Financial institutions are driven to measure and understand risk with greater accuracy and speed, which is fueling the need to invest in conventional hardware to boost processing speeds as well as increasing the need for a distributed processing system to serve as a new data management architecture.

In this context, Mizuho Information & Research Institute, Inc., Mizuho Financial Group’s IT strategy company, has adopted the MapR Converged Data Platform for its cost performance and data and processing scalability. Selected from among the available Hadoop distributions, MapR has no single point of failure(*2) and is the most suitable platform for enterprise use. To further boost efficiency, the Asakusa Framework(*3) is being used for batch processing development and deployment on MapR.

By virtue of this new parallel distributed processing technology, the MapR-based system achieves processing times 15X those of conventional systems. Increases in the amount and data and processing performed can also be achieved by simply adding server nodes, which promises to reduce TCO and enhance ROI over the medium/long term.