Large-scale data centers for cloud computing services consist ofa number of commodity servers, resulting in a huge amount ofpower consumption. In order to save power consumption, BEEMR(Berkeley Energy Efficient MapReduce), a MapReduce workload manager,is proposed. In a BEEMR-based data center, servers are allocated toeither of the interactive and batch zones. Arriving jobs of a small sizebegin to be processed immediately in the interactive zone, whilelarge-sized jobs are queued and served simultaneously at every fixedservice period in the batch zone. In this paper, we analyze the performanceof BEEMR-type job scheduling. We consider two queueing models forthe interactive and batch zones. The interactive zone is modeled asa single-server queueing system with processor-sharing (PS) service.In terms of the batch zone, we consider a queueing system withgated service in which arriving jobs are queued and begin to be servedwhen a fixed service period starts. For these models, the time-averagepower consumption and the mean response time are derived.Numerical examples show that the power consumption is significantlyaffected by the allocation of servers to both zones, while the powerconsumption is insensitive to the length of the batch-service period.