What's Your Favorite Web Server and Why?

What about python -m SimpleHTTPServer

It’s fine for very simple static stuff with no multithreading (as far as I know)

Apache + nginx is my preference but I’d settle on Apache simply for the fact that it’s a kitchen sink when you need it. Nginx is great but unfortunately it’s not as powerful, but I believe it’s about using the right tool for the job. No point using a spanner to drive nails if you have a hammer next to you.

I use Nginx but I wouldn’t say I love it. Recently tried out Caddy as an experiment and quite liked it.

Now do a little research on the primary author…

1 Like

NGINX for everything we use for internal projects, but Litespeed all the way.

With the resellers I threw out my nodes have been humming along and Litespeed couldn’t be happier.



I’ve just been having a look and haven’t really turned up anything that jumped out. The main thing seemed to be about them offering licenses and regarding the build server.


Is there anything specific you had in mind?

Well, first he decided he was going to dump the OSS version (even though the entire thing is based on Googles’ code), then when others decided to fork it, he then claimed that tools he checked into github weren’t under it, and were not allowed to be used to build forks. He then tried to threaten them, and he was gently told to fuck off.

Long story short, mholt is a huge prick, and you shouldn’t use anything of his which you can’t instantly replace.

1 Like

TBH I didn’t like the way he tried to monetize it. Pretty ridiculous. Still, I think he’s a very intelligent guy and he’ll learn the lesson.

Was the main issue just the volume of domains they were loading in? Or are those LS quirks still happening just less frequently?

I got a single node doing it still but I’m thinking it’s something else since Litespeed isn’t crashing during it. I’m going to be applying some adjustments to ethtool and see if that helps since we’ve seen some quirks with that chipset on some of our old storage nodes.

Past that, the rest are quiet as a bug in a rug and loads have fallen by half or more.

Part of my accounting script was wrong so we were missing a lot of it. In the end we had a single node that had 16,000 domains between 3 resellers alone, most the others were at least 10,000. Those entries accounted for around 70MB of our httpd.conf file.

I gave everyone backups and told them we couldn’t allow those types of sites on our services anymore and to not try to sneak back in. I shared my auditing script with a few hosts so they didn’t end up getting the same crap.



What the…

Amazon affiliate spam sites.

Each one was running a Wordpress install too and Litespeed cache refused to install so yeah.

It was pretty hard to tie it all together given how spread out things are.

cPanel doesn’t have a way to track/control this all so it’s annoying.



I quite like Lighttpd, no idea why people don’t use it more. That said, I also now use Nginx myself because of the shifting trend towards it. Still using Lighttpd for various applications because of the much saner config.

Also, Caddy is cancer and shouldn’t be used for anything that remotely matters. Use its fork (Wedge) if you must.

Nginx has given me the best experience out of anything… partially because it doesn’t drown with all attacks.

Dear fucking lord, all running WordPress too?!

And prob without proper caching…
Were they 16000 .com/.org/.net domains? If so someone is making big money to pay for that many…

I believe I did give you some incentive to boot them! <3

Apache it just works for me.

1 Like

Helllllll no.

Nothing but .gq .ga and all those free TLD’s.


I tried to at least install Litespeed cache but because of how they rigged things together it wasn’t possible.


1 Like

I’ve never figured out how they make money on those sites.

Affiliate sales.

They write a page about a product on Amazon & if you click on their link and buy the item - they get a small kickback.

1 Like