Closed cesarblum closed 7 years ago
I'm wary of increasing timeouts to "fix" tests. The average time it takes this test to pass is 21ms, it currently fails after 1s. Why is it reasonable to expect a test to sometimes take hundreds of times longer than normal? Did you try running the test in a loop/parallel loop to get a repro?
Did you try running the test in a loop/parallel loop to get a repro?
It's exactly what I did to repro. It ran fine in a loop with each iteration completing in a short amount of time, but as soon as I added load to the machine (I did it by opening a bunch of Windows Explorer windows very quickly 😛) it timed out.
Can we at least file a bug to investigate this bottleneck? Even if we don't consider it as a bug it's certainly something we could do to improve performance.
Sure. Can you file it?
CI link: http://aspnetci/project.html?tab=testDetails&projectId=XPlat&testNameId=-5004057257646002980&page=1
I could only repro it on a slow VM under load, and increasing the timeout interval seems to fix it.
cc @ryanbrandenburg