Github user davies commented on the pull request:
https://github.com/apache/spark/pull/6714#issuecomment-110432696
@airhorns That change is make by me (my mistake), to speed up the unit test
of external sorting. Because we adjust the batch size during external
aggregation (the size of objects vary from key to key), just borrowed that
logic here without notice that `elif not chunks:` below.
@pwendell Could we not block 1.4 release by this one? 1.4 release is
delayed much and many people are waiting for it. This issue could be workaround
by increasing the number of partitions during sorting (this could also improve
performance by avoid spilling).
@kevincox Is the current approach something similar what you had suggested?
---
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]
