Mailing-List: contact issues-help@tez.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@tez.apache.org
Date: Tue, 14 Jul 2015 10:55:04 +0000 (UTC)
From: "Tsuyoshi Ozawa (JIRA)" <jira@apache.org>
To: issues@tez.apache.org
Message-ID: <JIRA.12843542.1436368752000.180890.1436871304824@Atlassian.JIRA>
In-Reply-To: <JIRA.12843542.1436368752000@Atlassian.JIRA>
References: <JIRA.12843542.1436368752000@Atlassian.JIRA>
 <JIRA.12843542.1436368752838@arcas>
Subject: [jira] [Commented] (TEZ-2606) Cache-friendly data structure for
 sorting
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit


    [ https://issues.apache.org/jira/browse/TEZ-2606?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14626192#comment-14626192 ] 

Tsuyoshi Ozawa commented on TEZ-2606:
-------------------------------------

[~gopalv] Thank you for the clarification. I think my suggestion has been already implemented by using TezBytesComparator. Can we have TezIntComparator or TezTextComparator for accelerating MapReduce jobs? 

Additionally, I think we can implement SIMD-based bitonic merge sorting by using the prefix. I'd like to continue to work on TEZ-2607, but do you have any concern about it?

> Cache-friendly data structure for sorting
> -----------------------------------------
>
>                 Key: TEZ-2606
>                 URL: https://issues.apache.org/jira/browse/TEZ-2606
>             Project: Apache Tez
>          Issue Type: Sub-task
>            Reporter: Tsuyoshi Ozawa
>            Assignee: Tsuyoshi Ozawa
>
> Alphasort[1]  mentions prefix key sort is effective way. I'd like to suggest to change a layout of ring buffer to include prefix of key in meta data. This can improve the cache hit rate when sorting.
> [1] Alphasort: http://dl.acm.org/citation.cfm?id=615237


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