HTTP-Torrent

A simple P2P BitTorrent-like file sharing system implemented with HTTP protocol and Distributed Hash Table (DHT) support.

Technologies Used

• Python 3: Core programming language
• Flask: Web framework for the metrics server
• HTTP Server: Native Python http.server for client and tracker communication
• Kademlia: Implementation of the Kademlia distributed hash table for P2P functionality
• TailwindCSS: Frontend styling for the metrics dashboard

Components

1. Torrentify

Tool for generating torrent metadata files from source files.

2. Tracker

Central coordinator that maintains a registry of which clients have which file chunks.

3. Client

P2P client that downloads chunks from peers and shares them with others.

4. DHT (Distributed Hash Table)

Distributed system for storing chunk location information, providing redundancy when the tracker is unavailable.

5. Metrics Server

Dashboard for monitoring client activity, download/upload progress, and system health.

Communication Protocol

Tracker-Client Communication

Update Tracker with Available Chunks

PUT /chunk

Request Body:

{
  "client_host": "123.42.23.12",
  "client_port": 8000,
  "hashes": [
    "oc3kikx1o9x1oxo21k1x",
    "xyz123abc456def789",
    ...
  ]
}

Response:

Ok

Get Peers with Required Chunks

POST /chunk

Request Body:

[
  "oc3kikx1o9x1oxo21k1x",
  "xyz123abc456def789",
  ...
]

Response:

[
  {
    "client_host": "123.42.23.12",
    "client_port": 8000,
    "hash_list": [
      "oc3kikx1o9x1oxo21k1x",
      ...
    ]
  },
  {
    "client_host": "189.45.67.89",
    "client_port": 8001,
    "hash_list": [
      "xyz123abc456def789",
      ...
    ]
  },
  ...
]

Register DHT Node

POST /dht

Request Body:

{
  "host": "123.42.23.12",
  "port": 8468
}

Response:

Ok

Get DHT Nodes

GET /dht

Response:

[
  {
    "host": "123.42.23.12",
    "port": 8468
  },
  {
    "host": "189.45.67.89",
    "port": 8469
  },
  ...
]

Client-Client Communication

Request Chunks

POST /chunk

Request Body:

[
  "oc3kikx1o9x1oxo21k1x",
  "xyz123abc456def789",
  ...
]

Response:

[
  {
    "hash": "oc3kikx1o9x1oxo21k1x",
    "content": "base64_encoded_content_here"
  },
  {
    "hash": "xyz123abc456def789",
    "content": "base64_encoded_content_here"
  },
  ...
]

Health Check

GET /ping

Response:

pong

DHT Network Messages

The DHT implementation uses the Kademlia protocol with custom serialization for key-value storage:

Key-Value Operations

• Keys: File chunk hashes (e.g., "oc3kikx1o9x1oxo21k1x")
• Values: JSON stringified arrays of client addresses (e.g., ["123.42.23.12:8000", "189.45.67.89:8001"])

DHT Communication Flow

1. Bootstrap: New nodes contact known peers (from tracker) to join the network
2. Store: When a client has a chunk, it stores chunk_hash -> ["ip:port"] in the DHT
3. Retrieve: When a client needs a chunk, it queries the DHT for chunk_hash to get a list of peers
4. Update: When a client gets a new chunk, it adds itself to the list of providers for that chunk

Metrics Server

The metrics server collects and displays real-time information about all active clients including:

• Download/upload progress
• Number of chunks downloaded/uploaded
• Total file size
• DHT status (enabled/disabled)
• Number of DHT peers
• Tracker connectivity status
• Upload/download ratio

Metrics are sent to the server every 3 seconds and displayed in a real-time dashboard at http://localhost:8080/.

Usage Instructions

1. Install Dependencies

pip install -r requirements.txt

2. Start the Tracker

python tracker.py [port_number]

Default port is 5000 if not specified.

3. Generate a Torrent File

python torrentify.py <your_file> <tracker_ip:port>

Example:

python torrentify.py example.mp4 localhost:5000

This generates a torrent.json file containing metadata about the file chunks. Use the tracker information that is output from the previous command.

4. Start the Metrics Server

python metric_server.py

Access the dashboard at http://localhost:8080

5. Start Clients

To start a client with the source file (seeder):

python client.py has_file [dht_enabled]

To start a client without the source file (leecher):

python client.py [dht_enabled]

Add the dht_enabled parameter to enable DHT functionality incase when tracker is down.

6. Run a Stress Test

python stress_test.py

This will automatically start multiple clients and simulate a P2P network environment.

File Structure

• client.py: Implementation of the torrent client
• dht.py: Kademlia-based distributed hash table implementation
• metric_server.py: Monitoring dashboard server
• torrentify.py: Utility to create torrent metadata files
• tracker.py: Central tracker server implementation
• util.py: Helper functions for network operations
• templates/dashboard.html: Metrics visualization dashboard

Fault Tolerance

The system is designed with redundancy in mind:

1. If the tracker is offline, clients can continue to discover peers through the DHT network
2. If a peer disconnects, clients automatically find alternative sources
3. Health checks remove unavailable peers from the tracker registry

License

This project is licensed under the terms of the license included in the LICENSE file.