Site Recovery Manager (SRM) interacts with storage arrays where the VMware datastores are held, to protect VMs at the production site with replicas at a backup site. SRM does this by sending queries and commands to the storage arrays via a Storage Replication Adapter (SRA) which translates these into Protector REST API calls. Protector then either returns state information about the LDEVs where the datastores are held or performs: Pausing, Resuming, Secondary Failover, and Swapping; of the associated replication pairs for: SRM Test Failover, Test Cleanup, Failover and Reprotect; for both stretched and non-stretched storage SRM protection groups.
Figure. SRM, SRA and Protector interaction
| SRM Command | Protector Action |
|---|---|
| Discover paired storage arrays | List arrays in existing replications |
| Discover replicated devices | List replicated LUNs for those replications |
| Sync replicated device | Trigger the replication if replication is not in sync |
| Test a Failover by temporarily copying VMs from one vCenter to another |
|
| Failover by permanently moving VMs from one vCenter to another | Put the DR replication into a secondary failover state or swap the replication if it is GAD and there is no disaster recovery required |
| Reverse replication (reprotect) | Swap the replication if required |