hadoop-common-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Sam Pullara (JIRA)" <j...@apache.org>
Subject [jira] Commented: (HADOOP-3164) Use FileChannel.transferTo() when data is read from DataNode.
Date Mon, 21 Apr 2008 03:34:23 GMT

    [ https://issues.apache.org/jira/browse/HADOOP-3164?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12590837#action_12590837
] 

Sam Pullara commented on HADOOP-3164:
-------------------------------------

I tried enabling this on Mac OS X 10.5.2.  Throughput dropped by 8x and CPU only dropped by
2x.  My test was just cat'ing a 13G file out of a DataNode with 2 disks:

hadoop -fs cat wikipedia.xml > /dev/null

Without the patch I can get around 110-120MB/s (about the average speed of the two disks)
while with the patch enabled I get around 16MB/s.

> Use FileChannel.transferTo() when data is read from DataNode.
> -------------------------------------------------------------
>
>                 Key: HADOOP-3164
>                 URL: https://issues.apache.org/jira/browse/HADOOP-3164
>             Project: Hadoop Core
>          Issue Type: Improvement
>          Components: dfs
>            Reporter: Raghu Angadi
>            Assignee: Raghu Angadi
>             Fix For: 0.18.0
>
>         Attachments: HADOOP-3164.patch, HADOOP-3164.patch, HADOOP-3164.patch, HADOOP-3164.patch
>
>
> HADOOP-2312 talks about using FileChannel's [{{transferTo()}}|http://java.sun.com/javase/6/docs/api/java/nio/channels/FileChannel.html#transferTo(long,%20long,%20java.nio.channels.WritableByteChannel)]
and [{{transferFrom()}}|http://java.sun.com/javase/6/docs/api/java/nio/channels/FileChannel.html#transferFrom(java.nio.channels.ReadableByteChannel,%20long,%20long)]
in DataNode. 
> At the time DataNode neither used NIO sockets nor wrote large chunks of contiguous block
data to socket. Hadoop 0.17 does both when data is seved to clients (and other datanodes).
I am planning to try using transferTo() in the trunk. This might reduce DataNode's cpu by
another 50% or more.
> Once HADOOP-1702 is committed, we can look into using transferFrom().

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


Mime
View raw message