- by tomsg, Data Doctor
- by SaneIT, Data Doctor
- 7/17/2015 8:17:22 AM
"Running SAS on Hadoop allows you to process an analytics job strictly in memory, which enables the "discovery" stage of a big data application. Try something new with a large dataset, and if you aren't happy with the results, you can take the "fail fast" approach, abandon the idea, and try a new approach."
I think this is one of the biggest drivers for Hadoop, and not just failing fast but failing fast frequently. Not feeling like you just wasted a day processing a whole bunch of data that didn't come out the way you hoped is a big deal. Being able to process large chunks of data a half dozen times a day tweaking until you get what you want is a big step up from trying to mold that overnight batch job into what you needed and regroup for the next big run is a huge step.