[
https://issues.apache.org/jira/browse/HDFS-7005?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14284645#comment-14284645
]
Colin Patrick McCabe commented on HDFS-7005:
--------------------------------------------
[~zsl2007], it appears that the DataNode is setting both a write and a read
timeout on its sockets, but the DFSClient is only setting a read timeout. If
you want to file another JIRA to add a write timeout to DFSClient sockets, that
might be a good idea.
> DFS input streams do not timeout
> --------------------------------
>
> Key: HDFS-7005
> URL: https://issues.apache.org/jira/browse/HDFS-7005
> Project: Hadoop HDFS
> Issue Type: Bug
> Components: hdfs-client
> Affects Versions: 3.0.0, 2.5.0
> Reporter: Daryn Sharp
> Assignee: Daryn Sharp
> Priority: Critical
> Fix For: 2.6.0
>
> Attachments: HDFS-7005.patch
>
>
> Input streams lost their timeout. The problem appears to be
> {{DFSClient#newConnectedPeer}} does not set the read timeout. During a
> temporary network interruption the server will close the socket, unbeknownst
> to the client host, which blocks on a read forever.
> The results are dire. Services such as the RM, JHS, NMs, oozie servers, etc
> all need to be restarted to recover - unless you want to wait many hours for
> the tcp stack keepalive to detect the broken socket.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
