Github user aalexandrov commented on the pull request:
https://github.com/apache/flink/pull/605#issuecomment-93719337
:+1: thanks for the great work! I'll review that (probably over the
weekend) and will appreciate if some of the core Flink committers (@sewen,
@rmetzger, @fhueske) can also make a pass over the code.
One more caveat from me: this implements only the runtime aspect of the
statistics collecting logic.
A second PR which allows to configure the points where statistics should be
tracked in a programmatic way as part of the DataBag API shoud follow.
@tammymendt and me were discussing as syntax along the lines of:
```scala
A = // some dataflow assembly code
A.withStatistics( "statsForX", keySelectorFn )
env.execute()
// grab the statistics after the execution is done
env.getAccumulator("statsForX")
```
Once this is in place we will play around and implement some ideas on
incremental optimization.
---
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 infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---
