ID: 36724
User updated by: mberg at synacor dot com
Reported By: mberg at synacor dot com
-Status: Feedback
+Status: Open
Bug Type: Apache related
Operating System: Red Hat Linux 7.2 (ish)
PHP Version: 4.4.2
Assigned To: tony2001
New Comment:
Same behaviour. Same fix. Tested on CentOS 4.2 as well. Same
behaviour. Same fix.
Previous Comments:
------------------------------------------------------------------------
[2006-03-16 08:29:33] tony2001@php.net
I didn't say you need to move all your servers to 5.1.
I just asked you to test it and see if the same happens with 5.1 for
you, because I'm unable to reproduce it.
------------------------------------------------------------------------
[2006-03-15 18:28:05] mberg at synacor dot com
We have no plans to move to PHP 5.x in the near future.
------------------------------------------------------------------------
[2006-03-15 18:15:10] tony2001@php.net
Please try using this CVS snapshot:
http://snaps.php.net/php5.1-latest.tar.gz
For Windows:
http://snaps.php.net/win32/php5.1-win32-latest.zip
------------------------------------------------------------------------
[2006-03-15 17:47:47] mberg at synacor dot com
We're always getting a value of 0 logged for %{mod_php_memory_usage}
using stock 4.4.2 with Apache 1.3.34. It worked as expected with
4.3.11.
Reverting the change introduced in bug #35646 with the following patch
fixed it for us:
--- php-4.4.2.orig/Zend/zend_alloc.c Sun Jan 1 08:46:49 2006
+++ php-4.4.2/Zend/zend_alloc.c Fri Mar 10 16:59:18 2006
@@ -541 +540,0 @@
- AG(allocated_memory_peak) = 0;
This presumably breaks Apache 2.x though (_that_ I haven't tested),
which is where the second part of my suggested fix comes in. This is
just clearing the value in the bit of code in apache2handler that's
analagous to where it was cleared in mod_php4.c.
------------------------------------------------------------------------
[2006-03-15 17:33:10] tony2001@php.net
I still don't get it.
Did you test it? Does it work for you?
The current way is to zero twice:
1) it in the beginning of the request, when memory manager starts.
2) in the end of the request, when memory manager goes down.
What makes you think it's wrong if you didn't even test it?
------------------------------------------------------------------------
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/36724--
Edit this bug report at http://bugs.php.net/?id=36724&edit=1