Docker: container exits immediately with no logs

Step-by-step Solution

1. 1Run `docker logs <container-id>` (use `docker ps -a` to find the id). Even a crashed container keeps its stdout/stderr for a while.
2. 2Add `-it` to run interactively: `docker run -it my-image` — sometimes the image's default CMD is a REPL that exits when stdin closes.
3. 3Override the entrypoint and shell in: `docker run -it --entrypoint /bin/sh my-image` — you now have a shell inside the built image; run the intended command manually to see the real error.
4. 4Check the `Dockerfile`'s `CMD` and `ENTRYPOINT` — a common mistake is `CMD python` with no script, which runs Python, gets no input, and exits.
5. 5Confirm the app is a foreground process. Docker keeps a container alive **only as long as PID 1 is running**. Backgrounding your app with `&` or `nohup` makes the container exit immediately.
6. 6For servers, ensure they bind to `0.0.0.0` — binding to `127.0.0.1` inside a container means nothing outside can reach it, but the container still runs. Not a crash, just unreachable.
7. 7Add a `HEALTHCHECK` and monitor `docker inspect --format='{{json .State.Health}}' <id>` if the container starts but the app inside is unhappy.