Page 2 of 2

Re: How to increase NRPE socket timeout??

Posted: Tue May 05, 2015 4:04 pm
by abrist
Just to recap. You need to do 3 things:
1. Verify that service_check_timeout in the nagios.cfg on your nagios server is set high enough:
grep service_check_timeout /usr/local/nagios/etc/nagios.cfg
NOTE: Remember to restart nagios after any changes to nagios.cfg.
2. Add "-t" to the check_nrpe command on the nagios server:
command_line $USER1$/check_nrpe -H $HOSTADDRESS$ -t 30 -c $ARG1$
3. Verify that the command_timeout in the nrpe.cfg file on the remote host is configured to be long enough:

Code: Select all

grep command_timeout /usr/local/nagios/etc/nrpe.cfg
NOTE: Remember to restart the nrpe daemon on the remote host after changing any nrpe.cfg directive values.

Does the command run correctly with -t 30 from the cli?

Code: Select all

./check_nrpe -H <host> -t 30 -c check_cassandra

Re: How to increase NRPE socket timeout??

Posted: Tue May 05, 2015 4:30 pm
by xpac
Here's a snapshot:

Re: How to increase NRPE socket timeout??

Posted: Tue May 05, 2015 4:37 pm
by abrist
xpac wrote:Here's a snapshot:
Looks ok.

Does the command run correctly with -t 30 from the cli?

Code: Select all

 ./check_nrpe -H <host> -t 30 -c check_cassandra

Re: How to increase NRPE socket timeout??

Posted: Tue May 05, 2015 4:41 pm
by xpac
abrist wrote:Just to recap. You need to do 3 things:
1. Verify that service_check_timeout in the nagios.cfg on your nagios server is set high enough:
grep service_check_timeout /usr/local/nagios/etc/nagios.cfg
NOTE: Remember to restart nagios after any changes to nagios.cfg.
2. Add "-t" to the check_nrpe command on the nagios server:
command_line $USER1$/check_nrpe -H $HOSTADDRESS$ -t 30 -c $ARG1$
3. Verify that the command_timeout in the nrpe.cfg file on the remote host is configured to be long enough:

Code: Select all

grep command_timeout /usr/local/nagios/etc/nrpe.cfg
NOTE: Remember to restart the nrpe daemon on the remote host after changing any nrpe.cfg directive values.

Does the command run correctly with -t 30 from the cli?

Code: Select all

./check_nrpe -H <host> -t 30 -c check_cassandra
1, 2, and 3 are all correct.

If I run the command with -t 30 from the cli, it does execute properly

Re: How to increase NRPE socket timeout??

Posted: Tue May 05, 2015 4:46 pm
by jdalrymple
Your Nagios is straight up misbehaving.

Next thing that I personally would do is:

nagios.cfg:

Code: Select all

debug_level=-1
debug_verbosity=2
max_debug_file_size=20971520
Make sure you have 20MB of free disk space, restart Nagios then dig through debug to see why it's misbehaving.