[
https://issues.apache.org/jira/browse/HDFS-8959?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Rakesh R updated HDFS-8959:
---------------------------
Attachment: HDFS-8959-02.patch
> Provide an iterator-based API for listing all the snapshottable directories
> ---------------------------------------------------------------------------
>
> Key: HDFS-8959
> URL: https://issues.apache.org/jira/browse/HDFS-8959
> Project: Hadoop HDFS
> Issue Type: Improvement
> Components: hdfs-client
> Reporter: Rakesh R
> Assignee: Rakesh R
> Attachments: HDFS-8959-00.patch, HDFS-8959-01.patch,
> HDFS-8959-02.patch
>
>
> Presently {{DistributedFileSystem#getSnapshottableDirListing()}} is sending
> all the {{SnapshottableDirectoryStatus[]}} array to the clients. Now the
> client should have enough space to hold it in memory. There could be chance
> that the client JVMs running out of memory because of this. Also, some time
> back there was a
> [comment|https://issues.apache.org/jira/browse/HDFS-8643?focusedCommentId=14658800&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-14658800]
> about RPC packet limitation and a large number of snapshot list can again
> cause issues.
> I believe iterator based {{DistributedFileSystem#listSnapshottableDirs()}}
> API would be a good addition!
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)