SQLskills Insider Session 10: Statistics Wrap-Up + Q&A
We’ve spent two sessions talking about the histogram – how it’s used, how it’s built, and some of the concerns that stem from the fact that it has to be small to be effective (remember, the job of the optimizer is NOT to find the best plan but instead to find a good plan fast). As such, statistics (to be used QUICKL) must be small. And, they’re generally very good. However, there are definitely a few issues with large data sets or those that are very skewed.
For this “statistics wrap” session, I’d love for you to go back over the resources of BOTH sessions 8 and session 9 and also watch the PASS session referenced (you can use Jan 5’s timeslot for that!). Then, please send me any remaining questions. I’ll also plan a few summary slides (as I did in the other Q&A) and then we’ll go from there.
Kimberly L. Tripp
President and Founder, SQLskills
Read my full bio here: Kimberly L. Tripp and about my Celebrating 25 Years of SQLskills here: https://www.sqlskills.com/about/meet-the-team/celebrating25years/