Mailing-List: contact commits-help@cassandra.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@cassandra.apache.org
Date: Sun, 23 Mar 2014 07:05:45 +0000 (UTC)
From: "Benedict (JIRA)" <jira@apache.org>
To: commits@cassandra.apache.org
Message-ID: <JIRA.12694748.1392215110539.126586.1395558345225@arcas>
In-Reply-To: <JIRA.12694748.1392215110539@arcas>
References: <JIRA.12694748.1392215110539@arcas>
Subject: [jira] [Commented] (CASSANDRA-6694) Slightly More Off-Heap
 Memtables
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit


    [ https://issues.apache.org/jira/browse/CASSANDRA-6694?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13944357#comment-13944357 ] 

Benedict commented on CASSANDRA-6694:
-------------------------------------

FTR, with a very simple and short performance comparison, simulating writing a lot of small (integer) fields, using cassandra-stress write n=400000 -col size=fixed\(4\) n=fixed\(100\), I see a 25% throughput improvement using offheap_objects as the allocator type vs either on/off heap buffers.

I should expect to see performance improve further as the length of the test increases, as write amplification takes its toll more rapidly on the heap buffers.

> Slightly More Off-Heap Memtables
> --------------------------------
>
>                 Key: CASSANDRA-6694
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-6694
>             Project: Cassandra
>          Issue Type: Improvement
>          Components: Core
>            Reporter: Benedict
>            Assignee: Benedict
>              Labels: performance
>             Fix For: 2.1 beta2
>
>
> The Off Heap memtables introduced in CASSANDRA-6689 don't go far enough, as the on-heap overhead is still very large. It should not be tremendously difficult to extend these changes so that we allocate entire Cells off-heap, instead of multiple BBs per Cell (with all their associated overhead).
> The goal (if possible) is to reach an overhead of 16-bytes per Cell (plus 4-6 bytes per cell on average for the btree overhead, for a total overhead of around 20-22 bytes). This translates to 8-byte object overhead, 4-byte address (we will do alignment tricks like the VM to allow us to address a reasonably large memory space, although this trick is unlikely to last us forever, at which point we will have to bite the bullet and accept a 24-byte per cell overhead), and 4-byte object reference for maintaining our internal list of allocations, which is unfortunately necessary since we cannot safely (and cheaply) walk the object graph we allocate otherwise, which is necessary for (allocation-) compaction and pointer rewriting.
> The ugliest thing here is going to be implementing the various CellName instances so that they may be backed by native memory OR heap memory.


--
This message was sent by Atlassian JIRA
(v6.2#6252)