The Study of Hadoop Application across Multiple Data Centers

across multiple clusters Abstract. Hadoop is a reasonable tool for cloud computing in big data and MapReduce paradigm may be a highly successful programming model for large-scale data-intensive computing application. However, traditional Hadoop and MapReduce have been deployed over local or tightly-coupled cloud resources with one data center. This paper focuses on the issue of Hadoop application across multiple data centers. A hierarchical distributed computing architecture of Hadoop is designed and proposed. The job submitted by user can be decomposed automatically into several subtasks which are then allocated and executed on corresponding cluster by location-aware manner. The presentation of the workflow shows the operating principle of this architecture