Triggering a Force-Start
You can trigger a force-start to force a cluster with persistence enabled to restart and delete its persistence store when one or more members are missing after a cluster-wide restart.
A cluster whose recovery policy is set to
FULL_RECOVERY_ONLY can recover after a cluster-wide restart only if all members restart with a valid partition table.
If a member is missing after a cluster-wide restart or if the member’s partition table is invalid, the cluster cannot restart unless all members first delete their persistence stores.
A force-start deletes the persistence stores of all members in the cluster and then restarts the cluster.
Before you Begin
Go to Administration > Persistence and check the Cluster Status field. Only trigger a force-start if the cluster status is In_Progress, which means your cluster is missing a member.
To learn more about persistence in Hazelcast and what happens after a cluster-wide restart, see What Happens When the Whole Cluster Restarts in the Platform documentation.
This page is available only to admin users.