On 10/29/2012 10:09 PM, Josh Bowling wrote: > Cluster configuration: > My cluster consists of two machines: carbon and oxygen. > I use these two hostnames in the DRBD config and have defined their IPs
DRBD's configuration is separate from corosync, and it's corosync that handles cluster membership and fencing. DRBD should have the 'crm-fence-peer.sh' fence-handler defined. So in Pacemaker, are you using carbon or carbon.domain.com? > in /etc/hosts (crossover IPs - eth1): > carbon 10.128.1.1 > oxygen 10.128.1.2 > > Each of these hosts also has their fullname defined in /etc/hosts (lan - > eth0): > carbon.domain.com 192.168.10.151 > oxygen.domain.com 192.168.10.152 So whichever you used in corosync will determine which network is used for the cluster communications. > gethostip results: > I just ran "gethostip -d carbon" and got 127.0.1.1. > I check /etc/hosts and I noticed that on the second line (above my > handmade definitions), carbon was set to 127.0.1.1. You fixed this, I assume? > Also, when I run "uname -a", the hostname it returns is carbon, not > carbon.domain.com. Maybe I should change the hostname to > carbon.domain.com since IPMI is using eth0 and it is already configured > to use 192.168.10.151. > > Ping results: > I am able to ping both 192.168.10.151 and 10.128.1.1 from both machines > with no problem. Can you check the IPMI power status using 'ipmitool'? -- Digimer Papers and Projects: https://alteeve.ca/w/ What if the cure for cancer is trapped in the mind of a person without access to education? _______________________________________________ Linux-HA mailing list Linux-HA@lists.linux-ha.org http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems
