You are viewing a plain text version of this content. The canonical link for it is here.
Posted to jira@arrow.apache.org by "David Li (Jira)" <ji...@apache.org> on 2021/04/28 12:07:00 UTC
[jira] [Resolved] (ARROW-12386) [C++] Support file parallelism in
AsyncScanner
[ https://issues.apache.org/jira/browse/ARROW-12386?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
David Li resolved ARROW-12386.
------------------------------
Fix Version/s: 5.0.0
Resolution: Fixed
Issue resolved by pull request 10076
[https://github.com/apache/arrow/pull/10076]
> [C++] Support file parallelism in AsyncScanner
> ----------------------------------------------
>
> Key: ARROW-12386
> URL: https://issues.apache.org/jira/browse/ARROW-12386
> Project: Apache Arrow
> Issue Type: Improvement
> Components: C++
> Reporter: Weston Pace
> Assignee: Weston Pace
> Priority: Major
> Labels: pull-request-available
> Fix For: 5.0.0
>
> Time Spent: 2h
> Remaining Estimate: 0h
>
> Whether we pull from files in parallel or not is controlled by how we merge the batch streams in `AsyncScanner::ScanBatchesUnorderedAsync`. Currently we are relying on `MakeConcatenatedGenerator` which is incorrect. This is needed because `MakeMergedGenerator` pulls from its source (an `EnumeratingGenerator`) in an async reentrant fashion. `MakeMergedGenerator` should not do this. If some kind of readahead is truly necessary there then use `MakeReadaheadGenerator`.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)