I wrote that I am having the same problem a few days ago. Same OS as
well. I can add a little to the discussion, although I have not sorted
ot the problem yet either.
Regarding the zProties zCommandPath, you are correct that the rpm
install places the packages in /opt/zenoss/libexec. I set the value for
zCommandPath to /opt/zenoss/libexec under the zProperties page for the
/Devices root, so that all command templates inherit the value.
The response that I got from chet a few days ago indicated that it
depended on the data points. I've been looking at this document (
http://nagios.sourceforge.net/docs/3_0/pluginapi.html )and have begun to
wonder if the issue is that the plugin I'm monitoring (check_temptraxe)
doesn't output performance data (which would follow a "|" symbol). It
only outputs the text output, which you wouldn't really want to graph.
I'm going to play around with some sed commands to try an modify the
output to follow the Nagios Plugin API for performance data and see what
happens.
Sator81 wrote:
Hi,
I have created a custom RRD template wich runs a Nagios script (check_http) through and
proxy server and one that polls the web page without the proxy server. And after alot of
fiddling the "devices" I create inherit these templates default (I also had to
sacrifice the default template since I had to rename the template to Device wich
overwrote the default one). The graph is intended to graph both results to see if the
proxy server lags at times.
But Zenoss does not create the RRD archives, nor any folder for the new
"devices" for some reason, I have been reading dozens of forum posts and cant
figure out whats wrong. The only error message I have found is from the Render wich
doesnt find the RRD files to generate the graphs.
The commands wich is run (yes I used the wiki on Nagios scripts for Zenoss) is
like this:
/opt/zenoss/libexec/check_http -H x.x.x.x -p 8080 -u http://${devname}
and:
/opt/zenoss/libexec/check_http ${devname}
Both datasrouces has a datapoint called time.
And both crips run as intended with the zenoss user when run manualy or in the
WebUI (with the real URL instead of ${devname}).
One thing I have noted that might be a problem and that is the location of
Zenoss, as my CentOS 4.4 installation is located under /opt/zenoss and not
/usr/local/zenoss.
I have changed the zCommandPath to /opt/zenoss/libexec in Zproperties for both
the template and the devices. I have also restarted the zCommand service and
Zenoss several times while trying to make this work.
Other SNMP based devices works fine.
Any ideas?
Thanks in advance.
------------------------
Kai
-------------------- m2f --------------------
Read this topic online here:
http://community.zenoss.com/forums/viewtopic.php?p=5781#5781
-------------------- 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
