Operational Considerations
12.1 Cluster Rebalance
The following table lists the types of rebalance operations that can be performed on a vRealize Operations Manager cluster.
Table 8. vRealize Operations Manager Rebalance Operations
Operation | Description | Implication |
Rebalance adapter instances | Rebalance all adapter instances across nodes if experiencing poor collection or cluster performance. | vRealize Operations Manager will miss a few collection cycles. |
Rebalance disk | Repartition all data on disk to equally distribute the disk load. Use this option if you have added new nodes, but have not deployed any new adapter instances. | This is a disruptive operation that might take hours, but it balances the system and improves performance. |
Rebalance GemFire | Repartition GemFire data in memory to equally distribute the memory and network load. | |
Rebalance adapter instances only if the cluster is experiencing poor performance. Perform disk and GemFire rebalancing after adding new nodes to a cluster.
The rebalance disk operation redistributes FSDB data evenly across the cluster. It can take hours to complete, because many terabytes of data might need to be distributed. As such, VMware recommends that performing disk rebalance operations during a major maintenance window.
The GemFire rebalance operation redistributes the GemFire cache across all nodes. It is far less disruptive than the disk rebalance operation.