From: Joel VanderWerf <vjoel@...>
Date: 2009-05-19T08:39:20+09:00
Subject: fcgi.rb and lighttpd


I'm trying to get ruby code to listen to fastcgi from lighttpd... should 
be easy, but ...

The symptom: My ruby program is started up by lighttpd (4 instances, in 
fact), but it never gets requests.

The lighttpd config is:

   server.modules   += ( "mod_fastcgi" )

   fastcgi.debug = 1

   fastcgi.server    = ( "/test/" =>
           ((
                   "bin-path" => "/home/vjoel/simple.rb",
                   "socket" => "/tmp/ruby.socket"
           ))
   )

(I've also tried using ".rb" in place of "/test/".)

The ruby side is adapted from the fcgi documentation:

   #!/usr/bin/env ruby

   require "fcgi"

   File.open("/tmp/simple1.log", "a") {|f| f.puts "step 1"}

   FCGI.each {|request|
     File.open("/tmp/simple2.log", "a") {|f| f.puts "step 2"}

     out = request.out
     out.print "Content-Type: text/plain\r\n"
     out.print "\r\n"
     out.print Time.now.to_s
     request.finish
   }

The "step 1" is written 4 times, as expected, when lighttpd starts. So 
lighttpd is aware of fastcgi, and the fcgi.so lib is loading correctly, 
as is my simple.rb.

Lighttpd is listening at localhost:80, which is easily verified. (I get 
the placeholder page because I have put nothing there.)

However, when I browse to localhost:80/test/something, I get a 404, and 
the "step 2" is never written.

All that shows up in the lighttpd log is: the "starting" message, and 
the four spawning messages, like this one:

2009-05-18 16:12:44: (mod_fastcgi.c.1332) --- fastcgi spawning
         port: 0
         socket /tmp/ruby.socket
         current: 3 / 4

I feel as if I'm missing something obvious...

-- 
       vjoel : Joel VanderWerf : path berkeley edu : 510 665 3407