I have a multi-master kubernetes v1.1.3 setup. I use the hyperkube binary to run kubelet on the master and have the scheduler and controller manager running in pods with the podmaster acting as the elector. The master's are loadbalanced using HAproxy, and the kubelet on the minions uses that loadbalanced IP as their apiserver endpoint. I am using docker v1.8.2.

I am not having any errors in use, I can deploy pods that are running without issue across the minions and everything seems happy.

However, I am getting constant error spam (every 10 seconds) from the kubelet on the master:

[err] [kubelet] E0201 21:10:54 kubelet.go:1361] Mirror pod not available
[err] [kubelet] E0201 21:10:54 kubelet.go:1361] Mirror pod not available
[err] [kubelet] E0201 21:10:54 kubelet.go:1361] Mirror pod not available
[err] [kubelet] E0201 21:10:54 kubelet.go:1361] Mirror pod not available
[err] [kubelet] E0201 21:11:04 kubelet.go:1361] Mirror pod not available

I am concerned that this is symptomatic of something set up improperly that I am missing. Is this anything to be concerned of? If not, is there anything I could configure to silence the error spam?

edit: here are the options I am running kubelet with:

kubelet on master (mostly standard):

hyperkube kubelet --v=2 --config=/etc/kubernetes/manifests --kubeconfig=<path-to-config> --port=10250 --address=0.0.0.0 --cadvisor-port=4194 --allow-privileged=true

all pod manifests: http://pastebin.com/MJfQz78r