Specifications

33 Deploying VMware vCenter Site Recovery Manager with NetApp FAS/V-Series Storage Systems
9.3 RESYNCING: PRIMARY STORAGE LOST
In this case the outage that was cause for the failover of the environment to the DR site was such that it
caused the complete loss of data from the primary storage or the complete loss of the infrastructure at the
primary site. From a SnapMirror and SRM point of view, recovery from such an event is nearly identical to
the steps in the section above. An exception is that once the infrastructure at the primary site has been
recovered or replaced, a SnapMirror initialize action must be performed to replicate the entire content of
required data back from the DR site to the primary.
All of the steps necessary for recovery are not reiterated in detail here. It is important to note that if the WAN
link between the sites is sufficient to allow the complete transfer of data over the WAN, then this process
would be nearly identical to the process documented above with the exception that, once the primary
storage had been replaced and made available, the “snapmirror initialize –S <dr_controller>:<dr_volume>
<primary_controller>:<primary_volume>” command would be used in place of the “snapmirror resync –S”
command in the section above.
If a complete transfer of data over the WAN to the primary site is not possible, consult the SnapMirror Best
Practices Guide for information regarding other methods of a SnapMirror initialization, including SnapMirror
to tape, or cascading SnapMirror through an intermediate system:
www.netapp.com/us/library/technical-reports/tr-3446.html.
1. Recover or replace the infrastructure at the primary site as necessary, including but not limited to:
a. The VMware vCenter server, including the installation and configuration of the SRM plug-in
b. The VMware ESX hosts and any custom settings that were previously configured, time
synchronization, admin access, and so on
c. The NetApp primary storage, including volumes that data from the DR site will be recovered into
d. The igroups necessary for the ESX hosts to connect
e. The network infrastructure necessary
2. Configure network connectivity between the sites to allow SnapMirror relationships to be established.
3. Initialize the SnapMirror relationships. On the console of the primary site storage system the SnapMirror
initialize command is used specifying the DR site volume as the source with the –S switch.
4. Check the status of the transfer periodically until it is complete.
5. At this time create SnapMirror schedule entries to keep the relationship up to date until the failback is
scheduled.
6. Build reverse SRM relationships.
7. From the DR site vCenter server configure the array manager for the protection side.
8. Configure the recovery-side array manager.
9. Configure inventory preferences for the reverse SRM setup.
10. Create reverse protection groups at the DR site.
11. Create reverse recovery plans at the primary site.
12. Verify that the SnapMirror relationship has completed transferring and that it is up to date.
13. Perform a SnapMirror update if necessary and perform a test run of the recovery plan.
14. If all is well with the recovery plan and the other necessary infrastructure at the primary site, schedule
an outage to perform a controlled failover.
15. When appropriate, notify the users of the outage and begin executing the failback plan.
16. During the scheduled failover, if allowed to do so, SRM will shut down the virtual machines at the DR
site when the recovery plan is executed. However, this is not desired in most cases. Instead of allowing
SRM to shut down the virtual machines, perform a controlled proper shutdown of application and
infrastructure virtual machines at the DR site. This helps make sure that all applications are stopped,
users are disconnected, and no changes are being made to data that would not get replicated back to
the primary site.
17. Once the environment to be failed back is properly shut down, perform a final SnapMirror update to
transfer any remaining data back to the primary site.