Delay Service Problems

Hello forum,

I am searching for a way to delay Service Problems. I don’t want to see alarms (warn/crit) before a certain amount of time is gone. I already set up the “Delay service notifications” rule, but this only seems to delay notifications (such as mail) but not the alarm on the Dashboard itself. (Correct?)

[My use case: I am using the “APT Updates” plugin for Linux machines. I am updating all Linux machines automatically at night. That is: I only want to see “security updates available” if those updates are not installed during the last night. Hence I want to delay all those messages by 24 hours. But I don’t know how.]

Thanks for your help.
Johannes

Hello @webernetz
as far as I know this is not possible.
This Question was also discussed here:

You could make a feature request here:

I would vote for it, because I also need such a feature :slight_smile:

1 Like

Hi @webernetz ,

Dashboards are showing hard states and therefore there are some rules you can play around to may accomplish your need. A hard state is determined by your check frequency and the number of check attempts. So if you upper one or both of this value you can delay the change from OK to WARN.
The rule for this would be
Maximum number of check attempts for service (The maximum number of failed checks until a service problem state will be considered as hard. Only hard state trigger notifications.)

1 Like

Thanks for your replies, guys! I’ll have a look at the check attempts.

I am using the “APT Updates” plugin inside an /3600/ folder. That is: this plugin is only executed once an hour. Hence I should use a “24” to delay the hard state for a day:

Looking forward to it. :wink:

Hello again. It is still not working as I expected. What else am I doing wrong?

I configured the “Maximum number of check attempts for service” to only get alarms after one day:

Indeed this counter works on the service:
grafik

But the CRIT service problems are still shown immediately in the Dashboard and in the Host Matrix:

I do not find a filter that says something like “hard state” or the like. Shouldn’t these warnings change to CRIT status after 24 hours?

Thanks a lot!
Johannes

Please check your view, there should be a option to determine which state (hard or soft) you like to have on this view.
image

Seems like default is to show soft states.

Yeah, thanks again. I found it. I can now omit these soft warnings at the “Service Problems (unhandled)” view.

When I am using this filter on the “main dashboard”, the “Service Statistics” are updated (though omitting the acknowledged problems as well), while the “Host Matrix” on the Sidebar still shows the yellow/red boxes. Grr.

I think I need to rethink my intended use. Maybe I’m just dropping this disturbing service check. That’s probably easier. Or I will rule it to be always OK.

Thanks again for your help! Appreciate it.

At this point, maybe it is easier to write a small local check which is checking the waiting security updates against the last apt-get install run and only report CRIT if it’s more than 24h ago.
Up to how your are doing the nightly update you can write a file with the timestamp of update and check with Size and age of single files if the file is older than 24h. in this case you don’t need to fiddle around with display and hides of services.

Wow. Thanks for the ideas.

1 Like

This topic was automatically closed 365 days after the last reply. New replies are no longer allowed. Contact an admin if you think this should be re-opened.