[Linux-cluster] cluster not fencing after filesystem failure
Robert Jacobson
Robert.C.Jacobson at nasa.gov
Wed Apr 29 16:49:07 UTC 2015
Actually it does look like the failed node was fenced by the other node:
Apr 25 02:51:46 fenced fencing node sdo-dds-nfsnode2.dds.sdo
Apr 25 02:52:39 fenced fence sdo-dds-nfsnode2.dds.sdo success
However, even after fencing, the working node (sdo-dds-nfsnode1) did not
resume the HA_nfs service. The service was in a failed state:
[root at sdo-dds-nfsnode1 log]# clustat
Cluster Status for ddsnfs @ Sat Apr 25 03:45:11 2015
Member Status: Quorate
Member Name ID Status
------ ---- ---- ------
sdo-dds-nfsnode1.dds.sdo 1
Online, Local, rgmanager
sdo-dds-nfsnode2.dds.sdo 2
Online, rgmanager
Service Name Owner
(Last) State
------- ---- -----
------ -----
service:HA_nfs
(sdo-dds-nfsnode2.dds.sdo) failed
On 2015-04-29 11:31 AM, Vasil Valchev wrote:
> Hi,
>
> You can check in the log for "fenced" messages, if it tries to fence
> the node at all. Also for "cman".
> Is your cluster hanged after a node failure? That would indicate the
> fencing didn't succeed for some reason.
>
> >I've tested fencing from the command line and it works:
> >fence_vmware_soap --ip 192.168.50.9 --username ddsfence --password
> >secret -z --action reboot -U "423d288c-03ff-74bf-9a4f-bf661f8ed87b"
>
>
> You can also test fencing with "fence_node <node-to-be-fenced>" - that
> way it is tested with the exact arguments from the cluster.conf and
> you can see if it works or not.
>
> BR,
> Vasil
>
>
>
--
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Robert Jacobson Robert.C.Jacobson at nasa.gov
Lead System Admin Solar Dynamics Observatory (SDO)
Bldg 14, E222 (301) 286-1591
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listman.redhat.com/archives/linux-cluster/attachments/20150429/7d58792e/attachment.htm>
More information about the Linux-cluster
mailing list