Intelligent SLA Selection Through the Validation Cloud Broker System
Ihab Sekhi, Károly Nehéz · IEEE Access · 2024
Cloud computing has transformed digital service delivery by providing scalable, flexible access to computing resources, including servers, storage, and applications, under a pay-per-use model. This model utilizes geographically distributed data centers to enhance service delivery and dynamically adjust Service Level Agreement (SLA) pricing based on demand. However, effective resource allocation strategies remain challenging, especially for ensuring low latency and fast execution in real-time applications and interactive services. Increased data center load can degrade performance, impacting cost and productivity. . . To address these challenges, we developed the Intelligent Validation Cloud Broker System (IVCBS). Uses an algorithm to classify virtual machine (VM) resources and match them with users’ request sizes, relying on a mathematical model aligned with the trapezoidal membership function in fuzzy logic. This reduces fuzzy rules and improves decision-making accuracy. We tested 11 types of AWS General Purpose EC2 specifications across 31 data centers in various regions. Implementing and comparing IVCBS with a traditional method through two policies—optimize response time and dynamically reconfigure load—showed that the IVCBS with optimized response time policy outperformed in terms of overall response time, data center processing, and total VM cost. IVCBS addresses scalability and performance challenges by efficiently assigning VMs, managing workload distribution, and preventing data center overload. Improving the average data center request servicing time maintains a high quality of service (QoS) and energy optimization.