![alt text]Hi guys So i have been given the task to create a clustered environment for our main/critical database. I am new to the topic and thus far from my research would suggest a 3 node window cluster, 2 of which are a sql cluster with shared san storage and the 3rd node on the DR site as a standalone instance with its own storage. Availability groups would be configured on Node1 with secondary replica on the 3rd node in the DR site. San storage is already in place so hardware and cost is not a problem. The server admin is under the impression that the we should have 2 nodes in the cluster, one on the primary site the 2nd on the DR site, each with local storage, eliminating single point of failure on the san and eliminating the need for a 3rd node. Am i correct in saying that we need to use shared storage between the nodes? Appreciate any input on this topic. Thanks in advance. : /storage/temp/4574-sql-failover-and-availability-groups.png
Any chance the server admin is mixing up the Availability groups with Failover Cluster? The storage must be a shared storage between all nodes in the SQL Failover Cluster as all the data and log files of the databases resides in that shared storage. This is why the shared storage is vulnerable to be the single point of failure. Hence, you would need a shared storage for your 2 nodes in the cluster.