Has anyone come across setting up I/O fencing on the 2 EMC arrays disks,
which each connected a SAN Swich, and to a Solaris 8 system have 2 HBA card?
After I have follow the steps in the VCS installation guide on configurating
I/O fencing and do a reboot on my 2 node cluster systems, one of my node
"had" daemon will be kill off when the other node is up. I got to do a manual
"hastart" for the cluster to join the member and start my service groups.
The error msg that I can find on the killed node is : "VxFEN driver not configured,
VCS stopping. Restart the VCS manaully after configured the fencing".
Can anyone advise me wheter I/O fencing can be done with DMP on a SAN enviornemt
so that I can isolate the issue??
Thaks a lot...