0

I have the problem that the connection between the host running icinga2, and between the actually checked server machines, is a little bit... non-deterministic.

Thus, sometimes - every day for some minutes, in an unpredictable time - all the hosts and services seem went down by the icinga.

I don't want it. I think the best solution to work around this problem, if I could somehow set up a minimal downtime length, for example, 10 minutes.

Alarms should be created only if a downtime is longer as this value.

The obvious side-effect (alarms get this delay + lack of monitoring for the short network downtime) is not a problem for me.

peterh
  • 9,488
  • 16
  • 59
  • 88
  • Alternatively, any other solution is acceptable, if it solves the main problem (makes icinga to avoid false alarms). – peterh Feb 19 '18 at 10:22

1 Answers1

1

You can change check_interval and retry_interval in combination with max_check_attempts

template Host "10mindowntime" {
  max_check_attempts = 10
  check_interval = 1m
  retry_interval = 1m
}

You can change the times and attempts to fit any level of granularity you want after the ten mins is up.

123
  • 1,562
  • 7
  • 9