Q: time-saving tricks for Service Dependencies ?

Carroll, Jim P [Contractor] jcarro10 at sprintspectrum.com
Mon Feb 10 23:22:03 CET 2003


To answer your question first:  Yes, as far as I know.

To respond to your statement:  Yes, that could perhaps spark a philosophical
debate.  But perhaps I wasn't clear enough.

My goal:  To be able to create a servicedependency whereby I can merely add
a host (in the case of host_name / dependent_host_name) or a hostgroup (in
the case of hostgroup_name / dependent_hostgroup_name) such that the trivial
case of a service on a host is dependent on another service *on that same
host* (and not on any other host).

Allow me to attempt to illustrate by example.

define servicedependency{
		host_name				fizzgig,wingnut
		service_description		NRPE status
		dependent_host_name		fizzgig,wingnut
		dependent_service_description	Check swap
		other dependency directives ...
		}

I want 'Check swap' on fizzgig to be dependent on 'NRPE status', also on
fizzgig.  Similarly, I want 'Check swap' on wingnut to be dependent on 'NRPE
status', also on wingnut.  But I do *not* want 'Check swap' on fizzgig to be
dependent on 'NRPE status' on wingnut, nor do I want 'Check swap' on wingnut
to be dependent on 'NRPE status' on fizzgig.

Now that I'm reviewing the Service Dependency links (see URLs below) once
again after a long absence, I suspect that multiple hosts for either
host_name or dependent_host_name could only refer to template tricks, and
not any sort of matrix of dependency.  I will proceed with this suspicion.

Comments welcome.

jc

> -----Original Message-----
> From: Jamie [mailto:jamie at bclnz.net]
> Sent: Monday, February 10, 2003 3:23 PM
> To: Carroll, Jim P [Contractor]; Nagios-Users (E-mail)
> Subject: Re: [Nagios-users] Q: time-saving tricks for Service
> Dependencies?
> 
> 
> Hi Jim,
> 
> "..In short:  I'd like to take advantage of this time-saving 
> feature, but I
> don't want to end up creating logic where a vast selection of 
> services are
> dependent on one particular service being up or down.."
> 
> -Excuse me for stating the obvious, but I guess that depends 
> whether logic
> reflects reality or not...(Hmm. begin debate: "Is Reality 
> logic? - Or is
> Logic reality?" ;-)
> 
> Aren't service dependencies de-coupled from host dependencies?
> 
> jamie
> ----- Original Message -----
> From: "Carroll, Jim P [Contractor]" <jcarro10 at sprintspectrum.com>
> To: "Nagios-Users (E-mail)" <nagios-users at lists.sourceforge.net>
> Sent: Tuesday, February 11, 2003 8:23 AM
> Subject: [Nagios-users] Q: time-saving tricks for Service 
> Dependencies?
> 
> 
> > Just looking for some 'warm fuzzies' here.  In:
> >
> > http://nagios.sourceforge.net/docs/1_0/templatetricks.html
> >
> > under "Service Dependencies", there appears to be a great 
> time-saver.
> >
> > However, I'm concerned that creating this definition will 
> mean that *all*
> > hosts defined in the dependent_host_name declaration would become
> dependent
> > on *all* hosts defined in the host_name declaration.
> >
> > (I have a similar concern when using a
> > dependent_hostgroup_name/hostgroup_name approach.)
> >
> > Can anyone say with authority and conviction that (using 
> the example given
> > on that page) SERVICE2 on HOST3 is dependent on SERVICE2 on 
> HOST1, and
> > SERVICE2 on HOST4 is dependent on SERVICE2 on HOST2, but 
> that there's no
> > crossover where SERVICE2 on HOST3 becomes dependent on 
> SERVICE2 on HOST2,
> > and likewise with SERVICE2 on HOST4 crossing over to 
> SERVICE2 on HOST1?
> >
> > A 1-to-1 correlation would do quite nicely, so long as 
> there are N hosts
> > defined on each of the dependent_host_name and host_name lines.  A
> > 1-to-many, many-to-1, or many-to-many is *not* what I want.
> >
> > In short:  I'd like to take advantage of this time-saving 
> feature, but I
> > don't want to end up creating logic where a vast selection 
> of services are
> > dependent on one particular service being up or down.
> >
> > Thoughts?
> >
> > jc
> >
> >
> > -------------------------------------------------------
> > This SF.NET email is sponsored by:
> > SourceForge Enterprise Edition + IBM + LinuxWorld = Something 2 See!
> > http://www.vasoftware.com
> > _______________________________________________
> > Nagios-users mailing list
> > Nagios-users at lists.sourceforge.net
> > https://lists.sourceforge.net/lists/listinfo/nagios-users
> >
> 


-------------------------------------------------------
This SF.NET email is sponsored by:
SourceForge Enterprise Edition + IBM + LinuxWorld = Something 2 See!
http://www.vasoftware.com




More information about the Users mailing list