Protecting Database App
Protecting Database App
Julien Sauvanet
Kenneth Salerno
Markus Fehling
Redpaper
Draft Document for Review June 18, 2021 11:51 am 5640edno.fm
IBM Redbooks
June 2021
REDP-5640-00
5640edno.fm Draft Document for Review June 18, 2021 11:51 am
Note: Before using this information and the product it supports, read the information in “Notices” on page v.
Contents
Notices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .v
Trademarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vi
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii
Authors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii
Now you can become a published author, too! . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viii
Comments welcome. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viii
Stay connected to IBM Redbooks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viii
Notices
This information was developed for products and services offered in the US. This material might be available
from IBM in other languages. However, you may be required to own a copy of the product or product version in
that language in order to access it.
IBM may not offer the products, services, or features discussed in this document in other countries. Consult
your local IBM representative for information on the products and services currently available in your area. Any
reference to an IBM product, program, or service is not intended to state or imply that only that IBM product,
program, or service may be used. Any functionally equivalent product, program, or service that does not
infringe any IBM intellectual property right may be used instead. However, it is the user’s responsibility to
evaluate and verify the operation of any non-IBM product, program, or service.
IBM may have patents or pending patent applications covering subject matter described in this document. The
furnishing of this document does not grant you any license to these patents. You can send license inquiries, in
writing, to:
IBM Director of Licensing, IBM Corporation, North Castle Drive, MD-NC119, Armonk, NY 10504-1785, US
This information could include technical inaccuracies or typographical errors. Changes are periodically made
to the information herein; these changes will be incorporated in new editions of the publication. IBM may make
improvements and/or changes in the product(s) and/or the program(s) described in this publication at any time
without notice.
Any references in this information to non-IBM websites are provided for convenience only and do not in any
manner serve as an endorsement of those websites. The materials at those websites are not part of the
materials for this IBM product and use of those websites is at your own risk.
IBM may use or distribute any of the information you provide in any way it believes appropriate without
incurring any obligation to you.
The performance data and client examples cited are presented for illustrative purposes only. Actual
performance results may vary depending on specific configurations and operating conditions.
Information concerning non-IBM products was obtained from the suppliers of those products, their published
announcements or other publicly available sources. IBM has not tested those products and cannot confirm the
accuracy of performance, compatibility or any other claims related to non-IBM products. Questions on the
capabilities of non-IBM products should be addressed to the suppliers of those products.
Statements regarding IBM’s future direction or intent are subject to change or withdrawal without notice, and
represent goals and objectives only.
This information contains examples of data and reports used in daily business operations. To illustrate them
as completely as possible, the examples include the names of individuals, companies, brands, and products.
All of these names are fictitious and any similarity to actual people or business enterprises is entirely
coincidental.
COPYRIGHT LICENSE:
This information contains sample application programs in source language, which illustrate programming
techniques on various operating platforms. You may copy, modify, and distribute these sample programs in
any form without payment to IBM, for the purposes of developing, using, marketing or distributing application
programs conforming to the application programming interface for the operating platform for which the sample
programs are written. These examples have not been thoroughly tested under all conditions. IBM, therefore,
cannot guarantee or imply reliability, serviceability, or function of these programs. The sample programs are
provided “AS IS”, without warranty of any kind. IBM shall not be liable for any damages arising out of your use
of the sample programs.
Trademarks
IBM, the IBM logo, and ibm.com are trademarks or registered trademarks of International Business Machines
Corporation, registered in many jurisdictions worldwide. Other product and service names might be
trademarks of IBM or other companies. A current list of IBM trademarks is available on the web at “Copyright
and trademark information” at http://www.ibm.com/legal/copytrade.shtml
The following terms are trademarks or registered trademarks of International Business Machines Corporation,
and might also be trademarks or registered trademarks in other countries.
AIX® IBM® Redbooks®
Db2® IBM Services® Redbooks (logo) ®
DB2® IBM Spectrum® Tivoli®
The registered trademark Linux® is used pursuant to a sublicense from the Linux Foundation, the exclusive
licensee of Linus Torvalds, owner of the mark on a worldwide basis.
Microsoft, Windows, and the Windows logo are trademarks of Microsoft Corporation in the United States,
other countries, or both.
OpenShift, Red Hat, are trademarks or registered trademarks of Red Hat, Inc. or its subsidiaries in the United
States and other countries.
VMware, and the VMware logo are registered trademarks or trademarks of VMware, Inc. or its subsidiaries in
the United States and/or other jurisdictions.
Other company, product, or service names may be trademarks or service marks of others.
Preface
IBM® Spectrum Protect Plus is a data protection solution that provides near-instant recovery,
replication, retention management, and reuse for virtual machines, databases, and
application backups in hybrid multicloud environments.
Authors
This paper was produced by a team of specialists from around the world .
Bert Dufrasne
IBM Redbooks, San Jose Center
Siddharth Bhatt
Dominic Mueller
Jim Smith
Joerg Walter
Axel Westphal
IBM
Find out more about the residency program, browse the residency index, and apply online at:
ibm.com/redbooks/residencies.html
Comments welcome
Your comments are important to us!
We want our papers to be as helpful as possible. Send us your comments about this paper or
other IBM Redbooks publications in one of the following ways:
Use the online Contact us review Redbooks form found at:
ibm.com/redbooks
Send your comments in an email to:
[email protected]
Mail your comments to:
IBM Corporation, IBM Redbooks
Dept. HYTD Mail Station P099
2455 South Road
Poughkeepsie, NY 12601-5400
https://www.redbooks.ibm.com/Redbooks.nsf/subscribe?OpenForm
Stay current on recent Redbooks publications with RSS Feeds:
http://www.redbooks.ibm.com/rss.html
Preface ix
5640pref.fm Draft Document for Review June 18, 2021 11:57 am
Most of the configuration information included in this paper applies to all database
applications.
Note: IBM Spectrum Protect Plus offers data reuse functions in addition to backup and
restore. You can use the database backup data to create a permanent copy (or clone) of
your production database, or to temporarily establish a database copy directly from the
vSnap server volumes.
Backup, restore, and data reuse handling functions for the supported (relational) databases
are all similar in IBM Spectrum Protect Plus.
These are the steps which we cover in this chapter, enabling application’s backup:
Optional but considered as good practice: create an identity
Register the database application
Perform an inventory of the application (the system automatically triggers an inventory
following the registration process)
Assign the application to an SLA
– The SLA must exist, and contains, amongst other option, the Data backup frequency
(not the log backup frequency) and associated retention ( for both data and logs ).
Configure the application database Options
– At this step, you will have to specify how you want the application log to be handled
Start the backup manually or wait for the next schedule as per the SLA instructions
Note: Do not assign more than one application per machine as an application server to a
resource group. For example, if Microsoft SQL Server and Microsoft Exchange Server
occupy the same machine and both are registered with IBM Spectrum Protect Plus, only
one of the applications can be added as an application server to a given resource group.
As a reminder Resource Group is used for the Role Based Access Control (RBAC)
definition.
For more information about supported databases and environments, see IBM Spectrum
Protect Plus Installation and User’s Guide.
In the IBM Spectrum Protect Plus GUI, select Accounts → Identity → Add Identity to enter
the user definition for your specific databases, as shown in Figure 1-1.
Tip: If you use identical operating system users and passwords for multiple database
servers, IBM Spectrum Protect Plus allows you to manage these databases under one
identity.
With a list of identity entries, as shown in Figure 1-2, you can see that there are default
system identities (such as serveradmin) and identities explicitly created for database backup
and restore (DB administrator, operating system user, Mongo DB user, and Oracle DBA).
As shown in Figure 1-4, you then enter the database server’s IP name or address, enter the
database administration user or select an identity that you defined earlier. Click the Get
databases button to start a database discovery job on the server.
If you save this application server entry, IBM Spectrum Protect Plus automatically starts an
inventory job. This job confirms a network connection, adds the application server to the IBM
Spectrum Protect Plus database, and then catalogs the instance. You can follow the inventory
job or its status in the Jobs and Operations menu.
After you registered the application, an important step is to specify the options, this is where
you define how you want the logs to be managed for the database. This is not controlling the
retention but the frequency of the log backup as well as whether SPP will delete them from
the local system or not, and if yes, when.
Application Options is also where you specify how many streams will be used to perform the
database backup, for the applications which are supporting this feature (Oracle is one of
those). When enabling multiple streams for database backup, consider the performance
impact it can cause to the application server. Moreover, parallel stream is working at datafile
level, so consider the number of datafile to set parallel stream value. Tuning this performance
setting is likely to be an iterative process to find the right spot, specific to your environment.
Figure 1-5 below shows the possible Application Options. By checking the Enable Log
backup, additional options are proposed, to let you decide how log backup has to occur.
Figure 1-5 Application Database options selection - for log backup management & multi stream backup for supported
databases
Starting with Spectrum Protect Plus 10.1.8, the log backup can be done below the hour
frequency ( SubHourly option).
Moreover, beyond the frequency, you can decide whether the logs are being deleted from the
production system, upon Spectrum Protect Plus log backup, after some time (days or hours,
or never. Possible options are shown in Figure 1-6
Figure 1-6 How applications log are handled on the production database environment
To assign SLA, select the application resource and use the ‘Select an SLA policy’ button. You
can do it as either the host level for all databases, or at instance level.
Following the SLA assignment, your application or database is ready for backup. you can
decide to let the backup start as per the SLA schedule or manually trigger backup
immediately. In the latter, from the Application menu, select the application resource and Click
the Run button, as shown in Figure 1-7.
Figure 1-7 Start a manual backup for an application, after registration or at any later time
You can track the progress of a running backup from the Jobs & Operations menu.
Your application is now registered and is part of the Spectrum Protect Plus backup plan,
according to the assigned SLA.
1.2 IBM Spectrum Protect Plus database restore and data reuse
IBM Spectrum Protect Plus features a restore wizard that simplifies restore operations for
virtual machines and databases. The wizard guides you through the configuration of restore
types and parameters, and optionally schedules a job that performs the actual restore.
IBM Spectrum Protect Plus treats data reuse and data recovery as a restore activity. In either
case, you must create a restore job. The Databases and the Jobs and Operations menus in
IBM Spectrum Protect Plus have a button that is used to start creating a restore job. The
parameters that you select during job creation define which activity is performed.
The following list describes the parameters that control the final restore or data reuse activity:
Type of Restore:
– On-Demand Snapshot: one-time restore operation (you choose with-in the list of
backup date and time)
– On-Demand Point in Time: one-time restore by selecting a point-in-time backup of that
database (you specify a point in time or a transaction number)
– Recurring: repeating point-in-time restore job that runs on schedule
Restore Method:
– A production restore overwrites the original database or creates a database copy with
a different database name. In the database copy case, you must specify a new
database name and the destination paths.
– A test restore mounts the vSnap server directories with a database backup to a
database server, recovers and opens the database. You can rename the database.
– An instant access restore also mounts the vSnap server directories with a database
backup to a database server, but does not recover or open the database
Destination:
– Restore to the original instance
– Restore to an alternative instance
The following sections describe examples for these use cases. The sample database is
Oracle database 12c.
