Comments on: Why was(is) lighttpd slower? http://blog.init.hr/?p=28 Insight into thoughts of Init Crew Tue, 01 May 2012 11:32:26 +0000 hourly 1 http://wordpress.org/?v=3.3.2 By: GNa http://blog.init.hr/?p=28&cpage=1#comment-30 GNa Tue, 10 Mar 2009 17:00:13 +0000 http://blog.init.hr/?p=28#comment-30 Ofcourse "die" was meant with our config including mod_php. In our situation a dual apache setup was an option, but it was easier to deploy apache+php for the application, lighttpd + mysqlvhosts for static, and squid for a bit of reverse proxy caching. But am also interested in the Cherokee / nginx tests. Ofcourse “die” was meant with our config including mod_php. In our situation a dual apache setup was an option, but it was easier to deploy apache+php for the application, lighttpd + mysqlvhosts for static, and squid for a bit of reverse proxy caching.

But am also interested in the Cherokee / nginx tests.

]]> By: Duncan Lock http://blog.init.hr/?p=28&cpage=1#comment-29 Duncan Lock Tue, 10 Mar 2009 14:57:56 +0000 http://blog.init.hr/?p=28#comment-29 I'd be very interested to see how Cherokee stack up in the same benchmarks! Looking forward to more :) I’d be very interested to see how Cherokee stack up in the same benchmarks! Looking forward to more :)

]]> By: Ante Karamatić http://blog.init.hr/?p=28&cpage=1#comment-28 Ante Karamatić Tue, 10 Mar 2009 12:29:29 +0000 http://blog.init.hr/?p=28#comment-28 No, mod_php wasn't enabled in Apache. Object of testing was static objects, not an application. Well, as you can see, it doesn't die on me. Both apache and lighttpd are flexibile web servers. One should configure them to suite the purpose of requested task; not just use default configuration. No, mod_php wasn’t enabled in Apache. Object of testing was static objects, not an application.

Well, as you can see, it doesn’t die on me. Both apache and lighttpd are flexibile web servers. One should configure them to suite the purpose of requested task; not just use default configuration.

]]> By: GNa http://blog.init.hr/?p=28&cpage=1#comment-27 GNa Tue, 10 Mar 2009 10:30:21 +0000 http://blog.init.hr/?p=28#comment-27 I only have one question: Do you test involve a "bare-metal" apache without php support (i think of mod_php)? We currently use lighttpd for static content serving, and apache with mod_php for the dynamic, because apache on its own dies under the static content serving. I only have one question:

Do you test involve a “bare-metal” apache without php support (i think of mod_php)?
We currently use lighttpd for static content serving, and apache with mod_php for the dynamic, because apache on its own dies under the static content serving.

]]> By: Ante Karamatić http://blog.init.hr/?p=28&cpage=1#comment-26 Ante Karamatić Tue, 10 Mar 2009 08:51:44 +0000 http://blog.init.hr/?p=28#comment-26 Thank you, that's correct observation. Thank you, that’s correct observation.

]]> By: glandium http://blog.init.hr/?p=28&cpage=1#comment-25 glandium Tue, 10 Mar 2009 08:12:44 +0000 http://blog.init.hr/?p=28#comment-25 Please note that while you are trying to be fair by enabling cache in apache, you are not. What you enabled in lighttpd is a stat() cache, meaning all it caches is the stat() syscall return values. On the other hand, apache's cache will cache file *contents* to disk or memory. Please note that while you are trying to be fair by enabling cache in apache, you are not. What you enabled in lighttpd is a stat() cache, meaning all it caches is the stat() syscall return values. On the other hand, apache’s cache will cache file *contents* to disk or memory.

]]>