database must be put in hot backup mode.
l
Create backup controlfile must be selected in Create Copy phase.
Mount on standalone server and prepare scripts for manual recovery
This action overrides mount settings on a service plan. This section includes
prerequisites and details for performing a standalone mount of an Oracle copy for use
with script-assisted manual recovery steps.
Console field description:
RMAN cataloging feature
145
Protect Oracle
l
Mount to server: This field is used to specify the host where you want to mount the
Oracle copy.
l
Mount to path: The path on which to mount database files and filesystems. For ASM
RAC, this setting is unused/ignored.
l
Database name: This field represents the format of the mounted database name. To
specify the original database name use the token %DB% . For example: to use the
original name prefixed by TEST, use TEST%DB%.
l
SID name: This field represents the format of the mounted instance name. To specify
the original instance name use the token %SID%. For example: to use the original
name prefixed by TEST, use TEST%SID%.
l
ASM Diskgroup: This field represents the format of the ASM diskgroup. To specify the
original diskgroup name use the token %DG%. For example: to use the original name
prefixed by TEST, use TEST%DG%.
l
Custom initialization parameters: This field is a multi-line field which allows the you
to specify settings which will override any original database setting on the mounted
database copy. This is useful for editing options such as memory settings.
After the mount operations complete AppSync will create scripts on the mount host that
you must execute to recover the database. The scripts are RMAN scripts and SQL scripts.
The scripts are created in /tmp/<MOUNTED_SID_NAME>/RecoveryScripts. The
script files are named as Step-<number>_<operation>.<extension>. The
<number> represents the file thas must be run first and so on. The <operation>
signifies what the script does. The <extension> specifies the type of script, either
RMAN or SQL. Depending on the type of script, either execute it in RMAN or execute
through SQLPlus. The generated filenames follow:
Step-1_DatabaseRename.sql
Step-1_DatabaseFileRename.sql
Step-2_RecoverDatabase.rman
Step-3_RecoverDatabase.sql
Step-4_OpenDatabase.sql
There is only one Step-1 file created depending on whether the recovery operation was
performed using the production SID name or an altered SID name. In order to execute the
scripts, follow these steps as an Oracle user:
1. Export the Oracle SID as the SID used during recovery.
2. When executing an SQL script, login to SQLPlus using sqlplus / as sysdba. You
can then run the script: @/tmp/<MOUNTED_SID_NAME>/RecoveryScripts/
Step-<number>_<operation>.sql
3. When executing an RMAN script, login to RMAN using rman target=/. You can
then run the script as, @/tmp/<MOUNTED_SID_NAME>/RecoveryScripts/
Step-<number>_<operation>.rman.
Note
Make sure you follow the order of these steps during recovery.
146
EMC AppSync 2.2.3 User and Administration Guide
Protect Oracle
Mount on cluster and recover
This section includes prerequisites and details for performing a mount of a copy
containing an Oracle RAC database as a RAC database on another cluster or, if renamed,
back to the same cluster. The settings for this are as follows:
Console field descriptions:
l
Mount to cluster: This field is used to specify the cluster where you want to mount the
copy. Alternatively, it can be Original cluster to mount back to the production cluster.
l
Mount to servers: You can select a subset of nodes from the selected cluster, or
alternatively, all nodes in the cluster that have been added to AppSync.
Note
AppSync will only mount to cluster nodes which have been registered; unregistered
nodes will not be used.
l
Mount to path: For ASM RAC, ignore this setting
l
Database name: This field represents the format of the mounted database name. To
specify the original database name use the token %DB% . For example: to use the
original name prefixed by TEST, use TEST%DB%.
l
SID name: This field represents the format of the mounted instance name. To specify
the original instance name use the token %SID%. For example: to use the original
name prefixed by TEST, use TEST%SID%
Note
For RAC mounts, each node in the cluster receives a unique instance name, postfixed
by an numeral.
l
ASM Diskgroup: This field represents the format of the ASM diskgroup. To specify the
original diskgroup name use the token %DG%. For example: to use the original name
prefixed by TEST, use TEST%DG%.
l
Custom initialization parameters: This field is a multi-line field which allows you to
specify settings which will override any original database setting on the mounted
database copy. This is useful for editing options such as memory settings.
Mount/unmount VMAX 3 copies
Mount/unmount operations on VMAX 3 include masking/unmasking LUNs or a set of
LUNs to a host. AppSync relies on the VMAX 3 Auto-Provisioning capability.
The mount host must be zoned to the VMAX 3 array. Next, you can create a masking view
with the initiator group, port group, and storage group.
When AppSync performs a mount operation on VMAX 3, it discovers the host initiator for
the mount host first, then based on this host initiator, AppSync maps to (or from) the
masking view. This operation determines the storage group where the target LUNs are
masked/unmasked . For RDM or Vdisk mount/unmount, AppSync identifies the masking
view that is based on the host initiator for the ESX server.
You can select the wanted Service Level Objective (SLO) for the target LUN in the mount
phase of the service plan. If there is a storage group for the mount host with the wanted
SLO, AppSync adds the LUN to the storage group. If this storage group does not exist,
AppSync adds the LUN to any storage group that is masked to the host.
Mount on cluster and recover
147
Protect Oracle
If a storage group is configured to pick target devices, AppSync removes the devices from
the storage group at the time of mount and adds them to the storage group for the mount
host. The devices are added to the original storage group when the copy is expired.
Restoring an Oracle copy
You can perform a restore of an Oracle copy using the Oracle Restore wizard from the
AppSync console.
Before you begin
This operation requires the Data Administrator role in AppSync.
Note
For AppSync 2.2.2 and later and XtremIO 4.0 and later the Restore button is enabled in
the wizard for automated restore. If you are restoring two copies, one for XtremIO 3.x and
another for XtremIO 4.0 AppSync restores only the 4.0 copy (warning for 3.x copy
appears).
Procedure
1. On the AppSync console, click the Copy Management tab, then select Oracle from the
drop-down list.
A list of the Oracle database instances appears.
2. Select a database to open the Copies Page page for the selected Oracle database
which lists available copies with dates of copy.
3. Select the desired copy, and then click Restore to launch the Oracle Restore wizard.
You may receive the following warning message: You are attempting to
perform a restore on a cluster. Please follow the
instructions in the AppSync documentation for specific
cluster restore procedures.
4. Select the copy to restore, and then click Next.
The Restore Options page appears.
5. Click the Restore drop-down list and select one of the following options to restore:
Data, Archive logs, or Both Data and Archive logs.
If the database being restored affects any other database. you may receive an affected
entity warning message.
6. Click Next to display the Summary page.
7. Review your restore settings and click Finish to complete the restore.
On the Results page you can view the progress of the different phases that are part of
restoring a copy.
Results
Appsync only displays restore warnings for databases discoverable by AppSync that are
common to that host. No warnings display for any databases which either are not
common to the host or not discoverable.
Refer also to Restoring a RAC copy on page 150.
148
EMC AppSync 2.2.3 User and Administration Guide
Protect Oracle
Affected entities during restore
When restoring from a copy, you may be prompted to restore items in addition to the
ones you selected.
An affected entity is data that resides on your production host that unintentionally
becomes part of a replica because of its proximity to the data you intend to protect. You
can prevent affected entity situations by properly planning your data layout based on
replica granularity. The granularity of a replica depends upon the environment.
For Oracle, an affected entity can only be another Oracle Database data file(s) or archive
logs. You can choose to restore using one of three options. This will determine the level
to which affected entities are determined.
l
Data only
l
Archive Logs only
l
Data and Archive Logs
Affected entities only display according to the restore option. If you select, Data, Appsync
looks for affected entities with respect to the Oracle database data filesystems and
storage. AppSync does not use Oracle database(s) archive log storage for checking for
affected entities.
If you select, Archive logs, the reverse is true. Only the Oracle database archive logs
filesystems and storage are used for checking affected entities and not the Oracle
database(s) data filesystems.
If you select both Data and Archive logs, then filesystems and storage from both the
Oracle database(s) data files and archive logs will be used for checking for affected
entities.
If there are affected entities in your underlying storage configuration, the Restore Wizard
notifies you of these items. The following scenarios produce affected entities that require
you to acknowledge that additional items will be restored:
l
For RecoverPoint and ViPR Controller, if the databases are in the same consistency
group they become affected entities when the other database is protected.
l
For VNX, VNXe, VMAX, XtremIO, and ViPR Controller, if the databases are on the same
LUN they become affected entities when the other database is protected. For VNXe, if
the databases are in the same LUN group they become affected entities when the
other database is protected.
l
For vDISK/datastore - If data files of two data bases: DB1 and DB2 reside on
datastore [DS1] and or similarly archive logs of same two databases resides on
datastore [DS2], then both become affected entities.
If the affected entity was protected along with the Oracle database selected for restore,
AppSync restores it. Any other Oracle database that was not protected but is an affected
entity is overwritten.
AppSync calculates affected entities for the consistency groups or LUNs of the Oracle
database that is selected for restore. If the affected databases partially reside on other
consistency groups, LUN groups, or LUNs, AppSync does not calculate affected entities
on those consistency groups, LUN groups, or LUNs.
Affected entities are calculated on the basis of restore granularity. If both data and log
are selected for restore, then affected entities are calculated for all the consistency
groups, LUN groups, LUNs, or datastores on which the database resides. If only data or
only log restore is selected, then the affected entities are only calculated for the selected
component's consistency group, LUN group, LUN, or datastore.
If the database's data and log components reside on the same consistency group or LUN,
the option to restore only logs or restore only data is not available. You have the option
Affected entities during restore
149
Protect Oracle
only to restore data and logs. The only exception to this scenario is when you choose to
do a differential copy restore.
Vdisk restore with affected entities
Review this information for a Vdisk restore with affected entities.
l
During restore, if there are affected databases on virtual disks that are not protected
by AppSync, shutdown the these databases including all unmounted filesystems.
Additionally, remove Vdisks from VM before proceeding with LUN level restore.
l
If affected databases reside on any volume or disk groups, then deport or dismount
VGs and DGs before restore and then manually import and mount them post-restore.
(Since Appsync does not control these entities, a post storage LUN restore can fail
when attempting import/mount of affected VGs and DGs on the production host.)
l
Affected entity databases on Vdisks with VG or ASM are not supported.
Restoring a RAC copy
Follow this procedure to restore a RAC copy.
Before you begin
On remote nodes follow these steps:
1. Shutdown all impacted databases as oracle user: oracle> srvctl stop
instance -d <RACDB> -i <DbInstanceOnRemoteNodes>
2. Dismount all impacted ASM disk groups as grid user: grid> asmcmd umount
<DG>
On the restore node, perform the restore. Follow these steps:
Procedure
1. On the AppSync console, go to Copy Management > Oracle.
2. Select the desired database, and then from the drop-down menu in the lower center of
the screen click Recover and select Restore.
The Oracle Restore wizard launches.
3. On the Select Copy page, select the copy you want to restore, and then click Next.
4. Under Restore Options, select the option you want, and then click Next.
The Summary page opens.
5. Review your restore actions and click Finish.
6. Verify that the selected database is being shut down.
7. Verify the message that disk goups and devices are being unmounted.
8. Verify that the restore was successful.
Results
After the restore:
1. Remount the diskgroups on the remote nodes as grid user:grid> asmcmd mount
<DG>.
2. On any node, perform recovery of the restored database using redo or archive with
resetlogs:
Oracle > startup mount
Oracle > recover database
150
EMC AppSync 2.2.3 User and Administration Guide
Protect Oracle
3. Open the database on the recovery node:
Oracle > alter database open
4. Bring up the instances on the additional nodes:
srvctl start instance -d <RACDB> -i <DbInstanceOnRemoteNodes>
Restoring a RAC copy for affected entities
Follow these steps to create your restore.
Before you begin
On remote nodes follow these steps:
1. Shutdown all impacted databases as oracle user: oracle> srvctl stop
instance -d <RACDB> -i <DbInstanceOnRemoteNodes>
2. Shutdown other affected databases: oracle> srvctl stop database -d
<RACDB2>
3. Dismount all impacted ASM disk groups as grid user: grid> asmcmd umount
<DG>
On the restore node, perform the restore. Follow these steps:
Procedure
1. On the AppSync console, go to Copy Management > Oracle.
2. Select the desired database, and then from the drop-down menu in the lower center of
the screen click Recover and select Restore.
The Oracle Restore wizard launches.
Verify the Warning:
You are attempting to perform a restore on a cluster. Please
follow the instructions in the AppSync documentation for
specific cluster restore procedures.
3. On the Select Copy page, copy should already be selected. Click Next.
4. Under Restore Options, select Data, and then click Next.
Verify that the Affected Entities warning has been displayed that there is an database
that is impacted and that you need to shutdown the database manually.
Note
A database may have associated entities on the same storage but in a different node.
If this is the case this warning will not display.
