[
https://issues.apache.org/jira/browse/HDFS-8827?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Jing Zhao updated HDFS-8827:
----------------------------
Attachment: HDFS-8827.2.patch
Thanks [~tfukudom] for working on this! I agree with Walter that for the unit
test, the NPE is more like caused by the testing code. The .2 patch writes the
real data instead of using {{createStripedFile}}, but it exposes some other
bugs in the current writing code path so I have to change the block size from 1
cell to 4 cells. In the meanwhile the test passed with the change. So the NPE
you saw in the system test may be caused by some other issue. I will dig
further into that.
> Erasure Coding: When namenode processes over replicated striped block, NPE
> will be occur in ReplicationMonitor
> --------------------------------------------------------------------------------------------------------------
>
> Key: HDFS-8827
> URL: https://issues.apache.org/jira/browse/HDFS-8827
> Project: Hadoop HDFS
> Issue Type: Sub-task
> Reporter: Takuya Fukudome
> Assignee: Takuya Fukudome
> Attachments: HDFS-8827.1.patch, HDFS-8827.2.patch,
> processing-over-replica-npe.log
>
>
> In our test cluster, when namenode processed over replicated striped blocks,
> null pointer exception(NPE) occurred. This happened under below situation: 1)
> some datanodes shutdown. 2) namenode recovers block group which lost internal
> blocks. 3) restart the stopped datanodes. 4) namenode processes over
> replicated striped blocks. 5) NPE occurs
> I think BlockPlacementPolicyDefault#chooseReplicaToDelete will return null in
> this situation which causes this NPE problem.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
