I have Nagios installed and I'm working on getting the network map correct. The problem I have is that "Nagios" appears to be in the "internet" when it should be located on the MechNAS server.

enter image description here

What I want is Nagios Process to show up inside the local network. So it should show up at the same layer as MechNAS and development. Where exactly is that configured? I didnt see any place to set that up and it looks now like it's out there on it's own. Documentation and Googling didnt seem to turn up anything either.


This bugs the buhjeezuz out of me too.

According to the Nagios knowledge base, you can't:

When using one of the auto-layout modes of the statusmap or statuswrl CGIs you cannot hide the Nagios process icon, nor change its location. Its location is automatically determined, so that it is always at the "root" level of the map.


You might want to look into the Nagvis addon for more options for visualization. It is MUCH prettier then status map.

  • Yeah, I didnt want/need anything fancy, I just hate it when something visually looks wrong and you cant fix it. :) – Mech Software Feb 8 '11 at 22:47

Looks like it is not configurable. Which is a shame. Nagios FAQ F0081


You can't make "Nagios Process" actually appear on a server, but the graph should be coming out ordered by the parents fields of the hosts. To get this graph to draw the other way around, try making it so that the hosts directly reachable from nagios (MechNAS, gateway and development) have no parents. Then, dsl_gateway would have "gateway" as its "parents" setting, then all the external servers would have dsl_gateway as their parents.

  • While that might work it would also make the map just as inaccurate wouldn't it? Nagios is behind the gateway and dsl_gateway which means that if either of those goes down Nagios can only monitor Development and MechNAS, the other servers might be running but Nagios can't see it. I think I'll just have to live with the entire map being wrong or for "Illustrative purposes only". I think the Nagvis is the only suggestion that might work out. – Mech Software Feb 9 '11 at 14:49
  • 1
    @mech: The whole point of the parent mechanism is to allow nagios to tell the difference between a host being down and a host being un-reachable because the network between nagios and the monitored system is down. So in a nagios world, ideally, you put the gatways on nagios and make those the parents of the other things. – chris Apr 27 '11 at 11:11
  • @chris Yes, kind of. A true representation would have the host Nagios is running on at the center (or root node) of the graph, as it is the central observer. From there, as we all know, Nagios is talking to the switch(es) that system is connected to, other systems in that datacenter hung off of that switch, through a firewall (and ideally an (un-monitored greyed out) Internet cloud after that), to the FW at the next DC, to switches, to hosts, to services. Still, as it is, it's still a useful tool to hand a new-hire so they can see the connectivity and hence shape of the organization. – Andrew Beals Mar 30 '16 at 15:23

It's designed this way as the map is meant to be a graphical representation of the host dependencies.

All hosts have a dependency on the Nagios process so it always appears at the top (or centre) of the map, for instance, in your environment, access to "MechNAS" and "development" from your Nagios process relies on the DSL_Gateway and the Gateway device being "UP" whereas the other devices on your network are directly visible to Nagios (at least that's what your map is saying)

If you want an alternative visualisation, try out NagVis. You can get it to automap (and it looks better than the nagios one IMO) or you can specify your own layout (dependency independent) so you can either have the Nagios process listed or not.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.