I'm finding Polybase to be even more finicky than OPENROWSET, which is challenging enough.

I'm about to give up. Where Polybase has failed me, I've found success with and sp_execute_external_script. It has its own challenges, but the R packages are much more forgiving.

Microsoft releases SQL Server 2022, preferring hooks to Azure over enhancing the on-premises product.

Interesting report & analysis of the SQL Server 2022 release, including SQL Server tools and cloud vs on-prem thoughts.

devclass.com/2022/11/16/micros

Today I'll be evaluating clustered indexes. Do we choose the appropriate clustering key(s)?
Are there more cluster key scans than seeks?
Let's find out...

Sigh. Playing 'Whack-a-mole' again with heaps and uncompressed tables/indexes.

I'm trying to determine the trigger order for a table with eight triggers...and somehow find an answer to someone's question about a data change.

How's your day?

NEW POST:
✍ Read and Write files to Azure Data Lake with SQL Server

🔎 Write files to cloud storage with SQL Server's Machine Learning Services.

davemason.me/2022/11/08/read-a

NEW POST:
✍ Import and Export Parquet Files with SQL Server

🔎 T-SQL has no native support for parquet files. It's R, Python, and Machine Learning Services to the rescue!

davemason.me/2022/11/06/import

NEW POST:
✍ Machine Learning Services and File System Interaction

🔎 External Python and R scripts for SQL Server Machine Learning Services have limited access to the file system.

davemason.me/2022/11/04/machin

NEW POST:
✍ RStudio and Package Installation for Machine Learning Services

🔎 A walkthrough for installing R packages for SQL Server Machine Learning Services

davemason.me/2022/11/04/rstudi

My day concluded with me trying to explain that indexes have limits with regard to how much technical debt and bad design they can overcome.

😱 table with 850 columns. Every one is data type VARCHAR and storing dollar amounts.

T-SQL Tuesday #155 –The Dynamic Code Invitation

This is the monthly blog party where someone hosts and you all write a response.


voiceofthedba.com/2022/10/06/t

Native backups to disk and log shipping may not be glamorous, but they're reliable AF.

Also not glamorous: outages. ☹

Your dynamic TSQL might be clever, but you'll get side-eye instead of admiration if performance sucks.

BLOGGED: SQL Server Indexes With Unused INCLUDE Columns

How to find unused INCLUDE columns in nonclustered indexes for SQL Server.

davemason.me/2022/09/03/sql-se

Hey folks. This took a while to code and test...and an eternity to run the TSQL scripts to completion. But it seems to work. The first iteration looks really promising.

In our biggest/busiest SQL Instance, I counted 17,732 INCLUDE columns in all non-clustered indexes across all databases.

Only 6,645 of them have query plan references in Query Store or the plan cache.

That's a lot of "over indexing"!

github.com/BeginTry/SQL-Server

Today I'm learning about the joys of restoring a database that is a publisher for replication.

Show more

Dave Mason

CounterSocial is the first Social Network Platform to take a zero-tolerance stance to hostile nations, bot accounts and trolls who are weaponizing OUR social media platforms and freedoms to engage in influence operations against us. And we're here to counter it.