hadoop-hdfs-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Todd Lipcon (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HDFS-4111) Support snapshot of subtrees
Date Fri, 26 Oct 2012 14:33:12 GMT

    [ https://issues.apache.org/jira/browse/HDFS-4111?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13484954#comment-13484954
] 

Todd Lipcon commented on HDFS-4111:
-----------------------------------

This seems to implement the O(n) algorithm for snapshot creation, rather than the lazy method
discussed in the other JIRAs. I don't think it's feasible to use O(n) memory at snapshot creation
time, per my comments elsewhere. I had understood from that other discussion that you were
planning to implement O(1) usage at creation time, with up to O(m) usage after m modifications?

I can't imagine this working on a large subtree -- eg a few million files like a hive warehouse
dir.

Maybe we can reach agreement on the algorithm design before moving ahead with too much code?
                
> Support snapshot of subtrees
> ----------------------------
>
>                 Key: HDFS-4111
>                 URL: https://issues.apache.org/jira/browse/HDFS-4111
>             Project: Hadoop HDFS
>          Issue Type: Sub-task
>          Components: name-node
>            Reporter: Tsz Wo (Nicholas), SZE
>            Assignee: Tsz Wo (Nicholas), SZE
>              Labels: needs-test
>         Attachments: h4111_20121025.patch
>
>
> HDFS-4076 add the ability to take snapshot of single files.  In this JIRA, we add the
ability to take snapshot of subtree trees in the namesystem.

--
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

Mime
View raw message