[Pacemaker] Question about log level at monitor

[Kazunori INOUE]

2014-02-24 12:00 GMT+09:00 Andrew Beekhof <andrew at beekhof.net>:
>
> On 21 Feb 2014, at 9:35 pm, Kazunori INOUE <kazunori.inoue3 at gmail.com> wrote:
>
>> 2014-02-20 18:59 GMT+09:00 Andrew Beekhof <andrew at beekhof.net>:
>>>
>>> On 20 Feb 2014, at 8:37 pm, Kazunori INOUE <kazunori.inoue3 at gmail.com> wrote:
>>>
>>>> Hi,
>>>>
>>>> Is this by design although log levels differ with a stonith resource
>>>> and other resources in Pacemaker-1.1.11 ?
>>>>
>>>> "P1" is id of ocf:pacemaker:Dummy resource.
>>>> "F1" is id of stonith (ex. stonith:external/ipmi) resource.
>>>>
>>>> * log at "probe"
>>>> crmd[22860]:   notice: process_lrm_event: LRM operation P1_monitor_0
>>>> (call=9, rc=7, cib-update=56, confirmed=true) not running
>>>> crmd[22860]:     info: process_lrm_event: LRM operation F1_monitor_0
>>>> (call=5, rc=7, cib-update=55, confirmed=true) not running
>>>
>>> Looking at the code I can't see how this is possible.
>>> Only if the resource is found running is it set to LOG_INFO
>>>
>>>>
>>>> * log at "monitor NG"
>>>> crmd[22860]:   notice: process_lrm_event: LRM operation
>>>> P1_monitor_30000 (call=12, rc=7, cib-update=64, confirmed=false) not
>>>> running
>>>> crmd[22860]:    error: process_lrm_event: LRM operation
>>>> F1_monitor_30000 (call=13, status=4, cib-update=61, confirmed=false)
>>>> Error
>>>
>>> Thats an easy one... one completed but failed with OCF_NOT_RUNNING, the other didn't complete at all.
>>
>> This log should be LOG_NOTICE?
>> crmd[22860]:     info: process_lrm_event: LRM operation F1_monitor_0
>> (call=5, rc=7, cib-update=55, confirmed=true) not running
>
> Not really.
> There is too much context missing to know if this is worth bringing to anyone's attention.
>
>>
>> and, this log should be LOG_ERROR?
>> crmd[22860]:   notice: process_lrm_event: LRM operation
>> P1_monitor_30000 (call=39, rc=1, cib-update=92, confirmed=false)
>> unknown error
>
> Not really.
> The action completed (PCMK_LRM_OP_DONE) which is all this part of the code cares about.
>
> However, I have now tweaked the logging to have less special cases:
>
>     switch (op->op_status) {
>         case PCMK_LRM_OP_CANCELLED:
>             crm_info("Operation %s: %s (call=%d, confirmed=%s)",
>                      op_key, services_lrm_status_str(op->op_status),
>                      op->call_id, removed ? "true" : "false");
>             break;
>
>         case PCMK_LRM_OP_DONE:
>             crm_notice("Operation %s: %s (call=%d, rc=%d, cib-update=%d, confirmed=%s)",
>                        op_key, services_ocf_exitcode_str(op->rc),
>                        op->call_id, op->rc, update_id, removed ? "true" : "false");
>             break;
>
>         case PCMK_LRM_OP_TIMEOUT:
>             crm_err("Operation %s: %s (call=%d, timeout=%dms)",
>                     op_key, services_lrm_status_str(op->op_status), op->call_id, op->timeout);
>             break;
>
>         default:
>             crm_err("Operation %s (call=%d, status=%d, cib-update=%d, confirmed=%s) %s",
>                     op_key, op->call_id, op->op_status, update_id, removed ? "true" : "false",
>                     services_lrm_status_str(op->op_status));
>     }
>

I confirmed it.
Thank you.

>
>
>
>>
>> That is, should be as follows the log level?
>> crmd[22860]:   notice: process_lrm_event: LRM operation P1_monitor_0
>> (call=9, rc=7, cib-update=56, confirmed=true) not running
>> crmd[22860]:   notice: process_lrm_event: LRM operation F1_monitor_0
>> (call=5, rc=7, cib-update=55, confirmed=true) not running
>>
>> crmd[22860]:   notice: process_lrm_event: LRM operation
>> P1_monitor_30000 (call=12, rc=7, cib-update=64, confirmed=false) not
>> running
>> crmd[22860]:    error: process_lrm_event: LRM operation
>> P1_monitor_30000 (call=39, rc=1, cib-update=92, confirmed=false)
>> unknown error
>> crmd[22860]:    error: process_lrm_event: LRM operation
>> F1_monitor_30000 (call=13, status=4, cib-update=61, confirmed=false)
>> Error
>>
>> crmd[22860]:    error: process_lrm_event: LRM operation
>> P1_monitor_30000 (55) Timed Out (timeout=10000ms)
>> crmd[22860]:    error: process_lrm_event: LRM operation
>> F1_monitor_30000 (72) Timed Out (timeout=10000ms)
>>
>>>>
>>>> Regards,
>>>> Kazunori INOUE
>>>>
>>>> _______________________________________________
>>>> 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
>>>
>>>
>>> _______________________________________________
>>> 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
>>
>> _______________________________________________
>> 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
>
>
> _______________________________________________
> 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
>