Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[WIP] Fix flaky tests #3827

Draft
wants to merge 85 commits into
base: master
Choose a base branch
from
Draft

[WIP] Fix flaky tests #3827

wants to merge 85 commits into from

Conversation

kciesielski
Copy link
Member

No description provided.

@kciesielski
Copy link
Member Author

Issue still happens on Java 21

2024-06-07T10:54:12.5578064Z [TESTINFO] Test sttp.tapir.server.vertx.zio.VertxZioServerStubTest PASSED in 3 ms,601138
2024-06-07T10:54:12.5578953Z [TESTINFO] Finished test group: sttp.tapir.server.vertx.zio.VertxZioServerStubTest result: Passed,601138
2024-06-07T10:54:12.5579823Z �[0m[�[0m�[0minfo�[0m] �[0m�[0m�[36mRun completed in 33 seconds, 940 milliseconds.�[0m�[0m
2024-06-07T10:54:12.5580497Z �[0m[�[0m�[0minfo�[0m] �[0m�[0m�[36mTotal number of tests run: 271�[0m�[0m
2024-06-07T10:54:12.5581099Z �[0m[�[0m�[0minfo�[0m] �[0m�[0m�[36mSuites: completed 4, aborted 0�[0m�[0m
2024-06-07T10:54:12.5581851Z �[0m[�[0m�[0minfo�[0m] �[0m�[0m�[36mTests: succeeded 271, failed 0, canceled 0, ignored 0, pending 0�[0m�[0m
2024-06-07T10:54:12.5582541Z �[0m[�[0m�[0minfo�[0m] �[0m�[0m�[32mAll tests passed.�[0m�[0m
2024-06-07T11:06:20.2099155Z �[34mclientTestServer�[0m ... killing ...
2024-06-07T11:06:30.2053475Z ##[warning]Attempt 1 failed. Reason: Timeout of 900000ms hit

I'm curious about this line:

2024-06-07T11:06:20.2099155Z �[34mclientTestServer�[0m ... killing ...

it's the last line logged, and it is logged after a long gap. Then, timeout trigger kicks in.

@kciesielski
Copy link
Member Author

I found one problematic test - "empty client stream" in ServerWebSocketTests. It fails pretty frequently for VertX. The freezes aren't probably directly related, but fixing it should help a bit anyway.

@kciesielski
Copy link
Member Author

Spotted tests that fail sometimes, fixing them in a separate PR
#3847

@kciesielski
Copy link
Member Author

Disabling tests for all zio-related backends made the issue go away.
Restoring http4s-zio made it reappear. I saw a lot of exceptions in logs on closing WebSockets by http4s-zio, so I disabled only WS tests for this backend. It didn't help, after a few runs another timeout occured.
What's noteworthy, logs from this run show that all zio-http4s related tests pass. However, there seems something to be happening with verx-cats. There's a "starting" entry logged:

[TESTINFO] Starting test group: sttp.tapir.server.vertx.cats.VertxCatsServerStubStreamingTest

but no corresponding [TESTINFO] Finished test group: entry. There are no logs for this backend at all.

@adamw
Copy link
Member

adamw commented Jul 22, 2024

Did you find that separating the ZIO tests into separate build steps increases the number of builds that pass? I'm thinking of introducing only this change for now, even if it doesn't solve the issue, maybe it will improve our builds nonetheless

@kciesielski
Copy link
Member Author

This separation shows that it's clearly the zio part that fails with freezes.
In combination with pong issues #3944 these failures often exhaust retries.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants