k8s学习 — （运维）第十章 k8s 集群监控_kubernetes 应用,服务验活监控

1.1 Heapster

Heapster 是容器集群监控和性能分析工具，天然的支持Kubernetes 和 CoreOS。

Kubernetes 有个出名的监控 agent—cAdvisor。在每个kubernetes Node 上都会运行 cAdvisor，它会收集本机以及容器的监控数据(cpu,memory,filesystem,network,uptime)。
在较新的版本中，K8S 已经将 cAdvisor 功能集成到 kubelet 组件中。每个 Node 节点可以直接进行 web 访问。

1.2 Weave Scope

Weave Scope 可以监控 kubernetes 集群中的一系列资源的状态、资源使用情况、应用拓扑、scale、还可以直接通过浏览器进入容器内部调试等，其提供的功能包括：

1. 交互式拓扑界面
2. 图形模式和表格模式
3. 过滤功能
4. 搜索功能
5. 实时度量
6. 容器排错
7. 插件扩展

1.3 Prometheus

Prometheus 是一套开源的监控系统、报警、时间序列的集合，最初由 SoundCloud 开发，后来随着越来越多公司的使用，于是便独立成开源项目。自此以后，许多公司和组织都采用了 Prometheus 作为监控告警工具。

2 Prometheus 监控 k8s

2.1 自定义配置

2.1.1 创建 ConfigMap 配置

创建 prometheus-config.yml

apiVersion: v1
kind: ConfigMap
metadata:
  name: prometheus-config
data:
  prometheus.yml: |
 global:
 scrape\_interval: 15s 
 evaluation\_interval: 15s
 scrape\_configs:
 - job\_name: 'prometheus'
 static\_configs:
 - targets: ['localhost:9090']

创建 configmap

kubectl create -f prometheus-config.yml

2.1.2 部署 Prometheus

创建 prometheus-deploy.yml

apiVersion: v1
kind: Service
metadata:
  name: prometheus
  labels:
    name: prometheus
spec:
  ports:
  - name: prometheus
    protocol: TCP
    port: 9090
    targetPort: 9090
  selector:
    app: prometheus
  type: NodePort
---
apiVersion: apps/v1
kind: Deployment
metadata:
  labels:
    name: prometheus
  name: prometheus
spec:
  replicas: 1
  selector:
    matchLabels:
      app: prometheus
  template:
    metadata:
      labels:
        app: prometheus
    spec:
      containers:
      - name: prometheus
        image: prom/prometheus:v2.2.1
        command:
        - "/bin/prometheus"
        args:
        - "--config.file=/etc/prometheus/prometheus.yml"
        ports:
        - containerPort: 9090
          protocol: TCP
        volumeMounts:
        - mountPath: "/etc/prometheus"
          name: prometheus-config
      volumes:
      - name: prometheus-config
        configMap:
          name: prometheus-config

创建部署对象

kubectl create -f prometheus-deploy.yml

查看是否在运行中

kubectl get pods -l app=prometheus

获取服务信息

kubectl get svc -l name=prometheus

通过 http://节点ip:端口 进行访问

2.1.3 配置访问权限

创建 prometheus-rbac-setup.yml

apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRole
metadata:
  name: prometheus
rules:
- apiGroups: [""]
  resources:
  - nodes
  - nodes/proxy
  - services
  - endpoints
  - pods
  verbs: ["get", "list", "watch"]
- apiGroups:
  - extensions
  resources:
  - ingresses
  verbs: ["get", "list", "watch"]
- nonResourceURLs: ["/metrics"]
  verbs: ["get"]
---
apiVersion: v1
kind: ServiceAccount
metadata:
  name: prometheus
  namespace: default
---
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRoleBinding
metadata:
  name: prometheus
roleRef:
  apiGroup: rbac.authorization.k8s.io
  kind: ClusterRole
  name: prometheus
subjects:
- kind: ServiceAccount
  name: prometheus
  namespace: default

创建资源对象

kubectl create -f prometheus-rbac-setup.yml

修改 prometheus-deploy.yml 配置文件

spec:
  replicas: 1
  template:
    metadata:
      labels:
        app: prometheus
    spec:
      serviceAccountName: prometheus
      serviceAccount: prometheus

升级 prometheus-deployment

kubectl apply -f prometheus-deployment.yml

查看 pod

kubectl get pods -l app=prometheus

查看 serviceaccount 认证证书

kubectl exec -it <pod name> -- ls /var/run/secrets/kubernetes.io/serviceaccount/

2.1.4 服务发现配置

# 配置 job，帮助 prometheus 找到所有节点信息，修改 prometheus-config.yml 增加为如下内容
data:
  prometheus.yml: |
 global:
 scrape\_interval: 15s
 evaluation\_interval: 15s
 scrape\_configs:
 - job\_name: 'prometheus'
 static\_configs:
 - targets: ['localhost:9090']
 - job\_name: 'kubernetes-nodes'
 tls\_config:
 ca\_file: /var/run/secrets/kubernetes.io/serviceaccount/ca.crt
 bearer\_token\_file: /var/run/secrets/kubernetes.io/serviceaccount/token
 kubernetes\_sd\_configs:
 - role: node

