Abstract

Cloud systems are currently the most popular concept of information systems and are the result of the evolution of a chain of methods for their construction. The main task of cloud technologies is to create a virtual cloud system consisting of virtual distributed resources. These resources provide remote provisioning of cloud access services with the required level of customer service Analysis and load balancing in cloud systems is quite an urgent task, as most open access cloud systems use simple load schedulers for their physical servers. The problem of load balancing requires a solution not when the server unexpectedly failed in the process of working on the task, which discourages users from using such a product, but at the very beginning of the project. In the early stages of design, it is acceptable to increase capacity by connecting new servers or using code optimization algorithms. But when a certain limit is reached, these measures become insufficient. The article reviews the methods of load balancing in cloud systems. It is shown that the existing methods of load balancing of cloud systems have limited use and currently there is no universal load balancing system. In addition, none of the considered methods takes into account such important components of systems as network and disk subsystem. Load balancing methods for cloud systems require improvement, the purpose of which should be the ability to fully monitor the system to meet the requirements of users and developers.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call