[
https://issues.apache.org/jira/browse/HDFS-17182?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17765494#comment-17765494
]
ASF GitHub Bot commented on HDFS-17182:
---------------------------------------
Hexiaoqiao commented on code in PR #6029:
URL: https://github.com/apache/hadoop/pull/6029#discussion_r1326940963
##########
hadoop-hdfs-project/hadoop-hdfs/src/main/java/org/apache/hadoop/hdfs/server/datanode/DataSetLockManager.java:
##########
@@ -256,7 +256,7 @@ private synchronized void putThreadName(String thread) {
threadCountMap.putIfAbsent(thread, new TrackLog(thread));
}
- public void lockLeakCheck() {
+ public synchronized void lockLeakCheck() {
Review Comment:
@LiuGuH Thanks for your contribution here. IIRC, this is only invoked when
DataNode instance shutdown, and no more concurrent issue generally. So do you
meet some issues here? Theoretically, I think it is better to add
`synchronized`. Thanks again.
> DataSetLockManager.lockLeakCheck() is not thread-safe.
> -------------------------------------------------------
>
> Key: HDFS-17182
> URL: https://issues.apache.org/jira/browse/HDFS-17182
> Project: Hadoop HDFS
> Issue Type: Bug
> Reporter: liuguanghua
> Assignee: liuguanghua
> Priority: Minor
> Labels: pull-request-available
>
> threadCountMap is not thread-safe. Other functions add protected by
> synchronized expect lockLeakCheck(). Add synchronized on function
> lockLeakCheck().
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
