E4500 cluster reboot & A5200: SUMMARY
2007-12-25 4:43:00
Luca!!
Well it looks like this was some strange failure fencing where the
"failed" node could not see any of the disks mastered by the running
node. But I have never seen this before?
In a nutshell.
A "scadmin startnode" added the node to the cluster but it was not
mastering anything. Use "hastat" to confirm this.
Then did a "haswitch <node-name> <logical-host>"
Either be logged on at the console or tail the messages file to whats
going on. And bingo the servers are now back up and running as they
should be.
Thanks - Julian.
Comments
Got something to say?
You must be logged in to post a comment.
