Towards a Leaner Geo-distributed Cloud Infrastructure

Modern cloud infrastructures are geo-distributed. Geo-distribution offers many advantages but can increase the total cloud capacity required. To achieve low la-tency, geo-distribution forfeits statistical multiplexing of demand that a single data center could benefit from. Geo-distribution also complicates software design due to storage consistency issues. On the other hand, geo-distribution can lower costs through eliminating redun-dancies at individual sites or exploiting regional dif-ferences in energy prices. We discuss several factors that influence geo-distributed capacity provisioning, and quantify latency, availability, and capacity trade-offs that emerge. We describe open research challenges in design-ing software that efficiently uses cloud capacity.