🚀 Rust Order Book

A real-time, concurrent, fault-tolerant Order Book built in Rust.
Supports order matching, order insert/cancel, WebSocket API, structured logging, stress testing, and benchmarking.
Designed with Rust best practices, SOLID principles, security, and fault tolerance.

---

📦 Architecture

• src/main.rs — WebSocket server bootstrap & logging initialization
• src/order.rs — Order model and Side enum (Buy/Sell)
• src/order_book.rs — core matching engine (insert, cancel, match; price/size book)
• src/messages.rs — JSON commands: Insert { order }, Cancel { id, price, side }
• src/ws.rs — WebSocket routing and per-connection async session handling
• src/bin/client_test.rs — load/stress client (simulates inserts/cancels)
• benches/ — Criterion benches (insert, cancel, match, integration)

---

🛠️ How to Run

git clone <repo>
cd order_book

# start server (WebSocket on ws://127.0.0.1:3000/ws)
cargo run --bin ord_book

# run simulated clients (optional)
cargo run --bin client_test

---

📡 WebSocket API

// Insert order
{ "action": "insert", "order": { "id": 1, "price": 101, "quantity": 10, "side": "Buy" } }

// Cancel order
{ "action": "cancel", "id": 1, "price": 101, "side": "Buy" }

---

🧪 Testing

cargo test                 # run all tests
cargo test -- --nocapture  # show println!/tracing output
# examples:
#   cargo test partial_match
#   cargo test -- --list

• Unit tests: insert / cancel / match logic
• Integration tests: multi-threaded insert + cancel behavior
• Load tests: client_test simulating concurrent clients

---

📊 Benchmarking (Criterion)

cargo bench                                # run all benches
cargo bench --bench insert                 # run only the insert bench file
cargo bench --bench insert -- --list       # list benchmarks in the file
cargo bench --bench insert -- --bench "insert 1M orders"  # run by exact name

Reports: target/criterion/<bench_id>/report/index.html (SVG/HTML charts).
Example (1M orders, indicative): Insert ~35–40 ms • Cancel ~12 ms • Match ~40 ms

---

🧠 Concepts & Patterns

• Concurrency: async/await on Tokio; one async task per WebSocket connection
• Shared state: Arc<Mutex<OrderBook>> (short lock scopes)
• Deterministic structure: BTreeMap for ordered price levels (price-time priority)
• Serialization: serde for JSON I/O
• Observability: tracing with daily-rotating file logs (ANSI disabled for clean files)
• Design: Command pattern (Insert/Cancel), actor-like sessions, repository-style mutation boundary, SOLID module boundaries

---

🔒 Security & Fault Tolerance

• Validate incoming JSON commands and side values (Buy/Sell)
• Isolate each client session; a misbehaving client won’t affect others
• Bound message sizes and consider timeouts/rate limits if exposed publicly
• Avoid long-held locks; keep OrderBook critical sections minimal
• Fail-fast on unrecoverable states; log structured context for analysis

---

✅ Features

• Price priority matching (best price wins)
• Partial fills; limit order fallback when no match
• Cancel by id/price/side
• Structured logs for inserts, cancels, and matches
• Concurrent client handling via WebSockets
• Benchmarked core operations (Criterion)

---

🧭 Roadmap

• Live visualization (React or TUI)
• Persistence (Postgres/SQLite) for orders & trades
• Broadcast matched trades to all clients
• Sharded/lock-free design for high-throughput (HFT) workloads