Docker cheatsheet

📅 2017-Oct-19 ⬩ ✍️ Ashwin Nanjappa ⬩ 🏷️ cheatsheet, docker ⬩ 📚 Archive

Build

You need to create a Dockerfile that describes the plan to build an image. Here is a simple Dockerfile named Dockerfile that creates an image based off busybox Linux image into which we are copying a local file:

FROM busybox
COPY haha.cpp /somedir/haha.cpp

• To build this Docker image:

$ docker build .
$ docker build -f Dockerfile .

This looks for a Dockerfile in the current directory and also uses all the content of the current directory as the build context. Having a big build context with lots of files and directories in it can slow down a Docker build.

The build ends printing the hash of the image it has created. This image can be run using docker run commands (see below) by using this hash.

• To build a Docker image that picks the build context from a different directory:

$ docker build -f /path/to/Dockerfile path/of/build/context

Remote registry

• To list all the images and tags from a Docker Hub repository: there does not seem to be a command to do this.

• To pull an image imagename:

$ docker pull <imagename>

$ docker pull ubuntu
$ docker pull nvidia/cuda

For a list of Docker image names that I find useful, go here.

When no tag is specified, like above, docker automatically pulls the image with the tag latest.

Notice that we are not specifying the domain name or URL of any Docker registry. In such cases, Docker will look for and pull imagename from the default Docker registry: https://hub.docker.com/ The image name can be a single word (ubuntu) or a relative path (nvidia/cuda). For a single word image name, its tags can be seen listed at the https://hub.docker.com/_/ubuntu. The image name with a path has its repositories listed at https://hub.docker.com/u/nvidia and the tags of a particular image at https://hub.docker.com/r/nvidia/cuda.

• To pull an image with a specific tag, for example pulling ubuntu with 20.04 tag:

$ docker pull ubuntu:20.04

Note how the : character is used to separate the image name and the tag. In the case of Ubuntu, the list of available tags can be viewed here. Not all registries may support viewing tags like this.

• To pull an image from a specific registry:

$ docker pull <registry.domain>/imagename

You might need to first authenticate to the registry domain before the pull will work:

$ docker login <registry.domain>

After a successful login, Docker will store a key provided by the registry server in your ~/.docker/config.json. It will reuse this key for subsequent pull or push to this registry.

Local images

• To view Docker images available on your local computer:

$ docker images

This is the same as this command:

$ docker image ls

What is listed are the images that you either pulled from a registry or built locally.

• By default, only the images that you pulled by name are displayed. To view all images on your local computer, including the unnamed intermediate images that were pulled down to compose the named images that you pulled:

$ docker images -a

• To view the full repository name, tag name and ID of images, use the --no-trunc option:

$ docker images --no-trunc

• To remove an image with the latest tag:

$ docker rmi repo1/foobar2

• To remove an image with the tag123 tag:

$ docker rmi repo1/foobar2:tag123

• To remove dangling images:

$ docker image prune

• To remove all unused images:

$ docker image prune -a

• Filtering can be used with all the prune commands. For example, to remove all unused images older than a week:

$ docker image prune -a --filter "until=168h"

• To remove unused containers, images and data:

$ docker system prune

This is useful when your filesystem runs out of space due to docker images.

• To remove all unused containers, images and data:

$ docker system prune -a

This pretty much removes all Docker data on your system. This is useful when you really want to clean out everything.

• To look at Docker’s disk usage:

$ docker system df

In this output, the size of the Images is the disk space occupied by your Docker images.

Containers

• To run a Docker image as a container:

$ docker run image_name

This is only useful if running the container runs a service or a daemon inside it automatically.

• To run a Docker image as a container and get a shell into it:

$ docker run -it image_name /bin/bash

This is usually the method to use to have interactive access to a container.

• To view all running containers:

$ docker ps

• To view all containers, not just the running ones:

$ docker ps -a

• To view the full container ID, image name and command use the --no-trunc option:

$ docker ps --no-trunc

• To stop a running Docker container provide its container name (not image name) or its container hash:

$ docker stop container_name

• To remove a container, provide its hash:

$ docker rm 45235345

Copy

• To copy contents of a running container to the host filesystem:

$ docker cp ecstatic_moore:/etc/lsb-release my_lsb_release
$ docker cp 763338cb8aa1:/etc/lsb-release my_lsb_release

As shown above, either the container name or the container ID can be used as the source.

• To copy contents of the host filesystem into a running container:

$ docker cp my_details ecstatic_moore:/etc/my_details
$ docker cp my_details 763338cb8aa1:/etc/lsb-release

Tried with: Docker 18.09.6 and Ubuntu 18.04