Year: 2025
Type: article
Abstract: As enterprises increasingly migrate mission-critical applications to the cloud, achieving continuous availability across geographically distributed regions becomes essential for meeting service-level objectives and ensuring operational resilience.This paper presents high availability (HA) strategies for cloud-based active-active services, emphasizing the use of core infrastructure primitives such as global DNS load balancing, regionspecific load balancers, and smart health checks.Through a comparative analysis of leading cloud platforms-Google Cloud Platform (GCP), Microsoft Azure, and Amazon Web Services (AWS)-the paper outlines methodologies for designing robust, multi-region architectures capable of sustaining sub-second failover and seamless user experiences.It further explores the role of DNS-based routing, L4/L7 load balancing, CDN integration, and real-time observability tools in enabling automated recovery from regional and service-level failures.In addition to platform-specific evaluations, the paper provides actionable best practices for deploying HA systems, including low-TTL DNS configurations, comprehensive dependency-aware health checks, and fault injection testing.The paper concludes by reinforcing the importance of layered failover mechanisms and continuous monitoring in achieving operational excellence in modern cloud-native environments.
Language: English
Subfield: Information Systems
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W4410450383