System.Net.* cancellation tests fail in CI #72818
Comments
ghost
added
the
|
Tagging subscribers to this area: @dotnet/ncl Issue DetailsRunfo Creating Tracking Issue (data being generated)
|
noahfalk
added
the
blocking-clean-ci
|
WASM ones are some install failure. The Mac and Debian failures are assertion failures top of stack looks like or Note the fd is -1. Are the "long running" cases where the assertion has popped a dialog or something? Or different issue? |
|
This afternoon's rolling build just failed in the System.Net.Http.Functional.Tests test (http not mail), but the error in the log has a very similar pattern of 7 iterations of cancellation tests 2 minutes apart followed by a timeout. This morning it was Linux x64, now it is OSX x64. |
noahfalk
changed the title
|
@danmoseley - For the moment I am guessing we've potentially got multiple failures hitting the same test. I opened #72830 to track the Assertion `0 <= fd && fd < sysconf(_SC_OPEN_MAX) issue. If it turns out the cancellation and the fd have the same root cause we can always dupe it. |
|
@dotnet/ncl |
|
The failures in this issue also look similar to failures in System.IO.Pipes.Tests (#72585) which recently had tests disabled. |
noahfalk
changed the title
The HTTP failures are the same as #72596 (these are just the ValueTask versions of those same tests). @rzikm, can you please disable these variants as well against the original issue? |
antonfirsov
removed
the
untriaged
|
|
|
System.Net.Mail.Functional tests are still failing in CI as of this morning. https://runfo.azurewebsites.net/view/build/?number=1909685. |
|
System.Net.Mail.Functional tests are still failing in CI. I submitted PR #73340 to disable the test. |
This test is failing jobs in CI, see issue dotnet#72818
|
Ah, @antonfirsov accidentally removed untriaged without setting milestone, so it did not show up in our Triage queries nor 7.0 work :( |
|
@noahfalk we will need to somehow break down this mega issue - it is rather confusing. |
This test is failing jobs in CI, see issue #72818
|
At the moment this issue is intended to track exactly the two failures listed in first comment, and I am hoping that with PR #73340 it is now mitigated. If any other tests fail with a similar issue I can open a new issue to track them. There was also the assert 0 <= fd thing which was a different error occuring in the same tests which has a separate issue already tracking it.
I was relying on runfo's updating the GitHub issue with a list of builds where it occurred + looking at the most recent few rolling builds in runfo. I was also looking into potential runfo UI changes because I agree as the test monitor it isn't trivial to determine which issues remain hot at a glance. |
|
I'm waiting to see the afternoon rolling build, but assuming we see no failures there I will remove the blocking CI tag. |
|
Yeah, Runfo is really not useful for tracking these kind of failures (i.e. not just specific test) -- we did it by hand until now to avoid confusion. |
Closing this issue as it tracks basically anything wrong in Networking, which is not useful. |
System.Net.Mail.Functional.Tests test timed out in the rolling build 1901005
System.Net.Http.Functional.Tests test timed out with a very similar looking error in rolling build1902723
Between the two tests Runfo reports ~140 failures in the last 7 days. There are also some System.Net.Mail.Functional.Tests that failed in different ways and I am trying to filter those out. Every example from System.Net.Http.Functional.Tests has looked similar so far.
System.Net.Mail.Functional.Tests
https://helixre107v0xdeko0k025g8.blob.core.windows.net/dotnet-runtime-refs-heads-main-6bd0a6280a374277a1/System.Net.Mail.Functional.Tests/1/console.531221cb.log?%3Fhelixlogtype%3Dresult
System.Net.Http.Functional.Tests
https://helixre107v0xdeko0k025g8.blob.core.windows.net/dotnet-runtime-refs-heads-main-287f10d639984af991/System.Net.Http.Functional.Tests/1/console.8747f084.log?%3Fhelixlogtype%3Dresult
Runfo Tracking Issue: system.net.http.functional.tests work item
Build Result Summary
The text was updated successfully, but these errors were encountered: