8 Managing Backup and Recovery

Non-Cluster File System Backup Scheme

In a non-cluster file system environment, each node can back up only its own local archived redo logs. For example, node 1 cannot access the archived redo logs on node 2 or node 3 unless you configure the network file system for remote access. If you configure a network file system file for backups, then each node will backup its archived logs to a local directory.

Cluster File System Restore Scheme

The scheme that this section describes assumes that you are using the "Automatic Storage Management and Cluster File System Archiving Scheme". In this scheme, assume that node 3 performed the backups to a CFS. If node 3 is available for the restore and recovery operation, and if all of the archived logs have been backed up or are on disk, then run the following commands to perform complete recovery:

RESTORE DATABASE;
RECOVER DATABASE;

If node 3 performed the backups but is unavailable, then configure a media management device for one of the remaining nodes and make the backup media from node 3 available to this node.

Non-Cluster File System Restore Scheme

The scheme that this section describes assumes that you are using the "Non-Cluster File System Local Archiving Scheme". In this scheme, each node archives locally to a different directory. For example, node 1 archives to /arc_dest_1, node 2 archives to /arc_dest_2, and node 3 archives to /arc_dest_3. You must configure a network file system file so that the recovery node can read the archiving directories on the remaining nodes.

If all nodes are available and if all archived redo logs have been backed up, then you can perform a complete restore and recovery by mounting the database and running the following commands from any node:

RESTORE DATABASE;
RECOVER DATABASE;

Because the network file system configuration enables each node read access to the other nodes, then the recovery node can read and apply the archived redo logs located on the local and remote disks. No manual transfer of archived redo logs is required.

Using RMAN or Enterprise Manager to Restore the Server Parameter File (SPFILE)

RMAN can restore the server parameter file either to the default location or to a location that you specify. This procedure is described in Oracle Database Backup and Recovery Basics.

You can also use Enterprise Manager to restore SPFILE. From the Backup/Recovery section of the Maintenance tab, click Perform Recovery. The Perform Recovery link is context-sensitive and navigates you to the SPFILE restore only when the database is closed.

Single Node Failure in Real Application Clusters

Instance recovery in RAC does not include the recovery of applications that were running on the failed instance. Oracle clusterware restarts the instance automatically. You can also use callout programs as described in the example on OTN to trigger application recovery.

Applications that were running continue by using failure recognition and recovery. This provides consistent and uninterrupted service in the event of hardware or software failures. When one instance performs recovery for another instance, the surviving instance reads online redo logs generated by the failed instance and uses that information to ensure that committed transactions are recorded in the database. Thus, data from committed transactions is not lost. The instance performing recovery rolls back transactions that were active at the time of the failure and releases resources used by those transactions.

Multiple-Node Failures in Real Application Clusters

When multiple node failures occur, as long as one instance survives, RAC performs instance recovery for any other instances that fail. If all instances of a RAC database fail, then Oracle automatically recovers the instances the next time one instance opens the database. The instance performing recovery can mount the database in either cluster database or exclusive mode from any node of a RAC database. This recovery procedure is the same for Oracle running in shared mode as it is for Oracle running in exclusive mode, except that one instance performs instance recovery for all of the failed instances.

Using RMAN to Create Backups in Real Application Clusters

Oracle provides RMAN for backing up and restoring the database. RMAN enables you to back up, restore, and recover datafiles, control files, SPFILEs, and archived redo logs. RMAN is included with the Oracle server and it is installed by default. You can run RMAN from the command line or you can use it from the Backup Manager in Oracle Enterprise Manager. In addition, RMAN is the recommended backup and recovery tool if you are using Automatic Storage Management (ASM).

The procedures for using RMAN in RAC environments do not differ substantially from those for Oracle single-instance environments. Refer to the Oracle Backup and Recovery documentation set for more information about single-instance RMAN backup procedures.

Channel Connections to Cluster Instances

Channel connections to the instances are determined using the connect string defined by channel configurations. For example, in the following configuration, three channels are allocated using user1/pwd1@service_name. If you configure the SQL Net service name with load balancing turned on, then he channels are allocated at a node as decided by the load balancing algorithm.

