Procedures for Switching Replication Service Between Sites
Transferring activity from one replication site to another can occur through either a controlled or uncontrolled failover. The differences are as follows:
- A controlled failover occurs when activity is transferred from one replication site to its mate replication site in a planned manner for operational reasons. When you successfully follow the controlled failover procedure, no messages are lost, regardless of the replication mode of the messages.
- An uncontrolled failover occurs due to a sudden loss of connectivity to a replication site with Message VPNs with a replication-active status. When this occurs, only synchronous replication mode messages are guaranteed not to be lost. Messages remaining in the replication queue are not available at the Failed site.
The following sections provide the basic steps for performing a controlled failover and for recovering from an uncontrolled failover:
- Performing a Controlled Failover
- Performing an Uncontrolled Failover
- Failing Back to a Restored Site After an Uncontrolled Failover
If you are using replication with PubSub+ Cloud, see Using Replication for Disaster Recovery of Event Broker Services.