Hi,
Rancher 1.6.14 with Kubernetes 1.8.5 is displaying a very weird issue. When I want to connect to a pod to get the shell or view the logs, either through the Kubernetes dashboard or using kubectl, I get a ‘no route to host’. When I reboot the VM it’s fixed for all pods on that node, but the problem issue just moves to another host in the cluster. Feels like a rancher/k8s pod causing the issue. Anybody experienced this too?
I really don’t want to give devs SSH access to all nodes just to get to their pods.
Kind regards,
Eric V.