Litespeed servers = publishing fake benchmarks?

1

I don’t know if you guys use Reddit much but look at this recent discussion on Litespeed and they are allegedly creating anonymous-looking benchmarks that show Litespeed being faster than pretty much every single server stack and software in the world, in a rather dishonest fashion…

https://www.reddit.com/r/selfhosted/comments/f06vse/litespeed_servers_seem_like_a_marketing_scam_are/

I thought the public might be interested to know that although they tried hard to make the domain look like a neutral third-party it is actually managed by an employee at Litespeed Technologies:

  • Scam benchmark: http2benchmarkorg/results/benchmark-apache-caddy-h2o-litespeed-nginx-digitalocean.html
  • List of contributors: githubcom/http2benchmark/http2benchmark/graphs/contributors
  • The #1 top contributor work at: Litespeed Technologies

I’ve seen some websites on Litespeed that load fast, but it’s true that they are all shared servers so really no different from Apache2 or something where it might be fast sometimes and slow other times. It really depends on the way you have configured the server I think (and not overloading sites).

Anyway that being said, has anyone done a fair and reliable test that shows Litespeed “outperforms” a properly tuned Nginx + PHP-FPM stack with FastCGI Cache? I have never seen one and honestly I don’t think you can really compare these two stacks anyways… thoughts.

2 Likes
2

I honestly read LiteServer for a split second.

3

@eva2000 George is your man probably. He’s the guy that has the most benchmarks afaik :slight_smile:

I remember him saying that LSAPI was faster than PHP-FPM, regarding everything else I do not know.

Either way, LiteSpeed is worth it over Nginx + Apache imo, just because of the amount of headaches you’d save optimizing everything, making it work together and dealing with cPanel related software that wasn’t coded with Nginx in mind. LiteSpeed is plug and play with Nginx-like performance for shared environments.

Now outside of a shared environment, for personal stuff, Nginx is perfectly fine and actually better in my opinion just by the cheer amount of documentation and help articles that will help you in case you run into any issues.

1 Like
4

Who said anything about Nginx + Apache? No need for Apache if you run native Nginx. What headaches are you talking about? You assume everyone is running shared hosting and cPanel, idk why…

Comparing Litespeed to Nginx is absurd, they are in no way similar. Litespeed bundles its own cache system that emulates Varnish caching (and relies on janky htaccess… :face_vomiting:). Nginx has FastCGI Cache which is far superior when it comes to infinitely scaling and multiple hits per second on high-traffic sites, you don’t even need a benchmark to know that, just look at all the Alexa Top 1000 websites.

There are also no benchmarks showing LSAPI faster than PHP-FPM, hence this thread.

If you think documentation is the only advantage that Nginx has, mmm ok…

5

Human resource is cost much more than hardware and litespeed license afaik.

6

Not much to dislike with Litespeed.

7

LiteSpeed is mostly used in a shared environment, so I assumed you were referring to that. As I agree with you that there is no reason to use LiteSpeed for personal stuff unless you want the easyness of installing cPanel + Apache, performance similar to Nginx and dont care about the cost.

No it’s not, they are in the same market. People compare electric and gas cars all the time although they are different in many aspects. If two products can achieve the result you want, makes sense to compare which one can do it better, cheaper or faster.

If it doesn’t make sense for you… don’t buy it.

I’m not that much into LiteSpeed to know all it’s features and quirks, so I will assume what you are saying it’'s true.
Even so… have you ever ran any benchmarks when it comes to FastCGI performance vs LSCache? I’ve ran multiple on WordPress.
Both perform really well… to the point that you’re much more likely to run out of bandwidth first serving just HTML than any other resource.

LiteSpeed doesnt need to create a perfect product, they just need to create a product perfect for 99% of the market. Let the open-source get the other 1% that have reaaaally high traffic.

I love how you sound like Cenk Uygur, “JUST GOOGLE IT!”.

Correlation does not imply causation. Benchmarks are always necessary. As I said previously Nginx is fast, is open source, has many features and has a lot of documentation everywhere. If LiteSpeed performed 10% faster than Nginx those companies would (most likely) continue to use Nginx.

I’m not saying that Nginx is or not faster than LiteSpeed, I’m just saying that your premise is wrong.

What do you mean with janky?

I never said that. I gave one example that I consider to be the most important.

Why are you so mad about LiteSpeed’s marketing? No one is forcing you to buy it. Deceptive marketing has been around since ever. Take a deep breath, there are worse things going on in the world.

Having this said… I PERSONALLY USE NGINX AND I LOVE IT

2 Likes
8

Kind of old (2015), maybe George has newer benchmarks but I havent found them

What conclusions can be made ? From my experience using LiteSpeed Enterprise, OpenLiteSpeed, Apache and Nginx web servers, I already knew for best scalability and performance in high concurrency user loads, LiteSpeed’s LSAPI PHP implement will always win when it comes to non-cached PHP request serving.

