rbartz

Thank you for a great contribution.

I added the latest version (5) tp a site today, on Fedora Core 10, Apache 2.1.4, PHP 5.2.9, OSC 2.2-MS2. After installation, the apache server began throwing segmentations faults in the child processes. Within a few minutes the server load had increased significantly to the point of slowdown. Fedora was unable to kill the fault processes. Restarting httpd solved the problem for a few minutes, but as soon as any hits started on that site, the segmentation faults began again and overloading soon occurred.

Here are a few lines from the httpd error_log
[Mon Jan 02 03:05:20 2012] [notice] child pid 3192 exit signal Segmentation fault (11)
[Mon Jan 02 03:06:00 2012] [notice] child pid 3194 exit signal Segmentation fault (11)
[Mon Jan 02 03:07:21 2012] [notice] child pid 3191 exit signal Segmentation fault (11)
I tried stopping zend eaccelerator which was not the problem. I reset the .htaccess back to what it was before which was not the problem. I tried commenting out the require line first in the catalog and then in the admin, neither stopped the segmentation faults but they slowed. Finally I commented it out in both and the segmentation faults stopped. It seems something in the code itself is causing the faults.

I also have another site on a Centos 5 server, Apache 2.2.3, PH 5.1.6, OSC v2.2 RC2a so added it there. I got the same errors:

[Mon Jan 02 04:14:17 2012] [notice] child pid 20825 exit signal Segmentation fault (11)
[Mon Jan 02 04:15:11 2012] [notice] child pid 21063 exit signal Segmentation fault (11)
[Mon Jan 02 04:17:03 2012] [notice] child pid 21033 exit signal Segmentation fault (11)
On the Centos server, it handled the faults better and was able to kill then with SIGTERM. Again, the faults stopped when I commented out the osc_sec required lines in both application_top.php files.

Any ideas about what may be wrong? Thanks for any help you can give.

Richard