[Linux-cluster] cluster.conf not being propagated to all nodes

Neale Ferguson neale at sinenomine.net
Mon Dec 23 18:57:19 UTC 2013


Sorry, what *distro* and version? Sounds like RHEL / CentOS 6.something
though.

CentOS 6.4

> * Are the modclusterd and ricci daemons running?
> ricci     1439     1  0 11:18 ?        00:00:16 ricci -u ricci
> root      1486     1  2 11:18 ?        00:02:03 modclusterd
>
> * Did you set the local 'ricci' user password on your nodes?
> yes
>
> * Is ccs installed?
> No
>
> Thanks... Neale

Try installing ccs. I'm not sure it will help here, but it's worth a
try. I don't use luci myself so I push out updates using 'cman_tool
version -r'... Do you know if that works if you try from the node you
edited cluster.conf on?

luci is the one which updates cluster.conf and attempts to propagate. It's that propagation to cnode2 that doesn't happen (it doesn't fail - it just doesn't attempt it).




More information about the Linux-cluster mailing list