SnapCenter supports backup of different Oracle database configurations.

For the latest information about supported Oracle versions, see the NetApp Interoperability Matrix Tool.

Backup type specifies the type of backup that you want to create. SnapCenter supports online and offline backup types for Oracle databases.

"Resources" are Oracle databases on the host that are maintained by SnapCenter. You can add these databases to resource groups to perform data protection operations after you discover the databases that are available. You should be aware of the process that SnapCenter follows to discover different types and versions of Oracle databases.

In Oracle Real Application Clusters (RAC) setup, you can specify the preferred nodes on which the backup operation will be performed. If you do not specify the preferred node, SnapCenter automatically assigns a node as the preferred node and backup is created on that node.

The preferred nodes might be one or all of the cluster nodes where the RAC database instances are present. The backup operation will be triggered only on these preferred nodes in the order of the preference.

Example: The RAC database cdbrac has three instances: cdbrac1 on node1, cdbrac2 on node2, and cdbrac3 on node3. The node1 and node2 instances are configured to be the preferred nodes, with node2 as the first preference and node1 as the second preference. When you perform a backup operation, the operation is first attempted on node2 because it is the first preferred node. If node2 is not in the state to back up, which could be due to multiple reasons such as the plug-in agent is not running on the host, the database instance on the host is not in the required state for the specified backup type, or the database instance on node2 in a FlexASM configuration is not being served by the local ASM instance; then the operation will be attempted on node1. The node3 will not be used for backup because it is not on the list of preferred nodes.

In a Flex ASM setup, Leaf nodes will not be listed as preferred nodes if the cardinality is less than the number nodes in the RAC cluster. If there is any change in the Flex ASM cluster node roles, you should manually discover so that the preferred nodes are refreshed.

The backups of Oracle databases can be cataloged with Oracle Recovery Manager (RMAN) to store the backup information in the Oracle RMAN repository.

The cataloged backups can be used later for block-level restore or tablespace point-in-time recovery operations. When you do not need these cataloged backups, you can remove the catalog information.

The database must be in mounted or higher state for cataloging. You can perform cataloging on data backups, archive log backups, and full backups. If cataloging is enabled for a backup of a resource group that has multiple databases, cataloging is performed for each database. For Oracle RAC databases, cataloging will be performed on the preferred node where the database is at least in mounted state.

By default, the target database control file is used for cataloging. If you want to add external catalog database, you can configure it by specifying the credential and Transparent Network Substrate (TNS) name of the external catalog using the Database Settings wizard from the SnapCenter graphical user interface (GUI). You can also configure the external catalog database from the CLI by running the Configure-SmOracleDatabase command with the -OracleRmanCatalogCredentialName and -OracleRmanCatalogTnsName options.

If you enabled the cataloging option while creating an Oracle backup policy from the SnapCenter GUI, the backups are cataloged using Oracle RMAN as a part of the backup operation. You can also perform deferred cataloging of backups by running the Catalog-SmBackupWithOracleRMAN command. After cataloging the backups, you can run the Get-SmBackupDetails command to obtain the cataloged backup information such as the tag for cataloged datafiles, the control file catalog path, and the cataloged archive log locations.

If the ASM disk group name is greater than or equal to 16 characters, from SnapCenter 3.0, the naming format used for the backup is SC_HASHCODEofDISKGROUP_DBSID_BACKUPID. However, If the disk group name is less than 16 characters, the naming format used for the backup is DISKGROUPNAME_DBSID_BACKUPID, which is the same format used in SnapCenter 2.0.

You can perform crosschecks to update outdated RMAN repository information about backups whose repository records do not match their physical status. For example, if a user removes archived logs from disk with an operating system command, the control file still indicates that the logs are on disk, when in fact they are not. The crosscheck operation enables you to update the control file with the information. You can enable crosscheck by running the Set-SmConfigSettings command and assigning the value TRUE to the ENABLE_CROSSCHECK parameter. The default value is set to FALSE.

sccli Set-SmConfigSettings-ConfigSettingsTypePlugin-PluginCodeSCO-ConfigSettings "KEY=ENABLE_CROSSCHECK, VALUE=TRUE"

You can remove the catalog information by running the Uncatalog-SmBackupWithOracleRMAN command. You cannot remove the catalog information using the SnapCenter GUI. However, information of a cataloged backup is removed while deleting the backup or while deleting the retention and resource group associated with that cataloged backup.

If the cataloging and uncataloging fails because the operation time exceeded the time out value specified for the ORACLE_PLUGIN_RMAN_CATALOG_TIMEOUT parameter, you should modify the value of the parameter by running the following command:

/opt/Netapp/snapcenter/spl/bin/sccli Set-SmConfigSettings-ConfigSettingsType Plugin -PluginCode SCO-ConfigSettings "KEY=ORACLE_PLUGIN_RMAN_CATALOG_TIMEOUT,VALUE=user_defined_value"

After modifying the value of the parameter, restart the SnapCenter Plug-in Loader (SPL) service by running the following command:

/opt/NetApp/snapcenter/spl/bin/spl restart

The information regarding the parameters that can be used with the command and their descriptions can be obtained by running Get-Help command_name. Alternatively, you can also refer to the SnapCenter Software Command Reference Guide.

Backup frequency (schedule type) is specified in policies; a backup schedule is specified in the resource group configuration. The most critical factor in determining a backup frequency or schedule is the rate of change for the resource and the importance of the data. You might back up a heavily used resource every hour, while you might back up a rarely used resource once a day. Other factors include the importance of the resource to your organization, your Service Level Agreement (SLA), and your Recover Point Objective (RPO).

An SLA defines the level of service expected and addresses many service-related issues, including the availability and performance of service. An RPO defines the strategy for the age of the files that must be recovered from backup storage for regular operations to resume after a failure. The SLA and RPO contribute to the data protection strategy.

Even for a heavily used resource, there is no requirement to run a full backup more than once or twice a day. For example, regular transaction log backups might be sufficient to ensure that you have the backups you need. The more often you back up your databases, the fewer transaction logs SnapCenter has to use at restore time, which can result in faster restore operations.

Backup schedules have two parts, as follows:

You can either use the default Snapshot naming convention or use a customized naming convention. The default backup naming convention adds a timestamp to Snapshot names that helps you identify when the copies were created.

The Snapshot uses the following default naming convention:

resourcegroupname_hostname_timestamp

You should name your backup resource groups logically, as in the following example:

In this example, the syntax elements have the following meanings:

Alternatively, you can specify the Snapshot name format while protecting resources or resource groups by selecting Use custom name format for Snapshot copy. For example, customtext_resourcegroup_policy_hostname or resourcegroup_hostname. By default, the time stamp suffix is added to the Snapshot name.

You can choose either the number of days for which to retain backup copies or specify the number of backup copies you want to retain, up to a ONTAP maximum of 255 copies. For example, your organization might require that you retain 10 days of backup copies or 130 backup copies.

While creating a policy, you can specify the retention options for the backup type and the schedule type.

If you set up SnapMirror replication, the retention policy is mirrored on the destination volume.

SnapCenter deletes the retained backups that have retention labels that match the schedule type. If the schedule type was changed for the resource or resource group, backups with the old schedule type label might still remain on the system.

You can verify backup copies on the primary storage volume or on either the SnapMirror or SnapVault secondary storage volume. Verification using a secondary storage volume reduces load on the primary storage volume.

When you verify a backup that is either on the primary or secondary storage volume, all the primary and the secondary Snapshots are marked as verified.

SnapRestore license is required to verify backup copies on SnapMirror and SnapVault secondary storage volume.