[
https://issues.apache.org/jira/browse/HDFS-4141?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13489819#comment-13489819
]
Aaron T. Myers commented on HDFS-4141:
--------------------------------------
Patch looks pretty good to me, and it definitely helps to clear up what exactly
a diff is from the latest design doc posted by Suresh. My only suggestion for
the code is that both apply2Previous/apply2Current and
accessPrevious/accessCurrent are pretty similar to each other, and you could
probably reduce some code duplication there.
The current code doesn't actually create a Diff at all, but I assume that will
be added in another JIRA. I have just one question for you: how will multiple
snapshots of the same directory be handled? My guess is that a single
INodeDirectoryWithSnapshot will itself contain a list of Diffs, one per
snapshot where files are changed in the directory. Is the plan for those diffs
to be relative to each other?
> Directory diff in snapshots
> ---------------------------
>
> Key: HDFS-4141
> URL: https://issues.apache.org/jira/browse/HDFS-4141
> Project: Hadoop HDFS
> Issue Type: Sub-task
> Components: name-node
> Reporter: Tsz Wo (Nicholas), SZE
> Assignee: Tsz Wo (Nicholas), SZE
> Attachments: h4141_20121101.patch, h4141_20121102b.patch,
> h4141_20121102.patch
>
>
> With snapshots, the same directory could have different states, e.g.
> different sets of children, in different snapshots. We implement a directory
> diff algorithm in this JIRA.
--
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
