Hello All,
We have a two node veritas clustered VM and there is a physical RDM shared between these VMs using physical bus sharing. Now the pointer file is on a datastore where we are facing issues due to locking, we are unable to start vms from that particular datastore, one of these two VMs working folder is from the datastore and that VM is down, veritas cluster is running with secondary node where the pointer file is from affected datastore, which is up and running fine.
Hardwareassistedlocking is enabled and we are not sure why the lock is getting released, vmware suggested reboot of all the hosts, and storage team is suggesting unmap the lun from all ESXi hosts and remap it.
Now the question is unmapping the datastore which has the rdm pointer file and remapping will cause any signature issue at veritas level which will cause any data corruption/loss, I heard from few engineers that unmapping and remapping will cause issue at veritas level which is very sensitive, I am not sure how veritas work or handle this situation.
Please help...
Regards,
Suresh
https://vconnectit.wordpress.com/
