Datacenter net profit optimization with deadline dependent pricing

摘要

This paper considers deadline dependent pricing and its impact to datacenter net profit optimization. We formulate the problem by jointly maximizing total revenue as a function of individual job deadlines and minimizing electricity cost through job scheduling over different time and locations. These two complementary objectives-the maximization of revenue and the minimization of cost-are mutually-dependent due to the coupling of job completion time and scheduling decisions. Lever-aging a new approximation method for job completion time, we develop two low-complexity, distributed algorithms for the net profit optimization. Through numerical evaluations, we show the efficacy of the proposed algorithms as well as the net profit improvements.