Containers aren't entirely closed off from the outside world. You can use the
exec commands of the Docker CLI and Docker Compose to spawn an interactive
shell inside of a container. This can be particularly useful for debugging, but
you can also use this technique to turn containers into reproducible development
environments.
This example contains a Go web server. Inside the image for the service,
debugging tools have also been installed, allowing you to attach to the
ENTRYPOINT of the container and debug it.
Because debugging requires special permissions, certain security features have been disabled for the container. This is something you should only do in development.
Start the project with:
docker-compose up --build
You should be able to access the server on
http://localhost:8080.
Now try dropping into the container using:
docker-compose exec web /bin/sh
You'll be dropped into an interactive system shell inside the container. Start a
debugger and attach it to the ENTRYPOINT process. Then set a breakpoint and
let the process continue.
dlv attach 1
break handler /server.go:10
continue
Now try accessing the server again from your browser. If you hop back to the
terminal where you're running exec, you'll see that the break point has been
triggered. You'll have to type continue before the HTTP request will finish
being served.
This is just a basic example, but you can perform far more extensive debugging operations with this same technique.
To shut down the project, simply Ctrl+C the docker-compose up command and
then run:
docker-compose down --rmi=local
This will automatically terminate your exec session.