Load Balancing For Presence Server Architecture

Abstract- The importance of social network applications on mobile devices is becoming more popular. The mobile presence service is a vital component in social network through which it handles every mobile user information, such as the network address, present status and GPS location, and also updates the user online buddies with the information frequently. When plenty updates occur continuously more number of messages are sent to the servers, this may lead to scalability problem in large scale mobile presence services. To address this, we use a server architecture known as PresenceCloud, which supports large scale social network applications. In day to day life, the number of mobile users of social network applications is increasing rapidly, due to this the response time from the server may be decreased by getting more requests from the mobile users. In our proposed method we are introducing an algorithm for Load balancing, which allocate the work to the clusters of presence server. The load balancing algorithm is Transaction-Least-Work-Left(TLWL), used to allocate work to least values of the servers. This algorithm reduces the response time by distributing the requests to all other servers. Keywords- Distributed Presence servers, cloud computing, social networks, load balancing, mobile presence services I