Mailing-List: contact dev-help@hive.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@hive.apache.org
Date: Thu, 13 Nov 2014 05:51:34 +0000 (UTC)
From: "Xuefu Zhang (JIRA)" <jira@apache.org>
To: hive-dev@hadoop.apache.org
Message-ID: <JIRA.12725064.1404327606000.488482.1415857894818@Atlassian.JIRA>
In-Reply-To: <JIRA.12725064.1404327606000@Atlassian.JIRA>
References: <JIRA.12725064.1404327606000@Atlassian.JIRA>
 <JIRA.12725064.1404327606402@arcas>
Subject: [jira] [Commented] (HIVE-7333) Create RDD translator, translating
 Hive Tables into Spark RDDs [Spark Branch]
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit


    [ https://issues.apache.org/jira/browse/HIVE-7333?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14209321#comment-14209321 ] 

Xuefu Zhang commented on HIVE-7333:
-----------------------------------

Hi [~rxin], thanks for your input. This is rather interesting. Recently we run some performance testing on ORC file. It actcually performs worse when vectorization is on. I guess this stands to reason because rows are still read one at a time. 

We would like to make Spark RDD work with vectorization, though I don't have a concrete idea of how. It seems likely that work is needed on both Spark and Hive. Any further input would be grateful. Thanks!

> Create RDD translator, translating Hive Tables into Spark RDDs [Spark Branch]
> -----------------------------------------------------------------------------
>
>                 Key: HIVE-7333
>                 URL: https://issues.apache.org/jira/browse/HIVE-7333
>             Project: Hive
>          Issue Type: Sub-task
>          Components: Spark
>            Reporter: Xuefu Zhang
>            Assignee: Rui Li
>              Labels: Spark-M1
>
> Please refer to the design specification.


--
This message was sent by Atlassian JIRA
(v6.3.4#6332)