[
https://issues.apache.org/jira/browse/HADOOP-1707?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
dhruba borthakur updated HADOOP-1707:
-------------------------------------
Attachment: clientDiskBuffer15.patch
Found a race condition that was causing the client to close the connection
before the datanodes had a chance to process the end-of-packet. This caused the
datanode to treat it as an error condition, thus causing the client to do error
recovery and re-send the outstanding packets to the remaining good datanodes.
This was causing performance regression.
> Remove the DFS Client disk-based cache
> --------------------------------------
>
> Key: HADOOP-1707
> URL: https://issues.apache.org/jira/browse/HADOOP-1707
> Project: Hadoop
> Issue Type: Improvement
> Components: dfs
> Reporter: dhruba borthakur
> Assignee: dhruba borthakur
> Fix For: 0.16.0
>
> Attachments: clientDiskBuffer.patch, clientDiskBuffer10.patch,
> clientDiskBuffer11.patch, clientDiskBuffer12.patch, clientDiskBuffer14.patch,
> clientDiskBuffer15.patch, clientDiskBuffer2.patch, clientDiskBuffer6.patch,
> clientDiskBuffer7.patch, clientDiskBuffer8.patch, clientDiskBuffer9.patch,
> DataTransferProtocol.doc, DataTransferProtocol.html
>
>
> The DFS client currently uses a staging file on local disk to cache all
> user-writes to a file. When the staging file accumulates 1 block worth of
> data, its contents are flushed to a HDFS datanode. These operations occur
> sequentially.
> A simple optimization of allowing the user to write to another staging file
> while simultaneously uploading the contents of the first staging file to HDFS
> will improve file-upload performance.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
