Github user rxin commented on the pull request:
https://github.com/apache/spark/pull/11105#issuecomment-197570550
We have recently changed accumulators so they can satisfy the use case of
metrics (e.g. counting the size of bytes read) in
https://issues.apache.org/jira/browse/SPARK-10620
The definition of "consistent here" really doesn't resonate with me. The
design doc in SPARK-10620 also outlined the complexity beyond what this ticket
discusses (i.e. there are cases that I can't really come up with valid
semantics for accumulators, especially in shared dags and retries).
The implementation here also regresses performance due to the extra
wrapping of iterators, which are becoming increasingly expensive.
---
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.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
