[ 
https://issues.apache.org/jira/browse/HDFS-8460?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14557115#comment-14557115
 ] 

Walter Su commented on HDFS-8460:
---------------------------------

*Causes*
{code:title=DFSTestUtil.java}
1910   public static Block addStripedBlockToFile(List<DataNode> dataNodes,
       ...
1926       DatanodeStorage storage = new 
DatanodeStorage(UUID.randomUUID().toString());
{code}
If DN itself sends a blockreport, the storage with randomUUID will be 
considered zombie and be removed.

*reproduce problem*
{code}
     DFSTestUtil.createStripedFile(cluster, filePath, null, numBlocks,
         NUM_STRIPE_PER_BLOCK, false);
+    cluster.triggerHeartbeats();
     LocatedBlocks lbs = 
fs.getClient().namenode.getBlockLocations(filePath.toString(), 0, fileSize);
{code}


> Erasure Coding: stateful read result doesn't match data occasionally
> --------------------------------------------------------------------
>
>                 Key: HDFS-8460
>                 URL: https://issues.apache.org/jira/browse/HDFS-8460
>             Project: Hadoop HDFS
>          Issue Type: Sub-task
>            Reporter: Yi Liu
>            Assignee: Walter Su
>         Attachments: HDFS-8460-HDFS-7285.001.patch
>
>
> I found this issue in TestDFSStripedInputStream, {{testStatefulRead}} failed 
> occasionally shows that read result doesn't match data written.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Reply via email to