This article explores Kubernetes' core components—API Server, Controller Manager, Scheduler, and etcd—detailing their source‑code architecture, key mechanisms, extension points, performance tips, and practical steps for building operators, custom schedulers, and API extensions.
An in‑depth look at Kubernetes’ control plane reveals how the API Server, Scheduler, and Controller Manager work together to manage cluster state, handle authentication, schedule pods, and ensure convergence, with practical HA tips, advanced features, and real‑world deployment workflows.
In this walkthrough I show how, within minutes, I used AI‑driven code search and simple curl commands to locate, enable, and test the newly added API server feature in the open‑source Cherry Studio project, illustrating the power of an AI‑first workflow.
The article analyzes the December 11 OpenAI outage, revealing that a newly deployed telemetry service overloaded Kubernetes API servers, breaking DNS resolution and slowing recovery, and compares OpenAI’s approach with LoongCollector/iLogtail’s design to offer stability insights for cloud‑native environments.
An in‑depth post‑mortem reveals how OpenAI’s newly deployed telemetry service generated massive Kubernetes API requests, overloading the API server, breaking DNS resolution, and slowing recovery, while contrasting OpenAI’s approach with LoongCollector/iLogtail’s design to minimize API load and improve cluster stability.
This article explains how Kubernetes' API Priority and Fairness (APF) mechanism enhances kube‑apiserver traffic control by introducing FlowSchema and PriorityLevelConfiguration objects, allowing fine‑grained request prioritization, concurrency limits, and queue management beyond the basic inflight throttling flags.
This guide explains how to optimize Kubernetes API Server performance by configuring the max-mutating-requests-inflight limit and watch-cache-size, offering recommended values for different cluster sizes, monitoring metrics, and step‑by‑step adjustment strategies for stable, high‑throughput clusters.
This article explains how LMDeploy streamlines the deployment of Hugging Face transformer models by adding online conversion, offering an OpenAI‑compatible API server, a Gradio WebUI, and 4‑bit weight‑only quantization with AWQ, providing step‑by‑step commands, code examples, and performance insights.
This article explains the key components running on a Kubernetes master node—including the API Server, etcd, kube‑scheduler, kube‑controller‑manager, and Cloud Provider—detailing their roles, how they interact, and providing practical curl and kubectl commands for common operations.
The mixed version proxy, introduced as an alpha feature in Kubernetes 1.28, enables correct handling of HTTP requests across clusters with multiple API server versions by routing requests to a compatible peer, preventing 404 errors during upgrades and providing flags for configuration and security.
OpenAI’s engineering team details how they expanded a Kubernetes cluster to 7,500 nodes to support massive models like GPT‑3, CLIP, and DALL·E, describing workload characteristics, networking redesign, API server pressure, monitoring, health checks, resource quotas, and the remaining open problems.
The article explains how to enable Kubernetes audit logging and use its detailed fields—such as userAgent, responseStatus, requestURI, and object references—to detect CDK‑generated attacks and other threats like CVE‑2022‑3172, privilege escalation, and backdoor deployment, offering practical detection examples and security recommendations.
This article explains the key Kubernetes control‑plane components—including etcd, the API Server, Controller Manager, Scheduler, as well as worker‑node components like Kubelet, kube‑proxy, and the container runtime—detailing their roles, interactions, and the underlying mechanisms such as Raft consensus and admission control.
This article explains the core components of the Kubernetes control plane—including etcd, the API Server, Controller Manager, Scheduler—as well as key worker‑node components like Kubelet, kube‑proxy, and the container runtime, detailing their roles, interactions, and essential functions.
This article details the architectural background, challenges, and practical solutions implemented in SOFAStack CAFE for cloud‑native multi‑cluster deployment, including a custom multi‑topology CRD, an independent federation API server, enhanced KubeFed controller features, and network proxy integration to support hybrid‑cloud scenarios.
This guide explains the concepts of apiserver clients, authentication methods, service accounts, RBAC authorization, admission control, and provides step‑by‑step instructions for installing and accessing the Kubernetes Dashboard 2.0.
This article walks through building a simple Go web service, containerizing it, pushing the image to a registry, and then exploring how Kubernetes API Server authenticates clients with mutual TLS and how the scheduler’s pre‑filter and scoring phases decide the optimal node for a pod, complete with code snippets and log analysis.
This article details how Alibaba and Ant Financial tackled the performance and stability challenges of running Kubernetes at massive scale, describing their enhancements to etcd, API Server, scheduler, controller failover, load balancing, and other components that enabled a 10k‑node cluster to support the 2019 Tmall 618 promotion.
This article details Alibaba's experience deploying Kubernetes at massive scale, describing the performance bottlenecks encountered in etcd, API server, controller, and scheduler components, and presenting the concrete engineering improvements—such as storage sharding, lease‑based heartbeats, load‑balancing, watch bookmarks, and hot‑standby controllers—that enabled stable operation of clusters with tens of thousands of nodes.
