It's not absolute that you will have DL/DU if you lose two disks in the same pool - only if the two devices happen to hold the only two copies of any part of the data necessary required to present a consistent volume. Volumes don't appear to be necessarily split across every disk in the pool, not in my experience anyway.

Not aware of how to list volumes that are utilised by certain disks. I don't think it's something you can query - I raised a similar call with EMC some time back, and the response was that the system changes so much, a report taken one minute could be different a short time later - the implication being that such a report would largely be of no value.

So how to find out which volumes go offline if you lose two disks? Sadly, it appears to be that you'll only know when it happens.

Maybe the querying capability has changed in current version, will find out when I apply it.

Hi Andy;

Thanks for your feedback. Iwas raising this question regarding an overall design. Should we put all the drives in the same pools or not (same type of drivers of course). Putting them in the same pull will result in a large failure domain and in a case of double device failure we can end up with the hole platform going offline. This will result in a full restore. So imagine restoring 200 TB VS 20 TB. Would you use multiple storage pool or go with the recomanded one.

Remember pools are meant for different drive types. If you want a failure domains, you can always create at-least 2 Protection Domains. You can add some sds in PD1 and the rest in PD2.

As i understood from EMC presales is that the copy always happends in the same storage pool. If so having the failure domain is the Storage pool not the Protection Domain.