Dynamic resource provisioning is a main challenge in cloud computing due to distinct task resource requirements. An abnormal workload creates resource famine, resource wastage, haphazard resource and task allocation that influence task scheduling, and machine resource usage leads to SLA violation. To cope-up this issue, we propose a strategy Categorization of a Task with a Resource to assign VM (CTRV) scheduling approach for task consolidation. First, the Resource Requirement Rate (RRR) of each received task is asses to categorizes the tasks. Second, VMs are assorted based on resource capacity and maintained as CPU-set, memory-set, I/O-set, bandwidth-set, respectively. Subsequently, each task has been assigned to the respective VM when the maximum RRR value is equivalent to VM’s resource capacity. The effectiveness of our approach is described as theoretically and practically. We design three performance measurement metrics to validate the system, (1) resource utilization, (2) average response time, (3) deadline violation rates. The empirical outcomes confirm that CTRV enhances resource utilization efficiency by 30%, 25–35% diminishes energy consumption than extant algorithms without SLAs negotiation.

由于不同的任务资源需求，动态资源供应是云计算中的主要挑战。异常的工作负载造成资源饥荒、资源浪费、杂乱无章的资源和任务分配影响任务调度，机器资源使用导致违反SLA。为了解决这个问题，我们提出了一种策略分类具有资源的任务以分配 VM（CTRV）调度方法以进行任务整合。首先，评估每个接收到的任务的资源需求率（RRR）以对任务进行分类。其次，VM根据资源容量进行分类，分别维护为CPU-set、memory-set、I/O-set、bandwidth-set。随后，当最大 RRR 值等于 VM 的资源容量时，每个任务都已分配给相应的 VM。我们的方法的有效性被描述为理论上和实践上。我们设计了三个性能测量指标来验证系统，（1）资源利用率，（2）平均响应时间，（3）最后期限违规率。实证结果证实，CTRV 将资源利用效率提高了 30%，与没有 SLA 协商的现有算法相比，能耗降低了 25-35%。