Month: April 2021

How a row goal affects estimates

Most of our queries are intended to process a complete data set. But not all. There are cases where we restrict the number of rows. Or cases where SQL Server does this. In all such cases, and a few others, the query optimizer uses a mechanism called a “row goal” to ensure that the execution plan is optimized for a subset of the total result set, not all of it. Some examples Without trying to be complete, let’s list a few cases where a row goal would be set. Literal row limitation in the query The most simple is when…

Execution plans – learning opportunities

Most people know they have to invest in themselves, so that they can get better at their job. And if that job involves query performance on SQL Server, then part of that investment should definitely go towards execution plans. I happen to teach about execution plans. A lot. And there are a few opportunities in the near future that I want to make sure you are aware of. (If you are allergic to self-promotion, you might want to skip this post!) The basics Perhaps you are here because you have heard that execution plans are an important tool for query…

Execution Plan Video Training – Announcing the next block

It’s time! Time to formally announce the release date for the next block of videos for the SQLServerFast Execution Plan Video Training. The basic level of block 2, Reading data, will go live on May 15. What to expect? Block 2 of the SQLServerFast Execution Plan Video Training covers everything involved with reading data. The basic level, consisting of five videos (“chapter”) of 25 to 40 minutes each, focuses on what many people still see as the standard for SQL Server: disk-based rowstore indexes and heaps. The first chapter of the block gives a detailed overview of how this type…

T-SQL Tuesday #137 – Use notebooks every day? No way!

As you probably know, T-SQL Tuesday is a monthly event where bloggers are invited to write about a set topic. Our host for April 2021 is Steve Jones (b|t) this time, and he has asked us to write about Jupyter notebooks: do we use them, would we want to; and why or why not? The basics Let’s start with the basics, because I’m not convinced that everyone reading this already knows what Jupyter notebooks are. Simply put, a notebook is a single file where you can store a combination of blocks with formatted text and blocks with code, where the…

Plansplaining, part 17. Temporal tables (part 2)

Welcome to part seventeen of the plansplaining series, where I will continue my look at how temporal tables affect execution plans. The previous post focused on data modifications in temporal tables. Now let’s see what happens if we use temporal logic to our basic retrieval queries. This post focuses on the basics only; do keep in mind that as soon as you want to join data from multiple temporal queries you will run into some specific issues that you need to be aware of; this will be the subject for a later post. Sample tables For the code samples in…

By continuing to use the site, you agree to the use of cookies. more information

The cookie settings on this website are set to "allow cookies" to give you the best browsing experience possible. If you continue to use this website without changing your cookie settings or you click "Accept" below then you are consenting to this.

Close