It’s hard to find good 400-500 level SQL Server training at any price, let alone free. Dave Ballantyne’s past sessions from SQLBits are great examples of those.

Today, I’m going to review Dave’s most recent session on Estimation and Statistics. It’s from 2015, but it’s still very relevant today.

If you want to jump past the intro and bio, the technical stuff starts at about 2 minutes 15 seconds in.

Dave starts by explaining how building a query plan is a lot like building a driving route from point A to point B. I love that analogy because:

• There are a lot of options, and exploring all the possible routes is computationally intensive
• We have to make some assumptions about how the roads and traffic are going to look
• Along the way, things may go wrong, but at that point, it’s too late for us to replan our route from scratch

A few time marks if you want to jump ahead in the video recording:

• You can skip the intros by jumping to about 2m:15sec into the video.
• If you’ve watched How to Think Like the Engine, and you want to jump into how the statistics histogram is decoded, jump to 12m:30sec into the video.
• If you already know how to interpret a single histogram, and you want to know how SQL Server uses multiple histograms on different columns, jump to 21m in.
• When you hit the brief part on parameter sniffing and want to skip past it, jump to 43m:19sec. The parameter sniffing part is probably not new to regular readers here. At 43m:19sec, the next topic starts, the Plan Skyline problem.
• To learn about optimization’s ascending key problem, jump to 55m in

If you like training like this, consider attending the March conference either virtually, or in-person outside of London. Registration is open now, and you can use discount code Brent5 to save 5%. Here’s the spot to look for during registration:

As of this writing, there are literally 10 Early Bird seats left, so hustle!