We're planning to setup a 2-node (i.e. Node1 & Node2) cluster to host 4 SQL
Server DBs attached to SAN (J: Drive)
The default scenario is to make J: Drive a clustered resource and create a
clustered VIP(VIP1) and make Node1 as Primary owner of all resources and DBs.
In this scenario, Node2 will stay idle most of the time (except in case of
fail over)
In order to make use of Node2, I wonder if the below scenario works:
1- Create another SAN drive (L and make it clustered
2- Create another VIP (VIP2)
3- Split the DB files of the 4 DBs between J: and L:
4- Configure 2 DBs for Node1 and 2 for Node2
So L: drive and VIP2 will be owned by Node2 and will fail over to Node1 if
Node2 goes down.
Same for J: drive and VIP1 that fails over to Node 2.
In this case both nodes will be active but for different DBs.

Is this scenario doable?

Sort of.

The Instance is the unit of failover, not the database. OF course,
databases are owned by instances. You can create two instances and divide
the databases between them. Then you can normally place them on opposing
nodes. Whenever a node fails, the instance on that node automatically fails
over to the other node.


--
Geoff N. Hiten
Principal SQL Infrastructure Consultant
Microsoft SQL Server MVP


"Joseph" <Joseph (AT) discussions (DOT) microsoft.com> wrote