使用Prometheus Operator实现应用自定义监控

相关推荐：

使用Prometheus Operator监控kubetnetes集群

一、前言

在之前文章中为大家讲解了如何在Kubernetes集群中部署Prometheus，已经可以通过Prometheus监控Kubernetes中Pod的状态、核心组件的状态等数据。那么如何将应用自身的数据进行集成呢？

首先prometheus对收集的数据格式是有一定的要求的，具体格式如下,只有符合此格式的prometheus才会正常的采集，所以在应用中我们需要能把关键性的监控数据以此格式抛出来。

nginx_http_connections{state="active"} 2nginx_http_connections{state="reading"} 0nginx_http_connections{state="waiting"} 1nginx_http_connections{state="writing"} 1nginx_http_request_bytes_sent{host="10.46.0.4"} 11055968nginx_http_request_bytes_sent{host="testservers"} 4640nginx_http_request_time_bucket{host="10.46.0.4",le="00.005"} 3960...

至于怎么把监控指标数据抛出来，可能需要从应用本身动手。Prometheus社区也提供了大量的官方以及第三方Exporters，可以满足Prometheus的采纳者快速实现对关键业务，以及基础设施的监控需求。
官方以及第三方Exporters请参考此链接。

此处我们建议直接使用Prometheus的Client进行注册监控接口。Promehtues的Client目前支持大部分编程语言，支持列表可以参考如下文章。
prometheus client支持的语言列表。

二、应用监控

本次是使用nginx demo镜像来演示。

1. 创建deployment和service

apiVersion: apps/v1beta1kind: Deploymentmetadata:  name: nginx-demo  labels:    app: nginx-demospec:  replicas: 1  selector:    matchLabels:      app: nginx-demo  template:    metadata:      labels:        app: nginx-demo    spec:      containers:      - name: nginx-demo        image: billy98/nginx-prometheus-metrics:latest        ports:        - name: http-metrics          containerPort: 9527        - name: web          containerPort: 80        - name: test          containerPort: 1314        imagePullPolicy: IfNotPresent---apiVersion: v1kind: Servicemetadata:  labels:    app: nginx-demo  name: nginx-demo  namespace: defaultspec:  ports:  - name: http-metrics    port: 9527    protocol: TCP    targetPort: 9527  - name: web    port: 80    protocol: TCP    targetPort: 80  - name: test    port: 1314    protocol: TCP    targetPort: 1314  selector:    app: nginx-demo  type: ClusterIP 

2. 创建ServiceMonitor

由于prometheus里指定了serviceMonitor的标签必须release: p才会去抓取数据，所以在我们应用的ServiceMonitor中必须加上此标签。

[root@node-01 ~]# kubectl -n monitoring get prometheus p-prometheus -o yaml...  serviceMonitorSelector:    matchLabels:      release: p...(其余内容省略)

apiVersion: monitoring.coreos.com/v1kind: ServiceMonitormetadata:  labels:    app: nginx-demo    release: p  name: nginx-demo  namespace: monitoring  #prometheus的namespacespec:  endpoints:  - interval: 15s    port: http-metrics  namespaceSelector:    matchNames:    - default    #nginx demo的namespace  selector:    matchLabels:      app: nginx-demo

此处需要特别做些解释，serviceMonitor是Prometheus Operator中抽象的概念，他的作用就是讲配置Prometheus采集Target的配置变化成为动态发现的方式，可以serviceMonitor通过Deployment对应的Service配置进行挂钩，通过label selector选择Service，并自动发现后端容器。其中需要注意的是namespace字段永远为monitoring，而namespaceSelector中则是选择的应用所在的namespace。

创建完成可以看到endpoints

[root@node-01 ~]# kubectl get epNAME                       ENDPOINTS                                      AGEheketi                     10.45.0.3:8080                                 36dheketi-storage-endpoints   10.31.90.204:1,10.31.90.205:1,10.31.90.206:1   36dkubernetes                 10.31.90.201:6443                              38dnginx-demo                 10.46.0.4:9527,10.46.0.4:80,10.46.0.4:1314     6htomcat                     10.34.0.0:8080                                 38d

