Backing Up Applications With Networker Modules: Emc Proven Professional Knowledge Sharing 2009
Backing Up Applications With Networker Modules: Emc Proven Professional Knowledge Sharing 2009
Backing Up Applications With Networker Modules: Emc Proven Professional Knowledge Sharing 2009
NetWorker Modules
Including
EMC EMC
Proven Proven™ Professional
Professional Certification
Knowledge Sharing 2009
Aaron Kleinsmith
P&E Consultant, EMC Education
EMC²
[email protected]
Backing up Applications with NetWorker Modules
Aaron Kleinsmith
P&E Consultant, EMC Education
EMC²
[email protected]
Disclaimer: The views, processes or methodologies published in this compilation are those of the authors. They do not
necessarily reflect EMC Corporation’s views, processes, or methodologies
This article will not address Snap-shot back up and recovery, but rather the more
traditional online (hot) and offline (cold) backup methods. Also, we will not address how
to backup and protect applications or databases that are protected with clustering
software.
Audience
This article benefits EMC customers who are NetWorker backup administrators, EMC
system integrators, or EMC employees interested in backing up applications such as
Oracle, Microsoft SQL, Microsoft Exchange and SAP R/3 with Oracle. These are some
of the more commonly deployed NetWorker Modules.
Executive Summary
There is information available relating to individual NetWorker Modules, but no single
document offers get a complete picture of how to back up multiple types of databases
likely to be found in an environment. This article will help to fill a need by addressing
backup and recovery of some popular application types with NetWorker.
We will discuss traditional back up methodologies such as online (hot) backups, offline
(cold) backups, and transaction log backups. This article describes the backup
procedure to capture consistent copies of an online application from the source disks
using the application servers’ resources that manage the primary copy of the data.
Application Backups
Current management strategies and technologies for applications have evolved to a
point where it is commonly accepted that an application (application or database) will be
available all the time. There are exceptions to the seven days a week, 24 hours a day
(7x24) requirement, but 15-20 years ago it was common to shut down every application
server when it required a backup. Some related problems included not having enough
technologically advanced hardware in place, impact on server resources were
unbearable during the back up, cumulative days of log backups unreasonable for
recovery, or simply that the application did not need to be available 7x24. It was
generally acceptable to have a window of time when the application would be
unavailable.
Today, this notion is ludicrous; most applications are expected to be available 7x24.
There are many technologies available to ensure that an application is always available
such as clustering and replication with options to include offsite protection. However, it
is still important to perform data backups to ensure that a restore can be done in the
event that one of the availability solutions fails, or an application does not warrant a
more costly solution than backup.
Performing an online (hot) backup involves backing up the application while it is running
and available to users. Online backups prevent us from shutting the application down
during backup. NetWorker has many Modules that can perform supported online
backups of different application types through supported Application Programming
Interface (API) or backup utilities native to the different applications. Performing
backups through NetWorker modules allows the NetWorker software to control the
backup media, store save set and volume information, and perform all scheduling tasks.
We will cover common configuration steps with the NetWorker Modules in the first part of
this article. The second half covers the more detailed information and settings for each
of these modules:
The following table shows some common questions to help you configure a NetWorker
Module backup:
NetWorker and the NetWorker Modules are very flexible to help backup the different
types of server environments and meet the different backup requirements that come with
managing these environments. With flexibility, there are usually multiple ways to fill
requirements.
In the following scenarios, I’ll present the easiest or most common method to configure
an option to back up the applications. Please consult the appropriate Administration
Guide for the NetWorker Module in question to see different methods to perform
backups and recoveries.
Client resource
When configuring an application server for backup with NetWorker, it will need multiple
Client resources in the NetWorker configuration interface. The NetWorker client is the
application server that needs to be backed up. Each newly created client needs
identical client names to match the hostname of the application server being backed up.
We may have multiple client instances to accommodate for a client backing up file
system data, a client to perform full database backups, and one or more other clients to
perform transaction log or incremental backups at shorter intervals than the full database
backups. Each client resource will have different properties such as backup times,
different save set properties, and different NetWorker Module commands or scripts that
will need to run.
Group resource
You can start a regular file system backup of a NetWorker client from the client through
the NetWorker User GUI on a Windows client, or through the save command on other
operating system (OS) types such as UNIX or Linux. This type of backup is a client-
initiated or manual backup.
