Скачать или смотреть SQL Server Performance Office Hours Episode 33

**NEW PRESALE COURSE**
https://training.erikdarling.com/view...

Get it for 50% off ($250) during the pre-release sale. Head to the link for full details.

Become a member!    / @erikdarlingdata  

Ask me questions! https://erikdarling.com/office-hours/

Click here for 50% off a health check: https://training.erikdarling.com/darl...

If you like what you see here, you'll love my advanced performance tuning training:
https://training.erikdarling.com/?cou...

Questions:
What are your favorite sandwich condiments?

First off big fan. Also, Why does lt=1 on a query with row_number window function perform so different than = 1. I find its a trick that almost forces sql server to compute all row number data before doing something else like pushing down the predicate of a join if the window query is used as a join "table". I tried in (1,null) and gt0 and lt2 without the same results. Thanks!

If you could change something about the world, not related to Perf Tuning SQL Server, what would it be?

What do we need to be aware of when enabling snapshot isolation in a large production DB? Should we expect any downtime? Thanks!

I enjoyed hearing you beg a viewer to switch from DPA to SQL Sentry. Can I hear you say more about what DPA does wrong? I haven't used it in years.