[Linux-HA] Group resource does not become Master

Sun Nov 18 03:20:43 MST 2007

On Nov 18, 2007, at 12:15 AM, HIDEO YAMAUCHI wrote:

> Hi Andrew,
>
> I confirmed a revision in a version of CRM-Devel-cc23836a1fb3 today.
> (The problem of yesterday's ccm did not occur today.)
>
> However, the same problem is not broken off.
> The result was the same as the previous.

You're right.

I didn't actually check the transition graph to see that it was doing  
the right thing... I just looked at the logs and assumed the rest was  
correct which was lazy of me :-(

The PE tried to do the right thing:
pengine[11338]: 2007/11/18_08:13:06 info: master_promotion_order:  
Merging weights for ms-sf
pengine[11338]: 2007/11/18_08:13:06 info: master_color: Promoting  
grp_ms_sf:1
pengine[11338]: 2007/11/18_08:13:06 info: master_color: ms-sf:  
Promoted 1 instances of a possible 1 to master

But next_role wasn't being set recursively so nothing actually happened.

This time I have checked the output is correct and the resource is  
actually promoted.
Fixed:
    http://hg.beekhof.net/lha/crm-dev/rev/bc9c2bf5a8b7

>
>
> ============
> Last updated: Sun Nov 18 08:13:29 2007
> Current DC: rh44-1 (479ab737-3eb1-4f0a-bab6-b3e3f33f774a)
> 2 Nodes configured.
> 2 Resources configured.
> ============
>
> Node: rh44-1 (479ab737-3eb1-4f0a-bab6-b3e3f33f774a): online
> Node: rh44-2 (e7025b07-d393-40d0-a237-d33d6eda4d8c): online
>
> Resource Group: test
>    resource_1  (heartbeat::ocf:IPaddr):        Started rh44-1
> Master/Slave Set: ms-sf
>    Resource Group: grp_ms_sf:0
>        master_slave_Stateful:0 (heartbeat::ocf:Stateful):       
> Started rh44-2
>    Resource Group: grp_ms_sf:1
>        master_slave_Stateful:1 (heartbeat::ocf:Stateful):       
> Started rh44-1
>
> Regards,
> Hideo Yamauchi.
>
> --- HIDEO YAMAUCHI <renayama19661014 at ybb.ne.jp> wrote:
>
>> Hi Andres,
>>
>> I understood it.
>> I will confirm it again tomorrow.
>>
>> Regards,
>> Hideo Yamauchi.
>>
>> --- Andrew Beekhof <beekhof at gmail.com> wrote:
>>
>>>
>>> On Nov 17, 2007, at 5:04 AM, HIDEO YAMAUCHI wrote:
>>>
>>>> Hi Andrew,
>>>>
>>>> I confirmed a revision in a version of CRM-Devel-c7f778b5c26e.
>>>>
>>>> However, the same problem is not broken off.
>>>> The result was the same as the previous.
>>>
>>> there were two commits, you need the second one too
>>>    http://hg.beekhof.net/lha/crm-dev/rev/cc23836a1fb3
>>>
>>>>
>>>>
>>>> ============
>>>> Last updated: Sat Nov 17 12:59:09 2007
>>>> Current DC: rh44-1 (304025f9-0055-4811-964b-00620e2cc094)
>>>> 2 Nodes configured.
>>>> 2 Resources configured.
>>>> ============
>>>>
>>>> Node: rh44-1 (304025f9-0055-4811-964b-00620e2cc094): online
>>>> Node: rh44-2 (2477f736-e595-46b7-96c4-5c50df5a97bd): online
>>>>
>>>> Resource Group: test
>>>>   resource_1  (heartbeat::ocf:IPaddr):        Started rh44-1
>>>> Master/Slave Set: ms-sf
>>>>   Resource Group: grp_ms_sf:0
>>>>       master_slave_Stateful:0 (heartbeat::ocf:Stateful):
>>>> Started rh44-2
>>>>   Resource Group: grp_ms_sf:1
>>>>       master_slave_Stateful:1 (heartbeat::ocf:Stateful):
>>>> Started rh44-1
>>>>
>>>>
>>>> And, in the version of CRM-Devel-cc23836a1fb3, an error occurred.
>>>> ===================================================================
>>>> cib[21052]: 2007/11/17_12:48:26 WARN: ccm_connect: CCM Activation
>>>> failed
>>>> cib[21052]: 2007/11/17_12:48:26 WARN: ccm_connect: CCM Connection
>>>> failed 1 times (30 max)
>>>> ccm[21051]: 2007/11/17_12:48:28 info: Break tie for 2 nodes cluster
>>>> ccm[21051]: 2007/11/17_12:48:28 info: G_main_add_SignalHandler:
>>>> Added signal handler for signal 15
>>>> ccm[21051]: 2007/11/17_12:48:28 ERROR: socket_wait_conn_new: unlink
>>>> failure(/var/run/heartbeat/ccm/ccm): Permission denied
>>>> ccm[21051]: 2007/11/17_12:48:28 ERROR: socket_wait_conn_new: trying
>>>> to create in
>>>> /var/run/heartbeat/ccm/ccm bind:: Permission denied
>>>> ccm[21051]: 2007/11/17_12:48:28 ERROR: Can't create wait channel:
>>>> Permission denied
>>>> heartbeat[21031]: 2007/11/17_12:48:28 WARN: Managed /usr/lib/
>>>> heartbeat/ccm process 21051 exited with
>>>> return code 1.
>>>> heartbeat[21031]: 2007/11/17_12:48:28 EMERG: Rebooting system.
>>>> Reason: /usr/lib/heartbeat/ccm
>>>> cib[21052]: 2007/11/17_12:48:29 info: ccm_connect: Registering with
>>>> CCM...
>>>> cib[21052]: 2007/11/17_12:48:29 WARN: ccm_connect: CCM Activation
>>>> failed
>>>> cib[21052]: 2007/11/17_12:48:29 WARN: ccm_connect: CCM Connection
>>>> failed 2 times (30 max)
>>>> = 
>>>> ===================================================================
>>>
>>> you'll have to discuss this with alan and whoever is maintaining the
>>> ccm these days.
>>>
>>>>
>>>>
>>>>
>>>> I attached the log of the version of CRM-Devel-c7f778b5c26e.
>>>>
>>>> Regards,
>>>> Hideo Yamauchi.
>>>> <report-crm-devel-
>>>> c7f778b5c26e.tar.gz>_______________________________________________
>>>> Linux-HA mailing list
>>>> Linux-HA at lists.linux-ha.org
>>>> http://lists.linux-ha.org/mailman/listinfo/linux-ha
>>>> See also: http://linux-ha.org/ReportingProblems
>>>
>>> _______________________________________________
>>> Linux-HA mailing list
>>> Linux-HA at lists.linux-ha.org
>>> http://lists.linux-ha.org/mailman/listinfo/linux-ha
>>> See also: http://linux-ha.org/ReportingProblems
>>>
>>
>> _______________________________________________
>> Linux-HA mailing list
>> Linux-HA at lists.linux-ha.org
>> http://lists.linux-ha.org/mailman/listinfo/linux-ha
>> See also: http://linux-ha.org/ReportingProblems
>>
> <report-crm- 
> devel.tar.gz>_______________________________________________
> Linux-HA mailing list
> Linux-HA at lists.linux-ha.org
> http://lists.linux-ha.org/mailman/listinfo/linux-ha
> See also: http://linux-ha.org/ReportingProblems