Readme

Playground to mess around with python, electron, containerized.

Dockerfile syntax:

RUN: Runs a command. Has a shell form and exec form, exec preferred. Can have multiple RUN commands in a Dockerfile.
CMD: Only one of these allowed in a Dockerfile
CMD [ "param1", "param2" ]: Does not execute a shell, so no shell preprocessing or environment substitutions occur.
COPY vs ADD: prefer copy, does local files. ADD can do local files, but also does remote files.

Docker run

docker run -it <name> makes a container and starts an interactive session inside of it.
docker run -it <name> <shell> makes a container and starts an interactive session inside of it, specified by <shell>, e.g. /bin/bash.
--name flag to give the container a particular name
docker running a container that doesn't have a command at the end of the dockerfile will just make it run and exit.
-d flag allows it to run in the background, detached from your current terminal. Omitting this will have the non-interactive output appear as a child process of your terminal.
-p 8888:8888 flag maps ports from host:container, in this case 8888

To get results that are written to stdout/stderr, you need to use

docker logs [image name].

Docker exec:

docker exec <container> <command> to run stuff inside container.
docker exec -it /bin/bash to spawn a shell inside container. However, you'll be logged in as another shell of root. Need to use ps -ef to see other shell processes.

docker pause <name> pauses a container's execution.

stopping a docker with docker stop turns it off but keeps it hidden so you can save the state. See hidden docker images with docker ps -a. Use docker rm to delete it from this list.

Seems like they like to make the working directory in images either directly in the root dir or some folder in root dir.

Named mounts in docker run:

docker volume create <name>
docker volume inspect <name>
volumes are stored in the WSL VM.

Bind mounts in docker run:

-w /app sets working directory to "/app"
-v <dir>:/app binds host to the "/app" directory
anything that existed in that directory before the mount is "gone".

See image history:

docker image history <image> shows the layers of the image.

Dockerized node app should have .dockerignore with node_modules in it.

To-do:

~~Learn how to do named mount~~
~~Learn how to do bind mount~~
~~Need to set up docker compatibility with WSL 1.~~
- follow most of this link, particularly the step where you export the docker server line.
- you have to make a symbolic link. /c -> /mnt/c
- when you do bind mounts, you need to use the symbolic link /c/..../dir.
- Docker scan does not seem to work within the wsl 1 environment.

ryry0/serpens

Readme