Mailing-List: contact commits-help@cassandra.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@cassandra.apache.org
Date: Mon, 17 Aug 2015 23:30:46 +0000 (UTC)
From: "Blake Eggleston (JIRA)" <jira@apache.org>
To: commits@cassandra.apache.org
Message-ID: <JIRA.12769485.1421971503000.89490.1439854246400@Atlassian.JIRA>
In-Reply-To: <JIRA.12769485.1421971503000@Atlassian.JIRA>
References: <JIRA.12769485.1421971503000@Atlassian.JIRA>
 <JIRA.12769485.1421971503188@arcas>
Subject: [jira] [Commented] (CASSANDRA-8671) Give compaction strategy more
 control over where sstables are created, including for flushing and
 streaming.
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit


    [ https://issues.apache.org/jira/browse/CASSANDRA-8671?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14700430#comment-14700430 ] 

Blake Eggleston commented on CASSANDRA-8671:
--------------------------------------------

Here's the branch: https://github.com/bdeggleston/cassandra/tree/8671-2

and the last test runs: 
http://cassci.datastax.com/job/bdeggleston-8671-2-testall/2/
http://cassci.datastax.com/job/bdeggleston-8671-2-dtest/2/

Doesn't look like there's anything failing that isn't already failing on cassandra-3.0

> Give compaction strategy more control over where sstables are created, including for flushing and streaming.
> ------------------------------------------------------------------------------------------------------------
>
>                 Key: CASSANDRA-8671
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-8671
>             Project: Cassandra
>          Issue Type: Improvement
>            Reporter: Blake Eggleston
>            Assignee: Blake Eggleston
>             Fix For: 3.x
>
>         Attachments: 0001-C8671-creating-sstable-writers-for-flush-and-stream-.patch, 8671-giving-compaction-strategies-more-control-over.txt
>
>
> This would enable routing different partitions to different disks based on some user defined parameters.
> My initial take on how to do this would be to make an interface from SSTableWriter, and have a table's compaction strategy do all SSTableWriter instantiation. Compaction strategies could then implement their own SSTableWriter implementations (which basically wrap one or more normal sstablewriters) for compaction, flushing, and streaming. 


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