Domino on a SAN

I have read the numerous posts on domino in a SAN environment and I am very comfortable with how it operates. What I am wondering is can multiple front end servers utilize the same data set on the SAN? It would seem logical that it can be done since the front end servers running the domino server instances will access the data directories through a simple drive mapping. Using a clarion SAN on the back end for data storage with redundant fibre channel connectivity should eliminate the need for multiple data sets on the same SAN. Since the data set on the SAN achieves it’s redundancy through the SAN architecture then there is no need for multiple data sets. The front end servers will be configured for cluster failover or load balancing - we currently have the stand alone servers configured for failover only. Any experiences out there?

Subject: Domino on a SAN

Sounds like what you want to do is have two seperate Domino Servers (A and B) using Domino level clustering, with both servers working from a single data directory on the SAN. As far as I know, that will not work due to file contention. If Domino Server A has the files open, Server B will be unable to access them.

What you could do is use Windows service level clustering to have 2 front end boxes configured as Domino Server A, both using the same data directory on the SAN. This has to be an active/passive cluster. If the first instance of Domino is detected as down then instance B should respond and start serving the files centralized on your SAN.

I agree that having both members of a Domino cluster on the same SAN seems less than optimal. What we do is have cluster members reside on seperate SANs housed in seperate datacenters.

Keep in mind that Domino’s application level clustering can be used to offer data redundancy outside of what a single SAN offers. If you can house 2 Domino servers in different locations then you can provide clients with protection against SAN failure (or downtime), localized power failure, localized network failure, etc.