What you suggest sounds similar to what I've done in the past.

Create Backups

Attach new LUNs to both nodes, make sure they can fail back and forth in a different Cluster...

Make sure that Node 3 is the preferred owner for INS2, INS3, and INS4 in Cluster administrator, also make sure that you have the failback option enabled.

Hope that helps

That did the trick. Thank you so much, it's been frustrating the heck out of me.

I believe what I'm trying to do should be completely supported. Basically I'll have steps in a job. If any of the steps fail they will go to...

All three of the methods described in that article require restarting the SQLServer service. I want to know a way to do this without having to restart it.