web servers...

Load limits

A web server (program) has defined load limits, because it can handle only a limited number of concurrent client connections (usually between 2 and 80,000, by default between 500 and 1,000) per IP address (and TCP port) and it can serve only a certain maximum number of requests per second depending on:

• its own settings;
• the HTTP request type;
• content origin (static or dynamic);
• the fact that the served content is or is not cached;
• the hardware and software limits of the OS where it is working.

When a web server is near to or over its limits, it becomes overloaded and thus unresponsive.

Kernel-mode and user-mode web servers

A web server can be either implemented into the OS kernel, or in user space(like other regular applications).

An in-kernel web server(like TUX on Linux or Microsoft IIS on Windows) will usually work faster because, as part of the system, it can directly use all the hardware resources it needs, such as:

• non-paged memory;
• CPU time-slices;
• network adapters buffers.

Web servers that run in user-mode have to ask the system the permission to use more memory or more CPU resources. Not only these requests to the kernel take time, but they are not always satisfied because the system reserves resources for its own usage and has the responsibility to share hardware resources with all the other running applications.

Also applications cannot access the system internal buffers, which is causing useless buffer copies that create another handicap for user-mode web servers. As a consequence, the only way for a user-mode web server to match kernel-mode performances is to raise the quality of its code to much higher standards than the code used into another web server that runs in the kernel.

This is more difficult under Windows than under Linux where the user-mode overhead is 6 times smaller than under Windows.