9

Just because he’s the only guy on the internet mentioning the two servers doesn’t mean that it is a benchmark of any sort. Zero data shared, and nobody else claiming this…

Non-cached PHP requests, what website does this??

2015 is well before PHP 7+ also when Opcache is now default. No idea his environment, etc and if the websites he talks about using the same PHP scripts, or what… and LSAPI is only supporting Litespeed so you can’t really test directly anyways. Benchmarks would have to be very targeted at ONLY the PHP queries and try to replicate the rest of the server stack extremely similar as possible. Even then, it’s possible that the results were from the LS cache or something not even related, zero data is shared.

Nobody who says Litespeed is “faster” mentions any of these things!

10

You are the one making claims without sharing any data.

He literally shared all the data of his benchmark. The machine, the version of the software and their settings/configuration. You didnt even read it.

Uncached requests always happen even if you have cache enabled.

Found some of the George’s benchmarks from August 2019,
here is some reading material if you want to:
https://www.webhostingtalk.com/showthread.php?t=1775139&p=10167958#post10167958
https://www.webhostingtalk.com/showthread.php?t=1775139&p=10169869#post10169869
https://www.webhostingtalk.com/showthread.php?t=1775139&page=2&p=10169960#post10169960
https://community.centminmod.com/threads/is-lsapi-litespeed-faster-than-php-fpm-nginx-etc-or-not.19153/

You either benchmark it yourself and show your results or show the results other people got. You dont do either, you want us to take your opinion as a fact. Fuck off.
I wont reply any further. Too tired for this.

3 Likes
11

Keep it friendly, that’s not a request :slight_smile:

5 Likes
12

My personal and concrete albeit limited experience is that nginx and litespeed are about equal in terms of speed.

Both have this or that advantage (in terms of performance) over the other, e.g. litespeed’s faster PHP caching, but those advantages are largely theoretical because while both cache systems enhance performance the difference between them is ridiculous compared to the damage done by the pervert PHP monstrosity. Plus, for both the knowledge and experience of the users doing the config is by far more relevant than this or that technical advantage.
Oh and btw, unless you have an extremely complex and busy site even dumb multithreaded servers that are not event based are easily more than fast enough because even simple caching takes away typically 95+% of the servers workload.

@breakdan

I find it strange how rudely you treat @vovler who just tried to share his thoughts and experience. Looking at what (very little) information you provided I don’t see a basis for your behaviour.
Btw, the post you refer to (on reddit) clearly shows that that person is rather clueless and biased (e.g. the way (s)he describes lirespeeds caching), so I wouldn’t care too much about his opinion anyway.

3 Likes
13

Wrong.

You can run apache + LSAPI: Speed and Resource Improvements with EasyApache 4 and mod_lsapi | cPanel Blog

Note: Only on cPanel as they have a partnership with LiteSpeed.

14

My limited testing off it the performance of it is better than PHP-FPM.

1 Like
15

It surely is! :grin:

16

Or cloud linux and Da :slight_smile:

1 Like
17

A pointless post unless you are sharing details. I assume you are talking about Apache… strange how you worded it this way, it doesn’t sound like a confident statement.

No he didn’t and no they don’t, anyway its pointless (2015) claim because that is literally the only guy on the internet making this claim and his posts don’t include PHP-SPECIFIC benchmark data. If you don’t understand application-level benchmarking than stop clogging up this thread please…

Last time I will say it: you can’t benchmark an entire “stack” if you are claiming that only LSAPI is faster than PHP-FPM… you have to benchmark the application itself. Even if the claim is true, there are still no data suggesting that a Litespeed stack is faster than a Nginx LEMP stack!

Precisely. No data has been shared showing that LSAPI processes PHP queries faster, it is only theoretically faster based on the Litespeed environment and does not apply anywhere else.

Give me a break, the guy told me to F*ck Off and I never used that sort of language. I believe in sharing specific, verifiable data and not cruising around with vague claims and fake benchmarks.

Anyway as expected, a lot of angry cPanel guys on here…

I’m guessing most of the people here have never setup a LEMP server in fact.

18
19

Would’ve been a cool topic if passions were channeled into benchmarks instead of escalating feelings.

5 Likes
20

@eva2000 asked us to post a link to a relevant thread :slight_smile:

https://community.centminmod.com/threads/is-lsapi-litespeed-faster-than-php-fpm-nginx-etc-or-not.19153/

For those that don’t know, @eva2000 is the creator of the incredibly performant CentminMod.

5 Likes