-
Notifications
You must be signed in to change notification settings - Fork 14.1k
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 is extremely flaky when running Presto/Hive tests #17750
Comments
A little clue for this issue. I will try to update the images or increase Java heap memory for fixing it. https://github.com/apache/superset/runs/4530486125?check_suite_focus=true#step:9:5105 |
Just one more case: In this PR, I've had repeated failures of
I'm on my third run now... wish me luck 🤞 |
Dropping a link to the Slack thread where there's more discussion on this. |
I found that some large SQL insert tasks will cause Presto null-pointer error. I submit the same SQL through presto-cli or JDBC without any problems.
|
These tests are the same ones that run against other analytics DBs, but for some reason when we run them against Presto/Hive we get false positives (or are they false negatives? 🤔) at a much greater rate. These tests also seem to take much longer to run, which may be related.
If we can get to the bottom of this it will massively improve the experience of contributing to Superset.
How to reproduce the bug
Run CI a few times.
Expected results
CI should pass if the PR hasn't introduced any bugs.
Actual results
It fails a lot. Often multiple times in a row, with different tests.
Additional context
I want to start putting together a list of the tests that have failed, to see if there is some sort of pattern. Feel free to add new entries to this list (only if you are certain that it really is a false positive!). Please also link to a relevant line of the GitHub Action where the test failed.
The text was updated successfully, but these errors were encountered: