https://socake.github.io/tags/%E5%BC%B9%E6%80%A7%E4%BC%B8%E7%BC%A9/Recent content in 弹性伸缩 on 黄文卓 | DevOps EngineerHugo -- gohugo.iozh-CN17691281867@163.com (Wenzhuo Huang)17691281867@163.com (Wenzhuo Huang)© 2026 Wenzhuo HuangTue, 09 Dec 2025 10:00:00 +0800https://socake.github.io/docs/kubernetes/k8s-hpa%E5%BC%B9%E6%80%A7%E4%BC%B8%E7%BC%A9/Tue, 09 Dec 2025 10:00:00 +080017691281867@163.com (Wenzhuo Huang)https://socake.github.io/docs/kubernetes/k8s-hpa%E5%BC%B9%E6%80%A7%E4%BC%B8%E7%BC%A9/从 HPA v2 到 KEDA 事件驱动伸缩，覆盖 CPU/内存/自定义指标配置、防抖参数调优、VPA 推荐器集成和生产级弹性伸缩最佳实践。https://socake.github.io/docs/kubernetes/karpenter-%E5%BC%B9%E6%80%A7%E8%8A%82%E7%82%B9/Mon, 08 Dec 2025 13:00:00 +080017691281867@163.com (Wenzhuo Huang)https://socake.github.io/docs/kubernetes/karpenter-%E5%BC%B9%E6%80%A7%E8%8A%82%E7%82%B9/Karpenter 替代 Cluster Autoscaler 的完整实践：NodePool 约束配置、EC2NodeClass 实例选型、consolidation 节点整合降本、Spot 实例容错，以及多套集群配置的组织方式。https://socake.github.io/posts/karpenter-deep-dive/Wed, 11 Jun 2025 11:33:00 +080017691281867@163.com (Wenzhuo Huang)https://socake.github.io/posts/karpenter-deep-dive/从 Cluster Autoscaler 迁移到 Karpenter 之后，集群扩容速度和节点利用率都有明显提升。本文详细拆解 Karpenter 的核心机制、关键配置项，以及在多套生产集群运行中踩过的坑。https://socake.github.io/posts/keda-event-driven-autoscaling/Sat, 08 Feb 2025 10:12:00 +080017691281867@163.com (Wenzhuo Huang)https://socake.github.io/posts/keda-event-driven-autoscaling/HPA 只能看 CPU/内存，但生产环境真正的扩缩信号往往是 Kafka lag、RabbitMQ 队列深度、Prometheus 自定义指标、甚至 cron。本文把 KEDA 的架构、核心 CRD、常见 scaler 的坑和运维动作写成一份资深工程师的备忘录，不讲理论，只讲什么样的配置能在凌晨 3 点把你从告警里救出来。