4

I have a problem with my web application. Sometimes when people upload an image (seems to happen with large image only), I have mod_wsgi who crash with this message :

End of script output before headers: wsgi.py, referer:http:::...
Process 'wsgi' has died, restarting.
Process 'wsgi' terminated by signal 9

and so an internal error is raised.

Here is my wsgi file:

import os
os.environ.setdefault("DJANGO_SETTINGS_MODULE", "settings")

from django.core.wsgi import get_wsgi_application
application = get_wsgi_application()

and here is the wsgi conf file

LoadModule wsgi_module modules/mod_wsgi.so
WSGIPythonHome /opt/python/run/baselinenv
WSGISocketPrefix run/wsgi
WSGIRestrictEmbedded On

<VirtualHost *:80>

Alias /static/ /opt/python/current/app/static/
<Directory /opt/python/current/app/static/>
Order allow,deny
Allow from all
</Directory>


WSGIScriptAlias / /opt/python/current/app/wsgi.py


<Directory /opt/python/current/app/>
  Require all granted
</Directory>

WSGIDaemonProcess wsgi processes=1 threads=15 display-name=%{GROUP} \
  python-path=/opt/python/current/app:/opt/python/run/venv/lib/python2.7/site-packages user=wsgi group=wsgi \
  home=/opt/python/current/app
WSGIProcessGroup wsgi
</VirtualHost>

Do you know what is the problem and what I can do to solve it ?

Please note that I cannot migrate to nginx and gunicorn.

Thanks a lot

7
  • What version of mod_wsgi are you using? Does your system run a monitor which will kill processes if they use too much memory? Have you monitored the memory usage of the processes to see if they are growing to a very large value? Can you upgrade to the latest mod_wsgi version and try again? Apr 16, 2015 at 4:14
  • Hello, version I use is mod24_wsgi-py27.x86_64, 3.4-8.14.amzn1, I have tried to update, but it seems to be the last one. When monitoring with top, I do not see any over usage of memory :(
    – trnsnt
    Apr 16, 2015 at 7:48
  • Set 'LogLevel info' in the Apache configuration. This will cause mod_wsgi to log messages about why daemon processes are being restarted. This ail show whether the kill -9 is coming about because of a restart somehow triggered by mod_wsgi, or some external monitoring service. So change the Apache log level and then capture mod_wsgi log messages leading up to and after the messages you originally quoted. Apr 17, 2015 at 0:05
  • The messages I originally quoted were already with info log level.
    – trnsnt
    Apr 17, 2015 at 13:52
  • Can you provide greater context from the log so it is possible to see messages before and after that point then related to that event in time? Apr 17, 2015 at 20:54

0

You must log in to answer this question.

Browse other questions tagged .