LinuxQuestions.org

LinuxQuestions.org (/questions/)
-   Linux - Server (http://www.linuxquestions.org/questions/linux-server-73/)
-   -   Debian Corosync/Pacemaker Cluster Frustrations (http://www.linuxquestions.org/questions/linux-server-73/debian-corosync-pacemaker-cluster-frustrations-943754/)

mpapet 05-07-2012 04:29 PM

Debian Corosync/Pacemaker Cluster Frustrations
 
My goal is to have a clustered IP and an iSCSI disk that moves with the cluster IP address. I'm trying to get a Corosync/Pacemaker cluster going and nothing is quite as it seems based on the documentation.

The clustered IP address will be in 192.168.1.xxx.
I have a separate subnet for the cluster traffic 192.168.10.xx and one more for redundancy 192.168.11.xx.

Both hosts can ping each other over all their respective subnets. Nothing wrong with /etc/hosts to confuse the hosts.

Starting one node gets me the following:
Code:

============
Last updated: Mon May  7 14:24:06 2012
Last change: Mon May  7 14:06:40 2012 via cibadmin on mapper
Stack: openais
Current DC: mapper - partition WITHOUT quorum
Version: 1.1.7-ee0730e13d124c3d58f00016c3376a1de5323cff
2 Nodes configured, 2 expected votes
2 Resources configured.
============

Node mapper (be549e8f-9fa4-4317-894d-b2193135b7b3): UNCLEAN (offline)
Online: [ mapper ]

 kshqcla-stonith        (stonith:suicide):      Started mapper

Why is the same node listed twice? I start the other node and it joins the cluster vote count goes to 3. However everything is in in offline unclean status. How do you get a cluster node out of unclean offline status? I can't find anything that explains this.

Attempts to start the other node crashes both nodes.

Here's the only thing I've changed about a default corosync.conf
Code:

        # This specifies the mode of redundant ring, which may be none, active, or passive.
        rrp_mode: active

        interface {
                member {
                        memberaddr: 192.168.10.1
                }
                member {
                        memberaddr: 192.168.10.4
                }
                # The following values need to be set based on your environment
                ringnumber: 0
                bindnetaddr: 192.168.10.0
                mcastaddr: 226.94.1.1
                mcastport: 5405
        }
        interface {
                mamber {
                        memberaddr: 192.168.11.2
                }
                member {
                        memberaddr: 192.168.11.1
                }
        ringnumber: 1
        bindnetaddr: 192.168.11.0
        mcastaddr: 226.94.1.1
        mcastport: 10405
        }

service {
        name: pacemaker
        ver: 1
}


EricTRA 05-09-2012 12:40 AM

Hello,

First thing that pops out is the second interface (in the 192.168.11.x assignment). If you copy/pasted that from the configuration file, then change mamber to member and reload corosync. Can you post the missing part (totem definition, logging if any) of your corosync.conf file? I'm not sure about the service statement inside the corosync.log file. I always create a separate file for my services in /etc/corosync/service.d. Also post output from:
Code:

crm configure show
make sure to mask confidential info if applicable (public IP etc...)

Kind regards,

Eric


All times are GMT -5. The time now is 09:55 AM.