[ https://issues.apache.org/jira/browse/SPARK-4740?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14239102#comment-14239102
]
Aaron Davidson commented on SPARK-4740:
---------------------------------------
I tried to reproduce this on an EC2 cluster is 4 i2.8xlarge. These have 32 vCPUs, 8 SSDs,
but my interconnect is probably only 1GigE. I ran with the same parameters, except 10x fewer
values, to iterate more quickly. Unfortunately, I was unable to reproduce the issue -- Netty
ran with all nodes at the same speed and slightly faster than NIO.
However, looking at your graphs, the 1GigE may be the bottleneck here, so tomorrow morning
I will try to run with 10GigE and the normal data size.
> 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
> Assignee: Reynold Xin
> Priority: Blocker
> Attachments: (rxin patch better executor)TestRunner sort-by-key - Thread dump
for executor 3_files.zip, (rxin patch normal executor)TestRunner sort-by-key - Thread dump
for executor 0 _files.zip, Spark-perf Test Report 16 Cores per Executor.pdf, Spark-perf Test
Report.pdf, TestRunner sort-by-key - Thread dump for executor 1_files (Netty-48 Cores per
node).zip, TestRunner sort-by-key - Thread dump for executor 1_files (Nio-48 cores per node).zip,
rxin_patch-on_4_node_cluster_48CoresPerNode(Unbalance).7z
>
>
> 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
|