[
https://issues.apache.org/jira/browse/HDFS-6214?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13971581#comment-13971581
]
Kihwal Lee edited comment on HDFS-6214 at 4/16/14 3:59 PM:
-----------------------------------------------------------
Ok, if there is no performance degradation, I am fine with calling flush() for
the non-chunked case.
>So, if "io.file.buffer.size" is small enough, like 4K (the default), it may be
>overall slower, but there will be no difference for files > 2GB.
I take it back. It just means the client (block reader) will read 4K at a time
from datanodes. The 24KB response buffer will be filled in and flushed, so this
issue is still visible with the default buffer size.
+1 for the patch.
was (Author: kihwal):
Ok, if there was no performance degradation, I am fine with calling flush() for
the non-chunked case.
>So, if "io.file.buffer.size" is small enough, like 4K (the default), it may be
>overall slower, but there will be no difference for files > 2GB.
I take it back. It just means the client (block reader) will read 4K at a time
from datanodes. The 24KB response buffer will be filled in and flushed, so this
issue is still visible with the default buffer size.
+1 for the patch.
> Webhdfs has poor throughput for files >2GB
> ------------------------------------------
>
> Key: HDFS-6214
> URL: https://issues.apache.org/jira/browse/HDFS-6214
> Project: Hadoop HDFS
> Issue Type: Bug
> Components: webhdfs
> Affects Versions: 2.0.0-alpha, 3.0.0
> Reporter: Daryn Sharp
> Assignee: Daryn Sharp
> Attachments: HDFS-6214.patch
>
>
> For the DN's open call, jetty returns a Content-Length header for files <2GB,
> and uses chunking for files >2GB. A "bug" in jetty's buffer handling results
> in a ~8X reduction in throughput.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
