Hi Graham,

So, I took your advice and configured my httpd.conf to allow a daemon
process to be ran.

<VirtualHost 127.0.0.1>
ServerName www.blah.local
CustomLog /Users/blah/dev/logs/blahcustomlog common
ErrorLog /Users/blah/dev/logs/blaherrorlog

WSGIDaemonProcess www.blah.local user=www group=www
processes=2 threads=25
WSGIReloadMechanism Process

</VirtualHost>

As you can see I don't have the WSGIProcessGroup directive, I hope
that's not necessary.  I had initially put it in
WSGIProcessGroup www.blah.local
However was gettng a Internal Server Error to take it out.

Anyways, I made a change to an outside module to test. Ran the
command, touch index.py (my main script file),
and checked to see if the source code change would populate.  Sadly,
it didn't... =( ... Have you any idea of what
could be the problem?

Here's an example:

modules1/

modules2/

index.py

If I change a module in modules1/ and a module from modules2/ uses it,
then index.py uses that module from modules2/ i'd like for that change
to populate
upon request of the WSGI application.

Thanks,

Ian

On Sep 28, 11:51 pm, "Graham Dumpleton" <[EMAIL PROTECTED]>
wrote:
> 2008/9/29 MilesTogoe <[EMAIL PROTECTED]>:
>
>
>
> > ianmendiola wrote:
> >> Ok,
>
> >> So this has definitely isolated the problem.  A lot of the work I am
> >> doing is done in separate python modules that have been imported into
> >> the main script file.  In regards to the example of editing
> >> settings.py in django and having to restart apache for affects to take
> >> place, does this mean every time I make a change to outside modules I
> >> have to restart apache?
>
> > yeah, I'm finding that on WebFaction when you make changes to settings,
>
> I think because by default they have things run in embedded mode.
>
> > etc  it requires a stop and start (they advise against restart).
>
> I certainly would be cautious about using 'restart' a lot as some
> versions of Python leak memory when interpreter being destroyed.
> Normally this wouldn't be an issue for command line Python, but in
> mod_wsgi where new interpreter instance is then created within the
> same running process, this means that that memory leak occurs in
> Apache parent process and thus would be inherited from Apache child
> processes. Thus, use 'restart' or 'graceful' a lot and your Apache
> processes will grow in size.
>
> I have started looking at this memory leak, but because it is in
> Python itself, not sure how much I can do about it with the way that
> mod_wsgi is currently implemented. Details can be found at:
>
>  http://code.google.com/p/modwsgi/issues/detail?id=99
>
> Also, a 'graceful' restart is also a problem at the moment, although
> effect of the problem is limited. For this, when you do a 'graceful'
> restart at any point, it would seem that orderly shutdown of Apache
> daemon mode processes isn't occurring. The process still get killed
> off, but any Python callbacks registered with atexit module aren't
> being called. Details can be found at:
>
>  http://code.google.com/p/modwsgi/issues/detail?id=102
>
> Graham
>
> >   I'm
> > not sure what kind of caching goes on with the shared environment.
> > There also seem to be some quirks with routing  with the extra user user
> > site application level  (ie with a django setup the levels are home /
> > username / webapps / appname "django" / project name / then django app
> > name ).
--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups 
"modwsgi" 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/modwsgi?hl=en
-~----------~----~----~----~------~----~------~--~---

Reply via email to