[Pacemaker] Cluster Test Suite: Bad News

Andrew Beekhof andrew at beekhof.net
Wed Oct 28 19:25:25 UTC 2009


On Wed, Oct 28, 2009 at 4:33 PM, Schmidt, Torsten
<torsten.schmidt at tecdoc.net> wrote:
> Hi list,
>
> after manually (positive) testing of my new cluster i've initiated one round
> of the cluster test suite with the following command:
>
> ./CTSlab.py --at-boot 1 --nodes 'mysqlha1 mysqlha2' --stack ais
> --syslog-facility local6 --schema pacemaker-1.0 --logfile /var/log/messages
> --stonith 1 --standby 1 --fencing 1 --stonith-type ssh --once 1
>
> it seems there's a problem with openais but i'm not sure about the
> significance of these cts results;
> after cts finished, my crm on both nodes was not reachable (but openais was
> running)
>
> any hints or comments are very much appreciated.
>
> here my CTS:BadNews filtered for one node only:

yeah, um, please dont do that.
makes it near impossible to diagnose the problem.
best just just compress and attach the /var/log/messages from the test
host and tell us which test to look at.

btw. you know that one of the things CTS will do is call kill -9 on
various cluster processes?
we deliberately cause errors to make sure the cluster can handle
them... that may be what you're seeing.




More information about the Pacemaker mailing list