This project implements a distributed MongoDB cluster with sharding and replication to manage large-scale public transport data.

It simulates a real-world transportation system handling:

• Millions of tickets
• Thousands of passengers and buses
• Multi-city data distribution

The system is designed to ensure:

• Scalability
• High availability
• Fault tolerance

The system is based on a MongoDB Sharded Cluster:

• 3 Shards (data distribution)
• Replica Sets (fault tolerance)
• Config Servers (metadata)
• Mongos Router (query routing)

1. Client sends query to mongos
2. Mongos routes query to relevant shard(s)
3. Each shard processes locally
4. Results are aggregated and returned

This folder contains sample datasets used for demonstration purposes.

• sample/ → small data samples for testing
• schema/ → data structure definitions

Full datasets (millions of records) are not included due to size constraints.

• Shard Key: ville
• Strategy: Range-based sharding

• High cardinality
• Frequent filtering by city
• Uniform distribution
• Stable field

• Distributed storage using MongoDB
• Horizontal scaling via sharding
• Replication for high availability
• Fault tolerance testing
• Analytical queries (aggregation)

✔ Reading continues even if a node fails
✔ Data remains accessible via secondary nodes
✔ Automatic failover supported

• Number of tickets per city
• Most used transport line
• Time-based filtering of tickets

• MongoDB (Sharding + Replication)
• Docker & Docker Compose
• PyMongo (optional integration)
• Python (for frontend visualization)

For more details about the project setup and implementation:

• Setup guide: see docs/setup-guide.md
• Full command list: see scripts/commands.txt

These resources provide step-by-step instructions and cluster configuration details.