I was curious if anyone has run into a similar issue.
We have a 2 node cluster with a GFS file system running Oracle 10g (not RAC). The DB crashed and RHCS failed over the service as expected, but Oracle couldn't start correctly because of an exiting lock on one of the files. We had to unmount the GFS from both nodes, then mount. At that point, Oracle start up correctly.
I'm assuming that the lock was in place during the crash and DLM was honoring it and that's what cause Oracle to not start correctly. Then by unmounting DLM dropped all locks.
Any recommendations on how to avoid this? Is it a good idea to set up the GFS resource with a force unmount to avoid this scenario?
TIA
-Jeremy
-- Linux-cluster mailing list Linux-cluster@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/linux-cluster
