[
https://issues.apache.org/jira/browse/HDFS-13248?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16417737#comment-16417737
]
Íñigo Goiri commented on HDFS-13248:
------------------------------------
[~ajayydv], I agree.
I tried to go with the token approach but the tokens weren't sent at all.
I think we need some more basis like in HDFS-13358 for that.
If you have background on that, feel free to give it a try.
Thanks [~linyiqun], we can move the read/write logic into the NN side.
Let's see if we can do it with tokens though.
for the test, I tried to do something with the MiniDFSCluster but all the
addresses are the same.
I tried to force addresses like 127.0.0.2 and 127.0.0.3 for the Router and the
Client but it requires more wiring to set the origin IP in the client.
There is some logic to set it for the DNs but not for the user.
> RBF: Namenode need to choose block location for the client
> ----------------------------------------------------------
>
> Key: HDFS-13248
> URL: https://issues.apache.org/jira/browse/HDFS-13248
> Project: Hadoop HDFS
> Issue Type: Sub-task
> Reporter: Weiwei Wu
> Assignee: Íñigo Goiri
> Priority: Major
> Attachments: HDFS-13248.000.patch, HDFS-13248.001.patch,
> clientMachine-call-path.jpeg, debug-info-1.jpeg, debug-info-2.jpeg
>
>
> When execute a put operation via router, the NameNode will choose block
> location for the router, not for the real client. This will affect the file's
> locality.
> I think on both NameNode and Router, we should add a new addBlock method, or
> add a parameter for the current addBlock method, to pass the real client
> information.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org
