We are looking at aquiring NetApp as our SAN solution and our planned configuration for the SQL\cluster architecture was the same\similar as described here. Of course the sales people said it i was no issue. It sounds like it actually is. I am interested in more information on what you did as a solution, or was it just a work around? Since I am in the beginning stages I want to make sure I set this up right the first time.
I don't want to create any confusion and I mean no disrespect, but the statement from "Welch" that this is a supported configuration is incorrect. The IMT has no stated support for this configuration and I have run this post and question by 2 different Senior Support Engineers and this is their response:
"Combining a SQL FCI with an availability group is not supported in SnapDrive/SMSQL. This is due to a limitation within SnapDrive that attempts to communicate with all nodes defined within the cluster object.
Typically, we would suggest that for these three nodes, each node have dedicated storage, rather than use a failover instance (with shared storage) in the primary location."
In practical terms though - how do we provision LUNs with SnapDrive? AlwaysOn AG operates on top of a 3-node MS cluster, which includes 2 local & 1 remote node. SnapDrive plainly refuses to create a drive associated with local nodes only.