[jira] [Commented] (SPARK-4740) Netty's network throughput is about 1/2 of NIO's in spark-perf sortByKey

["Reynold Xin (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/SPARK-4740?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14235009#comment-14235009 ] 

Reynold Xin commented on SPARK-4740:
------------------------------------

One other thing that would be great to try is to disable transferTo. You can do that in FileSegmentManagedBuffer.

{code}
  @Override
  public Object convertToNetty() throws IOException {
    if (conf.lazyFileDescriptor()) {
      return new LazyFileRegion(file, offset, length);
    } else {
      FileChannel fileChannel = new FileInputStream(file).getChannel();
      return new DefaultFileRegion(fileChannel, offset, length);
    }
  }
{code}

You just need to return a ByteBuf (that reads the content of the file segment in) instead of returning a FileRegion.


> Netty's network throughput is about 1/2 of NIO's in spark-perf sortByKey
> ------------------------------------------------------------------------
>
>                 Key: SPARK-4740
>                 URL: https://issues.apache.org/jira/browse/SPARK-4740
>             Project: Spark
>          Issue Type: Improvement
>          Components: Shuffle, Spark Core
>    Affects Versions: 1.2.0
>            Reporter: Zhang, Liye
>         Attachments: Spark-perf Test Report.pdf, TestRunner  sort-by-key - Thread dump for executor 1_files (48 Cores per node).zip
>
>
> When testing current spark master (1.3.0-snapshot) with spark-perf (sort-by-key, aggregate-by-key, etc), Netty based shuffle transferService takes much longer time than NIO based shuffle transferService. The network throughput of Netty is only about half of that of NIO. 
> We tested with standalone mode, and the data set we used for test is 20 billion records, and the total size is about 400GB. Spark-perf test is Running on a 4 node cluster with 10G NIC, 48 cpu cores per node and each executor memory is 64GB. The reduce tasks number is set to 1000. 



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

---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@spark.apache.org
For additional commands, e-mail: issues-help@spark.apache.org