Designing of Resource Provisioning Framework Based on QoS SLA and Time in Cloud Environment

Abstract

In cloud computing, resource provisioning plays a pivotal role in facilitating timely and efficient access to computing resources for customers. This process involves allocating and managing essential resources like storage space, virtual machines and network capacity, which are essential for seamless application delivery. However, cloud providers encounter challenges related to quality of service (QoS), SLA negotiation, and scalability, impacting customer satisfaction and resource allocation efficiency. To address these challenges, this research focuses on designing a resource provisioning framework that offers flexible and cost-effective access to computing resources. The research begins with an in-depth study of QoS-based resource provisioning in cloud computing, aiming to highlight the importance of different QoS attributes in framework design. A comprehensive survey explores various provisioning frameworks and techniques, emphasizing scalability as a key QoS attribute. Our First research module introduces the Task-Side Service Level Agreement (TS2LBDP) model, employing the Elephant Herd Optimization algorithm to enhance task scheduling with deadline awareness. Using pattern analysis techniques increases the range of tasks and improves scheduling effectiveness, leading to better efficiency in scheduling, a broader array of tasks, and an increased rate of meeting deadlines The second module focuses on time-based resource provisioning frameworks to reduce long wait times for users. A Particle Swarm Optimization-based model is designed to improve scheduling efficiency and resource utilization. The third module addresses SLA violations detection and mitigation, utilizing a 1D Convolutional Neural Network (CNN) and Q-Learning framework. By analyzing cloudlogs, the model identifies patterns indicative of SLA violations and recommends mitigation strategies, reducing violations and scheduling delays under peak loads. The fourth module presents a Negotiation-Aware Service Level Agreement (SLA) Model