5. Click Next at the Affected Entities screen.
6. Review the Summary and then click Finish.
7. Verify that the selected database is being shut down.
8. Verify the message that disk goups and devices are being unmounted.
9. Verify that the restore was successful.
Results
After the restore:
1. Remount the diskgroups on the remote nodes as grid user:grid> asmcmd mount
<DG>.
Restoring a RAC copy for affected entities
151
Protect Oracle
2. On any node, perform recovery of the restored database using redo or archive with
resetlogs:
Oracle > startup mount
Oracle > recover database
3. Open the database on the recovery node:
Oracle > alter database open
4. Bring up the instances on the additional nodes:
srvctl start instance -d <RACDB> -i <DbInstanceOnRemoteNodes>
5. Repeat steps 2 and 3 to recover affected database <RACDB2>
6. Bring up affected database <RACDB2>
Restore a copy from XtremIO
Review this information regarding restoring an Oracle copy from XtremIO.
For AppSync 2.2.2 and later and XtremIO 4.0 and later the Restore button is enabled in
the Oracle Restore wizard for automated restore. If you are restoring two copies, one for
XtremIO 3.x and another for XtremIO 4.0 AppSync restores only the 4.0 copy (warning for
3.x copy appears).
To restore an Oracle database for XtremIO 3.x, mount the desired AppSync copy on the
production host, and then use RMAN to recover the database to the desired point in time
(includes RAC configurations).
If you cannot recover the database using RMAN, you might need to perform the following
restore steps to replace (restore) the Oracle database completely. Because Oracle
configurations on Linux environments vary (ASM, udev, MPIO, LVM, and so on), test and
document the full restore procedures on a non-production Oracle configuration that is
configured similarly to your production host before you use the procedures in a real
disaster recovery situation. Doing so ensures a successful restore scenario. The following
full restore procedure example is based on a stand-alone Oracle database on ASM that
uses udev devices.
The main concern to observe during the restore procedure is to prevent a conflict
between the production configuration and the mounted configuration. Ensure that you
shut down the production object (for example, Oracle database, ASM disk group, and so
on) before you attempt to convert the copy to production.
Restore Oracle database copy with AppSync and XtremIO
This section includes steps to perform a manual restore of an Oracle database copy
residing on an XtremIOarray for 2.2.1 and earlier and XtremIO versions before 4.0.
Before you begin
With AppSync 2.2.2 and later and XtremIO 4.0 and later use the Restore Wizard to restore
copies. Refer to Restoring an Oracle copy on page 148.
For manual restore, you need an Oracle database copy on XtremIO storage with an
AppSyncservice plan that replicates the database. Suspend any AppSync service plans
that might replicate the database that you are trying to restore
Procedure
1. In the AppSync console, mount the Oracle copy that you want to restore. Select Mount
on standalone server, and then select Original Server or specify the production host
where you want to mount the copy.
Because you are mounting to the original server, select Default Path for the mount
path.
152
EMC AppSync 2.2.3 User and Administration Guide
Protect Oracle
2. On the production host, stop any application that uses the database being restored.
3. On the production host, shut down the Oracle instance being restored.
ASM configuration is assumed.
4. Identify the ASM disk groups for the Oracle database being restored, and then
dismount them.
Figure 16 Dismount ASM disk groups
Figure 17 Change disk groups
5. Check the ASM disk string parameter (asm_diskstring), and then remove the
production ASM device pattern.
The mounted ASM devices have /dev/emc-appsync pattern as a prefix. In the
following example, the production ASM disks used /dev/asm-disk* as a device
pattern.
Restore a copy from XtremIO
153
Protect Oracle
Figure 18 Pattern = /dev/asm-disk*
6. Mount the ASM disk groups, which point to the mounted snapshot copy.
Figure 19 Mount ASM disk groups
7. Adjust the disk group bindings so that they do not use the EMC bindings. You can
perform this task now or later.
If you adjust the disk group bindings later, stop Oracle, and then dismount the disk
groups. For ASM using udev devices on SuSe Linux, includes changing the rules files
in /etc/udev/rules.d. Each version of Linux is different, so be sure to
understand the environment, and how Oracle is configured. If you perform this step
now, add the original ASM bindings back to asm_diskgroup.
8. Start Oracle, and then recover the database.
154
EMC AppSync 2.2.3 User and Administration Guide
Protect Oracle
Figure 20 Start, and then recover database
The restored database should be online.
9. Restart any database application that you stopped in step 2.
10. On the AppSync console, select Remove to remove the mounted copy.
This action removes the copy from the AppSync database but does not remove the
snapshot on the production mount host or on the XtremIO storage array. You can
remove a mounted copy, but you cannot set the expiration of a mounted copy.
11. Move the mounted snapshot from the /AppSyncSnapshots folder to the folder
where the production LUNs to be replaced reside.
Move the mounted snapshot (the snapshot that is currently masked to the ESX
initiator group) if the production host is a virtual machine. If the production host is a
physical host, move the snapshot to the folder where the production LUNs reside.
After you finish, you see the production LUNs and the snapshots in the same folder.
Figure 21 Before snapshot moves
Figure 22 After snapshot moves
Restore a copy from XtremIO
155
Protect Oracle
12. If the production host is a virtual machine, unmap the original production volumes
from the ESX server, rescan the ESX host, and then remove the RDMs from the
production virtual machine configuration. If the production host is a physical host,
unmap the original production volumes from the production host, and then perform a
device rescan on the host to remove the old production devices.
13. Remove the former production volumes from the XtremIO console.
The restored snapshots become volumes.
Figure 23 Restored snapshot becomes volumes
14. Rename the snapshots to the production volume names.
Figure 24 Rename snapshots
The other snapshots in the /AppSyncSnapshots folder for the old production
XtremIO volumes that you removed in Step 11 also become volumes.
Figure 25 Removed snapshots become volumes
15. Rediscover the production host to refresh the AppSync console view of the ASM
configuration. Next, run the service plan to create a snap of the restored database,
and then enable any service plans suspended in Step 1.
Repurposing overview
This topic explains how to use the AppSync repurposing feature for database and
Bookmark copies.
AppSync supports the ad-hoc creation of Oracle and SQL Server on page 117 database
copies, as well as RecoverPoint Bookmark copies, followed by the creation of copies of
those copies. This practice is referred to as repurposing. Repurposing serves many useful
functions including test-dev, break-fix, data mining, and reporting.
Review and consider the following information regarding repurposing features:
156
EMC AppSync 2.2.3 User and Administration Guide
Protect Oracle
l
Repurposing creates a multi-level tree of copies of the database.
l
AppSync identifies copies that are created from a repurpose action by a "generation
removed" number from the production source data, for example, Gen 1, Gen 2, and
so on.
l
There is no practical limit to the number of generations, but support is limited to two
generations removed from the production data.
l
A first generation(1st Gen) copy creates a copy that can be used as source for other
copies. This action creates a copy-of-copy.
l
Repurpose copies are meant to be mounted for extended periods of time for various
purposes.
l
After use, repurpose copies are either discarded or refreshed.
l
Repurpose copies do not figure into RPO calculations.
l
You can run 1st Gen copies now or schedule them to run at a future time.
l
AppSync-supported copy types include VNX Snap, TimeFinder clone of clone,
Snapshot of clone on VMAX, SnapVx of SnapVx for VMAX 3, XtremIO Snapshot, and
RecoverPoint Bookmarks.
l
Restore is not supported from 2nd Gen copies.
The console displays all scheduled repurpose actions. Select any of them, and then
select the Delete button. The scheduled activity is removed.
If you enable hot-backup, the 1st Gen copy of the database creates the application
consistent copy. It Includes application discovery, mapping, and database freeze/thaw.
AppSync does not set database freeze/thaw as the default for 1st Gen repurpose copies.
Second generation (2nd Gen) copies are created as copies-of-copies, using the 1st Gen
copy as the source. They do not include application discovery, mapping, and database
freeze/thaw. As a result, 2nd Gen copies are created much faster than 1st Gen copies.
Note
Repurposing copies are not supported for Oracle databases on VNX File storage, or NFS
File systems.
Repurposing overview
157
Protect Oracle
Repurpose refresh
Refresh means to discard the current copy (expire), and recreate the copy contents using
its parent.
l
1st Gen and 2nd Gen copies can be refreshed.
l
Refreshing a 1st Gen copy creates an application consistent copy with a new time.
l
2nd Gen copies are not modified if you refresh 1st Gen copy.
l
Refresh of 2nd Gen copy resynchronizes the 2nd Gen copy with the 1st Gen parent.
(Use for discarding changes of 2nd Gen copy and starting over.)
l
2nd Gen timestamp is the same as the 1st Gen copy.
l
With refresh, it is possible to have first and 2nd Gen copies with different copy time
stamps.
l
If you specified Unmount Previous Copy when creating the 2nd Gen copy, AppSync
unmounts the copy before creating the new copy.
l
Refresh of a 1st Gen copy on XtremIO: If the number of LUNs in the source database
changes, then create only new snapshots. For example, AppSync uses the refresh
provided by the XtremIO array from source LUNs to create the 1st Gen snapshot copy.
l
Refresh of 2nd Gen copy on XtremIO: If you refresh the source 1st Gen copy and the
number of LUNs in the source database changed, then create only new snapshots.
Otherwise, use the refresh provided by the XtremIO array from 1st Gen snapshots to
2nd Gen snapshot copies.
l
After the refresh, AppSync remounts the copy if the copy was mounted.
l
2nd Gen copy is recreated from 1st Gen parent. If 1st Gen parent no longer exists,
then the refresh fails.
l
1st Gen copy is recreated from the application.
l
The current copy that you want to refresh must be unmounted (if it is mounted). If you
selected Mount Copy in the Repurpose wizard, AppSync mounts the copy again.
You can refresh a repurposed copy at any time. To start the refresh:
1. From the Applications tab of the AppSync console, select the repurposed copy that
you want to refresh.
2. Click Refresh.
158
EMC AppSync 2.2.3 User and Administration Guide
Protect Oracle
Repurpose expire
Expiring first or 2nd Gen copies removes the ability to refresh the copy. Expiration of a 1st
Gen copy fails when a 2nd Gen copies is created from the 1st Gen copy.
Use expire if you are finished with a copy.
Using the Repurpose wizard
Use the Repurpose wizard to schedule or immediately create 1st Gen or 2nd Gen copies
as needed.
Before you begin
You need AppSync administrative privileges to Repurpose the database instance.
To display the list of available applications.
Procedure
1. Log in to the AppSync console and go to Copy Management.
Option
Description
To Repurpose an SQL copy:
Select SQL Server
To Repurpose an Oracle copy:
Select Oracle
A list of available databases for the application choice loads.
2. Click the database instance you want to Repurpose, and then select Repurpose from
the drop-down list in the lower left of the console screen.
This action launches the Repurpose wizard, and leads you to the Intentions page
where you can tell AppSync which action you want to perform:
l
Create First Generation copy
l
Create First Generation copy and a Second Generation copy
3. Select the desired copy type, and select Local or Remote in the Site drop-down to
continue creating the copy.
4. Click Next to launch the Settings screen.
From the Settings screen, you can define the specific options for 1st and 2nd gen
copies. Specifically you can
l
Define labels for each copy to help identify the copy purpose.
l
Select application-specific copy options for 1st gen copy only.
l
Choose appropriate copy type (wizard fails if incorrect type is chosen).
5. Select the desired options for the copy, and then click Next.
The schedule page of the wizard appears allowing you to identify when the 1st
generation copy should be created. Create the copy now or schedule the copy for a
convenient time.
6. Select Run now or Run later followed by typing run- time date and time, and then
press Next to complete the wizard.
After you finish
The Repurpose Monitor
The Repurpose Monitor allows you to view all currently running repurpose activities, and
monitor their progress. The Repurpose Monitor shows the item being repurposed (source)
Using the Repurpose wizard
159
Protect Oracle
and the label of the item being created or refreshed along with the application type. Refer
to The Repurpose Monitor on page 209.
View or cancel scheduled repurpose copies
You can view or cancel any scheduled first generation repurpose copy.
Procedure
1. From the Appsync console, navigate to Application > Copy Management, and then
select the appropriate database.
2. Select Repurpose > View Scheduled Repurpose Copy to view all scheduled repurpose
actions.
3. To delete an action, select one or more desired actions, and click Delete to remove the
action.
Mounting and recovering an Oracle clone of clone of RecoverPoint Bookmark
These steps show you how to mount and recover an Oracle database clone of clone for a
RecoverPoint Bookmark copy on VMAX.
Before you begin
Ensure the AppSync configuration has one Oracle database copy on RecoverPoint with
VMAX. Enable hot backup for 1st Gen copies for guaranteed recovery.
Procedure
1. On the AppSync console go to Applications > Oracle, select an Oracle database that is
on VMAX storage, and then select Repurpose.
The action launches the Intentions page of the Repurpose wizard.
2. On the Intentions page, select create 1st Gen and create 2nd Gen copies, select Use
Bookmark as an intermediate step, and then click Next.
