HP Matrix Operating Environment 7.3 Recovery Management User Guide

3 Testing and failover operations
This chapter describes Recovery Group testing, planned failovers, and unplanned failovers using
the Matrix recovery management Activate and Deactivate operations.
Testing Recovery Groups
There are two ways to test Recovery Groups:
• Using Maintenance Mode to test individual Recovery Groups.
NOTE: VC-hosted logical servers imported at the recovery site recovery group must be
activated and then deactivated using Visualization Manager before maintenance mode can
be disabled in Matrix recovery management.
• Performing a planned failover to test all Recovery Groups. See “Planned failover” (page 28)
for more information.
This section focuses on using Maintenance Mode to test individual Recovery Groups.
For testing purposes, you can manually place a Recovery Group into Maintenance Mode and then
activate the deactivated logical servers or inactive services belonging to that Recovery Group using
the HP Matrix Operating Environment.
Maintenance Mode temporarily stops Matrix recovery management from managing DR Protected
logical servers or DR Protected IO services. One use of Maintenance Mode is to perform a failover
rehearsal. By setting Maintenance Mode on a Recovery Group, all of the logical servers and IO
services in that Recovery Group can be activated or deactivated from the Matrix infrastructure
service user interface. Once you are satisfied with the failover rehearsal, a Recovery Group and
its corresponding logical servers or IO services can be brought back under the control of Matrix
recovery management by disabling Maintenance Mode on that Recovery Group.
At the Remote Site, the logical servers being managed by Matrix recovery management are known
as recovery logical servers. Normally, a recovery logical server cannot be activated except by
invoking an Activate operation from the Matrix recovery management user interface at the Remote
Site.
To use Maintenance Mode to test a Recovery Group:
1. At the Local Site, use the Logical Servers Deactivate operation in the Tools menu of the
Visualization tab in the HP Matrix Operating Environment to gracefully shut down the logical
servers in the Recovery Group. For IO services, use the Deactivate Servers IO service operation
in the Matrix infrastructure service to shut down the IO services in the Recovery Group.
NOTE: If the Matrix recovery management configuration includes Hyper-V logical servers
or IO services, bring the cluster disk resource used for logical server or IO service storage
offline prior to failing over storage to the remote site. If you did not bring the CSV offline, see
“Recover the CSV from online pending state” (page 71) for how to recover.
2. At the Remote Site, use the appropriate storage management tools, (for example, HP P6000
Continuous Access Software), to fail over the storage corresponding to the Recovery Group
to the Remote Site.
3. If there are logical servers or IO services in the Storage Replication Group that will be activated
on VM hosts during the test:
a. Rescan storage using VM host management tools, for example, VMware Virtual Center,
to ensure that the VM host recognizes the failed over storage.
b. Refresh Virtual Machine resources using the Logical Servers Refresh operation in the Tools
menu of the Visualization tab in Matrix OE visualization.
Testing Recovery Groups 27