Mailing-List: contact dev-help@hive.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@hive.apache.org
Date: Thu, 23 Jan 2014 20:09:39 +0000 (UTC)
From: "Xuefu Zhang (JIRA)" <jira@apache.org>
To: hive-dev@hadoop.apache.org
Message-ID: <JIRA.12689343.1389903227824.2808.1390507779812@arcas>
In-Reply-To: <JIRA.12689343.1389903227824@arcas>
References: <JIRA.12689343.1389903227824@arcas>
Subject: [jira] [Updated] (HIVE-6217) Refactor Beeline's JDBC connection to
 use property map instead of long url
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit


     [ https://issues.apache.org/jira/browse/HIVE-6217?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Xuefu Zhang updated HIVE-6217:
------------------------------

    Attachment: HIVE-6217.1.patch

> Refactor Beeline's JDBC connection to use property map instead of long url
> --------------------------------------------------------------------------
>
>                 Key: HIVE-6217
>                 URL: https://issues.apache.org/jira/browse/HIVE-6217
>             Project: Hive
>          Issue Type: Improvement
>          Components: CLI, JDBC
>    Affects Versions: 0.12.0
>            Reporter: Xuefu Zhang
>            Assignee: Xuefu Zhang
>         Attachments: HIVE-6217.1.patch, HIVE-6217.1.patch, HIVE-6217.patch
>
>
> To make a connection to HiverServer2, currently Beeline constructs a long url that contains hiveconf and hivevariables. At JDBC driver side, it parses the url into individual maps. Constructing and decomposing the url is complicated and unnecessary, using a different method from JDBC DriverManager that takes a property map makes the logic simpler, especially when hiveconf and hivevar variables can also come from Beeline command line in addition to those in connection url.
> This is identified in and required by HIVE-6173. 


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