3. On Settings > 1st Gen copy > Create a label select Clone copy type.
4. On the settings page for 2nd Gen copy, select a label, and then select Clone as copy
type.
5. Select mount settings for the 1st Gen copy, and then select Unmount phase for 1st
gen copy.
6. Select mount settings as Mount and Recovery for the 2nd Gen copy, and then select
Unmount phase for 2nd gen copy.
7. On the Schedule page, select Run Now to finish the wizard.
Mounting and recovering an Oracle Snap of Clone RecoverPoint Bookmark
These steps show you how to mount and recover an Oracle database Snap of Clone for a
RecoverPoint Bookmark copy on VMAX.
Before you begin
Ensure the AppSync configuration has one database on RecoverPoint VNX storage and
another on RecoverPoint VMAX storage. Snap of Snap for VNX and Snap of Snap for VMAX
is not supported.
Select 1st Gen copies with Hot backup enabled for guaranteed recovery.
Procedure
1. On the AppSync console go to Applications > Oracle, select an Oracle database that is
on VMAX storage, and then select Repurpose.
160
EMC AppSync 2.2.3 User and Administration Guide
Protect Oracle
The action launches the Intentions page of the Repurpose wizard.
2. On the Intentions page, select create 1st Gen and create 2nd Gen copies, select Use
Bookmark as an intermediate step, and then click Next.
3. Go to Settings > 1st Gen copy > Create a label and then select Clone copy type.
4. Go to Settings > 2nd Gen copy > Select Snap copy type, then select a label.
5. Select Mount Settings for the 2nd Gen copy.
6. On the Schedule page, select Run Now and Finish.
Mounting and recovering an Oracle Snap of Clone RecoverPoint Bookmark
161
Protect Oracle
162
EMC AppSync 2.2.3 User and Administration Guide
CHAPTER 8
Protect file systems
This chapter includes the following topics:
l
l
l
l
Overview of file system support...........................................................................164
Summary of file system service plan settings...................................................... 165
Mounting a copy with the File System Mount wizard............................................ 173
Restoring a file system........................................................................................ 175
Protect file systems
163
Protect file systems
Overview of file system support
Use AppSync to create and manage application-consistent copies of file systems.
File system features include:
l
Dynamic discovery of file systems during service plan run.
l
Protection of file systems with service plan or with copy now option. You can select
one or more file systems to protect at one time or click SELECT ALL to protect all the
file systems on the list of file systems page.
l
List copies that you can filter by time of creation, copy status, and service plan.
l
Mount on a standalone server
Protect NFS file systems on VNX storage
Learn how AppSync supports protection of NFS file systems on VNX storage.
AppSync supports protecting NFS file systems on Linux (RHEL, SUSE, and OEL) and AIX.
You can use these copies for operational recovery.
Use Bronze, Silver, and Gold service plans. For service plans configured for remote
protection, the NFS copy is created as a SnapSure Snapshot on the local and/or remote
file system. Copies of NFS data stores can be created from service plans configured for
local, remote, and local and remote protection. AppSync can also create copies for file
system on an Oracle database for Bronze, Silver, and Gold service plans.
During restore from an NFS copy, AppSync creates a roll back snapshot for every file
system that has been restored. The name of each roll back snapshot can be found in the
restore details. You can manually delete the roll back snapshot after verifying the
contents of the restore. Retaining these snapshots beyond their useful life can fill the
VNX snap cache and cause resource issues.
Review the following pre-requisites for Silver and Gold copies:
l
Register remote VNX arrays with AppSync.
l
Create Remote Replication sessions with corresponding remote arrays for each NFS
file system where you want creation of Silver and Gold copies. Ensure array status is
OK.
Host file systems page
The Host file systems page shows all the host instances currently registered with
AppSync.
You can add a host (Windows and UNIX ) from this page and can discover file systems
available on the hosts.
For further information on file systems, refer to:
l
File systems page on page 164
l
File system copies page on page 165
l
Mounting a copy with the File System Mount wizard on page 173
Filesystem page
The Filesystem page lists all the available filesystems that are discovered for the selected
server instance.
Click on a filesystem name to display copies of the filesystem.
164
EMC AppSync 2.2.3 User and Administration Guide
Protect file systems
Filesystem information includes:
l
Status of service plan run, for example checkmark in a green circle = successful
l
Name
l
Type, for example, NTFS
l
Format, for example MBR
l
Service plan, for example Bronze
Some filesystems can be subscribed to multiple serviceplans.
l
Storage size in GB
l
Send alerts to (if requested)
You can select one or more filesystems to protect at one time. Click SELECT ALL to protect
all the filesystems on this page (except a filesystem C:\ which contains host system
information). For further information on filesystems, refer to:
l
Filesystem copies page on page 165
l
Filesystem hosts page on page 164
l
Mounting a copy with the Filesystem Mount wizard on page 173
Filesystem Copies Page
In this page you can view the list of filesystem copies.
The list of copies can be filtered by time of creation, the status of the copies that are
created and service plan.
Select a copy to display events for that copy in the Details panel located on the bottom of
the Copies page.
From the Copies page you can select to mount, restore or expire copies.
For further information on filesystems, refer to:
l
Filesystems page on page 164
l
Filesystem hosts page on page 164
l
Mounting a copy with the Filesystem Mount wizard on page 173
Summary of file system service plan settings
Use this table to learn default file system settings for service plan phases including
startup, discovery, mapping, pre and post copy scripting, mount/unmount and copy.
Default service plan settings create an application-consistent copy every 24 hours. Only
the replication technology that is specified by the Copy type in the Create copy phase
varies among plans. The following table summarizes the default settings:
Table 20 Default file system Service Plan Settings
Setting
Enabled/No Default settings
t enabled
Schedule
Plan Startup
Enabled
Recurrence type: Creates
a copy every 24 hours,
with the first run at
midnight (00:00).
Automatic schedule
Summary of file system service plan settings
165
Protect file systems
Table 20 Default file system Service Plan Settings (continued)
Setting
Enabled/No Default settings
t enabled
Schedule
Recovery Point Objective
(RPO): A copy should be
created every 24 hours.
(Alert issued if objective
is not met.)
Application
discovery
Enabled
None
Determined by Plan
Startup phase.
Application
mapping
Enabled
None
Starts when Application
discovery phase
completes.
Pre-copy script
Not enabled
None
Starts when Application
mapping phase
completes.
Create copy
Enabled
Copy type:
Starts when Pre-copy
script phase completes.
l
Bronze
l
Silver
l
Gold
Also:
166
l
Storage Ordered Preference:
Snapshot, Clone, and
Bookmark.
l
Storage Settings: Include
RecoverPoint copies in
expiration rotation policy—
select this option to include
RecoverPoint copies when
calculating rotations. If you
do not select this option,
RecoverPoint copies
accumulate and remain until
the bookmarks for them "fall
off" the RecoverPoint
appliance.
Post-copy script
Not enabled
None
Starts when Create copy
phase completes.
Unmount
previous copy
Not enabled
None
Starts when Post-copy
script phase completes.
Mount copy
(A pre-mount
script phase is
available for file
system service
plans)
Not enabled
Mount Copy
Starts when Unmount
previous copy phase
completes.
EMC AppSync 2.2.3 User and Administration Guide
l
Mount on Server: Original
Host
l
Mount with access: Read/
write
Protect file systems
Table 20 Default file system Service Plan Settings (continued)
Setting
Enabled/No Default settings
t enabled
l
Mount Path: Default Path
Image
l
Access mode: Logged access
l
Copy to Mount: Local (Only
for Gold Plans)
l
Use Dedicated Storage
Group: Selected by default
Schedule
Post-mount script Not enabled
None
Starts when Mount copy
phase completes.
Unmount copy
None
Starts when Post-mount
script phase completes.
Not enabled
Subscribing a file system to a service plan
This sections shows you how to subscribe a file system to a service plan. Protection for all
file systems that are part of a service plan runs at the scheduled time.
Procedure
1. Browse to Copy Management > Filesystem.
2. Click the desired server.
The File Systems Page loads for the selected server.
3. Select the file systems that you want to protect, then click Subscribe to Service Plan
on the Protect drop-down list.
4. Select Gold, Silver, or Bronze service plan.
You can also select Subscribe to Service Plan and run for immediate subscription and
protection.
Overriding service plan schedules
You can set different schedules for individual applications that are subscribed to a
service plan, overriding the generic recurrence setting.
Before you begin
This operation requires the Data Administrator role in AppSync.
You can override only the settings of the recurrence type that is already selected for the
service plan.
Procedure
1. Browse to Service Plans and select one of the plans from the list.
2. From the Settings tab, select the Plan Startup phase.
3. In the Plan Startup Defaults pane on the right, note the Recurrence Type selected for
the plan.
A recurrence type can be set only if Scheduled is selected as the Startup Type.
Subscribing a file system to a service plan
167
Protect file systems
4. Click the Plan Startup Overrides tab.
You can see the list of all applications that are subscribed to the plan.
5. Select one or more applications and click Override Schedule.
6. In the Override Schedule dialog box, set the schedule that is based on your
requirement and click OK.
For example, if the default recurrence type is On specified days of the month, and the
rule setting is to Run at 12:00 AM on the 1st day of every month, you can override the
time and the day for individual applications.
A Pencil icon indicates that default settings have been overridden.
Service plan schedule
The schedule of a service plan is set in the Plan Startup phase.
The Startup Type (scheduled or on demand) determines whether the plan is run
manually, or configured to run on a schedule. Options for scheduling when a service plan
starts are:
l
l
l
l
Specify a recovery point objective (RPO)
n
Set an RPO of 30 minutes or 1, 2, 3, 4, 6, 8, 12, or 24 hours
n
Minutes after the hour are set in 5 minute intervals
n
Default RPO is 24 hours
Run every day at certain times
n
Select up to two different times during the day
n
Minutes after the hour is in 5 minute intervals
n
There is no default selected
Run at a certain time on selected days of the week
n
One or more days of the week (up to all seven days) can be selected
n
There is no default day of the week selected. Default time of day is 12:00 AM.
Run at a certain time on selected days of the month
n
Select one or more days of the month (up to all days)
n
Select one time of day. Available times are at 15 minute intervals.
n
Default is the first day of the month
Application discovery
Before creating a file system copy, AppSync examines the file system to look for changes
such as addition, deletion, renaming, or movement of file systems. If individual file
systems are being protected, AppSync rediscovers information about the selected file
system.
There are no user settings associated with this phase and it cannot be disabled.
Application mapping
After discovering the application, AppSync maps it to array storage, and protection
services such as RecoverPoint.
There are no user settings associated with this phase and it cannot be disabled.
168
EMC AppSync 2.2.3 User and Administration Guide
Protect file systems
Pre-copy script phase
To perform preparatory steps before creating a copy, specify a pre-copy script and
parameters on a service plan's Settings tab.
This operation requires the Service Plan Administrator role in.
The pre-copy script runs according to the schedule set in the Plan Startup phase. Valid
script formats are .bat, .exe, and .ps1 (PowerShell scripts) for Windows and .sh for UNIX.
AppSync does not support running of PowerShell scripts directly. You usually have to
wrap them in a .bat file. The other option is to make the default "Open" on ps1 files C:
\WINDOWS\system32\WindowsPowerShell\v1.0\powershell.exe. When the
PS script runs, you may get an error and you will need to set an appropriate execution
policy.
To run PowerShell commands from scripts:
1. Specify the full path name to your PowerShell command file in the .bat file:
powershell -command C:\PshellCommands.ps1 <nul
2. Set the PowerShell execution policy so you can run your script. For example, the first
line in the .bat file should look like the following for an unrestricted policy:
powershell -command set-executionpolicy unrestricted <nul
3. To ensure correct termination of your PowerShell session, add <nul to the end of the
line that calls your PowerShell script.
For Windows you can optionally enter credentials to run the script as a specific user. The
script runs as Local System by default. For UNIX the credentials are mandatory.
For Windows the default location of the script is %ProgramData%\EMC\AppSync
\scripts\ on the application host.
For UNIX the File field should have the full path to the scripts that are to be executed.
Exact parameters depend on your script. Parameters with spaces must be enclosed in
double quotes. This phase can be enabled or disabled. This operation requires the
Service Plan Administrator role in AppSync.
Create copy phase features freeze and thaw callout scripts
The Create copy phase creates a copy that is based on the replication technology that is
specified in the service plan. You can configure and run freeze and thaw callout scripts in
this phase
Configuring and running freeze and thaw callout scripts
AppSync provides two scripting opportunities during the execution of Create copy phase,
called the freeze and thaw callout scripts. Unlike pre-copy and post-copy script phases
which are run before and after the Create copy phase, freeze and thaw scripts are
unspecifiable by the GUI. The scripts are placed in a pre-defined location with a predefined name. You can use these scripts to quiesce (suspend I/O) and thaw on any
AppSync unsupported databases residing on the subscribed file systems for a short
period ( usually few seconds). During this time the copy is activated. The scripts are run
with the user credentials used to register host-plugin with AppSync.
