I’m running the native Linux Checkmk agent on a Kubernetes node. As Kubernetes creates filesystems and network interfaces for every new pod/container, I created a rule to disabled e.g. the service Filesystem /run/k3s/containerd/.*.
All new Filesystem /run/k3s/containerd/.* will be added to the node in Checkmk as disabled service, but there’s no cleanup when it vanishes. Is there a way to configure automatic cleanup of disabled vanished services?
Hi dahag.
you can config this rule “Periodic service discovery”
if you observe that it is not applicable to specific services but to hosts, keep in mind that a critical service for you may disappear and it may take you some time to realize it.
I added a service filter, so only some services are removed:
But still there are many disabled vanished filesystems:
Same problem here, thousands - leading to 413 errors when rediscovering services etc. Vanished disabled services should be removed automatically.
1 Like
Is there any solution in sight?
Easiest version is a “cmk -II hostname” on command line, then are all disabled & vanished services are gone. But a “tabula rasa” inside the setup should also lead to the same result.