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

ci changes #571

Merged
merged 3 commits into from
May 3, 2024
Merged

ci changes #571

merged 3 commits into from
May 3, 2024

Conversation

lread
Copy link
Collaborator

@lread lread commented May 2, 2024

Please complete and include the following checklist:

  • I have read CONTRIBUTING and the Etaoin Developer Guide.

  • This PR corresponds to an issue that the Etaoin maintainers have agreed to address.

  • This PR contains test(s) to protect against future regressions

  • I have updated CHANGELOG.adoc with a description of the addressed issue.

As the current primary maintainer of etaoin, I'm gonna choose a ci build
setup that I prefer. I don't need to build every time I push to a branch
that I have not PRed yet.
@lread
Copy link
Collaborator Author

lread commented May 3, 2024

Holy smokes, I'm fixing a macOS CI issue and feel the pain of Windows CI flakiness!

@lread
Copy link
Collaborator Author

lread commented May 3, 2024

Soo, seem to be failing lots on windows/chrome. Could be just a buggy release of chrome/chromedriver for windows? Dunno. From tools-versions task we see:

| Chrome            | 124.0.6367.61                                                                                       |                                                                                 |
| Chrome Webdriver  | ChromeDriver 124.0.6367.60 (8771130bd84f76d855ae42fbe02752b03e352f17-refs/branch-heads/6367@{#798}) | C:\SeleniumWebDrivers\ChromeDriver\chromedriver.exe                             |

And I see a bunch of orphaned chromes on "Complete Job" step:

Terminate orphan process: pid (6600) (conhost)
Terminate orphan process: pid (6596) (chrome)
Terminate orphan process: pid (4412) (chrome)
Terminate orphan process: pid (6296) (conhost)
Terminate orphan process: pid (1928) (chrome)
Terminate orphan process: pid (6748) (chrome)
Terminate orphan process: pid (2204) (conhost)
Terminate orphan process: pid (744) (chrome)
Terminate orphan process: pid (3308) (chrome)
Terminate orphan process: pid (8024) (conhost)
Terminate orphan process: pid (4944) (chrome)
Terminate orphan process: pid (5864) (chrome)

So maybe, wild guess, chromedriver isn't able to properly kill chrome and maybe chrome is busy doing some unkillable nonsense.

I'm beginning to think WebDriver work is not a ton of fun!

@lread
Copy link
Collaborator Author

lread commented May 3, 2024

There are a few disparate changes in this PR.
We ask contributors not to rebase, but I'm going to for this one and then merge instead of squash merge.
I'll leave the windows/chrome thing broken with the idea that the next release of the github action windows runner will solve this for us.

lread added 2 commits May 3, 2024 15:12
The new macos runner is arm-based and missing some software is missing
that all other runners have:
- ImageMagick
- Edge and msedgedriver
- Firefox and geckodriver
Take a shot at optionally installing the above when we need them.

There is also no JDK8 for arm-based macos.
We weren't testing that particular scenario on the intel-based macos runners
but I updated the code to reflect this fact.
I keep forgetting that log output to the terminal is normal for this
test and sometimes assume something has gone wrong. So mock the logger
and also test that it is getting called as expected.
@lread lread force-pushed the lread-ci-ci-oh branch from 9bc36af to 086e590 Compare May 3, 2024 19:13
@lread lread merged commit 08e7114 into master May 3, 2024
51 of 53 checks passed
@lread lread deleted the lread-ci-ci-oh branch July 28, 2024 18:14
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.

1 participant