I use docker just often enough to forgot the basic command-line arguments.

-v, mount a local folder as volume accessible within container

Example:

docker run -v $(pwd):/app alpine cat /app/hello.txt

mount app folder in current dir, run alpine container with cat command to output hello.txt file contents.

-it, run container in interactive mode docker run -it -v $(pwd):/app alpine /bin/sh. run in interactive mode, mount dir, run shell

-e, set environment vars

docker run -e NAME=World node:latest node -e "console.log('Hello' + process.env.NAME)", run and set environment var

-p, map ports localport:containerport

-d, detached to run in background

—name, make it easier to interact with container

—rm, automatically remove container, otherwise will need to run docker rm <container-name>

docker ps, list running containers

docker stop <name of container or first few chars of container id>, stop container

docker network create alpine-network, create docker network

--network alpine-network, add container to network

docker network inspect alpine-network, shows containers and ip addresses on network

Docker adds the container names to each containers host so you don’t need to get the internal IPs to have them communicate with each other.

Multistage build -> Use a container to build another container

Containerize

Identify all constant values, environment vars. Config vars, log files, etc.

Dockerfile

Always begins with FROM statement to tell it the BASE image

Specify WORKDIR which can be whatever you want.

Specify setup commands, create dirs, copy files, run npm install, etc.

Then run CMD to start the app. ex. CMD [“node”, “/app”]

Build the container:

docker build -t <tagname> <directory of Dockerfile>

ex. docker build -t k8scourse-back .

docker logs <name>, display logs from container, use name or ID