Migrating From Helm Chart to the Operator
Unfortunately there isn’t any way to migrate without re-creating the cluster as the Operator needs to take control of the custom resources which were created by the Helm chart, which it is not possible.
There are two ways to migrate from a cluster that is created by the Hazelcast Helm Chart.
For either solution to work, persistence must be enabled. You should also set the cluster-data-recovery-policy parameter’s value to PARTIAL_RECOVERY_MOST_COMPLETE to avoid any issues during the restore process.
|
Restoring from a Backup
Follow the steps below:
-
Install Management Center. (if you didn’t before)
-
Trigger a backup using the Management Center user interface.
-
Uninstall existing cluster. (Don’t forget to keep PVCs)
-
Follow the steps explained in the Restoring from Persistent Volumes section.
Attaching new Pods to Existing PVCs
You can also restore your data to the cluster that is installed by the Operator without creating a backup. You just need to adjust the newly created Hazelcast CR’s name correctly. The name of the new CR must be in the following format: <name>-hazelcast-persistence. For example, assuming that the older cluster was installed with helm install hz hazelcast/hazelcast-enterprise command, the new CR name must be hz-hazelcast-enterprise so that PVCs created by the Helm Chart will attach to the pods created by the Operator correctly.
