[
https://issues.apache.org/jira/browse/HDFS-1142?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12871416#action_12871416
]
sam rash commented on HDFS-1142:
--------------------------------
Hi Konstantin,
Sorry, I am basing this mostly on the 0.20 code base. I do not have a test
case--this is purely analytical at this point and meant to be an argument for
keeping this fix in 0.20 w/append + sync (hadoop 0.20 + hdfs-142 + hdfs-200).
From reading the design doc on hdfs-165, hflush would also be violated by this
since the client would have received an ack for bytes in the real last block
that will never be available for read.
Can you comment on this scenario with respect to 0.20? if I get some time, I
can try to create a test case for this--it seems doable.
> Lease recovery doesn't reassign lease when triggered by append()
> ----------------------------------------------------------------
>
> Key: HDFS-1142
> URL: https://issues.apache.org/jira/browse/HDFS-1142
> Project: Hadoop HDFS
> Issue Type: Bug
> Components: name-node
> Affects Versions: 0.21.0
> Reporter: Todd Lipcon
> Assignee: Todd Lipcon
> Attachments: hdfs-1142.txt, hdfs-1142.txt
>
>
> If a soft lease has expired and another writer calls append(), it triggers
> lease recovery but doesn't reassign the lease to a new owner. Therefore, the
> old writer can continue to allocate new blocks, try to steal back the lease,
> etc. This is for the testRecoveryOnBlockBoundary case of HDFS-1139
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
