[
https://issues.apache.org/jira/browse/HDFS-8829?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14709419#comment-14709419
]
Colin Patrick McCabe commented on HDFS-8829:
--------------------------------------------
thanks for this, [~He Tianyi].
* {{HdfsConstants#DEFAULT_DATA_SOCKET_SIZE}}: given that we are keeping this
around, let's remove the deprecated, and set the value of the variable in
{{DfsConfigKeys}} based on this.
* {{DfsConfigKeys}}: the variables should be named
{{DFS_DATA_SOCKET_SEND_BUFFER_SIZE}},
{{DFS_DATA_SOCKET_SEND_BUFFER_SIZE_DEFAULT}}, right?
* {{DataStreamer#createSocketForPipeline}}: we should cache the send buffer
size in the DfsClientConf.
> DataNode sets SO_RCVBUF explicitly is disabling tcp auto-tuning
> ---------------------------------------------------------------
>
> Key: HDFS-8829
> URL: https://issues.apache.org/jira/browse/HDFS-8829
> Project: Hadoop HDFS
> Issue Type: Improvement
> Components: datanode
> Affects Versions: 2.3.0, 2.6.0
> Reporter: He Tianyi
> Assignee: He Tianyi
> Attachments: HDFS-8829.0001.patch, HDFS-8829.0002.patch
>
>
> {code:java}
> private void initDataXceiver(Configuration conf) throws IOException {
> // find free port or use privileged port provided
> TcpPeerServer tcpPeerServer;
> if (secureResources != null) {
> tcpPeerServer = new TcpPeerServer(secureResources);
> } else {
> tcpPeerServer = new TcpPeerServer(dnConf.socketWriteTimeout,
> DataNode.getStreamingAddr(conf));
> }
>
> tcpPeerServer.setReceiveBufferSize(HdfsConstants.DEFAULT_DATA_SOCKET_SIZE);
> {code}
> The last line sets SO_RCVBUF explicitly, thus disabling tcp auto-tuning on
> some system.
> Shall we make this behavior configurable?
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
