#28556 [Fbk->Opn]: PHP as CGI becomes a zombie when loaded too often

2005-03-02 Thread floeff at arcor dot de
 ID:   28556
 User updated by:  floeff at arcor dot de
 Reported By:  floeff at arcor dot de
-Status:   Feedback
+Status:   Open
 Bug Type: CGI related
 Operating System: Linux 2.4
 PHP Version:  4.3.6
 New Comment:

It was the Debian 3.0 package, version 1.1.3-6.3


Previous Comments:


[2005-03-02 23:21:28] [EMAIL PROTECTED]

What libmm version do you have installed?
(I can't reproduce this with latest stable CVS and libmm-1.3.1)





[2005-03-01 19:23:06] floeff at arcor dot de

I already tried without binfmt_misc, same problem. Regarding the
snapshot: Unfortunately, I have no testing machine at the moment, and
after removing that mm stuff, it works fine. I will get back to you if
I have a testing machine again someday, but feel free to test it out
yourself and ask whatever question you have.



[2005-03-01 01:59:59] [EMAIL PROTECTED]

Please try using this CVS snapshot:

  http://snaps.php.net/php4-STABLE-latest.tar.gz
 
For Windows:
 
  http://snaps.php.net/win32/php4-win32-STABLE-latest.zip

And please try without using this binfmt_misc module to rule out THAT
possible cause of this bug.






[2005-02-06 14:27:36] floeff at arcor dot de

Okay, here we go:

./configure --enable-safe-mode --with-mysql --enable-discard-path
--with-exec-dir --enable-memory-limit --with-mm && make && make
install

cp php.ini-dist /usr/local/lib/php.ini

echo ':PHP:E::php::/usr/local/bin/php:' >
/proc/sys/fs/binfmt_misc/register
echo ':PHP3:E::php3::/usr/local/bin/php:' >
/proc/sys/fs/binfmt_misc/register
echo ':PHP4:E::php4::/usr/local/bin/php:' >
/proc/sys/fs/binfmt_misc/register

Changes in php.ini:
expose_php = Off
disable_functions = phpinfo
allow_url_fopen = Off

In httpd.conf of Apache2:

AddHandler cgi-script .cgi .pl .php .php3 .php4



[2005-02-06 06:57:15] [EMAIL PROTECTED]

Come up with configuration that isn't "security threat" to you and put
it here..




The remainder of the comments for this report are too long. To view
the rest of the comments, please view the bug report online at
http://bugs.php.net/28556

-- 
Edit this bug report at http://bugs.php.net/?id=28556&edit=1


#28556 [Fbk->Opn]: PHP as CGI becomes a zombie when loaded too often

2005-03-01 Thread floeff at arcor dot de
 ID:   28556
 User updated by:  floeff at arcor dot de
 Reported By:  floeff at arcor dot de
-Status:   Feedback
+Status:   Open
 Bug Type: CGI related
 Operating System: Linux 2.4
 PHP Version:  4.3.6
 New Comment:

I already tried without binfmt_misc, same problem. Regarding the
snapshot: Unfortunately, I have no testing machine at the moment, and
after removing that mm stuff, it works fine. I will get back to you if
I have a testing machine again someday, but feel free to test it out
yourself and ask whatever question you have.


Previous Comments:


[2005-03-01 01:59:59] [EMAIL PROTECTED]

Please try using this CVS snapshot:

  http://snaps.php.net/php4-STABLE-latest.tar.gz
 
For Windows:
 
  http://snaps.php.net/win32/php4-win32-STABLE-latest.zip

And please try without using this binfmt_misc module to rule out THAT
possible cause of this bug.






[2005-02-06 14:27:36] floeff at arcor dot de

Okay, here we go:

./configure --enable-safe-mode --with-mysql --enable-discard-path
--with-exec-dir --enable-memory-limit --with-mm && make && make
install

cp php.ini-dist /usr/local/lib/php.ini

echo ':PHP:E::php::/usr/local/bin/php:' >
/proc/sys/fs/binfmt_misc/register
echo ':PHP3:E::php3::/usr/local/bin/php:' >
/proc/sys/fs/binfmt_misc/register
echo ':PHP4:E::php4::/usr/local/bin/php:' >
/proc/sys/fs/binfmt_misc/register

Changes in php.ini:
expose_php = Off
disable_functions = phpinfo
allow_url_fopen = Off

In httpd.conf of Apache2:

AddHandler cgi-script .cgi .pl .php .php3 .php4



[2005-02-06 06:57:15] [EMAIL PROTECTED]

Come up with configuration that isn't "security threat" to you and put
it here..


----------------

[2005-02-05 19:07:17] floeff at arcor dot de

I don't post my (maybe security-related) configuration to the public,
sorry. Any e-mail address I could mail it to?



[2005-02-05 03:40:03] [EMAIL PROTECTED]

DO NOT email anything to me!! Put any details needed to reproduce this
bug here, via this url:

http://bugs.php.net/bug.php?id=28556&edit=2





The remainder of the comments for this report are too long. To view
the rest of the comments, please view the bug report online at
http://bugs.php.net/28556

-- 
Edit this bug report at http://bugs.php.net/?id=28556&edit=1


#28556 [Fbk->Opn]: PHP as CGI becomes a zombie when loaded too often

2005-02-06 Thread floeff at arcor dot de
 ID:   28556
 User updated by:  floeff at arcor dot de
 Reported By:  floeff at arcor dot de
-Status:   Feedback
+Status:   Open
 Bug Type: CGI related
 Operating System: Linux 2.4
 PHP Version:  4.3.6
 New Comment:

Okay, here we go:

./configure --enable-safe-mode --with-mysql --enable-discard-path
--with-exec-dir --enable-memory-limit --with-mm && make && make
install

cp php.ini-dist /usr/local/lib/php.ini

echo ':PHP:E::php::/usr/local/bin/php:' >
/proc/sys/fs/binfmt_misc/register
echo ':PHP3:E::php3::/usr/local/bin/php:' >
/proc/sys/fs/binfmt_misc/register
echo ':PHP4:E::php4::/usr/local/bin/php:' >
/proc/sys/fs/binfmt_misc/register

Changes in php.ini:
expose_php = Off
disable_functions = phpinfo
allow_url_fopen = Off

In httpd.conf of Apache2:

AddHandler cgi-script .cgi .pl .php .php3 .php4


Previous Comments:


[2005-02-06 06:57:15] [EMAIL PROTECTED]

Come up with configuration that isn't "security threat" to you and put
it here..


