jvlxz / octopus

Octopus Project

Octopus is an Ansible-based project to deploy a poll application across multiple machines without using containers. This application consists of a Python Flask web client, a Redis queue, a Java worker, and a Node.js web client that displays the results.

TABLE OF CONTENTS

1. Getting Started
2. Features
3. Environment Setup
4. Repository Structure
5. Usage
6. Important Notes

Getting Started

To get started with the project, make sure you have the following prerequisites:

Ansible installed on your local machine Access to 5 Debian 10 (Buster) virtual machines A cloud platform account (optional, but recommended)

Features

The project includes 6 Ansible roles:

1. base: Configures all machines and installs useful packages.
   
2. redis: Installs and sets up Redis.
   
3. postgresql: Installs PostgreSQL 12, psql tool, and creates the necessary user and schema.
   
4. poll: Uploads poll service, installs dependencies, and runs the poll web client.
   
5. worker: Uploads worker service, installs dependencies, builds the worker, and runs the worker.
   
6. result: Uploads result service, installs dependencies, and runs the result web client.
   

Environment Setup

Your inventory should have 5 groups, each containing 1 instance: redis, postgres, poll, result, and worker. You will need 5 virtual machines based on Debian 10 (Buster). You can run it locally, but it is recommended to use a cloud platform. Do not spend too much credit, as you may need cloud platforms for future DevOps projects.

Repository Structure

.
├── playbook.yml
├── poll.tar
├── result.tar
├── worker.tar
├── group_vars
│   └── all.yml
└── roles
    ├── base
    │   └── tasks
    │       └── main.yml
    ├── postgresql
    │   ├── files
    │   │   ├── pg_hba.conf
    │   │   └── schema.sql
    │   └── tasks
    │       └── main.yml
    ├── redis
    │   ├── files
    │   │   └── redis.conf
    │   └── tasks
    │       └── main.yml
    ├── poll
    │   ├── files
    │   │   └── poll.service
    │   └── tasks
    │       └── main.yml
    ├── result
    │   ├── files
    │   │   └── result.service
    │   └── tasks
    │       └── main.yml
    └── worker
        ├── files
        │   └── worker.service
        └── tasks
            └── main.yml

Usage

• Set the ANSIBLE_VAULT_PASSWORD_FILE environment variable and store your vault password: :
  

export ANSIBLE_VAULT_PASSWORD_FILE=/tmp/.vault_pass
echo verySecretPassword > /tmp/.vault_pass

• Run the Ansible playbook using your own inventory file named production:
  

ansible-playbook -i production playbook.yml

Important Notes

• Docker, Podman, and Ansible Galaxy are forbidden for this project.
• Clear-text passwords found in the repository will result in the project being considered as failed. Use Ansible Vault to secure sensitive information.
• All services must be managed by systemd and start automatically on boot.
• Services must be configured with environment variables, such as: host, port, user, password, database name, etc.
• Idempotence should be maintained. After applying the Ansible playbook twice, there should be as few changed tasks as possible in your PLAY RECAP.
• Avoid using the command / shell / raw modules as much as possible. Use dedicated modules like apt_key, apt_repository, or pip. Use the command module only if no dedicated module exists.

(back to top)