然后我们访问10.46.0.4:1314,生成一些测试指标。

[root@node-01 ~]# curl 10.46.0.4:1314hello world

访问10.46.0.4:9527/metrics,查看监控数据

[root@node-01 ~]# curl 10.46.0.4:9527/metrics# HELP nginx_http_connections Number of HTTP connections# TYPE nginx_http_connections gaugenginx_http_connections{state="active"} 2nginx_http_connections{state="reading"} 0nginx_http_connections{state="waiting"} 1nginx_http_connections{state="writing"} 1# HELP nginx_http_request_bytes_sent Number of HTTP request bytes sent# TYPE nginx_http_request_bytes_sent counternginx_http_request_bytes_sent{host="10.46.0.4"} 16885674nginx_http_request_bytes_sent{host="testservers"} 4800# HELP nginx_http_request_time HTTP request time# TYPE nginx_http_request_time histogramnginx_http_request_time_bucket{host="10.46.0.4",le="00.005"} 4437nginx_http_request_time_bucket{host="10.46.0.4",le="00.010"} 4437nginx_http_request_time_bucket{host="10.46.0.4",le="00.020"} 4437nginx_http_request_time_bucket{host="10.46.0.4",le="00.030"} 4437nginx_http_request_time_bucket{host="10.46.0.4",le="00.050"} 4437nginx_http_request_time_bucket{host="10.46.0.4",le="00.075"} 4437nginx_http_request_time_bucket{host="10.46.0.4",le="00.100"} 4437nginx_http_request_time_bucket{host="10.46.0.4",le="00.200"} 4437nginx_http_request_time_bucket{host="10.46.0.4",le="00.300"} 4437nginx_http_request_time_bucket{host="10.46.0.4",le="00.400"} 4437nginx_http_request_time_bucket{host="10.46.0.4",le="00.500"} 4437nginx_http_request_time_bucket{host="10.46.0.4",le="00.750"} 4437nginx_http_request_time_bucket{host="10.46.0.4",le="01.000"} 4437nginx_http_request_time_bucket{host="10.46.0.4",le="01.500"} 4437nginx_http_request_time_bucket{host="10.46.0.4",le="02.000"} 4437nginx_http_request_time_bucket{host="10.46.0.4",le="03.000"} 4443nginx_http_request_time_bucket{host="10.46.0.4",le="04.000"} 4443nginx_http_request_time_bucket{host="10.46.0.4",le="05.000"} 4443nginx_http_request_time_bucket{host="10.46.0.4",le="10.000"} 4443nginx_http_request_time_bucket{host="10.46.0.4",le="+Inf"} 4443nginx_http_request_time_bucket{host="testservers",le="00.005"} 30nginx_http_request_time_bucket{host="testservers",le="00.010"} 30nginx_http_request_time_bucket{host="testservers",le="00.020"} 30nginx_http_request_time_bucket{host="testservers",le="00.030"} 30nginx_http_request_time_bucket{host="testservers",le="00.050"} 30nginx_http_request_time_bucket{host="testservers",le="00.075"} 30nginx_http_request_time_bucket{host="testservers",le="00.100"} 30nginx_http_request_time_bucket{host="testservers",le="00.200"} 30nginx_http_request_time_bucket{host="testservers",le="00.300"} 30nginx_http_request_time_bucket{host="testservers",le="00.400"} 30nginx_http_request_time_bucket{host="testservers",le="00.500"} 30nginx_http_request_time_bucket{host="testservers",le="00.750"} 30nginx_http_request_time_bucket{host="testservers",le="01.000"} 30nginx_http_request_time_bucket{host="testservers",le="01.500"} 30nginx_http_request_time_bucket{host="testservers",le="02.000"} 30nginx_http_request_time_bucket{host="testservers",le="03.000"} 30nginx_http_request_time_bucket{host="testservers",le="04.000"} 30nginx_http_request_time_bucket{host="testservers",le="05.000"} 30nginx_http_request_time_bucket{host="testservers",le="10.000"} 30nginx_http_request_time_bucket{host="testservers",le="+Inf"} 30nginx_http_request_time_count{host="10.46.0.4"} 4443nginx_http_request_time_count{host="testservers"} 30nginx_http_request_time_sum{host="10.46.0.4"} 12.044999361038nginx_http_request_time_sum{host="testservers"} 0# HELP nginx_http_requests Number of HTTP requests# TYPE nginx_http_requests counternginx_http_requests{host="10.46.0.4",status="200"} 4443nginx_http_requests{host="testservers",status="200"} 30# HELP nginx_http_upstream_bytes_received Number of HTTP upstream bytes received# TYPE nginx_http_upstream_bytes_received counternginx_http_upstream_bytes_received{addr="10.12.13.14:80"} 0nginx_http_upstream_bytes_received{addr="127.0.0.1:80"} 4800# HELP nginx_http_upstream_connect_time HTTP upstream connect time# TYPE nginx_http_upstream_connect_time