Question : SQL Clustering

Correct me if I'm wrong, in a SQL Server cluster, there are multiple SQL nodes that all access a set of shared disks (SAN).  therefore, since the actual data resides in one location, the point of failure would be the SAN.

What are recommened approaches to protect from this point of failure?  should we use sql clustering along with mirroring?

Answer : SQL Clustering

>therefore, since the actual data resides in one location, the point of failure would be the SAN.
correct

> should we use sql clustering along with mirroring?
well, SAN already does mirroing and striping

what you are trying to protect from is if the SAN box and/or the path between the nodes to the SAN box fails.
you will need redundant fibre channel connections for the second part.
for the SAN box to fail itself, you would indeed need a full failover to a second SAN would be required (aka disaster recovery scenario)

http://searchstorage.techtarget.com/expert/KnowledgebaseAnswer/0,289625,sid5_cid573587,00.html