Most NetWorker Modules have the ability to perform manual backups. Execute a
NetWorker Module or native backup utility command instead of using the save command
for a backup. You can perform a manual backup on-demand from the client, or
integrated with scripts and third-party schedulers to schedule database backups.
We can use the same framework during a database backup with the NetWorker
Modules. The NetWorker Module command is executed on the application server during
the scheduled backup instead of the normal save command.
Test a file system (OS) scheduled backup first to make sure it can complete successfully
before configuring a scheduled backup for your NetWorker Module configuration. There
is a very strong possibility that an application scheduled backup will fail if a scheduled
file system backup of a client cannot be performed.
Full database backups are typically performed at least once a day early in the backup
window. There may be degraded performance during an online backup since the
running database and the backup processes are sharing server resources. Although a
database may offer 7x24 availability, it may have high peak usage only at certain times.
Choose backup start times so the backup can be performed during low-usage times.
You can configure a new Group resource with the Start time: specified in 24 hour time
and the Autostart attribute enabled for performing a full backup once a day.
You should perform transaction log (incremental) backups more frequently than full
application backups. Transaction logs recover a database or application to a specific
point in time. Different applications have different ways of handling incremental
backups, but it is always important to back up these changes frequently. This gives you
the option to recover an application to a different point-in-time than the last full backup.
Some scheduling products call scripts in a hierarchical order – one after the other. For
example, don’t start script2 until script1 is complete. This prevents an incremental
backup from starting until a full backup has completed, or prevents the file system (OS)
backups from overlapping with the application backups.
If you decide to perform manual data backups, the client file indexes (CFI) of these
clients will not be backed up automatically. Make sure that the CFIs are still backed up
automatically through a Group resource if you choose to perform manual backups on a
regular basis without scheduled backups.
Schedule resource
Each NetWorker Module handles NetWorker level backups differently. Most support one
or more level settings in a Schedule resource when used during a scheduled backup.
All of the NetWorker Modules support some kind of incremental or transaction log
backup whether set in the Schedule resource, or in a configuration or script file.
Consider the amount of time it will take to apply the logs during a recovery when
deciding on a schedule for incremental or log backups. A full weekly backup may be
sufficient in a smaller environment with fewer transactions. You may need to perform full
daily backups for a larger application server. Consider restore/recovery Service Level
Agreements (SLA), and test recoveries before deciding on a scheduling strategy.
Each NetWorker client resource has “Backup command attribute” used to perform a
regular save for a file system backup, savepnpc, a script or a NetWorker Module
command.
When you leave the Backup command attribute blank and a scheduled backup starts,
the save command on the client is used to backup the file systems and directories listed
in the “Save set:” attribute. When performing scheduled backups of the NetWorker
Module clients, the Backup command attribute will contain a command or script to
initiate the application backup on that client. Some of the commands require you to
include a special switch to point to a configuration file.
The NetWorker Modules that require configuration files or scripts usually include a
template file that you can use as a base for creating custom configuration files and
scripts with your own backup environment needs. Always copy the template files to a
new name and then modify the new copy. Make sure the new filename begins with
either nsr or save when you are modifying a script (for example, .sh or .bat files).
NetWorker will not execute any commands from the Backup command attribute that do
not follow this naming convention.
The browse policy allows you to configure how long to keep backup information in the
client file index. This will keep the information available for a simple, browse-able
recovery. The browse policy helps the NetWorker administrator control the size of the
client file indexes. Typically, client file indexes (CFI) for NetWorker Module clients will
be smaller than comparable NetWorker file system clients. For example, a NetWorker
Module client with a 2 TB Microsoft Exchange server will have a much smaller CFI than
a comparable 2 TB file server with millions of small files.
The retention policy allows you to configure how long to protect information from being
overwritten or removed from your backup volumes. Volumes that have expired save
sets are considered eligible for re-use and may be overwritten with new backup data.
Pool resource
The NetWorker server uses the pool resource to sort backup data to different groups of
backup volumes. A media pool is a collection of volumes (tapes, virtual tapes, and disk)
that contain backup data.
Using different pools for application backups is not required, but often needed for better
recycling efficiency and ease of management.
Application data can be subjected to different laws and policies than the file system data,
and may be subjected to legal discovery requirements that the file system data is not.
Therefore, the retention policy for the file system data is typically shorter and the
application data longer.
