GreenPlanning: Optimal Energy Source Selection and Capacity Planning for Green Datacenters
GreenPlanning: Optimal Energy Source Selection and Capacity Planning for Green Datacenters
复制标题
DOI：
10.1145/2591971.2592025
复制
发表时间：
2014-06
期刊：
影响因子：
--
通讯作者：
Fanxin Kong;Xue Liu
中科院分区：
文献类型：
--
作者：
Fanxin Kong;Xue Liu
微软和谷歌等云服务提供商正开始使用多种能源为其云计算中心供电。为了降低成本和排放，他们将绿色能源纳入电力供应，同时为了提高服务可用性，他们使用传统（通常是棕色）能源来支持电力中心。然而，由于用于不同目标的能源的不同特征，挑战出现了。如何选择最佳的能源和规划他们的能力，以满足成本，排放和服务的可用性要求，建设炼油厂仍有待充分探讨。这项工作为解决这一问题提供了一个整体解决方案。我们提出了绿色规划，一个框架，在多个能源，电网电源和能源存储设备之间的数据中心在上述三个目标的明智的平衡。绿色规划研究了可供零售商使用的各种绿色和棕色能源的不同特征和操作。该框架最小化了数据中心的生命周期总成本，包括资本和运营成本。我们进行了广泛的模拟，以评估绿色规划与现实生活中的计算工作量和气象数据的痕迹。结果表明，绿色规划可以减少生命周期的总成本和排放量的50%以上，相比传统的配置，同时仍然满足服务的可用性要求。
Cloud service providers such as Microsoft and Google are beginning to power up their datacenters using multiple energy sources. To reduce cost and emission, they incorporate green energy sources into the power supply, while to improve service availability, they back up datacenters using traditional (usually brown) energy sources. However, challenge arises due to distinct characteristics of energy sources used for different goals. How to select optimal energy sources and plan their capacity for constructing datacenters to meet cost, emission and service availability requirement remains to be fully explored. This work provides a holistic solution to address this problem. We present GreenPlanning, a framework to strike a judicious balance among multiple energy sources, grid power and energy storage devices for a datacenter in terms of the above three goals. GreenPlanning investigates different features and operations of a wide spectrum of green and brown energy sources available to datacenters. The framework minimizes the lifetime total cost including both capital and operational cost for a datacenter. We conduct extensive simulations to evaluate GreenPlanning with real-life computational workload and meteorological data traces. Results demonstrate that GreenPlanning can reduce the lifetime total cost and emission by more than 50% compared to traditional configurations, while still satisfying service availability requirement.