[ 
https://issues.apache.org/jira/browse/HDFS-10220?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15251150#comment-15251150
 ] 

Walter Su commented on HDFS-10220:
----------------------------------

1. isMaxFilesCheckedToReleaseLease is not requirted to be a function.
2. repeat [~vinayrpet] said, removeFilesInLease(leaseToCheck, removing); may 
not be required.
3. The LOG.warn("..") is kind of verbose.
4. I think the config should keep inside. It's about implemention detail. The 
re-check interval is 2s and is hard-coded too. Besides, It's too complicated 
for user to pick a right value. Instead of counting the files, I prefer 
counting the time. If it holds the lock for too long, log a warning and break 
out for a while.
5. btw, HDFS-9311 should solve this issue.


> Namenode failover due to too long loking in LeaseManager.Monitor
> ----------------------------------------------------------------
>
>                 Key: HDFS-10220
>                 URL: https://issues.apache.org/jira/browse/HDFS-10220
>             Project: Hadoop HDFS
>          Issue Type: Bug
>          Components: namenode
>            Reporter: Nicolas Fraison
>            Assignee: Nicolas Fraison
>            Priority: Minor
>         Attachments: HADOOP-10220.001.patch, HADOOP-10220.002.patch, 
> HADOOP-10220.003.patch, HADOOP-10220.004.patch, threaddump_zkfc.txt
>
>
> I have faced a namenode failover due to unresponsive namenode detected by the 
> zkfc with lot's of WARN messages (5 millions) like this one:
> _org.apache.hadoop.hdfs.StateChange: BLOCK* internalReleaseLease: All 
> existing blocks are COMPLETE, lease removed, file closed._
> On the threaddump taken by the zkfc there are lots of thread blocked due to a 
> lock.
> Looking at the code, there are a lock taken by the LeaseManager.Monitor when 
> some lease must be released. Due to the really big number of lease to be 
> released the namenode has taken too many times to release them blocking all 
> other tasks and making the zkfc thinking that the namenode was not 
> available/stuck.
> The idea of this patch is to limit the number of leased released each time we 
> check for lease so the lock won't be taken for a too long time period.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Reply via email to