A lightweight backend platform for automating cloud VM instance creation, tracking, and management via a RESTful API.

Watch a quick demo here 👉: my awsome video

CloudProvisioner is a simple server-side tool that allows users (or automated systems) to:

• Launch new cloud VM instances on demand
• Monitor instance creation status
• Store and track metadata in a cloud-hosted SQL database
• Retrieve server information through a lightweight REST API

It is similar in spirit to AWS EC2 instance management, but simplified for learning and educational purposes.

This project was completed for CSC 346: Cloud Computing at the University of Arizona (Spring 2025).

• Apache2 Server + CGI Scripts

  • Handles incoming requests
  • Talks to cloud provider API to create VMs
  • Connects to SQL database to store server metadata

• Cloud-hosted SQL Database

  • Stores VM instance metadata (id, owner, description, instance_id, ready status)

• Background Monitoring Script

  • Polls cloud provider API
  • Updates database when VM instance becomes "ready"

• Cloud Provider

  • Actual compute infrastructure for VMs: DigitalOcean

GET /api/servers

Returns JSON array of server objects with:

• id
• owner
• description
• public_ip (blank if not ready)

GET /api/servers/<id>

Returns JSON object for a specific server.

• SSH public key authentication is used to configure access to instances.
• No password-based logins are allowed to VM instances (mitigates brute-force attacks).
• Firewall rules restrict ports to necessary services only (e.g., 80 for web, 22 for SSH).

• Python 3 for CGI scripts
• Apache2 HTTP server
• MySQL cloud database (I use Railway)
• Cloud API SDKs (I use Digital Ocean)
• Bash scripting (for background monitoring)
• Linux/Ubuntu (typical VM OS)

1. Clone project files into your CGI directory (/usr/lib/cgi-bin/ if you use Ubuntu 22.04 (LTS) x64 version)
2. Set up environment variables or a config file for database credentials (never hardcode credentials into scripts).
3. Ensure Python3, MySQLdb, and necessary cloud CLI tools (e.g., doctl, aws, gcloud) are installed.
4. Deploy cloud SQL database according to provided schema (servers.txt).
5. Adjust CGI file permissions as needed (chmod 755).

• Cloud resource provisioning
• Backend REST API development
• Database integration with cloud services
• SSH key authentication for secure server access
• Basic cloud infrastructure automation
• System design thinking

Bao Nguyen
University of Arizona
Spring 2025 — CSC 346: Cloud Computing

• Support for terminating instances (DELETE endpoint)
• Add authentication