Re: [Nagios-devel] Why distinguish hosts from services?
Posted: Thu Aug 07, 2008 5:13 am
* matthias eble [2008-08-07 13:48]:
> On Wed, 2008-08-06 at 14:17 +0200, Holger Weiss wrote:
> > Note that I don't doubt the usefulness of syntactic configuration sugar,
> > such as the implicit service->host dependencies or the nice and simple
> > way of mapping the network topology using the "parents" directive. The
> > thing I don't really understand is why Nagios distinguishes hosts from
> > services internally (outside the configuration parser). However, I may
> > well be overlooking something, so I figured I'd ask what it is
> ???
> I think some stripped down host configuration should be kept. sth like
>
> define host {
> host_name foo
> address 127.0.0.1
> avail_service icmp_service
> }
>
> where icmp_service would be a common service
> and would prevent some redundant config parser code I think.
>
> But if you understood you correctly, you are asking for internal rather
> than configuration changes.
Yes, though I could also imagine a syntax which gives the user control
over the implicit dependencies of arbitrary user-defined object types.
But that's not my main point indeed.
> This would mean having a "monitoring object" rather than host or
> service structures, and automatically created object dependencies
> (services still rely on the monitored "host").
Yes.
> Beside that, the states for hosts and services differ (hosts have an
> unreachable state as well as state texts differ).
Yes, that's one of the differences in the current implementation, but I
don't really understand why this difference is required.
> With one monitoring object, unreachable states would automatically be
> valid for services, too.
Yes! It's a feature (see Ton's reply)!
Holger
This post was automatically imported from historical nagios-devel mailing list archives
Original poster: [email protected]
> On Wed, 2008-08-06 at 14:17 +0200, Holger Weiss wrote:
> > Note that I don't doubt the usefulness of syntactic configuration sugar,
> > such as the implicit service->host dependencies or the nice and simple
> > way of mapping the network topology using the "parents" directive. The
> > thing I don't really understand is why Nagios distinguishes hosts from
> > services internally (outside the configuration parser). However, I may
> > well be overlooking something, so I figured I'd ask what it is
> ???
> I think some stripped down host configuration should be kept. sth like
>
> define host {
> host_name foo
> address 127.0.0.1
> avail_service icmp_service
> }
>
> where icmp_service would be a common service
> and would prevent some redundant config parser code I think.
>
> But if you understood you correctly, you are asking for internal rather
> than configuration changes.
Yes, though I could also imagine a syntax which gives the user control
over the implicit dependencies of arbitrary user-defined object types.
But that's not my main point indeed.
> This would mean having a "monitoring object" rather than host or
> service structures, and automatically created object dependencies
> (services still rely on the monitored "host").
Yes.
> Beside that, the states for hosts and services differ (hosts have an
> unreachable state as well as state texts differ).
Yes, that's one of the differences in the current implementation, but I
don't really understand why this difference is required.
> With one monitoring object, unreachable states would automatically be
> valid for services, too.
Yes! It's a feature (see Ton's reply)!
Holger
This post was automatically imported from historical nagios-devel mailing list archives
Original poster: [email protected]