本文介绍在k8s集群中使用node-exporter、prometheus、grafana对集群进行监控。
其实现原理有点类似ELK、EFK组合。node-exporter组件负责收集节点上的metrics监控数据，并将数据推送给prometheus, prometheus负责存储这些数据，grafana将这些数据通过网页以图形的形式展现给用户。

在开始之前有必要了解下Prometheus是什么?
Prometheus （中文名：普罗米修斯）是由 SoundCloud 开发的开源监控报警系统和时序列数据库(TSDB).自2012年起,许多公司及组织已经采用 Prometheus,并且该项目有着非常活跃的开发者和用户社区.现在已经成为一个独立的开源项目。Prometheus 在2016加入 CNCF ( Cloud Native Computing Foundation ), 作为在 kubernetes 之后的第二个由基金会主持的项目。 Prometheus 的实现参考了Google内部的监控实现，与源自Google的Kubernetes结合起来非常合适。另外相比influxdb的方案，性能更加突出，而且还内置了报警功能。它针对大规模的集群环境设计了拉取式的数据采集方式，只需要在应用里面实现一个metrics接口，然后把这个接口告诉Prometheus就可以完成数据采集了，下图为prometheus的架构图。

Prometheus的特点：
1、多维数据模型（时序列数据由metric名和一组key/value组成）
2、在多维度上灵活的查询语言(PromQl)
3、不依赖分布式存储，单主节点工作.
4、通过基于HTTP的pull方式采集时序数据
5、可以通过中间网关进行时序列数据推送(pushing)
6、目标服务器可以通过发现服务或者静态配置实现
7、多种可视化和仪表盘支持

prometheus 相关组件，Prometheus生态系统由多个组件组成，其中许多是可选的：
1、Prometheus 主服务,用来抓取和存储时序数据
2、client library 用来构造应用或 exporter 代码 (go,java,python,ruby)
3、push 网关可用来支持短连接任务
4、可视化的dashboard (两种选择,promdash 和 grafana.目前主流选择是 grafana.)
4、一些特殊需求的数据出口(用于HAProxy, StatsD, Graphite等服务)
5、实验性的报警管理端(alartmanager,单独进行报警汇总,分发,屏蔽等 )

promethues 的各个组件基本都是用 golang 编写,对编译和部署十分友好.并且没有特殊依赖.基本都是独立工作。
上述文字来自网络！

现在我们正式开始部署工作。
一、环境介绍
操作系统环境：centos linux 7.2 64bit
K8S软件版本： 1.9.0（采用kubeadm方式部署）
Master节点IP： 192.168.115.5/24
Node节点IP： 192.168.115.6/24

二、在k8s集群的所有节点上下载所需要的image

# docker pull prom/node-exporter
# docker pull prom/prometheus:v2.0.0
# docker pull grafana/grafana:4.2.0

三、采用daemonset方式部署node-exporter组件

# cat node-exporter.yaml
---
apiVersion: extensions/v1beta1
kind: DaemonSet
metadata:name: node-exporternamespace: kube-systemlabels:k8s-app: node-exporter
spec:template:metadata:labels:k8s-app: node-exporterspec:containers:- image: prom/node-exportername: node-exporterports:- containerPort: 9100protocol: TCPname: http
---
apiVersion: v1
kind: Service
metadata:labels:k8s-app: node-exportername: node-exporternamespace: kube-system
spec:ports:- name: httpport: 9100nodePort: 31672protocol: TCPtype: NodePortselector:k8s-app: node-exporter

通过上述文件创建pod和service

# kubectl create -f  node-exporter.yaml 

四、部署prometheus组件
1、rbac文件

# cat rbac-setup.yaml
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRole
metadata:name: prometheus
rules:
- apiGroups: [""]resources:- nodes- nodes/proxy- services- endpoints- podsverbs: ["get", "list", "watch"]
- apiGroups:- extensionsresources:- ingressesverbs: ["get", "list", "watch"]
- nonResourceURLs: ["/metrics"]verbs: ["get"]
---
apiVersion: v1
kind: ServiceAccount
metadata:name: prometheusnamespace: kube-system
---
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRoleBinding
metadata:name: prometheus
roleRef:apiGroup: rbac.authorization.k8s.iokind: ClusterRolename: prometheus
subjects:
- kind: ServiceAccountname: prometheusnamespace: kube-system

