[ https://issues.apache.org/jira/browse/HDFS-13248?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16627596#comment-16627596 ]
Íñigo Goiri commented on HDFS-13248: ------------------------------------ Thanks [~ferhui], the approach in [^HDFS-13248.002.patch] is not very intrusive so it might be a good starting point. On the other hand, it changes the semantics slightly so I'd like to get feedback from others. I think we still need to tune {{NamenodeRpcServer#getClientName()}} but we can leave it for another JIRA. How does the rack locality and so on work in this case? Would it try to place it in the rack itself? Regarding the patch itself: * I think we should make a function to add to the {{favoredNodes}} and add a few unit tests just for it and a full unit test that actually involves a Router and a Namenode. * We should do {{new ArrayList<String>(favoredNodes.length + 1);}}. > 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, 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