hadoop-common-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "lohit vijayarenu (JIRA)" <j...@apache.org>
Subject [jira] Updated: (HADOOP-1979) fsck on namenode without datanodes takes too much time
Date Wed, 16 Apr 2008 22:37:22 GMT

     [ https://issues.apache.org/jira/browse/HADOOP-1979?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel

lohit vijayarenu updated HADOOP-1979:

    Attachment: HADOOP-1979.patch

This patch changes the way lines are flushed at server end. It tries to invoke flush for every
100 files. And at the client, wrapped InputStreamReader in BufferedReader and processing output
in lines. I tested this with a big image and I could see that this now reaches the time of

> fsck on namenode without datanodes takes too much time
> ------------------------------------------------------
>                 Key: HADOOP-1979
>                 URL: https://issues.apache.org/jira/browse/HADOOP-1979
>             Project: Hadoop Core
>          Issue Type: Bug
>          Components: dfs
>            Reporter: Koji Noguchi
>            Assignee: lohit vijayarenu
>            Priority: Minor
>         Attachments: HADOOP-1979.patch
> Once in a while , we want to debug by bringing up the namenode without any datanodes.
 dfs -ls/-lsr works, but fsck / takes too much time. (at least 10 times more)  Most likely
from HADOOP-943 that I requested.  If this is the case, it would be nice if we can have an
option for disabling it for this special case.

This message is automatically generated by JIRA.
You can reply to this email to add a comment to the issue online.

View raw message