[
https://issues.apache.org/jira/browse/HDFS-10742?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15436400#comment-15436400
]
Chris Douglas commented on HDFS-10742:
--------------------------------------
bq. there is no other advantage. It just eliminates the finally block
Since Hadoop already uses {{Lock}} (including the capabilities omitted from
{{AutoCloseableLock}}), it's unlikely we'd standardize on this abstraction. It
adds a special case for a very common pattern to a subset of the code, for some
mild syntactic sugar.
* This should have a unit test.
* The {{local\*}} variables in {{check}} appear to be intended for {{volatile}}
variables, but the fields are not
* Please fix the findbugs warnings by either using {{ConcurrentMap}} as
designed or using non-threadsafe {{Map}} classes and the existing
synchronization (though using {{get}}/null checks, rather than {{containsKey}}).
* If this can't use slf4j, then consider using {{isWarnEnabled()}} to guard the
string formatting.and/or ignore all of {{check}}
* Consider passing the logging thresholds as parameters, rather than
hard-coding them in the class
> Measurement of lock held time in FsDatasetImpl
> ----------------------------------------------
>
> Key: HDFS-10742
> URL: https://issues.apache.org/jira/browse/HDFS-10742
> Project: Hadoop HDFS
> Issue Type: Improvement
> Components: datanode
> Affects Versions: 3.0.0-alpha2
> Reporter: Chen Liang
> Assignee: Chen Liang
> Attachments: HDFS-10742.001.patch, HDFS-10742.002.patch,
> HDFS-10742.003.patch, HDFS-10742.004.patch, HDFS-10742.005.patch,
> HDFS-10742.006.patch
>
>
> This JIRA proposes to measure the time the of lock of {{FsDatasetImpl}} is
> held by a thread. Doing so will allow us to measure lock statistics.
> This can be done by extending the {{AutoCloseableLock}} lock object in
> {{FsDatasetImpl}}. In the future we can also consider replacing the lock with
> a read-write lock.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
