Github user tillrohrmann commented on the pull request:
https://github.com/apache/flink/pull/1047#issuecomment-134145892
I think both ways (increasing the number of times we sample as well as
softening the accuracy thresholds) will improve the test stability. However, as
@ChengXiangLi pointed out, increasing the number of samplings and the sample
size, will also increase the unit test costs (even though it should be in the
range of fractions of a second). Maybe we can also combine both approaches.
Other than that, LGTM.
---
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.
---
