Hi there, Just curious, can anyone provide a quick explanation or link to the source code of how Pig partitions data on a GROUP alias ALL operation? We're seeing some odd behaviour, likely caused by skew in our data, and was just curious how Pig will partition groups to reducers if there's no group key.
We've gotten around this already by providing our own partition key to reduce skew. Mike -- Mike Sukmanowsky Product Lead, http://parse.ly 989 Avenue of the Americas, 3rd Floor New York, NY 10018 p: +1 (416) 953-4248 e: m...@parsely.com
