Mailing-List: contact issues-help@spark.apache.org; run by ezmlm
Precedence: bulk
Date: Fri, 4 Mar 2016 02:44:40 +0000 (UTC)
From: "Michael Armbrust (JIRA)" <jira@apache.org>
To: issues@spark.apache.org
Message-ID: <JIRA.12946865.1457059424000.1393.1457059480784@Atlassian.JIRA>
In-Reply-To: <JIRA.12946865.1457059424000@Atlassian.JIRA>
References: <JIRA.12946865.1457059424000@Atlassian.JIRA>
 <JIRA.12946865.1457059424958@arcas>
Subject: [jira] [Created] (SPARK-13664) Simplify and Speedup
 HadoopFSRelation
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit

Michael Armbrust created SPARK-13664:
----------------------------------------

             Summary: Simplify and Speedup HadoopFSRelation
                 Key: SPARK-13664
                 URL: https://issues.apache.org/jira/browse/SPARK-13664
             Project: Spark
          Issue Type: Improvement
          Components: SQL
            Reporter: Michael Armbrust
            Assignee: Michael Armbrust
            Priority: Blocker


A majority of Spark SQL queries likely run though {{HadoopFSRelation}}, however there are currently several complexity and performance problems with this code path:
 - The class mixes the concerns of file management, schema reconciliation, scan building, bucketing, partitioning, and writing data.
 - For very large tables, we are broadcasting the entire list of files to every executor. [SPARK-11441]
 - For partitioned tables, we always do an extra projection.  This results not only in a copy, but undoes much of the performance gains that we are going to get from vectorized reads.

This is an umbrella ticket to track a set of improvements to this codepath.


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

---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@spark.apache.org
For additional commands, e-mail: issues-help@spark.apache.org