Ask HN: Clojure/Hadoop vs Python/Dumbo/Hadoop

Hello all,

I'm planning to shift some of my operations to hadoop, and given my intense dislike of Java, I'd prefer to operate on either one of Python or Clojure.

In order to use Hadoop with them, I have the option of either using Python-dumbo or a JVM based language like Clojure to directly use Hadoop. While I appreciate that the latter is likely to be a far superior option in terms of performance, I'm not very well versed with Clojure.

Can someone who's used Python-dumbo comment on whether it is something that can be used with reasonable performance on a production system? I don't want to have to spend time and effort learning the intricacies of Clojure if I can avoid it (for now!)..

Thanks!