Site Recovery Manager functions#
This information describes the Site Recovery Manager functions.
When you run a recovery plan in test mode, a snapshot of all VMs associated with the recovery plan is created, and the VM is booted off that snapshot. SRM imports the VMs and configures the networks based on the network mappings and recovery plan settings. SRM powers on the VMs based on the configured boot priority. After every recovery plan runs in test mode, you must run a cleanup.
Running the cleanup immediately powers off the recovered VMs and discards all changed data from the test.
In planned migration mode, SRM halts the recovery plan action if any of the steps fail. SRM attempts data replication to ensure the most recent data is available. SRM issues a shutdown command to all protected VMs in the recovery plan.
If any VMs are appliances or don't have VMware Tools, you should take special care.
Follow the steps in VM boot priority to configure the shutdown action for the VM.
The recovery process takes place in the following manner:
- If this was a previously-run recovery where VMs at the current source site
are configured to be suspended as part of the recovery, SRM attempts to resume those VMs.
- SRM reconfigures the protection group to prepare for migration to the target site.
- SRM initiates one more storage replication to ensure the target-side VMs are
in a mirrored state. This action ensures the guest OS virtual disks are in a clean state from a graceful shutdown.
- SRM suspends any non-critical VMs that are
preconfigured to suspend during a recovery operation.
- SRM reconfigures the target-side storage as writeable.
- SRM imports and powers on the VMs based on their preconfigured boot priority.
When you run a recovery plan in disaster recovery mode, SRM follows the exact steps outlined above but expects failure. After the source site is back online, you must run the recovery plan again. SRM steps through any failed steps and attempts to complete them successfully. You must rerun the recovery plan until all steps complete successfully. After the recovery plan is successfully listed as Recovered, you must proceed to run Reprotect.
You can only run a Reprotect after a recovery plan completes successfully. You must run a Reprotect to reconfigure the protection group and recovery plan to facilitate failback. SRM attempts to resume any suspended non-critical VMs. SRM reconfigures replication and the protection group in the reverse direction. This is where the configuration of reverse mappings is required. See Configuring Site Recovery Manager.
IP customization rules going back the other direction are also applied at this point. SRM begins replication and synchronizes the storage.