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-645) Namenode does not leave safe mode even if all blocks are available
Date Thu, 29 Oct 2009 00:50:59 GMT

    [ https://issues.apache.org/jira/browse/HDFS-645?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12771224#action_12771224
] 

Konstantin Shvachko commented on HDFS-645:
------------------------------------------

Zheng, could you please explain what your patch does and what it fixes?
Did you mean to reassign this to yourself?

> Namenode does not leave safe mode even if all blocks are available
> ------------------------------------------------------------------
>
>                 Key: HDFS-645
>                 URL: https://issues.apache.org/jira/browse/HDFS-645
>             Project: Hadoop HDFS
>          Issue Type: Bug
>          Components: name-node
>    Affects Versions: 0.20.1, 0.21.0, 0.22.0
>            Reporter: Qi Liu
>            Assignee: Qi Liu
>             Fix For: 0.20.2, 0.21.0, 0.22.0
>
>         Attachments: HDFS-645-0.18.3.patch, HDFS-645-0.20.1.patch, HDFS-645.patch, HDFS-645.patch
>
>
> When the safe mode threshold is set to 1.0f, sometimes the name node does not leave safe
mode even if all blocks are available. This is caused by floating point number division and
comparison issues.

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


Mime
View raw message