[
https://issues.apache.org/jira/browse/HDFS-16922?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17689503#comment-17689503
]
ASF GitHub Bot commented on HDFS-16922:
---------------------------------------
Hexiaoqiao commented on PR #5398:
URL: https://github.com/apache/hadoop/pull/5398#issuecomment-1432498891
Thanks to involve me here. It is interesting issue. I am confused about some
points of the description.
> dn3 is writting the blk_12345_002 , but dn2 is blocked by recoverClose
method and does not send ack to client.
is this another injects or related this write flow?
> dn3 writes blk_12345_003 successfully.
> dn3 writes blk_12345_002 successfully and notifyNamenodeReceivedBlock.
Here dn3 writes the same block replica twice, is it expected?
Sorry didn't dig deeply this logic, will trace it for a while.
@hfutatzhanghb Thanks again for your report and offer the solution.
> The logic of IncrementalBlockReportManager#addRDBI method may cause missing
> blocks when cluster is busy.
> --------------------------------------------------------------------------------------------------------
>
> Key: HDFS-16922
> URL: https://issues.apache.org/jira/browse/HDFS-16922
> Project: Hadoop HDFS
> Issue Type: Bug
> Components: datanode
> Reporter: ZhangHB
> Priority: Major
> Labels: pull-request-available
>
> The current logic of IncrementalBlockReportManager# addRDBI method could lead
> to the missing blocks when datanodes in pipeline are I/O busy.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org
