[
https://issues.apache.org/jira/browse/HDFS-73?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13083836#comment-13083836
]
Eli Collins commented on HDFS-73:
---------------------------------
Hi Uma,
Latest patch looks good. Wrt testing how about adding the following assert to
all non-final sockets in the file? The append tests will fail this currently
and it checks that we don't introduce this bug again.
{noformat}
try {
+ assert null == s : "Previous socket unclosed";
s = createSocketForPipeline(nodes[0], nodes.length, dfsClient);
{noformat}
Nit: please make the formatting (spacing) in closeStream consistent and
generate a new patch that will apply with test-patch.
> DFSOutputStream does not close all the sockets
> ----------------------------------------------
>
> Key: HDFS-73
> URL: https://issues.apache.org/jira/browse/HDFS-73
> Project: Hadoop HDFS
> Issue Type: Bug
> Reporter: Raghu Angadi
> Assignee: Raghu Angadi
> Priority: Blocker
> Fix For: 0.23.0
>
> Attachments: HADOOP-3071.patch, HDFS-73_0.23.patch
>
>
> When DFSOutputStream writes to multiple blocks, it closes only the socket
> opened for the last block. When it is done with writing to one block it
> should close the socket.
> I noticed this when I was fixing HADOOP-3067. After fixing HADOOP-3067, there
> were still a lot of sockets open (but not enough to fail the tests). These
> sockets were used to write to blocks.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
