[
https://issues.apache.org/jira/browse/HDFS-577?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12748937#action_12748937
]
Allen Wittenauer commented on HDFS-577:
---------------------------------------
We had a situation where the data node could send packets out, but packets
coming in were dropped. So while jobs were unable to write blocks to that data
node, the name node thought it was perfectly healthy. There should be some
sort of 'double check' to make sure the data node is telling the truth about
its health status.
> Name node doesn't always properly recognize health of data node
> ---------------------------------------------------------------
>
> Key: HDFS-577
> URL: https://issues.apache.org/jira/browse/HDFS-577
> Project: Hadoop HDFS
> Issue Type: Bug
> Reporter: Allen Wittenauer
>
> The one-way communication (data node -> name node) for node health does not
> guarantee that the data node is actually healthy.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