During the Create copy phase, when AppSync executes these scripts, a temporary XML
file is provided as the only argument to these callout scripts. This XML file has the list of
file systems being protected by the Create copy phase.
AppSync continues with normal copy creation if no callout script is found or if the callout
script in not executable. If either of the callout scripts fail (non-zero exit value), copy
creation fails and the Create copy phase ends with an error. If the freeze callout script
Pre-copy script phase
169
Protect file systems
runs successfully, and then copy creation fails due to any storage issue, the thaw callout
script is run before ending the create copy phase with an error.
For a Windows host-plugin, place the callout executable scripts in the %ProgramData%
\ EMC\AppSync\scripts folder and name it
appsync_freeze_filesystem_<service plan name in lower case>.bat
for the freeze callout. Name the thaw callout script:
appsync_thaw_filesystem_<service plan name in lower case>.bat
For example, AppSync runs this script as follows:
C:\ProgramData\EMC\AppSync\scripts
\appsync_freeze_filesystem_bronze.bat, C:\Windows\TEMP
\d575f2e6-7dc4-4389-87c9-491effc57318.xml
Where C:\Windows\TEMP\d575f2e6-7dc4-4389-87c9-491effc57318.xml
file content is in the following form:
<Application type='Filesystem'><sourceVolumePath>F:\</
sourceVolumePath><sourceVolumePath>G:\</sourceVolumePath></
Application>
For a UNIX host-plugin, place the callout executable scripts in the /var/opt/emc/
appsync/scripts folder name as appsync_freeze_filesystem_<service
plan name in lower case> for the freeze callout and
appsync_thaw_filesystem_<service plan name in lower case> for the
thaw callout. Do not use a file name extension such as .pl or .sh. The scripts should be
executable.
AppSync runs this script as follows:
/var/opt/emc/appsync/scripts/
appsync_freeze_filesystem_bronze /tmp/904f510f-47ce-402f-a27ab3a48840a279ybo61k.xml
Where /tmp/904f510f-47ce-402f-a27a-b3a48840a279ybo61k.xml file
content is in the following form: <Application
type='Filesystem'><filesystems><filesystem><name>/FS1</name></
filesystem><filesystem><name>/FS2</name></filesystem></
filesystems></Application>
Post-copy script phase
To perform cleanup or other post-copy steps after creating a copy, specify a post-copy
script and parameters in a service plan's Settings tab.
This operation requires the Service Plan Administrator role in AppSync.
The script runs on successful completion of the Create copy phase. Valid script formats
are .bat, .exe, and .ps1 (PowerShell scripts) for Windows and .sh for UNIX.
AppSync does not support running of PowerShell scripts directly. You usually must wrap
them in a .bat file. The other option is to make the default "Open" on ps1 files C:
\WINDOWS\system32\WindowsPowerShell\v1.0\powershell.exe. When the
PS script runs, you may get an error and you must set an appropriate execution policy.
To run PowerShell commands from scripts:
1. Specify the full pathname to the PowerShell command file in the .bat file:
powershell -command C:\PshellCommands.ps1 <nul
2. Set the PowerShell execution policy so you can run the script. For example, the first
line in the .bat file should look like the following for an unrestricted policy:
powershell -command set-executionpolicy unrestricted <nul
170
EMC AppSync 2.2.3 User and Administration Guide
Protect file systems
3. To ensure correct termination of the PowerShell session, add <nul to the end of the
line that calls your PowerShell script.
For Windows, you can optionally enter credentials to run the script as a specific user. The
script runs as Local System by default. For UNIX, the credentials are mandatory.
For Windows, the default location of the script is %ProgramData%\EMC\AppSync
\scripts\ on the application host.
For UNIX, the File field should have the full path to the scripts to be run.
Exact parameters depend on the script. Parameters with spaces must be enclosed in
double quotes. This phase can be enabled or disabled. This operation requires the
Service Plan Administrator role in AppSync.
Unmount previous copy
The service plan unmounts a previously mounted copy after creating the new copy. The
exception is a copy that was mounted on-demand as opposed to by the service plan; in
this case the on-demand mounted copy is not unmounted.
There are no user settings associated with this phase and it can be enabled or disabled.
Mount copy
The Mount copy phase mounts the copy. This phase can be enabled or disabled.
Field
Description
Mount on
Server
The server on which to mount the copy. Only the nodes of the cluster
or standalone hosts are available for selection.
Mount with
access
Type of access the copy should be mounted with.
Mount path
The Default Mount Path is %SystemDrive%\AppSyncMounts\%
%ProdServerName%%. To specify the value of a Windows
environment variable in the mount path, delimit the variable name
with single percent signs (%). The default path also contains an
AppSync variable (ProdServerName) which is delimited with two
percent signs (%%). The following characters are not valid in the
path:< > : " / | ? * . The mount path could also be Same as Original
Path. However, this option is not available when the mount host is
the same as production host.
Desired SLO
(VMAX 3 only)
Select the desired SLO for the target LUN. If there is a storage group
for the mount host with the desired SLO, the LUN will be added to the
storage group. If it does not exist, AppSync will add it to any storage
group that is masked to the host.
Image access
mode (during
RecoverPoint
mount)
l
Logged Access: Use this mount option if the integrity check
entails the scanning of large areas of the replicated volumes. This
is the only option available when you mount to the production
host.
l
Virtual Access with Roll: Provides nearly instant access to the
copy, but also updates the replicated volume in the background.
When the replicated volumes are at the requested point in time,
the RPA transparently switches to direct replica volume access,
Unmount previous copy
171
Protect file systems
Field
Description
allowing heavy processing. With RP-VMAX and RP-XtremIO, virtual
access with roll is not supported.
Use Dedicated
Storage Group
l
Virtual Access: Provides nearly instant access to the image; it is
not intended for heavy processing. Virtual access with RPVMAX
and RP XtremIO is not supported.
l
Applicable only for physical hosts or virtual machines with direct
iSCSI part of cluster.
l
Checked by default, enabling this option allows AppSync to
enforce a dedicated VMAX or VNX storage group for a mount. (A
dedicated VMAX or VNX storage group contains the selected
mount host only). The mount will fail if you are mounting to a
node of a cluster that is in a storage group shared with the other
nodes.
Note
Use this option to mount the copy to a node for copy validation or
backup to tape. In this scenario, you will need two storage
groups. One storage group is dedicated to the passive node being
used as a mount host and the other storage group is for the
remainder of the nodes in the cluster. Both storage groups
contain the shared storage for the cluster.
l
If unchecked, AppSync does not enforce the use of a dedicated
storage group for a mount.
Override mount settings in a service plan
If there are multiple file systems that are subscribed to the same plan, you can select
different mount settings for each file system, overriding the generic mount settings.
Before you begin
This operation requires the Data Administrator role in AppSync.
Procedure
1. Browse to Service Plans > FileSystems and click one of the plans from the list.
2. From the Settings tab, select the Mount copy phase.
3. On the right pane, select the Mount Copy Overrides tab.
The list of file systems includes all file systems subscribed to this plan. The mount
settings display the default settings.
4. Select the file system whose settings you want to override and click Set Overrides.
Press and hold the Shift or Ctrl keys to select multiple file systems.
5. On the Mount Copy Overrides dialog box, select options only for those mount settings
that you want to override.
For example, if you want to mount a copy to the production host, you would select Use
new signature from the Mount Signature drop-down.
Fields that do not have a selection retain their default settings.
172
EMC AppSync 2.2.3 User and Administration Guide
Protect file systems
6. Click OK.
A pencil icon appears in the first column of the file system's row whose default mount
settings you changed.
7. To revert to default settings, select the file systems and click Use Default Settings.
Post-mount script
Specify a post-mount script and parameters from the Post-mount script option in the
Settings tab of a service plan.
The script runs on successful completion of the mount copy or mount with recovery
phase. This script is typically used for backup.
From the Server list, select the server on which to run the script. You can optionally run it
on a registered host other than the mount host, and enter credentials to run the script as
a specific user.
The default location of the script is %ProgramData%\EMC\AppSync\scripts\ on
the application host.
Exact parameters depend on your script. Parameters with spaces must be enclosed in
double quotes.
This phase can be enabled or disabled. This operation requires the Service Plan
Administrator role in AppSync.
Mounting a copy with the File System Mount wizard
Use the File System Mount wizard to create any point-in-time mount for RecoverPoint
copies.
Before you begin
This task requires the Data Administrator role in AppSync.
For UNIX, the File field should have a full path to the scripts to be run. You need user
credentials to run the script.
Follow these steps:
Procedure
1. On the AppSync console, go to File systems > Copies to display available copies.
2. Select a copy to mount, and then click Mount.
The Select Copy to Mount page of the File System Mount wizard launches.
You can select any copy to be mounted. By default the copy that you selected before
selecting Mount is highlighted, however other copy instances also appear.
3. Select the copy to mount, and click Next.
The Specify Mount Settings page launches where you can select the Mount on Host
(mount host drop-down), and a mount path location either To original path, or Mount
to alternate path (drop-down). The mount path is the location where the copy is
mounted on the mount host. By default AppSync displays the path of the mount host
you selected. You can also edit and mount the copy to a user-defined location.
You can also select Mount Options including Mount with access (read-only or read/
write), and Image access mode (drop-down). Image access mode allows you to mount
the file system with proper access mode in case the selected copy is a RecoverPoint
bookmark.
For VMAX 3 arrays, you are presented with an SLO drop-down list. You can select the
Service Level Objective (SLO) for the mount copy. If there is a storage group for the
Post-mount script
173
Protect file systems
mount host with the desired SLO. AppSync adds the LUN to the storage group. If this
storage group does not exist, AppSync adds the LUN to any storage group that is
masked to the host. If a storage group is configured to pick target devices, AppSync
removes the devices from the storage group at the time of mount and adds them to
the storage group for the mount host. The devices are added to the original storage
group when the copy is expired.
4. Type mount location information and mount options, and click Next.
The Summary page launches.
5. Review the mount information and if it is correct click Finish.
The Results page launches with mount result information.
Changing the mount point for an affected file system
Follow this procedure to manually change the mount point for an affected file system.
Assume VG1 is the source volume group.
Procedure
1. Get the list of LVs using the lsvg -l VG1 command, and check which file systems
show mount point on /tmp/EMCAppsync ** directory.
2. Run chfs -m <Original MT Pt> /tmp/EMCAppsync6922/vg1_logs
command where <Original Mt Pt> is the mount point where the file system was
originally mounted.
3. Run fsck on the source Logical Volume fsck -y /dev/fslv01.
4. Run mount command using the log logical volume and make sure that the source has
been mounted successfully mount -v jfs2 -o rw,log=/dev/
loglv00 /dev/fslv01 <Orig Mt Pt>
Unmounting a file system copy
When you select a copy to unmount, other copies that were mounted along with the
selected copy will also be unmounted.
Before you begin
This operation requires the Data Administrator role in AppSync.
You can unmount a copy only from a list of copies made for a file system.
Procedure
1. Navigate to the Copies page from the Copy Management or Service Plan pages:
l
Copy Management > FileSystems > select the server which hosts the file system you
want to unmount, then select the file system with the copy to unmount.
l
Service Plans > File systems> select a service plan, then select the Copies tab.
2. From the list of copies, select the copy and click Unmount from the button in the lower
part of the page.
The Unmount Confirmation dialog displays all the copies of other file systems that
were mounted along with the selected copy to be unmounted.
3. Click Yes to confirm the unmount of all the copies shown in the dialog.
The Unmount page displays the progress of the unmount operation. All copies
associated with the selected copy will be unmounted.
174
EMC AppSync 2.2.3 User and Administration Guide
Protect file systems
Override mount settings in a service plan
If there are multiple file systems that are subscribed to the same plan, you can select
different mount settings for each file system, overriding the generic mount settings.
Before you begin
This operation requires the Data Administrator role in AppSync.
Procedure
1. Browse to Service Plans > FileSystems and click one of the plans from the list.
2. From the Settings tab, select the Mount copy phase.
3. On the right pane, select the Mount Copy Overrides tab.
The list of file systems includes all file systems subscribed to this plan. The mount
settings display the default settings.
4. Select the file system whose settings you want to override and click Set Overrides.
Press and hold the Shift or Ctrl keys to select multiple file systems.
5. On the Mount Copy Overrides dialog box, select options only for those mount settings
that you want to override.
For example, if you want to mount a copy to the production host, you would select Use
new signature from the Mount Signature drop-down.
Fields that do not have a selection retain their default settings.
6. Click OK.
A pencil icon appears in the first column of the file system's row whose default mount
settings you changed.
7. To revert to default settings, select the file systems and click Use Default Settings.
Restoring a file system
Use the File System Restore wizard to restore an existing file system copy.
