[ClusterLabs] Antw: ambiguity with ocf:pacemaker:HealthSmart

Peter Thurner p.thurner at blunix.org
Fri Apr 17 07:03:28 UTC 2015


On 04/17/2015 08:12 AM, Ulrich Windl wrote:
>>>> Peter Thurner <p.thurner at blunix.org> schrieb am 16.04.2015 um 17:18 in
> Nachricht <552FD2DA.3050502 at blunix.org>:
>> Hi Cluster-Lovers,
>>
>> I'm getting into corosync / pacemaker since a few days, currently
>> running a two-node Setup (Master and Failover). I'm running on ubuntu
>> 14.04 with corosync 2.3.3, pacemaker 1.1.10 and I'm using the latest
>> master branch of https://github.com/ClusterLabs/resource-agents.
>>
>> I'm trying to create the HealthSMART resource via:
>>
>> crm configure primitive Health_Smart_sda ocf:pacemaker:HealthSMART \
>>     meta \
>>         target-role="Started" \
>>     params \
>>         temp_upper_limit="60" temp_lower_limit="30" temp_warning="5"
>> drives="/dev/sda" state="/var/run/HealthSMARTsda.state"
>>
>> crm configure primitive Health_Smart_sdb ocf:pacemaker:HealthSMART \
>>     meta \
>>         target-role="Started" \
>>     params \
>>         temp_upper_limit="60" temp_lower_limit="30" temp_warning="5"
>> drives="/dev/sdb" state="/var/run/HealthSMARTsdb.state"
>>
>>
>> I'm using two resources because using drives="/dev/sda /dev/sdb" gives
>> me this error:
>>
>> /usr/lib/ocf/resource.d/pacemaker/HealthSMART: 173: [: /dev/sda: unexpected
>> operator
>>
>>
>> When I try to commit the two primitives I get the following error:
>>
>> WARNING: Resources Health_Smart_sda,Health_Smart_sdb violate uniqueness for
>> parameter "temp_warning": "5"
>> WARNING: Resources Health_Smart_sda,Health_Smart_sdb violate uniqueness for
>> parameter "temp_lower_limit": "30"
>> WARNING: Resources Health_Smart_sda,Health_Smart_sdb violate uniqueness for
>> parameter "temp_upper_limit": "60"
> 
> This sound to me like a "brain-free programmer" (TM) has written the parameter specification. Try to hack the XML part of the RA and remove the unique requirement where it's abolute nonsense (all the cases you saw).
Thank you for the hint, will do that.

> 
>>
>>
>> My Idea was to create clones of each ressource then, to use for each of
>> the two nodes in my cluster, like:
>>
>> crm configure clone CLO_Health_Smart_sda Health_Smart_sda \
>>     meta \
>>         clone-max="2" clone-node-max="1"
>>
>> crm configure clone Clo_Health_Smart_sdb Health_Smart_sdb \
>>     meta \
>>         clone-max="2" clone-node-max="1"
>>
>>
>> Whats the best practice way to monitor /dev/sda and /dev/sdb via
>> ocf:pacemaker:HealthSMART?
> 
> Other question: Are you sure /dev/sda is consistent after reboot?
This was just for the example, its also possible to use
/dev/disk/by-id/ata-SAMSUNG_MMCRE28G8MXP-0VB_DC0KP01012SE012B724...

> 
>>
>>
>>
>> PS: hope I didnt post twice, just subscirbed and saw that I had to
>> confirm it before posting.
>>
>> _______________________________________________
>> Users mailing list: Users at clusterlabs.org 
>> http://clusterlabs.org/mailman/listinfo/users 
>>
>> Project Home: http://www.clusterlabs.org 
>> Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf 
>> Bugs: http://bugs.clusterlabs.org 
> 
> 
> 
> 
> 
> _______________________________________________
> Users mailing list: Users at clusterlabs.org
> http://clusterlabs.org/mailman/listinfo/users
> 
> Project Home: http://www.clusterlabs.org
> Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf
> Bugs: http://bugs.clusterlabs.org
> 



Für Rückfragen stehe ich Ihnen jederzeit gerne zur Verfügung.

Mit freundlichen Grüßen,
Peter Thurner

--

Blunix.org Professional Linux Service
Linux: Consulting, Management, Schulungen

Web: www.blunix.org
Phone: +49 30 / 301 338 38
Fax: +49 30 / 233 288 28




More information about the Users mailing list