----------------

[2005-02-05 19:07:17] floeff at arcor dot de

I don't post my (maybe security-related) configuration to the public,
sorry. Any e-mail address I could mail it to?



[2005-02-05 03:40:03] [EMAIL PROTECTED]

DO NOT email anything to me!! Put any details needed to reproduce this
bug here, via this url:

http://bugs.php.net/bug.php?id=28556&edit=2



----------------

[2005-02-03 21:05:48] floeff at arcor dot de

Unfortunately, I have no test system to try.
If you want, I can mail you my documentation privately (via PM), so you
can check it out.



[2005-02-03 19:25:49] [EMAIL PROTECTED]

And it still happens with latest CVS snapshot?



The remainder of the comments for this report are too long. To view
the rest of the comments, please view the bug report online at
http://bugs.php.net/28556

-- 
Edit this bug report at http://bugs.php.net/?id=28556&edit=1


#28556 [Fbk->Opn]: PHP as CGI becomes a zombie when loaded too often

2005-02-05 Thread floeff at arcor dot de
 ID:   28556
 User updated by:  floeff at arcor dot de
 Reported By:  floeff at arcor dot de
-Status:   Feedback
+Status:   Open
 Bug Type: CGI related
 Operating System: Linux 2.4
 PHP Version:  4.3.6
 New Comment:

I don't post my (maybe security-related) configuration to the public,
sorry. Any e-mail address I could mail it to?


Previous Comments:


[2005-02-05 03:40:03] [EMAIL PROTECTED]

DO NOT email anything to me!! Put any details needed to reproduce this
bug here, via this url:

http://bugs.php.net/bug.php?id=28556&edit=2





[2005-02-03 21:05:48] floeff at arcor dot de

Unfortunately, I have no test system to try.
If you want, I can mail you my documentation privately (via PM), so you
can check it out.



[2005-02-03 19:25:49] [EMAIL PROTECTED]

And it still happens with latest CVS snapshot?



[2005-02-03 08:46:42] floeff at arcor dot de

Or was it --with-mm?



[2005-02-03 04:04:22] [EMAIL PROTECTED]

What --with-libmm do you mean? There is no such configure option in
PHP..




The remainder of the comments for this report are too long. To view
the rest of the comments, please view the bug report online at
http://bugs.php.net/28556

-- 
Edit this bug report at http://bugs.php.net/?id=28556&edit=1


#28556 [Fbk->Opn]: PHP as CGI becomes a zombie when loaded too often

2005-02-03 Thread floeff at arcor dot de
 ID:   28556
 User updated by:  floeff at arcor dot de
 Reported By:  floeff at arcor dot de
-Status:   Feedback
+Status:   Open
 Bug Type: CGI related
 Operating System: Linux 2.4
 PHP Version:  4.3.6
 New Comment:

Unfortunately, I have no test system to try.
If you want, I can mail you my documentation privately (via PM), so you
can check it out.


Previous Comments:


[2005-02-03 19:25:49] [EMAIL PROTECTED]

And it still happens with latest CVS snapshot?



[2005-02-03 08:46:42] floeff at arcor dot de

Or was it --with-mm?



[2005-02-03 04:04:22] [EMAIL PROTECTED]

What --with-libmm do you mean? There is no such configure option in
PHP..




[2004-12-13 09:18:50] floeff at arcor dot de

Sorry, seems I forgot to update this bug report.
The problem disappears as soon as I do *NOT* compile with --with-libmm

So it seem to be a libmm bug?



[2004-07-08 17:58:36] floeff at arcor dot de

I have tried on a lot of machines.



The remainder of the comments for this report are too long. To view
the rest of the comments, please view the bug report online at
http://bugs.php.net/28556

-- 
Edit this bug report at http://bugs.php.net/?id=28556&edit=1


#28556 [Fbk->Opn]: PHP as CGI becomes a zombie when loaded too often

2005-02-02 Thread floeff at arcor dot de
 ID:   28556
 User updated by:  floeff at arcor dot de
 Reported By:  floeff at arcor dot de
-Status:   Feedback
+Status:   Open
 Bug Type: CGI related
 Operating System: Linux 2.4
 PHP Version:  4.3.6
 New Comment:

Or was it --with-mm?


Previous Comments:


[2005-02-03 04:04:22] [EMAIL PROTECTED]

What --with-libmm do you mean? There is no such configure option in
PHP..




[2004-12-13 09:18:50] floeff at arcor dot de

Sorry, seems I forgot to update this bug report.
The problem disappears as soon as I do *NOT* compile with --with-libmm

So it seem to be a libmm bug?



[2004-12-13 01:06:13] [EMAIL PROTECTED]

Please try using this CVS snapshot:

  http://snaps.php.net/php4-STABLE-latest.tar.gz
 
For Windows:
 
  http://snaps.php.net/win32/php4-win32-STABLE-latest.zip





[2004-07-08 17:58:36] floeff at arcor dot de

I have tried on a lot of machines.



[2004-07-08 13:02:00] [EMAIL PROTECTED]

Have you tried on another machine?



