Yank: enabling green data centers to pull the plug
nsdi'13 Proceedings of the 10th USENIX conference on Networked Systems Design and Implementation
Incentivizing Advanced Load Scheduling in Smart Homes
Proceedings of the 5th ACM Workshop on Embedded Systems For Energy-Efficient Buildings
Enabling datacenter servers to scale out economically and sustainably
Proceedings of the 46th Annual IEEE/ACM International Symposium on Microarchitecture
Feedback Green Control for Data Centers Autonomy
UCC '13 Proceedings of the 2013 IEEE/ACM 6th International Conference on Utility and Cloud Computing
Hi-index | 0.00 |
Several companies have recently announced plans to build "green" datacenters, i.e. datacenters partially or completely powered by renewable energy. These datacenters will either generate their own renewable energy or draw it directly from an existing nearby plant. Besides reducing carbon footprints, renewable energy can potentially reduce energy costs, reduce peak power costs, or both. However, certain renewable fuels are intermittent, which requires approaches for tackling the energy supply variability. One approach is to use batteries and/or the electrical grid as a backup for the renewable energy. It may also be possible to adapt the workload to match the renewable energy supply. For highest benefits, green datacenter operators must intelligently manage their workloads and the sources of energy at their disposal. In this paper, we first discuss the tradeoffs involved in building green datacenters today and in the future. Second, we present Parasol, a prototype green datacenter that we have built as a research platform. Parasol comprises a small container, a set of solar panels, a battery bank, and a grid-tie. Third, we describe GreenSwitch, our model-based approach for dynamically scheduling the workload and selecting the source of energy to use. Our real experiments with Parasol, GreenSwitch, and MapReduce workloads demonstrate that intelligent workload and energy source management can produce significant cost reductions. Our results also isolate the cost implications of peak power management, storing energy on the grid, and the ability to delay the MapReduce jobs. Finally, our results demonstrate that careful workload and energy source management can minimize the negative impact of electrical grid outages.