Nevermind. I just found your post on this issue at http://code.google.com/p/modwsgi/wiki/ConfigurationIssues
On Sun, Sep 14, 2008 at 11:32 PM, Chris Spencer <[EMAIL PROTECTED]> wrote: > Ok, so this is what I put at the bottom of my httpd.conf: > > WSGIDaemonProcess myapp processes=1 threads=25 > WSGIProcessGroup myapp > Alias /myapp/static /usr/local/django/myapp/static/ > <Directory /usr/local/django/myapp/static/> > Order deny,allow > Allow from all > </Directory> > WSGIScriptAlias /myapp /usr/local/django/myapp/apache/django.wsgi > <Directory /usr/local/django/myapp/apache> > Order deny,allow > Allow from all > </Directory> > > Everything in /myapp/static/* loads fine. However, all other requests > that got to django.wsgi generate this error to my Apache log: > > [Sun Sep 14 23:26:09 2008] [error] [client 127.0.0.1] (13)Permission > denied: mod_wsgi (pid=344): Unable to connect to WSGI daemon process > 'myapp' on '/etc/httpd/logs/wsgi.28572.2.1.sock' after multiple > attempts. > > Any thoughts? The above is what the docs use as the standard recipe > for running Django with mod_wsgi. Am I missing anything? > > Regards, > Chris > > On Thu, Sep 11, 2008 at 9:06 PM, Graham Dumpleton > <[EMAIL PROTECTED]> wrote: >> >> >> >> On Sep 12, 10:35 am, "Chris Spencer" <[EMAIL PROTECTED]> wrote: >>> On Thu, Sep 11, 2008 at 6:16 AM, Graham Dumpleton >>> >>> >>> >>> <[EMAIL PROTECTED]> wrote: >>> > Be aware that Apache/mod_wsgi in embedded mode on UNIX is going to be >>> > multiprocess. Thus where you think it is being loaded on every >>> > request, it is more likely just the result of the various processes >>> > loading the application the first time it is used. After they are all >>> > loaded, you shouldn't see loading occurring. >>> >>> > Anyway this is it in simple terms, as it is actually more complicated >>> > than that as Apache can kill off processes and replace them in certain >>> > situations. >>> >>> > For some details of how processes are used in Apache/mod_wsgi see: >>> >>> > http://code.google.com/p/modwsgi/wiki/ProcessesAndThreading >>> >>> > To know what is definitively what is going on, following instructions >>> > in: >>> >>> > http://code.google.com/p/modwsgi/wiki/DebuggingTechniques >>> >>> > and set: >>> >>> > LogLevel info >>> >>> > in Apache configuration, in place of default 'warn'. This should >>> > result in mod_wsgi outputing a lot of information into main and per >>> > virtual host, as appropriate, error logs about when processes are >>> > being started/restarted and when WSGI application scripts being >>> > loaded. >>> >>> > So do that and report on what you see, including examples of Apache >>> > error logs which you believe shows the behaviour you are claiming. >>> >>> I set LogLevel, and after each request, the only thing I'm seeing in >>> the log is something like: >>> [Thu Sep 11 20:17:37 2008] [info] mod_wsgi (pid=18858): Create >>> interpreter 'localhost.localdomain|/myapp'. >>> [Thu Sep 11 20:17:37 2008] [info] [client 127.0.0.1] mod_wsgi >>> (pid=18858, process='', application='localhost.localdomain|/myapp'): >>> Loading WSGI script '/usr/local/django/myapp/apache/django.wsgi'., >>> referer:http://localhost/myapp >>> >>> Is there any way to force a specific WSGI app to run on a single >>> process, without changing the number of processes used otherwise on >>> the server? I've toyed around with WSGIDaemonProcess, but all this >>> seems to do is break the debugger without fixing the problem. Would I >>> use something like: >>> >>> WSGIDaemonProcess myapp processes=1 threads=1 >> >> Which debugger? Any specific reason you are using a single thread, >> such as using older Django version or your application is not >> multithread? >> >> Anyway, the in web browser debuggers I know of will only work if one >> process, but 'processes=1' still marks it as multiprocess in WSGI >> environment. Just drop 'processes=1' and let if use default of 1 >> process. If you don't do that the debugger may complain it is not >> running in valid environment. Thus: >> >> WSGIDaemonProcess myapp thread=1 >> WSGIProcessGroup myapp >> >> For more information on web browser debuggers and debugging in general >> see: >> >> http://code.google.com/p/modwsgi/wiki/DebuggingTechniques >> >> Also see comments about WSGI multiprocess/multithread flags in: >> >> http://code.google.com/p/modwsgi/wiki/ProcessesAndThreading >> >> For configuration of WSGIDaemonProcess see examples in: >> >> http://code.google.com/p/modwsgi/wiki/QuickConfigurationGuide >> http://code.google.com/p/modwsgi/wiki/ConfigurationGuidelines >> http://code.google.com/p/modwsgi/wiki/ConfigurationDirectives >> >> Sorry, but have to run somewhere. Will check later than I actually >> answered your question. I may not have. ;-) >> >> Graham >> >>> How does that directive differentiate between multiple WSGI apps >>> running on the same server? The docs mention that it specifies a >>> "process group", but how do you make a single WSGI Django app run in >>> that process group? >>> >>> > Also, you should still post the configuration you use so we can >>> > confirm it is correct. I have many times had people say they followed >>> > the documentation, but in fact they had tweaked it slightly and >>> > inadvertently caused problems for themselves. >>> >>> What I posted is my exact configuration. It's a simple minimal case >>> app for testing this problem. >>> >>> Regards, >>> Chris >> >> >> > --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "Django users" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/django-users?hl=en -~----------~----~----~----~------~----~------~--~---
