2010-03-17 11:48 AM
We're thin provisioning our storage and our safety net is first snapautodelete and then vol autogrow. We had some vols (VMware and Exchange) get close to full recently and snap autodelete kicked in and deleted some old snaps. Apparently this screws up the SnapMgr products and the snapshot delete retention.
See reasons #5 and #6 here https://now.netapp.com/Knowledgebase/solutionarea.asp?id=kb55578
NetApp support tells us that we shouldn’t have snap autodelete on any vols being backed up with SnapMgr products. That’s kind of a bummer…takes the least disruptive part of our safety net away for our most critical infrastructure.
Any suggestions or workarounds?
Any plans to make the snap autodelete function ‘SnapMgr aware’ or build some intelligence into it? This would be a nice feature. Seeing as we've made an investment in NetApp technology for our backups now with SnapManager products, they should be highly integrated and not break basic functions or create new limitations.
Please tell me this is in the works!
2010-03-17 01:57 PM
Are you using the snapmanager for exchange fractional_reserve settings to do this? It is builtin to SME to run on the host instead of on the FAS controller. For SMVI there is no integration for autodelete and I submitted a request to make SMVI do the same or recognize snapshots that have been autodeleted...it looks like this will happen but haven't seen a new roadmap on the next SMVI.