On Fri, Jan 13, 2012 at 03:40:13PM +0000, Christine Caulfield wrote: > On 13/01/12 15:37, David Teigland wrote: > >On Fri, Jan 13, 2012 at 10:37:54AM +0000, Christine Caulfield wrote: > >>>mandated it. In fact you can run cman in RHEL4 and RHEL5 without any > >>>nodes list if you don't need fencing. 'cman_tool join -X' > > > >... and don't need proper quorum. > > > > Sorry. What, in this context is "proper" quorum? and did VMS not > have it then? An expected_votes value the properly includes all possible votes. (You could say join -X -e <ev> but that's begging the question.) Correct me if I'm wrong, but I believe VMS has an explicit EXPECTED_VOTES setting, which we obviously allow also. I doubt anyone would argue that setting that number directly would be less error prone than computing it from a node list. > >Has the list of nodes specifically in cluster.conf been a source of > >mistakes and support calls over the years? My belief is that a lack of a > >node list would make the cluster more *difficult* to manage, not easier, > >which is mainly why I think it should be there. As I said before, where > >do you look for a list of all the nodes that *should* be in your cluster > >when you think there might be one or two missing? On your whiteboard? > > Yes we do get calls about people misunderstanding cluster.conf, > putting the wrong node name in there so that cman doesn't start or > using the 'wrong' interface. Right, but those problems would not go away without a list of nodes, in fact they'd probably be much much worse since you'd then have to have a different corosync.conf file on each node, each specifying different, locally correct values. I suspect the support call volume would explode. _______________________________________________ discuss mailing list discuss@xxxxxxxxxxxx http://lists.corosync.org/mailman/listinfo/discuss
