I had to reset my config and restore form a recent backup config, and while almost everything worked right again containers are not starting.

• Containers are all in stopped state
• Starting a container never changes the status. No error from the command, but immediately checking the status afterwards shows it remained stopped.
• Despite the containers having logging turned on, the only log messages found are the output from the first container downloading and starting successfully, except that container immediately stops afterwards.
• No other container log messages are ever logged after this. Nor are any other error messages logged in general.
• None of the containers after the first one appear to have even tried to download, as there is no log output for those at all.
• In all other experiences with containers not starting the behavior was very different. For example if there was a problem a container would show it was starting until it failed to start and transitioned back to stopped. There would also be log messages from the container that gave clues as to why. Not this time.

Has anyone run into this?

What version of RouterOS and also what container?

Has anyone run into this?

I did. Didn't restore from backup for a while, not sure on which ROS version, but when I did this issue occurred. I had to recreate all containers.

What version of RouterOS and also what container?

7.12

technitium/dns-server:latest is the first container in the list

I'm going to save my container settings and then nuke them and retry. It may be that the fact the folder and files already existed prior to the restore, due to the 'unclean erase' with the reset to defaults before the restore.

Cleaning up the files (or installing to another directory) worked.

Unfortunately I cannot copy my configs over from the old container, as RouterOS sets the wrong ownership permissions (making them inaccessible to the container) and absolutely ignores chown commands over SFTP/SSHFS.

Has anyone run into this?

Yes, I also experienced it. After a config reset the host (e.g. 7.12.1) boots with all containers stopped and they won't start until I delete and re-create them manually. What is more, when copying a container in winbox, it misses a remote image, so I must type it myself each time. It is very inconvenient and annoying.

As far as file copying is concerned, I mount folders to containers and use winscp to upload/download files to/from these folders. I have never faced any problems with permissions so far. But I didn't try to copy files from one container to another one, since mounted folders are persistent, i.e. never deleted on container rebuild.