For AtScale Clusters Only: Manually Promoting Standby Instances to Master Instances
Follow these steps when the master AtScale database instance in an AtScale cluster has failed and no standby instance is able to promote itself to be the master instance.
Procedure
-
Ensure that a standby instance is running and that no master (Leader) instance is running by running this command from any node in the cluster:
/opt/atscale/current/bin/database/postgres_nodes -
Fail over to a standby by running this command from any node in the cluster:
/opt/atscale/current/bin/database/patronictl failover
What to do next
Verify that the standby instance is now the master (Leader) instance by running this command from any node in the cluster:
/opt/atscale/current/bin/database/postgres_nodes