You know how this works: you posta the questions at https://pollgab.com/room/brento, and I giva the answers:

Here’s what you asked in this episode:

• 00:00 Start
• 03:18 DBA in VA: I recently discovered the “force order” query hint. I’m usually inclined to let the optimizer do its thing, but I have seen some of our code perform MUCH better with this hint. Are there gotchas/downsides I should know about?
• 04:29 RoJo: We call SQL from C# code, but it’s hard in SQL to trace where in code the call came from. Is adding a comment at the end of SQL statement, with module/method info a viable solution or are there better ways to connect SQL statements to actual method calls in a Live system.
• 06:34 OralceIsBetter: Hi, I have large database ~90 TB, full backup is taking almost 2 days (one time per month). Utilization off SAN network is around 2 gigabits/s. Do You have any tips how I can speed up this process ? Compressions is enabled and backup is running in many threads using couple disks.
• 07:16 Raj: When would you want to manually create statistics without a corresponding index?
• 08:13 Kimberly: I’ve got a 2 node AG cluster on SQL Server 2019. SQL Server sees 2 databases as added to the AG when they were actually removed. The primary shows synced and the secondary doesn’t have the db. Why does sql still think the db is in the AG when it is not?
• 09:06 Mike: Hi Brent, can you tell in which ways SQL Server is better than PostgreSQL ? Are there any bullet points ? And vice versa, are there any things in which PostgreSQL is better than SQL Server ?
• 11:26 RollbackIsSingleThreaded: Hi Brent! Writing articles about SQL Server does not earn much money. What do you think is the main advantage of writing articles?
• 14:49 Lysander: In boxed SQL, when should you use table partitioning vs partitioned views?
• 15:33 Mike: In which scenarios Failover Cluster Instances are preferable over Availability Groups ? When and why we should use FCI instead of AG ?
• 17:01 AnotherDataLayer: Linked Server vs Polybase: both are doing the similar things if not the same. which one to use when it comes to pull data from another MSSQL server and why. We are using entity framework.
• 18:35 Don’t Bother Asking: My friend has inherited a database which has lots of nonclustered PKs and very few clustered indexes. DUI query performance is not great. Will adding clustered indexes, or rebuilding PKs as clustered, improve query performance? And if so, are there any gotchas?
• 19:52 Bonnie: Do you have a good way to determine which operators in a query plan are contributing the most to a memory grant?
• 21:24 Perseus: Is there a good way to know why SQL Server ignores a given query hint?
• 22:29 Ophelia: What are the top signs that the SQL buffer pool is under pressure?
• 23:27 Chrysostomos: Does one database per customer model work well with SQL AG HADR?
• 25:32 Meryem: What’s the best resource for learning how to write efficient linked server queries?
• 25:46 Mandeep: What are the top things you see that break log shipping?
• 27:00 Pradeep: Given modern fast storage, is clustered column store index fragmentation as inconsequential as non-clustered index fragmentation?
• 27:46 Renaldo: For cloud SQL VM, what are top charge back mechanisms you see for billing SQL VM costs back to each customer on the SQL VM?
• 28:49 Olga: Have you ever had to rebuild all indexes? What was the use case?
• 30:09 Jessica: Hey brent, meta question. When building PollGab did you intentionally set out to build a site without trackers etc that would be blocked by uBlock Origin? Its amazing to see a clean site for the first time in a while.
• 30:43 muppet: Hey Brent, my friend has a fairly typically designed table with 1.2 billion row table stored as regular row store. It performs sluggishly so I was thinking of proposing partitioning but what do you think of switching to columnstore instead?
• 31:29 Brynjar: In sql server, how do you determine the optimal column order when creating a non-clustered column store index?
• 33:38 AllAboutSearch: When someone type the third character for the FName/MName/LName in the FE, it queries the DB’s computed column in the backend, all good. But when there is space in the name, entity framework query starts scanning the whole table and timeouts. Any tip?
• 34:53 Montro1981: Hi Brent, I hope you’re doing good. Have there been moments in your long career that you might have taken another path? If, your answer is yes (which is very likely), where might you have ended up in life?