[ https://issues.apache.org/jira/browse/HDFS-15768?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17261159#comment-17261159 ]
Janus Chow commented on HDFS-15768: ----------------------------------- I think you are raising this ticket because the performance is still not good even IBR interval set to 3s (default heartbeat interval), maybe the performance won't change too much even set to a bigger value but with a worse block acknowledgment. There are other solutions to mitigate the impact of NameNode lock, it's good to have a try. > Make the incremental blockreport interval can be bigger than heartbeat > interval > ------------------------------------------------------------------------------- > > Key: HDFS-15768 > URL: https://issues.apache.org/jira/browse/HDFS-15768 > Project: Hadoop HDFS > Issue Type: Improvement > Reporter: Yang Yun > Assignee: Yang Yun > Priority: Minor > Attachments: HDFS-15768.001.patch > > > Currently, if we can set incremental blockreport interval bigger than > heartbeat interval, it will use heartbeat interval. But in some case, we want > to reduce the pressure of incremental blockreport to namenode and send the > incremental block changes in more big batch. -- 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