selectively disable notifications for services inheriting a specific template?

[Marc Powell]

Please always respond on list so that others may learn from your  
experience. More below...

On Jan 26, 2009, at 7:20 PM, Owen LaGarde wrote:

>> How do you think it would alter the check behavior? Are you sure  
>> about
>> that?
>
> ... by causing a given service check to not fire because it's  
> dependencies
> are not met.  At least that's what it did when I tried it, and my  
> read of
> the docs is that this behavior is precisely what servicedependencies  
> and
> hostdependencies are for -- to alter the firing of checks based on  
> other
> checks' output, not to alter behavior of [non-sub-]processes  
> launched as a
> result of check failure, based on other checks' output.

http://nagios.sourceforge.net/docs/3_0/dependencies.html - See the  
examples and "Execution Dependencies"

You have two configurables. notification_failure_criteria and  
execution_failure_criteria. The first controls when notifications  
should go out and the last when the service checks for the dependent  
service should still be executed. This last can be set to 'n' to  
continue to run dependent service checks regardless of the status of  
the service being depended upon.


>> As far as doubling your config, it certainly doesn't have to if  
>> you're
>> smart about it. I think a servicegroup in the template + 1 dependency
>> might be all you need. I can't say for sure since I don't use
>> dependencies but the it looks workable from the documentation...
>
> Doesn't the documentation indicate that a servicedependency  
> definition must have
> the host and service fields populated (though the -dependent forms  
> can be left
> blank ala' wildcard inheritance)?  Of course I tried this route, I'm  
> not a
> *complete* idiot.  Ok, at least not often.  8-P`   When I left  
> either host or
> service blank nagios complained about an incomplete  
> servicedependency clause.

You may have tried this. My understanding is that you have a single  
service check (call it tgt-service) running on a single host (call it  
tgt-host). A bunch of services on many hosts depend on this being OK.  
You want the notifications to be suppressed if it's not OK but the  
checks to still continue regardless of it's status.

I would approach it as follows. For each service that has this  
dependency, add it to a servicegroup --

define service {
	host_name	remote-host
	service_description	remote-service
	check_command		check_remote ...
	servicegroups		d_service
	...
}

You may have a template that you already use for those services and  
can make the change in only one place. Then, define a service  
dependency that applies to the entire servicegroup --

define servicedependency{
	host_name			tgt-host
	service_description		tgt-service
	dependent_servicegroup_name	d_service
	execution_failure_criteria	n
	notification_failure_criteria	w,u,c,p
	}

Again, I don't use service dependencies but this looks like the way to  
do it.


--
Marc


------------------------------------------------------------------------------
This SF.net email is sponsored by:
SourcForge Community
SourceForge wants to tell your story.
http://p.sf.net/sfu/sf-spreadtheword
_______________________________________________
Nagios-users mailing list
Nagios-users at lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/nagios-users
::: Please include Nagios version, plugin version (-v) and OS when reporting any issue. 
::: Messages without supporting info will risk being sent to /dev/null