Litespeed Webserver, the latest addition to the hosting landscape, has been met with growing enthusiasm by numerous hosting companies and systems engineers over the past year. Litespeed is presented as the definitive solution to maximize performance, but the attraction for a system that promises excellence in exchange for a monthly subscription is perhaps not symptomatic of a lack of technical expertise on the part of hosting companies and self-styled systems engineers who should instead have any?
It would be interesting to hypothesize which direction the aforementioned hosting companies would have taken without the help of Litespeed, and above all what they sold to their customers up to a year ago. There is a real risk that they would have gotten lost in the quicksand of incompetence. However, with Litespeed, even those with limited knowledge of systems engineering can achieve very advanced results. It's a similar concept to the iPhone camera: it allows you to get high-quality shots without necessarily having to master technical settings such as ISO or white balance; however, on specific situations and demanding customers they will never be able to compete with a professional optical back and a photographer with a theoretical background worthy of respect and years of practice on high performance systems.
It is important to note that for the end customer the difference between a semi-professional system like Litespeed and a combination of higher-level technologies such as Varnish/NGINX, used by major organizations and projects such as Wikipedia or the New York Times, it may not be immediately apparent. Litespeed, in fact, does not offer the same level of customization and technical detail present in the VCL language of Varnish or in the sophisticated configurations possible thanks to the experience of an expert system administrator.
Certainly, Litespeed offers some enticing features like caching, compression, and support for WebP images. What about the brotli squeeze? Impressive, for sure. However, when comparing Litespeed to a standard Apache webserver or a basic NGINX webserver without advanced features like FastCGI Cache or Proxy Cache, it might look like a miracle solution. But are we really ready to set Litespeed as the standard for high-traffic, high-speed sites?
Litespeed is unable to compete with flagship solutions such as NGINX and Varnish, which are adopted by web giants such as Wikipedia, ANSA or the New York Times. These organizations have well-defined needs and are not satisfied with compromises, preferring superior solutions over products like Litespeed.
It is also essential to look critically at the benchmarks proposed by Litespeed. Even if these data are not false as some have suggested, they appear to present a partial picture. They don't take into account optimal NGINX configurations, such as enabling FastCGI Cache, nor do they provide an appropriate comparison to an enterprise-grade caching system like Varnish that isn't factored into their benchmarks at all.
In fact, their benchmarks could be perceived as partial and not entirely representative of reality and above all absolutely improbable as already pointed out by the Reddit community who have widely criticized the proposed benchmarks deriving from their technical support: https://www.reddit.com/r/selfhosted/comments/f06vse/litespeed_servers_seem_like_a_marketing_scam_are/
Regardless of the opinions we can find on Reddit however, market opinion does not seem to confirm Litespeed as the undisputed leader. An analysis of SimilarTech statistics would show that the number of sites in the top 10000 globally that actually use Litespeed is not that high.
In the analysis of the graph above “Varnish VS Litespeed“, it is evident that Varnish is used by almost 9% of the top 10.000 sites globally, a remarkable number. However, it is essential to note that SimilarTech's statistics are based solely on HTTP headers found during its scans, and not all Varnish installations use easily identifiable standard HTTP headers. This means that, in reality, there may be many more Varnish installations in the top 10.000 sites than indicated.
By contrast, LiteSpeed's statistics tend to be more accurate. In fact, LiteSpeed Webserver has a tendency to add specific and easily recognizable headers, making it easier for scanning tools like SimilarTech to identify it.
One more thing to consider is that Varnish does not terminate the HTTPS connection, which can further skew the data. To get a more accurate view of the relative use of these technologies, it would therefore be more appropriate to compare NGINX and LiteSpeed, as illustrated in the chart below.
Analyzing the graphNGINX VS Litespeed“, it is remarkably evident that NGINX dominates the scene by a large margin, accounting for 25% of installs on the top 10.000 websites globally. This contrasts with Litespeed's modest 1,37%, which is practically insignificant from a statistical point of view. In contexts where optimal performance is of vital importance, such as those represented by the top 10.000 sites in the world, the choice of using NGINX as a web server is a clear demonstration of its robustness, flexibility and its advanced customization capabilities. The limited use of Litespeed in this context underscores its relative positioning as a least favored solution for high performance implementations.
To further support and consolidate the conclusions reached so far, we can refer to the data provided by SimilarTech to identify the most relevant sites that use NGINX and LiteSpeed respectively as webserver technologies. When we look at the top 500 sites globally, we enter a context where budget constraints are no barrier to adopting solutions like LiteSpeed Web Server, should they be considered superior.
We can see from the graph below that at least 10 sites among the top 90 websites worldwide use NGINX as their WebServer, one of which ranks eighth in terms of traffic.
Instead, we find that Litespeed is used by practically anonymous sites starting from the 331st place with certainly interesting traffic of 142 million visits per month, but greatly reduced compared, for example, to the 2,8 billion of the eighth site in the world by traffic. managed by NGINX WebServer.
We must necessarily consider that if, in this context, the use of LiteSpeed remains marginal compared to NGINX, we must interpret this trend as a demonstration of the superiority of the latter in terms of performance, flexibility and customization capabilities. Organizations running websites of this size are looking for the best technologies available, and if they opt for NGINX over LiteSpeed, this reflects a technical assessment that NGINX is deemed better suited to meet their high-level needs .
Therefore, Litespeed's benchmarks shouldn't be considered as the only reference nor should they be the promises of hosting companies that for obvious reasons try to sell their solution that they know how to use. There are multiple aspects to consider that Litespeed seems to leave out.
Costs of Litespeed licenses passed on to the end customer.
Unlike NGINX and Varnish, LiteSpeed Web Server incurs significant licensing costs, which are inevitably reflected on the end customer. For example, it is common to find that the cost of the LiteSpeed Enterprise software license alone exceeds that of the dedicated server itself. This represents for the hosting provider and for the end customer, an extra expense that could be avoided by adopting complex solutions such as NGINX and Varnish, if properly configured. In a practical example, on a dedicated server, one could save around 1300 euros a year, including taxes, on the cost of the software license.
This raises a fundamental question: what are the reasons why a hosting provider would decide to invest such large sums for the purchase of licenses that could be superfluous, if they had the necessary system skills to configure an NGINX + Varnish software stack? After all, these are the same technologies adopted by major companies like Wikipedia or the New York Times. Presumably, the answer lies in a combination of LiteSpeed's ease of use and a lack of skills needed to effectively implement and manage a solution based on NGINX and Varnish.
Wikipedia, one of the most visited online resources in the world, uses a highly customized and optimized software stack to handle its massive volume of traffic and requests. The heart of the stack is based on Linux as an operating system, with web servers mainly run by NGINX, famous for its efficiency and scalability.
Also, Wikipedia relies on Varnish for HTTP caching, an extremely effective solution to accelerate the speed of content delivery. Varnish is known for its highly detailed configuration capabilities, allowing Wikipedia to customize the cache to fit the unique structure and requirements of its content.
Regarding value, the question of cost is relevant. While Litespeed may seem affordable at first glance, a $90 monthly license per server might make you reconsider. PWhy opt for a product that is unable to compete with its similarly priced rivals? Why not invest in enterprise-grade solutions like NGINX and Varnish, which offer superior performance and customization without incurring high monthly license fees?
In conclusion, while Litespeed may appear attractive to those looking for an “easy” option, those looking for exceptional performance and advanced customization should consider enterprise-grade solutions such as NGINX and Varnish. We shouldn't settle for less when it comes to technology, because we deserve the best.