Mailing-List: contact dev-help@hive.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@hive.apache.org
Date: Thu, 23 Jan 2014 19:32:37 +0000 (UTC)
From: "Sergey Shelukhin (JIRA)" <jira@apache.org>
To: hive-dev@hadoop.apache.org
Message-ID: <JIRA.12690860.1390505447955.2568.1390505557695@arcas>
In-Reply-To: <JIRA.12690860.1390505447955@arcas>
References: <JIRA.12690860.1390505447955@arcas>
Subject: [jira] [Created] (HIVE-6292) ponder doing stats-based aggregations
 in metastore
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit

Sergey Shelukhin created HIVE-6292:
--------------------------------------

             Summary: ponder doing stats-based aggregations in metastore
                 Key: HIVE-6292
                 URL: https://issues.apache.org/jira/browse/HIVE-6292
             Project: Hive
          Issue Type: Improvement
          Components: Metastore, Statistics
            Reporter: Sergey Shelukhin


StatsOptimizer currently fetches partition stats from metastore and then performs the aggregation. We could do some of these aggregations directly in metastore. It's probably not a very good idea to mess with SQL queries against underlying database, but at least if MS did the same loops we could avoid a bunch of network communication; that would be esp. good for tables with very large number of partitions.


--
This message was sent by Atlassian JIRA
(v6.1.5#6160)