Service are checked twice

Hey,

By verifying log, I’ve detected all Service check are executed by each Satellite server (and the same for EventCommand). Something there talk about that: https://github.com/Icinga/icinga2/issues/3733
But the bug is fixed.

After a long time of Googling, I guess you’re my last hope.

Here a part of my configuration:
version: r2.8.0-1 on all node.

zones.conf (the same for all icinga node)

object Endpoint "icinga-master1.localdomain" {
  host = "172.16.0.151"
}

object Endpoint "icinga-master2.localdomain" {
  host = "172.16.0.152"
}

object Endpoint "icinga-satellite1.localdomain" {
  host = "172.16.0.161"
}

object Endpoint "icinga-satellite2.localdomain" {
  host = "172.16.0.162"
}

object Zone "master" {
  endpoints = [ "icinga-master1.localdomain", "icinga-master2.localdomain" ]
}

object Zone "satellite" {
  endpoints = [ "icinga-satellite1.localdomain", "icinga-satellite2.localdomain" ]

  parent = "master"
}

object Zone "global-templates" {
  global = true
}

object Zone "director-global" {
  global = true
}

Enabled features on icinga-master1 and icinga-master2:
api checker command ido-mysql livestatus mainlog notification

Enabled features on icinga-satellite1 and icinga-satellite2:
api checker command mainlog

I deploy all Host objects into the “satellite” zone.
Template, command, CheckCommand, etc… are stored in director-global.

Hope these details are a good point to start the discussion :slight_smile:

Thanks
K.

Ensure that the satellite nodes are connected to each other (e.g. cluster health checks). Otherwise each instance thinks that it is in a split brain scenario and will attempt to execute checks on its own. You can also find out about that by checking the object authority, and respective active/paused attributes als shown in the troubleshooting docs with late check results.

https://www.icinga.com/docs/icinga2/latest/doc/15-troubleshooting/#late-check-results-in-distributed-environments