[MAPREDUCE-4949] Enable multiple pi jobs to run in parallel

• Type: Improvement
• Status: Closed
• Priority: Minor
• Resolution: Fixed
• Affects Version/s: 2.0.2-alpha
• Fix Version/s: 2.0.3-alpha, 0.23.7
• Component/s: examples
• Hadoop Flags: Reviewed
• Assignee: Sandy Ryza
• Reporter: Sandy Ryza
• Created: 18/Jan/13 22:29
• Updated: 15/Feb/13 13:10
• Resolved: 22/Jan/13 16:29

Description

Currently the hadoop pi example uses a hardcoded temporary directory to store its inputs and outputs. This makes it so that only one pi job can run at a time, and that if it is cancelled, the temporary directory must be manually deleted.

I propose using a temporary directory based on a timestamp and random number to avoid these conflicts