Ville Tuulos talks about Disco, the Map/Reduce framework for Python and Erlang, real-world data mining with Python, the advantages of Erlang for distributed and fault tolerant software, and more.
Ron Bodkin discusses big data architecture, real-time analytics, batch processing, map-reduce, and data science.
Adrian Cole discusses his jclouds project, which is an open source library that helps Java developers get started in the cloud and reuse their Java development skills. Cole also talks about some of the challenges of creating a cloud agnostic library, such as the use of different hypervisors and that various cloud implementations are written in different languages, such as VB, Python, Ruby, etc.
Billy Newport talks to InfoQ about the need for higher level abstraction to do parallel programming with multi-core systems effectively. The interview explores some approaches taken with MapReduce products such as Cascading and Pig for a Hadoop cluster, explores the limitations of the actor model and message passing, and touches on IBM's WebSphere eXtreme Scale (ObjectGrid) product.