On Thu, Nov 8, 2012 at 2:06 PM, Andrew Beekhof <and...@beekhof.net> wrote:
> The service directive that loads the pacemaker plugin appears to have been
> lost.
> Therefor Pacemaker is confused and thinks that you must be trying to
> start it with Heartbeat.
I've just made a change that will help Pacemaker log something more
accurate and relevant.
If you saw:
"error: Corosync is running, but Pacemaker could not find the CMAN or
Pacemaker plugin loaded"
would it have been more helpful?
>
> On Thu, Nov 8, 2012 at 1:38 PM, Jeff Johnson <echo5jul...@gmail.com> wrote:
>> Greetings,
>>
>> I had a running corosync/pacemaker two node configuration doing simple
>> filesystem failover and stonith fencing. I decided to update to 1.1.8-4
>> after seeing some odd behavior and someone suggested a bug in 1.1.7-6 not
>> playing nice with corosync 1.4.1-7. After updating my cluster will not
>> initialize. The first node starts fine, corosync and pacemaker start via
>> init scripts and resources show in good status. Rings are in good status as
>> well. When I start corosync and pacemaker on the second node corosync starts
>> fine but pacemaker fails to start. /var/log/messages reads:
>>
>> corosync[3344]: [MAIN ] Corosync Cluster Engine ('1.4.1'): started and
>> ready to provide service.
>> corosync[3344]: [MAIN ] Corosync built-in features: nss dbus rdma snmp
>> corosync[3344]: [MAIN ] Successfully read main configuration file
>> '/etc/corosync/corosync.conf'.
>> corosync[3344]: [TOTEM ] Initializing transport (UDP/IP Multicast).
>> corosync[3344]: [TOTEM ] Initializing transmit/receive security:
>> libtomcrypt SOBER128/SHA1HMAC (mode 0).
>> corosync[3344]: [TOTEM ] The network interface [10.0.0.102] is now up.
>> corosync[3344]: [SERV ] Service engine loaded: corosync extended virtual
>> synchrony service
>> corosync[3344]: [SERV ] Service engine loaded: corosync configuration
>> service
>> corosync[3344]: [SERV ] Service engine loaded: corosync cluster closed
>> process group service v1.01
>> corosync[3344]: [SERV ] Service engine loaded: corosync cluster config
>> database access v1.01
>> corosync[3344]: [SERV ] Service engine loaded: corosync profile loading
>> service
>> corosync[3344]: [SERV ] Service engine loaded: corosync cluster quorum
>> service v0.1
>> corosync[3344]: [MAIN ] Compatibility mode set to whitetank. Using V1
>> and V2 of the synchronization engine.
>> corosync[3344]: [TOTEM ] A processor joined or left the membership and a
>> new membership was formed.
>> corosync[3344]: [CPG ] chosen downlist: sender r(0) ip(10.0.0.102) ;
>> members(old:0 left:0)
>> corosync[3344]: [MAIN ] Completed service synchronization, ready to
>> provide service.
>> corosync[3344]: [TOTEM ] A processor joined or left the membership and a
>> new membership was formed.
>> corosync[3344]: [CPG ] chosen downlist: sender r(0) ip(10.0.0.101) ;
>> members(old:1 left:0)
>> corosync[3344]: [MAIN ] Completed service synchronization, ready to
>> provide service.
>> pacemakerd[3364]: notice: get_cluster_type: This installation of Pacemaker
>> does not support the 'heartbeat' cluster infrastructure. Terminating.
>> cibadmin[3376]: notice: crm_log_args: Invoked: cibadmin -l -Q
>> corosync[3344]: [SERV ] Unloading all Corosync service engines.
>>
>> I have no heartbeat rpms installed on the nodes in the cluster. I don't know
>> why pacemaker is deciding the configuration is 'heartbeat' instead of
>> corosync.
>>
>> I built the pacemaker rpms from the clusterlabs git tree. I used the same
>> combination of with/without options that 1.1.7-6 was compiled with. I
>> verified this by running a rpm rebuild on the pacemaker 1.1.7-6 source and
>> looked at the output to determine the with/without options.
>>
>> I have checked networking between the nodes and the function of the
>> multicast address as well. 1.1.7-6 works, 1.1.8-4 does not.
>>
>> OS/Software info:
>>
>> RHEL6.2 x86_64
>> corosync 1.4.1-7
>> cluster-glue 1.0.5-6
>> clusterlib 3.0.12
>> libqb 0.14.2-1
>> pacemaker 1.1.8-4
>> openais 1.1.1-7
>>
>> corosync.conf (if needed)
>>
>> # Please read the corosync.conf.5 manual page
>> compatibility: whitetank
>>
>> totem {
>> version: 2
>> secauth: off
>> threads: 0
>> interface {
>> ringnumber: 0
>> bindnetaddr: 10.0.0.0
>> mcastaddr: 226.94.1.1
>> mcastport: 5405
>> ttl: 1
>> }
>> }
>>
>> logging {
>> fileline: off
>> to_stderr: no
>> to_logfile: yes
>> to_syslog: yes
>> logfile: /var/log/cluster/corosync.log
>> debug: off
>> timestamp: on
>> logger_subsys {
>> subsys: AMF
>> debug: off
>> }
>> }
>>
>> aisexec {
>> user: root
>> group: root
>> }
>>
>> amf {
>> mode: disabled
>> }
>>
>>
>> Any help or suggestions are greatly appreciated!
>>
>> --Jeff
>>
>>
>> _______________________________________________
>> Pacemaker mailing list: Pacemaker@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://bugs.clusterlabs.org
>>
_______________________________________________
Pacemaker mailing list: Pacemaker@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://bugs.clusterlabs.org
