I've tried my best, but no luck:(
I have a Ubuntu box that I wish have lighttpd running. I've installed
packages:
lighttpd
php4-cgi
and probably a few others to resolve dependencies.
I've then called
$ lighty-enable-mod fastcgi
Okay, fastcgi is enabled. I check the path of php4-cgi and it's under
/usr/bin/php4-cgi. However, lighttpd dies whenever I call a .php file. A
non php file (static file) works fine, though. Error log says
2006-03-23 23:19:34: (response.c.452) Path : /var/www/index.html
2006-03-23 23:19:34: (response.c.459) -- file found
2006-03-23 23:19:34: (response.c.460) Path : /var/www/index.html
2006-03-23 23:19:34: (response.c.582) -- handling subrequest
2006-03-23 23:19:34: (response.c.583) Path : /var/www/index.html
2006-03-23 23:19:34: (mod_staticfile.c.386) -- handling file as static
file
2006-03-23 23:19:34: (response.c.594) -- subrequest finished
2006-03-23 23:20:18: (response.c.582) -- handling subrequest
2006-03-23 23:20:18: (response.c.583) Path : /var/www/index.php
2006-03-23 23:20:18: (mod_fastcgi.c.3549) handling it in mod_fastcgi
2006-03-23 23:20:18: (mod_fastcgi.c.2691) write-req: error 0x080a3df0 0
9000 0
I've tried looking at the configuration file, like so:
$ cat 10-fastcgi.conf
server.modules += ( "mod_fastcgi" )
## Start an FastCGI server for php4 (needs the php4-cgi package)
fastcgi.server = ( ".php" =>
( "localhost" =>
(
"bin-path" => "/usr/bin/php4-cgi",
"port" => 9000
)
)
)
Hm... I find no entry about "socket", so I try adding that.
"socket" => "/tmp/php4-fastcgi.socket"
This time, lighttpd doesn't even start. I never see the socket file
/tmp/php4-fastcgi.socket get created, if this information is of any
help.
on 23.03.2006 15:21
on 15.04.2006 22:37
I have the exact same problem with Debain Sid. Any news on that topic?
on 16.04.2006 18:53
darehanl wrote: > I've tried looking at the configuration file, like so: > $ cat 10-fastcgi.conf > server.modules += ( "mod_fastcgi" ) > > ## Start an FastCGI server for php4 (needs the php4-cgi package) > fastcgi.server = ( ".php" => > ( "localhost" => > ( > "bin-path" => "/usr/bin/php4-cgi", > "port" => 9000 > ) > ) > ) > > Hm... I find no entry about "socket", so I try adding that. > "socket" => "/tmp/php4-fastcgi.socket" > > This time, lighttpd doesn't even start. I never see the socket file > /tmp/php4-fastcgi.socket get created, if this information is of any > help. You can't use both UNIX-sockets and TCP/IP-sockets. If it's on localhost, UNIX-sockets are known to be faster. Anyway, what happens if you run a script through php4-cgi through some other means, say the cli?
on 16.04.2006 22:42
darealh0ba wrote: > I have the exact same problem with Debain Sid. > > Any news on that topic? Well, seems that someone opened a Debian bug report about it: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=362827
on 16.04.2006 22:50
Thomas Bergheim wrote: > You can't use both UNIX-sockets and TCP/IP-sockets. If it's on > localhost, UNIX-sockets are known to be faster. My config reads as follow: server.modules += ( "mod_fastcgi" ) ## Start an FastCGI server for php4 (needs the php4-cgi package) fastcgi.server = ( ".php" => (( "bin-path" => "/usr/bin/php-cgi", "port" => 9000, "max-procs" => 2, "idle-timeout" => 20, "bin-environment" => ( "PHP_FCGI_CHILDREN" => "4", "PHP_FCGI_MAX_REQUESTS" => "5000" ), "bin-copy-environment" => ( "PATH", "SHELL", "USER" ), ) ) ) and here is a strace output of calling a simple php with only phpinfo() in it: rt_sigaction(SIGPIPE, {SIG_IGN}, NULL, 8) = 0 rt_sigaction(SIGUSR1, {SIG_IGN}, NULL, 8) = 0 rt_sigaction(SIGINT, {0x804cfc0, [], SA_SIGINFO}, NULL, 8) = 0 rt_sigaction(SIGTERM, {0x804cfc0, [], SA_SIGINFO}, NULL, 8) = 0 rt_sigaction(SIGHUP, {0x804cfc0, [], SA_SIGINFO}, NULL, 8) = 0 rt_sigaction(SIGALRM, {0x804cfc0, [], SA_SIGINFO}, NULL, 8) = 0 rt_sigaction(SIGCHLD, {0x804cfc0, [], SA_SIGINFO}, NULL, 8) = 0 open("/dev/null", O_RDONLY|O_LARGEFILE) = 