Consecutive passive check with API treated as new alert

Service Monitoring
1

Hello everyone,

I am using the API (process-check-result) to do some passive checks :

  • Every 5 minutes, my client machine executes some scripts and sends the result to my Icinga2 by connecting to the API

Everything works fine except that each API call with “exit_status: 2” is considered as a new state.

For instance, when it goes critical, all subsequent criticals are treated as a new state and thus I receive a notification even if my notification setting has “interval = 0”.

Is there a way to make Icinga understand that as long as service did not recover, all consecutives critical are the “same”?

2

Hi & welcome,

I’d assume you need to add volatile = false to your service definition. More details could be found here.