Optimization mechanism for latency shortening and elasticity improvement in geographically distributed data centers .

摘要

In cloud computing network, a method for reassigning the load of the data center to another of the plurality of data centers site. Load, based on the derivative of the objective function that defines the performance characteristics of the cloud computing network in the data center sites, are reassigned. In this method, to obtain the respective derivative of a set of other data center sites, on the basis of the derivative obtained, in the set of a set of data center sites, and the data center increases the objective function is minimized Identify the site, based on the derivative and the identified other data center sites was determined to reassign the load between the data center sites, and other multiple data centers site. .