hadoop-hdfs-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Konstantin Shvachko (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HDFS-11087) NamenodeFsck should check if the output writer is still writable.
Date Tue, 01 Nov 2016 23:35:58 GMT

    [ https://issues.apache.org/jira/browse/HDFS-11087?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15627060#comment-15627060
] 

Konstantin Shvachko commented on HDFS-11087:
--------------------------------------------

The problem is that {{PrintWriter.ensureOpen()}} does not really check the status of the underlying
Writer to the HttpConnection, which has already been closed.
This is probably Jetty 6 problem. On trunk with upgraded Jetty I see that server side fsck
fails after client is interrupted. So this is applicable to all v.2 branches only.
I propose to fix this by calling {{PrintWriter.checkError()}} before each round of printing
in {{NamenodeFsck.check()}}.


> NamenodeFsck should check if the output writer is still writable.
> -----------------------------------------------------------------
>
>                 Key: HDFS-11087
>                 URL: https://issues.apache.org/jira/browse/HDFS-11087
>             Project: Hadoop HDFS
>          Issue Type: Bug
>          Components: namenode
>    Affects Versions: 2.6.5
>            Reporter: Konstantin Shvachko
>
> {{NamenodeFsck}} keeps running even after the client was interrupted. So if you start
{{fsck /}} on a large namespace and kill the client, the NameNode will keep traversing the
tree for hours although there is nobody to receive the result. 



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

---------------------------------------------------------------------
To unsubscribe, e-mail: hdfs-issues-unsubscribe@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-help@hadoop.apache.org


Mime
View raw message