The remainder of the comments for this report are too long. To view
the rest of the comments, please view the bug report online at
http://bugs.php.net/28556

-- 
Edit this bug report at http://bugs.php.net/?id=28556&edit=1


#28556 [Opn]: PHP as CGI becomes a zombie when loaded too often

2004-12-13 Thread floeff at arcor dot de
 ID:   28556
 User updated by:  floeff at arcor dot de
 Reported By:  floeff at arcor dot de
 Status:   Open
 Bug Type: CGI related
 Operating System: Linux 2.4
 PHP Version:  4.3.6
 New Comment:

Sorry, seems I forgot to update this bug report.
The problem disappears as soon as I do *NOT* compile with --with-libmm

So it seem to be a libmm bug?


Previous Comments:


[2004-12-13 09:18:50] floeff at arcor dot de

Sorry, seems I forgot to update this bug report.
The problem disappears as soon as I do *NOT* compile with --with-libmm

So it seem to be a libmm bug?



[2004-12-13 01:06:13] [EMAIL PROTECTED]

Please try using this CVS snapshot:

  http://snaps.php.net/php4-STABLE-latest.tar.gz
 
For Windows:
 
  http://snaps.php.net/win32/php4-win32-STABLE-latest.zip





[2004-07-08 17:58:36] floeff at arcor dot de

I have tried on a lot of machines.



[2004-07-08 13:02:00] [EMAIL PROTECTED]

Have you tried on another machine?



[2004-06-24 18:18:03] trevor at verite dot com

I have also experienced this problem, we have a nightly crontab that
uses the CGI version of PHP to run a weekly import, every week we need
to kill the script because even with a exit; command at the end it just
keeps running and taking up more and more memory/processor time. I'm not
sure if this is related but..

We're on a debian potato/apache 1.3/PHP 4.3.7 (Jun 7th Build).

Any ideas?



The remainder of the comments for this report are too long. To view
the rest of the comments, please view the bug report online at
http://bugs.php.net/28556

-- 
Edit this bug report at http://bugs.php.net/?id=28556&edit=1



#28556 [Fbk->Opn]: PHP as CGI becomes a zombie when loaded too often

2004-12-13 Thread floeff at arcor dot de
 ID:   28556
 User updated by:  floeff at arcor dot de
 Reported By:  floeff at arcor dot de
-Status:   Feedback
+Status:   Open
 Bug Type: CGI related
 Operating System: Linux 2.4
 PHP Version:  4.3.6
 New Comment:

Sorry, seems I forgot to update this bug report.
The problem disappears as soon as I do *NOT* compile with --with-libmm

So it seem to be a libmm bug?


Previous Comments:


[2004-12-13 01:06:13] [EMAIL PROTECTED]

Please try using this CVS snapshot:

  http://snaps.php.net/php4-STABLE-latest.tar.gz
 
For Windows:
 
  http://snaps.php.net/win32/php4-win32-STABLE-latest.zip





[2004-07-08 17:58:36] floeff at arcor dot de

I have tried on a lot of machines.



[2004-07-08 13:02:00] [EMAIL PROTECTED]

Have you tried on another machine?



[2004-06-24 18:18:03] trevor at verite dot com

I have also experienced this problem, we have a nightly crontab that
uses the CGI version of PHP to run a weekly import, every week we need
to kill the script because even with a exit; command at the end it just
keeps running and taking up more and more memory/processor time. I'm not
sure if this is related but..

We're on a debian potato/apache 1.3/PHP 4.3.7 (Jun 7th Build).

Any ideas?



[2004-05-28 14:37:06] floeff at arcor dot de

Description:

Sorry for posting this again and again, but I still experience this
problem and there seems to be no way for me to solve it. I've got
confirmation from others that this problem is not only mine, so please
take the time to read this.

I've tried Apache 1.3 and 2.0, both on Linux 2.4. I've tried using
suEXEC and not using suEXEC, and I even tried modules that stop script
execution at a specific load average (tested with 1.00!) or number of
processes (tested with 10!). But nothing seems to help in the following
case:

I run PHP as CGI because I don't want to have world-readable scripts
and mod_perchild is not ready yet. When I do a hard reload - i.e.
reloading the same script for about 10 seconds continously which should
open quite a lot of scripts - I can crash the server. PHP-CGI-processes
become zombies, I get a load average of about 90 (!) and it can take up
to 30 minutes until the system responds again. This happens even with
the simplest PHP scripts like a phpinfo call, but Perl scripts make
absolutely no problem.

The PHP developers say it's an Apache problem, the Apache developers
say it's a PHP problem. So *PLEASE* take the time to review this one
again - I'm helpless right now! :-( I know there must be a solution,
because some providers run PHP as CGI without problems, but I don't
know what it could be. :-(






-- 
Edit this bug report at http://bugs.php.net/?id=28556&edit=1


#28556 [Fbk->Opn]: PHP as CGI becomes a zombie when loaded too often

2004-07-08 Thread floeff at arcor dot de
 ID:   28556
 User updated by:  floeff at arcor dot de
 Reported By:  floeff at arcor dot de
-Status:   Feedback
+Status:   Open
 Bug Type: *General Issues
 Operating System: Linux 2.4
 PHP Version:  4.3.6
 New Comment:

I have tried on a lot of machines.


Previous Comments:


[2004-07-08 13:02:00] [EMAIL PROTECTED]

Have you tried on another machine?



[2004-06-24 18:18:03] trevor at verite dot com

I have also experienced this problem, we have a nightly crontab that
uses the CGI version of PHP to run a weekly import, every week we need
to kill the script because even with a exit; command at the end it just
keeps running and taking up more and more memory/processor time. I'm not
sure if this is related but..

We're on a debian potato/apache 1.3/PHP 4.3.7 (Jun 7th Build).

Any ideas?



[2004-05-28 14:37:06] floeff at arcor dot de

Description:

Sorry for posting this again and again, but I still experience this
problem and there seems to be no way for me to solve it. I've got
confirmation from others that this problem is not only mine, so please
take the time to read this.