CONFIGURE DEVICE TYPE sbt PARALLELISM 3;
CONFIGURE DEFAULT DEVICE TYPE TO sbt;
CONFIGURE CHANNEL DEVICE TYPE SBT CONNECT 'user1/pwd1@<service_name>'

However, if the service name used in the connect string is not for load balancing, then you can control at which instance the channels are allocated using separate connect strings for each channel configuration as follows:

CONFIGURE DEVICE TYPE sbt PARALLELISM 3;
CONFIGURE CHANNEL 1.. CONNECT 'user1/pwd1@node1';
CONFIGURE CHANNEL 2.. CONNECT 'user2/pwd2@node2';
CONFIGURE CHANNEL 3.. CONNECT 'user3/pwd3@node3';

In the previous example, it is assumed that node1, node2 and node3 are SQL Net service names that connect to pre-defined nodes in your RAC environment. Alternatively, you can also use manually allocated channels to backup your database files. For example, the following command backs up the spfile, controlfile, datafiles and archived logs:

RUN
{
    ALLOCATE CHANNEL CH1 CONNECT 'user1/pwd1@node1';
    ALLOCATE CHANNEL CH2 CONNECT 'user2/pwd2@node2';
    ALLOCATE CHANNEL CH3 CONNECT 'user3/pwd3@node3';
    BACKUP DATABASE PLUS ARCHIVED LOG;
}

During a backup operation, as long as at least one of the channels allocated has access to the archived log, RMAN automatically schedules the backup of the specific log on that channel. Because the control file, spfile, and datafiles are accessible by any channel, the backup operation of these files is distributed across the allocated channels.

For a local archiving scheme, there must be at least one channel allocated to all of the nodes that write to their local archived logs. For a CFS archiving scheme, assuming that every node writes to the archived logs in the same CFS, the backup operation of the archived logs is distributed across the allocated channels.

During a backup, the instances to which the channels connect must be either all mounted or all open. For example, if the node1 instance has the database mounted while the node2 and node3 instances have the database open, then the backup fails.

Node Affinity Awareness of Fast Connections

In some cluster database configurations, some nodes of the cluster have faster access to certain datafiles than to other datafiles. RMAN automatically detects this, which is known as node affinity awareness. When deciding which channel to use to back up a particular datafile, RMAN gives preference to the nodes with faster access to the datafiles that you want to back up. For example, if you have a three-node cluster, and if node 1 has faster read/write access to datafiles 7, 8, and 9 than the other nodes, then node 1 has greater node affinity to those files than nodes 2 and 3.

Deleting Archived Redo Logs after a Successful Backup

Assuming that you have configured the automatic channels as defined in section "Channel Connections to Cluster Instances", you can use the following example to delete the archived logs that you backed up n times. The device type can be DISK or SBT:

DELETE ARCHIVELOG ALL BACKED UP n TIMES TO DEVICE TYPE device_type;

During a delete operation, as long as at least one of the channels allocated has access to the archived log, RMAN will automatically schedule the deletion of the specific log on that channel. For a local archiving scheme, there must be at least one channel allocated that can delete an archived log. For a CFS archiving scheme, assuming that every node writes to the archived logs on the same CFS, the archived log can be deleted by any allocated channel.

If you have not configured automatic channels, then you can manually allocate the maintenance channels as follows and delete the archived logs.

ALLOCATE CHANNEL FOR MAINTENANCE DEVICE TYPE DISK CONNECT 'SYS/oracle@node1';
ALLOCATE CHANNEL FOR MAINTENANCE DEVICE TYPE DISK CONNECT 'SYS/oracle@node2';
ALLOCATE CHANNEL FOR MAINTENANCE DEVICE TYPE DISK CONNECT 'SYS/oracle@node3';
DELETE ARCHIVELOG ALL BACKED UP n TIMES TO DEVICE TYPE device_type;

Autolocation for Backup and Restore Commands

RMAN automatically performs autolocation of all files that it needs to back up or restore. If you use the non-cluster file system local archiving scheme, then a node can only read the archived redo logs that were generated by an instance on that node. RMAN never attempts to back up archived redo logs on a channel it cannot read.

During a restore operation, RMAN automatically performs the autolocation of backups. A channel connected to a specific node only attempts to restore files that were backed up to the node. For example, assume that log sequence 1001 is backed up to the drive attached to node 1, while log 1002 is backed up to the drive attached to node 2. If you then allocate channels that connect to each node, then the channel connected to node 1 can restore log 1001 (but not 1002), and the channel connected to node 2 can restore log 1002 (but not 1001).

Parallel Recovery with RMAN

With RMAN's RESTORE and RECOVER commands, Oracle automatically makes parallel the following three stages of recovery:

Restoring Datafiles When restoring datafiles, the number of channels you allocate in the RMAN recover script effectively sets the parallelism that RMAN uses. For example, if you allocate five channels, you can have up to five parallel streams restoring datafiles.

Applying Incremental Backups Similarly, when you are applying incremental backups, the number of channels you allocate determines the potential parallelism.

Applying Archived Redo Logs With RMAN, the application of archived redo logs is performed in parallel. Oracle automatically selects the optimum degree of parallelism based on available CPU resources.

Disabling Parallel Recovery

You can override this using the procedures under the following topics:

• Disabling Instance and Crash Recovery Parallelism

• Disabling Media Recovery Parallelism