MonoTests.System.Runtime.Caching.MemoryCacheTest.Trim fails in Windows arm64

[safern]

MonoTests.System.Runtime.Caching.MemoryCacheTest.Trim

Console Log Summary

Builds

Build	Pull Request	Test Failure Count
#636357	Rolling	1
#644655	Rolling	1
#645661	Rolling	1

Configurations

• net5.0-Windows_NT-Release-arm64-CoreCLR_release-Windows.10.Arm64.Open

Helix Logs

Build	Pull Request	Console	Core	Test Results	Run Client
#636357	Rolling	console.log		testResults.xml	run_client.py
#644655	Rolling	console.log		testResults.xml	run_client.py
#645661	Rolling	console.log		testResults.xml	run_client.py

testResults.xml -nologo -nocolor -notrait category=IgnoreForCI -notrait category=OuterLoop -notrait category=failing  
  Discovering: System.Runtime.Caching.Tests (method display = ClassAndMethod, method display options = None)
  Discovered:  System.Runtime.Caching.Tests (found 29 of 36 test cases)
  Starting:    System.Runtime.Caching.Tests (parallel test collections = on, max threads = 46)
    MonoTests.System.Runtime.Caching.MemoryCacheTest.Trim [FAIL]
      Assert.Equal() Failure
      Expected: 23
      Actual:   22
      Stack Trace:
        /_/src/libraries/System.Runtime.Caching/tests/System.Runtime.Caching/MemoryCacheTest.cs(992,0): at MonoTests.System.Runtime.Caching.MemoryCacheTest.Trim()
  Finished:    System.Runtime.Caching.Tests
=== TEST EXECUTION SUMMARY ===
   System.Runtime.Caching.Tests  Total: 29, Errors: 0, Failed: 1, Skipped: 0, Time: 4.611s

[KKhurin]

I understand why we failed here. One of the cache stores got 4 items (instead of expected 1) so trimming 75% of it resulted in 1 item surviving the trim. To greatly reduce the chance of seeing it again we can trim 76%...

[safern]

@KKhurin just to update here I ran the test without the fix in a loop 30 times and it failed 3 times.

I ran the test with the fix (changing to 76%) in a loop 50 times and it didn't fail, so I think it makes sense to put that fix. I will submit a PR. Thanks for the suggestion.

Run without the fix: https://helix.dot.net/api/2019-06-17/jobs/b7f84fb0-3804-49f6-97dc-d09fc2180585/workitems/NoFix/console

Run with the fix: https://helix.dot.net/api/2019-06-17/jobs/bc764363-3621-4740-a24c-d336546cb6c2/workitems/Fix/console

[akoeplinger]

@safern did you get a chance to submit a PR with the fix? I just saw the test failing on one of my PRs.

[safern]

No, actually thanks for reminding me. Will put up the PR.