[
https://issues.apache.org/jira/browse/HDFS-17707?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17911002#comment-17911002
]
ASF GitHub Bot commented on HDFS-17707:
---------------------------------------
hfutatzhanghb opened a new pull request, #7276:
URL: https://github.com/apache/hadoop/pull/7276
### Description of PR
Cause is race condition when multiple instances of
FileIoProvider#mkdirsWithExistsCheck are called for the same directory on
different DIR locks.
There can be two ways to solve this problem:
1. use DIR lock align with datanode two-level layout, like subdir0/subdir0
2. modify the logic of mkdirsWithExistsCheck like below:
> TestStateStoreFileSystem fails after HDFS-17496
> -----------------------------------------------
>
> Key: HDFS-17707
> URL: https://issues.apache.org/jira/browse/HDFS-17707
> Project: Hadoop HDFS
> Issue Type: Bug
> Reporter: Felix N
> Assignee: farmmamba
> Priority: Minor
>
> Cause is race condition when multiple instances of
> FileIoProvider#mkdirsWithExistsCheck are called for the same directory on
> different DIR locks.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