Before you start to create a restore job, you must first select the database and an associated
backup to restore, as shown in Figure 1-8.
In the IBM Spectrum Protect Plus restore wizard, you can set up such a user requirement by
choosing the following parameter settings:
Restore type: On-demand point in time (or On-demand snapshot, depending on the
available backups)
Restore method: Test
Destination: Original or alternative instance
First, select the database instance and an associated database backup, as shown in
Figure 1-8. In addition, select a site and a location for the instance to restore. These settings
depend on your specific environment, which can include a cloud or copy location, or a
secondary site that you use for replication.
In our example, we chose the primary site and we chose Recurring to create a repeating
restore job that runs on a schedule, as shown in Figure 1-9.
For our use case, we decided to create the test database in an alternative destination (which
means not on the original production server) and give the database a new name. Figure 1-10
and Figure 1-11 show the corresponding parameter selections.
Finally, we define the schedule: every Monday at 8:00 o’clock. Figure Example 1-12 on
page 9 shows an example. If you do not want to wait for the first test run, you can find the
scheduled job in the Jobs and Operations menu of IBM Spectrum Protect Plus. Select the
Schedule tab, find the job in the list, and start it manually.
IBM Spectrum Protect Plus does not reflect the new database name in the name of the
mounted directory or in the data file names, but it starts the database with the new database
identifier (System ID, SID).
The IBM Spectrum Protect Plus test restore job that you started stays active until you
manually terminate it. In the Job and Operations menu, the job status is shown as
“Resource active”. To terminate the job, select it and choose End instant disk restore.
In comparison to the test restore, an instant access restore job does not start a database;
therefore, you do not need to select a database instance as a restore target.
From the mounted file system you can use the data for custom recovery; for example:
Reload individual files such as control files, configuration files, and data files.
Rebuild a customized database copy.
As described in 1.2.1, “Test restore”, the instant access job remains active with the “Resource
active” state until you terminate it manually. To terminate the job, select it and choose the End
instant disk restore action.
In the IBM Spectrum Protect Plus restore wizard, the following parameters initiate a traditional
database restore that overwrites the existing database:
Restore type: On-Demand Point in Time
Restore method: Production
Destination: Original Instance
First, select the database instance and an associated database backup, as shown in
Figure 1-8 on page 8. In addition, select a site and a location for the instance to restore.
These settings depend on your specific environment, which can include a cloud or copy
location, or a secondary site that you use for replication.
The next two selections indicate what we are trying to achieve; that is, production restore to
the original instance, as shown in Figure 1-14 and Figure 1-15.
For a restore of a production database, the IBM Spectrum Protect Plus restore wizard
assumes a database roll forward to a specific point in time that you can configure in the next
menu.
You must also decide whether to overwrite an existing database, as shown in Figure 1-16 on
page 11. IBM Spectrum Protect Plus provides an auxiliary protection against an unintended
data overwrite: If the database still exists and you do not select the overwrite choice box, the
restore job fails.
Figure 1-16 Selecting the database roll forward and overwrite options
Carefully review the job summary that IBM Spectrum Protect Plus displays (Figure 1-17) and,
if the information describes what you are trying to achieve, run the restore job.
The IBM Spectrum Protect Plus restore job verifies whether a database exists or is even up
and running. In this case, the Allow database overwrite setting is relevant, as shown in the
restore job log shown in Example 1-1.
The intent here is to provide technical details about how the process works. For more
information about how to configure backup and run recovery operations for specific
databases with IBM Spectrum Protect Plus, see the following resources:
Chapter 2, “Protecting Oracle database applications” on page 21
For the database that includes log backup enabled, two volumes are created:
one volume for the data
one volume for the log management - which remains mounted at any time to the target
application server
MongoDB: Unlike the other supported databases, MongoDB does not have log
management configurable in IBM Spectrum Protect Plus.
The SLA policy that you assigned to your database defines the schedule time for the first
backup. If you did not define a schedule, or you do not want to wait for the first automatic
backup schedule, click the Run button or scroll down to the SLA policy that you provided for
this database and select the Actions button to start the database backup.
At this point, you also decide to perform a backup of a single database by clicking the Run
button) or a backup of all applications included in the SLA policy by clicking the Actions
button (see Figure 1-5 on page 5).
For each SLA and each database included in that SLA, the following resources are created
on the vSnap server to store backups:
One primary data vSnap server volume
One primary log vSnap server volume
These data and log volumes are made available to the database server, using different
access method depending on the platform and application, as shown in Figure 1-18.
Figure 1-18 How database server access their respective vSnap server volumes
Table 1-1 lists what actions take place for data backup and describes the mechanism that is
used by IBM Spectrum Protect Plus to handle the application logs backup.
Microsoft SQL VSS snapshot, copy all files to Windows task scheduler
iSCSI LUN
Microsoft Exchange VSS snapshot, copy all files to Windows task scheduler
iSCSI LUN
IBM Db2 LVM snapshot, copy all files to db2 archive log scheduler
share
MongoDB LVM snapshot, copy all files to Copy journal with data file
share
A point-in-time Production restore creates a temporary vSnap server clone volume of the last
vSnap server data volume before the selected point in time and mounts that clone to the
target server. A copy then occurs on the target server, from the clone volume to the
production volume. After the copy process completes, the clone volume is dismounted and
deleted from the vSnap server.
The next step is to create a temporary clone of the log volume that contains database logs
that are created after the selected point in time and mount that clone to the target. This clone
contains the log backup with database transactions that occurred after the data was restored
in the first step, and allows a roll forward recovery until the specified point in time.
The Test restore works the same way as point in time production restore, but the production
data is not copied back. The restore data is provided as a share from the vSnap server.
The Instant Access restore creates a temporary vSnap server clone volume of the selected
(point in time) backed up data and mounts that clone to the target application server for
access. The same clone and mount operation occurs on the log volume of that same point in
time.
These clones allow read/write access, so the application can work with the data. However,
when the instant access process completes, the data modifications are not persistent, and
any modifications that were made during the instant access are lost. The original backup
does not change.
Several environments exist in which the database or other application requires their backup to
be integrated into a sequence of actions, in a specific order. These tasks are usually
managed by a scheduler, such as Tivoli Workload Scheduler.
The intent of this section is to explain which steps and features can be used within IBM
Spectrum Protect Plus to synchronize the backup with an external scheduler. Although the
backup is still triggered by an IBM Spectrum Protect Plus policy, we can use the pre-script
option to have IBM Spectrum Protect Plus run a piece of code on the server host. In our
example, we show how to wait for an external scheduler flag before triggering the backup.
Example 1-2 shows the few lines of code that we use as a pre-script. This code is going to
loop until a specific file (acting as a flag) is placed in a specific location by an external
mechanism. This flag file is the signal for IBM Spectrum Protect Plus to trigger the backup.
Example 1-2 Pre-script to make an IBM Spectrum Protect Plus backup waiting for external signal
#! /bin/sh
while [ ! -f /tmp/external_scheduler.flag ]
do
date >> /tmp/SPP_wait4external_scheduler.log
echo " Flag /tmp/external_scheduler.flag not there, wait to start backup " >>
/tmp/SPP_wait4external_scheduler.log
sleep 60
done
echo " Flag /tmp/external_scheduler.flag is here, backup time ! " >>
/tmp/SPP_wait4external_scheduler.log
Tips: Supported scripts include shell scripts for Linux-based machines and batch and
PowerShell scripts for Windows-based machines. Scripts must be created by using the
associated file format for the operating system.
Running the dos2unix command before uploading the script in IBM Spectrum Protect Plus
might help you to ensure suitable format of a shell script if you encounter a formatting
problem (that is, ^M at end of line).
The use of the scripts with IBM Spectrum Protect Plus is an easy three-step process:
1. Define the script.
2. Define the script servers.
3. Update the SLA policy options from the SLA Policy Status page.
The first step is to define the script. From the IBM Spectrum Protect Plus GUI, select System
Configuration → Script to define the script.
Defining the script means uploading the script to IBM Spectrum Protect Plus, as shown in
Figure 1-19. Click Browse to select and upload the script that you plan to run on the
application or database server as part of the backup job.
The second step is to define the Script Server, as shown in Figure 1-20. Select System
Configuration → Script. Specify the Host address, login credentials, and operating system
type for which you plan to use script.
Figure 1-20 Define SCript Server by specifying its address and credentials
The third step is to update an SLA Policy or create an SLA Policy and enable the use of
pre-script in the policy options. Such an option can be accessed by editing the SLA Policy
Options by selecting Manage Protection → <Backup Workload of your choice> → SLA
Policy Status.
In our example, we want to run a script to check whether an Oracle backup ran. We select
Manage Protection → Databases → Oracle → SLA Policy Status. Then, we select Policy
Options, as shown in Figure 1-21.
Figure 1-21 Open the Policy Options to enable the use of a script for that SLA Policy run
When you click Policy Options, a pop-up window opens (see Figure 1-22) and you must
specify whether you want to enable pre-script or post-script, and which script to run.
Figure 1-22 Configure Policy Options to use a specific script for pre or post backup tasks
Whenever the SLA policy is used, completes the pre-script step before triggering the backup
commands. Moreover, if you disable the Continue job/task on script error option, the backup
does not run if the pre-script failed.
Figure 1-23 shows the output log that lists the execution of pre-script.
Figure 1-23 Job log showing that pre script action is executed on the target server t3-vm-lx
Figure 1-24 shows the output log when the pre-script completed so the backup job continues
and triggers the backup of (in our example) the Oracle database.
Figure 1-24 Job log showing that backup action is happening after the pre script completed successfully
Tip: For more information about pre-script or SLA policy options in the job log, enable the
Detail filter of the Job Log.
In this chapter, we demonstrated the use of pre-script with an external scheduler example.
However, be aware that schedule might also be triggered externally by using the IBM
Spectrum Protect Plus REST API.
Tip: If you are running a standby Oracle database, you can register the standby database
for backup instead of the production database to reduce the backup impact on business
operations.
Before jumping into the details of the above steps, and the Spectrum Protect Plus capabilities,
let’s review some Oracle contextual information.
RMAN
IBM Spectrum Protect Plus relies on Oracle RMAN (Recovery MANager) to perform the
backups and recovery. Any backup related metadata generated by Spectrum Protect Plus are
written to Oracle instance control files. Writing to a remote RMAN catalog is not supported at
the time of this writing.
Spectrum Protect Plus builds its own backup catalog. It is stored inside the Spectrum Protect
Plus virtual appliance and should be protected via Spectrum Protect Plus catalog backup. It is
not possible to use an application specific backup for the catalog.
All commands used by Spectrum Protect can be found in the Job logs, which includes the
command.log file. To find the RMAN command used, you can download the Job log (From the
Jobs and Operations → Job History , select the job you want and click the Download .zip
button).
Once you have extracted that zip file, browse the directory structure and look for the file
named command.log. In this file, you will have all the commands used for the specific
operation. Note that one Job log may have multiple logs (as many as sub-operations) and
therefore multiple directories in this structure.
