Double Failure

Failure of a node before the initial resynchronization operation is complete is a double failure and can result in data loss.

As long as both nodes stay up long enough for the resynchronization operation to succeed, there is none or minimal loss of data.

If a failover occurs too frequently, call TIBCO Support. This might be due to an unsupported hardware setup. The failover membership can trigger spurious failover if the network bandwidth is insufficient. It is good practice to use only Gigabit Ethernet connections for the HA replication traffic.