8 close(8) = 0 fcntl64(4, F_SETFD, FD_CLOEXEC) = 0 fcntl64(4, F_SETFL, O_RDWR|O_NONBLOCK) = 0 time(NULL) = 1145220294 poll([{fd=4, events=POLLIN}], 1, 1000) = 0 time(NULL) = 1145220295 poll([{fd=4, events=POLLIN}], 1, 1000) = 0 time(NULL) = 1145220296 poll([{fd=4, events=POLLIN, revents=POLLIN}], 1, 1000) = 1 accept(4, {sa_family=AF_INET, sin_port=htons(1088), sin_addr=inet_addr("192.168.1.2")}, [16]) = 8 brk(0x80d2000) = 0x80d2000 brk(0x80f3000) = 0x80f3000 fcntl64(8, F_SETFD, FD_CLOEXEC) = 0 fcntl64(8, F_SETFL, O_RDWR|O_NONBLOCK) = 0 ioctl(8, FIONREAD, [0]) = 0 read(8, 0x80db420, 4159) = -1 EAGAIN (Resource temporarily unavailable) accept(4, 0xbfaa3158, [112]) = -1 EAGAIN (Resource temporarily unavailable) time(NULL) = 1145220296 poll([{fd=4, events=POLLIN}, {fd=8, events=POLLIN, revents=POLLIN}], 2, 1000) = 1 ioctl(8, FIONREAD, [444]) = 0 read(8, "GET /test.php HTTP/1.1\r\nHost"..., 4159) = 444 stat64("/var/www/test.php", {st_mode=S_IFREG|0664, st_size=5841, ...}) = 0 open("/var/www/test.php", O_RDONLY|O_LARGEFILE) = 9 close(9) = 0 stat64("/etc/localtime", {st_mode=S_IFREG|0644, st_size=837, ...}) = 0 write(5, "2006-04-16 22:44:56: (mod_fastcg"..., 80) = 80 --- SIGSEGV (Segmentation fault) @ 0 (0) --- +++ killed by SIGSEGV +++ any ideas?
on 19.04.2006 10:07
Jason Niberg wrote: > darealh0ba wrote: >> I have the exact same problem with Debain Sid. >> >> Any news on that topic? > > Well, seems that someone opened a Debian bug report about it: > http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=362827 And also posted a solution, you might want to try it out.
on 29.04.2006 20:22
I have same problem on debian (testing/unstable) with lighttpd 1.4.11-4 even bug was closed with "We believe that the bug you reported is fixed in the latest version of lighttpd". Thomas Bergheim wrote: > Anyway, what happens if you run a script through php4-cgi through some > other means, say the cli? When ai try to run /usr/bin/php4-cgi -v I obtain the warning PHP Warning: mime_magic: type regex BEGIN[[:space:]]*[{] application/x-awk invalid in Unknown on line 0 I found a solution for this warning on php.net and I added in /etc/php4/cgi/php.ini [mime_magic] mime_magic.magicfile = "/etc/apache2/magic" After this... lighttpd + mod_fastcgi +php4-cgi work great.
on 17.06.2006 23:40
Ok, I had the same problem, with PHP 4.4.2-CGI logging the following error when the fastcgi module of lighttpd 1.4.11-4 (as is shipped w/Dapper) was enabled, and I was trying to load any ".php" file: (mod_fastcgi.c.2691) write-req: error 0x080a3d70 0 9000 0 The solution I found was to download the latest lighttpd package from Debian Unstable (lighttpd 1.4.11-6, get it from here: http://packages.debian.org/unstable/web/lighttpd) and replace my /etc/lighttpd/conf-available/10-fastcgi.conf file with the one in the Debian package. Hope this is useful to other Ubuntu users. Greetings from Argentina, home of Tango, Maradona & Dulce de Leche!
on 16.05.2007 11:09
Cristian R. Arroyo wrote: > Ok, I had the same problem, with PHP 4.4.2-CGI logging the following > error when the fastcgi module of lighttpd 1.4.11-4 (as is shipped > w/Dapper) was enabled, and I was trying to load any ".php" file: > > (mod_fastcgi.c.2691) write-req: error 0x080a3d70 0 9000 0 > > The solution I found was to download the latest lighttpd package from > Debian Unstable (lighttpd 1.4.11-6, get it from here: > http://packages.debian.org/unstable/web/lighttpd) and replace my > /etc/lighttpd/conf-available/10-fastcgi.conf file with the one in the > Debian package. > > Hope this is useful to other Ubuntu users. I just covered the same problem with the Ubuntu 6.06.1/Dapper Drake package. No need to use another packet, just use the syntax from PHP tutorial from the wiki: http://trac.lighttpd.net/trac/wiki/TutorialLighttpdAndPHP fastcgi.server = ( ".php" => (( "bin-path" => "/path/to/php-cgi", "socket" => "/tmp/php.socket" )))
