Natalia Kondratyeva

Stress tests are broken after 7.0 snap: [example log](https://dev.azure.com/dnceng/public/_build/results?buildId=1951513&view=logs&j=0da5d1d9-276d-5173-c4c4-9d4d4ed14fdb&t=45b31e99-b18c-506f-46cf-7d3a4d5140a9&l=17) ``` client_1 | Framework: 'Microsoft.NETCore.App', version '7.0.0' (x64) client_1 | .NET location: /live-runtime-artifacts/testhost/net7.0-Linux-Release-x64/ client_1 | client_1 | The following frameworks were...

Additional comment from user: >My current workaround is to use this class instead when Vector3.Cross is wrong. ```c# public class Vector3Workaround { [MethodImpl(MethodImplOptions.NoInlining)] public static Vector3 Cross(Vector3 vector1, Vector3 vector2)...

Sorry for the delay @sebastienros and thanks for the change! I've checked it, and there is a problem: while the machines are picked up correctly, the `serverAddress` variable is not...

1) the commit before the change -- works, the IP used is 10.0.0.104 `crank --config https://raw.githubusercontent.com/aspnet/Benchmarks/f544bbe8bf89aa29bf71861ff2790dee869dcc35/scenarios/httpclient.benchmarks.yml --scenario httpclient-kestrel-get --profile aspnet-perf-lin --variable warmup=3 --variable duration=3` 2) the commit after the change...

Also hit in #74669 ([log](https://helixre107v0xdeko0k025g8.blob.core.windows.net/dotnet-runtime-refs-pull-74669-merge-aebbdcb8ec304fd898/System.Net.Http.Functional.Tests/3/console.1eea56ed.log?helixlogtype=result)) and #74645 ([log](https://helixre107v0xdeko0k025g8.blob.core.windows.net/dotnet-runtime-refs-pull-74645-merge-d8e88aef58c34b8983/System.Net.Http.Functional.Tests.Attempt.3/3/console.be28fe05.log?helixlogtype=result)) with a bit different error: ``` System.Net.Http.Functional.Tests.SocketsHttpHandler_HttpClientHandler_MaxResponseHeadersLength_Http3.LargeSingleHeader_ThrowsException(maxResponseHeadersLength: 1) [FAIL] System.AggregateException : One or more errors occurred. (Assert.Contains() Failure Not found: 1024 In...

@MihaZupan I meant specifically a "timeout from inactivity". Because there were a bunch of different failures as well

Yes, that's https://github.com/dotnet/runtime/issues/72696. The fix is not merged yet, so it is bound to appear from time to time.

Http2OutputProducer seems like something from aspnetcore repo https://github.com/dotnet/aspnetcore/blob/32c9ec8d9c4b7f7f544cfb39dbf31f30a860d585/src/Servers/Kestrel/Core/src/Internal/Http2/Http2OutputProducer.cs#L15

There seem to be build failures @ManickaP

@jkoritzinsky The problem is that the build scripts (and therefore the produced binaries) are different, depending on the architecture (x64, arm, arm64) and openssl version (1.1, 3.0). So out of...