hadoop-hdfs-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Mingliang Liu (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HDFS-10306) SafeModeMonitor should not leave safe mode if name system is starting active service
Date Tue, 19 Apr 2016 00:35:25 GMT

    [ https://issues.apache.org/jira/browse/HDFS-10306?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15246844#comment-15246844
] 

Mingliang Liu commented on HDFS-10306:
--------------------------------------

Test failures are not related. We don't add new test because this is a follow-up of [HDFS-10192],
which added two unit test, and [HDFS-10284] refined them.

> SafeModeMonitor should not leave safe mode if name system is starting active service
> ------------------------------------------------------------------------------------
>
>                 Key: HDFS-10306
>                 URL: https://issues.apache.org/jira/browse/HDFS-10306
>             Project: Hadoop HDFS
>          Issue Type: Bug
>          Components: namenode
>            Reporter: Mingliang Liu
>            Assignee: Mingliang Liu
>         Attachments: HDFS-10306.000.patch
>
>
> This is a follow-up of [HDFS-10192].
> The {{BlockManagerSafeMode$SafeModeMonitor#canLeave()}} is not checking the {{namesystem#inTransitionToActive()}},
while it should. According to the fix of [HDFS-10192], we should add this check to prevent
the {{smmthread}} from calling {{leaveSafeMode()}} too early.



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

Mime
View raw message