Disaster Recovery with Emc Clariion and Mirrorview

[stormvm]

Hi

We have recently installed mirrorview on both our clariion cx3-40c and have set up some test replication of vmware luns.

The process in which I have been tested this is as follows

1. Set up vmware lun remote mirror from primary clariion cx3-40 to secondary clariion cx3-40

2. Stop I.O to source lun and remove from primary clariion storage group

3. Promote secondary image on secondary clariion to become the new Primary

4. Add new primary to storage group on secondary clariion storage group

5. On esxhost which will see the mirrored lun presentation change advanced parameter LVM.DisallowSnapshotLUN from 1 to 0 to allow datastore presentation

6. Re-inventory machines on host

This process works and all vm's can be seen on their secondary storage array, however ultimately my question is this I can't find much information or best practise regarding the LVM advanced paramaters this method works but I'm after whether I can leave the LVM.DisableSnapshotLun setting to 0 on all esxhosts or whether I should be setting LVM.Enableresignature to 0 and then renaming the datastore instead ?

can anyone assist in which settings are optimum and best practise for this kind of D.R scenario.

Points will be awarded many thanks

[chrisy]

It might depend on what your failback plans are. Suppose you fail to the remote mirror, and run with it for a few days. What are your plans for getting those changes back to the primary site - if they involve replication, you might want to stick with a strategy that avoids changing, renaming or resignaturing the LUN so you end up with the original name and signature.

Your current approach seems good.