lavenderwords/googleclusterdata

Identifcation of MapReduce jobs/tasks?

Opened this issue · 0 comments

I assume that these traces contain many MapReduce jobs. Since my research topic 
is the performance modeling of MapReduce jobs, I am very interested in 
identifying the MapReduce jobs in these traces, and in being able to 
distinguish the map and reduce tasks.

Would you (Google) be able and willing to provide a mapping for all MapReduce 
tasks to task type (eg., (job ID, task index) -> (map|reduce|...)), or if that 
is not feasible maybe just a list of logical jobnames (or job IDs) of MapReduce 
jobs?

Original issue reported on code.google.com by tho...@de-ruiter.cx on 8 Dec 2011 at 3:38