Hello,

I just realized why is this happening. I have a Opt-out cluster.
Switched it to Opt-in cluster and added some rules. Seems to work fine now ... 

Thank you.
--
Maxim Ianoglo

On Wed, 18 May 2011 08:03:08 +0300
Maxim Ianoglo <[email protected]> wrote:

> Hello,
> 
> I have: Node_A, Node_B, Node_C
> Have A MySQL Resource that should run on Node_A ( but with VIP ) and Node_B
> Node_C runs some other resources but this is not important now.
> on-fail monitor option for resource is "standby"
> 
> I have stopped MySQL on Node_A so, MySQL from Node_A and VIP went to Node_B.
> Restarted Heartbeat on Node_A and moved MySQL and VIP back to Node_B and 
> MySQL from Node_A Moved to Node_C even if I have location constraint for this 
> resource that says to keep this service on Node_A.
> 
> Resource definitions:
> group gr_mysql_master_primary VIP_10.10.1.235 mysql_master_primary 
> notif_gr_mysql_master_primary \
>         meta ordered="true" collocated="true" is-managed="true" 
> target-role="Started"
> group gr_mysql_master_secondary mysql_master_secondary 
> notif_gr_mysql_master_secondary \
>         meta ordered="true" collocated="true" is-managed="true" 
> target-role="Started"
> 
> Location Constraints:
> location loc_gr_mysql_master_primary_default gr_mysql_master_primary rule 
> 100: #uname eq Node_B
> location loc_gr_mysql_master_primary_failover gr_mysql_master_primary rule 
> 50: #uname eq Node_A
> location loc_gr_mysql_master_secondary gr_mysql_master_secondary rule inf: 
> #uname eq Node_A
> 
> Is this normal behavior ?
> 
> Thank you.
> --
> Maxim Ianoglo
> 

_______________________________________________
Linux-HA mailing list
[email protected]
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

Reply via email to