This article describe howto monitor an Asterisk server with Nagios or op5 Monitor. Pre-req to get it running is a working Nagios or op5 Monitor installation and an Asterisk 🙂
Theory
In my implementation of Asterisk I have a couple of softphones, 2 hardphones and connections to two SIP providers. I want to monitor the following:
- Possibility for a phone to be able to register at the Asterisk server
- The registration at the SIP providers are OK
- The Operating system is not overloaded
- The server where Asterisk is running is up
Implementation
Sip check
I found the following after Googling: http://bashton.com/osprojects/nagiosplugins/ Define it in commands.cfg
# command 'check_sip'
define command{
command_name check_sip
command_line $USER1$/custom/nagios-check_sip-1.2/check_sip -u "$ARG1$"
}
Define the sip check in services.cfg, I also created a service group called ip_telephony
# service 'Asterisk Check SIP'
define service{
use default-service
host_name dull
service_description Asterisk Check SIP
check_command check_sip!sip:XXXXX@dull.mynet
servicegroups ip_telephony
contact_groups it-slav_msn,it-slav_mail,call_it-slav
}
Monitor the Peers
First I tried to find a suitable plugin at Nagiosexchange, Google and other sites if there was anybody that has created and published a Nagios plugin to monitor the sip peers. I could not find any. With the asterisk command "sip show peers", information about the connected sip peers can be found:
[root@dull custom]# asterisk -rx "sip show peers" Name/username Host Dyn Nat ACL Port Status pulver 69.90.155.70 5060 OK (154 ms) digisip/XXXXX 82.209.165.194 5060 OK (44 ms) 6016 (Unspecified) D 0 UNKNOWN 6005 (Unspecified) D 0 UNKNOWN 6004/6004 10.1.1.168 D 5060 OK (139 ms) 6003/6003 10.1.1.168 D 5060 OK (136 ms) 6002/6002 10.1.1.152 D 5060 OK (8 ms) 6011 (Unspecified) D 0 UNKNOWN 6010 (Unspecified) D 0 UNKNOWN 6000 (Unspecified) D 0 UNKNOWN 6001 (Unspecified) D 0 UNKNOWN 11 sip peers [Monitored: 5 online, 6 offline Unmonitored: 0 online, 0 offline]
I discussed this issue with Marcus Rejås when we meet and he told me that he has written a Nagios plugin to monitor the sip peers. I got the script and modified to fit my needs, i.e. get performance data for graphing: /opt/plugins/custom/check_asterisk_sip_peers.sh
#!/bin/bash
#
# Simple Asterisk Peer Check.
# Copyright (C) 2008 Marcus Rejås / Rejås Datakonsult
#
# Modified with perfdata by Peter Andersson
# http://www.it-slav.net/blogs/?p=123
# peter@it-slav.net
#
# This program is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
#
# Very simple plugin that checks if a peer is ok. The peers needs "qualify=yes"
# in its configuration.
#
# A peer that is not registered or non-existent will result in error. If the
# peer is OK a short statusline (from Asterisk) is written. There is timing
# information suitable for graphing as well.
#
# You should have received a copy of the GNU General Public License
# along with this program. If not, see .
#
# Example use of this script:
#
# sip:~# ./sip_check_peer mysecretary-100
# mysecretary-100/461762501 62.80.200.53 5060 OK (10 ms)
# sip:~#
#
#
if [ $# == 0 -o "$1" == "-h" -o $# -gt 1 ]; then
echo "Usage: $0"
