[
https://issues.apache.org/jira/browse/HDFS-13248?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16391560#comment-16391560
]
Íñigo Goiri commented on HDFS-13248:
------------------------------------
I've tried to follow the logic in {{NamenodeRpcServer}} and it doesn't look
like it uses the RPC address but the {{clientName}} parameter.
{{NamenodeRpcServer#addBlock}} goes to {{FSNamesystem#getAdditionalBlock}}
which calls {{FSDirWriteFileOp#chooseTargetForNewBlock}}.
Inside {{chooseTargetForNewBlock}}, we call
{{DatanodeManager#getDatanodeByHost()}} with a parameter called
{{clientMachine}} which seems to come from
{{FSDirWriteFileOp#validateAddBlock()}}:
{code}
clientMachine =
pendingFile.getFileUnderConstructionFeature().getClientMachine();
{code}
I'm not 100% sure but I think this value is the parameter {{clientName}} which
we initially pass to {{RouterRpcServer#addBlock}}.
> 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
> Priority: Major
>
> 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 namennode 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]
