The main problem with depth=1 seems to be, that the query generator
still spans over too
many tables (watching the mysql.log i can see tables fully joined that
definitvly do not
belong to level 1). Thus the returned data set will surely confuse
django.
Looking at django/db/models/query.py reveals:
def fill_table_cache(opts, select, tables, where, old_prefix,
cache_tables_seen, max_depth=0, cur_depth=0):
"""
Helper function that recursively populates the select, tables and
where (in
place) for select_related queries.
"""
# If we've got a max_depth set and we've exceeded that depth, bail
now.
if max_depth and cur_depth > max_depth:
return None
[ ... ]
fill_table_cache(f.rel.to._meta, select, tables, where,
db_table, cache_tables_seen, max_depth, cur_depth+1)
So by just looking at it one can see that with max_depth=1 and
cur_depth initialized to zero the code will recurse
into level 2, generating a too broad query.
declaring
def fill_table_cache(opts, select, tables, where, old_prefix,
cache_tables_seen, max_depth=0, cur_depth=1):
fixes things for me. Interestingly enough the same observation doesnt
seem to yield for
def get_cached_row(klass, row, index_start, max_depth=0, cur_depth=0):
It seems to be neccessary within this function to go one level
deeper...
I am commenting rev 4661.
Boris
On 11 Jan., 13:52, "David Cramer" <[EMAIL PROTECTED]> wrote:
> It still seems to have a bug when just doing .select_related(depth=1),
> sometimes its filling the field w/ the wrong data, looking into it.
>
> On Jan 11, 11:31 am, "David Cramer" <[EMAIL PROTECTED]> wrote:
>
> > I retract my statement, we had some other random server configuration
> > error that just happened to appear at the same time as us putting this
> > live :)
>
> > On Jan 10, 8:34 pm, "David Cramer" <[EMAIL PROTECTED]> wrote:
>
> > > It seems there are some issues with the code. Although I'm not sure why
> > > it's happening, this was causing servers to get extremely loaded, and
> > > the SQL queries werent executing properly. I had no issues when running
> > > it on runserver before we pushed it onto a live environment though.
>
> > > Going to look into it some more tomorrow, if anyone has any ideas let
> > > me know.
>
> > > On Jan 10, 5:45 pm, Michael Radziej <[EMAIL PROTECTED]> wrote:
>
> > > > David Cramer schrieb:
>
> > > > > I can do the docs, but it'd be a great time saver if you could do the
> > > > > tests. I'm up to my elbows in work at the moment so :)Fine for me,
> > > > > then I take the tests.
>
> > > > Michael
>
> > > > --
> > > > noris network AG - Deutschherrnstraße 15-19 - D-90429 Nürnberg -
> > > > Tel +49-911-9352-0 - Fax +49-911-9352-100
>
> > > >http://www.noris.de-TheIT-OutsourcingCompany
--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups
"Django developers" group.
To post to this group, send email to django-developers@googlegroups.com
To unsubscribe from this group, send email to [EMAIL PROTECTED]
For more options, visit this group at
http://groups.google.com/group/django-developers?hl=en
-~----------~----~----~----~------~----~------~--~---
