Page 1 of 2
check_xi_service_snmp_linux_load issue
Posted: Sun Aug 04, 2013 2:09 pm
by ahmad.zuhd
I have an issue when using the check_xi_service_snmp_linux_load check as below:
Code: Select all
Command view
$USER1$/check_snmp_load_wizard.pl -H $HOSTADDRESS$ $ARG1$
Code: Select all
$ARG$ = -C public --v2c -w 80 -c 90 -T netsc -f
according to the help of the command check_snmp_load_wizard.pl,
netsc : cpu usage given by net-snmp (100-idle)
the problem is that a critical alarm is sent many times because the result is 100.
Is there any fix to the command or workaround?
Re: check_xi_service_snmp_linux_load issue
Posted: Mon Aug 05, 2013 9:55 am
by scottwilkerson
Can you try
Code: Select all
$ARG1$ = -C public --v2c -w 80 -c 90 -T stand -f
Re: check_xi_service_snmp_linux_load issue
Posted: Mon Aug 05, 2013 10:00 am
by abrist
Is the system in question running at 100% load most of the time?
Re: check_xi_service_snmp_linux_load issue
Posted: Mon Aug 05, 2013 11:10 am
by ahmad.zuhd
No its idle. As I mentioned previously this is the normal plugin output when the system is idle
Re: check_xi_service_snmp_linux_load issue
Posted: Mon Aug 05, 2013 11:17 am
by abrist
What was the output of Scott's command?
Code: Select all
$ARG1$ = -C public --v2c -w 80 -c 90 -T stand -f
Re: check_xi_service_snmp_linux_load issue
Posted: Mon Aug 05, 2013 11:23 am
by ahmad.zuhd
It returns error of not available object. If you need the exact output I can connect to site later today and get the output.
This option is not computable with the server I'm monitoring
Re: check_xi_service_snmp_linux_load issue
Posted: Mon Aug 05, 2013 11:43 am
by scottwilkerson
Can you send the full output you are getting with your original command
Re: check_xi_service_snmp_linux_load issue
Posted: Mon Aug 05, 2013 11:44 am
by lmiltchev
Yes, show the exact command from the command line, along with the output of it (hide sensitive info). For example, you may have something like this:
Code: Select all
/usr/local/nagios/libexec/check_snmp_load_wizard.pl -H 192.168.x.x -C <community> --v2c -w 80 -c 90 -T stand -f
1 CPU, load 12.0% < 80% : OK | cpu_prct_used=12%;80;90
Re: check_xi_service_snmp_linux_load issue
Posted: Mon Aug 05, 2013 12:10 pm
by ahmad.zuhd
Code: Select all
[root@nms libexec]# ./check_snmp_load_wizard.pl -H <ip-address> -C public --v2c -w 80 -c 90 -T netsc -f
CPU used 3.0% (<80) : OK | cpu_prct_used=3%;80;90
[root@nms libexec]#
[root@nms libexec]# ./check_snmp_load_wizard.pl -H <ip-address> -C public --v2c -w 80 -c 90 -T stand -f
Can't find CPU usage information : UNKNOWN
[root@nms libexec]#
Re: check_xi_service_snmp_linux_load issue
Posted: Mon Aug 05, 2013 1:08 pm
by lmiltchev
It seems the check below works fine in your environment:
Code: Select all
[root@nms libexec]# ./check_snmp_load_wizard.pl -H <ip-address> -C public --v2c -w 80 -c 90 -T netsc -f
CPU used 3.0% (<80) : OK | cpu_prct_used=3%;80;90
The load is 3.0... how often does it get 100?
the problem is that a critical alarm is sent many times because the result is 100.