Downloads: 7 | Views: 47 | Weekly Hits: ⮙6 | Monthly Hits: ⮙6
Research Paper | Information Technology | United States of America | Volume 14 Issue 3, March 2025 | Rating: 6 / 10
Framework for Kubernetes High Availability and Resilience in Public Cloud
Harshavardhan Nerella, Prasanna Sai Puvvada, Praveen Borra
Abstract: High Availability (HA) and Resilience is crucial for applications to minimize downtime, ensure business continuity, and maintain consistent reliability. Kubernetes, as the de facto container orchestration platform, has been widely adopted by startups, enterprises, and government agencies. While Kubernetes does not offer HA and Resilience for clusters and workloads by default, it provides several mechanisms that can be used in tandem to achieve them. Managed Kubernetes services (e.g., AWS EKS, Azure AKS, Google GKE) offload the responsibility of HA and resiliency for the control plane, which serves as the ?brain? of Kubernetes. However, the responsibility for the data plane and workloads still lies with cluster administrators and developers. This paper proposes a multi-layered approach to achieving HA and Resilience in managed Kubernetes clusters. We categorize the approach into two layers: workload and infrastructure layer. For each layer, we outline a series of mechanisms that must be adopted to ensure HA and Resilience. By applying these techniques sequentially at each layer, Kubernetes clusters can handle failures gracefully, scale efficiently, and maintain availability under varying workloads and infrastructure disruptions.
Keywords: Kubernetes, High availability, Cloud Computing, Resilience
Edition: Volume 14 Issue 3, March 2025,
Pages: 1705 - 1709