The art of simplicity

Today I lost a lot of time investigating a stupid(aren’t they all?) issue with some failing tests on the build server. The strange thing was that when I ran the same tests locally, they always succeeded...what was going wrong? Just for completeness, here is the test task configuration I was using: Nothing special I would think. It was only when diving deeper into the build output that I discovered what was going wrong. Here is the output that explains the problem: 2025-05-20T13:19:22.3855459Z vstest.console.exe 2025-05-20T13:19:22.3855564Z "D:\b\3\_work\210\s\IAM.Core.Tests\bin\Release\ net6.0\IAM.Core.Tests.dll " 2025-05-20T13:19:22.3855657Z "D:\b\3\_work\210\s\IAM.Core.Tests\bin\Release\ net8.0\IAM.Core.Tests.dll " 2025-05-20T13:19:22.3855761Z "D:\b\3\_work\210\s\Mestbank.Core.Tests\bin\Release\net6.0\Mestbank.Core.Tests.dll" 2025-05-20T13:19:22.3855857Z "D:\b\3\_work\210\s\Mestbank.Core.Tests\bin\Release\net8.0\Mestbank.Core.Tests....

I know, I know, .NET 10 is already in preview and I am still catching up on what was added to .NET 9.0. Today while upgrading an older application to .NET 9, I decided to have a look at the new static file handling introduced in .NET 9 through the MapStaticAssets feature. Static File middleware (before .NET 9) Before .NET 9, static files(Javascript, CSS, images, …) were handled through the UseStaticFiles middleware This middleware is still there as the new MapStaticAssets feature does not support all the features that the original middleware had. From the documentation : Serving files from disk or embedded resources, or other locations Serve files outside of web root Set HTTP response headers Directory browsing Serve default documents FileExtensionContentTypeProvider Serve files from multiple locations Serving files from disk or embedded resources, or other locations Serve files outside of web root Set HTTP response headers Direc...

This week a colleague contacted me with an issue he encountered after upgrading to .NET 8.0. On the project involved we were using the WsFederation middleware to authenticate and interact with ADFS. However after upgrading to .NET 8 and the 8.x version of the Microsoft.AspNetCore.Authentication.WsFederation middleware the trouble began. Using this version resulted in a change in behavior as suddenly the middleware starts to expect a SAML 2.0 token instead of a SAML 1.1 token that is now issued by our ADFS server: XmlReadException: IDX30011: Unable to read XML. Expecting XmlReader to be at ns.element: 'urn:oasis:names:tc:SAML:2.0:assertion.Assertion', found: 'urn:oasis:names:tc:SAML:1.0:assertion.Assertion'. Although I found a post online that it would be technically possible to let ADFS return a SAML 2.0 token through WSTrust, this doesn’t fit in the passive federation scenario we had, so time to look at some alternative solutions. Attempt 1 – Reverting to an ...

I got contacted by someone from my team that INSERTs were failing on one of our database instances. A look at Application Insights showed the following error message in the logs: System.Data.SqlClient.SqlException (0x80131904): Timeout expired.  The timeout period elapsed prior to completion of the operation or the server is not responding. For an unknown reason, the transaction timeout before it could be committed resulting in a rollback operation and a failing INSERT.  Time to open SQL Server Management Studio and first take a look at the Resource Locking Statistics by Objects report: Indeed, multiple sessions share a lock on the same table. If you want, you can also check the User Statistics report to find out what these sessions are related to:   To fix the issue, I killed the sessions that were causing the lock: KILL 236 KILL 210 KILL 257 After doing that, I could confirm that the locking was gone by refreshing the report: More information Sql...

Yesterday on my way home, I was listening to the SoftwareCaptains podcast episode with Mathias Verraes (sorry, the episode is in Dutch). One of the topics discussed was technical debt and the  question was raised why (most) organizations manage very carefully their financial debt, they don’t apply the same rigor for their technical debt. This triggered a train-of-thought that resulted in this post. Financial debt is meticulously tracked, reported, and managed. CFOs provide regular updates to boards about debt levels, leveraging ratios, and debt servicing costs. Detailed financial statements outline current liabilities, long-term obligations, and repayment schedules. Financial debt is visible, quantified, and actively managed. Yet technical debt—which can be just as crippling to an organization's future—often exists as an invisible, unquantified burden until it's too late. What if we managed technical debt with the same rigor as financial debt? The hidden cost of techni...

A while back I talked about a memory leak we had in one of our applications. As a consequence, it brought the full production environment to a halt impacting not only the causing application but all applications hosted on the same IIS instance. Although we found the root cause and fixed the problem, we did a post-mortem to discuss on how to avoid this in the future. In this post, I'll walk you through the practical strategies we implemented to limit and optimize memory usage for our applications running in Internet Information Services (IIS). n this guide, I'll walk you through practical strategies to limit and optimize memory usage for applications running in Internet Information Services (IIS). Understanding memory usage in IIS Before diving into solutions, it's important to understand how IIS manages memory. IIS runs web applications in application pools, which are processes (w3wp.exe) that host your web applications. Each application pool can consume memory ind...