For example in Figure 2-1, you can view an extracted log from the Spectrum Protect Plus
server, for an Oracle application backup job ID 1620121641156, which ran against the Oracle
server named 10.0.240.209 (IP address)
Note: If Block Change Tracking is not enabled, IBM Spectrum Protect Plus enables it
automatically during the first backup
Oracle Compression
IBM Spectrum Protect Plus uses its own compression and deduplication mechanisms. It does
not use the Oracle Advanced Compression feature (which requires an extra license).
Oracle multi-threading
Oracle 12c introduced the concept of multi-threading. In IBM Spectrum Protect Plus, a
multi-threaded database configuration requires Oracle credentials for backup processing.
The discovery process identifies if multi-threading processing is enabled, and prompts the
user for the credentials. Enter the credentials for multi-threaded databases at the time of
registration. IBM Spectrum Protect Plus passes on the credentials to the Oracle agent during
backup, and the agent uses the credentials to log in to the database.
Note: When restoring an Oracle database that was configured for multi-threading at the
time of backup, the restored database is non-multithreaded. The restored database must
be manually reconfigured to use multi-threading.
For an Oracle environment it is important that the user used for the registration process has
proper Oracle permissions as well as sudo permissions. Before registering the database on
the Spectrum Protect Plus server, on the system you want to protect, create a user and
configure the proper sudo permissions as shown in Example 2-1.
Example 2-1 Create sppagent user for Oracle server registration on Spectrum Protect Plus
itso-oracle:~ # useradd -m -G oinstall,dba,asmadmin sppagent
itso-oracle:~ # passwd sppagent
New password:
Retype new password:
passwd: password updated successfully
To reflect this configuration in Spectrum Protect Plus, you can then create a Spectrum Protect
Plus identity pointing to the user you just created on the machine you need to protect. Follow
the generic steps to register a database application as described in 1.1, “Database
application backup configuration basics” on page 2 .
Note that when registering Oracle Real Application Cluster (RAC) nodes, each node must be
registered using its physical name or address, not the virtual or SCAN address.
When Oracle mutithread is enabled, the Set Credential button appears at a database
instance level in the second registration step, as shown in Figure 2-2. You must specify the
credentials of an Oracle database user who has SYSDBA privileges. You can specify the
same credentials as those specified to register the application if the privileges associated
meet the SYSDBA requirement.
Application Inventory process, which is triggered right after the application registration
automatically (as well as run on a regular basis - every 24 hours by default), is collecting
extensive information about the Oracle database; that information is key to understand how
the database can be protected.
Application Server Inventory: this process runs by default every 24 hours and refreshes
Application inventory information, including Oracle information. It is advised to not
schedule this inventory within the backup window. You can configure it in the Jobs &
Operations → Schedule → Application Server Inventory menu.
Here after are the steps performed by the Application Inventory process:
Spectrum Protect Plus (SPP) deploys agents into each Oracle server
Agent discovers all Oracle Homes, databases, ASM disk groups (if any) and reports them
back to SPP.
– Collect installation details by reading the /etc/oraInst.loc which contains the path to
oraInventory
– Collect list of Oracle databases
– Collect database details
A set of SQL queries are executed to retrieve information specific to each database
such as DB Name, DB ID, current incarnation, character set, list of data files, list of
redo logs, list of control files, list of temp files, FRA location, list of archived log files ,
SPFILE location, log mode, open mode, size of the database, block change tracking
status and current SCN.
Example 2-2 List of SQL queries executed to retrieve information for an Oracle Instance
Based on this inventory job result, Spectrum Protect Plus displays all homes and databases
in the GUI and the user can select them for backup.
Following this inventory, some databases may be marked as "not eligible" and will therefore
not be selectable in the GUI.
Oracle log backup must be discussed with the database administrator to understand if the log
backup and truncation is managed by Spectrum Protect Plus or by another process
implemented by the Oracle database administrator.
Note: Do not combine the management of Oracle database log backup between Spectrum
Protect Plus and another tool.
Spectrum Protect Plus allows, for Oracle, the following possible scenarios:
Backup the Oracle logs and let them on the source system
Backup the Oracle logs and delete them from the source following a successful backup
and after the specified source retention
Do not backup the Oracle logs
Oracle definition of inconsistent backups: Any database backup that is not consistent is
an inconsistent backup. A backup that is made when the database is open or after an
instance failure or SHUTDOWN ABORT command is used is inconsistent. When a database is
restored from an inconsistent backup, the Oracle database must perform media recovery
before the database can be opened, applying any pending changes from the redo logs.
Spectrum Protect Plus performs an "online incremental update" backup of the Oracle
database. While Spectrum Protect Plus is copying the data files, the database accepts
updates. Consequently, it is likely that at the end of the backup, the datafiles are not
consistent. However, Spectrum Protect Plus cope with this situation by using archivelog
mode, meaning after restoring a specific backup, the collected log will be applied to recover
the database to this backup point in time and make it consistent again.
Therefore, IBM Spectrum Protect Plus requires that the Oracle database to be in
ARCHIVELOG mode. If the archive log is disabled the database will be displayed as not
eligible for backup in Spectrum Protect Plus.
Log backup management is controlled via crontab, Spectrum Protect Plus creates entries
(one per database), as shown in Example 2-3.
Example 2-3 crontab entries example for sub-hourly Oracle log backup
sppagent@itso-oracle:~> crontab -l
# DO NOT EDIT THIS FILE - edit the master and reinstall.
# (/tmp/crontab_1621230300 installed on Mon May 17 07:45:00 2021)
# (Cronie version 4.2)
0,15,30,45 * * * *
/opt/IBM/SPP/logbackup/ORATST/df125510cb9a87dbcbba902ea40f0061/logbackup.sh #
Added by SPP
0,55 * * * *
/opt/IBM/SPP/logbackup/SPP/df125510cb9a87dbcbba902ea40f0061/logbackup.sh # Added
by SPP
Spectrum Protect Plus Oracle log backup can be summarized as a six-step process:
1. Mount the vSnap volume to the target application server (only if this is the first log backup
or the mount was removed manually)
2. Start RMAN and connect to the target database
3. Verify that the target database is mounted or open
4. Connect to the database
5. Perform the log backup:
a. Perform the backup of the archive logs
BACKUP AS COPY ARCHIVELOG FROM SCN ##LATEST_KNOWN_SCN##
b. Perform the archive log truncate from the Oracle archive log path - as per the settings
you selected
DELETE NOPROMPT FORCE ARCHIVELOG UNTIL TIME ###YOUR SETTINGS ###
6. Disconnect from the database
Step 5.b depends on how you configured the log management from the Spectrum protect
Plus Oracle backup configuration options. If you decided to purge the log after 4 hours (see
Figure 2-3). Example 2-4 on page 27 shows the kind of command that Spectrum Protect Plus
will trigger.
Figure 2-3 Oracle database backup configuration options for log backup management
Example 2-4 Oracle archive log truncate management - post archivelog backup
Note: If you change the archive log management settings from the Oracle database
backup configuration options (Figure 2-3), the change you made will be applied on the
target system with the next SLA run.
That being said, beyond the 6 steps described above, Spectrum Protect Plus also does
metadata management such as :
Record the latest SCN in the SPP catalog so that this information will be available for the
next log backup.
Create Point in time snapshot of two vSnap volumes simultaneously. First vsnap volume
contains data files control files, spfile copies. The second vsnap volume contains the
oracle log files.
When an SLA backup runs ( data files, control file, SPFile backup ), Spectrum Protect Plus
deals also with archive log cleaning, but this time, for the archive log that has been copied to
the vSnap NFS share as part of the crontab archive log backup. Example 2-5 shows the
command that is run to cleanup the Oracle archive log no longer required in the log backup
destination, that are stored on the vSnap.
Example 2-5 SPP archive log cleanup from archive log backup destination
Incrementally updated backups were introduced in Oracle 10g. Using this feature, all changes
between the SCN (System Change Number) of the original image copy and the SCN of the
incremental backup are applied to the image copy, winding it forward to make the equivalent
of a new data file image copy, without the overhead of actually creating a new image copy.
As mentioned as part of the prerequisites, database archive logs are enabled, which implies
that local cleaning of these logs is required. IBM Spectrum Protect Plus can perform this task.
To control the local archive log cleaning by IBM Spectrum Protect Plus, select the Enable
Log Backup option and specify the Primary log retention parameter in the Backup Options
window, as shown in 1.1.3, “Configuring the database application options” on page 4.
Note: When the log backup is enabled, it will be executed as part of the incremental
backup
Figure 2-4 Specifying the Parallel Streams for Oracle datafiles backup
Note: Oracle Parallel Streams parameter shown in Figure 2-4 is satisfied for both backup
and simple Production restore. It is not used by Instant Access, Test restore, and
Production restore when changing the Oracle Resource Name and/or Destination Path. In
the latter, it means that restore is likely to be slower as it is using only one stream.
Restore modes
There are different restore modes available for Oracle database, that can address different
use cases. They are explained in 1.2, “IBM Spectrum Protect Plus database restore and data
reuse” on page 7.
For Oracle RAC, the restore operation is performed on all nodes in the cluster. The agent
spins up a new database using the data files directly from the vSnap volume. Spectrum
Protect Plus performs a point-in-time restore by first mounting the database from the
preceding backup and then applying archived logs to roll forward to the specified intermediate
point in time.
In this mode, a temporary database running directly off the vSnap backup repository can be
launched without performing any data copy back to production storage.
The production restore database can have its data restored to the same path (directory or
ASM disk group) as the original, or to an alternate path (directory or ASM disk group).
Spectrum Protect Plus performs a point-in-time restore by first restoring the database from
the preceding backup and then applying archived logs to roll forward to the specified
intermediate point in time.
Here are the RMAN steps and instructions being used for Production restore:
1. Mount the vSnap volume that holds the datafiles of the target database
2. Mount the vSnap volume that holds the log files of the target database
3. Copy the datafiles from the vSnap mount to the target directory
4. Recover the database based on archive log:
a. Run:
>STARTUP NOMOUNT PFILE=' oracle target database location '
b. Connect to database
c. Run the following commands:
>CREATE CONTROLFILE REUSE SET DATABASE ' oracle database name ' RESETLOGS
NOARCHIVELOG
>SET LOGSOURCE ' vSnap log backup mount '
>SET AUTORECOVERY ON
>RECOVER DATABASE USING BACKUP CONTROLFILE UNTIL CANCEL
>ALTER DATABASE ARCHIVELOG
>ALTER DATABASE OPEN RESETLOGS
5. Disconnect from database
If the database name and path are not modified, the recovery uses RMAN commands to copy
the data from vSnap to production system (see Example 2-6) and uses the parallel streams
option that was set when configuring the Oracle backup.
Example 2-6 Extract of RMAN logs for Production restore using 3 parallel streams
...
