I am using hadoop 0.16.3 in a production environment.
We have been using our system for a few weeks already
and are constantly running out of inodes.
We added cron job to clean the logs generated by tasktraskers/
but i am now doubting we are catching them all.
Anybody has an idea of what to clean periodically?
Are there any temp files generated by the jobs/tasks that need to be
cleaned as well?
Thanks a lot of any input.
This email has been scanned by the MessageLabs Email Security System.
For more information please visit http://www.messagelabs.com/email
Grokbase › Groups › Hadoop › common-user › August 2008