Perf regression: fortunes, fortunes

[pr-benchmarks]

Scenario	Environment	Date	Old RPS	New RPS	Change	Deviation	StDev	Dependencies
fortunes	Fortunes Blazor SSR 1- Trends Database Intel Linux	05/27/2025 09:32:48	79,711	1,574	-98.03 % (-78,136) 👎	-231.32 σ	337.79	Changes Name Version Diff Microsoft.AspNetCore.App 10.0.0-preview.5.25265.101 -> 10.0.0-preview.6.25276.106 ad85650...b53264c Microsoft.NETCore.App 10.0.0-preview.5.25265.101 -> 10.0.0-preview.6.25276.106 dotnet/runtime@ad85650...b53264c Microsoft.AspNetCore.Antiforgery.dll 10.0.0-preview.6.25276.106 (New ) n/a Microsoft.AspNetCore.Antiforgery.dll 10.0.0-preview.5.25265.101 (Removed) n/a
fortunes	Fortunes Blazor SSR 2- Trends Database Intel Windows	05/27/2025 10:26:23	103,334	2,045	-98.02 % (-101,289) 👎	-198.70 σ	509.77	Changes Name Version Diff Microsoft.AspNetCore.App 10.0.0-preview.5.25265.101 -> 10.0.0-preview.6.25276.106 ad85650...b53264c Microsoft.NETCore.App 10.0.0-preview.5.25265.101 -> 10.0.0-preview.6.25276.106 dotnet/runtime@ad85650...b53264c Microsoft.AspNetCore.Antiforgery.dll 10.0.0-preview.6.25276.106 (New ) n/a Microsoft.AspNetCore.Antiforgery.dll 10.0.0-preview.5.25265.101 (Removed) n/a

PowerBI Dashboard

@sebastienros

[DamianEdwards]

@lewing @javiercn @danroth27 huge Blazor Fortunes regression:

I suspected a benchmark break but no errors indicated:

Lock contention and LOH size went through the roof:

[lewing]

we're aware. look at the more recent data

[DamianEdwards]

Yep sorry, seems I wasn't looking at latest results.
We can close this then?

[DeagleGross]

perf regression reported in this issue is due to dotnet/runtime#116677 and dotnet/runtime#117444 I think; it has been fixed and benchmark is back to original values.

The latest improvement with a significant RPS increase is #62653.

Closing, let me know if there are reasons to keep it open and what should be the further steps