Performing failover v7

Failover is replacing the publication database with the subscription database if a failure occurs on the publication database or its host. Failover is an irreversible action, so the subscription database permanently takes over the role of the publication database.

Generally, to perform a failover, follow the same steps as for a controlled switchover, described in Performing controlled switchover. However, also take the following points into consideration:

  • If you can't salvage the control schema objects on the publication database (that is, schemas _edb_replicator_pub, _edb_replicator_sub, _edb_scheduler, and their objects) or restore them from a backup, then performing a failover might be possible only with the help of EnterpriseDB Technical Support Services.
  • Pending updates not yet applied to the subscription might be lost. The chances of this are greater if the interval between synchronizations is long.

If you determine that a failover is possible, follow the steps for a controlled switchover.