Github user fhueske commented on the pull request:
https://github.com/apache/flink/pull/1083#issuecomment-138825832
Hi @sachingoel0101,
I think there was no consensus in the discussion (FLINK-1730) for how to
implement this feature. @StephanEwen pointed that persisting data sets should
happen on the network stack level, i.e., not on the operator level. In fact,
this feature was considered and while redesigning and implementing the network
stack by @uce. By doing it on the network stack level, persisted data sets can
also be used for failure recovery or used across different jobs within a
session.
This PR is basically ignoring the previous discussion and features of the
network stack.
If you think that persisting on an operator level has advantages over the
network stack, it would be good to share your thoughts with the community.
Unless there are good reasons to do persisting on operator level, I would
not merge this PR.
---
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.
---
