Is it readily possible to integrate a Hadoop client with python (django) mapreduce apps/scripts remotely (on Heroku dynamos or from a free cluster) as is done locally in these examples:
- http://www.michael-noll.com/tutorials/writing-an-hadoop-mapreduce-program-in-python/
- http://blog.matthewrathbone.com/2013/11/17/python-map-reduce-on-hadoop---a-beginners-tutorial.html
- Hadoop and Django, is it possible?
This Heroku addon led me to believe this might be possible: https://devcenter.heroku.com/articles/treasure-data. But this app isn't free and the learning-curve-to-cost ratio is not an obvious investment to me.
My motivation to cross the Heroku/Django/Hadoop bridge is to upgrade my current Django apps with social media mining features.