arch_go

Creates docker images containing Arch Linux and Go (Golang) installation.

How does it work

This repo contains 3 files:

• A Dockerfile that is used to create the actual image;
• A bash script used to ensure that for the most of the available Go versions is a built Docker image;
• CircleCI config file that is used to start the bash script periodically;

Dockerfile

This can be found in here. This is receiving one argument called go_version which is should contains the Go version that has to be built in the current image. The only thing that is done here is to download the binary and extract it to /usr/local and update the PATH of the image in order to make the go binaries available right away.

Bash script

• This can be found in here.
• This is responsible with fetching the already built images from DockerHub.
• Lists most of the releases from the Golang git repo and for every one of those that has no image built in arch-go, is building it and push it to the DockerHub.
• Needed environment variables:
  • DOCKER_HUB_USER_ID - the user of the docker hub where the images will be pushed to;
  • DOCKER_HUB_USER_PWD - docker hub access token to have access to push the images;

CircleCI config

This can be found in here This is just a config file that gives to the script the ability to perform Docker related commands. Also in this file is a configuration for a CircleCI workflow to be executed once a week, on Sunday, meaning that if since the last Sunday a new Golang version was released, then the correspondent image will be built as well.

How to use it

You should just use the normal Dockerfile statement

FROM yottta/arch-go:<tag>

or using a command similar with this one

docker run --rm -it yottta/arch-go:<tag> bash

where <tag> should be replaced with the tag of the image that you want to use.