The article explains why native Elasticsearch throttling is insufficient, introduces node‑level traffic control provided by Infinilabs Gateway, shows detailed configuration examples, parameter meanings, FAQ solutions, advanced tuning tips, and performance comparisons to protect clusters from overload.
After the OpenAI outage caused massive Kubernetes API overload, Alibaba Cloud’s Container Service and Observability teams detail how they reinforce large‑scale K8s clusters with high‑availability control‑plane design, optimized Prometheus probing, out‑of‑band monitoring, and best‑practice guidelines for capacity planning, safe releases, and rapid incident response.
This article explains the stability challenges of large‑scale Kubernetes clusters, outlines ACK's high‑availability architecture and component optimizations, and details product features such as Prometheus, AIOps and managed node pools that together ensure reliable, performant cloud‑native workloads.
The article examines how an improperly set jute.maxbuffer in ZooKeeper can cause prolonged leader elections, server restarts, and high resource usage, explains the underlying code paths, and provides practical detection methods and configuration recommendations to ensure stable cluster operation.
By offloading thousands of short‑lived CronJob pods to Tencent Cloud EKS serverless virtual nodes, Zuoyebang isolated them from online services, eliminated IP waste, achieved millisecond‑level parallel scheduling and sub‑3‑second startup, freed 10 % of cluster resources and cut scheduling costs by roughly 70 % while markedly improving cluster stability.
This article analyzes why a company's Kubernetes clusters were constantly on the brink of failure and presents a comprehensive solution covering CI/CD pipeline reconstruction, federated monitoring with Prometheus, centralized logging via Elasticsearch, documentation centralization, and clarified request routing to achieve high reliability.
This article explains how Kubernetes nodes schedule pods based on total capacity, why lacking resource reservations can cause node failures and cluster avalanches, and provides step‑by‑step guidance on configuring Node Allocatable, kube‑reserved, system‑reserved, and eviction settings to ensure stable cluster operation.
The article explains how excessive HBase partitions can cause frequent flushes, compaction storms, high memory usage, long master assignment times, and reduced MapReduce concurrency, and provides formulas and guidelines for calculating a reasonable number of regions per RegionServer.
This article shares a detailed, experience‑driven guide on Elasticsearch tuning, covering data model fundamentals, storage cost reductions, cluster stability tricks, performance‑boosting settings, and custom kernel improvements, all illustrated with real‑world diagrams and Q&A insights.
