[ https://issues.apache.org/jira/browse/HDFS-15661?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17265745#comment-17265745 ]
Wei-Chiu Chuang commented on HDFS-15661: ---------------------------------------- I'm not familiar with the dead node detector -- it's a new feature. The patch itself looks good. If the goal is to ensure that closing one dfs client doesn't prematurely terminate the deadnode detector thread, maybe a simpler approach is to use the CloseableReferenceCount class (in hadoop-common) to keep the reference count. It seems like a good idea to keep a single dead node detector object. If I understand it correctly, the original design works in federated cluster. After this patch, the dead node information won't be shared between difference block pools. > The DeadNodeDetector shouldn't be shared by different DFSClients. > ----------------------------------------------------------------- > > Key: HDFS-15661 > URL: https://issues.apache.org/jira/browse/HDFS-15661 > Project: Hadoop HDFS > Issue Type: Bug > Reporter: Jinglun > Assignee: Jinglun > Priority: Major > Attachments: HDFS-15661.001.patch > > > Currently the DeadNodeDetector is a member of ClientContext. That means it is > shared by many different DFSClients. When one DFSClient.close() is invoked, > the DeadNodeDetecotor thread would be interrupted and impact other DFSClients. > From the original design of HDFS-13571 we could see the DeadNodeDetector is > supposed to share dead nodes of many input streams from the same client. > We should move the DeadNodeDetector as a member of DFSClient instead of > ClientContext. -- This message was sent by Atlassian Jira (v8.3.4#803005) --------------------------------------------------------------------- To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org