ShubhamRasal / django_todo

Requirements

• Install Python
• Install awscli (pip install botocore boto3 awscli)
• Configure aws cli (aws configure)

Problem Statement

---

Application Provisioning

• Create a simple Django/RoR rest app backed by RDS for production and docker-compose for local testing.
• Containerize the application using docker.
• Use make files to run tests, create docker images and push to docker registry.
• Use Ansible's dynamic inventory to provision the application on three service boxes.
• The application should be reachable behind ALB.

Build and Deploy Pipeline

Use the Jenkins setup to create a multi-branch CI/CD pipeline.

Create a declarative CI pipeline with the following stages:

• Lint code
• Run tests
• Build Docker image for the application
• Push to the docker registry

Re-use the make file created earlier in pipeline stages.

Extend the CD pipeline

• Deploy on AWS Infrastructure via Jenkins using Ansible.

---

---

Directory Structure

├── ansible                             # contains ansible playbook for deployment over aws ec2 instances behind ELB

├── docker                             #  contains docker files for diff environments such as local, dev and release

├── Jenkinsfile                       # Jenkins pipeline

├── Makefile                           # contains commands for test, build, publish and deploy

├── scripts                              # scripts used by docker container

├── src                                   # django application source code

└── venv                                # virtual environement

---

Quick Start

1. Build base docker image

• Change path to base Dockerfile

    $ cd docker/base

•  Build base docker image (use your account/registry name)

  $ docker build -t shubhamrasal/python-base:latest .

• Publish docker image to docker hub

$ docker push shubhamrasal/python-base:latest

2. Run unit test

$ make test

above command uses docker/dev/docker-compose.yml and docker/dev/Dockerfile and setup testing environment to run tests.

It will create reports folder in root directory with unit test and coverage report.

3. Generate

$ make build

This command will generate python wheels in target folder of our application using docker/docker-compose.yml file.

Also, it will create docker image with latest python wheels and tag the image with timestamp(default) using docker/release/Dockerfile.

4. Create docker image and publish to registry

$ export DOCKER_REGISTRY=<your registry url eg.docker.io>

$ make publish

above command will push latest docker image (created in build stage) to registry.

Note: Considering you already logged in to docker registry using $ docker login

5. Deploy App to EC2 servers behind ALB

Update ansible/app_vars.yml file. Add your values to given parametes.

aws_target_group: "<aws ec2 instance target group>"
service_container_name: "< container name>"
service_container_published_port: "< port on which service is deployed(you can find it in docker/release/Dockerfile) >"
host_container_published_port: "<port of ec2 instance from where your service will be accessible>"
database_name: <database name>
database_port: <database port eg. 5432 default port for postgress >

Add database variables in db_vars.yml

$ echo "yourpassword" > password.txt

$ chmod 700 password.txt

$ ansible-vault create db_vars.yml --vault-pass-file=password.txt

This command will open your default command line editor.

Add following variables into it.

database_user: <db user name>
database_password: < database password >
database_url: <database host url>

Save and close.

Note: do not add password.txt file to git. (keep it safe)

RUN Deploy command to deploy our docker container.

$ make deploy

This command will deregister EC2 instance from target group of loadbalancer, deploy our latest image container and if it is successful it will register again to target group.

Note: Update ansible/inventory/inventory_aws_ec2.yml file for selecting ec2 instances based on tags. In my case my ec2 instances are taged as 'server'.

To run this command, we need to configure awscli.

6. Clean the environment

$ make clean

It will delete all the container created while test and build phase.