On a 3 node RAC, are there problems with keeping one instance down and two up most of the time?
This is 11.2.0.1 RAC (ASM) on rhel 5. We're asked to create RAC databases with 3 instances, and then to shutdown one of these instances right away. The plan is to have the down instance available to start in case one of the other two instances fails; the customer always has a 2-node RAC. In my current dev databse, instance 1 is down; instances 2 and 3 are up.
RMAN backups are made daily. RMAN retention policy is a recovery window of 3 days, which RMAN uses to maintain the archived logs and backups.
Why do I have an uneasy feeling about this setup? Are my backups good, or can keeping a thread perpetually closed create problems on restore and recovery? I confirmed that thread #1 is status closed in gv$thread.