Summary
Because response time is important in an increasing number of applications, the distance between the user and the service is important. But, with a globally distributed base of users, if the cloud is to appear location independent, stringent response-time constraints can be met only with a distributed architecture.
The economics of building your own distributed service nodes or retail branches have rapidly diminishing returns: When the radius of a node is halved, the number of nodes required to cover a given area must be quadrupled. However, for many applications and scenarios, such as maintaining partitioned user data, the pay-per-use model of the cloud enables customers to enjoy latency reduction for free by exploiting dispersion, in the same way that, for many applications, the pay-per-use model of the cloud enables processing time acceleration for free.
Get Cloudonomics: The Business Value of Cloud Computing, + Website now with the O’Reilly learning platform.
O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.