Hi,
Dejan Muhamedagic wrote:
> This location is not very good. That's where the core files go.
I know... pls ignore it for the moment.
> This looks like an error from the pingd RA. There should've been
> one from the pingd program. You can also try to run this by hand
> and see what it says. Which shell do you have for the hacluster
> user? /bin/false won't do.
I can start it up with the command I found in the logs fine. It's the RA
for sure as you suspected.
I am still not sure whether the web is correct or not. I figured that if
I move the "instance_attributes" from the "primitive" section to the
"clone", it does start at least but stops.
It's the file I used last time:
<clone id="pingd">
<instance_attributes id="pingd-ia">
<attributes>
<nvpair id="pingd-ia-01" name="clone_node_max" value="1"/>
<nvpair id="pingd-ia-02" name="dampen" value="5s"/>
<nvpair id="pingd-ia-03" name="multiplier" value="100"/>
</attributes>
</instance_attributes>
<primitive id="pingd-child" provider="heartbeat" class="ocf" type="pingd">
<operations>
<op id="pingd-child-op-01" name="monitor" interval="20s"
timeout="40s" prereq="nothing"/>
<op id="pingd-child-op-02" name="start" prereq="nothing"/>
</operations>
</primitive>
</clone>
This starts but no pid file gets created, specifying it makes no
difference just as the user. It always starts by root and no pid hence
the resource cannot be stopped I reckon. I don't like it and note I
don't want to start it from ha.cf. Something is really not right here,
none of the examples work from the web and the one I showed above is
very close to the one that Alan R. published in his presentations but
then I have the pid and the stopping issue.
> What does it look like? Should be like this:
> <nvpair id="clone_1-meta-options-globally_unique" name="globally_unique"
> value="false"/>
Not sure, I had to put it back from backups coz' it's kinda semi-prod
system and my HA was quite sick after these failed attempts.
Kind regards,
Ivan
_______________________________________________
Linux-HA mailing list
[email protected]
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems
