[
https://issues.apache.org/jira/browse/HDFS-9422?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15009046#comment-15009046
]
stack commented on HDFS-9422:
-----------------------------
Patch just undoes the synchronize on dataset. This dataset synchronize in scan
has been in place for years.
As is, there are 'holes' given we let go of the dataset lock after we finish
scan and before we go into the checkAndUpdate. The synchronize on dataset in
scan seems gratuitous. The synchronize in checkAndUpdate seems sufficient
safety doing reconcile of memory and disk state.
Patch looks good to me.
> Unnecessary FsDatasetImpl locking in DirectoryScanner cause periodic datanode
> pauses
> ------------------------------------------------------------------------------------
>
> Key: HDFS-9422
> URL: https://issues.apache.org/jira/browse/HDFS-9422
> Project: Hadoop HDFS
> Issue Type: Bug
> Reporter: He Liangliang
> Assignee: He Liangliang
> Attachments: HDFS-9422.patch
>
>
> In DirectoryScanner, the scan call hold the global dataset lock for quite
> long time (typically > 5 secs for 500k blocks). This will stuck the client
> which need acquire this lock. For applications like HBase, this will affect
> the latency. In fact, this lock is unnecessary.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
