Re: Forking w/ mod_perl 2

[Richard F. Rebel]

IMHO, it would be better to put your report code into another perl
program and execute it.  From what I see from your snippet of code, it's
not important for the parent to know what the child is going, you are
even ignoring SIGCHLD.

Also, at some point in the future (I hope at least) mp2 + threaded mpm's
will become more than alpha (although I already use it extensively but
I'm crazy) and you might want to run you code in it.  Forking under
these circumstances would be a bad.

2c

On Fri, 2003-09-12 at 14:40, Cameron B. Prince wrote:
 Hi all,
 
 I have a report creation perl script that takes about 15 minutes to run and
 I need to fork it. I tried the code from v1:
 
   use strict;
   use POSIX 'setsid';
   use Apache::SubProcess;
 
 
   my  = shift;
   -send_http_header(text/plain);
 
   {CHLD} = 'IGNORE';
   defined (my  = fork) or die Cannot fork: \n;
   if () {
 print Parent 25481 has finished, kid's PID: \n;
   } else {
   -cleanup_for_exec(); # untie the socket
   chdir '/'or die Can't chdir to /: ;
   open STDIN, '/dev/null'  or die Can't read /dev/null: ;
   open STDOUT, '/dev/null'
   or die Can't write to /dev/null: ;
   open STDERR, '/tmp/log' or die Can't write to /tmp/log: ;
   setsid or die Can't start a new session: ;
 
   select STDERR;
   local $| = 1;
   warn started\n;
   # do something time-consuming
   sleep 1, warn sh\n for 1..20;
   warn completed\n;
 
   CORE::exit(0); # terminate the process
   }
 
 
 First problem, Apache::SubProcess doesn't seem to contain those methods
 anymore.
 Second problem is open.
 
 Can anyone tell me the proper way to fork with v2?
 
 Thanks,
 Cameron
-- 
Richard F. Rebel
[EMAIL PROTECTED]
t. 212.239.


signature.asc
Description: This is a digitally signed message part

RE: Forking w/ mod_perl 2

[Cameron B. Prince]

Hi Richard,

 IMHO, it would be better to put your report code into another perl
 program and execute it.  From what I see from your snippet of
 code, it's
 not important for the parent to know what the child is going, you are
 even ignoring SIGCHLD.

 Also, at some point in the future (I hope at least) mp2 +
 threaded mpm's
 will become more than alpha (although I already use it extensively but
 I'm crazy) and you might want to run you code in it.  Forking under
 these circumstances would be a bad.

Thanks for you reply. The report code is in another perl program that I'm
trying to execute. The code I included below was from the v1 docs:

http://perl.apache.org/docs/1.0/guide/performance.html#Forking_and_Executing
_Subprocesses_from_mod_perl

Here is the code I ended up with:

   $SIG{CHLD} = 'IGNORE';
   defined (my $pid = fork) or die Cannot fork: $!\n;
   unless ($pid) {
  exec $command;
  CORE::exit(0);
   }

This seems to work and no zombies are floating around. But I've not been
able to restart Apache while the forked program is running yet to see if
it's killed.

More comments or ideas welcome.

Thanks,
Cameron


 2c

 On Fri, 2003-09-12 at 14:40, Cameron B. Prince wrote:
  Hi all,
 
  I have a report creation perl script that takes about 15
 minutes to run and
  I need to fork it. I tried the code from v1:
 
use strict;
use POSIX 'setsid';
use Apache::SubProcess;
 
 
my  = shift;
-send_http_header(text/plain);
 
{CHLD} = 'IGNORE';
defined (my  = fork) or die Cannot fork: \n;
if () {
  print Parent 25481 has finished, kid's PID: \n;
} else {
-cleanup_for_exec(); # untie the socket
chdir '/'or die Can't chdir to /: ;
open STDIN, '/dev/null'  or die Can't read /dev/null: ;
open STDOUT, '/dev/null'
or die Can't write to /dev/null: ;
open STDERR, '/tmp/log' or die Can't write to /tmp/log: ;
setsid or die Can't start a new session: ;
 
select STDERR;
local $| = 1;
warn started\n;
# do something time-consuming
sleep 1, warn sh\n for 1..20;
warn completed\n;
 
CORE::exit(0); # terminate the process
}
 
 
  First problem, Apache::SubProcess doesn't seem to contain
 those methods
  anymore.
  Second problem is open.
 
  Can anyone tell me the proper way to fork with v2?
 
  Thanks,
  Cameron
 --
 Richard F. Rebel
 [EMAIL PROTECTED]
 t. 212.239.