Check gets executed twice with different options

Icinga 2
1

Hi,

i tried to change my check “disk” to not monitor stuff like /snap/core/…
so i set something like

apply Service “disk” {
import “agent-service”

check_command = “disk”

vars += config

vars.disk_cfree = "3%"
vars.disk_wfree = "5%"
vars.disk_partitions = [ "/" ]

vars.disk_all = “true”

vars.disk_local   = "true"

vars.disk_exclude_type = [“overlay”,“tmpfs”,“nsfs”,“sysfs”,“shm”,“debugfs”,“tracefs”,“nfs”,“squashfs”,“/dev/loop0”,“/dev/loop1”]

vars.disk_ignore_ereg_path = [“/daten/dddocker”,“/run/docker”]

vars.disk_partitions_excluded = [“/dev”,“/mnt/sugarcrm”]

vars.disks[“disk”] = {

disk_partitions = “/”

#}
assign where host.vars.os == “Linux”

}

already before my start the disk state was flapping every 30 seconds.
I see in Icinga Web, that the check should be executed every 60 seconds (what is also my config)
every 60 seconds the check is executed as intended only showing “/”
but after 30 seconds there is another disk check executed on this machine.
the timer for seconds since last check starts new.
the timer for seconds until next check stays.
After this check /snap/core/ is displayed again in the list of partions.

I’m searching for advise how to search source of this execution.
Its not happening on any other check then disk.
Service is only configured onetime in from what i can find.
Its happening to all Machines the service is assigned.

BR,
Daniel

2

On the system you execute the check, is the example configuration under conf.d still included? Or some other local configuration?

If there is some local configuration and the feature checker is enabled, it will locally schedule and execute the checks. When the system is now part of a cluster it will report check results to its parent. If now there is a service with the same name, check results will be accepted and you will have the same service twice with different parameters.

1 Like
3

That was the case, thank you alot :slight_smile: