On Sep 26, 2008, at 10:04 AM, Ioannis Feneris wrote:
Hi all,
Suppose i have 2 nodes running heartbeat v2, and 1 resource. Lets
say, for some reason, the resource that runs in node A fails. Then
resource fails-over at node B. Then, the resource fails again. In
ths situation, the resource can't failback to node A. It only fails-
back when i cleanup the resource from the LRM ( crm_resource -C -r
myresource -H nodeA )
crm_verify -LV says:
crm_verify[..]...: WARN: unpack_rsc_op: Processing failed op
(myservice_start_0) for myservice on nodeA
crm_verify[..]...: WARN: unpack_rsc_op: Handling failed start for
myservice on nodeA
crm_verify[..]...: WARN: unpack_rsc_op: Processing failed op
(myservice_start_0) for myservice on nodeB
crm_verify[..]...: WARN: unpack_rsc_op: Handling failed start for
myservice on nodeB
I found in http://www.linux-ha.org this about auto_failback:
*NOTE*: auto_failback does not have any effect on a Release 2 CRM-
style cluster (one configure
with crm_on). For CRM-style clusters, this has been replaced with
the <http://www.linux-ha.org/v2/dtd1.0/annotated#default_resource_stickiness
>default_resource_stickiness attribute in the CIB.
How is default_resource_stickiness attribute used and work?
http://clusterlabs.org/mw/Image:Configuration_Explained.pdf
Is there any way to make HA to auto cleanup a resource of the LRM?
In pacemaker 0.7 you can now time-out failed actions
Thanks!
Ioannis.
_______________________________________________
Linux-HA mailing list
[email protected]
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems
_______________________________________________
Linux-HA mailing list
[email protected]
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems
