Skip to content

Latest commit

 

History

History
144 lines (105 loc) · 4.64 KB

README.md

File metadata and controls

144 lines (105 loc) · 4.64 KB
Raw

Supported tags and respective Dockerfile links

Simple Tags

Pyskycoin CLI/DIND development image

This image (CLI) has the necessary tools to build, test, edit, lint and version the Pyskycoin source code. It comes with some versions of Python (2.7, 3.4, 3.5 and 3.6) and with Vim editor installed, along with some plugins to ease go development and version control with git.

Besides it is possible to use Docker in Docker (DIND) Pyskycoin development image, it is based on skycoin/skycoindev-cli:dind and provides all tools included in Pyskycoin CLI image.

How to use this image

Initialize your development environment.

$ mkdir src
$ docker run --rm \
    -v ${PWD}/src:/usr/local/src skycoin/skycoindev-python:develop \
    git clone https://github.com/simelo/pyskycoin.git \
$ sudo chown -R `whoami` src

This downloads the pyskycoin source to src/pyskycoin and changes the owner to your user. This is necessary, because all processes inside the container run as root and the files created by it are therefore owned by root.

Running commands inside the container

You can run commands by just passing them to the image. Everything is run in a container and deleted when finished.

Running tests

$ docker run --rm \
    -v ${PWD}/src:/usr/local/src skycoin/skycoindev-python:develop \
    sh -c "cd pyskycoin; make test"

Editing code

$ docker run --rm \
    -v ${PWD}/src:/usr/local/src skycoin/skycoindev-python:develop \
    vim

Switch python environment

Inside docker image is installed pyenv and pyenv-virtualenv, with one goal in mind:

  • Add a ready-to-use virtualenv in Docker dev image with stable version of pyskycoin package pre-installed from PyPI

The virtualenv installed are:

  • pysky27
  • pysky34
  • pysky35
  • pysky37

If you want to access to one specific virtualenv you just need run the next command: bash pyenv activate pysky<version> where version represent python version you need. Once activated, you may deactivate it with: bash pyenv deactivate

For more details with pyenv, visit pyenv doc

All of then contains pyskycoin package installed.

How to use Docker in Docker image

Start a daemon instance

$ docker run --privileged --name some-name -d skycoin/skycoindev-python:dind

Where to store data

Create a data directory on the host system (outside the container) and mount this to a directory visible from inside the container.

The downside is that you need to make sure that the directory exists, and that e.g. directory permissions and other security mechanisms on the host system are set up correctly.

  1. Create a data directory on a suitable volume on your host system, e.g. /my/own/var-lib-docker.
  2. Start your docker container like this:
$ docker run --privileged --name some-name \
    -v /my/own/var-lib-docker:/var/lib/docker \ 
    -d skycoin/skycoindev-python:dind

Build your own images

The build process relies on the following parameters

  • SOURCE_COMMIT: the SHA1 hash of the commit being tested.
  • IMAGE_NAME: the name and tag of the Docker repository being built.
  • DOCKERFILE_PATH: the dockerfile currently being built.

In order to build image from skycoindev-cli:develop execute the following shell command

$ cd skycoin
$ SOURCE_COMMIT=$(git rev-parse HEAD)
$ IMAGE_NAME=skycoin/skycoindev-python:develop
$ DOCKERFILE_PATH=docker/images/dev/Dockerfile
$ docker build --build-arg BDATE=`date -u +"%Y-%m-%dT%H:%M:%SZ"` \
               --build-arg SCOMMIT=$SOURCE_COMMIT \
               -f $DOCKERFILE_PATH \
               -t "$IMAGE_NAME" .

If you prefer to use skycoindev-cli:dind then run:

$ cd skycoin
$ SOURCE_COMMIT=$(git rev-parse HEAD)
$ IMAGE_NAME=skycoin/skycoindev-python:dind
$ DOCKERFILE_PATH=docker/images/dev/Dockerfile
$ docker build --build-arg IMAGE_FROM="skycoin/skycoindev-cli:dind" \
               --build-arg BDATE=`date -u +"%Y-%m-%dT%H:%M:%SZ"` \
               --build-arg SCOMMIT=$SOURCE_COMMIT \
               -f $DOCKERFILE_PATH \
               -t "$IMAGE_NAME" .

Automated builds

Docker Cloud is configured to build images from develop and master branch on every push made after merging. The same process is triggered for all feature branches matching the pattern /^([^_]+)_t([0-9]+)_.*docker.*/. The tag generated for such images will be of the form feature-{\1}-{\2}.