Question about SRM volume requirements

[hux2000]

I set up an SRM (5.0.1) proof of concept a few weeks ago and ran into this issue but ended up ignoring it at the time because it wasn't a major problem. Now, however, I need to resolve it. Basically, I had an NFS volume containing a bunch of VMs, only some of which I wanted to recover at the second site via SRM. However, when creating a protection group I was unable to add only those VMs; I could only make it work by adding all the VMs on that volume.

Is this the expected behavior? If not, how can I set up SRM so that I can recover only some VMs on a volume and not others?

Thanks.

[hux2000]

To clarify, when I do "Create New Protection Group", in the "Select one or more datastore groups" section I select my datastore, then all the VMs on that store are listed below, but there's no way to separate the VMs I want to recover from the VMs I don't want to recover. Is this by design? Or am I missing a setting somewhere that enables me to select specific VMs to recover?

[jeff9565 · Accepted solution]

Hux,

This is by design.  When using array based replication, SRM protects virtual machines at the datastore level.  There are really two ways around this:

1. Create a new NFS volume and move the virtual machines that you want to protect to it which will separate your protected from non protected.

2. You can use vSphere Replication which is a new feature inside SRM 5.  Make sure you read the requirements for this.  Quick note on this - Failback is not yet available for vSphere Replication though it is supposed to be implemented in a future release.

Hope this helps!

Jeff

[hux2000]

Thanks Jeff. I've already been separating them on different datastores anyway as a (presumed) workaround so I'll continue doing that. Kind of a drag in terms of NFS volume efficiency/managment, but it'll do.