I know this has been beat to death on this list but I have not found very clear steps in the archives or even all the doc's and FAQ linked to. Many times I've seen people say with "good" heuristics and voting settings but have not found examples of what numbers you should use, and why. The two node example in the FAQ is somewhat helpful, but what about a 3+ node? This test cluster is 3 nodes, RHEL 5, 200gb gfs shared disk, 100mb quorum disk So you should start out with: mkqdisk -c /dev/sdc -l nfsquorum chkconfig qdiskd on Add this to cluster.conf - is this what you should do for heuristics? What about the votes/scores? <quorumd interval="1" tko="10" votes="2" label="nfsquorum"> <heuristic program="ping 10.0.0.167 -c1 -t1" score="2" interval="2"/> <heuristic program="ping 10.0.0.168 -c1 -t1" score="2" interval="2"/> <heuristic program="ping 10.0.0.170 -c1 -t1" score="2" interval="2"/> </quorumd> Modify the cman statement, but what should expected votes be? <cman two_node="0" expected_votes="5"/> Then reboot one node at a time? Does qdiskd need to be started at a certain run level, I think it's trying to start before cman on this cluster I'm testing. -- Josh Gray Systems Administrator NIC, Inc Email: jgray@xxxxxxxxxx Desk/Mobile: 913-221-1520 "Talent is cheaper than table salt. What separates the talented individual from the successful one is a lot of hard work." - Stephen King -- Linux-cluster mailing list Linux-cluster@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/linux-cluster
