[Pacemaker] pcmk 1.1.8, some issues

[David Vossel]

----- Original Message -----
> From: laurent+pacemaker at u-picardie.fr
> To: pacemaker at oss.clusterlabs.org
> Sent: Friday, December 7, 2012 9:48:12 AM
> Subject: [Pacemaker] pcmk 1.1.8, some issues
> 
> 
> Hi,
> 
> I quite like 1.1.8 but I have some issues with it.
> 
> 
> 1) something wrong with stonith-timeout
> 
> stonith-ng[24816]:    error: get_capable_devices: stonith-timeout
> duration 0 is too low, raise the duration to 80 seconds
> 
> with <nvpair id="cib-bootstrap-options-stonith-timeout"
> name="stonith-timeout" value="120s" />

When do you see this?  Is pacemaker fencing a node when this occurs, or are you manually doing it using stonith_admin?

> 
> 2) crm_mon crashing with this message : "Your current configuration
> could only be
> upgraded to <null>... the minimum requirement is
> pacemaker-1.0. .....Connection to the CIB
> terminated Reconnecting...Upgrade failed: Update does not conform to
> the configured schema"

I have seen some odd things with crm_mon in the past month or two.  Some recent work was done to address how crm_mon dynamically updated the latest cib changes.  This may be fix what you are seeing here.

> 
> 3) a lot of "attrd[30769]:  warning: attrd_cib_callback: Update
> shutdown=(null) failed: No such device or address" messages
> 
> or "warning: attrd_cib_callback: Update
> last-failure-vm-elasticsearch-02=(null) failed: No such device or
> address"
> 
> is that harmful ?
> 
> 4) crm_mon -d -p /tmp/t.pid -h /tmp/t.html not creating the pid file
> anymore. (it breaks ocf:pacemaker:ClusterMon)
> 
> 
> Can you reproduce any of these issues ?
> (been able to reproduce 1 and 3 with yesterday's head, i've not paid
> attention to 2 and 4 at that time
> 
> I'd be ok to provide more info and fill tickets if you do.
> 
> --
> Laurent
> 
> _______________________________________________
> Pacemaker mailing list: Pacemaker at oss.clusterlabs.org
> http://oss.clusterlabs.org/mailman/listinfo/pacemaker
> 
> Project Home: http://www.clusterlabs.org
> Getting started:
> http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf
> Bugs: http://bugs.clusterlabs.org
>