It is not possible to recycle parts of a tape if tape media or virtual tape media is being
used for backups. When everything on a particular tape expires, the whole tape is
eligible to be reused and recycled. If there is a small amount of data on a tape that will
not expire for a long time, the whole tape cannot be re-used until that data expires. For
example, if there is a tape that is full and is 90% file system data that expires in 1 year
and 10% database data that will not expire for 7 years – the 90% of the tape that will
expire in one year will still not be used again (recycled) until after the database data has
expired after 7 years. This is inefficient use of tape media.
An alternative is to have two separate NetWorker pools, one pool for the 1 year file
system data and one pool for the 7 year application data. Separate media pools for
application data also prevents database backups inter-mixed (multiplexed) with other
client or file system backups. This provides quicker recovery performance with physical
tape recoveries.
Pool resource
NetWorker pools can dedicate a particular pool to use specific backup devices. Backing
up log files from a database can be a critical function to free up disk space for new
transaction logs. Usually, removing a transaction log file from the disk is not acceptable
until we have a backup copy. Having dedicated backup devices for log backups ensures
that they are always available for backup and restore operations of log files. This way,
older log files can be removed from the disk.
Virtual tape devices and disk arrays are ideal targets for transaction log backups since
the random-access nature of disk storage provides quick restores. The device attribute
of the Pool resource allows a pool to be directed to specific device types, such as virtual
tape or disk. Creating a pool specifically for transaction log backups on these devices
allows data to be directed away from the physical tape devices.
Pools for clone copies of backup data can then be written to physical tape devices
creating off-site tape copies to meet any disaster recovery requirements.
You can configure a NetWorker directive resource to exclude the application files during
the file system backup. That way you can use an “All” save set to capture all local file
systems, including any newly added file systems that may be accidentally excluded if we
were listing them in the save set attribute.
Restore is the process that the NetWorker Module will perform to bring data back from
backup media. Recovery is a process invoked by the NetWorker Module or native
backup utility after data is restored back from media.
You must update the Remote Access: attribute in the Client resource when planning to
restore data to a NetWorker client that the data was not originally backed up from. This
gives privileges to the receiving client to access the original backup client’s CFI.
The following sections of the article detail with specific NetWorker Modules. Each
section is organized to focus on the first five questions in the Business requirements
table. The NetWorker settings related to these five questions are generally different for
each NetWorker Module.
RMAN is the native backup and recovery utility for Oracle and is responsible for
capturing consistent copies of the database for backup including datafiles, archive log
files and control files. RMAN performs recovery and restore operations with the help of
the Oracle RMAN recovery catalog. RMAN maintains the recovery catalog, but you
should configure this data as part of your NetWorker backup outside of any steps
included here.
Client resource
For NMO, you must create multiple Client resources using the Oracle server hostname
in order to perform different types of scheduled backups. These include full database
backups, archived redo log backups, block-level incremental or differential backups and
file system (OS) backup data.
Group resource
You can start an NMO backup as a manual or scheduled backup. With a manual
backup, the RMAN command is called directly on the Oracle database server from the
command line or a script. Alternatively, you can perform a manual backup through the
Oracle Enterprise Manager (OEM) Backup Management Tools that provide a GUI
interface to RMAN.
You can perform a scheduled backup using a Group resource to configure a time that
you would like the database to back up. Each type of backup requires a different RMAN
script that is called through a client resource. You should perform archived redo log
backups more frequently than full database backups, so you may need to create multiple
groups for the different start times.
Use archived redo logs to recover a database to a specific point in time. Without
archived redo logs, a database could only be recovered to the last time a full or
incremental backup was performed.
You can configure the RMAN script to perform any of the supported Oracle RMAN
backup types including Backup full, Block level incremental, Archived redo log backups,
Control file autobackup, Automatic channel allocation, Backup copies (during manual
backups only), Retention policies, Backup and restore optimization, Backup of backup
sets, Restartable and offline backups.
You can perform Oracle full, incremental and differential backups using appropriate
RMAN settings. A level 0 backup set in the RMAN script will perform a full backup; level
1 backup will perform an incremental backup. Performing archived redo log backups is
another way to capture changes on a database.
The RMAN script name is placed in the Save set attribute of the Client resource so it can
be passed to the savefs command on the NMO client during a scheduled backup.
Some of the required variables that you need to modify in the nsrnmo script are included
in the following table:
NMO has an additional feature, save set bundling, that you can enable to automatically
create a save set bundle for each schedule backup cycle (full backup of the database
object and all subsequent incremental backups) of an Oracle database object. This
feature groups the dependent save sets from the same backup cycle into one save set
bundle, and automatically names and records it in the NetWorker media database.
