Github user IceMan81 commented on the issue:
https://github.com/apache/spark/pull/7379
@zzeekk Would you mind explaining how your workaround works.
> A Workaround is to build blocks and add them as equi-join condition
Not sure I understand what you are suggesting here.
@marmbrus Inability to do range join efficiently results in very poor
performance. Are there plans on addressing this directly in an upcoming
release? I've scenarios where the optimizer sorts the results into the single
partition for the join (all other partitions are empty) because the sort does
not include the columns in the range condition. And this task will run for more
than a day which a forced broadcast version of it will run in 3 hours. And here
I'm only able to do the boradcast because I'm using a smaller data set on one
side of the join.
---
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]
