We like to monitor several cluster resources. A resource can be active on one of the clusternodes and the resource itself has it's own ip-address and a san-disk.
At this moment we use one big nrpe.cfg file that has to be the same on all clusternodes.
Our goal is to have a nrpe.cfg file on a clusternode only containing the checks needed for the server itself. Within the file we would like to refer to include files. When the nrpe daemon is starting it has to load the local checks and (if the resource disk is available) the extra checks for the resource.
/etc/nagios/nrpe.cfg contains the lines:
# System checks
command[check_swap]=/usr/lib/nagios/plugins/check_swap -w 5% -c 10%
# Application checks
/apps/ict/nrpe.cfg contains the line:
command[check_disk]=/usr/lib/nagios/plugins/check_disk -w 10% -c 5% /apps/ict
Unfortunately this concept doesn't work as expected. I also tried include_dir=/apps/ict.
Only the local checks work, the checks in the include file result in "NRPE: Command 'check_disk not defined ".
Where did I go wrong ?