Github user mdaxini commented on the issue:
https://github.com/apache/flink/pull/6302
@indrc
In addition to removing the additional dependency, I think there should be
a test for validating the randomness of the chose algorithm, and to make sure
there are no conflicts. For a Flink Job with large number of Task Managers and
state in Terrabytes resulting in several files, there could be a possibility of
conflicts with the current random string generation method.
fyi @StefanRRichter @StephanEwen
---
