escalations and NRPE

[Carroll, Jim P [Contractor]]

I've discovered (and I should have realized this at the time of
implementation) that when a host running with the NRPE daemon is rebooted
(and NRPE isn't configured to start automatically on boot), Nagios will
squawk about all the NRPE checks configured for that host.  Which makes
sense, of course.

But to get notified for each/every NRPE check when it's simply the NRPE
daemon which is unavailable?  Hhhmmm.  Methinks there's a better approach.

I was thinking of configuring dependancies, such that if (let's say) a
check_tcp on port 5666 for that host fails, don't bother me with
notifications for the unavailiability of the N NRPE checks which fail
because they can't connect to the daemon.

I've read the dependancies documentation, and feel a wee bit that I'm
lacking clarity, and so wouldn't mind if someone could give me a simple
example of how to configure it, assuming Service B (and C and D) depends on
Service A.  In particular, I'm unclear on the function and benefit of
defining execution_failure_criteria.

Before someone suggests "why don't you just run it from inetd or install a
startup script", some background:

- we don't have root on all hosts
- we want consistency as much as possible across all hosts

For the hosts where we have root, yes, I will most likely install a startup
script.  But for the hosts where we don't have root, I'd like to simply be
notified that NRPE is down, and not separate pages for each partition,
memory, cpu and logfile check which fails.

Feel free to share your $0.02.

jc


-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf