The Rim node is taking over central node immediately at service start instead of waiting for the default 10 minutes failover timeout.
Environments:
Qlik Sense Enterprise on Windows February 2021 and later
Resolution
This is working as designed. If central node and rim node that is a failover candidate have been stopped for more than 10 minutes (default timeout for failover timeout). The first node that starts will immediately take over the role as the timeout calculation is based on the last timestamp recorded in the ServerNodesHeartbeat in the repository database.
If the timestamp is older than 10 minutes and the rim node starts faster than the central node then it will take over the central node role.