[Video] Office Hours: Answering Your Microsoft Data Platform Questions

1 Comment

Ask questions at https://pollgab.com/room/brento and upvote the ones you’d like to see me cover.

Here’s what we covered today:

  • 00:00 Start
  • 00:59 neil: is it dangerous/risky to expand a drive in azure with SQL data files on it? sql on azure vm. dont know what happens behind the scenes
  • 02:07 Brett: I am a consultant at a large cloud provider. I have to protect my job and not upset our customers too much. How do you recommend balancing the line when delivering bad news? At times I have to soften my words, but in reality everything is in the toilet.
  • 02:35 Madisynn: Does PostgreSQL handle xml / Json data any better than SQL server?
  • 03:14 Tobias: What is your opinion of the new buffer pool parallel scan feature in SQL 2022? When is this a compelling reason to upgrade from SQL 2019?
  • 04:41 Fyodor: What is your opinion of the Microsoft Azure Premium V2 cloud storage improvements (125 – 1,200 MBPS, 80,000 IOPS, 64TB)?
  • 05:40 Yitzchak: What kinds of performance issues have you run into with SQL Change Tracking?
  • 07:00 DBAInHiding: After 10y as a prod support DBA, I moved to ETL dev with T-SQL + SSIS for 5y. I still tinker with my DBA skills with a home lab setup and training up on PowerShell/dbatools. I’d like to go back to being a DBA but will my recent developer track be seen as an asset or hindrance?
  • 08:09 Efraim: What conditional debug message printing techniques do you like to use in your stored procedures?
  • 08:42 Pyjamarama: Hi Brent, SQLServer 2014 with LARGE amounts of multiple plans. I used your query to find the top10 queries involved (your “Why Multiple Plans for 1 Query Are Bad”post). They are SELECT statements by our PowerBuilder application.Parameter sniffing?What do you suggest we do? Thanx
  • 09:18 Magnús: Which windows server admin concepts / technologies should a SQL DBA be familiar with when running SQL Server 2019 on a Windows Server 2019 cloud VM?
  • 10:22 Haydar: What are common mistakes you see regarding the use of TSQL cursors?
  • 11:34 Jagelman: My Friend needs a Reporting Database on AWS RDS (SQL 2016). He plans to use replication, however there is a problem. The fact table is a partitioned, and has a clustered columnstore index, so it has no primary key (cannot be replicated). Any suggestions on how to deal with it?
  • 12:50 Hans_Niemann_is_innocent : Hey Brent, I’m 5 years into the DBA career and never had to deal with any complex corruption issues. Is this still an important skill? I’m thinking of complex corruption as something more than just doing backup/restore to get online. Go Hans Niemann!
  • 15:16 Yousef: What is your opinion of Azure Cosmos DB for PostgreSQL?
  • 16:24 Yitzchak: Can having a [UNIQUIFIER] hidden column on clustered index affect query plan / query performance?


Previous Post
[Video] Office Hours: Nine Minutes of Answers
Next Post
Who’s Hiring in the Microsoft Data Platform Community? November 2022 Edition

1 Comment. Leave new

  • danielle.paquette-harvey
    October 27, 2022 4:47 pm

    OMG we’re using Change Tracking in prod. Over time, I’ve been asking the developers to stop using it and using columns to return the changes instead. Because Change Tracking locks the whole table the time it resolves the query, and if your system is an OLTP (like mine), then you can’t afford for Change Tracking to lock the table. Using a column to return the changes instead allows SQL to lock less data (with the right indexes) and have less locks and poison waits.


Leave a Reply

Your email address will not be published. Required fields are marked *

Fill out this field
Fill out this field
Please enter a valid email address.