      - job\_name: 'kubernetes-service'
        tls\_config:
          ca\_file: /var/run/secrets/kubernetes.io/serviceaccount/ca.crt
        bearer\_token\_file: /var/run/secrets/kubernetes.io/serviceaccount/token
        kubernetes\_sd\_configs:
        - role: service

      - job\_name: 'kubernetes-endpoints'
        tls\_config:
          ca\_file: /var/run/secrets/kubernetes.io/serviceaccount/ca.crt
        bearer\_token\_file: /var/run/secrets/kubernetes.io/serviceaccount/token
        kubernetes\_sd\_configs:
        - role: endpoints

      - job\_name: 'kubernetes-ingress'
        tls\_config:
          ca\_file: /var/run/secrets/kubernetes.io/serviceaccount/ca.crt
        bearer\_token\_file: /var/run/secrets/kubernetes.io/serviceaccount/token
        kubernetes\_sd\_configs:
        - role: ingress

      - job\_name: 'kubernetes-pods'
        tls\_config:
          ca\_file: /var/run/secrets/kubernetes.io/serviceaccount/ca.crt
        bearer\_token\_file: /var/run/secrets/kubernetes.io/serviceaccount/token
        kubernetes\_sd\_configs:
        - role: pod

升级配置

kubectl apply -f prometheus-config.yml

获取 prometheus pod

kubectl get pods -l app=prometheus

删除 pod

kubectl delete pods <pod name>

查看 pod 状态

kubectl get pods

重新访问 ui 界面

2.1.5 系统时间同步

查看系统时间

date

同步网络时间

ntpdate cn.pool.ntp.org

2.1.6 监控 k8s 集群

# 往 prometheus-config.yml 中追加如下配置
      - job\_name: 'kubernetes-kubelet'
        scheme: https
        tls\_config:
          ca\_file: /var/run/secrets/kubernetes.io/serviceaccount/ca.crt
        bearer\_token\_file: /var/run/secrets/kubernetes.io/serviceaccount/token
        kubernetes\_sd\_configs:
        - role: node
        relabel\_configs:
        - action: labelmap
          regex: __meta_kubernetes_node_label_(.+)
        - target\_label: __address__
          replacement: kubernetes.default.svc:443
        - source\_labels: [__meta_kubernetes_node_name]
          regex: (.+)
          target\_label: __metrics_path__
          replacement: /api/v1/nodes/${1}/proxy/metrics

升级资源

kubectl apply -f prometheus-config.yml

重新构建应用

kubectl delete pods <pod name>

利用指标获取当前节点中 pod 的启动时间

kubelet_pod_start_latency_microseconds{quantile="0.99"}

计算平均时间

kubelet_pod_start_latency_microseconds_sum / kubelet_pod_start_latency_microseconds_count

2.1.6.1 从 kubelet 获取节点容器资源使用情况

# 修改配置文件，增加如下内容，并更新服务
        - job\_name: 'kubernetes-cadvisor'
          scheme: https
          tls\_config:
            ca\_file: /var/run/secrets/kubernetes.io/serviceaccount/ca.crt
          bearer\_token\_file: /var/run/secrets/kubernetes.io/serviceaccount/token
          kubernetes\_sd\_configs:
          - role: node
          relabel\_configs:
          - target\_label: __address__
            replacement: kubernetes.default.svc:443
          - source\_labels: [__meta_kubernetes_node_name]
            regex: (.+)
            target\_label: __metrics_path__
            replacement: /api/v1/nodes/${1}/proxy/metrics/cadvisor
          - action: labelmap
            regex: __meta_kubernetes_node_label_(.+)

2.1.6.2 Exporter 监控资源使用情况

# 创建 node-exporter-daemonset.yml 文件
apiVersion: apps/v1
kind: DaemonSet
metadata:
  name: node-exporter
spec:
  template:
    metadata:
      annotations:
        prometheus.io/scrape: 'true'
        prometheus.io/port: '9100'
        prometheus.io/path: 'metrics'
      labels:
        app: node-exporter
      name: node-exporter
    spec:
      containers:
      - image: prom/node-exporter
        imagePullPolicy: IfNotPresent
        name: node-exporter
        ports:
        - containerPort: 9100
          hostPort: 9100
          name: scrape
      hostNetwork: true
      hostPID: true