exit 3
fi
LINE=`asterisk -r -x "sip show peers" | grep $1 | grep "OK ("`
#
# This is a uggly. Just to check that the expression above does not match more
# then one line.
#
HITS=`asterisk -r -x "sip show peers" | grep $1 | grep "OK (" | wc -l`
if [ $HITS -gt 1 ]; then
echo "ERROR: Multiple match, tweak your arguments or fix $0 :-)"
exit 3
fi
if [ "$LINE" ]; then
echo -n "OK: "
echo -n $LINE
#Create perdata
echo -n "|time="
echo $LINE | awk '{gsub(/\(/,"")};{gsub(/\)/,"")};{print $(NF-1)$NF}'
exit 0
elif [ -z "$LINE" ]; then
echo "CRITICAL: Something is wrong with $1";
exit 2
else
echo $LINE
exit 2
fi
The command run by hand looks like this:
[root@dull /]# /opt/plugins/custom/check_asterisk_sip_peers.sh pulver OK: pulver 69.90.155.70 5060 OK (166 ms)|time=166ms
The plugin is started by nrpe at the asterisk server and configured in /etc/nrpe.d/mycommands.cfg
command[asterisk_peer_digisip]=sudo /opt/plugins/custom/check_asterisk_sip_peers.sh digisip command[asterisk_peer_pulver]=sudo /opt/plugins/custom/check_asterisk_sip_peers.sh pulver command[asterisk_peer_6002]=sudo /opt/plugins/custom/check_asterisk_sip_peers.sh 6002 command[asterisk_peer_6003]=sudo /opt/plugins/custom/check_asterisk_sip_peers.sh 6003 command[asterisk_peer_6004]=sudo /opt/plugins/custom/check_asterisk_sip_peers.sh 6004
/opt/plugins/custom/check_asterisk_sip_peers.sh must run as a high privileged user so I’m using sudo, modify /etc/sudoers with visudo:
# visudo -f /etc/sudoers
--snip-- nobody ALL= (root) NOPASSWD: /opt/plugins/custom/check_asterisk_sip_peers.sh --snip-- Turn of requiretty, because it will run without a console --snip-- #Defaults requiretty --snip--
The service checks are defined in services.cfg and also put into the same servicegroup
# service 'Asterisk Peer Pulver'
define service{
use default-service
host_name dull
service_description Asterisk Peer Pulver
check_command check_nrpe!asterisk_peer_pulver
servicegroups ip_telephony
contact_groups it-slav_sms,it-slav_mail,call_it-slav
}
I also defined the other peers, i.e. digisip, 6002, 6003, 6004
The servicegroup is defined in servicegroups.cfg
# servicegroup 'ip_telephony'
define servicegroup{
servicegroup_name ip_telephony
alias IP - Telephony
}
This is the result, a screenshot from my op5 Monitor system:
I leave to the reader to define the standard checks for load, ping and total processes. If you use op5 Monitor or pnp4nagios the graphs are automatically created: 
Links
12 Responses to “Asterisk monitoring with Nagios or op5 Monitor”
Leave a Reply
You must be logged in to post a comment.
March 13th, 2009 at 3:25 pm
Hello.
Which nagios version do you use?
This skin looks great, is it in standard nagios package?
March 13th, 2009 at 3:44 pm
It is op5 Monitor, based on Nagios 3.x.
op5 has changed the look and feel, what we are doing right now is developing a total new Gui and that will be GPL:ed.
http://www.op5.org/community/projects/ninja
If you want to try op5 Monitor, there is both a downloadable vmware image and a package for RHEL5 or CentOS5.
https://shop.op5.com/free-trial-versions-c-13.html
Regards
Peter
June 5th, 2009 at 10:20 am
Did everything listed here including the visudo but i keep getting this error:
NRPE: Unable to read output
I am calling the check from a Nagios Monitoring server and i get that error, when i run the check on commandline on the monitored box, i get the output fine, any pointers, been scratching my head for a week now.
August 4th, 2009 at 5:32 pm
I changed to “asterisk” the user and group used to start nrpe daemon in my init.d file.
Now I am able to receive info from asterisk:
./check_nrpe -H 10.X.X.X -p 5666 -c check_asterisk_sip_peer -a OVH
OK: OVH/003——– 91.121.129.17 N 5060 OK (21 ms)|time=21ms
Reminder: to send argument to nrpe, you need to compile it with: ./configure –enable-command-args
and enable i(n the nrpe.cfg file) this argument:
dont_blame_nrpe=1
June 11th, 2010 at 2:25 pm
Any ideas on the newest Network Management system freeware?
I searched the web and discovered the following:
Kaseya.com
Logmein.com
They all look different… Does anyone has experience with any of them?
In addition did anybody else try that software:
N-able IT automation software ?
June 11th, 2010 at 2:41 pm
I’m totally amazed of all these different softwares that try to solve the same issues.
Is there room for all of them on the market, my guess is NO!
No I have not tried them.
June 15th, 2010 at 7:38 am
Hi Peter, great site! and great article.
However I have a slight problem in understanding something, to which I am sure I’m not alone.
Basically the above tasks are easy to implement, such as updating the commands.cfg (which on Debian BTW, is linux-commands.cfg), and linux-services.cfg.
The problem I have is that this script is looking for the actual plugin and although I’ve downloaded it (Nagios check_sip plugin v1.2), I don’t know where to install it.
I’ve tried putting it in /usr/src/nagios-plugins-1.4.13/plugins
(this is because the plugin file (Nagios check_sip plugin v1.2) contains a simple script.
Is it possible for you to add the steps to actually install the plugin please? or am I totally on the wrong track altogether?
I’m sure that once that is shown, it would help many other people including me 😉
June 17th, 2010 at 1:46 pm
In op5 Monitor the macro $USER1$ is equal to /opt/plugins/ and we recommend our customers to put their homebrown or downloaded scripts into /opt/plugins/custom
You can put it wherever you want just update the commands definition to suit your needs.
December 17th, 2010 at 10:12 am
got the NRPE: Unable to read output error
nagios ALL=(ALL) NOPASSWD: /usr/lib/nagios/plugins/sip_check_peers
added that line to sudoers and got it working
December 22nd, 2010 at 5:31 am
#HITS=`asterisk -r -x “sip show peers like phone*” | grep “$1.*OK (” | wc -l`
HITS=`echo $LINE|wc -l`
Well,asterisk -rx uses to much time,
this can reduce half time ^_^
December 29th, 2010 at 8:51 am
Yes, you are right!