This article explains why scaling GPU inference services on Kubernetes is challenging and presents a multi‑layer control architecture, metric upgrades, and production‑ready implementations using HPA, KEDA, KServe, and Karpenter to achieve stable, cost‑effective autoscaling.
Argo CD 3.3 introduces a suite of production‑ready features—including native PreDelete hooks for safe resource cleanup, an enhanced Source Hydrator with Git notes and inline parameters, automatic OIDC token refresh, shallow Git cloning for large repos, and built‑in KEDA autoscaling—plus detailed upgrade guidance and best‑practice recommendations.
Argo CD v3.3 introduces long‑awaited features such as PreDelete hooks for cleanup before resource removal, resource‑name‑based ClusterResourceWhitelist, shallow Git clone support, first‑class KEDA integration with pause/resume and health checks, plus numerous UI, CLI, and performance enhancements.
This article explains the fundamentals of event‑driven architecture, introduces Knative Eventing as a cloud‑native serverless framework, details its broker/trigger model, event sources, orchestration patterns such as Sequence and Parallel, and shows how to integrate it with Alibaba Cloud EventBridge while comparing it to KEDA.
This article explores Kubernetes horizontal pod autoscaling mechanisms, comparing HPA and KEDA, their implementation principles, limitations, and practical deployment scenarios for cloud-native applications.
This article explains pod‑level elasticity in Kubernetes by detailing the principles, metric types, and limitations of the Horizontal Pod Autoscaler (HPA) and then introduces KEDA as an event‑driven extension that adds true scale‑to‑zero capabilities, complete with configuration examples and code snippets.
This article explores the business scenario of data archiving in autonomous driving platforms, explains why a serverless architecture was chosen, and details how OpenFunction combined with Dapr, KEDA, and microservices addresses challenges such as diverse storage types, scaling, and observability, providing a flexible, elastic solution.
This article details how the QingJiao Classroom platform evolved from a single‑cluster setup to a multi‑cluster architecture using KubeSphere on QingCloud (QKE), covering selection criteria, deployment strategies, automation, monitoring, authentication, elastic scaling with Keda, and future plans.
This article explains how to scale Kubernetes clusters and applications using tools like Cluster Autoscaler, Virtual Kubelet, HPA, VPA, and especially KEDA, while showcasing Alibaba Cloud EDAS's integration and future AI‑driven autoscaling enhancements.
KEDA 2.0, now a CNCF sandbox project, adds many new event‑driven triggers, introduces a dedicated ScaledJob resource, supports both Deployments and Jobs, and provides fresh scalers such as Azure Log Analytics, IBM MQ, a CPU scaler, and a Metrics API extender, making Kubernetes autoscaling far more responsive to external workloads.
This article explains the concepts of vertical and horizontal elastic scaling, explores how Spring Cloud and Kubernetes native features like HPA and KEDA enable precise scaling, and presents EDAS‑based optimizations for rule triggering, instance scheduling, and serverless Kubernetes deployments.
