Run-time workload driven structure in statistics for the cloud

Run-time statistical reporting function is popular in enterprise information system. However, most of these functions have the same problems about how to run more quickly, and decrease occupation of server resources in practice. The emergence of cloud computing provides a different approach to solve the problems. Recent trends show that cloud computing is becoming a robust paradigm for supporting modern data-intensive enterprise applications. In this paper, we provide a run-time workload driven structure. Our applications and experiments show that the structure is efficient, and estimable. It is applicable in the larger-scale and real-time required statistics particularly.