Optimizer methods...

We have been experimenting with rule-based and cost-based over the last couple of weeks and have noticed conflicting results. After analyzing a schema (with compute), we ran our typical PeopleSoft processes and they were incredibly slow. It would revert to full scans when it was should of used a concatenated index on a large table in a nested query. However, in some situations, cost-based dramatically increase performance (we have only noticed this on the NT so far). We are running 7.3.2 (NT) and 7.2.3 (AIX). Any input would be greatly appreciated.

-greg.jackson
gjackson_at_pier1.com Received on Mon Jul 14 1997 - 00:00:00 CDT