综合论文：OnBigDataBenchmarking（英文）.pdfVIP

On Big Data Benchmarking 1Rui Han and 2Xiaoyi Lu 1Department of Computing, Imperial College London 2Ohio State University r.han10@imperial.ac.uk, luxi@ Abstract mainstream application domains, namely search engine, social network, and e-commerce, of internet service workloads), and the Big data systems address the challenges of capturing, storing, emergence of new systems driven by the exploration of big data managing, analyzing, and visualizing big data. Within this value, covering diversity of workloads is the prerequisite to context, developing benchmarks to evaluate and compare big data perform successful and efficient benchmarking tests. Within this systems has become an active topic for both research and industry context, we propose our insights into the requirements and communities. To date, most of the state-of-the-art big data challenges in developing big data benchmarks. We also present benchmarks are designed for specific types of systems. Based on the methodology on big data benchmarking, which represents our our experience, however, we argue that considering the thinking about how to address these challenges. Finally, we complexity, diversity, and rapid evolution of big data systems,