hadoop-hdfs-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Alejandro Abdelnur (JIRA)" <j...@apache.org>
Subject [jira] Commented: (HDFS-1131) each DFSClient instance uses a daemon thread for lease checking, there should be a singleton daemon for all DFSClient instances
Date Wed, 05 May 2010 07:38:03 GMT

    [ https://issues.apache.org/jira/browse/HDFS-1131?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12864217#action_12864217
] 

Alejandro Abdelnur commented on HDFS-1131:
------------------------------------------

A solution would be to have a singleton variable  =ScheduledThreadPoolExecutor= in the DFSClient
and all DFSClients should schedule their =LeaseChecker= with it.

The =LeaseChecker= =run()= method should be changed not to loop/sleep letting the =ScheduledThreadPoolExecutor=
to take care of that.

On close the =LeaseChecker= should remove from the =ScheduledThreadPoolExecutor=



> each DFSClient instance uses a daemon thread for lease checking, there should be a singleton
daemon for all DFSClient instances
> -------------------------------------------------------------------------------------------------------------------------------
>
>                 Key: HDFS-1131
>                 URL: https://issues.apache.org/jira/browse/HDFS-1131
>             Project: Hadoop HDFS
>          Issue Type: Improvement
>            Reporter: Alejandro Abdelnur
>            Priority: Critical
>
> When accessing HDFS from a server application that acts on behalf of several users each
user has its own file system handle (DFSClient), this creates one daemon thread per file system
instance.

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


Mime
View raw message