Special Considerations in Clustered Operation
In an environment where two or more Process Engines are operated in a cluster, you must ensure that Storage Cleaner is enabled on at least one node in the cluster, and preferably on multiple nodes.
This is especially true if the cluster configuration is not stable over time, that is, nodes are often taken offline and then brought back online.
Important:
If a cluster node that has Storage Cleaner enabled is taken offline or removed from the cluster, be sure that Storage Cleaner is enabled on at least one other node in the cluster. If you do not, Process Engine database size limits may be reached, and problems may be experienced if you add the offline node back into the cluster. The previously offline node will attempt to sync to the Process Engine database, and if no Storage Cleaner has been running, the data transfer is likely to be large and out-of-memory errors may result. When Storage Cleaner eventually does run, it is likely to take a long time to complete.
In situations where cluster operation is reliably stable, you can enable Storage Cleaner on multiple nodes in a cluster to share load. Process Engine uses a lock-management system to serialize the operation of Storage Cleaners on multiple nodes, so no two nodes will be running the Storage Cleaner at the same time.