8 Overview | Nortel Networks Confidential |
|
|
Database recovery
Recovery restores a database to its original state after a failure. See “Database recovery” on page 84.
Manage replication transaction errors
The OEM Console provides tools for resolving replication errors, should they occur. Replication errors can result from a lack of available space in a table targeted for an update or other unresolved replication conflicts.
For additional information, see “Resolving replication errors” on page 89.
Resynchronization
Replicated databases should always be in synchronization. In the unlikely event that changes made to one database are not successfully propagated to the other, the two databases become out of synchronization and must be manually resynchronized. See “Resynchronization” on page 92.
Disk space management
Ascript called optimize_dbspace drops and recreates Database Module indexes to reduce disk space usage. See “Disk space management” on page 93.
Fault management
Use the OEM Console to monitor and respond to Database Module faults in the categories listed in this section. Note the recommended frequency for each task.
•Replication: Replication jobs ensure that the primary and secondary databases remain synchronized.
Therefore it is critical that you use the OEM Console to monitor replication jobs once every 24 hours to ensure the databases remain synchronized.
See “Monitor replication jobs” on page 27 and “Resolving replication errors” on page 89.
•Backups: Once backup jobs have been scheduled using the Oracle Enterprise Manager (OEM) Console, the system backs up the database at predefined periods and sends a status email to the operator.
It is recommended that you monitor backup status to ensure the backup was successful. See “Monitoring backups” on page 20.
Copyright © 2003, Nortel Networks |