As far as I know, Zenoss goes by the return code of the script - so 1 or
something instead of 0. This is mostly because not all nagios plugins
return
STATUS something | something=foo ....
--
James Pulver
Information Technology Area Supervisor
LEPP Computer Group
Cornell University
ctso wrote, On 4/8/2009 4:36 PM:
> Hi there,
>
> I've been creating my own Nagios-style COMMAND plugins and they've worked
> great on Zenoss so far - I have always used them in conjunctions with
> thresholds. However, I noticed that Zenoss seems to ignore the status code
> that is given back in the message.
>
>
> Code:
>
> /bin/check_haproxy" was "STATUS CRITICAL | "
> DEBUG:zen.zencommand:Queueing event {'manager': 'localhost', 'eventKey':
> 'check_haproxy', 'device': 'localhost', 'eventClass': '/Cmd/Fail', 'summary':
> 'STATUS CRITICAL', 'component': '', 'agent': 'zencommand', 'severity': 0}
>
>
>
>
> As you can see, the severity is 0. It seems that I only ever get a severity
> > 0 whenever the command actually times out, as opposed to also when a
> critical status code is returned. Is this the expected behavior? If so, I
> would have to modify all of my nagios-style plugins to also return an error
> code, and then build a threshold of that data point (not a big problem, but
> just not what I would have expected the behavior to be).
>
> Thanks
>
>
>
>
> -------------------- m2f --------------------
>
> Read this topic online here:
> http://forums.zenoss.com/viewtopic.php?p=33226#33226
>
> -------------------- m2f --------------------
>
>
>
> _______________________________________________
> zenoss-users mailing list
> [email protected]
> http://lists.zenoss.org/mailman/listinfo/zenoss-users
_______________________________________________
zenoss-users mailing list
[email protected]
http://lists.zenoss.org/mailman/listinfo/zenoss-users
