New on LowEndTalk? Please Register and read our Community Rules.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Comments
I still use it for serving cgi (e.g., baycam.ca) or for when I want only minimal http capabilities without php, etc.
Nginx uses less CPU, memory, and serve more requests/sec, why would you use lighttpd? I also find lighttpd crumbles under high load.
Wheres that based on?
Experience I guess. here's some simple status from one of my LEBs, which I use as a download server. Debian 6 stripped down, Nginx/PHP-FPM and SSHd are all that run.
One example more of this kind of questions...
http://serverfault.com/questions/114895/why-is-nginx-more-popular-than-lighttpd
But you have toooo much things to do, so don't read
LIGHTTPD configuration is indeed 'clean' and easier to understand imo. But because I love PHP-FPM more than fast-cgi, so I switched to nginx.
Using LIGHTTPD with PHP-FPM is pain in the ass if your web is WordPress with pretty permalinks, because you have to use mod rewrite and it's not easy for me.
Thats just plain false.
I used Lighttpd a year back, caused huge memory leaks.
Since then, I've only ever used it for HTTP Keep Alives, which NGINX doesn't seem to like too much.
Proof? I've been using Lighttpd for years and haven't had a single memory issue. I currently manage a client's server with 500k+ uniques a day and it runs Lighttpd without issues as well. Before you go pushing that memory leak rumor (that was based off an actual memory leak from like 4-5 years ago), you better have some proof.
It won't need proof, every software has memory leaks. Nginx does, just everything does. And no professional developer would ever claim that his software doesn't have memory leaks.
I'm rolling my eyes at that retarded answer.
http://redmine.lighttpd.net/boards/2/topics/4009 (notice he's one of the lead developers)
The "OMG LIGHTTPD HAS MEMORY LEAKS!!!11!!" crap originally started when people misunderstood this bug report: http://redmine.lighttpd.net/issues/758
If there was some giant leak that caused RAM to be consumed to the point of crashing a server, there would be at least ONE bug report with actual proof. As it is, the only people who bitch about a Lighttpd memory leak, are people on forums like this, who haven't even used it.
@Roph thats a nice status page.. Do you want to share the source? It looks prettier than mine: http://raymii.org/status.html
Noone talked about a giant leak, i was refering to a 2 or 3 byte memory leak.
@Raymii How bout the source of yours?
@bijan588 http://raymii.nl/status.txt
In a cronjob every 5 minutes, running as root. Adapt where necessary.
It's this: http://linux.dashed.ca/index.cgi/2011-01
@OneTwo
Congratulations you create a holy war on LEB
# Created by D. Robbins.'
# Last revised: 17 october 2011'
# by Raymii.org'
# CC Unported Attribution 3.0'
There's a good chance of thttpd (and uhttpd, as @yomero noted) bringing an LEB to its knees once they start getting more than one or two simultaneous connections.
I used to prefer thttpd for my impromptu-console-download-server needs, but it often goes OOM if using aria2 to increase throughput from a slow route. I've switched to webfs since and been very happy.
As for the main discussion, yes, the nginx paradigm/config has a learning curve for anyone used to apache, but once you spend a little time understanding it, it's no harder. nginx FTW.
Funnily enough it was posted to 4chan's /g/ board.
http://pastebin.com/L5C39ZTx
I plan to edit it up a little to fix it for OVZ when I can be bothered
Separate logs per vhost in nginx, "alive" and in heavy development nginx, http://redmine.lighttpd.net/projects/4/wiki/Howto_WSGI - more simple configs?!
Not relevant right now, but Nginx is going ahead with implementing SPDY. I'm definitely interested in how it pans out when it's better supported in browsers =o
Can't say I've ever had that experience. After running thttpd for over a year I think I musta had more then 1 or 2 simultaneous connections at some point
Though I do use only to serve cgi, behind nginx.
Because it's Russian.
http://royal.pingdom.com/2010/02/23/nginx-the-little-russian-web-server-taking-on-the-giants/
I don't see much VPS/Dedi providers offering pre-installed Ngnix though.
I have used Lighttpd on a Shared Hosting server once before it and did cause a huge memory leak aswell.. I've always used Apache, it is light and does what I need it to do. My AlienVPS $19/Year VPS idles at 33MB Ram and runs Apache.
I only use Lighttpd, never had any issues with it.
I never have issues with nginx except when people decide to run slowloris from multiple systems.
While we were talking about exotic webservers like microhttpd, anyone has experience with the Cherokee Webserver?
http://www.cherokee-project.com/
I tried it, but didn't see the point.
Ah okay, I will try myself once and post back some results. Name sounds like Apache's brother lol, but I did like what I saw from the admin interface built in (for non-cli people)
Founder of it seems a big guy from OpenSolaris, and reading this post it is (just) another Apache alternative:
http://blog.dynamichosting.biz/2011/04/07/cherokee-vs-apache-benchmarks/