Details
-
Bug
-
Status: Open
-
Major
-
Resolution: Unresolved
-
None
-
None
-
None
-
None
Description
We are seeing some random issues in unit tests such as https://issues.apache.org/jira/browse/DRILL-5925. While we should fix these issues, we may want to have different ways to handle such situation:
1, we may want to continue running the unit tests esp. those in storage plugins regardless non-fundamental random failures happened in exec module.
2, we may want to re-run these failed tests individually. If we saw these failed tests failed the first time but passed the second run when it was run individually, we could mark these as 'random failures' and decide to continue the whole set of unit tests.