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-736) commitBlockSynchronization() should directly update block GS and length.
Date Fri, 30 Oct 2009 00:34:59 GMT

    [ https://issues.apache.org/jira/browse/HDFS-736?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12771721#action_12771721
] 

Konstantin Shvachko commented on HDFS-736:
------------------------------------------

No new tests required, modified functionality is checked by the existing tests TestLeaseRecovery
and TestLeaseRecovery2.

> commitBlockSynchronization() should directly update block GS and length.
> ------------------------------------------------------------------------
>
>                 Key: HDFS-736
>                 URL: https://issues.apache.org/jira/browse/HDFS-736
>             Project: Hadoop HDFS
>          Issue Type: Improvement
>          Components: name-node
>    Affects Versions: 0.21.0
>            Reporter: Konstantin Shvachko
>            Assignee: Konstantin Shvachko
>             Fix For: 0.21.0
>
>         Attachments: commitBlockSync.patch
>
>
> {{FSNamesystem.commitBlockSynchronization()}} updates block's generation stamp and length
by first removing the old block instance from blocksMap and then inserting the new instance
back. This was necessary when GS was a part of the block key. After HDFS-512 the GS along
with the block length can be updated directly in the blocksMap entry.

-- 
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