Paul White explains why the IGNORE_DUP_KEY index option on a clustered index can result in much poorer performance than for a nonclustered unique index.
Jonathan Kehayias shares insight about new UDF information in showplan, and shows that only some tools will show it (SSMS 17+ & SentryOne Plan Explorer).
Continuing his series on row goals, Paul White explains how they can impact certain types of semi joins, and how you can spot these effects in showplan and trace flag output.