Please reference the NetWorker Module for Oracle Multiplatform Version Administration
Guide for more details on save set bundling.
We can perform restores of Oracle data files, specific table spaces, control files and
archived redo log files With the NetWorker Module for Oracle.
Please reference the NetWorker Module for Oracle Multiplatform Version Administration
Guide for more specific details on restore RMAN scripts and the process of restoring the
Oracle database with NMO and RMAN.
The NetWorker Module for Microsoft SQL provides additional programs on the database
server to perform online and transaction log backups of the SQL databases. NetWorker
User for SQL Server GUI is installed with NMSQL and is used for manual backup and
restores. NMSQL can perform backups of the online database including files, file
groups, file streams, SQL database and transaction logs.
Client resource
Create multiple Client resources using the SQL Server hostname to perform different
types of scheduled backups (full database backups, transaction log backups and file
system backup data) when performing NMSQL backups. You can also create optional
clients to backup specific SQL databases at different times.
Group resource
You can start an NMSQL backup as a manual or scheduled backup. With a manual
backup, you can choose a collection of files, filegroups, databases or transaction logs
from the NetWorker User for SQL Server GUI.
You can perform scheduled backups using Group resources to configure the time that
you would like the database to back up. Perform transaction log backups more
frequently than full database backups. You can create Multiple Group resources with
different times to start transaction log backups at different times of the day. This provides
a number of point-in-time options to recover the database in-between full backups.
Schedule resource
There are four supported levels when configuring a Schedule for NMSQL backups: full
backup, incremental backup, differential backup and skip a backup:
There are some variations on the supported NetWorker levels. They are documented in
the NetWorker Module for Microsoft SQL Server Administration Guide if you are
planning to take advantage of the filestream and filegroup backup options.
The filestream feature of SQL 2008 must be enabled on the SQL Server during a backup
or restore operation.
Use backup striping when backing up a SQL Server to enable the use of multiple backup
devices. Stripes of data can be extracted from the SQL database in parallel to enable a
quicker backup by using multiple backup devices. You can enable backup striping during
a manual or scheduled group backup.
Append the “-Sn” option to the snrsqlsv command to enable striping during a scheduled
backup. The upper-case “S” informs NMSQL to perform the backup striping. Replace
the lower-case “n” with a number to specify the appropriate number of backup stripes.
The nsrsqlrc program, installed with the NMSQL software, performs the restore process.
The API translates the object names requested for restore by nsrsqlrc into a format
NetWorker recognizes so that the data can be requested to be received from the storage
node nsrmmd process.
You must enable the SQL Server filestream feature on the destination recovery server to
restore SQL Server 2008 filestream data.
Backups with NME 5.1 can perform backup and restore through the traditional Exchange
aware Backup API or using the Backup Utility Exchange agent. With the traditional
backup method, NME will contact the Exchange database through a local command and
request a consistent backup through the supported Exchange Backup API.
The Backup Utility Exchange agent can be used for mailbox level (brick-level) and public
folder backup and restore operations. The Backup Utility Exchange agent uses MAPI to
connect to each mailbox and public folder the same way Outlook does (instead of using
the API). Download and install the Microsoft Exchange Server MAPI Client and
Collaboration Data Objects kit from Microsoft to have support for mailbox and public
folder backup options for NME 5.1 running on Windows 2008.
Snap-shot type backups with Microsoft Volume Shadow Copy Service (VSS) or
NetWorker PowerSnap module are not currently supported by NME version 5.1. You
can achieve Snapshot VSS backups of Exchange through the NetWorker Module for
Microsoft Applications.
Client resource
When performing NME backups, create multiple Client resources using the Exchange
server hostname to perform different types of scheduled backups at different times.
These client resources can start full information store backups, transaction log backups
and file system backup data. You could also create optional clients for mailbox and
public folder backups.
Group resource
You can start an NME backup as a manual or scheduled backup. With a manual
backup, the administrator can choose different components of the Exchange server to
backup through the NetWorker User for Exchange Server GUI.
It might be sufficient to backup the Exchange server as a Full backup once a week with
a transaction log when creating groups for a smaller Exchange environment. You could
then perform differential or incremental backups in the middle of the week. You should
perform transaction log backups more frequently than full database backups to bridge
the gap of time from one full backup to another.
Schedule resource
You can request five Exchange Backup levels through a configured schedule resource.
