Decentralized Container Orchestration System

This project implements a decentralized container orchestration system that uses:

• Hyperswarm DHT for peer-to-peer node discovery (all nodes join a common hash-based topic).
• PBFT Consensus for fault-tolerant agreement on state changes (task assignments and failure handling).
• Hyperbee (Autobee CRDT) for a leaderless, multi-writer shared state store. Autobase + Hyperbee
• Dockerode for executing tasks within isolated Docker containers.
• Ed25519 Signatures (Keypear) for secure, authenticated inter-node communication.

Features

• Decentralized Architecture: No single master; every node is equal in scheduling and state replication.
• Fault Tolerance: PBFT consensus ensures correct decisions even if some nodes fail or act maliciously.
• Automatic Task Assignment & Reassignment: Tasks are scheduled via consensus and automatically reassigned if a node goes down.
• Secure Communication: All messages are digitally signed using Ed25519; Hyperswarm provides encrypted channels.
• Shared State: Cluster state is maintained in Hyperbee using a multi-writer CRDT approach.
• Automated Cleanup: Docker containers auto-remove upon exit, and the deploy script cleans up all processes.
• Single Script Deployment: deploy.js reads cluster-config.yaml and spins up all nodes.

Repository Structure

. ├── deploy.js # Deployment script for launching nodes ├── cluster-config.yaml # Cluster configuration (swarm name, node count, ports, cleanup flag) ├── package.json # Dependencies and project metadata ├── README.md # This documentation file ├── server │ └── index.js # Orchestrator node: networking, consensus, scheduling, and state sync └── worker └── worker.js # Worker process that runs tasks in Docker containers

markdown

Edit

Requirements

• Node.js (>=14)
• Docker (running locally and accessible via /var/run/docker.sock)
• NPM

Setup and Deployment

1. Clone the Repository:

   git clone https://your-repo-url.git
   cd decentralized-container-orchestration

Install Dependencies:

bash

Edit npm install Configure the Cluster: Edit cluster-config.yaml if needed. The sample configuration:

swarm: test-name
nodes:
  count: 10
  httpPort: 8080
  wsPort: 8585
  randomPortRange:
    min: 3000
    max: 10000
cleanup: true
Deploy the Cluster: Start the deployment script:

npm start

This spawns the specified number of node processes. Each node joins the common swarm, participates in consensus, and periodically simulates task assignments. This project is licensed under the Apache-2.0 License.

---

Final Notes

This codebase is detailed and “production‑inspired” but remains a prototype. In a production system you would:

• Use persistent storage for Hyperbee (e.g., using random-access-file instead of RAM).
• Secure key management (avoid storing private keys in plain text).
• Enhance the PBFT consensus implementation for a larger number of nodes and more rigorous fault tolerance.
• Extend failure detection, logging, and monitoring.
• Optionally expose HTTP or WebSocket endpoints for external APIs.

You can now deploy the entire system by running npm install and then npm start from the project root. Each node process will self-organize, accept simulated tasks, run them in Docker containers, and update shared state via Hyperbee—all while communicating securely over Hyperswarm.

Feel free to adjust, extend, or integrate further features as required.