[
https://issues.apache.org/jira/browse/HDFS-4499?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13582815#comment-13582815
]
Tsz Wo (Nicholas), SZE commented on HDFS-4499:
----------------------------------------------
The javadoc is excellent. Some comments:
- In INodeDirectoryWithSnapshot.cleanSubtree(), remove "this instanceof
INodeDirectoryWithSnapshot" since it is always true.
- INodeDirectoryWithSnapshot.isPrior(), may be remove and check the condition
in cleanSubtree? Also, only need to check "newPrior != null" for newPrior.
- INodeDirectoryWithSnapshot.ChildrenDiff.destroyCreatedList() is not needed
since the nodes in created are the current children, which will be destroyed.
- getLastBeforeSnapshot: may rename it to getPrior and also compute the index
before calling diffs.get(..).
- INodeFileWithSnapshot.cleanSubtree: collectBlocksAndClear should be called at
the end.
- INodeFileWithSnapshot.destroyAndCollectBlocks add a clear method to
AbstractINodeDiffList. Also,
INodeDirectoryWithSnapshot.destroyAndCollectBlocks should also call it.
> Fix file/directory/snapshot deletion for file diff
> --------------------------------------------------
>
> Key: HDFS-4499
> URL: https://issues.apache.org/jira/browse/HDFS-4499
> Project: Hadoop HDFS
> Issue Type: Sub-task
> Components: datanode, namenode
> Reporter: Jing Zhao
> Assignee: Jing Zhao
> Attachments: HDFS-4499.000.patch, HDFS-4499.001.patch,
> HDFS-4499.002.patch
>
>
> After recording file diffs, the original file/directory/snapshot deletion
> process needs to updated. This jira revisits and fixes the whole deletion
> process. New unit tests will also be added to cover different cases.
--
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
