🚀 High-Performance Multi-threaded Web Server with LRU Cache

A blazing-fast, multi-threaded HTTP web server implementation in C featuring an intelligent LRU cache system for optimal performance! 🔥

📋 Table of Contents

• ✨ Features
• 🏗️ Architecture
• 🚀 Quick Start
• 🔧 Usage
• 📊 Performance Testing
• 📈 Benchmarking
• 🛠️ Components
• 📦 Build System

✨ Features

• 🧵 Multi-threaded Architecture - Configurable thread pool for concurrent request handling
• 🗄️ LRU Cache System - Intelligent caching with customizable memory limits
• ⚡ High Performance - Optimized for speed and efficiency
• 📊 Built-in Benchmarking - Comprehensive performance testing tools
• 🔧 Configurable Parameters - Flexible server configuration
• 📈 Performance Visualization - Gnuplot integration for data visualization
• 🛡️ Robust Error Handling - Comprehensive error management and logging

⚠️ Known Issue: There's a current bug in the hashtable eviction logic that can cause linked list corruption when removing the first node in a hash chain during cache eviction.

🏗️ Architecture

┌─────────────────┐    ┌─────────────────┐    ┌─────────────────┐
│   HTTP Client   │────│   Web Server    │────│   File System   │
└─────────────────┘    └─────────────────┘    └─────────────────┘
                              │
                       ┌─────────────────┐
                       │   LRU Cache     │
                       │   (Hash Table)  │
                       └─────────────────┘

The server implements:

• Producer-Consumer Pattern with thread pool
• Hash Table + Doubly Linked List for O(1) LRU cache operations
• Non-blocking I/O with poll() for connection management

🚀 Quick Start

Prerequisites

• GCC compiler with pthread support
• Make build system
• POSIX-compliant system (Linux/Unix)

Build & Run

# Build the project
make

# Start the server
./server <port> <nr_threads> <max_requests> <max_cache_size>

# Example: Start server on port 8080 with 8 threads, 100 max requests, 1MB cache
./server 8080 8 100 1048576

🔧 Usage

Server Parameters

Parameter	Description	Example
port	Port number (≥1024)	8080
nr_threads	Number of worker threads	8
max_requests	Maximum concurrent requests	100
max_cache_size	Cache size in bytes	1048576

Client Testing

# Simple client test
./client_simple localhost 8080 /index.html

# Multi-threaded client benchmark
./client [-t] <host> <port> <nr_times> <nr_threads> <fileset>

📊 Performance Testing

Automated Experiments

# Test different thread counts
./run-experiment 8080

# Test various cache sizes
./run-cache-experiment 8080

# Single experiment run
./run-one-experiment <port> <threads> <requests> <cache_size> <fileset>

Generate Test Files

# Create file set for testing
./fileset -d fileset_dir

📈 Benchmarking

The project includes comprehensive benchmarking tools:

• 🧵 Thread Scaling: Tests performance with 0-32 threads
• 📊 Request Load: Varies concurrent request counts
• 🗄️ Cache Impact: Analyzes cache size effects (0-8MB)
• 📉 Visualization: Automatic Gnuplot chart generation

Output Files

• plot-threads.out - Thread performance data
• plot-requests.out - Request load data
• plot-cachesize.out - Cache performance data
• *.pdf - Generated performance charts

🛠️ Components

Core Files

File	Purpose	Description
server.c	🖥️ Main server	HTTP server entry point
server_thread.c	🧵 Thread management	Worker threads & cache logic
request.c	📝 Request handling	HTTP request processing
client.c	👤 Test client	Multi-threaded client for testing
common.c	🔧 Utilities	Shared utility functions

Cache Implementation

• Hash Table: O(1) lookup with chaining
• LRU List: Doubly-linked list for cache eviction
• Thread-Safe: Mutex protection for concurrent access

📦 Build System

# Available targets
make all        # Build all executables
make clean      # Remove build artifacts  
make realclean  # Deep clean including logs
make depend     # Generate dependencies
make tags       # Generate TAGS file

Dependencies

• pthread - POSIX threads
• popt - Command line parsing
• math - Mathematical functions

---

Happy Coding! 🎉 Feel free to contribute, report issues, or suggest improvements!