Analysis of issues with load balancing algorithms in hosted (cloud) environments

In order to provide valuable information and influence the decision-making process of a load balancer, thus maintaining optimal load balancing in hosted (or cloud) environments, it is not enough just to provide information from networking part of the computer system or from external load balancer. Load balancing models and algorithms proposed in the literature or applied in open-source or commercial load balancers rely either on session-switching at the application layer, packet-switching mode at the network layer or processor load balancing mode. The analysis of detected issues for those load balancing algorithms is presented in this paper, as a preparation phase for a new load balancing model (algorithm) proposition. The new algorithm incorporates information from virtualized computer environments and end user experience in order to be able to proactively influence load balancing decisions or reactively change decision in handling critical situations.