log creating in /tmp of mapreduce job.

[#1]

I have one query. When we run any any mapreduce job that time it will create the job logs in /tmp on client only or it will create the logs on datanode also at /tmp location? And also same query for the yarn?

[#2 · Accepted solution]

@Ashnee Sharma

For YARN applications log aggregation takes case of log management

Please read below blog for more details

http://hortonworks.com/blog/simplifying-user-logs-management-and-access-in-yarn/