[
https://issues.apache.org/jira/browse/HDFS-7729?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14303827#comment-14303827
]
Zhe Zhang commented on HDFS-7729:
---------------------------------
bq. There may exist some problems about client and namenode collaboration, so
this unit test tries to find and fix these problems.
Once you allocate enough DNs to the cluster, like {{cluster = new
MiniDFSCluster.Builder(conf).numDataNodes(5).build();}}, NN will return a full
block group. I quickly tried that but {{TestDFSOutputStreamStripingLayout}}
still failed. Could you take a look?
The basic logic of {{addBlock}} looks correct. Please also remember to set the
storage policy of the test file, like
{{cluster.getFileSystem().setStoragePolicy(new Path("/"),
HdfsConstants.EC_STORAGE_POLICY_NAME);}}
{{convertGroupToBlocks}} needs to take care of calculating block IDs from the
group ID. Put simple, it should use {{group_ID + i}} to name the ith block in
the group.
Let's fix the above so the test runs. I'll hold on to more detailed reviews
until then.
> Add logic to DFSOutputStream to support writing a file in striping layout
> --------------------------------------------------------------------------
>
> Key: HDFS-7729
> URL: https://issues.apache.org/jira/browse/HDFS-7729
> Project: Hadoop HDFS
> Issue Type: Sub-task
> Reporter: Li Bo
> Assignee: Li Bo
> Attachments: Codec-tmp.patch, HDFS-7729-001.patch, HDFS-7729-002.patch
>
>
> If client wants to directly write a file striping layout, we need to add some
> logic to DFSOutputStream. DFSOutputStream needs multiple DataStreamers to
> write each cell of a stripe to a remote datanode.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
