DevOps Sandbox Platform

A self-service platform for spinning up isolated temporary environments, deploying apps, simulating outages, monitoring health, and auto-destroying everything. Think miniature internal Heroku with a chaos engineering toggle.

---

Architecture

+--------------------------------------------------+
|            Linux VM (Single Host)                |
|                                                  |
|  User --HTTP:80--> [ Nginx Container ]           |
|                           |                      |
|              +------------+------------+         |
|              |                         |         |
|   [ net-env-abc / app-env-abc ]  [ net-env-xyz ] |
|   [ demo-app:latest           ]  [ demo-app    ] |
|   [ GET /    GET /health      ]                  |
|                                                  |
|  User --HTTP:8080--> [ Flask API Container ]     |
|                      wraps all bash scripts      |
|                                                  |
|  [ Cleanup Daemon ]    [ Health Poller ]         |
|    every 60s               every 30s             |
|    auto-destroys           marks degraded        |
|                                                  |
|  envs/*.json   logs/   nginx/conf.d/             |
+--------------------------------------------------+

---

Prerequisites

• Linux VM (Ubuntu 22.04+)
• Docker Engine 24+
• Docker Compose plugin (docker compose)
• jq, curl, make
• Port 80 and 8080 open in security group

sudo apt-get update && sudo apt-get install -y jq curl make

---

Quick Start

git clone https://github.com/asanteedith/devops-sandbox.git
cd devops-sandbox
docker build -t demo-app:latest ./demo-app
cp .env.example .env
make up
make create

---

Full Demo Walkthrough

# Start platform
make up

# Create environment
make create

# Test it
curl -H 'Host: <env-id>.localhost' http://localhost:80/
curl -H 'Host: <env-id>.localhost' http://localhost:80/health

# Check health
make health

# Simulate crash
make simulate ENV=<env-id> MODE=crash

# Recover
make simulate ENV=<env-id> MODE=recover

# Auto-destroy demo
bash platform/create_env.sh shortlived 70
tail -f logs/cleanup.log

# API
curl -s http://localhost:8080/envs | jq .

# Shut down
make down

---

Makefile Reference

Command	Description
make up	Start Nginx, API, daemon, poller
make down	Stop everything
make create	Create new env
make destroy ENV=<id>	Destroy specific env
make logs ENV=<id>	Tail env logs
make health	Show all env health
make simulate ENV=<id> MODE=<mode>	Run outage simulation
make clean	Wipe all state and logs

---

API Reference

Method	Endpoint	Description
POST	/envs	Create env
GET	/envs	List envs + TTL
DELETE	/envs/:id	Destroy env
GET	/envs/:id/logs	Last 100 log lines
GET	/envs/:id/health	Last 10 health checks
POST	/envs/:id/outage	Trigger simulation

---

Outage Modes

Mode	Effect
crash	Kills container immediately
pause	Freezes all processes
network	Disconnects from network
recover	Restores whatever was broken
stress	CPU spike for 60s

---

Known Limitations

• Single VM only
• Host-header routing — real DNS needed for browser access
• No API authentication
• Flask dev server — use Gunicorn for production
• No log rotation