SQL Catalog
Next, I needed to gain some detailed information about what they were doing and how they were doing it. Once I had done this, then I could look for patterns and ultimately recommend an aggregation strategy (more on that later). To do this I cataloged each one using their generated SQL from the OBI Query log. The result of this helped me to understand the following:
- Tables used
- Joins
- Which tables had filters and selects
- Based on the above, determine each report’s granularity
- Baseline each report’s performance to measure improvement.
Understanding the grain was the ultimate goal of this exercise – it will become useful for identifying aggregate candidates later on. Note I was not interested in the metric calculations or other report logic – this is more about tables than fields. Complex logic will be dealt with later after the bases are covered.
No comments:
Post a Comment