I've tried Apache 1.3 and 2.0, both on Linux 2.4. I've tried using
suEXEC and not using suEXEC, and I even tried modules that stop script
execution at a specific load average (tested with 1.00!) or number of
processes (tested with 10!). But nothing seems to help in the following
case:

I run PHP as CGI because I don't want to have world-readable scripts
and mod_perchild is not ready yet. When I do a hard reload - i.e.
reloading the same script for about 10 seconds continously which should
open quite a lot of scripts - I can crash the server. PHP-CGI-processes
become zombies, I get a load average of about 90 (!) and it can take up
to 30 minutes until the system responds again. This happens even with
the simplest PHP scripts like a phpinfo call, but Perl scripts make
absolutely no problem.

The PHP developers say it's an Apache problem, the Apache developers
say it's a PHP problem. So *PLEASE* take the time to review this one
again - I'm helpless right now! :-( I know there must be a solution,
because some providers run PHP as CGI without problems, but I don't
know what it could be. :-(






-- 
Edit this bug report at http://bugs.php.net/?id=28556&edit=1


#28556 [NEW]: PHP as CGI becomes a zombie when loaded too often

2004-05-28 Thread floeff at arcor dot de
From: floeff at arcor dot de
Operating system: Linux 2.4
PHP version:  4.3.6
PHP Bug Type: Reproducible crash
Bug description:  PHP as CGI becomes a zombie when loaded too often

Description:

Sorry for posting this again and again, but I still experience this
problem and there seems to be no way for me to solve it. I've got
confirmation from others that this problem is not only mine, so please
take the time to read this.

I've tried Apache 1.3 and 2.0, both on Linux 2.4. I've tried using suEXEC
and not using suEXEC, and I even tried modules that stop script execution
at a specific load average (tested with 1.00!) or number of processes
(tested with 10!). But nothing seems to help in the following case:

I run PHP as CGI because I don't want to have world-readable scripts and
mod_perchild is not ready yet. When I do a hard reload - i.e. reloading
the same script for about 10 seconds continously which should open quite a
lot of scripts - I can crash the server. PHP-CGI-processes become zombies,
I get a load average of about 90 (!) and it can take up to 30 minutes
until the system responds again. This happens even with the simplest PHP
scripts like a phpinfo call, but Perl scripts make absolutely no problem.

The PHP developers say it's an Apache problem, the Apache developers say
it's a PHP problem. So *PLEASE* take the time to review this one again -
I'm helpless right now! :-( I know there must be a solution, because some
providers run PHP as CGI without problems, but I don't know what it could
be. :-(


-- 
Edit bug report at http://bugs.php.net/?id=28556&edit=1
-- 
Try a CVS snapshot (php4):  http://bugs.php.net/fix.php?id=28556&r=trysnapshot4
Try a CVS snapshot (php5):  http://bugs.php.net/fix.php?id=28556&r=trysnapshot5
Fixed in CVS:   http://bugs.php.net/fix.php?id=28556&r=fixedcvs
Fixed in release:   http://bugs.php.net/fix.php?id=28556&r=alreadyfixed
Need backtrace: http://bugs.php.net/fix.php?id=28556&r=needtrace
Need Reproduce Script:  http://bugs.php.net/fix.php?id=28556&r=needscript
Try newer version:  http://bugs.php.net/fix.php?id=28556&r=oldversion
Not developer issue:http://bugs.php.net/fix.php?id=28556&r=support
Expected behavior:  http://bugs.php.net/fix.php?id=28556&r=notwrong
Not enough info:http://bugs.php.net/fix.php?id=28556&r=notenoughinfo
Submitted twice:http://bugs.php.net/fix.php?id=28556&r=submittedtwice
register_globals:   http://bugs.php.net/fix.php?id=28556&r=globals
PHP 3 support discontinued: http://bugs.php.net/fix.php?id=28556&r=php3
Daylight Savings:   http://bugs.php.net/fix.php?id=28556&r=dst
IIS Stability:  http://bugs.php.net/fix.php?id=28556&r=isapi
Install GNU Sed:http://bugs.php.net/fix.php?id=28556&r=gnused
Floating point limitations: http://bugs.php.net/fix.php?id=28556&r=float


#28550 [WFx]: limit number of simultaneous processes

2004-05-27 Thread floeff at arcor dot de
 ID:   28550
 User updated by:  floeff at arcor dot de
 Reported By:  floeff at arcor dot de
 Status:   Wont fix
 Bug Type: Feature/Change Request
 Operating System: Linux 2.4
 PHP Version:  4.3.6
 New Comment:

I did not find any way to limit this in Apache, and this problem really
worries me. So if you know anything, PLEASE let me know!


Previous Comments:


[2004-05-27 23:42:09] [EMAIL PROTECTED]

That's already handled (appriately so) in the webserver.  Check your
webservers documentation for how to implement.




[2004-05-27 22:52:46] floeff at arcor dot de

Description:

I would like to see a feature to limit the number of simultaneous
PHP-as-CGI processes, maybe per user.






-- 
Edit this bug report at http://bugs.php.net/?id=28550&edit=1


#28550 [NEW]: limit number of simultaneous processes

2004-05-27 Thread floeff at arcor dot de
From: floeff at arcor dot de
Operating system: Linux 2.4
PHP version:  4.3.6
PHP Bug Type: Feature/Change Request
Bug description:  limit number of simultaneous processes

Description:

I would like to see a feature to limit the number of simultaneous
PHP-as-CGI processes, maybe per user.


