[Pacemaker] Debug message granularity

[Ron Kerry]

I think I did not phrase my question correctly. It appears that pacemaker and subfunctions like 
crmd, lrmd, pengine and resource agent script themselves inherit their logging level from corosync 
as defined in corosync.conf. The logger_subsys directive in that file offers some help but it 
appears the subsys it talks about only refer to various corosync subsystems (and there are over a 
dozen of them I have counted so far).

What I want to be able to do is to leave corosync message levels at the default info level, but 
drive the message level for things like crmd, lrmd, pengine and/or the resource agent scripts down 
to the debug level. I do not see how to use the logger_subsys{} setting to do this and I see nothing 
in pacemaker itself that gives me any separate controls over its logging verbosity.

- Ron



On 5/15/12 7:30 PM, Andrew Beekhof wrote:
> You should be able to define
> PCMK_trace_functions=nction1,function2,... as an environment
> variable to get additional information from just those functions.
> It might take a bit of searching through source code to find the ones
> you care about, but it is possible.
>
> On Wed, May 16, 2012 at 5:58 AM, Ron Kerry<rkerry at sgi.com>  wrote:
>> I have a cluster running SLES11 SP2 including SLE-HAE-SP2. This includes the
>> following software levels:
>>   resource-agents-3.9.2-0.25.5
>>   openais-1.1.4-5.6.3
>>   corosync-1.4.1-0.13.1
>>   pacemaker-1.1.6-1.29.1
>>   pacemaker-mgmt-2.1.0-0.8.74
>>   pacemaker-mgmt-client-2.1.0-0.8.74
>>
>> The only way I know to get debug level messages is to set debug:on in
>> /etc/corosync/corosync.conf
>>   logging {
>>         #Whether or not turning on the debug information in the log
>>         debug:  on
>>   }
>>
>> Is there any way to get more granularity? I would like to get debug level
>> information out of my resource agents and perhaps other parts of pacemaker
>> like pengine, crmd and lrmd, but not from corosync. Turning debug:on is very
>> noisy, but without doing that it is very hard to troubleshoot resource
>> issues in production clusters. Other times I may be troubleshooting
>> membership issues and I need detailed corosync messaging. Ideally I would
>> like to be able to set a (potentially) different message level for each of
>> the major components of the HA solution but I see no way to do that
>> currently. Any advice?
>>
>> --
>>
>> Ron Kerry         rkerry at sgi.com
>>
>>
>>
>> _______________________________________________
>> 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://bugs.clusterlabs.org
>
>


-- 

Ron Kerry         rkerry at sgi.com
Global Product Support - SGI Federal