[
https://issues.apache.org/jira/browse/SPARK-27025?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16784892#comment-16784892
]
Sean Owen commented on SPARK-27025:
-----------------------------------
You'll want to cache() the thing you call toLocalIterator() on no matter what
in this case. If it's not helping, then I think the delay remains the
transferring of data to the driver, as it will all be computed and cached
before you start. The 2-at-a-time implementation could help that and I'd be
curious if it works out.
> Speed up toLocalIterator
> ------------------------
>
> Key: SPARK-27025
> URL: https://issues.apache.org/jira/browse/SPARK-27025
> Project: Spark
> Issue Type: Wish
> Components: Spark Core
> Affects Versions: 2.3.3
> Reporter: Erik van Oosten
> Priority: Major
>
> Method {{toLocalIterator}} fetches the partitions to the driver one by one.
> However, as far as I can see, any required computation for the
> yet-to-be-fetched-partitions is not kicked off until it is fetched.
> Effectively only one partition is being computed at the same time.
> Desired behavior: immediately start calculation of all partitions while
> retaining the download-a-partition at a time behavior.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