These are full backup, copy backup, incremental backup, differential backup and skip a
backup. These options match the NetWorker Schedule resource in the following chart:
The mailbox level backups are useful for recovering individual mailboxes when a specific
user loses email, and that is the only mailbox that must be restored. Use this strategy
for a targeted number of individuals since this method of backup is slow.
The nsrxchsv program interfaces with the Exchange backup API to request the
traditional information store, mailbox and public folder backups of Exchange 2000 and
Exchange 2003. For Exchange 2007, the nsrxchsv command is only used for
Information Store and public folder backups. The nsrxchmbsv program interfaces with
the Exchange 2007 MAPI backup agent to perform mailbox level backups.
If you are backing up Exchange running on a virtual server, you will also want to specify
the “-a virtual_server_name” option used to tell NetWorker the client is a virtual server.
The nsrsqlrc program installed with the NME software performs the restore process. The
Exchange API translates the object names requested for restore by nsrxchrc into a
format that NetWorker recognizes when restoring data backed up through a traditional
method. The nsrxchmbrc command is used instead of the normal nsrxchrc command
when performing a restore of a mailbox level object or item for Exchange 2007.
When restoring a storage group on your Exchange server (2003, 2007), you can choose
to restore to a Recovery Storage Group (RSG). This allows you to recover a storage
group to an alternate location within the Exchange server (the RSG) and then choose
relevant information from the recovered copy.
When recovering the entire information store or a storage group, the NetWorker User for
Exchange Server interface gives an option to also “Complete Exchange recover
process” (replay logs). It puts the database online after a successful restore/recover
process is completed. If complete replay of logs is not performed by the NetWorker
restore process, it will need to be done afterwards using the eseutil Microsoft Exchange
Server utility.
VSS was first introduced in Windows 2003 and is now part of Windows 2008. The VSS
framework has three major components: writers, providers and requestors. The writer
is the component of the application being backed up that guarantees there is a
consistent copy of the data to backup. The provider is the storage component that
keeps the different point-in-time copies (snapshots) of the data, and the requestor is the
software that initiates the creation of shadow copies and performs the backup.
The NetWorker software requests a Shadow Copy snapshot be created with the NMM.
For every application or component that has a writer, NetWorker can request creation of
a snapshot to capture the data for backup. NetWorker uses the snapshot to create a
single consistent point-in-time copy of the application that can then be copied to
traditional backup media such as tape, virtual tape or disk.
Some of the applications that can be backed up and protected by NMM include SQL
Server, Exchange Server, SharePoint Server, Data Protection Manager, Active
Directory, Hyper-V, and File system snapshots.
The NMSQL and NME Modules use more traditional methods and APIs for backing up
the databases and do provide some features that are not supported with NMM such as
mailbox level backup and restore.
There is no question there is some overlap between NMSQL, NME and the NMM, but
the approach for backup and recovery is different (API vs. VSS). There is a very good
chance that NMM and VSS backup will be the future of Microsoft Application backup and
restore as more features and applications are added, but for now we have more than
one option available for SQL and Exchange backups.
Please check powerlink.emc.com for more information on NMM such as the “NetWorker
Module for Microsoft Applications 2.1 Administration Guide” and the “Backup and
Recover Using the EMC NetWorker Module for Applications” whitepaper.
The NetWorker Module for SAP R/3 provides additional programs on the SAP server
used to perform online, offline and archived redo log backups. NMSAP can also be
used in conjunction with the sapclone feature of BR*Tools to make additional copies of
the backup data for offsite protection, or to copy backup data between disk and tape
media. The main supporting programs for SAP BR*Tools that interface with backint and
NetWorker are brbackup, brarchive, brrecover and brrestore. You can also use an SAP
BrGui and character based “brtools” program to interface and call the other BR*Tools
commands.
Client resource
Create multiple client resources using the SAP server hostname to perform different
types of scheduled backups when configuring NMSAP. Examples of the different
backup types include full online database backups, archive redo log backups and
backing up file system data such as the backint log files or other operating system data.
Group resource
You can start an NMSAP backup as a manual or scheduled backup. You must perform
three configuration tasks before this can happen.
Before the NMSAP backint can be used for any backup and restore operations the first
task is to copy the backint (UNIX/Linux) or backint.exe (Windows), which is part of the
NMSAP software to the same location as the BR*Tools programs.
