There are two things in the RPMs we should tweak to get them working better with RedHat Cluster Suite.
First, we shouldn't automatically run 'chkconfig openvista on' and 'chkconfig openvista-databases on' in the postinst for openvista-utils. While automatically configuring these two scripts to run when the system is booted is good when you're on a single machine, it's not as good when you're in a clustered environment, because rgmanager is supposed to start/stop services. Technically, it shouldn't matter because both scripts are no-ops when /opt/openvista is empty, but we should be consistent with most of the other RPMs out there and not automatically run chkconfig.
The other thing we need to do is support the 'status' argument in both init scripts. This would help us comply with Debian policy, too. For openvista-databases, we can probably just print 'running' if there are any GT.M processes running on the system. I'm not sure what the openvista script should do... maybe print listening ports and the associated PIDs? We should adhere to LSB return codes where possible -- both RedHat and Ubuntu appear to follow LSB for the status action. See http://dev.linux-foundation.org/betaspecs/booksets/LSB-Core-generic/LSB-Core-generic/iniscrptact.html
We ended up punting on implementing the status argument in a meaningful way -- see http:// medsphere. org/thread/ 1482