https://www.bartlannoeye.com/series/app-service-performance/ Recent content in App Service Performance on Bart Lannoeye Hugo -- gohugo.io en-us Thu, 22 Jan 2026 19:45:00 +0000 https://www.bartlannoeye.com/blog/appinsights-long-duration-requests/ Thu, 22 Jan 2026 19:45:00 +0000 https://www.bartlannoeye.com/blog/appinsights-long-duration-requests/ When inheriting a solution, you learn quite a few things and try to fix some of them. This particular problem was impacting our monitoring and capability to find issues in the application. Some of our background jobs were triggered every minute, and every request (and request duration in Application Insights) showed as a full 1 minute. Of course that skews the performance numbers quite a lot and makes identifying and troubleshooting slow requests a lot harder. https://www.bartlannoeye.com/blog/appservice-scaling-for-cost/ Mon, 01 Dec 2025 19:50:00 +0000 https://www.bartlannoeye.com/blog/appservice-scaling-for-cost/ No matter what you do, if it comes to performance and scale, there’s always a cost involved. It is easy to go bigger and throw more compute at it, but those dollars start adding up very fast. So we need to look at a solution to keep costs in check while we do the necessary troubleshooting and fixes. Most likely, not every single minute of the day requires the same power. https://www.bartlannoeye.com/blog/appinsights-code-optimizations/ Sun, 16 Nov 2025 21:40:00 +0000 https://www.bartlannoeye.com/blog/appinsights-code-optimizations/ In the last post we already looked at memory dumps and profiler traces, but going that deep can be time consuming. Luckily Application Insights provides somewhat similar tooling out of the box by giving you a built-in .NET Profiler and automatic “AI” analysis under the name Code optimizations Ensure your logging is configured correctly, start capturing data and analyse the issues. Note: When the Application Insights Profiler for .NET is actively running and collecting traces, it typically adds between 5% to 15% of CPU and memory overhead to your server. https://www.bartlannoeye.com/blog/appinsights-diagnose-and-solve/ Sat, 08 Nov 2025 21:40:00 +0000 https://www.bartlannoeye.com/blog/appinsights-diagnose-and-solve/ When things are burning, you typically don’t have a lot of time. Or the performance issues might have been recurring on a certain interval, but you’re always failing to capture the right moment. It would be handy if you had a memory dump of the exact moment when the compute was choking. This can be done by capturing that memory dump either at the moment itself or planning it to be captured through ‘proactive CPU monitoring’. https://www.bartlannoeye.com/blog/appservice-instance-load/ Tue, 04 Nov 2025 18:15:00 +0000 https://www.bartlannoeye.com/blog/appservice-instance-load/ There are multiple places to get started in troubleshooting. This post already covers a few. App Service Plan Metrics A first good start is to look at the App Service (single site/service) or App Service Plan (see it as the IIS of old days, combines multiple services) Metrics page and dig into both CPU and memory. As we can see, there is clearly an issue. However, the question is: is this an application wide problem, or a very specific bottleneck. https://www.bartlannoeye.com/blog/appservice-performance-series/ Sun, 02 Nov 2025 20:10:00 +0000 https://www.bartlannoeye.com/blog/appservice-performance-series/ When you inherit an existing brownfield application, there are usually plenty of challenges to be tackled. The past year and a half, I have been assisting a team in “keeping the application from burning down”. Often this results in: Customer support tickets “The application is slow”. Graphs like the one below. Up until some point the team simply restarted the compute and reported ‘solved’. But as we all know, this doesn’t fix issues in the long term, so issues needed to be tackled.