Procedure
1. On the AppSync console go to Copy Management > File systems.
The file system Hosts page launches displaying available servers.
2. Select the server which has the file systems for restore, to launch the file systems
page.
3. Select the file system that you want to restore (Select All option appears in lower left
of page), and then click Restore from the Recover menu in the lower left area of the
file system page.
This action launches the File System Restore wizard. All protected copies are listed in
the wizard. You can only select one file system copy at a time for restore. The copy
date, file system name, server name, service plan, and copy type appear for each
copy.
4. Select the file system copy for restore and click Next.
The Restore Warnings page launches. The Restore Warnings page lists the file system
which belongs to the same consistency group or volume group where the copy is
restored. If this file system is protected as well as the selected file system copy, this
file system is overwritten.
Override mount settings in a service plan
175
Protect file systems
The restore warning page also lists any application that is installed in the file system
that is being restored.
5. Read the warnings and click the "I have read warnings" checkbox and click Next to
continue with the restore.
This action launches the Summary page of the wizard.
6. If the Summary page looks correct, click Finish.
The Results page loads where you can view the restore results.
Restore a file system copy manually on XtremIO
Automated restore of an XtremIO 4.0 (minimum version) copy with AppSync 2.2.2 and
higher is supported. This topic shows you how to manually restore a copy on an XtremIO
version lower than 4.0 with AppSync 2.2 through 2.2.1.
Before you begin
Refer to Restoring a file system copy on page 175 for automated restore.
To manually achieve a granular restore of files and folders in a file system, mount a file
system copy and use selective file/folder copy, you need the following configuration:
l
Host file system that resides on XtremIO storage
l
AppSync service plan that replicates the file system
Procedure
1. On the AppSync console, select Service Plans > Filesystem and the wanted service
plan, for example Bronze.
The Copies page displays existing copies.
2. Disable any service plans scheduled to run, which includes the file system that you
plan to restore.
This action prevents a copy from being made during the restore procedure.
3. To mount the copy that you want to restore to production host, click mount to default
location.
Mount the copy with read/write access.
4. Perform one of the following steps depending on the operating system:
a. For Windows: Record the NAA Identifier for the production volume and the
mounted snapshot, and then remove the drive letter of the production device on
the production host.
b. Linux/AIX: Unmount the file system.
5. Assign the drive letter of mounted snapshot copy on production host (Windows) or
remount file system (Linux/AIX).
6. Select Remove for the mounted copy.
This action removes the copy from the AppSync database without changing the
snapshot on the mount host or the XtremIO storage array. The system displays a
message that discusses this action.
7. If the production host is a virtual machine, use the NAA Identifier that you previously
recorded to remove the old production LUN from the host inventory.
This action removes the volume from the host (host device rescan might be
necessary). No more action is required on the production host.
176
EMC AppSync 2.2.3 User and Administration Guide
Protect file systems
Figure 26 Virtual machine: Remove old production LUN
8. Remove the old production LUN from the initiator group.
9. Move the mounted snapshot from the /AppSyncSnapshots folder to the folder of
the production LUN that you are replacing.
Move the snapshot that is mounted (the snapshot that is currently masked to the
production host initiator group or the ESX initiator group).
After you finish, you will see the production LUN and the snapshot.
Figure 27 Production LUN and snapshot
One is a snapshot. The other is a volume.
Restore a file system copy manually on XtremIO
177
Protect file systems
Figure 28 One snapshot and one volume
10. Remove the volume that you want to replace.
The snapshot becomes a volume.
Figure 29 Snapshot becomes volume
11. Rename the snapshot (now a volume) to the name of the volume that you replaced,
and then run the file system service plan to create a snapshot of the "restored" file
system.
178
EMC AppSync 2.2.3 User and Administration Guide
CHAPTER 9
Protect VMware Datacenters
This chapter includes the following topics:
l
l
l
l
l
l
Configuration prerequisites ................................................................................ 180
Discovering datacenters ..................................................................................... 183
Considerations when mounting a VMFS copy ......................................................192
Restoring a datastore from a copy....................................................................... 194
Restoring a virtual machine from a copy.............................................................. 198
File or folder restore with VMFS or NFS datastores............................................... 201
Protect VMware Datacenters
179
Protect VMware Datacenters
Configuration prerequisites
AppSync can create, mount, and restore copies in VMware vStorage VMFS and NFS data
store configurations. Configuration prerequisites are required to integrate AppSync with
VMware vStorage VMFS protection. Configure RecoverPoint and VMware according to the
product documentation.
VMware configuration prerequisites
l
VMware vCenter Server must be used in the environment.
l
AppSync supports VMware’s use of VSS with VM snapshots when a supported
version of vSphere is installed and the VMware Tools facility is present on the virtual
machine on the VMFS you are replicating. Refer to VMware documentation for
information on the VSS-related characteristics in an AppSync copy. Contact VMware
regarding considerations that are related to VSS in this configuration.
l
When there is a configuration change in the vCenter Server, perform a discovery of
data centers in the vCenter Server from the AppSync console before you protect a
data store. Ensure that the VMFS UUID is unique in the virtual center inventory across
all data centers.
Note
You can also create and manage copies of VMware data stores with the VSI for VMware
vSphere web Client. The plug-in is available as a separate download from the AppSync
Online Support page at support.EMC.com.
RecoverPoint configuration prerequisites
l
Configure RecoverPoint protection (CDP/CRR/CLR) for the production LUNs before
deploying AppSync. Refer to RecoverPoint documentation to create consistency
groups and define replication sets.
l
In an ESX cluster, target LUNs should be made visible to all the ESX hosts in the
cluster.
l
The AppSync server must connect to the RPA through the network.
VMware vStorage VMFS requirements
Some considerations apply when AppSync is introduced into a VMware environment for
protecting VMware data stores.
All VMware specific operations occur through the VMware vCenter Server.
AppSync can be configured to require vCenter Server login credentials to allow protection
of a certain VMFS for security purposes. Unless you instruct AppSync to omit this feature,
AppSync takes a VMware Snapshot for each virtual machine that is online and residing
on the VMFS before protection. This action ensures operating system consistency for the
resulting replica. The following user roles for a Virtual Center ESX cluster are allowed with
AppSync:
180
l
Administrator
l
VM power user
l
VM user
l
Resource pool Administrator
l
VMware consolidated backup user
EMC AppSync 2.2.3 User and Administration Guide
Protect VMware Datacenters
l
Data store consumer
l
Network Administrator
The following Edit Role screen capture shows privileges that you need to select for roles.
VMware vStorage VMFS requirements
181
Protect VMware Datacenters
Figure 30 Privileges needed for VC roles for AppSync
182
EMC AppSync 2.2.3 User and Administration Guide
Protect VMware Datacenters
AppSync supports VMware’s use of VSS with VM snapshots when a supported version of
vSphere is installed and VMware Tools are present on the virtual machine on the VMFS
you are protecting. Refer to VMware documentation for use of the VSS-related
characteristics in the AppSync copy and contact VMware regarding considerations that
are related to VSS in this configuration.
If virtual machines in the data store have RDMs or iSCSI LUNs visible to them, the
resulting copy does not contain those LUNs. If the virtual machine has virtual disks other
than the boot drive located in other data stores, it is possible to capture these disks by
configuring the service plan to include virtual machine disks.
Discovering datacenters
To keep AppSync up to date, discover datacenters on the VMware vCenter Server when
there is a change in the configuration of the vCenter Server.
Before you begin
l
This operation requires the Data Administrator role in AppSync.
l
At least one vCenter server must be added to AppSync.
Procedure
1. Browse to Copy Management > VMware Datacenters.
2. On the VMware Datacenters page, click Discover Datacenters > On Virtual Center and
select a vCenter server to discover its datacenters.
Optionally, you can also add a vCenter server by clicking Add vCenter Server. See Add
a vCenter Server on page 184.
List of datacenters
The top level of the VMware Datacenters page shows all datacenters registered with
AppSync.
Column
Protection status
of datacenter
Description
l
Green: Latest copies of all datastores on the datacenter
protected successfully
l
Yellow: One or more of the latest datastore copies on the
datacenter completed with errors
l
Red: One or more of the latest datastore copies on the
datacenter failed to complete
l
"i" symbol: One or more datastores on the datacenter are either
not subscribed to service plans or do not have copies
associated with them
Name
Name of the datacenter on the vCenter server.
vCenter Server
Name of the vCenter server that hosts the datacenter.
Last Discovery
Time when a discovery was last performed on the vCenter server.
Alert Recipients
List of email aliases to receive email alerts.
Clicking on a datacenter name shows the datastores.
Discovering datacenters
183
Protect VMware Datacenters
Adding a VMware vCenter Server
Add a VMware vCenter Server to AppSync when a virtual machine is used as a mount host
for a VNX Snapshot copy.
Before you begin
l
This operation requires the Resource Administrator role in AppSync.
l
Ensure that you know the credentials of an account with Administrator privileges on
the vCenter Server.
Procedure
1. Select Settings > VMware vCenter Servers.
2. Click Add.
3. Type the vCenter Server name.
4. Type the credentials for an account that has Administrator privileges on the vCenter
Server.
List of VMware datastores
The list contains VMware datastores that have been discovered and stored in the
AppSync database.
Clicking on the datastore name displays the copies of the datastore.
The Service Plan column shows the plans that the datastore is subscribed to. Other
details include the type of datastore (VMFS or NFS), and name of the ESX server.
Protect a VMware datastore
Protect a VMware datastore by subscribing it to an AppSync VMware service plan.
AppSync's protection mechanism for datastores is by means of service plans. You
subscribe a datastore to a service plan and run the service plan immediately, or schedule
the service plan to run at a later time.
l
Choose Subscribe to Plan and Run when you want to protect selected datastores
immediately. The service plan is executed for the datastores alone.
l
Choose Subscribe to Plan when you want to schedule the protection for later.
Protection for datastores that are part of the service plan are executed at the
scheduled time.
l
Choose an appropriate service plan from Create a copy using in the datastore Copies
page.
l
Choose Run from the VMware Datacenters Service Plan page to run the whole plan
immediately.
Subscribing VMware datastores to a service plan
The Subscribe to Plan operation schedules the protection for later. Protection for all
datastores that are part of the service plan are executed at the scheduled time.
Before you begin
This operation requires the Data Administrator role in AppSync.
184
EMC AppSync 2.2.3 User and Administration Guide
Protect VMware Datacenters
Procedure
1. Navigate to Copy Management > VMware Datacenters.
2. Click a datacenter to display its datastores.
3. From this list, select the datastore to protect.
Select multiple datastores by holding down the Shift or Ctrl keys on your keyboard.
4. From the Protect list, select the appropriate service plan from Subscribe to Plan.
The selected plan appears in the Service Plan column for the datastore.
Protecting VMware datastores immediately
The Subscribe to Plan and Run operation adds datastores to an existing service plan and
runs the service plan immediately for the selected datastores only.
Before you begin
This operation requires the Data Administrator role in AppSync.
Procedure
1. Navigate to Copy Management > VMware Datacenters.
2. Click a datacenter to display its datastores.
3. From this list, select a datastore to protect.
Select multiple datastores by holding down the Shift or Ctrl keys on your keyboard.
4. From the Protect list, select the appropriate service plan from Subscribe to Plan and
Run.
The Subscribe to Plan and Run dialog appears displaying the progress through the
different phases.
List of protected virtual machines
The list contains virtual machines belonging to datastores that are protected as part of a
service plan run.
Click on the virtual machine name to display copies of the virtual machine. To perform a
restore operation, select a virtual machine and click Restore.
Other details include the OS platform on the virtual machine, the version of the virtual
machine, the ESX host on which the virtual machine resides, as well as the path to the
virtual machine file. In the path, the name of the datastore that the virtual machine
resides on is within the [ ] parentheses.
Unsubscribing datastores from a service plan
You can unsubscribe datastores from a service plan or from all service plans that they are
subscribed to.
Before you begin
This operation requires the Data Administrator role in AppSync.
Procedure
1. Navigate to Copy Management > VMware Datacenters.
2. Click a datacenter to display its datastores.
3. From this list, select the datastore to remove from a service plan.
You can select multiple datastores if they are subscribed to the same service plan.
Select multiple datastores by holding down the Shift or Ctrl keys on your keyboard.
Protect a VMware datastore
185
Protect VMware Datacenters
4. From the Protect list, select a service plan from Unsubscribe from Plan, or select All to
remove the datastore(s) from all plans.
The service plan name is removed from the Service Plan column for the datastore(s).
VMware snapshots
When the VM consistency option is selected, AppSync creates snapshots of all the virtual
machines that are in powered on state while the datastore is being replicated.
AppSync creates a Quiesced snapshot of the virtual machines that are in powered on
state. VMware Tools is used to quiesce the file system in the virtual machine. Quiescing a
file system is a process of bringing the on-disk data of a physical or virtual computer into
a state suitable for backups. This process might include operations such as flushing dirty
buffers from the operating system's in-memory cache to disk, or other higher-level
application-specific tasks. If the VM consistency option is not set, AppSync skips the
process of creating the virtual machine snapshots.
