[ClusterLabs] RHEL 7.4 cluster cannot commit suicide (sbd)
kwenning at redhat.com
Sun Aug 27 06:33:48 EDT 2017
On 08/26/2017 09:53 PM, strahil nikolov wrote:
> Hello everyone,
> as this is my first usage (writing) to mailing list , please excuse me.
> Here is the reason I'm writing to you. I have 3 VM machines (kvm/qemu),
> watchdog of type 'i6300esb' with RHEL 7.4 and iscsi target as a shared
> I have created the 3 node cluster and poison pill (pcs stonith fence
> node_name) works, but I can't make the sbd daemon to self-suicide the
> node once the network is being cut-off (firewall-cmd --panic-on).
> The strange thing is that sbd daemon detects that the storage is off-
> line via (I've stripped out the clutter):
> sbd[pid]: warning: inquisitor_child: Servant <iSCSI Disk> is outdated
> (age: 4)
> sbd[pid]: warning: inquisitor_child: Majority of devices lost -
> surviving on pacemaker
> sbd[pid]: <iSCSI Disk>: error: header_get: Unable to read header
> from device 6
The log says it: surviving on pacemaker
If you have pacemaker-observation activated sbd won't self-fence
as long as it sees a (quorate) cluster.
> The servant keeps restarted but no self-fencing. I thought that the
> issue is in the watchdog , but immediately after killing the sbd main
> pid - the node gets reset (as expected).
> This is the configuration in "/etc/sysconfig/sbd":
> SBD_OPTS="-n harhel1"
> I have used the following example for setting up the sbd: https://acces
> Thank you for reading this long e-mail. I would be grateful if someone
> finds out my mistake.
> Best Regards,
> Strahil Nikolov
> Users mailing list: Users at clusterlabs.org
> Project Home: http://www.clusterlabs.org
> Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf
> Bugs: http://bugs.clusterlabs.org
More information about the Users