Issues on the state of containers

[szzdmj]

Hi all. My container can work very well after deployed, and can successfully stop according to sleepafter specified, then can restart.

Seems everything is ok, but the problem is containers can't restart as requested, like the requests were sent directly to a container actually stopped but the worker or DO thought running, which can easily check the state on the dashboard, then got an error like "Error proxying request to container.....", until a container restarted and was really running but no idea why and when it would do. Especially, if you have some containers showed all running but actually several not, it's like a lotto to get a correct response.

And the worse is after containers become inactive or stopped for a long time, it's much more difficult to restart them, like always fail healthcheck so that errors like "Error checking if container is ready...." and "There is no Container instance available at this time......." appear again and again.

I think those may be issues updating the state of a container. Any ideas or suggestions?

[widavies]

This is happening to me too, the service is basically unusable right now - really excited for it to release!

[widavies]

This was a PICNIC (person in chair, not in computer) problem - I was on an old template with a compatibility date set to March 2025. Updating the compatibility date to the current date fixed this issue.