Viewing datastore copies
The copies are named with the time at which they were made.
Before you begin
This operation requires the Data Administrator role in AppSync.
Procedure
1. Navigate to Copy Management > VMware Datacenters.
2. Select a datacenter, then a datastore to view its copies.
You can see other copy details such as copy type, mount status, the VM consistency
status and the service plan associated with the copy.
3. Select a copy to see more details in the Details pane.
l
Virtual Machines tab: lists the virtual machines that are part of the selected
datastore copy.
l
Virtual Disks tab: lists the virtual disks that are part of the selected datastore copy.
l
Events tab: lists the events that occurred when the datastore copy was created.
Viewing virtual machine copies
View the list of copies for a virtual machine based on the time they were made.
Before you begin
This operation requires the Data Administrator role in AppSync.
Procedure
1. Navigate to Copy Management > VMware Datacenter.
2. Select a datacenter, then the Protected Virtual Machines tab.
3. Select a virtual machine from the list to view its copies.
You can see copy details such as the size of the copy, storage on the disk, mode of
the disk, and so on.
4. Select a copy to see additional details in the Details pane.
l
186
Virtual Disks tab: lists the virtual disks that are part of the selected virtual machine
copy.
EMC AppSync 2.2.3 User and Administration Guide
Protect VMware Datacenters
l
Events tab: lists the events that occurred when the virtual machine copy was
created.
Creating a datastore copy from the Copies page
Create a copy of a datastore by subscribing it to an AppSync VMware service plan from
the Copies page.
Before you begin
This operation requires the Data Administrator role in AppSync.
Procedure
1. Navigate to Copy Management > VMware Datacenters.
2. Click a datacenter to display its datastores.
3. From this list, click the datastore to view its copies.
4. From the Create a copy using list, select the appropriate service plan.
The service plan runs immediately for the datastore.
Expiring a datastore copy on demand
Expiring a copy removes it from the AppSync database and can free up storage,
depending on the replication technology and the copy state.
Before you begin
This operation requires the Data Administrator role in AppSync.
Expiring a copy that was made with RecoverPoint does not remove the corresponding
bookmark from RecoverPoint itself.
Procedure
1. Select Copy Management > VMware Datacenters.
2. Click a datacenter to display its datastores.
3. Click the datastore whose copies you want to expire.
4. From the Copies page, select one or more copies to expire.
You can also perform this action from the Service Plan's Copies tab.
5. Select Expire.
6. Verify that you want to expire the copy you selected and any associated copies listed
and confirm.
Service plan schedule
The schedule of a service plan is set in the Plan Startup phase.
The Startup Type (scheduled or on demand) determines whether the plan is run
manually, or configured to run on a schedule. Options for scheduling when a service plan
starts are:
l
l
Specify a recovery point objective (RPO)
n
Set an RPO of 30 minutes or 1, 2, 3, 4, 6, 8, 12, or 24 hours
n
Minutes after the hour are set in 5 minute intervals
n
Default RPO is 24 hours
Run every day at certain times
Protect a VMware datastore
187
Protect VMware Datacenters
l
l
n
Select up to two different times during the day
n
Minutes after the hour is in 5 minute intervals
n
There is no default selected
Run at a certain time on selected days of the week
n
One or more days of the week (up to all seven days) can be selected
n
There is no default day of the week selected. Default time of day is 12:00 AM.
Run at a certain time on selected days of the month
n
Select one or more days of the month (up to all days)
n
Select one time of day. Available times are at 15 minute intervals.
n
Default is the first day of the month
Overriding service plan schedules
You can set different schedules for individual datastores subscribed to a service plan,
overriding the generic recurrence setting.
Before you begin
This operation requires the Data Administrator role in AppSync.
You can override only the settings of the recurrence type already selected for the service
plan.
Procedure
1. Navigate to Service Plans and select one of the plans from the list.
2. From the Settings tab, select the Plan Startup phase.
3. In the Plan Startup Defaults pane on the right, note the Recurrence Type selected for
the plan.
A recurrence type can be set only if Scheduled is selected as the Startup Type.
4. Click the Plan Startup Overrides tab.
You can see the list of all datastores subscribed to the plan.
5. Select one or more datastores and click Override Schedule.
6. In the Override Schedule dialog, set the schedule based on your requirement and
click OK.
For example, if the default recurrence type is On specified days of the month, and the
rule setting is to Run at 12:00 AM on the 1st day of every month, you can override the
time and the day for individual datastores.
A Pencil icon indicates that default settings have been overidden.
Application discovery
Before creating the copy, AppSync performs discovery on the selected datastores and
updates the AppSync database if there is any change in configuration of the vCenter
server.
There are no user settings associated with this phase and it cannot be disabled.
188
EMC AppSync 2.2.3 User and Administration Guide
Protect VMware Datacenters
Application mapping
After discovering the application, AppSync maps it to array storage, and protection
services such as RecoverPoint.
There are no user settings associated with this phase and it cannot be disabled.
Create copy
The Create Copy phase creates a copy based on the preferred storage type specified by
the user.
This phase specifies the type of datastore copy to make, and the storage settings for the
copies. The copy phase creates a local copy, remote copy, or a local and remote copy
based on whether you have chosen the bronze, silver, or gold service plan.
Review Overview: Service Plan on page 10 for more service plan copy information.
Datastore copy options
Select the copy type, the virtual machines to ignore for snaps, storage preferences, and
the number of snapshot copies to retain.
l
Copy Consistency
VM Consistent creates a copy of the datastores in the service plan including
running programs, processes, and even windows that were open at the time of the
snapshot. Maximum Simultaneous VM Snapshots is the number of
simultaneous snapshots of all VMs present. The default value is four snapshots.
Crash Consistent creates a copy of the datastores in the service plan. Crash
consistent copies have everything except data from the memory at the time of
taking the snapshot.
Ignore VM Snapshots for VMs link allows you to select virtual machines
from the datastores added to the service plan. The selected VMs are ignored while
taking VMware snapshots during the service plan run.
Include Virtual Machine Disk includes all the datastores that are
associated with the virtual machines running on the datastores being protected.
For example, Datastore DS1 is subscribed to the service plan. Virtual Machine
VM1 which is a part of DS1 has virtual disks in Datastores DS2 and DS3. When the
service plan runs, datastores DS2 and DS3 are protected along with DS1.
However, datastores DS2 and DS3 are not subscribed to the service plan.
l
Storage Ordered Preference- the preferred order of storage technology to use while
creating copies. Use the Move Up and Move Down buttons. Copies are made using
the first technology preference when possible. If conditions are such that the first
technology can no longer be used, then any remaining copies are processed in the
next preference instead. For example, if the first preference was a bookmark but not
all the application data in the service plan could be mapped to RecoverPoint, then
AppSync uses VNX snapshots instead.
Note
A single service plan can contain a mix of VNX block, VNX file, and RecoverPoint
replication objects. For example, if you have a Bronze service plan for VMware, the
datastores can be a mix of RecoverPoint, VNX file, and VNX block replication.
l
Expiration - the maximum desired number of RecoverPoint, VNX and VNX File
Snapshot copies that can exist simultaneously.
Protect a VMware datastore
189
Protect VMware Datacenters
Automatic expiration of RecoverPoint, XtremIO, VNX and VNX File, VMAX, and ViPR
Controller snapshot copies
The automatic expiration value in a service plan's Create Copy phase specifies the
maximum number of snapshot copies that can exist simultaneously.
When the "Always keep x copies" value is reached, older copies are expired to free
storage for the next copy in the rotation. Failed copies are not counted. AppSync does not
expire the oldest copy until its replacement has been successfully created. For instance,
if the number of copies to keep is 3, AppSync does not expire the oldest copy until the
fourth copy is created successfully.
This setting is independent of the VNX pool policy settings in Unisphere for automatic
deletion of oldest snapshots. The service plan administrator should work with the storage
administrator to ensure that the VNX pool policy settings enable the support of the
specified number of snapshot copies for the application residing in that pool.
AppSync does not expire copies under the following circumstances:
l
Mounted copies are not expired.
l
A copy that contains the only replica of a datastore is not expired.
Include RecoverPoint copies in expiration rotation policy: Check this option to include
RecoverPoint copies when calculating rotations.
Note
If this option is not selected, then RecoverPoint copies accumulates, and will remain until
the bookmarks expire from the RecoverPoint appliance.
Unmount previous copy
The service plan unmounts a previously mounted copy after creating the new copy. The
exception is a copy that was mounted on-demand as opposed to by the service plan. The
on-demand mounted copy is not unmounted.
There are no user settings associated with this phase and it can be enabled or disabled.
Mount copy
The Mount copy phase mounts all the datastore copies created by that service plan run.
The Mount Copy Defaults settings for the copy to mount depends on the service plan.
Other mount settings determine the mount host, access mode and mount signature.
This phase can either be enabled or disabled.
General Settings:
l
Mount on host: lists all the ESX servers discovered on the registered vCenter servers.
l
Mount Signature: lists Use original signature and Use new signature to select from.
When Use new signature is selected, AppSync resignatures the VMFS volume on
mount. Applicable only for VMware VMFS datastores.
l
Cluster Mount: Select Yes or No .
RecoverPoint Settings:
190
l
Image access mode (during RecoverPoint mount):
l
Logged Access:
Use this mount option if the integrity check entails the scanning of large areas of the
replicated volumes.
l
Virtual Access with Roll:
Provides nearly instant access to the copy, but also updates the replicated volume in
the background. When the replicated volumes are at the requested point in time, the
EMC AppSync 2.2.3 User and Administration Guide
Protect VMware Datacenters
RPA transparently switches to direct replica volume access, allowing heavy
processing.
l
Virtual Access:
Provides nearly instant access to the image; it is not intended for heavy processing.
l
Desired Service Level Objective (SLO): Specifies the required VMAX 3 Service Level
Objectives. SLO defines the service time operating range of a storage group.
VNX File Settings: This option is available only for VMware VNXFile datastores.
l
Mount Copy with access: Select the type of access the copy should be mounted with Read-only or Read-Write.
Overriding mount settings in a service plan
If there are multiple VMware datastores subscribed to the same plan, you can select
different mount settings for each datastore, overriding the generic mount settings.
Before you begin
This operation requires the Data Administrator role in AppSync.
Procedure
1. Navigate to Service Plans > VMware Datacenters and click one of the plans from the
list.
2. From the Settings tab, select the Mount copy phase.
3. On the right pane, select the Mount Copy Overrides tab.
The list of datacenters includes all vCenter datacenters whose datastores are
subscribed to this plan. The mount settings display the default settings. Additionally,
for VMAX v3 Datastores, SLO Service Level Objective appears as another option.
4. Select the datastore whose settings you want to override and click Set Overrides.
Select multiple datastores by holding down the Shift or Ctrl keys on your keyboard.
5. On the Mount Copy Overrides dialog, select options only for those mount settings that
you wish to override.
For example, if you want to mount a copy to the production host, you would select Use
new signature from the Mount Signature drop-down.
Fields that do not have a selection retain their default settings.
6. Click OK.
A pencil icon appears in the first column of the datastore's row whose default mount
settings you changed.
7. To revert back to default settings, select the datastore(s) and click Use Default
Settings.
Unmount copy
The final phase in the service plan unmounts the copy.
This phase is disabled if the Unmount previous copy phase is enabled. There are no user
settings associated with this phase.
Protect a VMware datastore
191
Protect VMware Datacenters
Considerations when mounting a VMFS copy
When you mount a VMFS copy to an alternate ESX Server, AppSync performs all tasks
necessary to make the VMFS visible to the ESX Server.
l
After these tasks complete, further administration tasks such as restarting the virtual
machines and the applications must be completed by scripts or manual intervention.
l
For datastore and virtual disk mounts on ESXi 5.x and RecoverPoint 4.0
environments, disable hardware acceleration to ensure successful virtual access type
mounts. For more details, refer VMware Knowledge Base article 2006858.
Mounting a datastore copy on-demand
You can initiate an on-demand mount of a datastore copy from the datastore's Copies
page, service plan's Copies tab or from a datacenter's Datastore page.
Before you begin
This operation requires the Data Administrator role in AppSync.
Procedure
1. From the Recover popup button, select Mount a Copy in the Datastore or Copies page
or Service Plan Copies tab.
2. Use the Copies or Service Plan filters to select the appropriate copy to mount.
The copies list is refreshed based on the filters selected.
3. Select the copy to mount.
For a RecoverPoint copy, you also have the option to select a bookmark based on a
specific time. However, there should be a copy available in AppSync prior to the time
you select.
Click Select a point in time to select a copy with a specific time stamp and select the
location of the copy (remote or local). Select remote to mount remote copy or local to
mount local copy. The time shown here is the AppSync console's time. If the console
is in a different time zone from the RecoverPoint Appliance (RPA), specify the time as
per the server's time zone to mount the copy.
