[
https://issues.apache.org/jira/browse/HDFS-11890?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16027124#comment-16027124
]
Surendra Singh Lilhore commented on HDFS-11890:
-----------------------------------------------
Thanks [~umamaheswararao] for review..
bq. From next iteration, you will not find any moverTaskFeatures as it was
already cleared. So, this newly added trackID will be leaked right?
yes, you are right. It is race condition..
I feel we can add double check in {{if (blocksMoving.isEmpty()) {}} some thing
like this...
{code}
if (blocksMoving.isEmpty() || moverTaskFutures.get(trackId) == null) {
.........
.........
movementResults.remove(trackId);
}
{code}
> Handle NPE in BlockRecoveryWorker when DN is getting shoutdown.
> ---------------------------------------------------------------
>
> Key: HDFS-11890
> URL: https://issues.apache.org/jira/browse/HDFS-11890
> Project: Hadoop HDFS
> Issue Type: Bug
> Components: datanode
> Affects Versions: 2.7.2
> Reporter: Surendra Singh Lilhore
> Assignee: Surendra Singh Lilhore
> Priority: Minor
> Attachments: HDFS-11890-001.patch
>
>
> {code}
> Exception in thread
> "org.apache.hadoop.hdfs.server.datanode.BlockRecoveryWorker$1@1c03e6ae"
> java.lang.NullPointerException
> at
> org.apache.hadoop.hdfs.server.datanode.BlockRecoveryWorker$RecoveryTaskContiguous.recover(BlockRecoveryWorker.java:131)
> at
> org.apache.hadoop.hdfs.server.datanode.BlockRecoveryWorker$1.run(BlockRecoveryWorker.java:596)
> at java.lang.Thread.run(Thread.java:748)
> {code}
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
---------------------------------------------------------------------
To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org
