[ https://issues.apache.org/jira/browse/HDFS-9040?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14875837#comment-14875837 ]
Walter Su commented on HDFS-9040: --------------------------------- bq. I think the key point here is to bump the GS, which is necessary to identify stale/corrupted internal blocks Client is a good child. He bumps GS every time there's a failure. One day, he got killed before tried to bump GS the last time. he also didn't get the chance to commit the block. Oops, all the left healthy internal replicas have the same GS. We don't know who is the bad replica he tried to report before he got killed. What's the different from never report once? > Erasure coding: Refactor DFSStripedOutputStream (Move Namenode RPC Requests > to Coordinator) > ------------------------------------------------------------------------------------------- > > Key: HDFS-9040 > URL: https://issues.apache.org/jira/browse/HDFS-9040 > Project: Hadoop HDFS > Issue Type: Sub-task > Reporter: Walter Su > Attachments: HDFS-9040-HDFS-7285.002.patch, > HDFS-9040-HDFS-7285.003.patch, HDFS-9040.00.patch, HDFS-9040.001.wip.patch, > HDFS-9040.02.bgstreamer.patch > > > The general idea is to simplify error handling logic. > Proposal 1: > A BlockGroupDataStreamer to communicate with NN to allocate/update block, and > StripedDataStreamer s only have to stream blocks to DNs. > Proposal 2: > See below the > [comment|https://issues.apache.org/jira/browse/HDFS-9040?focusedCommentId=14741388&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-14741388] > from [~jingzhao]. -- This message was sent by Atlassian JIRA (v6.3.4#6332)