You posted and upvoted SQL Server questions at https://pollgab.com/room/brento, and I went out for a walk on the Seltjarnarnes peninsula to the lighthouse to answer ’em.

Here’s what we covered:

• 00:00 Introductions
• 00:58 Gio: Any known gotchas from using Windows Update to apply CUs to SQL
• 02:17 AlwaysLearningDBA: Hello Brent. A friend of mine wanted to know what camera/system do you use when you are outdoor doing Q&A office hour sessions? He is just curious as the camera follows your movement as if someone is moving the camera physically and he thinks it’s awesome.
• 03:34 PGA: Hi Brent If I have firstname lastname in one index and city,age on other index if my query is Order by City , Age My second index will be used?
• 04:47 Inpasta DBA: Hi Brent! Do execution plans replicate over to the secondary in an AG? Wondering if a failover, planned/unplanned, occurred would the plans be the same. Thanks for what you do for the community!
• 05:49 Jack: Hi Brent, with all the certifications available now for Azure, AWS and GCP has your opinion on the value of certifications changed?
• 06:02 I’m a potato: Hi Brent! Is there a way to know the health of your statistics on SQL server and prevent their most common issues?
• 06:44 Thomas Horner: In your experience, is using Always-On Availability Groups require more DBA monitoring and maintenance (i.e. babysitting)?
• 07:26 Rick Lively: How should we capture a “baseline”, or a series of captures, with the Blitz scripts that you wish would have been collected to help understand performance problems when the server is under stress at a later time?
• 10:07 Willem: Do you know of any success stories of companies that have made a the move to Azure with a large SQL Estate(+- 80 Enterprise Servers and up)? We’re figuring out how to attempt this, but there seems to be so many pitfalls to be aware off.
• 11:47 Stockburn: Hi Brent any advice on deadlocks where the two processes are an update and a delete, they both end up hitting the clustered primary key and then as you put it, one of them gets a bullet from the deadlock monitor. Thanks for these office hours!
• 13:04 catmanjan: Applications now have tech like kubernetes and are more or less hands off. Why don’t database engines do this too?
• 15:17 LazyD: Hi Brent In upgrading from SQL2014 to SQL2019 what can be done about CLRs so that they don’t “break” due to changes in SQLCLR security since SQL2017.
• 15:43 VegasDBA: What do you think of RedGate SQL Toolbelt? A friend insists that it’s a necessary purchase to help production DBAs manage an environment. Any other similar tools or utilities you’re a personal fan of?
• 17:13 NullPointer: I have been fighting with tuning a query/view, this view joins to another view that when run by itself is super fast (with good estimates) but when I join it to a larger query all the estimates are way off
• 17:58 Dave: What would you think about a disaster recovery plan that involved taking differential backups of the production server every 2 hours? Asking for a friend.
• 20:03 Scott McFadden: Hi Brent, appreciate your SQLServer book recommendations. They are spot on. Do you have any book or training recommendations for learning PostgreSQL?
• 21:28 Champaign DBA: Brent, I read your 5 part blog on foreign keys, and your concern about locking when using cascade delete. Wouldn’t that not be an issue when using read committed snapshot?
• 22:14 Pat: Hi Brent! I’m a accidental DBA. My DB monitoring software often alerts on “OS paging”. What’s “normal” to hone in on?
• 23:17 Seven: How does sampling work for statistics? If a FULLSCAN has been specified previously…
• 23:37 SpongeBob SquarePants: Do you have any recommendation for tools specific for ODS in SQL Server?
• 23:49 David: Do you have an tips to deal with people who take other peoples advice or tips as there own in the workplace? Here I am with 1.5 years experience in SQL, teaching people with 5+ Years experience. But when they tell the Boss it them who came up with query. Some don’t know union
• 26:01 Khaled Budajaja: Duplicates on Primary Key Column. Some of my customers on SQL Server 2014 have this weird issue.
• 26:45 SQL_Deadwood: How do I perform index maintenance on a synchronous AG database with heavy fragmentation (there was no maint. job) without, well, ruining everything? On-prem VM.
• 28:18 SQL_Deadwood: According to the tricksy documentation hobbit, “There are specific but uncommon scenarios when…index maintenance may be needed in Azure SQL Database” – what are they?
• 29:23 Recap