Github user mxm commented on the pull request:
https://github.com/apache/flink/pull/2056
I'm not sure whether the changes fix the problem. I think you were lucky
with the build machine on Travis :) The issue on Travis is that the
`ExecutionEnvironment` defaults to using the number of cores reported by
`Runtime.getRuntime().availableProcessors()` as task slots. On Travis this can
be `32`. We need to adjust the number of network buffers correctly. In
addition, setting an upper limit for the parallelism for tests would also make
sense.
---
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.
---