2、以configmap的形式管理prometheus组件的配置文件

# cat configmap.yaml
apiVersion: v1
kind: ConfigMap
metadata:name: prometheus-confignamespace: kube-system
data:prometheus.yml: |global:scrape_interval:     15sevaluation_interval: 15sscrape_configs:- job_name: 'kubernetes-apiservers'kubernetes_sd_configs:- role: endpointsscheme: httpstls_config:ca_file: /var/run/secrets/kubernetes.io/serviceaccount/ca.crtbearer_token_file: /var/run/secrets/kubernetes.io/serviceaccount/tokenrelabel_configs:- source_labels: [__meta_kubernetes_namespace, __meta_kubernetes_service_name, __meta_kubernetes_endpoint_port_name]action: keepregex: default;kubernetes;https- job_name: 'kubernetes-nodes'kubernetes_sd_configs:- role: nodescheme: httpstls_config:ca_file: /var/run/secrets/kubernetes.io/serviceaccount/ca.crtbearer_token_file: /var/run/secrets/kubernetes.io/serviceaccount/tokenrelabel_configs:- action: labelmapregex: __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/${1}/proxy/metrics- job_name: 'kubernetes-cadvisor'kubernetes_sd_configs:- role: nodescheme: httpstls_config:ca_file: /var/run/secrets/kubernetes.io/serviceaccount/ca.crtbearer_token_file: /var/run/secrets/kubernetes.io/serviceaccount/tokenrelabel_configs:- action: labelmapregex: __meta_kubernetes_node_label_(.+)- target_label: __address__replacement: kubernetes.default.svc:443- source_labels: [__meta_kubernetes_node_name]regex: (.+)target_label: __metrics_path__replacement: /api/v1/nodes/${1}/proxy/metrics/cadvisor- job_name: 'kubernetes-service-endpoints'kubernetes_sd_configs:- role: endpointsrelabel_configs:- source_labels: [__meta_kubernetes_service_annotation_prometheus_io_scrape]action: keepregex: true- source_labels: [__meta_kubernetes_service_annotation_prometheus_io_scheme]action: replacetarget_label: __scheme__regex: (https?)- source_labels: [__meta_kubernetes_service_annotation_prometheus_io_path]action: replacetarget_label: __metrics_path__regex: (.+)- source_labels: [__address__, __meta_kubernetes_service_annotation_prometheus_io_port]action: replacetarget_label: __address__regex: ([^:]+)(?::\d+)?;(\d+)replacement: $1:$2- action: labelmapregex: __meta_kubernetes_service_label_(.+)- source_labels: [__meta_kubernetes_namespace]action: replacetarget_label: kubernetes_namespace- source_labels: [__meta_kubernetes_service_name]action: replacetarget_label: kubernetes_name- job_name: 'kubernetes-services'kubernetes_sd_configs:- role: servicemetrics_path: /probeparams:module: [http_2xx]relabel_configs:- source_labels: [__meta_kubernetes_service_annotation_prometheus_io_probe]action: keepregex: true- source_labels: [__address__]target_label: __param_target- target_label: __address__replacement: blackbox-exporter.example.com:9115- source_labels: [__param_target]target_label: instance- action: labelmapregex: __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'kubernetes_sd_configs:- role: ingressrelabel_configs:- source_labels: [__meta_kubernetes_ingress_annotation_prometheus_io_probe]action: keepregex: 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.example.com:9115- source_labels: [__param_target]target_label: instance- action: labelmapregex: __meta_kubernetes_ingress_label_(.+)- source_labels: [__meta_kubernetes_namespace]target_label: kubernetes_namespace- source_labels: [__meta_kubernetes_ingress_name]target_label: kubernetes_name- job_name: 'kubernetes-pods'kubernetes_sd_configs:- role: podrelabel_configs:- source_labels: [__meta_kubernetes_pod_annotation_prometheus_io_scrape]action: keepregex: true- source_labels: [__meta_kubernetes_pod_annotation_prometheus_io_path]action: replacetarget_label: __metrics_path__regex: (.+)- source_labels: [__address__, __meta_kubernetes_pod_annotation_prometheus_io_port]action: replaceregex: ([^:]+)(?::\d+)?;(\d+)replacement: $1:$2target_label: __address__- action: labelmapregex: __meta_kubernetes_pod_label_(.+)- source_labels: [__meta_kubernetes_namespace]action: replacetarget_label: kubernetes_namespace- source_labels: [__meta_kubernetes_pod_name]action: replacetarget_label: kubernetes_pod_name

3、Prometheus deployment 文件

# cat prometheus.deploy.yml
---
apiVersion: apps/v1beta2
kind: Deployment
metadata:labels:name: prometheus-deploymentname: prometheusnamespace: kube-system
spec:replicas: 1selector:matchLabels:app: prometheustemplate:metadata:labels:app: prometheusspec:containers:- image: prom/prometheus:v2.0.0name: prometheuscommand:- "/bin/prometheus"args:- "--config.file=/etc/prometheus/prometheus.yml"- "--storage.tsdb.path=/prometheus"- "--storage.tsdb.retention=24h"ports:- containerPort: 9090protocol: TCPvolumeMounts:- mountPath: "/prometheus"name: data- mountPath: "/etc/prometheus"name: config-volumeresources:requests:cpu: 100mmemory: 100Milimits:cpu: 500mmemory: 2500MiserviceAccountName: prometheus    volumes:- name: dataemptyDir: {}- name: config-volumeconfigMap:name: prometheus-config       