-- 
Edit bug report at http://bugs.php.net/?id=28550&edit=1
-- 
Try a CVS snapshot (php4):  http://bugs.php.net/fix.php?id=28550&r=trysnapshot4
Try a CVS snapshot (php5):  http://bugs.php.net/fix.php?id=28550&r=trysnapshot5
Fixed in CVS:   http://bugs.php.net/fix.php?id=28550&r=fixedcvs
Fixed in release:   http://bugs.php.net/fix.php?id=28550&r=alreadyfixed
Need backtrace: http://bugs.php.net/fix.php?id=28550&r=needtrace
Need Reproduce Script:  http://bugs.php.net/fix.php?id=28550&r=needscript
Try newer version:  http://bugs.php.net/fix.php?id=28550&r=oldversion
Not developer issue:http://bugs.php.net/fix.php?id=28550&r=support
Expected behavior:  http://bugs.php.net/fix.php?id=28550&r=notwrong
Not enough info:http://bugs.php.net/fix.php?id=28550&r=notenoughinfo
Submitted twice:http://bugs.php.net/fix.php?id=28550&r=submittedtwice
register_globals:   http://bugs.php.net/fix.php?id=28550&r=globals
PHP 3 support discontinued: http://bugs.php.net/fix.php?id=28550&r=php3
Daylight Savings:   http://bugs.php.net/fix.php?id=28550&r=dst
IIS Stability:  http://bugs.php.net/fix.php?id=28550&r=isapi
Install GNU Sed:http://bugs.php.net/fix.php?id=28550&r=gnused
Floating point limitations: http://bugs.php.net/fix.php?id=28550&r=float


#28457 [WFx]: PHP should be able to check certain conditions before invoking the script

2004-05-20 Thread floeff at arcor dot de
 ID:   28457
 User updated by:  floeff at arcor dot de
 Reported By:  floeff at arcor dot de
 Status:   Wont fix
 Bug Type: Feature/Change Request
 Operating System: Linux
 PHP Version:  4.3.6
 New Comment:

Having mod_php can be a security risk, too.


Previous Comments:


[2004-05-20 20:42:43] [EMAIL PROTECTED]

Having a windows box on the internet is a security risk too, and people
still do it even if they are warned.



[2004-05-20 19:47:03] floeff at arcor dot de

You should at least provide some tips for stopping these problems, as I
consider this a security risk.



[2004-05-20 18:55:03] [EMAIL PROTECTED]

This is not something that PHP should implement, as it's not PHPs job
to make sure your system "doesn't overload". It's an operating systems
feature, or at most done by the webserver.

----

[2004-05-20 17:14:22] floeff at arcor dot de

Description:

PHP should be able to check certain conditions before invoking the
script. I 
would like to be able to check

 - number of already running PHP(-as-CGI) instances
 - load average of the system

If some user-definable values are exceed, PHP should NOT invoke the
required 
script, but instead printing out a "System overload, please try again
later" 
message.






-- 
Edit this bug report at http://bugs.php.net/?id=28457&edit=1


#28457 [WFx]: PHP should be able to check certain conditions before invoking the script

2004-05-20 Thread floeff at arcor dot de
 ID:   28457
 User updated by:  floeff at arcor dot de
 Reported By:  floeff at arcor dot de
 Status:   Wont fix
 Bug Type: Feature/Change Request
 Operating System: Linux
 PHP Version:  4.3.6
 New Comment:

You should at least provide some tips for stopping these problems, as I
consider this a security risk.


Previous Comments:


[2004-05-20 18:55:03] [EMAIL PROTECTED]

This is not something that PHP should implement, as it's not PHPs job
to make sure your system "doesn't overload". It's an operating systems
feature, or at most done by the webserver.

----

[2004-05-20 17:14:22] floeff at arcor dot de

Description:

PHP should be able to check certain conditions before invoking the
script. I 
would like to be able to check

 - number of already running PHP(-as-CGI) instances
 - load average of the system

If some user-definable values are exceed, PHP should NOT invoke the
required 
script, but instead printing out a "System overload, please try again
later" 
message.






-- 
Edit this bug report at http://bugs.php.net/?id=28457&edit=1


#28457 [NEW]: PHP should be able to check certain conditions before invoking the script

2004-05-20 Thread floeff at arcor dot de
From: floeff at arcor dot de
Operating system: Linux
PHP version:  4.3.6
PHP Bug Type: Feature/Change Request
Bug description:  PHP should be able to check certain conditions before invoking the 
script

Description:

PHP should be able to check certain conditions before invoking the script.
I 
would like to be able to check

 - number of already running PHP(-as-CGI) instances
 - load average of the system

If some user-definable values are exceed, PHP should NOT invoke the
required 
script, but instead printing out a "System overload, please try again
later" 
message.


-- 
Edit bug report at http://bugs.php.net/?id=28457&edit=1
-- 
Try a CVS snapshot (php4):  http://bugs.php.net/fix.php?id=28457&r=trysnapshot4
Try a CVS snapshot (php5):  http://bugs.php.net/fix.php?id=28457&r=trysnapshot5
Fixed in CVS:   http://bugs.php.net/fix.php?id=28457&r=fixedcvs
Fixed in release:   http://bugs.php.net/fix.php?id=28457&r=alreadyfixed
Need backtrace: http://bugs.php.net/fix.php?id=28457&r=needtrace
Need Reproduce Script:  http://bugs.php.net/fix.php?id=28457&r=needscript
Try newer version:  http://bugs.php.net/fix.php?id=28457&r=oldversion
Not developer issue:http://bugs.php.net/fix.php?id=28457&r=support
Expected behavior:  http://bugs.php.net/fix.php?id=28457&r=notwrong
Not enough info:http://bugs.php.net/fix.php?id=28457&r=notenoughinfo
Submitted twice:http://bugs.php.net/fix.php?id=28457&r=submittedtwice
register_globals:   http://bugs.php.net/fix.php?id=28457&r=globals
PHP 3 support discontinued: http://bugs.php.net/fix.php?id=28457&r=php3
Daylight Savings:   http://bugs.php.net/fix.php?id=28457&r=dst
IIS Stability:  http://bugs.php.net/fix.php?id=28457&r=isapi
Install GNU Sed:http://bugs.php.net/fix.php?id=28457&r=gnused
Floating point limitations: http://bugs.php.net/fix.php?id=28457&r=float


#28238 [NEW]: --enable-discard-path prevents PHP from being run as CGI with Apache

2004-04-30 Thread floeff at arcor dot de
From: floeff at arcor dot de
Operating system: Linux
PHP version:  4.3.6
PHP Bug Type: CGI related
Bug description:  --enable-discard-path prevents PHP from being run as CGI with Apache

Description:

It seems that --enable-discard-path prevents PHP from being run as CGI
with Apache.

