Press "Enter" to skip to content

Day: December 4, 2025

Join Planning in PostgreSQL 19

Published 2025-12-04 by Kevin Feasel

Robins Tharakan notes an upcoming performance boost:

The hidden cost of knowing too much. That’s one way to describe what happens when your data is skewed, Postgres statistics targets are set high, and the planner tries to estimate a join.

For over 20 years, Postgres used a simple O(N^2) loop to compare (equi-join) Most Common Values (MCVs) during join estimation. It worked fine when statistics targets are small (default_statistics_target defaults to 100). But in the modern era – we often see Postgres best-practices recommend cranking that up. Customers are known to be using higher values (1000 and sometimes even higher) to handle complex data distributions + throw a 10 JOIN query to the mix – and this “dumb loop” can easily become a silent performance killer during planning. 

That changes in Postgres 19.

Read on for an example of the problem and what is coming out to mitigate issues that currently exist.

Leave a Comment

Exposing Materialized View in Microsoft Fabric Lakehouses

Published 2025-12-04 by Kevin Feasel

Ed Lima makes some data available to other tools:

In today’s data-driven world, the ability to quickly expose data through modern APIs is crucial. Microsoft Fabric’s API for GraphQL combined with Materialized Lake Views offers a powerful solution that bridges the gap between your Fabric LakeHouse data and application developers who need fast, flexible access to your data.

In this guide, we’ll walk you through how to create a materialized view in a Lakehouse and expose it through a GraphQL API—all within the Microsoft Fabric ecosystem. This approach gives you the best of both worlds: the performance optimization of materialized views and the developer-friendly querying capabilities of GraphQL.

I’d say one interesting reason for why you might want to do this is to feed data to products like Teams, Power Automate, or Copilot Studio. In those cases, having the data be accessible via GraphQL makes it easier than working with finicky connectors that may or may not exist.

Leave a Comment

Backing up a Microsoft Fabric Workspace

Published 2025-12-04 by Kevin Feasel

Gilbert Quevauvilliers finds a gap and fills it:

In the high-stakes world of data architecture, where downtime can cascade into real business disruptions, I’ve learned that even the most robust platforms have their blind spots. Just last month, while collaborating with a client’s Architecture team on their disaster recovery strategy, we uncovered a subtle but critical gap in Microsoft Fabric: while OneLake thoughtfully mirrors data across multiple regions by default, other workspace items—like notebooks, semantic models, and pipelines—aren’t directly accessible in a failover scenario without extra steps. For the nitty-gritty on Fabric’s built-in reliability features, check out this Microsoft Learn guide.

That’s the spark that led me down this rabbit hole, and in this post, I’ll walk you through a practical solution: a Python Notebook that automates backing up your entire Fabric workspace to OneLake and an Azure Storage Account for that extra layer of redundancy. Whether you’re prepping for the worst or just embracing the “better safe than sorry” mindset, this approach gives you portable, versioned copies you can restore quickly.

Click through for the notebook, as well as instructions on how to use it.

Leave a Comment

Default Constraints and User-Defined Functions

Published 2025-12-04 by Kevin Feasel

Erik Darling has a new video. Erik shows how SQL Server handles default constraints that use user-defined functions and how this behaves under a variety of circumstances. There’s also a dive into parallelism and constraints. We also learned Erik’s ability to perform fractional math and how he actually differentiates “scalar” from “scaler,” proving once again that he is not midwestern from his use of extraneous vowel sounds.

Leave a Comment

Running SQL Server on KubeVirt

Published 2025-12-04 by Kevin Feasel

Andrew Pruski builds a virtual machine:

With all the changes that have happened with VMware since the Broadcom acquisition I have been asked more and more about alternatives for running SQL Server.

One of the options that has repeatedly cropped up is KubeVirt

KubeVirt provides the ability to run virtual machines in Kubernetes…so essentially could provide an option to “lift and shift” VMs from VMware to a Kubernetes cluster.

Read on to learn a bit more about KubeVirt, including how to set up a Windows-based virtual machine with it. Andrew does document some performance woes, so that’d be a big concern to work out the why behind this.

Leave a Comment

Reverse Engineering a Physical Model Diagram with Redgate Data Modeler

Published 2025-12-04 by Kevin Feasel

Steve Jones gives the new Regate acquisition a try:

I recently wrote about a logical diagram with Redgate Data Modeler. That was interesting, but creating all the objects is a pain. I decided to try creating a physical diagram from an existing database. This post looks at the experience.

Click through for Steve’s thoughts. I appreciate how he’s willing to call out the pain points that exist in the product today.

Leave a Comment