Today i go the cluster up with two nodes having local hard disk. I had
some more hack which corrected a node number corruption node number
during system calls. I am getting a Oops at three points now.But that
doesn't cause the system to crash. In fact all the cluster tools
commands are working . The Oops i think is due to the similar
corruption that I was talking before . Once i get those Oops fixed i
will send the clean patch .