You are viewing a plain text version of this content. The canonical link for it is here.

Posted to hdfs-issues@hadoop.apache.org by "Jing Zhao (JIRA)" <ji...@apache.org> on 2013/04/15 03:07:13 UTC

[jira] [Updated] (HDFS-4694) The SecondaryNameNode may change the inode id of root while doing checkpoint

     [ https://issues.apache.org/jira/browse/HDFS-4694?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Jing Zhao updated HDFS-4694:
----------------------------

    Attachment: HDFS-4694.001.patch

A simple patch that passes an inode id to FSDirectory#createRoot. Also update test cases in TestSecondaryNameNodeUpgrade to verify the root id.
                
> The SecondaryNameNode may change the inode id of root while doing checkpoint
> ----------------------------------------------------------------------------
>
>                 Key: HDFS-4694
>                 URL: https://issues.apache.org/jira/browse/HDFS-4694
>             Project: Hadoop HDFS
>          Issue Type: Sub-task
>          Components: namenode
>    Affects Versions: 3.0.0
>            Reporter: Jing Zhao
>            Assignee: Jing Zhao
>         Attachments: HDFS-4694.001.patch
>
>
> SNN will call FSNamesystem#clear while merging the fsimage to its local FSNamesystem, and the FSDirectory will be reset in this process. However, in FSDirectory#reset, a new inode id will be allocated and assigned to the new root.
> To fix the problem, we can either pass in an inode id in FSDirectory#createRoot, or update the root id in FSImageFormat#updateRootAttr.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira