Essentially I think the problem is this...
I am trying to do an ORDER BY on a pre-populated (and indexed) column
called random_number... I created this column because RAND() is so
slow...
It is the ORDER BY which is causing all the slow downs... If I do a
select without the order by it is
Really what I need to do is "re-order" the database rows (for good)
by the random_number column... note.. I am not using RAND() at all...
Hey, this gives me an idea...
Given I have price ranges 1-2, 2-5, 5-10 etc... and the prices are
all like 1.99, 1.96 etc... What if I precalculate a value
On 18 Nov 2006, at 19:35, Jeremy Dunck wrote:
> This isn't a Django question, but indexes only work well when there's
> a large distribution of values. If lots of your products are in that
> price range, you're still ordering by loads of randoms. Try assigning
> (as a column value) a random int
On 18 Nov 2006, at 18:59, tonemcd wrote:
>
> Perhaps something from here will help?
>
> http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/59883
>
> Where your list is an xrange(min...max) of id's from your database
> table. That is, do it in python, not with your database...
Ah sorry, for
Tom Smith wrote:
> Hello.
>
> When in mysql I do a...
>
> explain select title from burningahole_product where ( price >10 and
> price<20) order by random_number limit 10;
>
>
> ... I get...
>
> ++-+--+---
> ++-
On 11/18/06, Tom Smith <[EMAIL PROTECTED]> wrote:
>
> Hello.
>
> When in mysql I do a...
>
> explain select title from burningahole_product where ( price >10 and
> price<20) order by random_number limit 10;
This isn't a Django question, but indexes only work well when there's
a large distribution
Perhaps something from here will help?
http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/59883
Where your list is an xrange(min...max) of id's from your database
table. That is, do it in python, not with your database...
Or maybe this, http://www.petefreitag.com/item/466.cfm - which *does
7 matches
Mail list logo