4、Prometheus service文件

# cat prometheus.svc.yml
---
kind: Service
apiVersion: v1
metadata:labels:app: prometheusname: prometheusnamespace: kube-system
spec:type: NodePortports:- port: 9090targetPort: 9090nodePort: 30003selector:app: prometheus

5、通过上述yaml文件创建相应的对象

# kubectl create -f  rbac-setup.yaml
# kubectl create -f  configmap.yaml
# kubectl create -f  prometheus.deploy.yml
# kubectl create -f  prometheus.svc.yml 

Node-exporter对应的nodeport端口为31672，通过访问http://192.168.115.5:31672/metrics 可以看到对应的metrics

prometheus对应的nodeport端口为30003，通过访问http://192.168.115.5:30003/target 可以看到prometheus已经成功连接上了k8s的apiserver

可以在prometheus的WEB界面上提供了基本的查询K8S集群中每个POD的CPU使用情况，查询条件如下：

sum by (pod_name)( rate(container_cpu_usage_seconds_total{image!="", pod_name!=""}[1m] ) )

上述的查询有出现数据，说明node-exporter往prometheus中写入数据正常，接下来我们就可以部署grafana组件，实现更友好的webui展示数据了。

五、部署grafana组件
1、grafana deployment配置文件

# cat grafana-deploy.yaml
apiVersion: extensions/v1beta1
kind: Deployment
metadata:name: grafana-corenamespace: kube-systemlabels:app: grafanacomponent: core
spec:replicas: 1template:metadata:labels:app: grafanacomponent: corespec:containers:- image: grafana/grafana:4.2.0name: grafana-coreimagePullPolicy: IfNotPresent# env:resources:# keep request = limit to keep this container in guaranteed classlimits:cpu: 100mmemory: 100Mirequests:cpu: 100mmemory: 100Mienv:# The following env variables set up basic auth twith the default admin user and admin password.- name: GF_AUTH_BASIC_ENABLEDvalue: "true"- name: GF_AUTH_ANONYMOUS_ENABLEDvalue: "false"# - name: GF_AUTH_ANONYMOUS_ORG_ROLE#   value: Admin# does not really work, because of template variables in exported dashboards:# - name: GF_DASHBOARDS_JSON_ENABLED#   value: "true"readinessProbe:httpGet:path: /loginport: 3000# initialDelaySeconds: 30# timeoutSeconds: 1volumeMounts:- name: grafana-persistent-storagemountPath: /varvolumes:- name: grafana-persistent-storageemptyDir: {}

2、grafana service配置文件

# cat grafana-svc.yaml
apiVersion: v1
kind: Service
metadata:name: grafananamespace: kube-systemlabels:app: grafanacomponent: core
spec:type: NodePortports:- port: 3000selector:app: grafana
component: core

3、grafana ingress配置文件

# cat grafana-ing.yaml
apiVersion: extensions/v1beta1
kind: Ingress
metadata:name: grafananamespace: kube-system
spec:rules:- host: k8s.grafanahttp:paths:- path: /backend:serviceName: grafanaservicePort: 3000

通过访问traefik的webui可以看到k8s.grafana服务发布成功

修改hosts解析，访问测试


也可以直接访问nodeport端口

默认用户名和密码都是admin

配置数据源为prometheus

导入面板，可以直接输入模板编号315在线导入，或者下载好对应的json模板文件本地导入，面板模板下载地址https://grafana.com/dashboards/315

导入面板之后就可以看到对应的监控数据了。



这里要说明一下，在测试过程中，导入编号为162的模板，发现只有部分数据，且pod的名称显示不友好。模板地址https://grafana.com/dashboards/162，详见下图。

六、后记
这里存在一些问题后续要继续研究解决。
1、prometheus的数据存储采用emptydir。如果Pod被删除，或者Pod发生迁移，emptyDir也会被删除，并且永久丢失。后续可以在K8S集群外部再配置一个Prometheus系统来永久保存监控数据， 两个prometheus系统之间通过配置job自动进行数据拉取。
2、Grafana的配置数据存储采用emptydir。如果Pod被删除，或者Pod发生迁移，emptyDir也会被删除，并且永久丢失。我们也可以选择将grafana配置在k8s外部，数据源选择K8S集群外部的prometheus即可。
3、关于监控项的报警（alertmanager）尚未配置。

参考文档，感谢作者分享！
https://www.kubernetes.org.cn/3418.html
https://blog.qikqiak.com/post/kubernetes-monitor-prometheus-grafana/
https://github.com/giantswarm/kubernetes-prometheus/tree/master/manifests
https://segmentfault.com/a/1190000013245394