unethical information on the litespeed website

#1
hello,

I didn't know where to put this post apologies in advance

i'm using litespeed like 2 years ago in a lot of servers and is a good web server, but
in litespeed website I think it shows misleading information:

check.png

in this comparison of webservers with wordpress show litespeed as an incredibly optimal webserver because yes but the tests performed are not fair:

check2.png
in tests show litespeed with wordpress cache (not installed default) and nginx and apache with sad fastcgi cache and w3totalcache,

why nginx without w3totalcache (like apache)
why nginx without nginx cache or varnish cache (not installed for default like litespeed equally easy to implement)
why apache without other cache (like litespeed show)

this gave me trouble at the beginning of using litespeed sell me something that doesn't exist,
litespeed It is a fantastic webserver but sometimes it seems that I am using the product of the devil
I understand that it is a company but it is very creeping without understanding at all because that should be so
If someone can explain to me why this is so and I do not know, I would be delighted

thanks,
 

mistwang

LiteSpeed Staff
#2
Thank you for the feedback.

We certainly want to compare the performance using the best available cache system for each web server.
Nginx FCGI cache is a server based cache, similar to litespeed cache, which is server based cache, people always believe Nginx FCGI cache is the fastest cache for Nginx, and want an Apple to Apple comparison, that's the reason why we choose that. Nginx + w3total cache is likely slower.
Due to the fact that we are comparing HTTP/2 performance, Varnish does not support HTTP/2 itself, setup will be complicate when another layer of front-end proxy involved, and it will further slow down the performance.

Running the benchmark on your own server wont take much time. using the open source project

https://github.com/http2benchmark/http2benchmark

BTW: all server configurations are in that project, feel free to inspect, tune the configuration if you like.

If you are good at setting things up, and truly believe it could change the outcome of the benchmark test, you can contribute a test setup to that project. We'd be happy to verify the results and update our published results. :)
 
Top