No Outbound Check Transfer...

[mroter]

Thanks - the fix is working.
Can you give me an example of the ocsp command using NRPE (e.g. send_nrpe.sh)?

[scottwilkerson]

The master Nagios machine must use OCSP and OCHP in order to send the proper updates to
the central server.

The default nagios.cfg has these settings.

Code: Select all

obsess_over_services=0
#ocsp_command=somecommand
obsess_over_hosts=0
#ochp_command=somecommand

Change the nagios.cfg settings to those below. The timeout prevents your Nagios server from spending too much time
on one command.

Code: Select all

obsess_over_services=1
ocsp_command=service_check
ocsp_timeout=5
obsess_over_hosts=1
ochp_command=host_check
ochp_timeout=5

You will need to create two scripts in the /usr/local/nagios/libexec/eventhandlers directory replacing <SERVER_TO_SEND_TO>.

service_check

Code: Select all

#!/bin/bash
# Service Check
cmd="/usr/local/nrdp/clients/send_nrdp.sh"
host=$1
srv=$2
result=$3
output=$4
/bin/echo -e "$host\t$srv\t$result\t$output\n" | $cmd -u <URL of the NRDP server> -t <shared token>

host_check

Code: Select all

#!/bin/bash
# Host Check
cmd="/usr/local/nrdp/clients/send_nrdp.sh"
host=$1
result=$2
output=$3
/bin/echo -e "$host\t$result\t$output\n" | $cmd -u <URL of the NRDP server> -t <shared token>

Then add 2 commands under Configure -> CCM -> Commands:

Code: Select all

command=service_check
command line*=$USER1$/eventhandlers/service_check $HOSTNAME$ '$SERVICEDESC$' $SERVICESTATEID$ '$SERVICEOUTPUT$'
Command type=misc command

Code: Select all

command=host_check
command line*=$USER1$/eventhandlers/host_check $HOSTNAME$ $HOSTSTATEID$ '$HOSTOUTPUT$'
Command type=misc command

[mroter]

Thank you very much for your great support

Regarding:

Code: Select all

$cmd -H <SERVER_TO_SEND_TO>

I think the use of -H is wrong. it should be:

Code: Select all

$cmd -u <URL of the NRDP server> -t <shared token>

Is the use of pipe (STDIN) better than the use of the explicit parameters -H -s -S -o ?

Do I need to set the "Obsess over host"/"Obsess over service" to "on" on the required hosts/services?

[scottwilkerson]

You are correct. I was modifying this from some old docs I had that use NSCA. I'm going to modify my previous post to reflect the correct setup.

[mroter]

Can you please answer my 2 questions from the last post?

Two other corrections (if I may):
1. The path to the cmd is /clients/ and not /client/
2. The service_check is working perfectly but the host_check returens the following as if the $SERVICESTATEID$ and '$SERVICEOUTPUT$' are undefined

Code: Select all

host_check SNG-VC $SERVICESTATEID$ '$SERVICEOUTPUT$'

[scottwilkerson]

Is the use of pipe (STDIN) better than the use of the explicit parameters -H -s -S -o ?

It doesn't matter either way.

Do I need to set the "Obsess over host"/"Obsess over service" to "on" on the required hosts/services?

It should use the defaults from the nagios.cfg if you haven't specifically set it per host/service, so the answer would be no.

The service_check is working perfectly but the host_check returens the following as if the $SERVICESTATEID$ and '$SERVICEOUTPUT$' are undefined

Boy, I am starting to think I didn't put enough time in that first post, too much copy/pasting...
The host check should be like this

Code: Select all

command=host_check
command line*=$USER1$/eventhandlers/host_check $HOSTNAME$ $HOSTSTATEID$ '$HOSTOUTPUT$'
Command type=misc command

The path to the cmd is /clients/ and not /client/

Thanks again...

[mroter]

Where is the default in nagios.cfg ? I want to obsess only the services with no perfdata otherwise i might get duplicate reporting.
One from perfproc and one from obsess.

[scottwilkerson]

The defaults are being set when you issue

Code: Select all

obsess_over_services=1
obsess_over_hosts=1

So you could set the obsess over service to no in your templates, and then override with yes on the specific services.