Mailing-List: contact dev-help@storm.incubator.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@storm.incubator.apache.org
Date: Wed, 26 Feb 2014 16:31:22 +0000 (UTC)
From: "Adam Lewis (JIRA)" <jira@apache.org>
To: dev@storm.incubator.apache.org
Message-ID: <JIRA.12695342.1392434585127.109524.1393432282422@arcas>
In-Reply-To: <JIRA.12695342.1392434585127@arcas>
References: <JIRA.12695342.1392434585127@arcas>
Subject: [jira] [Commented] (STORM-241) "storm jar" underperforms compared
 to scp
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit


    [ https://issues.apache.org/jira/browse/STORM-241?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13913108#comment-13913108 ] 

Adam Lewis commented on STORM-241:
----------------------------------

I'm with you on keeping configs trim...Taylor originally suggested including a config option for this when I filed the JIRA, I'm kind of ambivalent.  As Nathan said on the mailing list, the 15k default was a completely arbitrary number...300k is another arbitrary number that works better for me :)  I just want the out-of-the-box experience to be smooth for people and this issue was/is a thorn in my side in that regard.  My only concern is that 300k was specific to my particular network configuration (on that day of the week)...and it is a number I arrived at by increasing 50k at a time until I saw diminishing returns and speeds approaching SCP.

Perhaps the best way to handle this is incrementally.  Hard code 300k now, and if we start seeing requests to change the number _again_ then make it configurable.

> "storm jar" underperforms compared to scp
> -----------------------------------------
>
>                 Key: STORM-241
>                 URL: https://issues.apache.org/jira/browse/STORM-241
>             Project: Apache Storm (Incubating)
>          Issue Type: Improvement
>         Environment: Internet connected dev machine deploying to storm cluster on AWS.  Storm 0.9.0.1
>            Reporter: Adam Lewis
>              Labels: configuration, newbie
>
> "storm jar" is very slow at sending the jar file, achieving roughly 10% of the throughput compared to SCP for the same network configuration.  This is especially problematic when deploying jars from over the Internet.
> This appears to be related to the chunk size used to transmit the JAR file over the thrift protocol.  Anecdotally, a 300KB chunk size (compared to the currently used default of 15KB), achieved performance in line with scp.
> Default can be overridden in the BufferFileInputStream ctor: https://github.com/apache/incubator-storm/blob/master/storm-core/src/jvm/backtype/storm/StormSubmitter.java?source=cc#L154


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