Recovery Manager: Release 12.2.0.1.0 - Production on Tue May 4 14:10:18 2021
3> run {
4> configure controlfile autobackup off;
5> set command id to 'SPP_BACKUP_1056_1620130218';
6> allocate channel spp1 type disk;
7> allocate channel spp2 type disk;
8> allocate channel spp3 type disk;
9> restore database from tag 'SPP_BACKUP_1056';
10> release channel spp1;
11> release channel spp2;
12> release channel spp3;
13> }
14> exit;
echo set off
However, If either the Instance Name or Path is modified (see Figure 2-5), then the process is
using the operating system copy command “cp”. In the latter case, the recovery will use only
one parallel copy stream (see Example 2-7 on page 33).
Figure 2-5 Spectrum Protect Plus Oracle production redirect restore wizard
Example 2-7 shows the kind of command you will observe when doing a restore of the SPP
instance to an alternate location, here in the /home/ORA2 path, as specified in the Restore
wizard shown in Figure 2-5.
It is possible to perform a PIT restore of the Oracle database based on a time stamp or on
a SCN. To recover a database PIT both sources are required, the datafiles and the archive
log files. SPP performs a roll forward recovery of the target database.
Therefore, the most recent datafiles are needed that were created in the past relatively to
the PIT (SCN). In addition the archive log data is needed that contains the PIT (SCN).
Because of this dependency the most recent vSnap snapshot that was taken from the
datafiles, in the past relatively to the PIT (SCN) is used for recovery and the next vSnap
snapshot that was taken from the archive log backup in the future relatively to the PIT
(SCN).
The instant access restore method is meant to have the ability to verify the backup, but not
the recovery of the database.
The user can then perform any custom recovery action using those files, such as manually
copy some files or use RMAN catalog to be able to work with these file and do further specific
recovery, which you cannot do with other Spectrum Protect Plus restore modes.
A good use case for Instant Access could be to retrieve a specific file, such as control file, as
illustrated in Example 2-8, or to execute “CATALOG START WITH <mount point>” to make
RMAN scan the contents of the Instant Access mount, then perform any RMAN supported
granular restore operation.
Figure 2-6 shows the wizard Review menu when doing an Oracle instant access process.
This operation can be done through Create Job → Restore from the Manage Protection →
Databases → Oracle menu, for example.
From the Oracle system side, an NFS share is mounted and can be browsed with regular
operating system commands. You can then read or copy any of the desired files back to your
environment.
Example 2-8 Instant access allows you to access file directly from the Spectrum Protect Plus backup
itso-oracle:~ # df |grep vsnap
10.0.240.223:/vsnap/vpool1/fs181 95060480 195456 94865024 1%
/mnt/spp/vsnap/vpool1/fs181/10_0_240_223
10.0.240.223:/vsnap/vpool1/fs243 96222976 1357952 94865024 2%
/mnt/spp/vsnap/vpool1/fs243/10_0_240_223
itso-oracle:~ # ls -l /mnt/spp/vsnap/vpool1/fs243/10_0_240_223
total 21
drwxr-xr-x 3 oracle oinstall 10 Jun 1 08:06 ORATST
drwxr-xr-x 3 oracle oinstall 10 Jun 1 08:06 SPP
itso-oracle:~ # cp /mnt/spp/vsnap/vpool1/fs243/10_0_240_223/ORATST/controlfile.txt
/tmp/restore_controlefile.txt
Here are the few restore use cases along with the suggested Spectrum Protect Plus restore
mode.
Database duplication
Native RMAN provides a function called RMAN duplicate, Production restore mode of
Spectrum Protect Plus can be used to create a PIT clone of an existing Oracle database.
for more information about requirements, IBM Spectrum Protect Plus Installation and User’s
Guide.
The NFS mount failure can have multiple root causes. For example:
Network issues
vSnap goes down for maintenance
vSnap error
Depending on when the NFS mount failure occurs, the result can be different:
When the database processes the log backup the vSnap share is already invalid (vSnap
down). This situation will cause an alert in SPP, but SPP won't lose log backup data.
When the database processes the log backup, the vSnap goes down in the middle of the
log backup processing, which will cause an alert in SPP and log data can get lost. If the
log files or metadata files are damaged or incomplete, SPP loses log backup data. In this
case PIT restores will fail.
When the database has finished the log backup the vSnap goes down before it takes a
snapshot on the log volume, the data was save written to disk and the snapshot will be
taken later depending on the next snapshot schedule.
Note: If the vSnap server goes down for any reason, such as maintenance for example,
the persistent NFS mount point for log backup becomes invalid and must be unmounted.
The next backup (either SLA or log backup) would mount the NFS share again.
Note: Although database configuration and handling is widely similar for databases in IBM
Spectrum Protect Plus, some differences exist for the supported database systems. We
describe that information in Chapter 4, “Backing up and restoring Db2 databases” on
page 51, and Chapter 5, “Backing up and restoring SQL Server” on page 69
For more information about generic test restore or DevOps use cases, see Chapter 1.,
“Protecting database applications” on page 1. This chapter also describes database
backup, restore, and DevOps use cases in general, but refers specifically to an Oracle
database whenever necessary.
Current restrictions
In IBM Spectrum Protect Plus version 10.1.6, MongoDB is configured as a stand-alone
instance or replica set. Currently, IBM Spectrum Protect Plus does not support backup
operations of MongoDB shared cluster instances. A backup always includes all databases in
the instance.
IBM Spectrum Protect Plus offers a two-stage process to access a MongoDB database. First,
you register the database server with an IP name or address, an operating system user, and
a corresponding password. IBM Spectrum Protect Plus initiates a database discovery job on
this server. If you run your MongoDB without authentication, the database registration in IBM
Spectrum Protect Plus is complete at this point.
Also, if you secured your databases on the database level, you specify more user credentials
for each secured database that IBM Spectrum Protect Plus discovered.
3.1.4, “Register a MongoDB server” on page 43 describes the MongoDB registration in IBM
Spectrum Protect Plus.
If your MongoDB database is configured without credentials, you should secure it. There
are many MongoDB databases open on the internet, providing the opportunity for massive
data breaches.
For more information about available authentication options, see the MongoDB manuals,
which are available at this website.
MongoDB authentication requires the definition of at least one MongoDB user. If database
authentication is enabled, IBM Spectrum Protect Plus must provide a user name and a
password to run backup and restore activities.
For each MongoDB user that you plan to use for backup and restore with IBM Spectrum
Protect Plus, specify MongoDB access roles by using the db.grantRolesToUser() command,
as shown in Example 3-2.
> db.grantRolesToUser("mdbuser",
[ { role: "hostManager", db: "admin" },
{ role: "clusterMonitor", db: "admin" } ] )
> db.grantRolesToUser("mdbuser",
[ { role: "clusterManager", db: "admin" } ] )
The clusterManager role is required only for running test restore operations of replica sets.
If you decide to create a new or dedicated user for backup and restore purposes, you can use
the db.createUser() command, as shown in Example 3-3. According to the MongoDB
manuals, the ClusterAdmin role includes the clusterManager, clusterMonitor, and
hostManager roles.
Example 3-3 Create a MongoDB user with the permissions required by IBM Spectrum Protect Plus
> show dbs
admin 0.000GB
config 0.000GB
local 0.000GB
> use admin
switched to db admin
> db.createUser(
{
user: "mdbuser",
pwd: "mypasswd",
roles: [ "readWrite", "dbAdmin","clusterAdmin" ]
}
)
Note: Enhanced database administration permissions are required to create users and
grant roles. The roles that are required for backup and restore with Spectrum Protect Plus
are not sufficient.
For MongoDB authentication to take effect, restart the MongoDB daemon (mongod) with the
“--auth” option. Example Example 3-4 on page 42 shows how to start the daemon on a Linux
command line.
Create identities
Based on your decision to run your MongoDB database with or without authentication, one or
two user definitions are required: an operating system user and optionally a MongoDB user.
You can specify the users in the Add application server menu, but we recommend explicitly
creating a so-called Identity with a customized name first. Figure 3-1 and Figure 3-2 show
Identities for an operating system and a MongoDB user. The two user names can be identical.
If you want to start a database discovery job on the server, click Get Instances. If IBM
Spectrum Protect Plus discovers databases, it shows the connection data for these
databases: IP name or address, and IP port.
Important: If you run your MongoDB database without authentication, the registration
procedure is complete. However, you should secure your database. If your database is
secured, you must specify more user credentials to access the database. The IBM
Spectrum Protect Plus GUI provides a Set Credential option for the discovered databases
(see Figure 3-3).
For more information about handling of MongoDB databases with authentication, see
section 3.1.3, “MongoDB databases with authentication enabled”.
For more information about required configuration steps and parameters, see IBM Spectrum
Protect Plus Installation and User’s Guide, which is available at IBM Knowledge Center.
For more information about the configuration of other use cases, see 1.2, “IBM Spectrum
Protect Plus database restore and data reuse” on page 7.
After you set up an SLA policy for your MongoDB backup job, you can choose to configure
extra options for that job. More SLA options include running scripts, and forcing a full base
backup.
For more information, see 1.4, “Database backup with pre-script and post-script” on page 16.
Now, you also decide whether to perform a backup of a single database (click Run), use the
Create Job wizard, or perform a backup of all applications that are included in the SLA policy
(click Actions).
Figure 3-4 MongoDB instance discovered by IBM Spectrum Protect Plus with an SLA policy assigned
Wait until a backup is automatically scheduled or scroll down to the SLA policy section in the
window and select Actions → Start to manually start a backup. This process is IBM
Spectrum Protect Plus standard handling, and not specific to MongoDB environments.
To run an on-demand backup job for multiple MongoDB databases that are associated with
an SLA policy, click Create job, select Ad hoc backup, and follow the instructions.
Note: Do not run inventory jobs at the same time that MongoDB backup jobs are
scheduled.
IBM Spectrum Protect Plus mounts a vSnap server directory to the database server to copy
the backup data (see Example 3-5). During the initial backup operation, IBM Spectrum
Protect Plus creates a vSnap server volume and NFS share.
During incremental backups, the created volume is reused. The IBM Spectrum Protect Plus
MongoDB agent mounts the share on the MongoDB server where the backup is performed.
Switch to the Jobs and Operations menu to display the job protocol and optionally download
the job logs and command files.
For more information about the configuration of other use cases, see 1.2, “IBM Spectrum
Protect Plus database restore and data reuse” on page 7.
First, select the database instance and an associated database backup, as shown in
Figure 3-6. Select the available Source Snapshot that needs to be restored, as shown in
Figure 3-7.
The next two selections express what we are trying to achieve: A production restore to the
original instance, as shown in Figure 3-8 and Figure 3-9 on page 48.
For an on-demand snapshot restore of a production database, the IBM Spectrum Protect
Plus restore wizard assumes a subsequent database roll forward to the end of logs included
in the backup (see Figure 3-10).
You must also decide about overwriting a database. IBM Spectrum Protect Snapshot provides
an auxiliary protection against an unintended data overwrite; that is, if the database still exists
and you do not select the overwrite option, the restore job fails.
In IBM Spectrum Protect Plus, an on-demand snapshot restore is not scheduled. Spectrum
Protect Plus runs it only once, as shown in Example 3-11.