4. In the Mount Additional Copies page, select one or more additional copies to mount.
The copies listed here are of other datastores that were protected at the same time
and on the same datacenter as the copy you selected in the previous step.
5. On the Select Mount Settings page:
a. From the Mount on Host list, select the host on which to mount the copy.
All ESX Servers under the vCenter Server registered with AppSync are listed as
possible mount host candidates. If a selected ESX is part of an ESX cluster,
another field, Clutser Mount, is displayed, which can be set to Yes or No. If the ESX
Server is part of a cluster you also have the option of mounting to all nodes of the
cluster or just to the mount host you choose.
b. From the Mount Signature list, select from one of these options: Use new
signature, Use original signature. Applicable only for VMware VMFS datastores.
c. In case of a RecoverPoint copy, from the Image Access Options list, select from
one of these options:
192
EMC AppSync 2.2.3 User and Administration Guide
Protect VMware Datacenters
l
Logged access: Use this mount option if the integrity check entails the
scanning of large areas of the replicated volumes. Logged access is the only
option available when you mount to the production host.
l
Virtual access with roll: Provides nearly instant access to the copy, but also
updates the replicated volume in the background. When the replicated
volumes are at the requested point in time, the RPA transparently switches to
direct replica volume access, allowing heavy processing. With RP VMAX, and RP
XtremIO, virtual access with roll is not supported.
l
Virtual access: Provides nearly instant access to the image. Virtual access is
not intended for heavy processing. Virtual access with RP VMAX and RP
XtremIO is not supported.
d. From the Mount copy with access drop-down list, select the type of access the
copy should be mounted with: Read-only or Read-write. Applicable only for
VMware VNXFile datastores.
e. For VMAX 3 arrays, you are presented with an SLO drop-down menu. You can select
the desired Service Level Objective (SLO) for the mount copy.
If there is a storage group for the mount host with the desired SLO. AppSync will
add the LUN to the storage group. If this storage group does not exist, AppSync
adds the LUN to any storage group that is masked to the host. If a storage group is
configured to pick target devices, AppSync removes the devices from the storage
group at the time of mount and adds them to the storage group for the mount host.
The devices will be added to the original storage group when the copy is expired.
An example of the Desired SLO menu follows:
Figure 31 For VMAX 3 select Desired SLO
6. From the Summary page, review the copy and mount settings that you chose in the
previous pages and click Finish to mount the copy.
7. In the Results page, click Details link to see the progress of the different phases that
are part of mounting a copy.
Mounting a datastore copy on-demand
193
Protect VMware Datacenters
The last phase completed is displayed at the bottom of the list of phases.
Unmounting a VMware datastore copy
When you select a copy to unmount, other copies that were mounted along with the
selected copy are unmounted.
Before you begin
This operation requires the Data Administrator role in AppSync.
You can unmount a copy only from a list of copies made for a datastore.
Procedure
1. Navigate to the Copies page from the Protection or Service Plan pages:
l
Copy Management > VMware Datacenters > select the VMware datacenter that
hosts the datastore, then select the datastore with the copy to unmount.
l
Service Plans > VMware Datacenters> select a service plan, then select the Copies
tab.
2. From the list of copies, select the datastore copy and click Unmount from the button
below.
The Unmount Confirmation dialog displays all the copies of other datastores that were
mounted along with the selected copy to be unmounted.
3. Click Yes to confirm the unmount of all the copies shown in the dialog.
The Unmount window displays the progress of the unmount operation. All copies that
were mounted along with the selected copy will be unmounted.
Restoring a datastore from a copy
You can perform a restore of a datastore copy from the datastore's Copies page, service
plan's Copies tab or from a datacenter's Datastore page.
Before you begin
l
This operation requires the Data Administrator role in AppSync.
l
Prior to restoring a datastore, it is recommended that you power off the VMs in the
datastore.
Procedure
1. Select Recover > Restore from Datastore or Copies page or Service Plan Copies tab.
The Datastore Restore wizard launches.
2. Select the copy to restore.
Use the Copies or Service Plan filters to select the appropriate copy to restore. The
copies list is refreshed based on the filters selected.
For a RecoverPoint copy, you also have the option to select a bookmark based on a
specific time. However, there should be a copy available in AppSync prior to the time
you select.
Click Select a point in time to select a copy with a specific time stamp. The time
shown here is the console's time. If the console is in a different time zone from the
RPA, specify the time as per the server's time zone to restore the copy.
3. Click Next.
194
EMC AppSync 2.2.3 User and Administration Guide
Protect VMware Datacenters
If the selected copy has affected entities, the Restore Warnings page is displayed.
4. Read the warning messages for the affected datastores. Select the checkbox to
indicate your agreement to restore other entities along with the selected copy.
You can manually unmount the datastores that will be overwritten prior to restore.
Only RecoverPoint copies have affected entities.
5. In the Virtual Machine Operations step, select the appropriate actions that you want
AppSync to perform before and after restore. See Virtual Machine Operations on page
195 for details.
6. In the Summary page, review the settings that you selected in the previous pages and
click Finish to perform the restore.
7. In the Results page, click View Details to see progress of the different phases that are
part of restoring a copy.
The last phase completed is displayed at the bottom of the list.
Virtual Machine Operations during restore
AppSync can perform operations on the virtual machines associated with the datastores
selected for restore.
Table 21 Virtual Machine operations
Virtual Machine
Operation
Description
VMs present at start
of restore
Power down VMs at start of restore: If the virtual machines are present
Perform VM
operations after
restore
at the start of restore, AppSync shuts them down prior to beginning the
restore operation. This is important for a successful restore.
l
Return VMs back to state found at start of restore: After restore,
AppSync powers on the virtual machines and returns them to the
same state they were at prior to restore.
l
Register all virtual machines: After restore, AppSync registers all
virtual machines to the vCenter inventory.
l
Register and power up all virtual machines: After restore,
AppSync powers on the virtual machines and registers them to the
vCenter inventory.
You can select only one of these options.
VMs not present at
start of restore.
Perform VM
operations after
restore.
l
Register all virtual machines: After restore, AppSync registers all
virtual machines to the vCenter inventory.
l
Register and power up all virtual machines: AppSync powers on
the virtual machines and registers them to the vCenter inventory.
Datastore affected entities during restore
When you restore a datastore, AppSync calculates affected entities for other datastores
that share the same storage.
An affected entity is data that resides on your ESX server that unintentionally becomes
part of a replica because of its proximity to the data you intend to protect. You can
prevent affected entity situations by properly planning your data layout.
Virtual Machine Operations during restore
195
Protect VMware Datacenters
In case of RecoverPoint or ViPR Controller, the granularity is at the consistency group (CG)
level. If the CG is selected for restore, AppSync identifies other datastores residing on the
same CG that were also protected alongside, and restores them. If the affected entity was
not protected, AppSync will not be able to restore it properly. This is displayed as a
warning in the Restore wizard.
There are no affected entities for VNX because multiple datastores cannot span the same
LUN and multiple datastores cannot be hosted on the same File System.
If there are affected entities in your underlying storage configuration, the Restore Wizard
notifies you of these items requiring you to acknowledge that additional items will be
restored.
Restoring a VMware Datastore copy from XtremIO
Learn to semi-manually restore a Datastore copy on XtremIO with AppSync. Automated
restore is available with AppSync version 2.2.2 and above and XtremIO 4.0 and above.
Before you begin
For automated restore, refer to Restoring a Datastore copy using the Restore Wizard on
page 198.
For manual restore, ensure the following pre-requisites exist, then follow the steps:
l
VMware Datastore copy residing on an XtremIO array.
l
An AppSync service plan that replicates the Datastore.
Procedure
1. On the AppSync console, select Service Plans > VMware Datacenters, and then click
the wanted service plan, for example, Bronze.
The Copies page displays existing copies.
2. Select the Mount copy to restore to production the ESX server.
Figure 32 Select Mount Settings dialog box
3. Stop any virtual machines from the Datastore that are being restored.
If you do not want a virtual machine to be restored, use VMware vMotion to move it to
another database. it to another Datastore.
4. Remove virtual machines on the Datastore that you are restoring from vCenter
inventory.
5. Unmount the Datastore that you are restoring.
196
EMC AppSync 2.2.3 User and Administration Guide
Protect VMware Datacenters
Figure 33 Unmount Datastore
6. Delete the Datastore that you are restoring.
7. Rename the Datastore that is mounted by AppSync to the name of the production
Datastore.
8. Select the Remove action against the mounted copy.
This action removes the copy from the AppSync database without removing the
snapshot on the mount host or on the XtremIO array. You receive a message that
explains the Remove action. If the service plan has multiple Datastores, the storage
for all Datastores in the service plan remain on the XtremIO storage array.
9. Remove the old production LUNs from the initiator group for the ESX.
10. Move the mounted snapshot from the /AppSyncSnapshots folder to the folder
where the production LUN that you are replacing resides.
Move the mounted snapshot (that is, the snapshot that is currently masked to the ESX
initiator group). After you finish, the production LUN and the snapshot appear in the
same folder.
Figure 34 Move mounted snapshot
After the move one Datastore becomes a snapshot, the other becomes a volume.
Restoring a VMware Datastore copy from XtremIO
197
Protect VMware Datacenters
Figure 35 Snapshot and volume
11. Remove the volume that you want to replace with the snapshot.
The snapshot becomes a volume.
Figure 36 Snapshot becomes volume
Note
All other snapshots in the /AppSyncSnapshots folder for the Datastore XtremIO
volume that you just removed have also become volumes.
12. Rename the snapshot (now a volume) to the name of the volume that you replaced,
and then perform any necessary virtual machine actions for the "restored" Datastore,
such as VMware vMotion or add to the inventory.
13. Start all virtual machines in the Datastore that you want to run.
The restored Datastore has a new Datastore ID in vCenter.
14. Unsubscribe the old Datastore from the service plan, subscribe the new Datastore ,
and then run the Datastore service plan to create a snapshot of the "restored"
Datastore.
Restoring a virtual machine from a copy
You can perform a restore of a virtual machine from the Protected Virtual Machines tab or
the virtual machine's Copies page.
Before you begin
198
l
This operation requires the Data Administrator role in AppSync.
l
You must be using vSphere Enterprise Edition.
l
All datastores used by the virtual machine must be protected by selecting the Include
Virtual Machine Disk option in the Create copy phase of the service plan.
l
The virtual machine should not have any pre-existing snapshots.
l
Virtual machines with RDMS cannot be restored.
EMC AppSync 2.2.3 User and Administration Guide
Protect VMware Datacenters
Procedure
1. Select Restore from the Protected Virtual Machines tab or the virtual machine's
Copies page.
The virtual machine Restore Wizard launches.
2. Select the copy to restore.
Use the Copies or Service Plan filters to select the copy to restore. The copies list is
refreshed based on the filters selected.
For a RecoverPoint copy, you also have the option to select a bookmark based on a
specific time. However, there should be a copy available in AppSync before the time
you select.
Click Select a point in time to select a copy with a specific timestamp. The time shown
here is the console's time. If the console is in a different time zone from the RPA,
specify the time according to the server's time zone to restore the copy.
3. Click Next.
If other VMs were also protected along with the selected virtual machine, the Multiple
VM Restore page is displayed.
Select one of the following options:
l
Continue to restore only one virtual machine
l
View and/or select the other VMs for restore
4. In the Select Restore Location page, make the selections. See Virtual Machine restore
options on page 200 for details.
5. In the Select Mount Host page:
a. Select the mount ESX.
If production data resides on RecoverPoint storage, the target devices should be
visible to the selected mount host. In case of VNX, the mount host should be
registered to the VNX.
b. Select the RecoverPoint image access mode from the list - Logged Access, Virtual
Access, or Virtual Access with Roll.
c. Click Next.
Note
AppSync employs VMware vMotion technology to move the virtual machine from
mount host to restore location. Therefore, the mount host and host at the restore
location should satisfy the VMware vMotion prerequisites such as network
requirement.
In the Choose Instant Restore page, you can make a selection only if one of the
following conditions is met:
l
The mount and restore hosts are the same.
l
The mount and restore hosts are different but are nodes of the same ESX cluster.
6. In the Choose Instant Restore page, select Yes or No for the Do you want to perform
an instant restore option option, based on whether you want to perform an instant
restore.
Restoring a virtual machine from a copy
199
Protect VMware Datacenters
During instant restore, you can continue to use the virtual machine. Though the virtual
machine is powered on, the VMs are restored in the background.
If you select No, and if you had chosen to restore multiple virtual machines in Step 2
of this wizard, specify a number in the Maximum number of simultaneous virtual
machines to be restored box. By default, the number is 2.
Note
If you are restoring multiple virtual machines belonging to a vApp, set Maximum
number of simultaneous virtual machines to be restored to 1.
If the source devices are think, the Instant restore option is not available for VMAX
copies.
