php.net@domainofdarkness.com
29-Jan-2001 03:26

OK, WRT to the p* functions opening a new connection when one already
exists. It is my understanting that (under Apache anyways) this is on a
per-process basis. If you do a 'ps auxw|grep httpd' on your server you
will see more than one process. What p* does is make a p-connection on one
of those processes only, the one that actually handles your request.
Chances are that when you hit the page again it will be answered by a
different process. I'm guessing if you keep hitting reload you'll get
around to the original process again and there will be no error message or
second connection open. Anyhow, this is true of all p* functions; they
open not one connection per server, but one connection per server
_process_.

Hope this explains something (it's from the manual at http://www.php.net/manual/en/function.pfsockopen.php)
I'm keeping this open as a doc problem though, cause the documentation is very skimmy.

Derick

I know, that p* functions opens 1 connection per process.
But in any case p* functions should not close this connection, but hold it opened in hashed table for next time, same script will be executed on this apache process, isnt it ? Unfortunately this doen not happend ;( PHP really closes the socket right after script end ;(

And in tests apache was configured like:
MinSpareServers 1 
MaxSpareServers 1 
StartServers 1    
MaxClients 1      

That means that only 2 apache processes are in memory, (parent and a child) and only one of them handles php scripts. But all tests faied, as i talk. any pfsockopen connections are closed after end of script ;(

I think this locate the problem and try to test it as soon as possible.
----
http://marc.theaimsgroup.com/?l=php-dev&m=100842297427750&w=2

[2000-09-27 22:17:24] wbrack@mmm.com.hk

A UNIX socket connection is made using pfsockopen().  On the initial script entry, a message is sent to our server and the reply is correctly received by the script.  On 
subsequent script entries, messages to the server are still sent successfully, but upon attempting to reply a SIGPIPE is received by server, and the php script receives a 0-length reply to its fgets() read.

We traced the problem to ext/standard/file.c, in the routine _file_socket_dtor().  In that routine, the macro SOCK_FCLOSE is used, which calls php_sock_close() in fsock.c. This routine correctly handles the persistent socket.  However, after that call, _file_socket_dtor() then incorrectly calls the C routine shutdown(), which is what 
caused the problem.  In fact, php_sock_close() already completely takes care of the shutdown() (for the non-persistent case), so in any event the shutdown() call in 
_file_socket_dtor() is not necessary.

Fixed in CVS.
It looked to me like pfsockopen never worked properly before.
Please try http://snaps.php.net/php4-latest.tar.gz and
re-open this report if the problem is still evident.