I'm trying to use the aggregate framework but in a non-streaming fashion. I detailed what I'm doing pretty well here:
http://stackoverflow.com/questions/20085532/trying-to-make-hadoop-1-2-1-aggregate-framework-work-non-streaming I have a feeling there's a simple solution (involving setting the appropriate classpath somewhere) but I don't know what it is. Any help appreciated. There's remarkably little information available on non-streaming use of aggregates that I can find. Thanks, Nick