I got some error when scheduling pod though ReplicationController:
failedSync {kubelet 10.9.8.21} Error syncing pod, skipping: API error (500): Cannot start container 20c2fe3a3e5b5204db4475d1ce6ea37b3aea6da0762a214b9fdb3d624fd5c32c: [8] System error: Activation of org.freedesktop.systemd1 timed outThe pod is scheduled but cannot run unless I re-deploy it with another image.
I'm using kubelet 1.0.1, CoreOS v773.1.0
The part that says Error syncing pod, skipping: API error means that kubelet got an error when trying to start a container for your Pod.
Since you use CoreOS, I think you are using rkt, not docker.
I think that rkt uses systemd to start containers.
And I think systemd crashes when the "unit" name starts with an underscore: https://github.com/coreos/go-systemd/pull/49
So, maybe one of your pods or containers has a name that starts with an underscore. Change that.