CRFTS: a cluster-centric and reservation-based fault-tolerant scheduling strategy to enhance QoS in cloud computing.

Cloud systems supply different kinds of on-demand services in accordance with client needs. As the landscape of cloud computing undergoes continuous development, there is a growing imperative for effective utilization of resources, task scheduling, and fault tolerance mechanisms. To decrease the user task execution time (shorten the makespan) with reduced operational expenses, to improve the distribution of load, and to boost utilization of resources, proper mapping of user tasks to the available VMs is necessary. This study introduces a unique perspective in tackling these challenges by implementing inventive scheduling strategies along with robust and proactive fault tolerance mechanisms in cloud environments. This paper presents the Clustering and Reservation Fault-tolerant Scheduling (CRFTS), which adapts the heartbeat mechanism to detect failed VMs proactively and maximizes the system reliability while making it fault-tolerant and optimizing other Quality of Service (QoS) parameters, such as makespan, average resource utilization, and reliability. The study optimizes the allocation of tasks to improve resource utilization and reduce the time required for their completion. At the same time, the proactive reservation-based fault tolerance framework is presented to ensure continuous service delivery throughout its execution without any interruption. The effectiveness of the suggested model is illustrated through simulations and empirical analyses, highlighting enhancements in several QoS parameters while comparing with HEFT, FTSA-1, DBSA, E-HEFT, LB-HEFT, BDHEFT, HO-SSA, and MOTSWAO for various cases and conditions across different tasks and VMs. The outcomes demonstrate that CRFTS average progresses about 48.7%, 51.2%, 45.4%, 11.8%, 24.5%, 24.4% in terms of makespan and 13.1%, 9.3%, 6.5%, 21%, 22.1%, 26.3% in terms of average resource utilization compared to HEFT, FTSA-1, DBSA, E-HEFT, LB-HEFT, BDHEFT, HO-SSA, and MOTSWAO, respectively.