HP P6000 Replication Solutions Manager Administrator Guide (T3680-96069, June 2012)

• A suspended DR group cannot be failed over and its members cannot be removed. The
exception is that if the intersite links are broken, a suspended DR group can be failed over.
• A DR group cannot be deleted if a member of the destination DR group is presented.
To be added to a DR group, a virtual disk:
• Cannot be a member of another DR group
• Cannot be a snapshot or a mirrorclone
• Must be in a normal operational state
• Must use mirrored cache
• Should not be greater than 2 TB (virtual disks greater than 2,048 GB are not supported as
DR group members)
• Cannot be a thin provisioned virtual disk
Failover
In the context of remote replication, failover reverses replication direction in a DR group pair. The
destination DR group (DR group B) assumes the role of the source, and the source DR group (DR
group A) assumes the role of the destination. For example, if DR group A is replicating to DR group
B, a failover would cause data to be replicated from DR group B to DR group A. You can fail over
a single DR group or you can fail over multiple DR groups simultaneously. You must use a managed
set to fail over multiple DR groups.
NOTE: Failover can take other forms in the array environment:
• Controller failover occurs when one controller in a pair assumes the workload of the failed or
redirected controller in the same cabinet.
• Fabric or path failover occurs when I/O operations transfer from one fabric or path to another.
For more information about failover, see the HP P6000 Continuous Access Implementation Guide.
Vraid hierarchy and redundancy
Vraid levels provide the following relative redundancy:
• Vraid6 offers the highest redundancy level and provides the best level of data protection.
• Vraid1 offers a high redundancy level and uses the most raw capacity.
• Vraid5 is a medium redundancy level. It uses less raw capacity than Vraid1 but more raw
capacity than Vraid0.
• Vraid0 is the lowest redundancy level and uses the least raw capacity.
NOTE: For optimum availability in an HP P6000 Continuous Access configuration, HP recommends
that you do not use Vraid0 at any time and use Vraid5 only on arrays with eight or more disk
enclosures. For more information, see the configuration best practices white paper for your array
model.
Round robin replication
You can use the Host Volume Local Replication Wizard to create a job that will create snapshots
and snapclones of host volumes, and then schedule the job to run periodically. The Host Volume
Replication Wizard allows you to establish the round robin replication and specify the maximum
number of replicas to be retained before automatic deletion occurs. You can restore data from any
of the replicas.
Remote replication concepts 17