In some environments, users have reported that the "exec into pod" feature in the Kubernetes web dashboard will hang indefinitely.
We have found that a common cause of this behavior is running multiple instances of the kubernetes-dashboard pod. The Kubernetes Dashboard does not support HA deployments at this time, so doing so may result in unexpected behaviors such as this. Reducing the replica count to 1 should resolve the issue.
Otherwise, the workaround we recommend is to use "kubectl exec" from the CLI to exec into running containers. More information can be found in the Kubernetes documentation:
https://kubernetes.io/docs/tasks/debug-application-cluster/get-shell-running-container/