[
https://issues.apache.org/jira/browse/HDFS-17280?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17809007#comment-17809007
]
ASF GitHub Bot commented on HDFS-17280:
---------------------------------------
hfutatzhanghb commented on PR #6336:
URL: https://github.com/apache/hadoop/pull/6336#issuecomment-1902486982
> > @hfutatzhanghb Thanks for your contribution! Sorry I didn't get this
proposal clearly. Would you mind to offer some more information about what
issue do you meet, and what this PR could do? Thanks again.
>
> @Hexiaoqiao Sir, so sorry for repsonsing too late. Let me describe this PR
in detail.
>
> ### The goal of this PR
> Since we have
[HDFS-16348](https://issues.apache.org/jira/browse/HDFS-16348), we can kick out
SLOW node in pipeline when writing data to pipeline. If we call
addDatanode2ExistingPipeline() method, it will trigger transfer block process.
>
> Think about below situation : we have a cluster with block size equals to
512MB, if we have already written 500MB and one datanode was kicked out from
pipeline, then add a new datanode to pipeline. It will transfer 500MB data to
the new chosen datanode. This is not efficient.
>
> So this PR is trying to alleviate this effect, if we have already written
over a half of block size, we can end this block in advance to avoid tranfering
data.
@Hexiaoqiao @zhangshuyan0 Sir, could you please take a look at this PR when
you are free? Thanks ahead.
> Pipeline recovery should better end block in advance when bytes acked greater
> than half of blocksize.
> -----------------------------------------------------------------------------------------------------
>
> Key: HDFS-17280
> URL: https://issues.apache.org/jira/browse/HDFS-17280
> Project: Hadoop HDFS
> Issue Type: Improvement
> Components: hdfs-client
> Reporter: farmmamba
> Assignee: farmmamba
> Priority: Major
> Labels: pull-request-available
>
--
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
