Project template for a python docker project

This project is intended as an initial template for composing dockerized project with all the base requirements we would like to see baked in.

What does this template do out of the gates:

Static Analysis using pylint
Security scanning using bandit
Unit tests with code coverage and minimum coverage settings using pytest
Docker container composition
Execution of acceptance tests against the docker container using Cucumber [NOTE: I want to create a variant that uses lettuce or some other python based acceptance tests instead].
Deployment of the container.

Getting Started

Assert you have all the pre-requisites for development!
Clone the project ( git clone git@github.com:adammartin/py_docker_template.git)
Rename the root folder to your project name and cd into it.
Update .python-version if you do not wish to use your own python version.
Execute the following command at root of project to reset git and make it your project rm -rf .git && git init
Edit the file bin/build_variables.sh
Update the settings below to your desired values (and you are cooking with fire!):

export MIN_COVERAGE_PERCENTAGE=0
export BASE_IMAGE=[REPLACE_ME_WITH_IMAGE_YOU_WANT_AS_BASE]
export DOCKER_REPO=[DOCKER_REPO_YOU_WILL_COMMIT_TO]
export VERSION=[VERSION_OF_IMAGE]
export PROJECT_NAME=[YOUR_PROJECT_USED_TO_NAME_IMAGE]
# DEPLOY ONLY
export CERT_PROVIDER=[letsencrypt-prod|letsencrypt-staging]
export HOST_URL=[KUBERNETES_HOST_URL]
export ROUTE=[ROUTE_ON_KUBERNETES_CLUSTER]
export SECRET_NAME=[CREDENTIAL_NAME_ABLE_TO_ACESS_PRIVATE_REPO]

BASE_IMAGE: This is the base image you are going to use for your project. An example might be tiangolo/uwsgi-nginx-flask:python3.6. It is expected this image contains flask by default. The container must have the following:
- Python 3 or better
- Flask
- A mechanism to bind traffic to port 80 (you can do others but you'll have to configure it) [Example: ]
DOCKER_REPO: This is YOUR repository you will be pushing to.
PROJECT_NAME: This is what you want to call your new container.
VERSION: Docker image version (ex. v1.0) that the image is tagged with.
MIN_COVERAGE_PERCENTAGE: This is the setting that changes the percentage of code coverage to fail the build. Currently it is set to 0% so we do not fail the build ever unless a test fails. We recommend 90% as a starting point.
CERT_PROVIDER: (DEPLOY ONLY) Which lets encrypt provider to use for certs [letsencrypt-prod|letsencrypt-staging].
HOST_URL: (DEPLOY ONLY) Kubernetes host to deploy to.
ROUTE: (DEPLOY ONLY) Route in the kubernetes from root ('/')
SECRET_NAME: (DEPLOY ONLY) Kubernetes cluster secret name that has the credentials necessary to connect to the private container registry source.

For deployment if you are pulling from a private repository there are two additional settings you likely need to adjust in the bin/build_variables.sh file that are commented out and one that is not.

# export REPO_SERVER=[PRIVATE REPOSITORY SERVER]
# export TAG="$REPO_SERVER/$DOCKER_REPO/$PROJECT_NAME:$VERSION"
export TAG="$DOCKER_REPO/$PROJECT_NAME:$VERSION"

Simply uncomment and populate REP_SERVER, uncomment the TAG line that references it, and delete or comment out the old export TAG="$DOCKER_REPO/$PROJECT_NAME:$VERSION". This will allow the deploy script to operate.

Prerequisites

Docker host
Bash shell

Production requirements

Something to deploy Docker Containers to! (For example AKS)

Local development requirements

Requires:

Docker
Bash for shell execution
RVM to manage Ruby (Acceptance Tests in Cucumber)
Ruby version 2.5.1 (Acceptance Tests in Cucumber)
pyenv to manage which python version you are managing for local development and build activities (default is python 3.6.3)
module venv must be installed on your python version!

Building

It is assumed you have already followed steps described in Getting Started. From there you now have the following scripts at your fingertips!

bin/build.sh: This script is the base script that will set up the environment, execute static analysis, execute security scan, execute unit tests, build the container, and execute acceptance tests.
bin/deploy.sh: This script is the base script to deploy the composed and pushed image.
bin/run_local.sh: This script will execute the image with the latest tag. Tailer to your needs.
bin/cleanup_environment.sh: This script will clean up excess images and containers laying around that are not tagged with latest tag. Tailer to your needs.
bin/build_variables.sh: Variables used through out the scripts.
bin/run_unit_tests.sh: This script will run the unit tests.
bin/analyze_python.sh: This script will do local static analysis and security scans.
bin/analyze_and_test.sh: This script will execute the tuple of analyze_python.sh and run_unit_tests.sh
bin/run_acceptance_tests.sh: This script will execute the acceptance tests. Note that if you run this outside of the bin/build.sh script you will need to set the IMAGE_ID environment variable to be passed into the acceptance tests scripts for firing up the correct image in a container.

Running the Tests

Deployment

Things to Uncover

Is application_ingress.yaml and application.yaml the right names. I think that creates a problem and we should be using the real application name for those files?
How do I update routes after initial deployment?
How do I use alternate routes other then '/' I tried it once and it didn't go so well?

adammartin / py_docker_template