You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@drill.apache.org by "Bhallamudi Venkata Siva Kamesh (JIRA)" <ji...@apache.org> on 2015/08/28 11:55:46 UTC
[jira] [Commented] (DRILL-3720) Avro Record Reader should process
Avro files by per block basis
[ https://issues.apache.org/jira/browse/DRILL-3720?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14718296#comment-14718296 ]
Bhallamudi Venkata Siva Kamesh commented on DRILL-3720:
-------------------------------------------------------
Attached patch in review board. Review request : [37885|https://reviews.apache.org/r/37885/]
> Avro Record Reader should process Avro files by per block basis
> ---------------------------------------------------------------
>
> Key: DRILL-3720
> URL: https://issues.apache.org/jira/browse/DRILL-3720
> Project: Apache Drill
> Issue Type: Improvement
> Components: Storage - Other
> Affects Versions: 1.0.0, 1.1.0
> Reporter: Bhallamudi Venkata Siva Kamesh
> Assignee: Jacques Nadeau
> Fix For: 1.2.0
>
>
> Currently Avro Record Reader processes entire file in single drill bit. This is fine as long as files are of size HDFS block size. However, if the size of the file is *large*, it causes a lot of performance issues.
> To address this, Avro Record Reader should process Avro files by per block basis.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)