On 8/10/07, Rob Aronson <[EMAIL PROTECTED]> wrote:
> I have SLES 10 with the following versions fo heartbeat
> heartbeat-stonith-2.0.7-1.2
> heartbeat-cmpi-2.0.7-1.2
> yast2-heartbeat-2.13.10-1.3
> heartbeat-2.0.7-1.2
> heartbeat-ldirectord-2.0.7-1.2
> heartbeat-pils-2.0.7-1.2

do please upgrade to SP1
(SUSE support will send you packing until you do)

>  The line you couldn't parse correctly was meant to convey:
> I was trying to emulate loss of network connectivity in a cluster node by
> shutting down the primary network interface.

to the outside world or to the other machine?

> In my two node cluster with
> resource groups assigned to each node I had thought the correct behavior of
> the cluster would be upon loss of connectivity in a node, the cluster would
> remove the resources from the node with out connectivity and then the
> stonith daemon would take over and restart the disconnected node to make
> sure it couldn't have a split brain condition. This is what happened when I
> did the test the first time. The second time my 2 stonith resources were
> stopped and the node without network access was left running.

In a yet-to-be-released version, we ensure that all stonith-related
operations are complete before stopping stonith resources.  This is
probably what you're seeing.

> This is a sample of the log files off of both cluster nodes

please make sure logs are sent as file attachments - most mail clients
mangle them beyond recognition
_______________________________________________
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