The second task is to update the SAP initialization file. The SAP initialization file is
normally named initSID.sap where SID is the Oracle system identifier (SID) and is
typically located in $ORACLE_HOME/dbs on UNIX and Linux systems and
%ORACLE_HOME%\database directory on Windows.
The backup device type parameter and the backup utility parameter file are two required
settings that need to be updated in the initialization file.
For the backup utility parameter file, update it to point to the NMSAP utl file:
util_par_file = $ORACLE_HOME/dbs/initSID.utl
or
util_par_file = %ORACLE_HOME%\database\initSID.utl
The third task is to create a proper parameter file for NMSAP. When performing a
manual or scheduled backup, the backint and BR*Tools programs use parameters in the
parameter file, initSID.utl. Be sure to have the initSID.utl file properly configured before
attempting any backint backup. Examples of some parameters that can be set include
client, server, pool, group, etc. Most of the parameters are optional or should be left at
With a manual backup, the administrator can use the BrGui interface to begin a backup
from the SAP server or run a BR*Tools command (brbackup/brarchive) directly from the
command line or a script. When performing a manual brbackup with backint it will send
Oracle datafiles, control files, existing profiles and SAP internal catalogs as backup
sessions to NetWorker. When using brarchive the archived redo logs will be passed for
backup to NetWorker.
For smaller SAP environments, it might be sufficient to backup the server as a Full
backup once a week with archived redo logs backing up in the middle of the week. You
should perform archived redo log backups more frequently than the full database
backups and bridge the gap of time between the full database backups.
When backing up SAP on Windows in a scheduled backup, you will need to update the
Remote User and Password attributes in the SAP client resource for proper
authentication to the SAP application.
When SAP is running with Oracle, the RMAN utility is also an option for capturing
database data and archived redo logs for backup. However, using RMAN in this
scenario is optional since the NMSAP backint program can be used to capture all
necessary data from the database. Although RMAN can be used for performing the
backup, recovery and restore operations, this article focuses strictly on using the backint
for performing these operations.
For NMSAP, we use the configuration file (nsrsapsv.cfg) covered in the next section to
determine the actual data to backup.
A template of the NMSAP configuration file, nsrsapsv.cfg is installed with the NMSAP
client software. On UNIX and Linux, it will be located in /etc; on Windows you will find it
in C:\Program Files\Legato\nsr\res. Copy this template file and update it to contain
proper settings for your setup and specific Oracle SID. You will likely have multiple
nsrsapsv.cfg files for different types of backups. For example, you might have an
nsrsapsv_SID_full_database.cfg file to call brbackup to perform a full database backup
and an nsrsapsv_SID_archive.cfg for calling brarchive to perform archived redo log
backups.
Command syntax:
# nsrsapadm –c nsrsapsv_custom_file.cfg
where nsrsapsv_custom_file.cfg is the configuration file you want updated.
Backup session reports can be a useful tool for monitoring or troubleshooting your SAP
manual and scheduled backups. The backup session reports can be found in:
C:\Program Files\Legato\applogs\nsrsapsv.<process_id>SID for Windows
or for UNIX and Linux in: /nsr/applogs/ nsrsapsv.<process_id>SID
You can also check the backintSID.log.raw found in the same location.
Verify that both the SAP initialization file, initSID.sap and the initSID.utl file are
configured correctly before attempting a restore or recover operation.
Verify that a required backup exists before initiating the restore or recovery operation.
Perform this by using the “-verify” option on the brrestore command.
After a restore is complete, you can find results in the backintSID.log.raw file in the
applogs sub-directory wherever the NetWorker nsr directory is located on the client.
Researching the required NetWorker Modules in great detail for each database is ideal,
but requesting outside help from a services organization such as EMC Global Services
would ensure that you get the best solution. Sometimes, these options are just not
possible. This article was written to give guidelines on the similar properties of the
NetWorker Modules and a more detailed look at NMO, NMSQL, NME, NMM and
NMSAP when an administrator might need to deploy the backups themselves. It is also
a useful reference after an external services organization has configured the backups.
Every application has transactional data that needs to be captured through backups. We
have covered some of the strategies to perform these backups. Deployments of
NetWorker Modules not covered in detail in this article can also benefit from some of the
information covered on configuring clients, backup times, backup levels, retentions,
pools, devices and directives.
Always remember that backups are being done to address situations when a restore or
recovery operation is needed. Consider the restore and recovery SLAs of a particular
database and make sure these requirements can be met. Information in this article can
help to build and configure your backup strategy, but also remember to regularly test
your application restores.