Possibly related to my other thread, not sure how to reproduce consistently.
When I try to scale a single container that is part of an app that was created with docker-compose, I will get the “Error scaling container 500 Internal Server Error: Could not get container for (container name)” message, where (container name) is the name of one of the other containers that is part of the app.
When I go back to the application view after this error occurs, I will see a few “dead” instances of the container is was trying to scale. Their icon will be colored red but the count of “exited” containers will still be 0. When I investigate the container, it will have a status of “created” but no logs will display for that container.
I ran the Docker compose command from the node with the UCP leader and the dead containers it creates are on one of the other follower nodes in the cluster. I have confirmed that all the images associated with the application are present on that follower node so it should have no trouble getting the images for the new containers.
I have no clue why this is happening or what else to try. Please let me know any further information I can provide to help debug.
Docker Engine 1.10
Swarm 1.1.0-rc2 (installed by the UCP install)
All on EC2 instances.