创建 daemonset

kubectl create -f node-exporter-daemonset.yml

查看 daemonset 运行状态

kubectl get daemonsets -l app=node-exporter

查看 pod 状态

kubectl get pods -l app=node-exporter

# 修改配置文件，增加监控采集任务
      - job\_name: 'kubernetes-pods'
        kubernetes\_sd\_configs:
        - role: pod
        relabel\_configs:
        - source\_labels: [__meta_kubernetes_pod_annotation_prometheus_io_scrape]
          action: keep
          regex: true
        - source\_labels: [__meta_kubernetes_pod_annotation_prometheus_io_path]
          action: replace
          target\_label: __metrics_path__
          regex: (.+)
        - source\_labels: [__address__, __meta_kubernetes_pod_annotation_prometheus_io_port]
          action: replace
          regex: ([^:]+)(?::\d+)?;(\d+)
          replacement: $1:$2
          target\_label: __address__
        - action: labelmap
          regex: __meta_kubernetes_pod_label_(.+)
        - source\_labels: [__meta_kubernetes_namespace]
          action: replace
          target\_label: kubernetes_namespace
        - source\_labels: [__meta_kubernetes_pod_name]
          action: replace
          target\_label: kubernetes_pod_name

# 通过监控 apiserver 来监控所有对应的入口请求，增加 api-server 监控配置
      - job\_name: 'kubernetes-apiservers'
        kubernetes\_sd\_configs:
        - role: endpoints
        scheme: https
        tls\_config:
          ca\_file: /var/run/secrets/kubernetes.io/serviceaccount/ca.crt
        bearer\_token\_file: /var/run/secrets/kubernetes.io/serviceaccount/token
        relabel\_configs:
        - source\_labels: [__meta_kubernetes_namespace, __meta_kubernetes_service_name, __meta_kubernetes_endpoint_port_name]
          action: keep
          regex: default;kubernetes;https
        - target\_label: __address__
          replacement: kubernetes.default.svc:443

2.1.6.3 对 Ingress 和 Service 进行网络探测

# 创建 blackbox-exporter.yaml 进行网络探测
apiVersion: v1
kind: Service
metadata:
  labels:
    app: blackbox-exporter
  name: blackbox-exporter
spec:
  ports:
  - name: blackbox
    port: 9115
    protocol: TCP
  selector:
    app: blackbox-exporter
  type: ClusterIP
---
apiVersion: apps/v1
kind: Deployment
metadata:
  labels:
    app: blackbox-exporter
  name: blackbox-exporter
spec:
  replicas: 1
  selector:
    matchLabels:
      app: blackbox-exporter
  template:
    metadata:
      labels:
        app: blackbox-exporter
    spec:
      containers:
      - image: prom/blackbox-exporter
        imagePullPolicy: IfNotPresent
        name: blackbox-exporter

创建资源对象

kubectl -f blackbox-exporter.yaml

# 配置监控采集所有 service/ingress 信息，加入配置到配置文件
    - job\_name: 'kubernetes-services'
      metrics\_path: /probe
      params:
        module: [http_2xx]
      kubernetes\_sd\_configs:
      - role: service
      relabel\_configs:
      - source\_labels: [__meta_kubernetes_service_annotation_prometheus_io_probe]
        action: keep
        regex: true
      - source\_labels: [__address__]
        target\_label: __param_target
      - target\_label: __address__
        replacement: blackbox-exporter.default.svc.cluster.local:9115
      - source\_labels: [__param_target]
        target\_label: instance
      - action: labelmap
        regex: __meta_kubernetes_service_label_(.+)
      - source\_labels: [__meta_kubernetes_namespace]
        target\_label: kubernetes_namespace
      - source\_labels: [__meta_kubernetes_service_name]
        target\_label: kubernetes_name

      - job\_name: 'kubernetes-ingresses'
        metrics\_path: /probe
        params:
          module: [http_2xx]
        kubernetes\_sd\_configs:
        - role: ingress
        relabel\_configs:
        - source\_labels: [__meta_kubernetes_ingress_annotation_prometheus_io_probe]
          action: keep
          regex: true
        - source\_labels: [__meta_kubernetes_ingress_scheme,__address__,__meta_kubernetes_ingress_path]
          regex: (.+);(.+);(.+)
          replacement: ${1}://${2}${3}
          target\_label: __param_target
        - target\_label: __address__
          replacement: blackbox-exporter.default.svc.cluster.local:9115
        - source\_labels: [__param_target]
          target\_label: instance
        - action: labelmap
          regex: __meta_kubernetes_ingress_label_(.+)
        - source\_labels: [__meta_kubernetes_namespace]
          target\_label: kubernetes_namespace
        - source\_labels: [__meta_kubernetes_ingress_name]
          target\_label: kubernetes_name

2.1.7 Grafana 可视化

Grafana 是一个通用的可视化工具。‘通用’意味着 Grafana 不仅仅适用于展示 Prometheus 下的监控数据，也同样适用于一些其他的数据可视化需求。在开始使用Grafana之前，我们首先需要明确一些 Grafana下的基本概念，以帮助用户能够快速理解Grafana。

2.1.7.1 基本概念