What is a Web Server

1. The traditional web server as an individual, dedicated, and central computer
2. Web server operating systems overview
3. IPv6 as an efficient protocol for data transfer
4. Web server as a program - Apache and Nginx
5. Optimising web servers through caching

Both in colloquial language and in the IT industry, the term web server refers to two different levels. On one hand, it refers to a program like Apache or Nginx, whose task is to enable communication between a client in the form of a browser and a server as a host. Additionally, a dedicated server or VPS hosting is also referred to as a web server - in this case, the term web server encompasses operating systems like Linux or Microsoft Windows in conjunction with a suitable environment. They must in any case support addressing via IPv4 and IPv6 as well as protection against server failure. A special form of web server is found in complex and distributed networks like Content Delivery Networks (CDN), where requests are accepted decentrally via servers or VPS hosting, but content management, updating, and organisation are done centrally.

The traditional web server as an individual, dedicated, and central computer

In the traditional model, web server operating systems such as Microsoft Windows Server or Linux are used, on which necessary services for dynamic websites like Apache, Nginx, PHP, and MySQL are locally installed. This approach is suitable for small and medium-sized companies, as it enables a cost-effective and efficient infrastructure in conjunction with a VPS hosting or web space. In this case, a single dedicated server handles all the tasks of a web server. It also allows for flexible and gradual expansion by outsourcing certain services to a growing IT infrastructure. In the event of expansion at a later stage, load distribution on local web servers or VPS hosting, caching on DB servers within a CDN is possible.

Web servers require not only suitable web server operating systems but also powerful and modern hardware at the physical level. The workload and resources needed for timely processing of requests fluctuate throughout the day and seasonally - hence the importance of sufficient reserves in performance that can reliably handle spontaneous peaks in computing load. As a standalone, universal web server hosts various necessary services such as databases and an interpreter for scripting languages like PHP or Python, each active connection consumes multiple running processes, rapidly increasing the demand for performance. Certain techniques such as data and content caching can only partially alleviate this issue.

Overview of Web Server Operating Systems

The web server operating systems must meet specific requirements. For instance, long-term support for web server operating systems is crucial, as transitioning to a new version without server downtime is nearly impossible. Additionally, web server operating systems need to be regularly updated to patch new security vulnerabilities. A quick response from administrators to critical and potentially dangerous weaknesses is essential to prevent cyberattacks following their disclosure.

In general, web server operating systems such as Microsoft Windows Server or various Linux distributions are suitable choices, optimised for use as standalone web servers and offering high stability with fast and reliable support. Due to their direct connection to the public internet, web servers are at high risk of cyberattacks and server downtime, necessitating special and robust protection against external intrusion. For this reason, many companies opt for web server operating systems with commercial support - such as Red Hat Enterprise Linux (RHEL) in addition to Microsoft Windows Server - which supplements the free Fedora with professional, paid assistance. This approach minimises the risk of server downtime due to external factors and ensures continuous accessibility with a low server response time.

Due to their stability, security, and efficiency, specialised Linux distributions such as the mentioned RHEL or Debian have established themselves as web server operating systems and have achieved a leading position in the market. These web server operating systems based on the Linux kernel are divided into different distributions, each with their own priorities and characteristics. The individual features are less about the selection of utilities like the database and web server, but more about the operation and handling of third-party software that is not licensed under an open-source license. This concerns less commonly used programs like the Apache or Nginx web server, but primarily the support of hardware through proprietary drivers that enable features such as IPv6 addressing of network interfaces. Linux is highly modular, allowing for significant customisation of the web server operating systems. Options include varying degrees of server monitoring, where the administrator can set priorities.

The web server operating systems based on Microsoft Windows Server, on the other hand, are characterised by high user-friendliness, continuous server monitoring, and good documentation. This is complemented by a clear and step-by-step installation and configuration process, as well as excellent support for the web server, which Microsoft ensures with a professional IT security department and its own search for potential security vulnerabilities. Microsoft's web server operating systems have the additional advantage that they often explicitly warn against incorrect settings and avoid configurations that could lead to server downtime. Furthermore, these web server operating systems simplify optimisation, considering various aspects from resource management to content caching and the configuration of network protocols such as TCP and IPv6.

