In some cases container agent may try and scheduled the pod before a node is alive and healthy. It would be great to add a small delay and/or retries to container agent to allow the node to fully come up.
Created by Logan Godsey