My configure options are as follows:
./configure --enable-safe-mode --with-mm --with-mysql
--enable-discard-path --with-exec-dir --enable-memory-limit
--enable-force-cgi-redirect

PHP has been enabled in Apache with these settings in httpd.conf:


Order allow,deny
Allow from all


ScriptAlias /.php-script/ /usr/local/bin/
AddHandler php-script .php
Action php-script /.php-script/php

This does not work. I receive parse errors for ASCII characters, it seems
that no PHP script is being parsed, but some binary data. As soon as I
remove --enable-discard-path it works fine. I run Apache 2.0.49.

Seems to be related to Bug #18371.

Reproduce code:
---

or any other PHP code.

Expected result:

Parse error.

Actual result:
--
Correct PHP output.

-- 
Edit bug report at http://bugs.php.net/?id=28238&edit=1
-- 
Try a CVS snapshot (php4):  http://bugs.php.net/fix.php?id=28238&r=trysnapshot4
Try a CVS snapshot (php5):  http://bugs.php.net/fix.php?id=28238&r=trysnapshot5
Fixed in CVS:   http://bugs.php.net/fix.php?id=28238&r=fixedcvs
Fixed in release:   http://bugs.php.net/fix.php?id=28238&r=alreadyfixed
Need backtrace: http://bugs.php.net/fix.php?id=28238&r=needtrace
Need Reproduce Script:  http://bugs.php.net/fix.php?id=28238&r=needscript
Try newer version:  http://bugs.php.net/fix.php?id=28238&r=oldversion
Not developer issue:http://bugs.php.net/fix.php?id=28238&r=support
Expected behavior:  http://bugs.php.net/fix.php?id=28238&r=notwrong
Not enough info:http://bugs.php.net/fix.php?id=28238&r=notenoughinfo
Submitted twice:http://bugs.php.net/fix.php?id=28238&r=submittedtwice
register_globals:   http://bugs.php.net/fix.php?id=28238&r=globals
PHP 3 support discontinued: http://bugs.php.net/fix.php?id=28238&r=php3
Daylight Savings:   http://bugs.php.net/fix.php?id=28238&r=dst
IIS Stability:  http://bugs.php.net/fix.php?id=28238&r=isapi
Install GNU Sed:http://bugs.php.net/fix.php?id=28238&r=gnused
Floating point limitations: http://bugs.php.net/fix.php?id=28238&r=float


#28065 [Bgs]: resource limits do not work

2004-04-22 Thread floeff at arcor dot de
 ID:   28065
 User updated by:  floeff at arcor dot de
 Reported By:  floeff at arcor dot de
 Status:   Bogus
 Bug Type: Reproducible crash
 Operating System: Linux 2.4.26
 PHP Version:  4.3.6
 New Comment:

> Don't run PHP as CGI

I need to run PHP as a user, so there only seems to be the CGI method.
Does memory limitation not work for CGI?


Previous Comments:


[2004-04-22 15:11:51] floeff at arcor dot de

> Don't run PHP as CGI

I need to run PHP as a user, so there only seems to be the CGI method.
Does memory limitation not work for CGI?



[2004-04-22 09:29:21] [EMAIL PROTECTED]

1. Don't run Apache 2
2. Don't run PHP as CGI

Do use apache 1.3 with the apache SAPI and this works fine; apache 2 is
NOT ready for production.

----

[2004-04-21 21:36:55] floeff at arcor dot de

I checked some more things, here some more results:

- On my test machine, I've set MaxClients in httpd.conf to 20 and I
couldn't kill the machine anymore. This is only a workaround, as it
knocks out a lot of users on higher load machines, but at least it
shows that MaxClients
works for that.

- When a "normal access" (i.e. not my "attack") accesses a PHP script,
the PHP parser is only shown during the run of the script, checked with
ps auxw | grep php. However, when I "attack", it runs for quite a while
several times. It does seem to use the 300 seconds timeout, as I
lowered this to 10
seconds, and the scripts got killed with "(70007)The timeout specified
has expired: ap_content_length_filter: apr_bucket_read() failed".
However, it does not help to lower it, the load rises even more. Had
about 97 (!) some
seconds ago :-( "killall php" or "killall -9 php" doesn't help.
However, shutting down Apache brings down the PHP interpreters, so I
think that Apache just "forgets" about the running instances when it is
flooded and cannot handle them anymore.

Do you have any idea? Do you need more information?

--------

[2004-04-20 19:05:25] floeff at arcor dot de

If PHP does not terminate after the set amount of time, then for me it
is a bug. Why shouldn't it be one?



[2004-04-19 23:23:34] [EMAIL PROTECTED]

Sorry, but your problem does not imply a bug in PHP itself.  For a
list of more appropriate places to ask for help using PHP, please
visit http://www.php.net/support.php as this bug system is not the
appropriate forum for asking support questions. 

Thank you for your interest in PHP.

.



The remainder of the comments for this report are too long. To view
the rest of the comments, please view the bug report online at
http://bugs.php.net/28065

-- 
Edit this bug report at http://bugs.php.net/?id=28065&edit=1


#28065 [Bgs]: resource limits do not work

2004-04-22 Thread floeff at arcor dot de
 ID:   28065
 User updated by:  floeff at arcor dot de
 Reported By:  floeff at arcor dot de
 Status:   Bogus
 Bug Type: Reproducible crash
 Operating System: Linux 2.4.26
 PHP Version:  4.3.6
 New Comment:

> Don't run PHP as CGI

I need to run PHP as a user, so there only seems to be the CGI method.
Does memory limitation not work for CGI?


Previous Comments:


[2004-04-22 09:29:21] [EMAIL PROTECTED]

1. Don't run Apache 2
2. Don't run PHP as CGI

Do use apache 1.3 with the apache SAPI and this works fine; apache 2 is
NOT ready for production.



