This paper addresses the problem of composition service scheduling and resource allocation in cloud. In the development of cloud computing with data-intensive and compute-intensive features, many applications are based on Map/Reduce model to enhance the performance. A hierarchical scheduling scheme strategy, i.e., PPA2-level scheduler, is proposed in this paper for the composition services architecture to achieve planning of composition service scheduling with Map/Reduce model. While executing PPA2-level scheduler, it decomposes the total progress into three steps, i.e., preprocessing, pooling and allocating. It decides the job priority according to both service-level and task-level in the resource side. And the resources are allocated by taking the data locality and total job completion rate into account on the basis of cloud framework. In the numerical evaluation, it uses a Markov model to generate various scenarios of client requested composition services. Then it evaluates the strategy by applying the scheduling strategy mentioned above to compare with the default first-in-first-out scheduling (FIFO) of Hadoop. In different distribution of composition services, the proposed strategy also performs well, especially in I/O-Bound services. It achieves better performance (about 45 %) and efficiently decreases the probability of disk spill.