Mailing-List: contact issues-help@hbase.apache.org; run by ezmlm
Precedence: bulk
Date: Tue, 7 Mar 2017 11:00:40 +0000 (UTC)
From: "Yu Li (JIRA)" <jira@apache.org>
To: issues@hbase.apache.org
Message-ID: <JIRA.13048834.1488882695000.490.1488884440077@Atlassian.JIRA>
In-Reply-To: <JIRA.13048834.1488882695000@Atlassian.JIRA>
References: <JIRA.13048834.1488882695000@Atlassian.JIRA> <JIRA.13048834.1488882695746@jira-lw-us.apache.org>
Subject: [jira] [Commented] (HBASE-17744) Implement simple embedded mode
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit
archived-at: Tue, 07 Mar 2017 11:00:44 -0000


    [ https://issues.apache.org/jira/browse/HBASE-17744?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15899239#comment-15899239 ] 

Yu Li commented on HBASE-17744:
-------------------------------

We could use this simple embedded mode for performance testing of:
1. The core part without both client and RPC logics, with DirectOperator
2. The core and client performance without RPC logics, with TableOperator

And with distributed mode, we could test the end-to-end performance. In this way, we could easily check and make sure no performance regression for each part (client, RPC and "core") and as a whole.

We have found some performance and even correctness issue in current master codes, and will open new JIRAs to fix them.

We've also modified YCSB to test hbase in embedded mode, will share the method and data later.

> Implement simple embedded mode
> ------------------------------
>
>                 Key: HBASE-17744
>                 URL: https://issues.apache.org/jira/browse/HBASE-17744
>             Project: HBase
>          Issue Type: Sub-task
>            Reporter: Yu Li
>            Assignee: Yu Li
>         Attachments: HBASE-17744.patch
>
>
> In simple embedded mode, we will
> 1. Start only one HMaster thread, which also serves as regionserver, and could serve tables other than meta
> 2. Implement direct access w/o RPC
> 3. Keep using the existing region assignment, procedure and recovery mechanism
> 4. Keep using zookeeper (mini-cluster)
> More details please refer to the patch.


--
This message was sent by Atlassian JIRA
(v6.3.15#6346)