Re: pro C versus PL/SQL
From: Pablo Sanchez <pablo_at_dev.null>
Date: Tue, 25 Mar 2003 08:02:58 -0600
Message-ID: <Xns934947B65FBC7pingottpingottbah_at_216.166.71.233>
Date: Tue, 25 Mar 2003 08:02:58 -0600
Message-ID: <Xns934947B65FBC7pingottpingottbah_at_216.166.71.233>
lolitagroux_at_hotmail.com (Lolita) wrote in news:8027ad1.0303231736.5f93cf4d_at_posting.google.com:
> The code does the following:
> 1- Select the data from the database
> 2- For each record seleted it looks through a serie of elsif
> statement, if the record meets the criteria of one elsif statement,
> the data of this record is updated into the database. There can be
> up to 300 elsif statements.
The algorithm above is a classic 'row at a time' processing. This is one of the slowest ways to process data.
I would suggest that you change it to handle sets of data and you'll find that it'll take about 1/10 the time it does now.
-- Pablo Sanchez, High-Performance Database Engineering http://www.hpdbe.comReceived on Tue Mar 25 2003 - 15:02:58 CET