Y’all post questions at https://pollgab.com/room/brento, upvote the ones you’d like to see me cover, and then I talk through ’em.

Here’s what we covered today:

• 00:00 Start
• 00:22 CKI: Could please give me pointers how to answer the question “Do we need to create a new instance for the new application or can use the existing one?” Thank you very much
• 01:18 Chris May: Within an index, if you can guarantee unique selectivity with 2 key columns, is there a fundamental difference between adding an extra key column or an included column? Which would you typically prefer?
• 03:15 Pyjamarama: Hello Brent. In one of our customers our User databases (Greek_CI_AI) are in different Collation than TempDB (Latin1_General_CI_AS). What is the best way to fix that? (We prefer to keep the CI_AI). Thanx!
• 04:41 Maksim Bondarenko: What would you do first if two sql servers with the same version and CU generate almoust identical execution plan with one exception? Second server generates one additional block (in my case it is Index Spool after Cl scan) and that’s why 10 times work faster. Thx
• 05:47 Piotr Rasputin: What’s your opinion of azure vm host caching for SQL vm’s?
• 07:30 Peter: What does your op love ones at home say when you talk like Clippy all the time? 🙂 Thanks for all the knowledge you share with us all
• 08:00 Gerardo: What is your top story for when you were burned by missing / incorrect SQL Server documentation?
• 10:30 Too Much Spare Time: Last time there was a question about how to deliver bad news as a consultant. What type of bad news do you find delivering the most?
• 11:49 chris: With a SQL Server built on a virtual machine on VMWare is there a way to determine CPU and/or memory pressure on the host without having access to VCentre?
• 12:38 LockedUp: I have a proc that is causing deadlocks. I have tracked it down to an Update statement with a ROWLOCK hint. Could the ROWLOCK hint be causing the deadlock? Update statement is well written and indexed, not sure why ROWLOCK is even needed.
• 14:24 Leonard: What is your opinion of the SQL Server data collector? Any good for performance troubleshooting?
• 15:12 Maciej: Hi Brent, do you have a missing feature in SSMS and/or SQL Server that you treat (at this point) as an old friend? 😉 In addition if you could “influence” Microsoft to implement one which one would it be?
• 17:11 Leif Hole: Hi Brent ! Really enjoy your “think like the engine” and the fundamentals ! My friend have this theory that Mr. Codd meant that all joins between tables is the best way of getting data. Never denormalize ! Is there any thumb of rules when it comes to denormalizing ?
• 18:09 chris: I understood Azure SQL DB was always current; however, when I view the compatibility level of some of my databases they’re not current. Is it up to me to change this? Any gotchas with doing so?