Skip to content

test: mark two WASI tests flaky on windows #33403

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.

Sign up for GitHub

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

Jump to bottom
Closed
cjihrig wants to merge 1 commit into from
Closed

test: mark two WASI tests flaky on windows #33403

cjihrig wants to merge 1 commit into from

Conversation

@cjihrig
Copy link
Contributor

@cjihrig cjihrig commented May 14, 2020

These tests have become flaky on Windows in the CI. The failures do not appear to be consistent, or even in the same parts of the tests. They do seem to be related to trying to open or stat files though.

Checklist
  • make -j4 test (UNIX), or vcbuild test (Windows) passes
  • commit message follows commit guidelines

@cjihrig
test: mark two WASI tests flaky on windows
d69c8f2 
These tests have become flaky on Windows in the CI. The failures
do not appear to be consistent, or even in the same parts of the
tests. They do seem to be related to trying to open or stat
files though.
@nodejs-github-bot nodejs-github-bot added the test Issues and PRs related to the tests. label May 14, 2020
@cjihrig cjihrig mentioned this pull request May 14, 2020
Closed
@cjihrig
Copy link
Contributor Author

cjihrig commented May 14, 2020 •
edited
Loading

FWIW, I'm experimenting with applying the revert from #33364 to see if it makes a difference.

From some testing last night, I saw failures on 2 out of 5 runs - both on Windows compiled with VS 2019. With the revert applied, I'm currently seeing success for 3 out of 3 4 out of 4 runs, but I'm going to make a few more runs.

@cjihrig
Copy link
Contributor Author

cjihrig commented May 14, 2020 •
edited
Loading

I ran test-wasi through a stress test on Windows with the revert applied. It did fail two times, but it certainly seems a lot less flaky with the revert applied.

Across the various Windows machines, the test ran 500 times. It failed once on win2012r2-vs2019 in a WASI stat call. It failed in the same way once on win2016-vs2017 as well. I'll have to do some more debugging but I'm OK with labeling these as flaky for now since there wasn't a 100% pass rate. This might also be a data point in favor of landing #33364.

EDIT: I also plan to look into the nature of the stat failure since Windows is prone to that IIRC from other issues.

@cjihrig cjihrig mentioned this pull request May 16, 2020
Merged
@cjihrig cjihrig deleted the flake branch May 17, 2020 15:25
@cjihrig cjihrig mentioned this pull request May 17, 2020
Closed
2 tasks
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@lpinca lpinca lpinca approved these changes

@targos targos targos approved these changes

@gengjiawen gengjiawen gengjiawen approved these changes

@richardlau richardlau richardlau approved these changes

Assignees

No one assigned

Labels

test Issues and PRs related to the tests.

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

6 participants

@cjihrig @lpinca @targos @gengjiawen @richardlau @nodejs-github-bot