You are viewing a plain text version of this content. The canonical link for it is here.

Posted to jira@arrow.apache.org by "Weston Pace (Jira)" <ji...@apache.org> on 2021/04/14 16:55:00 UTC

[jira] [Created] (ARROW-12386) [C++] Support file parallelism in AsyncScanner

Weston Pace created ARROW-12386:
-----------------------------------

             Summary: [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


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)