Press "Enter" to skip to content

Category: Bugs

A Bug in a Frameless Window Aggregate

Hugo Kornelis digs into an execution plan:

The OVER clause, that can be added to aggregate functions to turn them into window aggregates, can come with or without a frame specification, in the form of an ORDER BY clause, plus an explicit or implied ROWS or RANGE clause. If there is no frame specification, then every row in a partition, or window, can see all other rows in the same partition or window for the purpose of the aggregation.

In plansplaining part 6, I looked at this specific form of window aggregation, and explained in detail all the steps that the execution plan takes to compute the aggregation result and add it to each of the rows, for every row in each window.

Click through for another look at the plan.

Leave a Comment

Dropping Users and Stale EXTERNAL MODEL Permissions

Andreas Wolter sees a cache invalidation issue:

When identity or permission boundaries behave inconsistently – even under specific lifecycle conditions – that is more than a product bug. It becomes a security-relevant design issue, because security depends not just on how access is granted, but also on how reliably it is revoked.

While investigating the new permission model introduced alongside SQL Server 2025’s AI integration and vector search capabilities (Article: New Permissions in SQL Server 2025), I encountered a case where EXTERNAL MODEL permissions can persist after a user is dropped, creating stale authorization state.

Click through to learn more about this issue and what it means.

Comments closed

Bug in sys.dm_exec_query_plan_stats

Brent Ozar finds a bug:

When you turn on last actual plans in SQL Server 2019 and newer:

ALTER DATABASE SCOPED CONFIGURATION SET LAST_QUERY_PLAN_STATS = ON;

The system function sys.dm_exec_query_plan_stats is supposed to show you the last actual query plan for a query. I’ve had really hit-or-miss luck with this thing, but my latest struggle with it is that two of the numbers are flat out wrong. It mixes up CPU time and elapsed time.

Here’s a simple query to prove it:

Click through for a demonstration of the bug.

Comments closed

pg_prewarm and a Bug in PostgreSQL 16.8

Warda Bibi explains an issue:

Recently, we encountered a production incident where PostgreSQL 16.8 became unstable, preventing the application from establishing database connections. The same behavior was independently reproduced in a separate test environment, ruling out infrastructure and configuration issues. Further investigation identified the pg_prewarm extension as the source of the problem.

This blog post breaks down the failure, the underlying constraint, why it manifests only under specific configurations, and the corresponding short-term mitigation and long-term fix.

Click through for those details. The bug is no longer an issue as of PostgreSQL 16.10.

Comments closed

No Buffer Pool Memory or OS Memory Available

Paul Randal re-tells a story:

Jonathan had a client issue recently where SQL Server’s buffer pool had been forced down to a ridiculously small size, only a few hundred MB, but the OS also showed basically no free memory. Page Life Expectancy was zero! What was going on?

From investigating SQL Server’s memory usage, the memory manager showed that target and total memory were the same, at only 1.2GB, and lock pages in memory was correctly set.

Read on for Jonathan’s troubleshooting steps and what he discovered.

Comments closed

What’s New in SQL Server 2025 CU2

Rebecca Lewis keeps track:

SQL Server 2025 CU2 released on February 12, 2026. Build 17.0.4015.4, six fixes. I’ve been running 2025 since RTM, and have been watching things pretty closely — looking for fixes, what to flag for clients, and what’s still sitting in the ‘known issues’ category, pending correction.

Here’s the short version.

Click through for that short version.

Comments closed

SESSION_CONTEXT and Parallelism Bug in SQL Server

Rebecca Lewis lays out the consequences of an existing bug:

If you use SESSION_CONTEXT() in any query that can run with parallelism, you may be getting wrong results right now and not know it. This is not new. It has been a documented known issue since January 2022. It shipped unfixed in SQL Server 2019, 2022, and 2025 — and as of 2025 CU2 (February 12, 2026), it is still not resolved.

This is easy to miss. It’s buried in the Known Issues section of CU release notes, and the symptoms — wrong results or dump files — do not obviously point back to SESSION_CONTEXT.

Read on to learn more about the issue, the current workaround, and how you can discern whether you are experiencing the issue today.

Comments closed

Excessive Crash Dumps in Power BI Report Server January 2026

Deepthi Goguri troubleshoots an issue:

I first updated the lower environments and then prod, but since most of the reports were used only in production, I didn’t see the issue coming. So, the issue with this release was that crash dump files were generated in the logfiles (C:\Program Files\Microsoft Power BI Report Server\PBIRS\LogFiles).

It’s bad enough that we who need to use Power BI Report Server are always at the end of the line when it comes to functionality, but this is adding insult to injury.

Comments closed

Solid Black Icons in Power BI Desktop January 2026

Marco Russo clues us in on a bug:

In mid-January 2026, many Power BI Desktop users began noticing a strange visual regression: several icons in the interface—including the Data, Model, and DAX icons—appeared as solid black shapes instead of their normal outlines. The issue appeared suddenly, even on machines where Power BI Desktop had not been updated. Because it happened the same week that Power BI Desktop January 2026 was released, this led to understandable confusion.

The root cause was not Power BI Desktop itself but an unexpected regression in the WebView2 runtime, one of the external dependencies used to render parts of the UI. Microsoft confirmed the issue and is preparing an official fix.

Marco does have a temporary fix that works until Microsoft has the official fix in place.

Comments closed