Downloads: 1
Research Paper | Information Technology | Volume 13 Issue 3, March 2024 | Pages: 1977 - 1983 | United States
A Scalable and Resilient Cloud Architecture for Next-Generation Intelligent Applications
Abstract: The demands of intelligent applications of the next generation require cloud architectures which are both elastically scalable, resilient to failure, and manage heterogeneous AI workloads with predictable performance. The paper will provide a resilient and scalable cloud architecture that provides edge-assisted data ingestion, containerized microservices, distributed model serving, and policy-limited autoscaling to address these needs. Its architecture uses stateful service copies, consensus-based control planes, and disaggregated replicated data stores to eliminate single points of failures as well as incremental growth. Adaptive placement schemes reduce the end-to-end latency by matching compute, and data locality to achieve better results; canary rollouts to achieve transparent rollback would increase the reliability in continuous deployment. Operational governance of the multi-cloud environment is made possible by integrated observability, security-by-design, as well as cost-aware resource management. The services of transparent model updates, fault-isolated multi-tenancy, and graceful degradation under overload are support services to provide mission-critical applications with service continuity. Representative workload evaluation indicates a scale-out performance that is quite linear, shorter recovery times, and a limit curve in latency, will provide a viable blueprint in implementing resilient cloud platforms that lead to the next generation of AI-driven services. Future work is based on even further automated policies of resilience and cross-layer optimization.
Keywords: Cloud computing, scalable architecture, fault tolerance, resilient systems, microservices, intelligent applications
How to Cite?: Sunil Netra, Anil Vijarnia, "A Scalable and Resilient Cloud Architecture for Next-Generation Intelligent Applications ", Volume 13 Issue 3, March 2024, International Journal of Science and Research (IJSR), Pages: 1977-1983, https://www.ijsr.net/getabstract.php?paperid=SR24315103949, DOI: https://dx.dx.doi.org/10.21275/SR24315103949