Effective Load Balancing with Adaptive Task Scheduling in the Cloud using the Weighted-Priority Approach

In this decade, "Cloud-Computing (CC)" is growing in popularity as a means to facilitate the pooling of computing resources across networks. The cloud is becoming more popular among businesses due to its many advantages, including lower costs and fewer maintenance requirements. Even so, the issues of "Task-Scheduling (TS)" and "Load-Balancing (LB)" arise as a consequence of increasing traffic, which happens when the range of "Cloud-Users (CUs)" increases. Many unique, dynamic approaches were created and placed into use to deal with these TS. Because CC resources and tasks are so diverse, TS is considered a full-blown issue. This is because the efficiency of such dynamic approaches fluctuates with variable factors. Finding an ideal or effective outcome to the TS issues is still a difficult task. For TS in a CC context, this research combines the innovative "Rank-based Weighted-Priority (RWP)" method alongside the established "Divide & Conquer (DC)" data-mining approach. Here, it uses the DC technique to sort and rank the tasks through the creation of "Task-Groups (TGs)" according to their size and duration. Following their respective CPU and memory capacities, the "Virtual-Machines (VMs)" need to be divided into the corresponding "VM-Groups (VMGs)". The suggested DC-RWP technique dynamically assigns work to VMs based on their priority. Moving workloads from overworked VMs to VMs that are underutilized requires a similar procedure. The simulated scenarios are run with Cloud-Sim with different cloud resource settings. In comparison to the current MOPTSA3C techniques, the suggested DC-RWP technique outperforms it across all parameters, including LB Efficiency, Response-Time, and Waiting-Time, according to the simulation findings. Managing tasks with a higher priority has also been far superior.