Replication requirements for EMC SRDF
Before it performs any tasks, the wizard validates the array configuration as follows:
-
On the primary site, the wizard verifies that no devices are RDF2.
-
On the secondary site, the wizard verifies that no devices are RDF1.
Otherwise, the wizard displays an invalid configuration message and is unable to proceed.
The DR wizard does not start or stop replication. Array replication configuration is not a prerequisite for the wizard to perform storage cloning or service group cloning.
After the service group cloning task is complete, the DR wizard displays a screen describing the following replication requirements:
-
All disks in SFW disk groups must belong to the same device group.
-
The device group must not span more than one array (no composite device groups).
-
A device group can contain one or more disk groups.
-
Dynamic swap must be enabled on both sites.
-
On the primary site:
-
All devices must be RDF1 and part of an RDF1 device group.
-
Devices must have write access.
-
-
On the secondary site:
-
All devices must be RDF2 and part of an RDF2 device group.
-
Write access must be disabled.
-
It is recommended that you ensure that these requirements are met before proceeding with the wizard. The wizard then validates the array replication configuration.
If replication is configured correctly, the wizard populates the resource configuration screen with the required replication settings for the SRDF resource.
If the replication configuration does not meet the requirements, the wizard leaves the fields on the resource configuration screen blank. You can optionally enter the resource configuration information in the wizard and configure the array replication requirements later. The information you enter is applied only to the SRDF resource, not to the array configuration. However, the SRDF resource will be unable to come online in the service group until replication has been configured correctly.
Note: In addition, the agent requires that the device group configuration must be the same on all nodes of the cluster.
Related information