My production server is managed by WHM. I installed apache, python, etc on
a virtual machine to test. I can reproduce the problem on my virtual
machine. The Django app causes the seg fault every time with 4.4 and works
fine with 4.3. This is a pure python program. While developing it with
the Django dev server I didn't see any problems.
Everything except mysql is compiled from source.
For apache cfg I replaced:
#ServerName www.example.com:80 with ServerName host.centos6-02
User daemon with User nobody
Group daemon with Group nobody
I added this to the bottom of the apache cfg:
-------------------------------
LoadModule wsgi_module modules/mod_wsgi.so
WSGIPythonHome /usr/local/python2p7
<VirtualHost *:80>
ServerAdmin [email protected]
ServerName tst01.local
DocumentRoot "/home/brian/server/tst01/public_html"
<Directory "/home/brian/server/tst01/apache">
Require all granted
</Directory>
<Directory /home/brian/server/tst01/public_html>
AllowOverride None
Require all granted
Options Indexes Includes FollowSymLinks
</Directory>
ErrorLog /home/brian/server/tst01/logs/error.log
CustomLog /home/brian/server/tst01/logs/access.log combined
<IfModule mod_alias.c>
Alias /robots.txt
/home/brian/server/tst01/public_html/media/robots.txt
Alias /favicon.ico
/home/brian/server/tst01/public_html/media/favicon.ico
Alias /media /home/brian/server/tst01/public_html/media
Alias /adminmedia
/home/brian/server/tst01/public_html/media/adminmedia
Alias /static /home/brian/server/tst01/public_html/static
</IfModule>
<IfModule mod_wsgi.c>
WSGIScriptAlias / /home/brian/server/tst01/apache/django.wsgi
WSGIDaemonProcess tst01.local threads=15 display-name=%{GROUP}
WSGIProcessGroup tst01.local
WSGIApplicationGroup %{GLOBAL}
</IfModule>
</VirtualHost>
-------------------------------
django.wsgi
-------------------------------
import os, sys
pathList = ['/home/brian/server/tst01/django',
'/home/brian/server/tst01/django/yCrmProj']
for path in pathList:
if path not in sys.path:
sys.path.append(path)
os.environ['PYTHON_EGG_CACHE'] = '/home/brian/server/tst01/.python-eggs/'
#from django default wsgi.py file
os.environ.setdefault("DJANGO_SETTINGS_MODULE", "yCrmProj.settings")
from django.core.wsgi import get_wsgi_application
application = get_wsgi_application()
-------------------------------
I downloaded the wsgi source 4.3 and unpacked to
/usr/src/modWsgi/mod_wsgi-4.3.0
and did the same with 4.4: /usr/src/modWsgi/mod_wsgi-4.4.0.
To downgrade to 4.3 I did:
rm -f /usr/local/apache/modules/mod_wsgi.so
cd /usr/src/modWsgi/mod_wsgi-4.3.0
make install
httpd -k graceful
Then I did this to be sure apache got the correct version:
service httpd status
I'll look into setting up GDB and post back about it.
Thanks for the quick response!!!
Brian
On Friday, December 12, 2014 12:19:47 AM UTC-5, [email protected] wrote:
>
> When I upgraded from 4.3.0 to 4.4.0 I've started to get errors “Truncated
> or oversized response headers received from daemon process” and seg faults.
>
>
> I downgraded and everything went back to working.
>
>
> I'm seeing this on a Django app that uses AJAX to send to a custom view
> in the admin. I've also seen this with a bot that is claiming to be Google.
> But since it is only visiting one of my sites I'm expect it's not from
> Google.
>
>
> I'm using Centos 6.6, python 2.7.8, and Apache 2.4.10.
>
>
> Is there a new setting I need to change or is this a bug? I didn't see
> anything in the change log that I need to change
>
--
You received this message because you are subscribed to the Google Groups
"modwsgi" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To post to this group, send email to [email protected].
Visit this group at http://groups.google.com/group/modwsgi.
For more options, visit https://groups.google.com/d/optout.
