FailureReason Trace System: Node Metrics node metrics kubelet metrics daemonset metrics node load slo metrics csi metrics dirty data With huge amount of metrics data collected, statistical methods Besides, node delivery capacity can also be evaluated via historical data. With dirty data metrics which consists of • escaped/zombie/uninterruptible process • orphaned containers • orphaned pod data or alert cluster admins to process dirty data manually. Unhealthy node SLO Trace NPD Metrics Event Source …… runtimeErrorCo ntoller failedPodContr oller Detector Strategy Unhealthy node

proxy cluster destinationrule:
 proxy-helloworld 10 90 Think in Cloud . 北北京 监控系统 Kube-State-Metrics Kubernetes APIServer cAdvisor Node-Exporter KUN-Agent Grafana Blackbox Exporter Prometheus 外部探测 Probe Exporter AZ 2 (Local File) Alert Manager gossip Kubernetes 互相监控 AZ 1 Kube-State-Metrics Kubernetes APIServer cAdvisor Node-Exporter KUN-Agent Alert Manager Prometheus (Local File) K8s 集群中，使⽤用 HostPath 存储数据； • Metrics 采集： A. 采集 apiserver、controller-manager、scheduler、etcd、kube-proxy、Kubelet 等组件提供的 metrics B. Kubelet ⾃自带的 cAdvisor 采集容器器 Metrics C. 每个 Node 上以 DaemonSet 的形式部署

先提前下载以下2个镜像（具体版本根据实际使用的配置yaml文件来决定） docker.io/kubernetesui/dashboard:v2.6.0 docker.io/kubernetesui/metrics-scraper:v1.0.8 # wget h�ps://raw.githubusercontent.com/kubernetes/dashboard/v2.6.0/aio/deploy/recom grafana/grafana:8.3.3 kubeStateMetrics-deployment.yaml: image: k8s.gcr.io/kube-state-metrics/kube- state-metrics:v2.3.0 kubeStateMetrics-deployment.yaml: image: quay.io/brancz/kube-rbac- proxy:v0 k8s.gcr.io/kube-state-metrics/kube-state-metrics:v2.3.0 docker pull k8s.gcr.io/prometheus-adapter/prometheus-adapter:v0.9.1 #替换为： docker pull docker.io/lbbi/kube-state-metrics:v2.3.0 docker pull

CVEs kubectl apply RBAC Project Project Project Replic ation Developer Structured Data Metrics Alerts Events VMware vRealize Operations Capacity, Performance and Configuration Management Advanced Analytics Engine Metrics Collection and Storage Iterate & Troubleshoot Issues Trend & Alert on Anomalies Visualize Metrics at Scale Self-Service Metrics Analytics for All Engineering Engineering & Business Wavefront By VMware SaaS-Based Metrics Monitoring and Analytics Platform Constant Compatibility Between PKS and Google Container Engine (GKE) 33 CONFIDENTIAL Google Container

Scale Horizontal Pod Autoscaler Horizontal Pod Autoscaler 获取metrics CPU Mem已经支持指标自定 义metrics(用户自定义指标) 获取metrics CPU Mem已经支持指标自定 义metrics(用户自定义指标) metric上报 metric上报 • 基于有状态服务链路的扩缩容 Redis后端管理平台 应用1

application-aware scaling of the Operand Monitoring • Operator exposing metrics about its health • Operator exposes health and performance metrics about the Operand Alerting and Events • Operand sends useful heal unhealthy Operands based on Operand metrics/alerts/logs • Operator can prevent the Operand from transitioning into an unhealthy state based on Operand metrics Auto-tuning • Operator is able to automatically

Pod Get Pod status Service/Ingress DNS Entry SLB Private Zone ASK-Scheduler K8S API Server Metrics API CloudMonitor, Prometheus HPA Controller Cloud Controller Manager Serverless Scheduler 优的调度效率。 • 支持多可用区、多规格调度。 • 支持批量调度。 • AZ亲和性和反亲和性。 • ECI Pod生命周期管理：直通ECI Pod（Pod状态查询、logs、exec、metrics链路），降低ECI管控压力。 Elastic Scalling ECI ASK-Scheduler K8S API Server • Pod（N） : Node（1） • 单集群支持1万Pod Pod 的基本调度单位和标准、开放的API接口 ECI ASK ACK 云上k8s集群 线下k8s集群 Creae/Delete/Update/Describe/Logs/Exec/Metrics ECI关键技术选择 - 基于安全沙箱技术的容器运行时 ECI Elastic Container Instance Pod container agent Container Container

/proxy/namespaces/kube-system/services/heapster/api/v1/model/namespaces/de fault/pods/cpu-demo/metrics/cpu/usage_rate 即可看到监控信息。 在本例中，尽管容器启动时，尝试使⽤2个CPU单位，但由于配置了只允许使⽤1个CPU单位，因此，最终最多只能使 ⽤1个CPU单位。 CPU单位 oxy/namespaces/kube-system/services/heapster/api/v1/model/namespaces/de fault/pods/memory-demo/metrics/memory/usage 可看到如下结果： { "timestamp": "2017-06-20T18:54:00Z", "value": 162856960 } 由结果可知，Pod正使⽤⼤约162 Heapster。 需在集群上部署Heapster并在kube-system这个namespace中运⾏。 有关使⽤REST客户端访问的详细信息，请参阅 Support for custom metrics 。 autoscaler通过 scale sub-resource访问相应的Replication Controller、Deployment或ReplicaSet。Scale是⼀个允许您

eleases In-progress work ● Migrating from ng1 to ng2 (#3152) ● Migrating metrics from Heapster to Kubernetes Metrics API (#2986) ● Apps list page (#2980) Demo Future of Dashboard How do people

Controller Scaling Controller Rollout Controller kubectl apply Workload Controller Kubernetes metrics traffic Workloads (YAML) Continuous Delivery is in k8s now! code 三者结合呢？ • 基于 CUE 的客户端抽象 • Kubernetes traffic code Raw k8s API resources 面向应用开发者的 appfile • 基于 CUE 进行抽象 • 兼容 OAM Spec metrics Deployment Controller 持续交付 KubeVela “The Extensible Application Platform Based on Kubernetes