Providing Quality of Service on Services Selection using Anycast Techniques

Over the last years, the complexity and variety of services available on the Internet increased. This fact is leading to the search for efficient techniques of routing client requests to the best server available. A known technique is the application layer anycast (ALA). The main goal of this work is to elaborate efficient ways to provide ALA with quality of service in the context of cloud computing. To achieve this goal, we proposed a new algorithm (GALA, Global Application Layer Anycast). It inherits characteristics from another existing system (GAA, Global Application-layer Anycast), and uses geolocation as a differential. The results of the experiments show that GALA, compared to the inherited algorithm, maintains the client requests efficiencies and substantially lowers their latencies.