Research And Implement Of Load Balancing On Cloudstack

As a kind of bussiness model that base on net and shared resource pools,The Cloud Compute has following characteristic,convenience and purchase on demand.Receiving much attention of late makes it getting rapid development.The users love it because of its limitless resources and its simple way to use.But it does not mean this technology is perfect,there’re still many aspectes to research and develop.Just like the load balancing is a important kind of strategy in the cloud platform which has limitless resources.A good load balancing strategy can not only improves the computing performance of the cloud platform,but also its stability.If we want to improve the performance of the whole cloud platform,Load balancing is a question.In this paper I design and implement a loda balancing strategy on the platform of Cloud Stack and Xen Server.Firstly,research the existing load balancing stratgey, analysis what shortages these stratgeies still have and come up whit the way to solve these shortages.What’s more,I design and implement this load balancing stratgey which consistes of trigger module,choose module and location module.In location moudle I use the ant algorithm for it has an advantage over greed algorithm in sloving the NP hard problemes,So the performence of load balancing is improved in this module.And then I make some optimizing about the memory used and energy consumption.These optimizing not only improves the performence of load balancing,but also the time of task execution and the energy consumption.At last,I implement this stratgey in the platform of Cloud Stack and Xen Server and prove its advantage by expermentes with other kinds of stratgey.