Github user fhueske commented on the issue:
https://github.com/apache/flink/pull/3095
Thanks for looking into this issue @KurtYoung
I agree that this is rather an issue with the tests and not with the actual
code. However, I would fix the tests a bit differently.
The goal of the code that you removed was to validate that each partition
is correctly sorted and that the partitions themselves are correctly sorted,
i.e., for a descending sort, the highest values should be in partition 0 and
the lowest in partition n.
In order to ensure parallel execution, we cannot execute the sort tests in
a collection environment but need a cluster environment. Moreover, we should
explicitly set a default parallelism on the ExecutionEnvironment to avoid that
the program is executed with parallelism 1 (a parallelism of 3 should suffice).
Once we do that we must ensure that the started minicluster offers enough slots
to run the program.
I'll add a few more inline comments to the tests.
Thanks, Fabian
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
