Some Joules Are More Precious Than Others: Managing Renewable Energy in the Datacenter∗

Large cloud-computing datacenters now host a wide range of business applications, ranging from e-commerce websites to search engines to data mining. Increasingly, these datacenters use renewable energy from wind turbines or solar panels to reduce their dependence on costly and less clean energy from the grid [17, 4, 1]. This paper argues that datacenter management should be revised to maximize the use of such off-grid renewables.1 Our argument is similar in spirit to the case for energyefficient management [18, 7, 11]. But efficiency alone may not take full advantage of renewables. For instance, energy-efficient management may batch writes so disks can sleep for long durations. But sometimes renewables are intermittently unavailable, long sleeps may miss opportunities to use them. Operating systems and cluster management software already manage energy, CPU cycles, and other low-level resources. To be sure, many past techniques will probably carry over to renewables. However, unlike other managed resources, renewables are available only when the wind blows or the sun shines. Such intermittency makes it hard for datacenters to use renewables effectively. For example, windy nights that follow calm afternoons would produce renewables only during traditionally light workload times for datacenters [15, 2]. In the first part of this paper, we model a datacenter that gets power from intermittent wind renewables. Our re-