[Pacemaker] Some questions about corosync-cfgtool

Dan Frincu dfrincu at streamwide.ro
Wed Jan 19 07:34:38 UTC 2011


Hi,

https://lists.linux-foundation.org/pipermail/openais/2011-January/015626.html

xin.liang at cs2c.com.cn wrote:
> hi everybody!
>
>     I have some questions about corosync-cfgtool
>    
>     1. What should I do when "corosync-cfgtool -s" return "Could not 
> initialize corosync configuration API error" ? Restart corosync ?(I 
> don't think it's a good idea)
>
>     2. How can the process happen automatically when network problems 
> is repaired, instead of using "corosync-cfgtool -r" manually?
>
>
>
>     My testing environment is :
>     2 PC (on virtualbox-3.2.12)
>     Double network card and double heart-beat link(eth0 and eth1)
>     OS: RHEL 5.3 x86
>     primary rpms: corosync-1.3.0 and pacemaker-1.0.10
>     corosync.conf:(relevant portion)
>    
>         compatibility: whitetank
>
>         totem {
>                 version: 2
>                 secauth: off
>                 threads: 0
>                 rrp_mode: active
>                 interface{
>                         ringnumber:0
>                         bindnetaddr:10.10.10.0
>                         mcastaddr:235.3.4.5
>                         mcastport:9876
>                         }
>                 interface{
>                         ringnumber:1
>                         bindnetaddr:20.20.20.0
>                         mcastaddr:235.3.4.6
>                         mcastport:9877
>                 }       
>         }
>
>
>
>    
>     When something bad happen on one of the double-heart-beat-link
>     --------just like: ifdown eth0, or pull out one of the netting 
> twine-----
>
>     I use:(ifdown eth1 before)
>         corosync-cfgtool -s
>     found that:
>         Printing ring status.
>         Local node ID 185207306
>         RING ID 0
>         id    = 10.10.10.11
>         status    = ring 0 active with no faults
>         RING ID 1
>         id    = 20.20.20.11
>         status    = Marking seqid 14089 ringid 1 interface 20.20.20.11 
> FAULTY - adminisrtative intervention required.
>     then:
>         ifup eth1
>     and:
>         corosync-cfgtool -r
>
>     The problem is repaired.
>
>     BUT, I want this process happen automatically when network 
> problems is repaired, I write this shell, and start it when service 
> corosync start:
>
>  ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~    
>
>     #!/bin/bash
>     local_is_down=0
>     ip_res=`corosync-cfgtool -s|grep id|awk '{print $3}'`
>     while [ -z "$ip_res" ]
>     do
>             sleep 5
>             ip_res=`corosync-cfgtool -s|grep id|awk '{print $3}'`
>     done
>
>     ip_num=`corosync-cfgtool -s|grep "no faults"|wc -l`
>
>     while true
>     do
>             sleep 10
>             res=`corosync-cfgtool -s`
>             echo "$res" |grep FAULTY &> /dev/null
>             if [ "$?" -ne 0 ];then
>                     tmp_num=`echo "$res"|grep "no faults"|wc -l`
>                     if [ "$tmp_num" -eq "$ip_num" ];then
>                             local_is_down=0
>                     else
>                              continue
>                      fi
>             else
>              mii-tool |grep "no link" &> /dev/null  
>                     if [ $? -eq 0 ];then
>                            local_is_down=1         #pull out one of 
> the netting twine
>                     else
>                            for IP in $ip_res
>                            do
>                                    ifconfig|grep "$IP" &> 
> /dev/null            
>                                    if [ $? -ne 0 ];then
>                                            local_is_down=1    #ifdown
>                                     fi
>                             done
>                        fi
>
>                           if [ $local_is_down -eq 1 ];then
>                               corosync-cfgtool -r &> /dev/null
>                        fi
>              fi
>     done &
>  ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>
>     The shell works well mostly, however, it does not work sometimes 
> because command "corosync-cfgtool -s" return
>     "Could not initialize corosync configuration API error 6"
>
> and the pacemaker process seems also done.
>
>
>  THANKS!
> ------------------------------------------------------------------------
>
> _______________________________________________
> Pacemaker mailing list: Pacemaker at oss.clusterlabs.org
> http://oss.clusterlabs.org/mailman/listinfo/pacemaker
>
> Project Home: http://www.clusterlabs.org
> Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf
> Bugs: http://developerbugs.linux-foundation.org/enter_bug.cgi?product=Pacemaker
>   

-- 
Dan FRINCU
Systems Engineer
CCNA, RHCE
Streamwide Romania

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.clusterlabs.org/pipermail/pacemaker/attachments/20110119/0788ca2d/attachment-0002.htm>


More information about the Pacemaker mailing list