Application Version Information Services

The application_version_info is developed with Python, Docker and Travis CI

Table of Contents

• Introduction
  • Python3
  • Docker Image Build
  • TravisCI
  • GitHub
• Build with TravisCI
• Build Locally
  • Prerequisite
  • Build
  • Test
  • Push
• Enhancement Production Readiness

Introduction

This repository builds containerised application as a single deployable artifact along with all the dependencies. This simple application provides detailed versioning information with single endpoint /version.

Python3

This application is written in Python utilising python version 3.8.0.

Docker Image Build

This image uses python:alpine as base image for thin size and install required packages along with copying application code. Dockerfile is here.

Image Build:

• Install required packages and
• Copy the source code
• Build the image

TravisCI

TravisCI pipeline is [here] .travis.yml file here which conatins three stages as:

• Build: Build the containerised application image and will run for all the branches
• Test: This will test by hitting API endpoint to get expected results in JSON format and also stage will run for all the branches.
• Push: Once the Test stage is completed successfully Push stage will run to push the final image to docker repository. This stage will run for only master branch. This stage need to access docker repository to push the image, so need to set docker login credentials as environment variables.

GitHub

All the application code along with CI pipeline sits in GitHub repo which will allow TravisCI to auto build for each and every commit.

Build with TravisCI

• Login to https://travis-ci.com with GitHub login details and from settings a lclick to Manage repositories on GitHub and select the repository to Approve and Install TravisCI.
• Once the TravisCI is installed for the repository, it will build for every push commit by running all stages in pipeline [here] .travis.yml.

Build Locally

Build and Run application locally

Prerequisite

• python3
• docker
• docker-compose
• git

Build

Clone the git repo locally.

git clone git@github.com:aparnamane/application-version-info.git

Change directory to the application directory

cd application-version-info

Build the image

docker-compose build

This will build the application image with aparnamane/app-version-info:abc57858585 tag.

As in the above command commit SHA is the static value from the .env file while running locally, but when run by TravisCI the file will be overwrite the commit SHA with actual last git commit SHA.

Test

Run the container using the image

docker run -d --name=aplication_version_info -p 5000:5000 aparnamane/app-version-info:abc57858585

Check the container is running

docker ps

Test the application with API call using curl

docker exec aplication_version_info curl -i http://0.0.0.0:5000/application-version-info/api/v1.0/version

This should get response in JSON format as:

  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100   157  100   157    0     0  78500      0 --:--:-- --:--:-- --:--:-- 78500
HTTP/1.0 200 OK
Content-Type: application/json
Content-Length: 157
Server: Werkzeug/0.16.0 Python/3.8.0
Date: Thu, 14 Nov 2019 04:40:48 GMT

{
  "myapplication": [
    {
      "description": "pre-interview technical test", 
      "lastcommitsha": "abc57858585", 
      "version": "1.0"
    }
  ]
}

Push

Login to docker registry before pushing the image.

Push the image to the image registry

docker-compose push

Image will be availabe in docker registry as:

Enhancement Production Readiness

To deploy the application in production/production like environment following statergies need to be considered.

• Use certified image by security team.
• Branching statergy need to implement.
• Deployment statergy need implement.
• Use third party tools for authentication, identity management.
• CI/CD tool should be managed internally.
• User should provided limited/previlaged access for all tools.
• Aplication packages and images should be saved and extracted from private repositories.