Power-utilization provisioning for data centers

The applications hosted in a datacenter share more than just servers; they also share electrical circuits. Datacenter managers provision the power capacity of these circuits to hosted applications, often based on their peak power needs. In this work, we studied the actual and peak power needs of 3 real datacenters, using data from 1) hardware manufacturers and 2) actual, observed power needs to estimate peak needs. We found that actual power needs were nonmonotonic relative to peak needs. That is, some applications with low actual power needs had large peak needs-stemming from the diverse power utilization of datacenter applications. Such diversity caused surprising order inversions where applications with smaller peak power had larger actual needs. Based on these results, we propose a power provisioning approach that considers power-utilization diversity. Our approach provides 1) predictable monotonic results as power capacity increases and 2) performs better than approaches commonly used in practice.