How Nginx Handles Millions of Requests: Event‑Driven Model and Performance Tuning

Nginx is a critical middleware for large‑scale architectures and a must‑know skill in big‑tech companies.

Event‑Driven Model

The event‑driven model is the key to Nginx’s high performance and ability to handle massive concurrency.

In a traditional thread model each request gets its own thread or process, which consumes a lot of memory. For example, handling 10,000 concurrent connections would require creating and destroying 10,000 threads, using roughly 10 GB of memory—unacceptable for most servers.

Nginx typically starts a small number of worker processes (usually equal to the number of CPU cores). Each worker can handle thousands of connections without creating a separate thread per connection.

Asynchronous Non‑Blocking I/O

Nginx uses an asynchronous non‑blocking I/O model, avoiding the blocking that occurs while waiting for I/O operations to complete.

When a connection initiates an I/O operation, the worker process does not block; it continues processing other events until the operating system notifies that the I/O is complete.

I/O Multiplexing

Asynchronous non‑blocking I/O is usually combined with I/O multiplexing techniques such as epoll, select, and poll. epoll has no limit on the number of file descriptors and does not need to traverse the entire descriptor set, so its performance does not degrade with increasing concurrency.

Therefore, most high‑concurrency deployments use epoll for I/O multiplexing.

Nginx Performance Optimization

Properly configuring Nginx’s cache mechanism can dramatically reduce backend load and improve response speed.

Static resources such as images, CSS, and JavaScript can be cached to boost site performance.

# Set cache time
location ~* \.(jpg|jpeg|png|gif|ico|css|js)$ {
    expires 30d;
    add_header Cache-Control "public, max-age=2592000";
}

Proxy caching can be used when Nginx acts as a reverse proxy, caching backend responses to relieve backend pressure.

proxy_cache_path /var/cache/nginx levels=1:2 keys_zone=my_cache:10m max_size=10g inactive=60m;
server {
    location / {
        proxy_cache my_cache;
        proxy_cache_key "$scheme$host$request_uri";
        proxy_cache_valid 200 301 302 10m;
        proxy_cache_valid 404 1m;
        proxy_pass http://backend_server;
    }
}

Beyond these settings, additional Nginx configuration tweaks can further optimize performance for high‑concurrency scenarios.