Enhancing Performance and Scalability in Microservices

Advanced solutions are required to improve the performance and scalability of microservices due to the dynamic and more complex nature of backend systems. In order to overcome these difficulties, this paper presents a Quantum Gaussian Process Regression (QGPR) model that makes use of quantum-enhanced machine learning methods. Key metrics like prediction accuracy, resource utilisation, workload flexibility, computational efficiency, and Service Level Objective (SLO) compliance are used to compare the suggested model to more conventional methods like Neural Networks (NN), Gaussian Process Regression (GPR), and Quantum Support Vector Machines (QSVM). By obtaining the lowest Mean Absolute Percentage Error (MAPE) of 5.5% and the highest R2 score of 0.97, the results show how well QGPR performs in resource demand forecasting. With a CPU utilisation of 79%, network utilisation of 84%, and minimum overprovisioning of 4%, QGPR also maximises resource utilisation. With a SLO violation rate of less than 1% and a quick adaption period of 10 seconds, it also demonstrates remarkable flexibility. With the highest SLO compliance rate of 99% and the fastest prediction time of 25 milliseconds, QGPR shows that it can achieve demanding performance standards while reducing computational overhead. These results demonstrate how QGPR has the ability to transform microservices for backend applications by providing a scalable and effective architecture that can manage fluctuating workloads, improve resource efficiency, and guarantee dependable system performance. The paper emphasises the benefits of using quantum enhanced techniques to increase the functionality of contemporary backend systems.