First of all, let’s take a look at how a web server works. When someone makes a request to open a webpage, the browser contacts the server of that website. Then, the server looks for the requested files for the page and sends them to the browser. This is only the simplest kind of request.

The example above is also considered as a single thread. Traditional web servers create a single thread for every request, but NGINX does not work that way. As stated before, it performs with an asynchronous, event-driven architecture. It means that similar threads are managed under one worker process, and each worker process contains smaller units called worker connections. This whole unit is then responsible for handling request threads. Worker connections deliver the requests to a worker process, which will also send it to the master process. Finally, the master process provides the result of those requests.

That may sound simple, but one worker connection can take care of up to 1024 similar requests. Because of that, it can process thousands of requests without any difficulties. This is also the reason why it became excellent for busy websites like e-commerce, search engines, and cloud storage.