[
https://issues.apache.org/jira/browse/HDFS-6948?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14125728#comment-14125728
]
Kihwal Lee commented on HDFS-6948:
----------------------------------
We may not need this change in {{createRbw()}}, which is for new block
creation. We see this kind of conflict after block creation. A copy is
attempted as part of pipeline recovery, replication after closing or balancing.
So there shouldn't be any conflict during initial creation of a block. I think
we only need to clean-up in {{createTemporary()}}.
What happens if the thread writing using older gen stamp is still running? We
can kill the thread as it is done in {{recoverRbw()}}.
> DN rejects blocks if it has older UC block
> ------------------------------------------
>
> Key: HDFS-6948
> URL: https://issues.apache.org/jira/browse/HDFS-6948
> Project: Hadoop HDFS
> Issue Type: Bug
> Affects Versions: 2.0.0-alpha, 3.0.0
> Reporter: Daryn Sharp
> Assignee: Eric Payne
> Attachments: HDFS-6948.201409052147.txt
>
>
> DNs appear to always reject blocks, even with newer genstamps, if it already
> has a UC copy in its tmp dir.
> {noformat}ReplicaAlreadyExistsException: Block
> XXX already
> exists in state TEMPORARY and thus cannot be created{noformat}
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
