2009/6/25 Andrew Fong <fongand...@gmail.com>: > > To clarify, you're displaying a list of 1000+ models on a single page, > yes? >
Yes, i will change that as soon as I could. > My understanding is that the debug_toolbar analyzes each SQL query and > keeps each result in memory before rendering everything in the > template. I don't know how your query works, but let's take an extreme > case and say your request executes 1000+ queries. In addition to > keeping 1000 models in memory, the debug_toolbar now has to also > contain the query data for 1000 queries. If you have a small amount of > memory, this could eat it up pretty quick. > I understand that. But, what's bother me is that this very same list have worked well before (with an average of two seconds of loading) on the same server. Indeed when I has activated debug_toolbar on my PC, lists were very slow to display (about ten seconds and more) but there was not a Memory error. My validation server is virtualized. I will see with my admin the amount of memory on it. > I don't think setting MaxRequestsPerChild = 1 would do anything to > cause a MemoryError. It should actually help, since if you're > restarting the process after just 1 request, any memory leaks won't > have time to build up. No, indeed. A MaxRequestPerChild set to 10000 don't change anything Many thanks for your reply, > > -- Andrew > Frédéric > On Jun 25, 1:15 pm, Frédéric Hébert <fg.heb...@gmail.com> wrote: >> Hi there, >> >> I'm facing with a curious problem. >> >> I've an project which is hosted on three differents servers : >> >> one (my local PC) : >> debian lenny / >> python 2.5.2 / django 1.0.2 final >> apache 2.2.9 / mod_python 3.3.1-7 >> >> the two others (production and validation server) : >> OpenSUSE 11.1 / >> python 2.6.2 / django 1.0.2 final >> apache 2.2.10 / mod_python 3.3.x-x >> >> Two months ago, I've launched this project both on production server >> and validation server. >> Neither validation server nor production server caused me any >> trouble. >> >> Today, to test something, I've installed debug_toolbar extension and >> set MaxRequestPerChild to 1 in Apache's conf file, both on my local >> dev PC and on the validation server. >> >> When I want to display two long list (1000+ items each) i get a >> MemoryError one time out of two. The remainder of project goes fine. >> >> I've resetted MaxRequestPerChild to 10000 (default value on OpenSUSE) >> and commented out debug_toolbar's things in my settings, restart >> Apache. But error happened again on the validation server. >> >> On my local PC, it works well, with debug_toolbar and MaxReq.. set or >> not. >> >> I've pasted the traceback herehttp://dpaste.com/hold/59667/ >> >> AFAICS, an exception occurs when django tries to render a debug >> template >> >> Here is an exerpt of my vhost conf file on OpenSUSE : >> DocumentRoot /srv/www/enseignements-dev.ehess.fr/ue >> >> <Location "/"> >> SetHandler python-program >> PythonHandler django.core.handlers.modpython >> SetEnv DJANGO_SETTINGS_MODULE ue.settings_dev >> PythonDebug On >> PythonPath "['/srv/www/enseignements-dev.ehess.fr/', '/srv/www/ >> enseignements-dev.ehess.fr/ue/'] + sys.path" >> </Location> >> >> Many thanks in advance, >> >> Frédéric > > > -- http://www.openidfrance.fr/fhebert --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "Django users" group. To post to this group, send email to django-users@googlegroups.com To unsubscribe from this group, send email to django-users+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/django-users?hl=en -~----------~----~----~----~------~----~------~--~---
