Enhancing Cloud Performance through HLD_ASG by using a Two-Tier Load Balancing Strategy

Cloud computing has significantly reformed modern IT infrastructure by providing scalable access to computing resources over the Internet. Proper organization of job tasks and keeping loads even are vital for retaining fast performance, scalability and low power use in cloud computing. Traditional methods like FCFS, Round Robin and LD_ASG pay very little attention to the overall load of each host in favor of tracking VM statistics. Because of these issues, some jobs are performed less efficiently, take longer to complete and burn more energy. Since similar algorithms have certain shortcomings, this study introduces HLD_ASG as a Host-Level and VM-Level scheduler that improves task placement and utilization of computing resources. At the first step, the system calculates the heavy and light workloads of every host by looking at how much CPU and memory is used; at the second step, it compares all VMs running on the selected host and picks the one being used the least. Trying out HLD_ASG on CloudSim, we found that it can reduce the time it takes by 25%, increase throughput by 20% and consume 17% less power than LD_ASG. The outcomes demonstrate that HLD_ASG is an efficient and convenient choice for load balancing in fast-changing cloud environments.