Nginx source code analysis (1)-Origin

Nginx is an open-source high-performance WEB server system. The event-driven request processing method and extremely demanding resource usage method make nginx a real high-performance server. The quality of nginx source code is also quite high, and the author has made a lotCodeHas created many wheels, such as the memory pool, buffer zone, String, linked list, red/black tree, and other classic data structures, event-driven models, HTTP parsing, and various seed processing modules, even the automatically compiled scripts are written by the author based on his own understanding. This makes nginx more efficient than other Web servers.

Nginx code is very delicate and compact. Although all the code has only 0.1 million lines, it is not inferior to Apache with hundreds of thousands of lines. However, the coupling between various parts is very powerful, and it is difficult to split the implementation of a part of it for use. It is difficult to analyze the source code of such a medium and large complex system, and it is difficult to find the entry point at the beginning, therefore, the goal and plan must be clearly defined first.

At first, I decided to do this to challenge myself and make my life more interesting. After reading this article for a few days, I thought it should not be so simple. There are too many attractive things, and it is worth planning for system learning and analysis. First of all, this system covers almost all the essential technologies for implementing high-performance servers, such as epoll, kqueue, master-workers, pool, buffer ......, it also covers many web service development technologies, such as SSI, SSL, proxy, Gzip, RegEx, load balancing, reconfiguration, hot code swapping ......, there are also some commonly used sophisticated data structure implementation, all things are very mainstream; secondly, the first-class code organization structure and clean and concise code style, especially the name of the entire system is just right, the readability is quite high and it is very kiss. This style is worth learning and imitating. The third is that by reading the source code, you can feel the rigorous style and excellent capabilities of the author, and you can increase your motivation, set an example.

On the other hand, to achieve these objectives, it is very difficult to develop detailed plans and adopt effective methods.

For such a large system, it is impossible to know all the details in one breath, and the implementation of each part of nginx is closely related, so it is impossible to have a glimpse of it, the proper approach seems to be to start from main, first understand the sequence of the nginx startup process, then break down the problem, and then analyze each important part one by one.

It is also important to record and organize the key parts of each understanding in detail, which is also the task to be completed in this source code analysis Log Series.

To better understand the key to code implementation, it is inevitable to modify the code and write some test cases. Therefore, we need to build a debugging environment, which is easier, the Linux system is a natural development and debugging environment.

My personal abilities are limited. Fortunately, a group of colleagues on the Internet are still tirelessly doing the same thing. They will also promote each other by reducing their detours, more in-depth and accurate understanding of the source code.

Start a pleasant trip, go!