Oracle performance diagnosis has a relatively short but interesting history. From its inception to the early 1980s, Oracle performance diagnosis was just getting started, and Oracle systems were relatively small and simplistic. As a result, SQL tuning and resolving lock contention were the focus.

From the mid-1980s to mid-1990s, Oracle systems began to grow in size and complexity. As a result, performance analysis was needed urgently. Oracle ratio analysis was born, and the performance industry started to flourish, giving rise to technical papers, books, vendor products, and consulting services. It was kind of the cowboy days of performance analysis. Authors and consultants had a good thing going. The books on the topic were massive because of the sheer number of performance ratios. And because there were so many ratios to examine, consultants had long engagements, generating a lot of money. And since the analyses were not as exact as today, they tended to take a few more cycles to complete. In truth, not only authors and consultants benefited; anyone involved in performance analysis had plenty of work.

Now, don't get me wrong. Ratio analysis works, and at that time, it was the best method we had at our disposal. But times have changed.