[Pacemaker] pacemaker colocation after one node is down

Andreas Kurz andreas at hastexo.com
Thu May 16 15:46:27 UTC 2013


On 2013-05-16 13:42, Wolfgang Routschka wrote:
> Hi Andreas,
> 
> thank you for your answer.
> 
> solutions is one coloation with -score

ah, yes .... only _one_ of them with a non-negative value is needed.
Scores of all constraints are added up.

Regards,
Andreas

> 
> colocation cl_g_ip-address_not_on_r_postfix -1: g_ip-address r_postfix
> 
> Greetings Wolfgang
> 
> 
> On 2013-05-15 21:30, Wolfgang Routschka wrote:
>> Hi everybody,
>>  
>> one question today about colocation rule on a 2-node cluster on
>> scientific linux 6.4 and pacemaker/cman.
>>  
>> 2-Node Cluster
>>  
>> first node haproxy load balancer proxy service - second node with
>> postfix service.
>>  
>> colocation for running a group called g_ip-address (haproxy lsb-resouce
>> and ipaddress resource) on the other node of the postfix server is
>>  
>> cl_g_ip-address_not_on_r_postfix -inf: g_ip-address r_postfix
> 
> -INF == never-ever ;-)
> 
>>  
>> The problem is now that the node with haproxy is down pacemaker cannot
>> move/migrate the services to the other node -ok second colocation with
>> lower score but it doesn?t works for me
>>  
>> colocation cl_g_ip-address_on_r_postfix -1: g_ip-address r_postfix
>>  
>> Whats my fault in these section?
> 
> Hard to say without seeing the rest of your configuration, but you can
> run "crm_simulate -s -L" to see all the scores taken into account.
> 
> Regards,
> Andreas
> 


-- 
Need help with Pacemaker?
http://www.hastexo.com/now


-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 287 bytes
Desc: OpenPGP digital signature
URL: <https://lists.clusterlabs.org/pipermail/pacemaker/attachments/20130516/0e52f113/attachment-0004.sig>


More information about the Pacemaker mailing list