Press "Enter" to skip to content

Day: February 28, 2025

Spark Connector for Fabric Data Warehouse

Published 2025-02-28 by Kevin Feasel

Arshad Ali announces a connector:

We are pleased to announce the availability of the Fabric Spark connector for Fabric Data Warehouse (DW) in the Fabric Spark runtime. This connector enables Spark developers and data scientists to access and work with data from Fabric DW and the SQL analytics endpoint of the lakehouse, either within the same workspace or across different workspaces, using a simplified Spark API. The connector will be included as a default library within the Fabric Runtime, eliminating the need for separate installation.

Click through to check out its capabilities. This is a tiny step toward where I think Microsoft Fabric should go: any tool accessing the same data, eliminating separate lakehouses vs warehouses and having to remember that you can’t use this syntax in this scenario unless you connect to it this way and sacrifice one live chicken.

Leave a Comment

Seeding an Availability Group via T-SQL Snapshot Backup

Published 2025-02-28 by Kevin Feasel

Anthony Nocentino jump-starts an availability group:

In this post, the fifth in our series, I want to illustrate an example of using the T-SQL Snapshot Backup feature in SQL Server 2022 to seed Availability Groups (AGs) with storage-based snapshots. Efficiently seeding an Availability Group is essential for maintaining high availability and ensuring effective disaster recovery. With the introduction of T-SQL Snapshot Backup in SQL Server 2022, snapshots can now be created at the storage layer. This advancement significantly speeds up the initialization of secondary replicas, particularly in environments that handle large databases.

This post will walk through a PowerShell script that effectively seeds an AG using T-SQL Snapshot Backup, dbatools, and Pure Storage FlashArray.

Click through for the script and the process.

Leave a Comment

Debugging SSIS Packages

Published 2025-02-28 by Kevin Feasel

Andy Brownsword wrote code with an error in it:

When constructing or investigating a SSIS package we can run into a variety of issues. To help resolve there are various techniques which can be used to troubleshoot the package.

Whilst we have the Progress tab for the package to tell us what’s happened during execution, it’s usually more effective to debug packages in flight. Below we’ll look at a few ways we can achieve this:

Never having written code that doesn’t work perfectly the first time, this post is, of course, merely academic for all of us perfect developers. But for those people who may have let a little something sneak in or have to deal with the errors that others have seeded into our beautiful programs, Andy provides three separate techniques for troubleshooting packages in flight.

Leave a Comment

Removing Unnecessary Delete Operations

Published 2025-02-28 by Kevin Feasel

Monica Rathbun truncates some data:

SQL Server performance issues often stem from easy fix bottle necks that can be fixed with the right tuning strategies. This short blog will focus on the DELETE statement.

Click through for the full story. One minor correction that I’d offer is that the TRUNCATE TABLE command is logged, which is why you can roll it back in a transaction. The logging process is much less intensive than with DELETE because of the deferred drop logic that Paul talks about in the link.

That aside, Monica’s key point is absolutely correct: DELETE operations tend to be quite slow, especially as the number of records you need to delete increases. This is where techniques like batching delete operations can help reduce the pain level, but if you’re deleting every (or almost every) row from a table, there’s probably a better method. Unless replication is involved in the mix—in that case, there are no good methods for anything because replication hates us.

Leave a Comment

Finding what Changed in a SQL Server Cumulative Update

Published 2025-02-28 by Kevin Feasel

Brent Ozar does some sleuthing:

Over the last several years, Microsoft has been putting less and less effort into Cumulative Update documentation. We used to get full-blown knowledge base articles about fixes, but these days, we get a collection of footnotes with deceiving hyperlinks that look like they’re going to lead to more information – but they simply lead back to themselves.

So whenever a new Cumulative Update drops at SQLServerUpdates.com, before I install it, I like to:

  1. Log my dev server’s list of current sys.all_objects, all_columns, messages, configurations, etc to tables
  2. Apply the update
  3. Query the new contents of sys.all_objects, all_columns, etc to see what new ones were added

Click through for a script that does just this. It’s not as good as having detailed patch notes, but it beats having nothing.

Leave a Comment

Scraping SQL Server Version Updates

Published 2025-02-28 by Kevin Feasel

Rod Edwards scrapes patch information:

(just for info, i’ve always refer to the SQL Server Version List (https://sqlserverbuilds.blogspot.com) as Blogspot for some reason)

Doesn’t take long, but it would be nice to not have to bother at all, and I don’t mean by moving everything to the cloud and let MS handle it either before anyone says it. 🙂

The details are just websites, so the html is just plain text, AND the awesome people who provide Blogspot also add the details to a publicly shared google doc, in multiple formats, legends.

So what can we do to make our lives even easier. How about downloading/scraping all of the details that you might ever need, and dropping them into tables somewhere? So when its in SQL, you can then choose to trigger off that however you like?

Read on to learn how.

Leave a Comment

Trying out fabric-cicd

Published 2025-02-28 by Kevin Feasel

Kevin Chant tries a Python package:

In this post I want to cover my initial tests of fabric-cicd. In order to provide some tips for those looking to work with this new offering.

Just so that everybody is aware, fabric-cicd is a Python library that allows you to perform CI/CD of various Microsoft Fabric items into Microsoft Fabric workspaces. At this moment in time there is a limited number of supported item types. However, that list is increasing.

Read on for the test. It currently supports a limit amount of functionality, but it looks promising.

Leave a Comment