topic Re: P4500 Replication Question in StoreVirtual Storage

P4500 Replication Question

L1nklight — Mon, 14 May 2012 18:34:48 GMT

I got off the phone with an HP engineer and I didn't really think much about it until just now. So the engineer mentioned that if I am remote replicating via snapshots to my DR P4500 I cannot use that P4500 to say, run a VMware environment. For example, my production environment in Virginia is replicating to my enviroment in Colorado. I am doing replication on 3 volumes. All the volumes are in the same cluster. The cluster is in a single management group (obviously). I then replicate these to a mangement group in Colorado. The 3 volumes are small in size, only 500 gigs each. This leaves me a sizeable storage space left. I want to run a development environment in Colorado, to take advantage of the 2 TBs of available storage and the inactive vSphere environment. Under that scenario, I was told by the HP engineer, that the cluster in that management group (the DR mangement group), could not be used for a VMFS store for the dev environment because it is currently recieving snaps from Virginia.

Is this true? Does this scenario make sense? If this is true, I feel like I have been sold a bag of goods that doesn't work for what I had initially specc'd out to my sales engineering group.

Re: P4500 Replication Question

David_Tocker — Mon, 14 May 2012 21:24:31 GMT

Doesnt sound right to me - the P4000s are very good at multitasking - I think the engineer may have it wrong.

I would take care not to over provision however.

Re: P4500 Replication Question

L1nklight — Tue, 15 May 2012 12:22:12 GMT

It was something about having an iSCSI session active to a cluster for replicating. Like replication locked out a cluster or something... I put a call into HP for clarification.

Re: P4500 Replication Question

BulkRate — Tue, 15 May 2012 14:17:45 GMT

That's...not right. If you get a chance to, tell the engineer that they're in need of spending some quality time with the product he or she supports.

I've got a DR cluster in a VSA-based management group that's the receiving end of all my remote office's Remote Snapshots. There's nothing that prevents me from mounting one of those snapshots in a running VM to read data, or creating other volumes local to that cluster in the remaining space and using them to host VM's. Now, given that the P4000 eats free space like popcorn to protect your data across nodes you do need to watch out for overcommitment.

Wow...if that engineer were right, how'd you be expected to backup/restore information from those remote snaps if you're expected to be unable to access them?

Now, having a Remote Volume's snapshot mounted and in use when your schedule's retention limit is reached could cause some hilarity to ensue. I'm not sure how that particular conflict would be resolved by SANiQ...my bet is that you'd continue to accrue snapshots and potentially run out of space in the cluster.