Shangshu Qian created HDFS-17838:
------------------------------------

             Summary: Potential Feedback Loop in Block Deletion
                 Key: HDFS-17838
                 URL: https://issues.apache.org/jira/browse/HDFS-17838
             Project: Hadoop HDFS
          Issue Type: Bug
    Affects Versions: 3.4.1
            Reporter: Shangshu Qian


We find a potential feedback loop involving MarkedDeleteBlockScrubber.

Delay in block deletion (BlockManager$MarkedDeleteBlockScrubber)
-> Contention causes IOException in the write pipeline
-> Due to high load caused by the block deletion, DN times out.
-> The deletion needs to be retried, causes even more delays.



--
This message was sent by Atlassian Jira
(v8.20.10#820010)

---------------------------------------------------------------------
To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org

Reply via email to