4.1

3 Reconfigure SRM to Enable Failback to the Protected Site on page 45
Before you can run a failback, you must create the protection groups and recovery plans required to
migrate protected inventory from the recovery site back to the protected site.
4 Restore the Original Configuration on page 45
After a failback
is complete, you can restore the original configuration so that the protected and recovery
sites resume the roles they had before the failover.
Review and Execute Post-Failover Cleanup Tasks
Before you can execute a failback, you must remove artifacts such as invalid protection groups and unneeded
placeholders that are left over from the previous configuration.
If the original protected site is intact after a failover, all the protected virtual machines are still part of its
inventory, although they are powered off. You must remove those virtual machines before you can create the
protection groups that are needed by the failback. You must also remove various other artifacts at both sites
as part of preparing the sites to assume their new roles.
Procedure
1 If the protected site is intact after a failover and you want to restore it to its former status, clean up the
protected site.
a Verify that the SRM server host is operational. Verify that the SRM installation, including its database
and vCenter server, has not been affected by the events that motivated the failover.
b Open a vSphere Client and connect to the vCenter server at the protected site. Log in as a vCenter
administrator.
c Rescan the host bus adapters (HBAs) on the SRM server host.
After the rescan completes, SRM lists the failed-over virtual machines and their protection groups as
invalid because their storage is no longer replicated. (Replication was turned off by the failover.)
d Delete the invalid virtual machines and protection groups.
2 Clean up the recovery site.
a Open a vSphere Client and connect to the vCenter server at the recovery site. Log in as a vCenter
administrator.
b Remove the placeholder virtual machines from vCenter inventory.
Reconfigure Replication
Failover stops replication. Failback requires you to configure replication in reverse, from the recovery site to
the protected site. Restoring the protected and recovery sites to their original roles requires you to configure
replication from the protected site to the recovery site, as it was before the original failover was executed.
Reconfiguring replication is likely to require help from the team that manages vSphere storage for the two
sites. The operations required are specific to the arrays that you are using. Generally, you must take the
following steps:
n
To prepare for a failback, configure the arrays so that the source devices are the ones located at the recovery
site and the target devices are the ones located at the protected site.
n
After the failback is complete and you are ready to have the protected site and recovery site resume their
original roles, configure the arrays so that the source devices are the ones located at the protected site and
the target devices are the ones located at the recovery site.
After you have configured replication as needed, force an immediate, one-time replication from the source to
the target. This step is always required during a failback, but might not be needed when you are reconfiguring
the protected and recovery sites to resume their original roles.
Site Recovery Manager Administration Guide
44 VMware, Inc.