Loss of connectivity to OCR from all nodes in cluster
644720Sep 23 2008 — edited Dec 13 2011Hello,
I recently created a 2 node RAC cluster using Oracle 10G on windows 2003. Shared storage is an XP SAN using raw disk partitions for the OCR , Voting Disks and ASM storage. My problem is that I have now lost 'connectivity' with the OCR from both nodes and a result none of the dependent services will now start.This 'appears' to have been as a result of rebooting both nodes at various times
Running CLUVFY COMP -SSA -n racnode1,racnode2 returns the expected output suggesting that the shared storage is accessable from both nodes.
Running OCRCHECK returns a PROT-206 error saying that the OCR cannot be read or is not accessible
I am unable to recover the OCR from any backups that exist using the OCRCONFIG -restore command
CRSCTL Stop/Start CRS doesnt work as it cant communicate with the daemon.
Has anyone ever experienced anything similiar to this situation or can offer any advice on how I can attempt to resolve the problem ? My feeling is that permissions on the disk storage may be involved, but as we are using Windows and not UNIX in this case I dont really know where to start looking.
Thanks,
Shaun.