[2004-04-21 21:36:55] floeff at arcor dot de

I checked some more things, here some more results:

- On my test machine, I've set MaxClients in httpd.conf to 20 and I
couldn't kill the machine anymore. This is only a workaround, as it
knocks out a lot of users on higher load machines, but at least it
shows that MaxClients
works for that.

- When a "normal access" (i.e. not my "attack") accesses a PHP script,
the PHP parser is only shown during the run of the script, checked with
ps auxw | grep php. However, when I "attack", it runs for quite a while
several times. It does seem to use the 300 seconds timeout, as I
lowered this to 10
seconds, and the scripts got killed with "(70007)The timeout specified
has expired: ap_content_length_filter: apr_bucket_read() failed".
However, it does not help to lower it, the load rises even more. Had
about 97 (!) some
seconds ago :-( "killall php" or "killall -9 php" doesn't help.
However, shutting down Apache brings down the PHP interpreters, so I
think that Apache just "forgets" about the running instances when it is
flooded and cannot handle them anymore.

Do you have any idea? Do you need more information?

------------

[2004-04-20 19:05:25] floeff at arcor dot de

If PHP does not terminate after the set amount of time, then for me it
is a bug. Why shouldn't it be one?



[2004-04-19 23:23:34] [EMAIL PROTECTED]

Sorry, but your problem does not imply a bug in PHP itself.  For a
list of more appropriate places to ask for help using PHP, please
visit http://www.php.net/support.php as this bug system is not the
appropriate forum for asking support questions. 

Thank you for your interest in PHP.

.

------------

[2004-04-19 19:49:35] floeff at arcor dot de

Description:

Maybe I misunderstand something, but for me it seems that the resource
limits do not work. In my php.ini, I set

max_execution_time = 10
max_input_time = 10

which - to my understanding - should limit executing time for scripts
to 10 seconds.

I have PHP running as CGI with Apache 2.0.49 on Linux 2.4.26 here, and
with a hughe PHP file involving some diagram creation, I can kill the
machine if I re-load the script for five seconds continuously. It soaks
up all my memory and runs for nearly a minute.

What could be wrong in here? If you need more information, please let
me know. Thanks!







-- 
Edit this bug report at http://bugs.php.net/?id=28065&edit=1


#28065 [Bgs->Opn]: resource limits do not work

2004-04-21 Thread floeff at arcor dot de
 ID:   28065
 User updated by:  floeff at arcor dot de
 Reported By:  floeff at arcor dot de
-Status:   Bogus
+Status:   Open
 Bug Type: Reproducible crash
 Operating System: Linux 2.4.26
 PHP Version:  4.3.6
 New Comment:

I checked some more things, here some more results:

- On my test machine, I've set MaxClients in httpd.conf to 20 and I
couldn't kill the machine anymore. This is only a workaround, as it
knocks out a lot of users on higher load machines, but at least it
shows that MaxClients
works for that.

- When a "normal access" (i.e. not my "attack") accesses a PHP script,
the PHP parser is only shown during the run of the script, checked with
ps auxw | grep php. However, when I "attack", it runs for quite a while
several times. It does seem to use the 300 seconds timeout, as I
lowered this to 10
seconds, and the scripts got killed with "(70007)The timeout specified
has expired: ap_content_length_filter: apr_bucket_read() failed".
However, it does not help to lower it, the load rises even more. Had
about 97 (!) some
seconds ago :-( "killall php" or "killall -9 php" doesn't help.
However, shutting down Apache brings down the PHP interpreters, so I
think that Apache just "forgets" about the running instances when it is
flooded and cannot handle them anymore.

Do you have any idea? Do you need more information?


Previous Comments:
--------------------

[2004-04-20 19:05:25] floeff at arcor dot de

If PHP does not terminate after the set amount of time, then for me it
is a bug. Why shouldn't it be one?



[2004-04-19 23:23:34] [EMAIL PROTECTED]

Sorry, but your problem does not imply a bug in PHP itself.  For a
list of more appropriate places to ask for help using PHP, please
visit http://www.php.net/support.php as this bug system is not the
appropriate forum for asking support questions. 

Thank you for your interest in PHP.

.

----------------

[2004-04-19 19:49:35] floeff at arcor dot de

Description:

Maybe I misunderstand something, but for me it seems that the resource
limits do not work. In my php.ini, I set

max_execution_time = 10
max_input_time = 10

which - to my understanding - should limit executing time for scripts
to 10 seconds.

I have PHP running as CGI with Apache 2.0.49 on Linux 2.4.26 here, and
with a hughe PHP file involving some diagram creation, I can kill the
machine if I re-load the script for five seconds continuously. It soaks
up all my memory and runs for nearly a minute.

What could be wrong in here? If you need more information, please let
me know. Thanks!







-- 
Edit this bug report at http://bugs.php.net/?id=28065&edit=1


#28065 [Bgs]: resource limits do not work

2004-04-20 Thread floeff at arcor dot de
 ID:   28065
 User updated by:  floeff at arcor dot de
 Reported By:  floeff at arcor dot de
 Status:   Bogus
 Bug Type: Reproducible crash
 Operating System: Linux 2.4.26
 PHP Version:  4.3.6
 New Comment:

If PHP does not terminate after the set amount of time, then for me it
is a bug. Why shouldn't it be one?


Previous Comments:


[2004-04-19 23:23:34] [EMAIL PROTECTED]

Sorry, but your problem does not imply a bug in PHP itself.  For a
list of more appropriate places to ask for help using PHP, please
visit http://www.php.net/support.php as this bug system is not the
appropriate forum for asking support questions. 

Thank you for your interest in PHP.

.



[2004-04-19 19:49:35] floeff at arcor dot de

Description:

Maybe I misunderstand something, but for me it seems that the resource
limits do not work. In my php.ini, I set

max_execution_time = 10
max_input_time = 10

which - to my understanding - should limit executing time for scripts
to 10 seconds.

