Job scheduling reservations on cloud resources

The current application of cloud computing focuses more on research problems. One of the main problems in the cloud is job allocation. Jobs are dynamically allocated to server processors. All cloud virtualized hardware is available to users on demand and can be dynamically upgraded. Resource scheduling is critical in research in the cloud, due to its large execution time and resource costs. The differences in resource scheduling criteria and parameters used cause various categories of Resource Scheduling Algorithms. Resource scheduling has a goal, identifying the right resources to schedule workloads in a timely manner and improving the effectiveness of resource utilization. In other words, minimizing workload completion time. Mapping the right workloads to resources will result in good scheduling. Another goal of resource scheduling is to identify adequate and appropriate workloads. So it can support scheduling of multiple workloads, to meet various QoS needs in cloud computing. The aim of this research is to determine the value of waiting time, idle time and makespan on cloud resources. The proposed method is to sort the arrival times of jobs with the least workload and place the jobs on a virtual view, before scheduling them on cloud resources. Experimental results show that the proposed method has an idle time of 25.3%, FCFS is 43.1% while for bacfilling it is 31.5%. The average makespan reduction for FCFS is 16.73%, for bacfilling it is 12.87%. The average decrease in AWT for FCFS was 13.3% for bacfilling of 12.03%. The results of this research can be applied to cloud rentals with flexible times.