-
Notifications
You must be signed in to change notification settings - Fork 91
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
Maven 4 support #1674
Comments
We cannot remove |
Flaky tests have negative value. They must go away. Sooner than later. If not by removing them, then by fixing them (or rewriting or whatever). Or imagine this. I create a new PR x in 15 minutes. Then I have to wait 40? Minutes on average for you to approve the test run. The test fail because they are flaky. I have to wait another 40? Minutes or so until the tests are re-run. That's why they provide negative value. They cost way more than they actually help. "several completely green" is not good enough. It must be reliable. Related: https://gradle.com/blog/do-you-regularly-schedule-flaky-test-days/ https://www.atkinsondev.com/post/improve-developer-productivity/ |
@bmarwell I agree with you to an extent. It absolutely is costing us money, and our contributors money, and aggravation. However, if we remove the "flaky" test, there will be no tests for dev mode. We do not know how to fix the test. It passes on the majority of the builds (e.g. it will fail on 2 out of 8 builds). It doesn't consistently fail on the same build (combination of OS, Java, OL vs WL, etc). Perhaps we can break up the |
One more thing on flaky tests...when a Windows build fails...it is almost always due to a file locking issue which is OS specific and I have not been able to find a way to guard our tests against it. I did several things a while ago that reduced the number of times we run into it, but I do not know of a full proof solution for that. |
Do you know the source (class) of the file lock? If it's maven, you could try the new resolver. IIRC, there were some changes. Maybe a split repo, workspace-local repo etc would help, too? Any stack trace appreciated, I can see if I can boot up a windows VM to get behind it. |
Thanks Ben. It is usually this file and it is related to the Liberty server itself. When the server is stopped, it should release the lock on the file. It either doesn't or it takes too much time to do so.
See this log file for a recent example. It happens on different random tests when trying to clean after the Liberty server is stopped. |
@bmarwell I refreshed the |
A couple of warnings are left:
|
@bmarwell Do you have a pointer on how/where to fix that? Those are all in the pom.xml as |
@cstamas did a check and did not get the warnings. Maybe my mirror refused to update the snapshot. I'll try again from home. |
Maven is working on its v4 release. There are already a few alpha releases available and a GA release is likely to come later this year. Liberty Maven plugin should support Maven 4 after it reaches GA.
Early work has been done with thanks to contribution from @bmarwell as noted in #1651.
The text was updated successfully, but these errors were encountered: