On Apr 13, 10:59 pm, "Luciano Adamiak" <[EMAIL PROTECTED]> wrote: > It's for the whole thing... I guess > > But it will prevent your restarting every change..
But it still happens automatically and as a result everything runs very slowly. At that point you may as well use CGI which may even be a bit quicker as you aren't restarting all Apache processes, you instead are only incurring the cost of fork/execing one new process instead. FWIW, in mod_wsgi you can optionally enable a mode whereby when the main WSGI application scripts modification time changes, it will destroy and then recreate the actual sub interpreter within the existing running process leaving alone any other sub interpreters and any other Apache modules. The idea is you make the changes you need, touch the main script file and the reload of the whole Django application and all modules it uses will happen automatically on the next request, all without having to restart Apache. Whether this mechanism in mod_wsgi will work for you with Django depends a bit on what additional third party Python C extension modules you use, especially database modules. A basic Django application with sqlite seems to handle interpreter reloading okay but more complex database modules may have issues. As an example, neither Pylons or TurboGears can handle interpreter reloading because they both rely on PyProtocols which uses some pyrex generated modules. These pyrex generated modules will cause Apache to crash if a sub interpreter is destroyed and a new one created and the pyrex generate module reused. It isn't known yet whether the problem in this case is with how pyrex generates code or how PyProtocols uses it. Either way, the basic problem is the C extension module code isn't correct for working with multiple sub interpreters and doesn't handle destruction of sub interpreters. If reloading interpreters doesn't work for your specific application, the only other choice at this point is to use something like mod_fastcgi. At least that way when you restart the process, you are only restarting the process the Python code is running in and you will not affect other stuff running in Apache. The use of separate daemon processes for handling just Python requests is being add to mod_wsgi, but still a bit more work to complete it. Graham > On 4/13/07, durumdara <[EMAIL PROTECTED]> wrote: > > > > > Hi! > > > On Apr 13, 2:00 pm, "Luciano Adamiak" <[EMAIL PROTECTED]> > > wrote: > > > Just set "MaxRequestsPerChild 1" in your httpd.conf > > > And this parameter is used for whole webserver, or I can set it by > > virtualhosts? > > (The last solution is better, because the server is handle 15 site... > > (python and php)). > > > dd --~--~---------~--~----~------------~-------~--~----~ 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 -~----------~----~----~----~------~----~------~--~---