I have PHP running as CGI with Apache 2.0.49 on Linux 2.4.26 here, and
with a hughe PHP file involving some diagram creation, I can kill the
machine if I re-load the script for five seconds continuously. It soaks
up all my memory and runs for nearly a minute.

What could be wrong in here? If you need more information, please let
me know. Thanks!







-- 
Edit this bug report at http://bugs.php.net/?id=28065&edit=1


#28065 [NEW]: resource limits do not work

2004-04-19 Thread floeff at arcor dot de
From: floeff at arcor dot de
Operating system: Linux 2.4.26
PHP version:  4.3.6
PHP Bug Type: Reproducible crash
Bug description:  resource limits do not work

Description:

Maybe I misunderstand something, but for me it seems that the resource
limits do not work. In my php.ini, I set

max_execution_time = 10
max_input_time = 10

which - to my understanding - should limit executing time for scripts to
10 seconds.

I have PHP running as CGI with Apache 2.0.49 on Linux 2.4.26 here, and
with a hughe PHP file involving some diagram creation, I can kill the
machine if I re-load the script for five seconds continuously. It soaks up
all my memory and runs for nearly a minute.

What could be wrong in here? If you need more information, please let me
know. Thanks!



-- 
Edit bug report at http://bugs.php.net/?id=28065&edit=1
-- 
Try a CVS snapshot (php4):  http://bugs.php.net/fix.php?id=28065&r=trysnapshot4
Try a CVS snapshot (php5):  http://bugs.php.net/fix.php?id=28065&r=trysnapshot5
Fixed in CVS:   http://bugs.php.net/fix.php?id=28065&r=fixedcvs
Fixed in release:   http://bugs.php.net/fix.php?id=28065&r=alreadyfixed
Need backtrace: http://bugs.php.net/fix.php?id=28065&r=needtrace
Need Reproduce Script:  http://bugs.php.net/fix.php?id=28065&r=needscript
Try newer version:  http://bugs.php.net/fix.php?id=28065&r=oldversion
Not developer issue:http://bugs.php.net/fix.php?id=28065&r=support
Expected behavior:  http://bugs.php.net/fix.php?id=28065&r=notwrong
Not enough info:http://bugs.php.net/fix.php?id=28065&r=notenoughinfo
Submitted twice:http://bugs.php.net/fix.php?id=28065&r=submittedtwice
register_globals:   http://bugs.php.net/fix.php?id=28065&r=globals
PHP 3 support discontinued: http://bugs.php.net/fix.php?id=28065&r=php3
Daylight Savings:   http://bugs.php.net/fix.php?id=28065&r=dst
IIS Stability:  http://bugs.php.net/fix.php?id=28065&r=isapi
Install GNU Sed:http://bugs.php.net/fix.php?id=28065&r=gnused
Floating point limitations: http://bugs.php.net/fix.php?id=28065&r=float


#28021 [Fbk->Opn]: user and group of the tar archived files should be set to nobody:nogroup

2004-04-16 Thread floeff at arcor dot de
 ID:   28021
 User updated by:  floeff at arcor dot de
 Reported By:  floeff at arcor dot de
-Status:   Feedback
+Status:   Open
 Bug Type: Feature/Change Request
 Operating System: Linux
 PHP Version:  4.3.6RC3
 New Comment:

http://www.php.net/get/php-4.3.6.tar.bz2/from/a/mirror

and

http://www.php.net/get/php-4.3.6.tar.gz/from/a/mirror


Previous Comments:


[2004-04-16 15:08:47] [EMAIL PROTECTED]

what tar file?





[2004-04-16 05:53:28] floeff at arcor dot de

Description:

The user and group of the tar archived files should be set to
nobody:nogroup. Right now they are 1003:1003, which can cause quota
warning error messages (and more!) on some systems where this user
exist.






-- 
Edit this bug report at http://bugs.php.net/?id=28021&edit=1


#28021 [NEW]: user and group of the tar archived files should be set to nobody:nogroup

2004-04-16 Thread floeff at arcor dot de
From: floeff at arcor dot de
Operating system: Linux
PHP version:  4.3.6RC3
PHP Bug Type: Unknown/Other Function
Bug description:  user and group of the tar archived files should be set to 
nobody:nogroup

Description:

The user and group of the tar archived files should be set to
nobody:nogroup. Right now they are 1003:1003, which can cause quota
warning error messages (and more!) on some systems where this user exist.


-- 
Edit bug report at http://bugs.php.net/?id=28021&edit=1
-- 
Try a CVS snapshot (php4):  http://bugs.php.net/fix.php?id=28021&r=trysnapshot4
Try a CVS snapshot (php5):  http://bugs.php.net/fix.php?id=28021&r=trysnapshot5
Fixed in CVS:   http://bugs.php.net/fix.php?id=28021&r=fixedcvs
Fixed in release:   http://bugs.php.net/fix.php?id=28021&r=alreadyfixed
Need backtrace: http://bugs.php.net/fix.php?id=28021&r=needtrace
Need Reproduce Script:  http://bugs.php.net/fix.php?id=28021&r=needscript
Try newer version:  http://bugs.php.net/fix.php?id=28021&r=oldversion
Not developer issue:http://bugs.php.net/fix.php?id=28021&r=support
Expected behavior:  http://bugs.php.net/fix.php?id=28021&r=notwrong
Not enough info:http://bugs.php.net/fix.php?id=28021&r=notenoughinfo
Submitted twice:http://bugs.php.net/fix.php?id=28021&r=submittedtwice
register_globals:   http://bugs.php.net/fix.php?id=28021&r=globals
PHP 3 support discontinued: http://bugs.php.net/fix.php?id=28021&r=php3
Daylight Savings:   http://bugs.php.net/fix.php?id=28021&r=dst
IIS Stability:  http://bugs.php.net/fix.php?id=28021&r=isapi
Install GNU Sed:http://bugs.php.net/fix.php?id=28021&r=gnused
Floating point limitations: http://bugs.php.net/fix.php?id=28021&r=float