Rehearse operations for VMware virtual machines - array-based replication
The requirements for rehearse operations for VMware virtual machines depend on the replication type.
Rehearse operations with EMC SRDF-based replication
Rehearse operations with NetApp SnapMirror based replication
When the rehearse operation is initiated, Resiliency Platform creates point-in-time snapshots, since rehearsal cannot work with existing snapshots. These snapshots are created using the underlying snapshot technology and varies according to the replication configuration. The required snapshot LUNs have to be pre-configured and also mapped to the appropriate VMware ESXi servers.
For EMC SRDF replication configured using device groups, Resiliency Platform supports TimeFinder Snap and TimeFinder Mirror as snapshot technologies. The appropriate LUNs for snapshot have to be associated with the device group.
For example: For Snap, VDEV devices and Mirror, BCV devices have to be associated with the device group in order to perform rehearsal. If the device group is replicated in asynchronous mode, then only TimeFinder Mirror technology is supported.
For EMC SRDF replication configured using Storage Groups and Composite Groups, Resiliency Platform supports TimeFinder Snapvx as snapshot technology.
For rehearsal of SRDF Storage Groups, you have to create an additional Storage Group containing the TGT devices for linking the snapvx snapshot. The name of this Storage Group has to follow a simple convention: it should be similar to the replicated Storage Group name, with a suffix '_vrpsnap' appended to it. E.g. if the replicated Storage Group's name is 'finance_sg', then you need to create a Storage Group with name 'finance_sg_vrpsnap', containing the TGT devices.
The TGT devices have to be exact same in number and size as the RDF devices.
During rehearsal, the Resiliency Platform takes a new snapshot using the configured snapshot technology on the associated snapshot LUNs.
The datastores on the snapshot LUNs are mounted on the hosts in disaster recovery site.
Resiliency Platform registers the virtual machines in the disaster recovery data center for rehearsal. They have identical configuration to the production virtual machines, except these virtual machines consume storage from the datastore mounted using the snapshot volumes and have and have a suffix '-vrprehearsal' in their names. These virtual machines are disconnected from the network and are unregistered during cleanup.
NetApp SnapMirror based replication uses FlexClone for the rehearse operation, so the NetApp storage server must be enabled with the FlexClone license.
When the rehearse operation is initiated, Resiliency Platform creates a point-in-time volume snapshot as part of the rehearsal operation. The snapshot volume is exported and mounted on the DR host.
Note:
The rehearse operation breaks any ongoing replication between the source and destination storage server as the FlexClone operation cannot be performed on the destination read-only volume. SnapMirror replication resumes after the rehearsal cleanup operation.
Resiliency Platform registers the virtual machines in the production data center for rehearsal. They have identical configuration to the DR virtual machines, except these virtual machines consume storage from the datastore that is mounted using the snapshot volumes. These virtual machines are disconnected from the network and are unregistered during cleanup.
More Information