[ 
https://issues.apache.org/jira/browse/HDFS-13248?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16812464#comment-16812464
 ] 

Ayush Saxena commented on HDFS-13248:
-------------------------------------

Thanx [~hexiaoqiao] and [~elgoiri]

I guess if we use the {{IpcConnectionContext}} approach, we may even solve the 
problem of auditing? and I feel this approach if implemented carefully wouldn't 
have any compatibility issues too. Regarding security also I don't see as such 
any threat, shall dig in more to verify.

So my personal bet is on this!!!

Regarding client protocol changes, This is like in the present scenario only 
router shall be using it and it shall be an optional parameter, but a normal 
client shall also be able to use it. We can't stop that being used there so 
easily and even this whole shan't be achieved through single change whole read 
and write the elementary protocols are the target.
Not against it but, I feel we should make it a second option as of now.There is 
no doubt regarding the work-ability of the solution(We are pretty sure this 
will work) but would like to offer a chance if we tend to get a better and 
safer solution.

Rest, Whichever solution everyone feels good. I shall be happy helping out in 
the way in. 

> 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, 
> HDFS-13248.002.patch, HDFS-13248.003.patch, HDFS-13248.004.patch, 
> HDFS-13248.005.patch, HDFS-Router-Data-Locality.odt, RBF Data Locality 
> Design.pdf, 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: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to