Apache Spark uses Hadoop underneath and is doing very well. We use it in our production server for all of our machine learning algorithms