Web Servers: Looking at Options
Estimated Reading Time: 4 minutes
Last updated on 4th March 2026
Without web servers, we would never have websites and the web would not exist. While the machines are a variation on the PC theme, it is the software that does the job is what I want to highlight here. Microsoft's Internet Information Server (IIS) would be an example of a proprietary option, but what I wish to mention are its open-source alternatives, and it is these that largely dominate the market because they will work on operating systems like Linux or BSD UNIX.
Apache is more than just a web server because it is a whole collection of software projects that have their own open-source licences, which are more permissive than GPL (GNU Public Licence). One of these is HTTPD (HyperText Transport Protocol Daemon) and it is that which gains the Apache name here. My impression is that it was the first of a series of projects that now includes OpenOffice among their number as well as Tomcat, which is a Java server page platform that sounds as if it is based on HTTPD. There also is a story that the Apache name came from the nickname of "a patchy server" but I cannot confirm the truth of that.
As well as serving (X)HTML files, HTTPD usefully allows server-side web scripting using various languages such as Perl, PHP, Python and Ruby. With one of those, databases like MySQL, PostgreSQL or MariaDB can be used to power content management systems like WordPress and Drupal as well as other web applications. NoSQL databases probably are possibilities too. Client side scripting with the likes of JavaScript was made possible to complete the set, with CSS being used to make things look better. The basic platform certainly allows plenty of possibilities that were not imagined at its inception around twenty years ago.
Developed in 2003 and maintained by the Lighttpd Project, this open-source web server is designed for efficiency, prioritising low memory consumption and high concurrency to handle numerous connections with minimal resource use. Built on an event-driven architecture, it excels in environments such as embedded systems, small virtual servers and high-traffic websites, offering features like fast static file delivery, support for modern protocols including HTTP/2 and TLS/SSL and compatibility with dynamic content through FastCGI and similar interfaces. While its role has been largely overtaken by Nginx in high-performance deployments, it remains relevant in scenarios where simplicity and limited resource requirements are prioritised, often appearing in embedded devices, router interfaces and legacy infrastructure. Configuration is typically managed through a single file and its modular design allows for tailored setups, though its usage has diminished in modern Linux hosting stacks compared to alternatives like Apache or Nginx.
Nginx (pronounced Engine X) is a more recent upstart that has grown in favour because of its different way of handling multiple processes at the time. It appears to support the same technologies as Apache HTTPD, and I have had a go with it on a VirtualBox virtual machine running CentOS. So far, that has not got so far, but this experiment is set to continue. Notable users include WordPress.com and Fasthosts so it certainly is gaining acceptance. The perceived robustness of performance is the driver for that, though some websites that I have are not great demonstrations with more outages experienced than is ideal, one I have running on Apache appears to have less downtime. The offline server that I have appears to exert a greater load on a system than Apache too, but this needs further investigation because it may be caused by a suboptimal set-up due to a lack of experience on my part.