by Nathan Marz
A new breed of technologies has emerged to handle the scale of Big Data, including Hadoop, HBase, Cassandra, and other "NoSQL" databases. However, these technologies are not one-size-fits-all solutions. Using them effectively requires a whole new set of data processing and data management techniques. In combination, however, these tools can solve any Big Data problem in a way that's robust and highly performant. The key is to embrace both batch processing and stream processing for building these systems. By doing the heavy lifting with batch processing and augmenting those results with fast computations done by a stream processing system, you can build realtime Big Data systems that are robust, scalable, and extensible.
23rd–25th March 2011