I use Kubernetes version 1.15.3 . I use calico as cni plugin.

I have a deployment that has service configured to as Nodeport on a port serviced by HTTPS protocol.

For some reason when I look at one of the pods of the deployment I see this output:

2020-01-07T08:45:02.905Z [INFO]  http: TLS handshake error from 10.233.92.0:60286: tls:first record does not look like a TLS handshake
2020-01-07T08:44:42.884Z [INFO]  http: TLS handshake error from 10.82.83.213:45654: tls: first record does not look like a TLS handshake

The IP 10.82.83.213 is the IP of the node and the second IP is some internal IP of Kubernetes.

These errors are not interfering with the actual work of hte pod since it is up but I makes it hard to look for pother errors since this lines appears every 4 seconds.

Initially I tried removing the Probes but that did not help.

Once I remove the NodePort service or changes the label these error stops but ofcourse then the pod is not reachable.

Does any one knows the root cause for this? I would assume kube-proxy but I couldn't find evidence of that.

thanks