IPv6 as an efficient protocol for data transfer

The main task of IPv6 is to address a web server on the internet and establish a bidirectional connection between the web server and client. It is a protocol that operates across different layers, encapsulating important information for routing within a network and transitioning between different subnets. IPv6 was standardised in 1998 as its predecessor, IPv4, showed weaknesses over time and the available space of around 4.3 billion unique addresses would soon be exhausted. In contrast, IPv6 allows direct addressing of up to 340 undecillion different participants with their own identifier - a number with a total of 39 zeros.

Although web servers and operating systems have been supporting the transition from IPv4 to IPv6 since the turn of the millennium, many users still continue to use the old standard. One reason for this is that the implementation of IPv6 should be gradual, without a specific timeline or mandatory framework for implementation defined. Even in 2018, less than half of all connections used IPv6 for addressing, with a common practice in many parts of the internet being the parallel use of both IPv6 and IPv4. However, due to the significant increase in participants through embedded PCs and mobile devices, experts anticipate that IPv6 will rapidly evolve and establish itself as a universal standard. To ensure compatibility with older IT infrastructure, it remains necessary for web servers and operating systems to continue supporting IPv4 in the long term.

Web Servers as Software - Apache and Nginx

While Windows or Linux are options for the physical web server's operating system, the most commonly used alternatives for the virtual web server as software for bilateral communication between client and host are usually Apache or Nginx.

The Apache web server is considered a standard in many areas and has held a leading position with a market share of over 50% since the turn of the millennium. In comparison, the Nginx web server is a relatively young web server - the first version was released in 2004, and it only gained significant popularity since around 2010. Both servers differ significantly in their structure and characteristics. Some of the features of Apache include:

• Modular web server with flexible functionality extension
• Each connection is assigned its own process
• Support for extensions and plugins
• High functionality with good performance
• Efficiency decreases with increasing number of clients
• Support for local web server configuration through .htaccess

In contrast, Nginx is designed for large and complex networks with numerous clients, such as CDNs, which is why large portals often prefer this web server. Typical features of Nginx include:

• Static web server with limited expandability through modules
• Connections are bundled and collected in one process
• Plugins cannot be loaded without server downtime
• Excellent performance with numerous parallel connections
• Low resource consumption with increasing requests
• No support for local configuration through .htaccess

In addition to Apache and Nginx, the LiteSpeed web server has gained market share recently. On Windows servers, the IIS web server from Microsoft is also preferred.

Optimising Web Servers through Caching

For continuous accessibility, a web server must run 24/7/365 and be connected to the global Internet at all times. At the same time, the requirements and computational load fluctuate significantly, while the web server and web server operating systems can only influence these conditions to a limited extent. To keep response times low and avoid computations as much as possible, optimising web servers and web server operating systems is essential when dealing with a higher number of clients. One method for this is content caching, where the web server stores frequent requests in a cache after their initial processing. However, a disadvantage of caching is that it is primarily suitable for websites with static content - if individual elements need to be adjusted, for example after a login in a non-public area, a complete recalculation is necessary. Therefore, caching is primarily suitable for optimising a web presence and is limited to universal areas when using social media and individual requests.

Server Monitoring and Server Downtime

Due to their importance for public presentation, web servers require intensive care including server monitoring and reliable security. Any server downtime in this area - whether caused by the web server operating systems, hardware failure, or misconfiguration of a VPS hosting - immediately leads to a loss of public reputation and damages the brand building of a start-up. Servers and VPS hosting should therefore be monitored for server downtime or unexpected behaviour through frequent server monitoring. This monitoring includes checking the accessibility of web servers and operating systems as well as the functionality of relevant services like Apache or Nginx.

VPS hosting offers cost optimisation

Compared to a Dedicated Server, a VPS hosting has an excellent cost-benefit ratio and is therefore suitable as a primary instance or as a CDN web server with distributed locations. The only issues with a VPS hosting as a web server arise from heterogeneous web server operating systems that prohibit central data caching. For this reason, it is advisable to make an overarching decision for VPS hosting and IT infrastructure and to stick to this decision when it comes to VPS hosting.

In our web server overview, you can find detailed information on the current distribution of market shares of the most popular web servers.

Photo: kreatikar pixabay.com