I am using fail over cluster with share storage for Oracle Db and another application. Actually by putting reasonable time delay on the script, I am able to delay the service to start after the mount point completed.
The service script are in the /etc/init.d in each system but the database and another application are running on the share storage.
On Sat, 2006-02-18 at 14:00 -0500, DYM3FYEHE aol com wrote:
> I am using RHCS V4 for Oracle database and another application running
> on the two node failover cluster system. The application has different
> services and installed in the shared storage. The cluster manager
> starts the services before the shared drives mounted to the second
> node. I use script to start and stop the service with time delay but
> still the service start before the shared storage moved to the second
> node and failed. Is any body knows how I can control the storage
> resource and the service? I want to make the service to wait until the
> shared storage available in the second node.
I'm not exactly sure what you're trying to do. Shared storage is
expected to be concurrently accessible by all nodes at the same time.
Are you trying to use one service as shared storage for another service?
Having a delay can be racy...
I'm sure there's an answer to your question. :)
Linux-cluster mailing list
Linux-cluster redhat com