7. On the Summary page, review the settings that you selected in the previous pages
and click Finish to perform the restore.
8. In the Results page, click View Details to see progress of the different phases that are
part of restoring a virtual machine.
The last phase completed is displayed at the bottom of the list.
Virtual Machine Restore options
You can select the restore location as well as restore operations.
Table 22 Virtual machine restore options
Restore
Option
Description
Original
location
Restores to the location where the virtual machine was present at the time of
protection.
Note
For a RecoverPoint copy, restoring to the original location is not recommended.
AppSync displays an appropriate warning when you select this option.
Alternate
location
Options if the
VM being
restored
already exists
in the restore
location
200
Restores to a location selected from the following options. All are mandatory.
l
vCenter Server: You can select either the same vCenter Server where the
datastore with the virtual machine was at the time of protection or a
different server.
l
Datacenter
l
Host
l
Datastore
l
Fail the restore: AppSync checks for the existence of the virtual machines in
the restore location. For those virtual machines that exist in the restore
location, the restore operation is aborted. For the rest, the restore operation
continues. This is a precautionary option.
l
Create a new virtual machine: AppSync creates a new virtual machine
before restoring.
EMC AppSync 2.2.3 User and Administration Guide
Protect VMware Datacenters
Table 22 Virtual machine restore options (continued)
Restore
Option
Description
l
Unregister the virtual machine: If the virtual machines selected for restore
exist in the restore location, AppSync unregisters them from the inventory
before restoring.
l
Delete from disk before performing restore: If the virtual machines being
restored exist in the restore location, AppSync deletes them before
restoring.
Note
It is recommended you take a backup of the virtual machine before
proceeding with the restore operation.
l
Delete from disk after performing restore: If the virtual machines being
restored exist in the restore location, AppSync deletes them after restoring.
File or folder restore with VMFS or NFS datastores
Files or folders stored on virtual disks on a virtual machine in VMFS and NFS datastores
can be restored through AppSync.
The virtual disks stored in a VMFS or NFS datastore that are protected by an AppSync
service plan can be used for file or folder level restore by specifying the location for
mounting the virtual disk copy.
Within AppSync, file or folder level restore is a three phase process: To complete the
restore, the final step is performed manually outside of AppSync. You must copy the files
or folders from the location where the virtual disk is mounted to a location of your choice.
1. AppSync mounts the datastore snapshots to the ESX server on which the virtual
machine with the AppSync agent resides.
2. The vCenter server adds the virtual disks from the datastore snapshots to the mount
VM without powering off the VM.
3. AppSync agent performs a filesystem mount to the mount VM.
Restore of files or folders from virtual disks with multiple partitions is supported.
If the ESX server version is 5.0 and higher, the original VM can also be the mount VM.
Restrictions
l
File or folder level restore is not possible on dynamic disks.
l
Virtual disks belonging to same phase pit cannot be mounted to the same virtual
machine even when created using the Gold service plan. In other words, if a virtual
disk from a local copy is mounted; then the same virtual disk from a remote copy
cannot be mounted.
l
To perform an Any Point in Time (APiT) file restore, you must first perform an APiT
mount of the datastore and then launch the Granular File Restore wizard from the
APiT copy.
File or folder restore with VMFS or NFS datastores
201
Protect VMware Datacenters
Restoring a file or folder from a virtual disk
You can perform the restore of a file or folder of a virtual disk from the Protected Virtual
Machines tab or the virtual machine's Copies page.
Before you begin
l
This operation requires the Data Administrator role in AppSync.
l
You must be using vSphere Enterprise Edition.
l
The virtual machine on which the copy is mounted and restored must be 64-bit with
Windows 2008 or Windows 2012 as the operating system. The AppSync host plugin
must be installed on it and it should be registered with the AppSync server.
Procedure
1. Select Restore > File from the Protected Virtual Machines tab or the virtual machine's
Copies page.
The Granular File Restore Wizard launches.
2. Select the copy to restore.
Use the Copies or Service Plan filters to select the appropriate copy to restore. The
copies list is refreshed based on the filters selected.
3. Click Next.
The Select Virtual Disk page appears displaying the virtual disks in the selected
virtual machine.
4. Select the virtual disks whose files or folders must be restored and click Next.
5. In the Select Host page:
a. Select the virtual machine on which the copy must be mounted.
In addition, specify the location in the selected virtual machine where the disk
must be restored to. By default, the files are restored to the following location:
%system drive%\AppSyncMounts\<VM_name>\<copy_id>\Hard
disk# where:
l
%system drive% is system drive of the selected virtual machine on which
the copy is to be mounted
l
<VM_name> is the name of the virtual machine whose virtual disks are being
restored
l
<copy_id> is an AppSync generated ID
l
Hard disk# is the number of the hard disk in the virtual machine. This
number is the same as on the original virtual machine.
b. Select the RecoverPoint image access mode from the list - Logged Access, Virtual
Access or Virtual Access with Roll.
6. In the Summary page, review the settings that you selected in the previous pages and
click Finish to start the restore of the disk.
7. In the Results page, click View Details to see progress of the different phases that are
part of restoring a virtual disk.
The last phase completed is displayed at the bottom of the list.
8. Next, perform the manual step of copying the required files or folders from the mount
location to a location of your choice.
202
EMC AppSync 2.2.3 User and Administration Guide
Protect VMware Datacenters
9. Optionally, unmount the datastore. Unmount a datastore copy on page 194.
Restoring a file or folder from a virtual disk
203
Protect VMware Datacenters
204
EMC AppSync 2.2.3 User and Administration Guide
CHAPTER 10
Monitor AppSync
This chapter includes the following topics:
l
l
l
l
RPO concepts and best practices.........................................................................206
Alerts and associated events...............................................................................207
Sending alerts via email...................................................................................... 208
Repurpose Monitor..............................................................................................209
Monitor AppSync
205
Monitor AppSync
RPO concepts and best practices
A recovery point objective (RPO) is one of several scheduling options that can be selected
as part of a service plan's Plan Startup phase.
When you subscribe an object (such as an Exchange database) to a service plan that
uses RPO as its recurrence type, the object acquires the recovery point objective
specified in the service plan.
Since you can subscribe an object to more than one service plan, it is possible for an
object to have more than one recovery point objective. When an object has more than
one RPO, the service plan with the highest RPO frequency (that is, the lowest RPO hours
value) is used for calculation and reports.
As a best practice, you should subscribe an object to only one RPO-enabled service plan.
If you subscribe an object to additional service plans, they should not use the RPO-based
recurrence type.
Recovery point compliance report
The recovery point compliance report shows the recoverability for all objects that are
subscribed to service plans with an RPO recurrence type. The report is at Monitoring >
Recovery Point Compliance Report.
Table 23 Recovery Point Compliance Report
Column
Description
Server
Host level object, such as a Microsoft Exchange Mailbox Server
Application
Name of the protected object, such as a Microsoft Exchange database.
Click the name to go to the list of copies for the object.
Recovery Point
Objective
The recovery point objective as defined in the Start phase of the
associated service plan.
Time Since Last
Recovery Point
Amount of time since the last copy or bookmark created by the associated
service plan. A green icon indicates the copy is RPO compliant. A red icon
indicates non-compliance.
Service Plan
Name of the service plan. Click the name to go to the service plan
definition.
Exporting an RPO compliance report to CSV
You can create a recovery point objective (RPO) compliance report in comma-separated
value format.
Before you begin
No particular AppSync role is required for this operation.
Procedure
1. Navigate to Monitoring > Recovery Point Compliance Report.
2. Sort and arrange columns as desired for the report.
3. Click Export to run the Export wizard.
206
EMC AppSync 2.2.3 User and Administration Guide
Monitor AppSync
You have the option to include table headers and export only selected rows. The
default file name is "Recovery Point Objectives_mm_dd_yyyy_hh_mm_ss.csv".
Summary of RPO compliance
The Recovery Point Objectives (RPO) summary on the dashboard shows the percentage of
RPOs met across all objects that are subscribed to RPO-enabled service plans.
Alerts and associated events
AppSync generates an alert when a service plan phase fails, when a recovery point
objective (RPO) is not met, or when a mount or restore fails.
Service plan failure alerts are generated immediately on failure of a service plan phase.
When an application goes out of RPO compliance, the associated alert is generated
within 1 hour. Refer to Acknowledging alert icons for database, file system, and data
store service plan runs on page 208, for information.
AppSync displays alerts in the console at Monitoring > Alerts.
Table 24 Details of alerts
Column
Description
Alert (!)
Level of alert
Time
Date and time of the alert.
Server
Application server, such as a Microsoft Exchange server.
Application
Replicated object, such as a Microsoft Exchange database.
Category
Phase Failure, RPO
Service Plan Service plan that was running when the alert was generated, or the service plan that
created the copy that failed a mount or restore.
Message
Describes the cause of the alert.
You can filter alerts by the time they were generated and by the associated service plan.
View the associated events that led up to the alert by clicking the alert. Expand the toplevel events to see additional details. You can filter associated events by any column.
Acknowledging alerts
With AppSync version 2.2.2 and later you can acknowledge alerts. Refer to
Acknowledging alerts on page 207, and Acknowledging alert icons for database, file
system, and data store service plan runs on page 208 for more information.
Acknowledging alerts
These steps show you how to acknowledge alerts for monitoring.
You can choose to acknowledge alerts that are shown in the console, Monitoring > Alerts.
The alerts page display shows a column for Alert Acknowledged.
A value of No is the default. When you acknowledge an alert, the value of the alert
changes to YES from the default value NO.
Summary of RPO compliance
207
Monitor AppSync
Procedure
1. Go to Monitoring > Alerts and select an alert from the alerts table with a current value
of NO in the Alert Acknowledged column.
2. Click Acknowledge Alert.
Results
The alert displays a value of YES in the Alert Acknowledged column of the Alert table.
Acknowledging alert icons for database, file system, and data store service plan runs
You can acknowledge an alert icon within the AppSync console for Oracle and SQL
databases, file systems, and data stores.
An alert icon indicates the status of the most recent service plan run. The icon appears
beside a database, file system, or data store after the run.
After you acknowledge the icon in the Acknowledge column, AppSync changes the icon to
an information icon. Also, you can acknowledge the alert after every Service Plan run. You
are not restricted to acknowledge an alert only once.
Note
If a delete is pending, then the Acknowledge button becomes disabled for the database,
file system, or data store.
The following procedure shows you how to view and acknowledge the alert.
Procedure
1. Select an alert from the database, file system, or data store table that has an alert
icon that is associated with the last service plan run.
2. Click the Acknowledge Alert button that is located below the database, file system, or
data store table.
The alert icon changes to an information icon.
3. If wanted, re-run the service plan on the same database, file system, or data store to
display the alert icon for this run.
Sending alerts via email
You can configure AppSync to send failure alerts via email to a list of recipients.
You enable alert emails and add recipients per application instance, such as an Exchange
mailbox server.
Note
You can also configure AppSync to send email after successful completion of a scheduled
service plan. This is disabled by default. Contact Customer Support if you want this
enabled.
Configuring server settings for email alerts
Configure SMTP services on a machine that the AppSync server can access.
Before you begin
This operation requires the Resource Administrator role in AppSync.
Refer to SMTP documentation for configuration procedures.
208
EMC AppSync 2.2.3 User and Administration Guide
Monitor AppSync
Procedure
1. Select Settings > AppSync Server Settings.
2. Under Send Mail settings, enter values for SMTP server, port, and sender.
Use Test to validate the settings. Enter a recipient's address and then check the
recipient's inbox. You can also select Insert Default Values.
3. Click Apply.
Specifying email alert recipients
Configure email alerts per application instance level.
Before you begin
l
This operation requires the Data Administrator role in AppSync.
l
SMTP services must be configured on a machine accessible by the AppSync server.
See Configuring server settings for email alerts on page 208.
Procedure
1. Select Copy Management > Application ( select application).
2. Select the application object for which to set alert settings.
3. Select Alert Settings.
4. Enter one or more email recipients, separated by commas (,) and click OK.
A test email is sent to recipients for verification.
Repurpose Monitor
This topic describes the Repurpose Monitor.
The Repurpose Monitor shows all in-progress Repurpose and Refresh processes. It
enables you to view currently running repurpose activities and to monitor their progress.
It shows the current copy you are repurposing (making a snap of a snap copy), and
related copy labels such as 1st Gen (generation) or 2nd Gen.
You typically use the Repurpose Monitor for long-running repurpose activities (for
example, VMAX Clone synchronization). When you select this monitor, it lists all the
repurpose activities in progress. Select any repurpose activity, and then select Monitor to
launch a progress dialog to directly view the progress of the selected activity.
To launch the Repurpose Monitor:
1. On the AppSync console select Monitoring.
2. Select Repurpose Monitor on the Monitoring page to launch.
Specifying email alert recipients
209
Monitor AppSync
210
EMC AppSync 2.2.3 User and Administration Guide
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertising