Guy Rouillier wrote:
The application is fairly straightforward, but as you say, what is working okay with BigDBMS isn't working as well under PG. I'm going to try other configuration suggestions made by others before I attempt logic changes. The core logic is unchangeable; millions of rows of data in a single table will be updated throughout the day. If PG can't handle high volume updates well, this may be brick wall.
I understand your reluctance to change your working design in the change over to PostgreSQL but -
1. Your table definitions may or may not be the issue and a small change in design (even only choice of datatype) may be all that is needed to get the needed performance out of PostgreSQL. These changes would be done before you put PostgreSQL into production use so the amount of current usage is not relevant when deciding/analyzing these changes but they may affect your ability to use PostgreSQL as an alternative.
2. I think that the idea of logic changes suggested earlier was more aimed at your select/update commands than the structure of your tables. You should expect to have some SQL changes between any database and using select/update's designed to take advantage of PostgreSQL strengths can give you performance improvements.
-- Shane Ambler pgSQL@xxxxxxxxxxxxxxxx Get Sheeky @ http://Sheeky.Biz
