Script timed out before returning headers
Fini Decima
I have an installation of MediaCore, a media serving application built atop the Django framework I have i running on Apache2 using mod_wsgi.
Sometimes it works, other times, it also works, but it takes forever to load, and I get some error messages in Apache's error file like these: =
<code>
[error] [client IP-ADDR] Script timed out before returning headers: mediacore.wsgi, referer: http://IP-ADDR/
[error] [client IP-ADDR] Script timed out before returning headers: mediacore.wsgi, referer: http://IP-ADDR/
[error] [client IP-ADDR] Script timed out before returning headers: mediacore.wsgi, referer: http://IP-ADDR/
[error] [client IP-ADDR] Script timed out before returning headers: mediacore.wsgi, referer: http://IP-ADDR/
[error] [client IP-ADDR] Script timed out before returning headers: mediacore.wsgi
[error] [client IP-ADDR] Script timed out before returning headers: mediacore.wsgi, referer: http://IP-ADDR/
[error] [client IP-ADDR] Script timed out before returning headers: mediacore.wsgi, referer: http://IP-ADDR/
[info] [client IP-ADDR] (70007)The timeout specified has expired: core_output_filter: writing data to the network
[info] [client IP-ADDR] (70007)The timeout specified has expired: core_output_filter: writing data to the network
[error] [client IP-ADDR] mod_wsgi (pid=25066): Exception occurred processing WSGI script '/var/www/mediacore.wsgi'.
[error] [client IP-ADDR] IOError: failed to write data
[info] [client IP-ADDR] (32)Broken pipe: core_output_filter: writing data to the network
[info] [client IP-ADDR] (32)Broken pipe: core_output_filter: writing data to the network
</code>
Even as I'm writing this, a video I clicked several minutes ago is still trying to load.
Here's my media.wsgi file
<code>
activate_this = '/var/www/virtualECK/bin/activate_this.py'
execfile(activate_this, dict(__file__=activate_this))
deployment_config = '/var/www/deployment.ini'
temp_dir = '/var/www/data/tmp'
# NOTE: Before running MediaCore, you will need to update the two paths
# above to point to the appropriate locations for your installation.
import os
os.environ['TMPDIR'] = temp_dir
if __name__.startswith('_mod_wsgi_'):
# Set up logging under mod_wsgi
from paste.script.util.logging_config import fileConfig
fileConfig(deployment_config)
# Load the app!
from paste.deploy import loadapp
application = loadapp('config:'+deployment_config)
</code>
And here's my VirtualHost configuration:
<code>
<VirtualHost *:80>
ServerAdmin *
DocumentRoot /var/www
# For best performance the number of processes should equal the number of CPU
# cores (but please note that each process may use about 500 MB RAM).
WSGIDaemonProcess eckmedia user=www-data group=www-data \
processes=1 \
threads=1 \
display-name=%{GROUP} \
python-path=/var/www/virtualECK/lib/python2.7/site-packages \
python-eggs=/var/www/data/python-egg-cache
WSGIProcessGroup eckmedia
# Intercept all requests to /* and pass them to mediacore.wsgi
WSGIScriptAlias / /var/www/mediacore.wsgi process-group=eckmedia application-group=%{GLOBAL}
<Directory />
Options FollowSymLinks
AllowOverride None
</Directory>
<Directory /var/www/>
Options Indexes FollowSymLinks MultiViews
AllowOverride None
Order allow,deny
allow from all
</Directory>
# Make all the static content accessible.
<Directory /var/www/eckmedia/mediacore/public/*>
Order allow,deny
Allow from all
Options -Indexes
WSGIScriptReloading On
</Directory>
ErrorLog ${APACHE_LOG_DIR}/error.log
LogLevel info
CustomLog ${APACHE_LOG_DIR}/access.log combined
</code>
And in Apache2.conf, I have:
<code>
# prefork MPM
<IfModule mpm_prefork_module>
StartServers 7
MinSpareServers 5
MaxSpareServers 10
MaxClients 150
MaxRequestsPerChild 0
</IfModule>
Fini Decima
Graham Dumpleton
--
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 modwsgi+u...@googlegroups.com.
To post to this group, send email to mod...@googlegroups.com.
Visit this group at http://groups.google.com/group/modwsgi.
For more options, visit https://groups.google.com/groups/opt_out.
Graham Dumpleton
You have a very poorly setup configuration. Your WSGI application can only handle one request at a time. If one request takes a long time to run, it will block out everything else and if it blocks for too long, it will timeout.I would thus very much suggest not using processes=1 and threads=1 for daemon mode.You could also be suffering issues due to third party C extension modules which will not working in sub interpreters and which block and cause process to hang, thus invoking mod_wsgi deadlock timeout.I would suggest using:WSGIApplicationGroup %{GLOBAL}