Carefully review the job summary that IBM Spectrum Protect Plus displays. If the information
describes what you trying to achieve, run the restore job.
Finally, switch to the Job and Operations menu to check the job results (see Figure 3-12).
IBM Db2 Version 10.5, 11.1, 11.5 and later maintenance levels: Enterprise Server Edition are
supported at the time of this writing.
To manage Db2 databases with IBM Spectrum Protect Plus the following prerequisites must
be met:
Define a dedicated IBM Spectrum Protect Plus agent user, for example sppagent, on every
Db2 server with the required privileges for sudo, as shown in Example 4-1.
Db2 archive logging is activated and Db2 is in recoverable mode, which requires that at
least LOGRETAIN is enabled.
Logical volumes holding IBM Db2 table spaces (data and temporary table spaces), the
local database directory, and IBM Db2 log files are managed by Logical Volume
Management system (LVM2) on Linux and by the Journaled File System (JFS2) on AIX.
LVM2 on Linux and JFS2 on AIX are used for creating temporary volume snapshots.
Ensure that there is at least 10% free capacity for logical volume snapshots.
Each Db2 host has to be registered in IBM Spectrum Protect Plus. In a Db2 DPF
environment with multiple hosts, every Db2 host has to be registered in IBM Spectrum
Protect Plus.
In this publication, the Db2 database example consists of a multi-partitioned Db2 Database
Partitioning Feature (DPF) database version 10.5 that is running on two Red Hat Enterprise
Linux Server hosts, as shown in Figure 4-1.
In our example, the Db2 partitions 0, 1, 2, and 3 are spread over the two servers kansasprod1
and floridaprod1, as shown in the db2nodes.cfg file in Example 4-2.
To be able to manage the Db2 DPF database with IBM Spectrum Protect Plus the parallel
backup mode, as shown in Figure 4-2 has to be enabled. To run parallel backup processing of
partitions in your Db2 environment, ensure that one of the following prerequisites is met:
The Db2 registry variable DB2_PARALLEL_ACS is set to YES, for example: db2set
DB2_PARALLEL_ACS=YES
In earlier versions of Db2, the backup mode is determined by the Db2 registry variable
DB2_WORKLOAD. To enable parallel backup mode, run the Db2 command db2set
Db2_WORKLOAD=SAP. Check with the Db2 command db2set -all Db2_WORKLOAD.
Note: Db2 serial backup mode is not supported with IBM Spectrum Protect Plus because
of the fact that logs included in the backup can be inconsistent across partitions.
Figure 4-2 Parallel backup mode with Db2 Advanced Copy Services (ACS)
IBM Spectrum Protect Plus triggers the Db2 agent once per host, and if there is more than
one partition on the host, Db2 will trigger ACS for each partition individually. A dedicated
protocol file is available per partition that is later stored on the vSnap server volume. The Db2
agent can handle the multiple invocations of its ACS scripted part through Db2.
In parallel backup mode, which is the default mode for an SAP Db2 database, all partitions
are suspended before Db2 issues snapshot requests. The requests are then performed in
parallel on all partitions, as shown in Figure 4-2. IBM Spectrum Protect Plus runs the Db2
backup command on the Db2 catalog partition. The main Db2 ACS processes are:
1. Prepare phase: The write operations of the database are suspended; that is, WRITE
SUSPEND is set automatically on the database. Db2 prepares the file systems, checks
space requirements in the storage system and does other things to keep the database
consistent.
2. Snapshot phase: Db2 instructs the Db2 agent to perform a software snapshot on each
partition in parallel. The snapshot request is done by taking software snapshots of the
corresponding volumes.
3. Verify phase: Db2 checks if the snapshot was taken successfully. If the snapshot is
correct, the data is moved to the vSnap server by the Db2 agent.
For more information about updates to the Db2 database prerequisites, see Spectrum Protect
Plus- All Requirements, which is available at this web page.
After assigning the Db2 database to an SLA policy, as shown in Figure 4-5, you can optionally
click the Select Options button, to enable Log Backup, as shown in Figure 4-6, “Select
options to enable log backup of Db2” on page 58. With log backup enabled, IBM Spectrum
Protect Plus will automatically create a log backup volume and mount it to the application
server.
For IBM Spectrum Protect Plus, the Db2 archive logging must be enabled and Db2 must be in
recoverable mode. If log backup is enabled in IBM Spectrum Protect Plus, one of the Db2
parameters, LOGARCHMETH1 or LOGARCHMETH2, is updated with the path of the vSnap
pool for the log files, as shown in Example 4-3. Therefore, it is important that one of the
LOGARCHMETH parameters includes the value OFF and can be used for a vSnap log
volume assignment.
Note: To successfully enable Db2 log backup in Spectrum Protect Plus, the Db2 agent
expects (and verifies) that all partitions have unique settings for logarchmeth1 and
logarchmeth2.
In the Db2 Backup window, select the Db2 database and click Select Options → Enable Log
Backup → Save, as shown in Figure 4-6, to allow roll forward recovery when you set up a
backup job or SLA policy. When selected for the first time, you must run a backup job for the
SLA policy to activate log archiving to Spectrum Protect Plus on the database.
Log backup transaction files are copied to this share according to the schedule created for log
backup.
If the DB2 backup job is running, we can see an NFS share on the file system that is
associated with the SLA. As shown in Example 4-4, running the vSnap CLI command vsnap
share show lists the active share, in which the Volume ID 3671 and the share name
/vsnap/vpool1/fs148 can be identified.
ID: 3671
NAME: /vsnap/vpool1/fs148
SHARE TYPE: nfs
VOLUME ID: 148
PARTNER ID: N/A
CREATED: 2020-06-30 11:58:46 UTC
UPDATED: 2020-06-30 11:58:46 UTC
SHARE OPTIONS:
ALLOWED HOSTS:
192.168.122.1
192.168.5.94
READ ONLY: No
The share is used to transfer the backup data from the database to the vSnap server.
After the backup of the log completes, log backup transaction files are copied to this share
according to the schedule that was created for log backup, as shown in Example 4-5.
Example 4-5 Log backup transaction files copied into NFS shared VSnap in DB2 guest spp-db2-01
[root@spp-db2-01 ~]# df -h
Filesystem Size Used Avail Use% Mounted on
/dev/mapper/rhel-root 38G 9.8G 28G 27% /
devtmpfs 1.9G 0 1.9G 0% /dev
tmpfs 1.9G 12K 1.9G 1% /dev/shm
tmpfs 1.9G 25M 1.9G 2% /run
tmpfs 1.9G 0 1.9G 0% /sys/fs/cgroup
/dev/sda1 1014M 143M 872M 15% /boot
/dev/mapper/rhel-home 19G 844M 18G 5% /home
/dev/mapper/db2-data 15G 240M 14G 2% /db2_data
/dev/mapper/db2-log 9.8G 116M 9.1G 2% /db2_log
tmpfs 380M 12K 380M 1% /run/user/42
192.168.5.234:/vsnap/vpool1/fs148 898G 128K 898G 1% /mnt/spp/vsnap/vpool1/fs148/192_168_5_234
tmpfs 380M 0 380M 0% /run/user/0
Note: The Run button is enabled only for a single database backup. Also, the database
must have an SLA policy applied.
To run an on-demand backup job for multiple Db2 databases that are associated with an SLA
policy, click Create job. Then, select Ad hoc backup and follow the instructions.
Log in to one of the Db2 database server by using SSH and check where the backup is
created. Run the df -h command, as shown in Example 4-6, and review the vSnap server
volumes.
Example 4-6 vSnap server volumes for data and log backup
[root@spp-db2-01 C0000000]# df -h
Filesystem Size Used Avail Use% Mounted on
/dev/mapper/rhel-root 38G 9.8G 28G 27% /
devtmpfs 1.9G 0 1.9G 0% /dev
tmpfs 1.9G 12K 1.9G 1% /dev/shm
tmpfs 1.9G 26M 1.9G 2% /run
tmpfs 1.9G 0 1.9G 0% /sys/fs/cgroup
/dev/sda1 1014M 143M 872M 15% /boot
/dev/mapper/rhel-home 19G 989M 18G 6% /home
/dev/mapper/db2-data 15G 240M 14G 2% /db2_data
/dev/mapper/db2-log 9.8G 116M 9.1G 2% /db2_log
tmpfs 380M 12K 380M 1% /run/user/42
192.168.5.234:/vsnap/vpool1/fs148 898G 128K 898G 1% /mnt/spp/vsnap/vpool1/fs148/192_168_5_234
tmpfs 380M 0 380M 0% /run/user/0
tmpfs 380M 0 380M 0% /run/user/1003
192.168.5.234:/vsnap/vpool1/fs113 898G 25M 898G 1% /mnt/spp/vsnap/vpool1/fs113/192_168_5_234
One vSnap server log volume is used for multiple Db2 partitions. A single log archive volume
on vSnap server is sufficient because the log paths are orthogonal because of the Db2
NODEXXXX element in each of the log paths. Log volumes stay mounted on the Db2
application server. When the backup completes, you \ see the status Completed, as shown in
Figure 4-8 on page 61.
After you select an SLA policy for your Db2 backup job, you can choose to configure extra
options for that job. Other SLA options include running scripts and forcing a full base backup.
For more information, see 1.4, “Database backup with pre-script and post-script” on page 16.
IBM Spectrum Protect Plus automatically deletes older transactional logs after a successful
database backup. This action ensures that the capacity of the log archive volume is not
compromised by retention of older log files. These truncated log files are stored in the vSnap
server repository until the corresponding backup expires and is deleted. The retention period
of database backups is defined in the assigned SLA policy.
Db2 database restore with IBM Spectrum Protect Plus supports several restore methods that
are explained in the following sections.
Important: For all restore operations, Db2 must be at the same version level on the source
and target hosts. In addition to that requirement, you must ensure that an instance with the
same name as the instance that is being restored exists on each host. This requirement
applies when the target instance has the same name, and when the names are different. In
order for the restore operation to succeed, both instances must be provisioned, one with
original name and the other with the new name.
The combination of these selections define which action to perform, including the following
examples:
Restore a database restore and optionally overwrite an existing database
Establish a copy of a previously backed up database (DevOps)
Get access to the database files (data and metadata) of a previous backup
Note: When you are restoring a multi-partitioned database to an alternate location, ensure
that the target instance is configured with the same partition numbers as the original
instance. All of those partitions must be on a single host.
For more information about database examples that show a test restore or instant access,
see “IBM Spectrum Protect Plus database restore and data reuse” on page 7.
As shown in Figure 4-9, the user must select the Db2 database that requires a restore.
By selecting the blue plus sign, a backup is associated with the database, as shown in
Figure 4-10.
In the IBM Spectrum Protect Plus restore wizard, the following parameters must be selected
to start a traditional database restore that overwrites the database:
Restore type: On-Demand Point in Time
Restore location type and location can vary. Here, we use Site and Primary.
Restore method: Production
Destination: Restore to original Instance
Restore Method: Production
Job options: Overwrite existing database
The next step is to select the type of restore, as shown in Figure 4-11. Here, On-demand
Point in Time was selected.
