Cloudonomics and the Implications for Integrated Telecom Service Providers

摘要

A debate is raging between those who posit that cloud computing represents the future of enterprise IT, and those who suggest that cloud infrastructure services are too expensive relative to "do-it-yourself." The truth--determined by simulation analysis and basic mathematics-- turns out to lie somewhere in between: cloud infrastructure services are optimal for handling variable and unpredictable demand profiles, such as spikes due to seasonal demand, one-time events, or aperiodic product launches. In addition, dispersed cloud infrastructure can well complement consolidated enterprise IT, improving the response time of highly interactive applications ranging from on-line auctions to remote virtual desktops to immersive videoconferencing. rnThese basic economics of the cloud in turn drive a number of preferred architectural configurations, such as virtual private clouds that elastically grow from enterprise data centers into public cloud service provider infrastructure, and two-tier solutions where the interactive portion of the application is geo-dispersed but the legacy application exists as a single instance. rnTherefore, the economics of the cloud inexorably lead to the conclusion that telecommunications companies, who can integrate a variety of access methods from optical transport through VPN through Internet, coupled with optimized anycast routing, coupled with minimal hops and propagation delay through on-net services, are best poised to offer a range of voice, video, data, and multi-modal cloud services.rnWe will review and puncture a number of myths of the cloud, explain the ten laws of cloudonomics, and argue why telcos are best-suited to create value for a broad variety of customer segments.