Thoughts/experience reordering gl_code_combinations raw data to improve clustering factor and perfor
Hi,
I'm curious on if anyone has any experience/thoughts they would be interested in sharing, on reordering the raw data in gl_code_combinations to align with the ordering approach of building the gl_code_combinations_cat index. The goal would be to improve performance whenever the gl_code_combinations_cat (a recommended concatenated index) is used.
For background, the gl_code_combinations_cat index (described in MOS Note 858725.1) is a concatenated index recommended by Oracle development that is built with the most selective segment# columns in descending order. Per 858725.1, an example is:
Consider that gl_code_combinations has the following number of distinct values, per segment: