Github user ash211 commented on the pull request:
https://github.com/apache/spark/pull/2401#issuecomment-57760828
This looks very reasonable. Counting the executor's bookkeeping core
against the resources also seems much more correct than pretending it doesn't
exist like before.
Regarding the fraction-vs-absolute debate: if down the line users find
setting the overhead with absolute numbers to be suboptimal, we can always add
a new parameter named `spark.mesos.executor.memoryOverheadPercent` that
defaults to 7 and can be changed to 15 or higher as needed.
Considering Brenden's test on the cluster and experience with Mesos this
seems good to merge pending extremely minor typos
---
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.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
