The Response Rank based Fault-tolerant Task Scheduling for Cloud System
Sheikh Umar Mushtaq, Sophiya Sheikh, Ajay Nain · Advances in intelligent systems research/Advances in Intelligent Systems Research · 2024
A cloud computing platform has higher failure rates because of its highly dynamic nature and running of concurrent applications.However, the outcomes of running concurrent applications won't be accurate without VM synchronization.The issue of coordination between many virtual machines (VMs) is their synchronization in working is rarely considered by existing solutions.Moreover, fault tolerance constitutes one of the most crucial components for cloud computing architecture to ensure high reliability.In this research, Reserved Fault Tolerance and Ranked Task Scheduling (RFRTS) is proposed.Initially, the proposed ranked based scheduling approach is used for task allocation, and later reservation-based reactive fault tolerance method is suggested for a cloud system.To achieve the highest level of cloud computing infrastructure reliability, the suggested technique considers CPU faults and the VM reservation will ensure the assignment of alternative VM to the affected task.The proposed fault-tolerant approach has been compared with three existing reliable fault-tolerant approaches namely multi-objective scheduling algorithm with Fuzzy Resource utilization (FR-MOS), Cost-effective Workflow Scheduling Algorithm (CWS), and Fault-tolerant Cost-effective Workflow Scheduling Algorithm (FCWS) based on reliability.The outcomes unequivocally show that our suggested RFRTS algorithm surpasses the current FR-MOS, CWS, and FCWS considering the reliability in all the states.