Up/down state of hosts without real observation possible?

I’m playing around with check_mk_raw 1.6 and I’m a bit overwhelmed by the flexibility and wide range of functions. Most configuration is quite easy but I have a task which I don’t know excatly how to realize.

I like to have an overview about the up/down state of some hostes/devices in one network segment. What is the best way to configure these hosts, so I can see them and their current state e.g. in the host overview, but not in the section “Host problems (unhandled)”? Notifications are already disabled with special labels for these hosts and a corresponding rule [Enable/disable notifications for hosts].
Thanks for any advice.


So are you looking for a view for a specific group of hosts?

Hello ,

no special view, I think. I don’t like to see these hosts in my standard area “Host problems” because their down state is no real problem but only an information for me. Imagine a special printer which is only used sometimes. While monitoring this I can see the state without walking to the end of the building, otherways it could be a nice way to log the operating times.

But with alle of these devices in the list of down hosts it is possible to lose sight of other important machines which are down. So I think I need a special flag “down but don’t care anyway” or something like this.

What you could do is:
Create a Host tag called something like “Down but don’t care” (use your imagination :slight_smile:) and then edit the Host Problems view to ignore any host that has the tag you just created.

Hmm, I see. Maybee I can reach my goal with “Context/Search Filters” in a customized view. I will try and give you some feedback. Editing views and using them was not on my agenda 'til today :wink:

If your main goal is to not see that specific group of machines in a view, I’m pretty sure you’ll have to edit it :slight_smile:

Let me know how it went or if you need further help!

I’ve tried to customize my view(s) but I’m not very happy with it.

In the dashboard I could hide the devices with additional search filters (Auxiliary Tags) in the “Host problems” view. But for host and service statistics view desklet (the coloured sphere) unfortunately there are no such filter rules available. Maybe there’s another way I can achieve my goal. But in the end this makes my configuration complicated and not reproducible for others.

A global rule which affects the view(s) for all devices in a folder or with special tags/labels would be the smartest solutions. I’m still learning and will try a little further. If I find a satisfactory configuration I’ll let you know.

Any other suggestions are of course welcome.

there is currently no global filter to get rid of some hosts in all dashlets. I see two solutions for your problem:

  1. Use an additional site on the same host to monitor this hosts. I use this sometimes for big sites with multiple test hosts to not have them in productive monitoring at all.
  2. Create a user that is not allowed to see all hosts, but only those in a defined host group. If you login with this user you only see the relevant hosts. I use this for NOC display on dedicated wall mounted monitors to only see productive outages there.
    Best regards

Thank you Udo for your ideas. A second site for special devices is a simple solution. It allows most flexibility with only a little extra configuration work. I’ll think about it, also about your second proposal.

Meanwhile I played around a bit and found another solution that works for me. I will describe my procedure here for discussion respectively as basis for anyone has a similar challange.

  1. I created on top of my site two folders “Business Critical” and “Just for info”. For “Just for info” I defined at least the option “No Check Agent” and a label “notification:no”.

  2. I created a global rule “Disable host notifications” with condition “Host label has notification:no”

Every device which ist created in or moved to folder “Just for info” will inherit this label and so no notification/alarm will be createt in down state.

  1. Now I had to modify 3 desklets in Main Overview dashboard: Host statistics, Service statistics and Host problems (unhandled). All desklets allow Context/Search filters. I added a filter condition: WATO folder = Business Critical.

As a result only hosts and devices in folder “Business Critical” will be displayed here. The desklet “Service Problems (unhandled)” does not need to be adjusted, because services of hosts in down state are not raised as default.

Any other view can optionally display devices from one or the other folder or both together, too.
Perhaps in large setups with many groups or departments my configuration will be not the best way.