[Pacemaker] failed actions after resource creation

andreas graeper agraeper at googlemail.com
Thu Jun 6 13:49:43 UTC 2013


hi,
i started
 lsb:nfs-common
 lsb:nfs-kernel-server
 ocf:heartbeat:exportfs

everything running on Node0 (drbd:Master, clusterip,filesystem,..)
nothing running on Node1

crm_mon shows:
 Failed actions:
    p_nfscommon_monitor_0 (node=linag, call=189, rc=5, status=complete):
not installed


'linag' is Node1 (drbd:Slave)

that cannot cleaned up :
 crm_resource -P
 crm_resource --cleanup --resource p_nfscommon

what can i do ?
thanks in advace
andreas


2013/6/6 andreas graeper <agraeper at googlemail.com>

> hi,
> in examples using crmsh to create a resource with constraints is an
> interactive mode is used
>
> crm configure edit
>   > primitive   B
>   > order        B_after_A
>   > colocation B_on_A
>   > commit
>   > end
>  quit
>
> when resource B depends on another resource A that is running once on
> Node0
> (or as master/slave on both nodes, as master on Node0, slave on Node1)
> and i call
> `crm configure primitive B ...`
>
> will in first attempt (not necessarily) the new resource A started on
> Node1 till i define the constraints
> `crm configure colocation c_B_on_A inf: B A`
> ?
>
> and is this the reason for a 'failed action' on Node1
> that can get cleaned up with `crm_resource -P` for instance ?
>
> how can i achieve such a session-like creation of a resource with
> constraints
> from out of a script ? ( i have prepared for each resource a s_xxx.sh and
> k_xxx.sh (start/kill)
> to add and remove .. )
>
> thanks in advance
> andreas
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.clusterlabs.org/pipermail/pacemaker/attachments/20130606/b8f7ae0e/attachment.htm>


More information about the Pacemaker mailing list