Then, choose a restore location, as shown in Figure 4-12. These settings depend on your
specific environment, which can include an object storage or vSnap server location, or a
secondary site that you use for replication. In our example, we chose Site and Primary, as
shown in Figure 4-12.
There are three restore methods available, as shown in Figure 4-13. In our scenario we are
choosing a Production restore.
Production restore
A production restore either overwrites the original database or creates a database copy on an
alternate host and optionally in an alternate database instance.
As shown in Figure 4-14, click Restore to original instance to restore to the Db2 production
server.
For Devops scenarios, it is possible to create a restore job that runs periodically at a specific
time. In our scenario, we create an on-demand restore job that runs only once. As a further
restore job option, we select Overwrite existing databases, as shown in Figure 4-15.
Another option, as shown in Figure 4-16, is to provide pre- and post-scripts that perform
specific actions before and after the Db2 restore. Those scripts must be uploaded to IBM
Spectrum Protect Plus before creating the restore job.
Finally, the Review page is displayed and after checking all values, the on-demand restore job
can be submitted. See Figure 4-16. To start the on-demand restore job, click Submit.
The restore job can be monitored by selecting Jobs and Operations → Running Jobs, as
shown in Figure 4-18. When the restore job finishes, it is removed from the Running Jobs list.
Figure 5-1 Windows Remote Shell configured into Microsoft SQL Server
A Microsoft iSCSI Initiator service must be enabled and running on the Microsoft SQL
server system, as shown in Figure 5-2.
An IBM Spectrum Protect Plus agent user must have “Log on as a service” rights on the
SQL application server.
The login credentials must have public and sysadmin permissions enabled, plus
permission to access cluster resources in a SQL Server AAGs environment.
To perform log backups, the SQL Server agent service user must be a local Windows
administrator and must have the sysadmin permission enabled to manage SQL Server
agent jobs.
The host name of the IBM Spectrum Protect Plus appliance should be resolvable from the
SQL application servers.
The Microsoft SQL Server Guest Network Adapter Backup must have the option “Client for
Microsoft Networks” enabled to prevent CIFS share issues, when Databases SQL Backup
Logs are defined and configured, as shown in Figure 5-3.
Figure 5-3 Client for Microsoft Networks option enabled into MIcrosoft SQL Server
For more information about the SQL Server database prerequisites, see IBM Spectrum
Protect Plus- All Requirements, which is available at this web page.
In the IBM Spectrum Protect Plus GUI navigation pane, click Manage Protection →
Databases → SQL → Manage Application Servers → Add Application Server.
Enter the required login credentials for the SQL application server, as shown in Figure 5-4. In
this example, the IBM Spectrum Protect Plus admin includes predefined the SQL Server
Admin in Accounts → Identity → Add Identity.
Perform a configuration test of the newly assigned SQL Server in IBM Spectrum Protect Plus,
as shown in Figure 5-5 on page 73.
If SQL application servers are attached to a domain, a user name in the format domain\Name
must be used. If a user is a local administrator, the format .\<local administrator> must be
used.
For failover clusters and AAGs, each node must be registered by name or IP address. If fully
qualified domain names are used, they must be resolvable and routeable from IBM Spectrum
Protect Plus.
Assign the selected SQL Server database to an SLA policy to create a backup job. SQL
Server backups run in a “Base-Once-Incremental-Forever” scheme. During the initial base
(full) backup, IBM Spectrum Protect Plus creates a vSnap server volume and mounts it to the
SQL application server over iSCSI.
Note: An iSCSI route must be enabled between the SQL Server and vSnap server. For
more information, see this web page.
Optionally, the SQL Server admin can click the Select Options button to enable Log Backup,
as shown in Figure 5-6. With log backup enabled, IBM Spectrum Protect Plus manages the
log backup by using the SQL Server agent service.
To complete log backups, the SQL Server Agent service user must be a local Windows
administrator and must have the sysadmin permission enabled to manage SQL Server agent
jobs. Also, the SQL VSS Writer service running on the local SQL Server system must be
started from a local system user.
The agent uses the administrator account to enable and access log backup jobs. The IBM
Spectrum Protect Plus SQL Server agent service user must also be the same as the SQL
Server service and SQL Server agent service account for every SQL Server instance to be
protected.
Set the maximum number of data streams per database to the backup storage. This setting
applies to each database in the job definition. Databases can be backed up in parallel if the
value of the option is set to 1. Multiple parallel streams might improve backup speed, but high
bandwidth consumption might affect overall system performance.
The SQL Server backup job status can be monitored in the Jobs and Operations →
Schedule panel, as shown in Figure 5-7.
It also can be monitored by selecting Jobs and Operations → Running Jobs → Progress,
as shown in Figure 5-8.
Note: The Microsoft SQL Server agent sets the VSS backup type to COPY_ONLY for all
database backups.
Depending on what type of SQL backup log is required, it can be configured by using one of
the following methods:
With Truncate SQL Logs option activated on Virtualized Systems wizard
Note: If you multiple backup solutions are performing log truncation, you can establish
discontinuity in the log chain. It must be ensured that the log truncation occurs only once
during a backup.
With this option activated, logs might be truncated during the VM Backup as a result of log
clearing. In this case, you can restore a VM only; a roll forward of the transaction log data
cannot be performed.
The option to truncate SQL logs can be defined under Manage Protection → Virtualized
Systems → VMware or Hyper-V → Select VM Server → Select Options, as shown in
Figure 5-9.
Under Agent Options, select the Truncate SQL Logs option, as shown in Figure 5-10. Click
Save.
Note: For more information about how to enable Log Truncation, see Protecting Virtualized
Systems - Backing up Vmware / Hyper-V data Guides, which is available at IBM
Knowledge Center:
Backing up VMware data
Backing up Hyper-V data
The option to enable SQL log backup can be defined under Manage Protection →
Databases → SQL → Select SQL Instance → Select Options, as shown in Figure 5-11.
Under Options, select Enable Log Backup and define a Log Backup Frequency, as
shown in Figure 5-12.
The enabled SQL log backup schedule option can also be reviewed in the Microsoft SQL
Server system, as shown in Figure 5-13 on page 80 under Task Scheduler → Task
Scheduler Library → IBM → SPP Windows Agent.
Note: To run the Windows log backup task, the IBM Spectrum Protect Plus agent user
must have the Log On As Batch Job assignment privilege.
Note: For more information about how to enable Log backup, see Backing Up SQL Server
Data Guide, which available at IBM Knowledge Center.
If the SQL backup SLA job is running, you can see a share smb on the file system, which is
associated with the SLA. The vSnap CLI command vsnap share show, as shown in
Example 5-1, lists the active share where the Volume ID 1 and the file system name
/vpool1_fs2 can be identified.
ID: 1
NAME: vpool1_fs2
SHARE TYPE: smb
VOLUME ID: 2
PARTNER ID: N/A
CREATED: 2020-06-03 12:06:15 UTC
UPDATED: 2020-06-18 06:00:57 UTC
SHARE OPTIONS:
ALLOWED HOSTS:
10.0.250.46
READ ONLY: No
The shared volume is used to transfer the backup data from the database to the vSnap
server.
After the log backup completes, log backup transaction files are copied to this share, as
shown in Example 5-2.
One ad-hoc job can be started from Manage Protection → Databases → SQL → Create
Job → Ad hoc Backup → Select SLA Policy → Select Source, as shown in Figure 5-14.
Figure 5-14 Ad Hoc SQL Backup showing Name, Location and SLA Policy
Multiple sessions from the same SLA policy can be started from Manage Protection →
Databases → SQL → Create Job → Ad hoc Backup → Select SLA Policy → Select
Source. The sessions can be monitored from Jobs and Operation → Running Jobs, as
shown in Figure 5-15.
Figure 5-15 Multiple running sessions from the same SLA policy
Note: Only users with administrator credentials can manage global preferences.
Enable SQL Server databases restored in test mode eligible for backup
When this option is selected, SQL Server databases that were restored in test mode are
available for selection in the SQL Backup pane or ad hoc backup wizard.
Allow SQL database backup when transaction log backup chain is broken
Run a database SLA backup job when IBM Spectrum Protect Plus detects a break in the
log backup chain for a database.
Rename SQL data and log files when database is restored in production mode with new
name
This options allows to rename SQL database and log files files during a production or test
restore job. This field applies only when a new database name is provided during an SQL
database restore job.
IBM Spectrum Protect Plus treats data reuse and data recovery as a restore activity. In both
cases, you must create a restore job. A restore job can be started by making one of the
following selections in IBM Spectrum Protect Plus:
Manage Protection → Databases → SQL → Create Restore Job
Jobs and Operations → Create Restore Job → Restore
The parameters that you select during backup job creation define which is performed.
The following main parameters control the final restore or data reuse activity:
Type of Restore:
– On-Demand Snapshot
– On-Demand Point in Time
– Recurring
Restore Method:
– A production restore either overwrites the original database or creates a database copy
with a different database name. In the latter case you must specify a new database
name and the destination paths.
– A test restore mounts the vSnap server directories with a database backup to a
database server, recovers and opens the database. You can chose to rename the
database.
– An instant access restore also mounts the vSnap server directories with a database
backup to a database server, but does not recover or open the database. An instant
access restore of an Always On database is restored to the local destination instance.
Note: The SQL Server system databases (master, msdb, model) can be restored only
with Instant Access mode in IBM Spectrum Protect Plus.
Destination:
– Restore to the original instance
– Restore to an alternate instance
The combination of these selections define which action to perform, including the following
examples:
Perform a database restore and optionally overwrite an existing database
Establish a copy of a previously backed up database (DevOps)
Get access to the database files (data and metadata) of a previous backup
In the first example scenario, a Production restore of a SQL Server stand-alone database is
performed by using SQL Server version 2012. As shown in Figure 5-17 on page 85, the
databases ESCC and IBM2 are selected for the restore. By selecting the blue plus sign, a
backup is associated with the database.
For more information about database examples that show a test restore or instant access,
see “IBM Spectrum Protect Plus database restore and data reuse” on page 7.
Other restore parameters that must be specified are shown in Figure 5-18:
Restore type: On-Demand: Snapshot
Runs a one-time restore job from a database snapshot. The restore job starts immediately
upon the completion of the wizard.
Restore location type: Site
The site where snapshots were backed up. The site is predefined in IBM Spectrum Protect
Plus.
Location = Primary
The primary site location from which to restore snapshots.
In production mode, the agent first restores the files from the vSnap server volume back to
primary storage and then creates the new database by using the restored files. Select
Production, as shown in Figure 5-19 and then, click Next.
When selecting production mode, you can also specify a new folder for the restored database
by expanding the database section and entering a new folder name.
Enable the restore job to overwrite the selected database. By default, this option is not
enabled, as shown in Figure 5-21 on page 87.
Note: Before you run restore operations in an SQL Server Always On environment by
using the production mode with the Overwrite existing databases option, ensure that the
database is not present on the replicas of the target availability group. As a prerequisite,
manually clean up the original databases (to be overwritten) from all replicas of the target
availability group.
In the Review page, check all entered restore job parameters, as shown in Figure 5-22. Click
Submit to start the on-demand restore job.
IBM Spectrum Protect Plus mounts the vSnap server backup volume at the SQL application
server and copies the backup data to the source.
In our example that is shown in Figure 5-23, the vSnap server backup volume is mounted as
Disk1 during the restore job.
Figure 5-23 Mount of vSnap server volume for the restore on the SQL application server
As shown in Figure 5-24, we select the new Standby mode Job Options and perform the
restore similar to the previous example.
IBM Spectrum Protect Plus mounts the vSnap server backup volume at the SQL application
server and copies the backup data to the source. Figure 5-25 shows that the new Database
IBM2_TEST was restored with Standby/Read-Only mode.
This chapter describes how to set up IBM Spectrum Protect Plus to protect Microsoft
Exchange Servers, and explores common scenarios and best practices. It includes the
following topics:
6.1, “Microsoft Exchange server” on page 92
6.2, “Prerequisites for protection in IBM Spectrum Protect Plus” on page 94
6.3, “IBM Spectrum Protect Plus configuration for Exchange” on page 101
6.4, “Backup jobs overview” on page 103
6.5, “Restore jobs” on page 111
Edge/Transport X X X
Client Access X
Mailbox X X X
The Edge/Transport role is used to transport mail from external sources into the Exchange
infrastructure. A server with installed Edge/Transport is usually placed in a specific secured
firewall zone because it is directly connected to the internet. If this role is the only role that is
installed on the server, the server needs no Exchange-specific protection (because it has no
persistent user data; it acts only as a proxy).
The Client Access role is a separate role in Exchange 2013 and was merged into the mailbox
role in Exchange 2016 and 2019. If a server is installed with Client Access Server role only,
the same type of protection applies that is used for Edge/Transport only servers.
Only Microsoft Exchange servers with installed Mailbox role are protected by IBM Spectrum
Protect Plus Backup and Restore for Microsoft Exchange. These servers are usually called
mailbox servers. In the IBM Spectrum Protect Plus GUI, they are referred to as Application
Servers.
Database Availability Groups (DAG) are a group of Mailbox Servers in the same Exchange
domain that share multiple copies of Mailbox databases. Up to 16 copies of a single
Exchange database can exist. However, only one copy is active, meaning that users are
working on this copy and changes are applied to this copy. The other copies are updated by
shipping the committed Exchange log files from the active copy to the other copies. The
inactive copies are showing a healthy status if the log replication is working.
A ReplayLagTime and TruncationLagTime can be defined for every copy to ensure that the
copy does not commit or truncate the replicated logs before the ReplayLagTime and
TruncationLagTime are reached. The default value of these two parameters is 0 seconds and
the maximum value is 14 days.
A database copy with default settings is a nearly real-time copy (there is always the gap of the
active log file, which is not shipped to inactive copy yet) of the active copy.
For example, a database copy with a ReplayLagTime of 7 days is a copy that lags the active
copy by 7 days. A lagged copy ensures that if the active database copy becomes corrupted, a
working copy (7 days back in time) is still available that can be used to fix the corruption or be
used as a new base to apply the logs until the corruption occurred.
IBM Spectrum Protect Plus adds data protection capabilities that can be used whenever the
built-in solutions are not satisfying or in case of a disaster.
IBM Spectrum Protect Plus is a zero touch data protection product; therefore, no installation
on the Exchange Mailbox Servers is needed. However, some requirements must be met to
enable IBM Spectrum Protect Plus to access Exchange Mailbox Servers and perform backup
or restore tasks.
This Windows system can be one of the Exchange Mailbox servers, but Microsoft advises
against installing Outlook on an Exchange Mailbox Server. Therefore, it is best to use a
separate Windows server.
To use the remote management features, you must first install and enable Windows
PowerShell 3.0, or later, on all IBM Spectrum Protect Plus protected Exchange servers and
the remote server from which you intend to run the IBM Spectrum Protect Plus MMC GUI.
To download, install, and enable the software, follow the instructions in Microsoft Windows
Management Framework 3.0 Downloads. The remote server and Application server must be
in the same domain.
This installation is called Granular remote package. The installation steps be found in the
readme file for the Spectrum Protect Plus MMC GUI, which we included here for
convenience.
Installation steps
Deploy the granular restore package to a remote server that has Microsoft Outlook installed.
The following installation steps are performed only once. After the granular restore package is
installed, you can continue to use it to perform later granular restore operations:
1. Copy the granular restore package, which is in C:\Program Files\IBM\IBM Spectrum
Protect Plus\tools\exchange\imr\<version>TIV-TSMEXC-Win.ex, from the Application
Server to the remote server from where you manage the granular restore operations. Also,
note that <version> indicates the version.
2. On the remote server, run the following commands to install the package (these
commands assume that you copied to the C:\temp directory):
a. Create the installation diagnostic folder:
mkdir C:\temp\diag
b. Install MMC GUI and granular components:
C:\temp\imr\install_imr.bat *-TIV-TSMEXC-Win.exe 10.1.7 c:\temp
Where * is the MMC GUI version.
3. Configure the remote connection between the remote server and Application server:
a. Verify that the Windows Firewall allows inbound connections on the remote server.
b. Set the hostnames for the remote server and respective for the Application server.
The Application server runs the Exchange server and the remote server performs the
granular restore operation.
$remote_server_host_name = "outlook1.domain.org"
$app_server_host_name = "exchange1.domain.org"
4. Enable remote management for the MMC GUI that is deployed with IBM Spectrum Protect
Plus entering the following Windows PowerShell command:
Enable-PSRemoting -Force
Depending on your environment, you might need to add trusted hosts to the Exchange
Server and server where the MMC GUI is deployed:
a. Add the Application Server and remote server to the trusted hosts list by running the
following command on each system:
Set-Item WSMan:\localhost\Client\TrustedHosts -Value
“$remote_server_host_name,$app_server_host_name” -Force
b. Restart the winrm service by running the following command:
Restart-Service winrm
5. Enable the Windows PowerShell Remoting feature with Credential Security Support
Provider (CredSSP) authentication. Complete the following steps:
a. On the remote server, run the following command to enable the Windows PowerShell
Remoting feature with CredSSP:
Enable-WsmanCredSsp -Role Client -DelegateComputer $app_server_host_name
-Force
b. On the Application Server that runs the granular restore operation, run the following
command to enable the Windows PowerShell Remoting feature with CredSSP:
Enable-WsmanCredSsp -Role Server -Force
6. Verify that the Windows PowerShell Remoting feature is configured by using one of the
following methods: (use the Test-WSMan cmdlet to test whether the WinRM service is
running on the remote computer):
a. On the remote server, run the following cmdlet to verify that the Windows PowerShell
Remoting feature is configured correctly:
Test-WSMan $app_server_host_name
b. On the Application Server, run the following cmdlet to verify that the Windows
PowerShell Remoting feature is configured correctly:
Test-WSMan $remote_server_host_name
Optionally, for more remote configuration verification, complete the following steps:
1. Set the credentials object you used. Usually, this credential is a domain administrator:
$creds = Get-Credential
a. On the Application Server and remote server, run the following cmdlet to verify basic
remote connection:
Invoke-Command -ComputerName $remote_server_host_name -ScriptBlock { pwd }
-Credential $creds
For our example, we show the commands that run in our test environment, which consists of
the following servers:
Windows 10 server ("windows10.xxxxxxx.lab,192.168.111.66"), as shown in
Example 6-1
Exchange server ("epc-exchange.xxxxxxx.lab,192.168.111.167"), as shown in
Example 6-2 on page 97
wsmid : http://schemas.dmtf.org/wbem/wsman/identity/1/wsmanidentity.xsd
ProtocolVersion : http://schemas.dmtf.org/wbem/wsman/1/wsman.xsd
ProductVendor : Microsoft Corporation
ProductVersion : OS: 0.0.0 SP: 0.0 Stack: 3.0
wsmid : http://schemas.dmtf.org/wbem/wsman/identity/1/wsmanidentity.xsd
ProtocolVersion : http://schemas.dmtf.org/wbem/wsman/1/wsman.xsd
ProductVendor : Microsoft Corporation
ProductVersion : OS: 0.0.0 SP: 0.0 Stack: 3.0
cfg : http://schemas.microsoft.com/wbem/wsman/1/config/client/auth
lang : en-US
Basic : true
Digest : true
Kerberos : true
Negotiate : true
Certificate : true
CredSSP : true
Path PSComputerName
---- --------------
C:\Users\Administrator.xxxxxxx\Documents epc-exchange.xxxxxxx.lab
Path PSComputerName
---- --------------
C:\Users\spp\Documents epc-exchange.xxxxxxx.lab
WSManConfig: Microsoft.WSMan.Management\WSMan::localhost\Client
System.String TrustedHosts
windows10.xxxxxxx.lab,192.168.111.66
wsmid : http://schemas.dmtf.org/wbem/wsman/identity/1/wsmanidentity.xsd
ProtocolVersion : http://schemas.dmtf.org/wbem/wsman/1/wsman.xsd
ProductVendor : Microsoft Corporation
ProductVersion : OS: 0.0.0 SP: 0.0 Stack: 3.0
wsmid : http://schemas.dmtf.org/wbem/wsman/identity/1/wsmanidentity.xsd
ProtocolVersion : http://schemas.dmtf.org/wbem/wsman/1/wsman.xsd
ProductVendor : Microsoft Corporation
ProductVersion : OS: 0.0.0 SP: 0.0 Stack: 3.0
cfg : http://schemas.microsoft.com/wbem/wsman/1/config/service/auth
lang : en-US
Basic : false
Kerberos : true
Negotiate : true
Certificate : false
CredSSP : true
CbtHardeningLevel : Relaxed
Path PSComputerName
---- --------------
C:\Users\spp\Documents windows10.xxxxxxx.lab
Path PSComputerName
---- --------------
C:\Users\Administrator\Documents windows10.xxxxxxx.lab
Running the MMC GUI on the remote server to perform granular restore
To perform a granular restore by using the MMC GUI, complete the following steps:
1. Start the MMC GUI application:
C:\Program Files\Tivoli\FlashCopyManager\FlashCopyManager.exe
2. To add the Application Server in MMC GUI, click Actions → Manage Computers to open
Manage Computers window.
Click the plus-sign icon (+) in the Computers pane and enter an Application Server name.
Click Set Account and enter user credentials for the Application Server, as shown in
Figure 6-1.
Use the same credentials as used by Spectrum Protect Plus GUI. Select Manage
Protection → Databases → Exchange → Backup → Manage Application Servers →
Add Application Server.
3. Test the connection to the Application Server. Select Application Server node and then,
click the Test Connection tab.
From this output, consider the information about CredSSP, which indicates that the
configuration is successful and a connection is possible.
4. Click OK to close the Manage Computers window.
To configure an exchange server in IBM Spectrum Protect Plus, start the IBM Spectrum
Protect Plus Server GUI and log in to the dashboard.
