Github user StephanEwen commented on the pull request:
https://github.com/apache/flink/pull/408#issuecomment-77774282
The vertex-centric iterations use semantic annotations a lot to save
sorting/partitioning where possible. This results in a plan for the
vertex-centric iterations that partitions one and sorts one, despite the fact
that it uses wither two coGroups or 2 joins plus a reduce. I think this would
help big time here as well. In general, plan optimization may be something that
Gelly could take a look at.
---
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.
---
