[ https://issues.apache.org/jira/browse/HDFS-9761?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15133362#comment-15133362 ]
Chris Trezzo commented on HDFS-9761: ------------------------------------ +1. The patch looks good to me. I agree, I think the intent was to have the sleep call outside of the for loop. > Rebalancer sleeps too long between iterations > --------------------------------------------- > > Key: HDFS-9761 > URL: https://issues.apache.org/jira/browse/HDFS-9761 > Project: Hadoop HDFS > Issue Type: Bug > Components: balancer & mover > Reporter: Adrian Bridgett > Priority: Blocker > Attachments: HDFS-9761.000.patch > > > Whilst reading the code to try and determine why I'm seeing bad performance I > think I've spotted an inadvertent move of a sleep() to inside a for loop. > This means that instead of sleeping once per iteration the balance is > sleeping once per namenode per iteration. If I've done my maths correctly > the sleep time default is 18seconds so this is a nice little speedup for HA > clusters. > Commit d31a41c35927f02f2fb40d19380b5df4bb2b6d57 moved the sleep in > org/apache/hadoop/hdfs/server/balancer/Balancer.java:run() -- This message was sent by Atlassian JIRA (v6.3.4#6332)