Figure 6-5 shows how to add or edit an Application Server. The Host address can be the
Server name or IP address. The User ID must be entered for the first Exchange Server with
the domain or user ID and password. For any other Application Servers, the same user ID can
be used by selecting the Use existing user option and then, selecting it from the drop-down
menu.
The Maximum concurrent database number (default: 10) is used to reduce or raise the
number of mailbox databases that are backed up concurrently. In production environments,
the default of 10 is a good starting point and is raised incrementally only to avoid overloading
the Microsoft Exchange Server.
After the Application Server is registered, the Mailbox databases on the server can be
browsed and SLAs can be assigned.
Figure 6-6 shows an example of a Microsoft Exchange server with two Mailbox databases
that are not part of a Database Availability Group (DAG). These databases can be selected
and assigned to an SLA individually.
The Run Inventory button can be used to immediately query the Microsoft Exchange Server
for a list of databases and their status. The list also indicates whether the Mailbox database is
using circular logging or not. Databases with disabled circular logging are flagged with Yes in
the Eligible for Log backup column. Databases can also be filtered by using the Search box or
the view can be switched from Standalone/Failover Cluster to a list of DAG enabled Mailbox
databases.
To configure the protection of DAG-enabled Mailbox databases, the view must be changed to
“Database Availability Groups”.
By default, the backup is performed on the active copy, which might interfere with the
Exchange user workload. To switch the protection to a passive copy in Options, select the
Backup preferred node. This option can be pointed to the Exchange Mailbox Server with the
lowest activation preference. This setting ensures that the backup is performed on the
passive copy, which is the last copy to take over the Active copy role in the cluster.
Generally, it is preferred to create dedicated SLA policies for single databases or for groups of
logically related databases.
Select an Exchange instance to back up all the data in that instance. Optionally, you can click
an instance name and then, select individual databases that you want to back up.
Three choices are available: Run, Select an SLA policy, and Select options, as shown in
Figure 6-9.
Click Select an SLA Policy. Predefined choices are: Gold, Silver, and Bronze. Each choice
includes different frequencies and retention rates, as shown in Figure 6-10.
Gold is the most frequent with the shortest retention rate. You can also create a custom SLA
policy or edit a policy, as we did by selecting the Exchange_Silver SLA policy, as shown in
Figure 6-10. Click Save to confirm your choice.
Now, the SLA selection can be verified and options can be defined for the scheduled backup
job by clicking Select Options, as shown in Figure 6-11.
You can define options for your backup, such as enabling log backups for future recovery and
specifying the parallel streams to reduce the time necessary to back up large databases (see
Figure 6-12). Click Save.
Configure the SLA policy by clicking the icon in the Policy Options column of the SLA Policy
Status table, as shown in Figure 6-13.
After clicking the icon, a pop-up window appears, as shown in Figure 6-14, in which you can
configure more policy options.
To run the policy outside of the scheduled job, select the instance or database and then, click
Actions → Start.
The status changes to Running for your chosen SLA. To pause the schedule, click Actions →
Pause Schedule. To cancel a job after it starts, click Actions → Cancel (see Figure 6-15).
2. You are presented with a choice for Ad hoc backup or Restore. Select Ad hoc backup
(see Figure 6-17).
4. Select a predefined SLA policy. After clicking the SLA policy, the defined values for that
policy are shown (see Figure 6-19).
5. Select the database to back up. If many databases are available, use the search function
to easily find the wanted database. Now, the database can be added to the backup job list
by clicking the blue plus sign (+), as shown in Figure 6-20.
6. Review the backup job options. Then, click Submit to start the job, as shown in
Figure 6-21.
7. As shown in Figure 6-22, a message is displayed to confirm that the job was submitted.
Click OK to close the message.
The Ad hoc backup job can be monitored under the Jobs and Operations pane, as
shown in Figure 6-23.
Two options are available to restore Microsoft Exchange data. It is possible to recover a
complete Exchange Database into any database or Recovery Database (RDB) or to recover
individual items, such as mailboxes or individual emails.
In both cases, you find the entry point for the procedure in the Jobs and Operations panel or
the Manage Protection panel, as shown in Figure 6-24.
To create the restore job, select Create Job in the Manage Protection - Exchange menu.
Then, select Restore, as shown in Figure 6-25.
– On-Demand Point-in-Time
In this case, the available restore location sites are: Demo, Primary, and Secondary
(see Figure 6-29).
– Recurring
Choose a restore location type. As shown in Figure 6-30, different location types from
where the restore can be taken are also available: Site, Cloud service, Repository
server, Cloud service archive, and Repository server archive.
3. For our scenario, we select On-Demand: Snapshot. After selecting one of the available
backups, click Next. You are presented with the next step, which is to select amongst two
restore methods: Complete Restore or Item Recovery.
In Production or Test mode, enter a new database name. In the panel for production, the
destination path can be changed, as shown in Figure 6-32.
In our example, we proceed with the restore into production, as can be the case in a situation
where the source database is corrupted and must be replaced.
After clicking Next, set the destination and choose restore into the original instance, as shown
in Figure 6-33.
By clicking Next, other job options are available that are necessary for the recovery. The
choice here is: No Recovery or Recover until end of backup.
The options Recover until end of available logs and Recover until specific point-in-time
are not available for this type of restore because no log backups are available.
Click Next to proceed. The last panel displays a summary for review, as shown in
Figure 6-35. Click Submit to start the restore job.
Under Jobs and Operations, you can follow the progress of the operation, as shown in
Figure 6-36.
After completion, the job information is moved into the Job History window, as shown in
Figure 6-37.
Refer to the prerequisites, described in 6.2.1, “Granular restore remote package installation”
on page 94.
In both cases, the restore procedure is started as described in 6.5, “Restore jobs” on
page 111.
This type of restore uses a Recovery Database (in our case, MDB1.RDB). The rest of the
procedure is similar to what is described in 6.5.1, “Complete Restore” in successively setting
destination and job options and then submitting the job.
After the restore job is started, it can be monitored in the Jobs and Operations panel.
The recovery database is created and the snapshot is mounted as Recovery Database
(RDB). An excerpt from the job login Example 6-3 shows the steps that are performed.
Clicking the information icon (as indicated by the arrow in Figure 6-39) in the Type column
provides more information about how to start the IBM Spectrum Protect Plus MMC GUI. This
GUI is automatically installed during the restore procedure on the Exchange server (see
Figure 6-40).
Figure 6-40 Information how to start the IBM Spectrum Protect MMC GUI
You must decide which target to use to proceed with the item recovery: install the MMC GUI in
combination with Outlook 2016 on the Exchange Server or run it on a separate server.
3. The IBM Spectrum Protect MMC GUI opens. The next step is to check with the wizard that
all prerequisites are fulfilled. Click IBM Spectrum Protect Plus → Dashboard →
Manage → Configuration and start the Configuration Wizard, as shown in Figure 6-42.
4. Click Wizards and the configuration option IBM Spectrum Protect Plus configuration is
shown. Click Start to run the wizard. The result should display failed: 0, as shown in
Figure 6-43.
5. Click Next and the wizard proceed and completes the process.
The warning about VSS Provider Check can safely be ignored because no IBM VSS
Hardware Provider is installed and it is not necessary when restoring from IBM Spectrum
Protect Plus.
The restart required warning often is caused by a pending restart (most likely after a
Windows patch update on the operating system).
6. After completing the configuration wizard, proceed to recover single mailbox items.
7. Expand the Protect and Recover Data tab and select the Exchange server. On the right
side of the display, three tabs are available: Protect, Recover, and Automate. Click the
Recover tab.
A Configuration Error appears, as shown in Figure 6-44. This error is shown because it is
not recommended to perform the recovery with the exchange server.
8. Click ReadMe to see information that is similar to the information that was described in
6.2.1, “Granular restore remote package installation” on page 94 (see Figure 6-45).
9. The Recovery Database (RDB) opens, but no mailbox is selected. The mailboxes appear
as closed. Proceed with the recovery by selecting the Mailbox Restore Browser view, as
shown in Figure 6-46.
By clicking the mailbox icon (in our example, SPP), the mailbox is populated and the items
are provided for recovery. This process can take some time.
10.The populated mailbox now shows the mailbox items, such as inbox (see Figure 6-47).
Click the inbox and all mail objects are shown. By selecting individual mail items, the
content is shown in the middle part of the window.
In the Actions column on the right side of the display, the choices for the recovery are
listed. The column is divided into Folder Actions and Message Actions sections. We can
recover folders or single messages.
11.Click the Restore Messages to Original Mailbox entry. The restore from the Recovery
Database (RDB) goes done into the active database. The restore progress and the result
are displayed in a separate window, as shown in Figure 6-48.
After successful recovery, a cleanup procedure must be completed on the IBM Spectrum
Protect Plus Server. This cleanup can be done in Jobs and Operations → Active Jobs, by
cancelling the running job, as shown in Figure 6-49.
Item Recovery restore job also can be stopped is in Jobs and Operations → Active
Resources by clicking the three vertical dots and selecting Cancel job, as shown in
Figure 6-50.
The Job History Job Logs includes the detailed log of the cleanup procedure and is confirmed
with a success message, as shown in Figure 6-51.
The Exchange server must be added as a managed computer so that it appears in the
Group → Dashboard view, as shown in Figure 6-52.
After expanding the Protect and Recover Data entry, the Mailbox Restore Browser shows the
available mailbox items in Recovery Database (RDB) that are connected to the Exchange
Server and provided through the PowerShell communication.
The recovery procedure on a remote system is identical to the recovery procedure on the
Exchange Server, as described in “Item Recovery by using an Exchange Server” on
page 119.
Related publications
The publications listed in this section are considered particularly suitable for a more detailed
discussion of the topics covered in this paper.
IBM Redbooks
The following IBM Redbooks publications provide additional information about the topic in this
document. Note that some publications referenced in this list might be available in softcopy
only.
IBM Spectrum Protect Plus Practical Guidance for Deployment, Configuration, and
Usage, REDP-5532.
You can search for, view, download or order these documents and other Redbooks,
Redpapers, Web Docs, draft and additional materials, at the following website:
ibm.com/redbooks
Online resources
The following websites are also relevant as further information sources:
Featured Documents for IBM Spectrum Protect Plus:
https://www.ibm.com/support/pages/featured-documents-ibm-spectrum-protect-plus
IBM Spectrum Protect Plus BluePrints:
https://ibm.biz/IBMSpectrumProtectPlusBlueprints
IBM Spectrum Protect Plus Documentation:
https://www.ibm.com/docs/en/spp
IBM Spectrum Protect Plus Support:
https://www.ibm.com/mysupport/s/topic/0TO50000000IQWtGAO/spectrum-protect-plus?
language=en_US&productId=01t50000004uZGc
IBM Spectrum Protect Plus - All Requirements Doc:
https://www.ibm.com/support/pages/ibm-spectrum-protect-plus-all-requirements-do
c
REDP-5640-00
ISBN DocISBN
Printed in U.S.A.
®
ibm.com/redbooks