HPC2-ARS: An Architecture for Real-Time Analytic of Big Data Streams

Abstract

HPC2-ARS supports a high performance cloud computing (HPC2) based streaming data analytic system, which ensures real-time response on unpredictable and fluctuating Big Data Streams by provisioning and scheduling computing resources autonomously. It focuses on parallel high-volume streaming applications, which have stringent real-time constraints and bring Big Data issues. It is a brand-new three-layered architecture, which solves three essential problems: (a) how many resources are needed for each application to achieve real-time analytic on streaming Big Data, (b) where to best place the allocated resources to minimize resource consumption, and (c) how to minimize response time for parallel applications. In summary, HPC2-ARS provides high performance streaming services.