openerp-expert-framework team mailing list archive

[Ferdinand Gassauer <office@xxxxxxxxxx>]

Am Dienstag 12 Januar 2010 14:02:07 schrieb Albert Cervera i Areny:
> - Checking available stock requires a sequential scan on stock_move table 
> which may be really large.
> 
IMHO the concept of reading stock_move to get product_qty has to be redesigned 
too introducing period sums
(I have done such a module already for account_moves which I will publish 
shortly)

This reduces the amount of resources to be read to 13 (balance + 12 periods) 
per year and account/product to get  the total.

I am not a python / ORM / OpenERP / postgres expert to judge if the current 
solution is sufficiently fast in a multi user environment with millions or 
moves.

We had already corresponding discussions with Fabien who showed some 
impressive performance values for access of account tree. 

Nevertheless I believe that the overall performance should increase 
significantly if the number of record reads is reduced significantly, 
especially if postgres and OpenERP server are running on different logical or 
physical servers

-- 
Best Regards

ChriCar Beteiligungs- und Beratungs- GmbH
http://www.chricar.at/ChriCar/index.html
Dr. Ferdinand Gassauer
Official Tiny Partner