I could use a little help. I know I made this work in a prior version of
Zenoss, but I can't make it work now and can't refer to my old installation to
figure out why.
I've got Zen 2.4.5 on Centos. I've downloaded the Cacti plugin for Asterisk
from ITConnection.ru. It's a python script that creates output in the standard
Cacti format - for example:
Code:
sip.peers.total:13
sip.peers.online:0
iax.peers.total:0
iax.peers.online:0
sip.users.total:13
iax.users.total:0
sip.channels:0
iax.channels:0
pstn.channels:0
zap.channels:4
agents.loggedin.yes:0
agents.loggedin.no:0
voicemail.all:1
core.threads:29
core.uptime:3729863
As far as I understand, Zenoss should be able to take this and create a graph.
But it isn't doing that for me.
I associated a perf template with my Asterisk host and designated the python
script as a data source of type command. When I hit the test button, the
script returns good data.
I then created data points named after the metrics I was interested in (i.e.
core.uptime, sip.peers.online, etc). Each point is of type gauge.
Finally, I defined a graph with some of the datapoints on them.
I've tried setting the parser dropdown to Automatic, Cacti, and Nagios, without
success.
As far as I understand, I shouldn't have to do anything more (like set a create
command for each datapoint). Shouldn't Zen just take the output, parse it, and
graph it?
Instead, I get nothing - the graph I created stays collapsed.
When I turn up debugging for the zencommand daemon, I get the following entries
in the log:
Code:
2009-09-05 22:15:47 DEBUG zen.zencommand: The result of
"/usr/share/cacti/site/scripts/asx-stats.py 0" was "'Asterisk Call
Manager/1.0\nResponse: Success\nMessage: Authentication
accepted\n\nsip.peers.total:13\nsip.peers.online:0\niax.peers.total:0\niax.peers.online:0\nsip.users.total:13\niax.users.total:0\nsip.channels:0\niax.channels:0\npstn.channels:0\nzap.channels:4\nagents.loggedin.yes:0\nagents.loggedin.no:0\nvoicemail.all:1\ncore.threads:29\ncore.uptime:3730343\n'"
2009-09-05 22:15:47 DEBUG zen.zencommand: Dropping useless clear event
{'severity': 0, 'performanceData': '', 'component': '', 'agent': 'zencommand',
'summary': 'Asterisk Call Manager/1.0', 'manager': 'localhost', 'eventKey':
'AsteriskPlugin', 'device': '192.168.1.240', 'message': 'Asterisk Call
Manager/1.0', 'eventClass': '/Perf'}
2009-09-05 22:15:47 DEBUG zen.zencommand: Next command in 59 seconds
I also get the following in event.log:
Code:
2009-09-05T22:15:37 ERROR RenderServer Failed to generate a graph
Traceback (most recent call last):
File "/usr/local/zenoss/zenoss/Products/ZenRRD/RenderServer.py", line 190, in
render
rrdtool.graph(*gopts)
error
------
2009-09-05T22:15:37 WARNING RenderServer
/tmp/renderserver/graph-1df2734ca40c65821861107e51c9578a129600.png
--start=1252073437 --end=1252203037 COMMENT:None\c --imgformat=PNG -F -E
--height=100 --vertical-label= COMMENT:MISSING RRD FILE\: 192.168.1.240
AsteriskPlugin_sip.peers.total COMMENT:MISSING RRD FILE\: 192.168.1.240
AsteriskPlugin_sip.peers.online --width=500
My guess is that something in between the execution of the script and the
graphing is breaking down, but the docs, forums, and my logs don't really point
to a problem or help me to understand what should happen after the command is
executed.
Can anyone shed some light on this? What am I doing wrong?
Thanks.
-------------------- m2f --------------------
Read this topic online here:
http://forums.zenoss.com/viewtopic.php?p=38988#38988
-------------------- m2f --------------------
_______________________________________________
zenoss-users mailing list
[email protected]
http://lists.zenoss.org/mailman/listinfo/zenoss-users
