This repository was archived by the owner on Dec 18, 2018. It is now read-only.
Repository navigation
This repository was archived by the owner on Dec 18, 2018. It is now read-only.
Flaky Test: MemoryPoolBlock ObjectDisposedException #2486
Description
Kestrel is timing out while running tests.
At first glance it seems like this might be caused by this:
Log Critical[0]: Unexpected exception in HttpConnection.ProcessRequestsAsync. System.ObjectDisposedException: Cannot access a disposed object.
[20:27:03] Object name: 'MemoryPoolBlock'.
[20:27:03] at System.Buffers.ThrowHelper.ThrowObjectDisposedException(ExceptionArgument argument) in D:\b\w\1e8cd6334ef22651\packages\microsoft.extensions.buffers.sources\2.1.0-preview3-32110\contentFiles\cs\netstandard1.0\Microsoft.Extensions.Buffers.Sources\PipelinesThrowHelper.cs:line 51
[20:27:03] at System.Buffers.MemoryPoolBlock.Dispose() in D:\b\w\1e8cd6334ef22651\packages\microsoft.extensions.buffers.sources\2.1.0-preview3-32110\contentFiles\cs\netstandard1.0\Microsoft.Extensions.Buffers.Sources\MemoryPoolBlock.Release.cs:line 54
[20:27:03] at System.IO.Pipelines.BufferSegment.ResetMemory()
[20:27:03] at System.IO.Pipelines.Pipe.CompletePipe()
[20:27:03] at System.IO.Pipelines.Pipe.CompleteWriter(Exception exception)
[20:27:03] at Microsoft.AspNetCore.Server.Kestrel.Core.Adapter.Internal.AdaptedPipeline.ReadInputAsync(Stream stream)
[20:27:03] at Microsoft.AspNetCore.Server.Kestrel.Core.Adapter.Internal.AdaptedPipeline.RunAsync(Stream stream) in D:\b\w\1e8cd6334ef22651\src\Kestrel.Core\Adapter\Internal\AdaptedPipeline.cs:line 45
[20:27:03] at Microsoft.AspNetCore.Server.Kestrel.Core.Internal.HttpConnection.ProcessRequestsAsync[TContext](IHttpApplication`1 httpApplication) in D:\b\w\1e8cd6334ef22651\src\Kestrel.Core\Internal\HttpConnection.cs:line 183
But I don't know the area well so that might be a red haring.
Activity
Tratcher commentedon Apr 23, 2018
dotnet/extensions#342
mikeharder commentedon Apr 23, 2018
The test hang is likely a dup of #2503. The
ObjectDisposedExceptionforMemoryPoolBlockmay also be a valid issue, but it's likely unrelated to the hang which was caused by a GC bug (https://github.com/dotnet/coreclr/issues/17716).[-]Test timeout: MemoryPoolBlock ObjectDisposedException[/-][+]Flaky Test: MemoryPoolBlock ObjectDisposedException[/+]pakrym commentedon May 14, 2018
We refactored the code that caused this exception and I don't see it failing any recent tests.