histogramnginx_http_upstream_connect_time_bucket{addr="10.12.13.14:80",le="03.000"} 6nginx_http_upstream_connect_time_bucket{addr="10.12.13.14:80",le="04.000"} 6nginx_http_upstream_connect_time_bucket{addr="10.12.13.14:80",le="05.000"} 6nginx_http_upstream_connect_time_bucket{addr="10.12.13.14:80",le="10.000"} 6nginx_http_upstream_connect_time_bucket{addr="10.12.13.14:80",le="+Inf"} 6nginx_http_upstream_connect_time_bucket{addr="127.0.0.1:80",le="00.005"} 30nginx_http_upstream_connect_time_bucket{addr="127.0.0.1:80",le="00.010"} 30nginx_http_upstream_connect_time_bucket{addr="127.0.0.1:80",le="00.020"} 30nginx_http_upstream_connect_time_bucket{addr="127.0.0.1:80",le="00.030"} 30nginx_http_upstream_connect_time_bucket{addr="127.0.0.1:80",le="00.050"} 30nginx_http_upstream_connect_time_bucket{addr="127.0.0.1:80",le="00.075"} 30nginx_http_upstream_connect_time_bucket{addr="127.0.0.1:80",le="00.100"} 30nginx_http_upstream_connect_time_bucket{addr="127.0.0.1:80",le="00.200"} 30nginx_http_upstream_connect_time_bucket{addr="127.0.0.1:80",le="00.300"} 30nginx_http_upstream_connect_time_bucket{addr="127.0.0.1:80",le="00.400"} 30nginx_http_upstream_connect_time_bucket{addr="127.0.0.1:80",le="00.500"} 30nginx_http_upstream_connect_time_bucket{addr="127.0.0.1:80",le="00.750"} 30nginx_http_upstream_connect_time_bucket{addr="127.0.0.1:80",le="01.000"} 30nginx_http_upstream_connect_time_bucket{addr="127.0.0.1:80",le="01.500"} 30nginx_http_upstream_connect_time_bucket{addr="127.0.0.1:80",le="02.000"} 30nginx_http_upstream_connect_time_bucket{addr="127.0.0.1:80",le="03.000"} 30nginx_http_upstream_connect_time_bucket{addr="127.0.0.1:80",le="04.000"} 30nginx_http_upstream_connect_time_bucket{addr="127.0.0.1:80",le="05.000"} 30nginx_http_upstream_connect_time_bucket{addr="127.0.0.1:80",le="10.000"} 30nginx_http_upstream_connect_time_bucket{addr="127.0.0.1:80",le="+Inf"} 30nginx_http_upstream_connect_time_count{addr="10.12.13.14:80"} 6nginx_http_upstream_connect_time_count{addr="127.0.0.1:80"} 30nginx_http_upstream_connect_time_sum{addr="10.12.13.14:80"} 12.015nginx_http_upstream_connect_time_sum{addr="127.0.0.1:80"} 0.006# HELP nginx_http_upstream_header_time HTTP upstream header time# TYPE nginx_http_upstream_header_time histogramnginx_http_upstream_header_time_bucket{addr="10.12.13.14:80",le="03.000"} 6nginx_http_upstream_header_time_bucket{addr="10.12.13.14:80",le="04.000"} 6nginx_http_upstream_header_time_bucket{addr="10.12.13.14:80",le="05.000"} 6nginx_http_upstream_header_time_bucket{addr="10.12.13.14:80",le="10.000"} 6nginx_http_upstream_header_time_bucket{addr="10.12.13.14:80",le="+Inf"} 6nginx_http_upstream_header_time_bucket{addr="127.0.0.1:80",le="00.005"} 30nginx_http_upstream_header_time_bucket{addr="127.0.0.1:80",le="00.010"} 30nginx_http_upstream_header_time_bucket{addr="127.0.0.1:80",le="00.020"} 30nginx_http_upstream_header_time_bucket{addr="127.0.0.1:80",le="00.030"} 30nginx_http_upstream_header_time_bucket{addr="127.0.0.1:80",le="00.050"} 30nginx_http_upstream_header_time_bucket{addr="127.0.0.1:80",le="00.075"} 30nginx_http_upstream_header_time_bucket{addr="127.0.0.1:80",le="00.100"} 30nginx_http_upstream_header_time_bucket{addr="127.0.0.1:80",le="00.200"} 30nginx_http_upstream_header_time_bucket{addr="127.0.0.1:80",le="00.300"} 30nginx_http_upstream_header_time_bucket{addr="127.0.0.1:80",le="00.400"} 30nginx_http_upstream_header_time_bucket{addr="127.0.0.1:80",le="00.500"} 30nginx_http_upstream_header_time_bucket{addr="127.0.0.1:80",le="00.750"} 30nginx_http_upstream_header_time_bucket{addr="127.0.0.1:80",le="01.000"} 30nginx_http_upstream_header_time_bucket{addr="127.0.0.1:80",le="01.500"} 30nginx_http_upstream_header_time_bucket{addr="127.0.0.1:80",le="02.000"} 30nginx_http_upstream_header_time_bucket{addr="127.0.0.1:80",le="03.000"} 30nginx_http_upstream_header_time_bucket{addr="127.0.0.1:80",le="04.000"} 30nginx_http_upstream_header_time_bucket{addr="127.0.0.1:80",le="05.000"} 30nginx_http_upstream_header_time_bucket{addr="127.0.0.1:80",le="10.000"} 30nginx_http_upstream_header_time_bucket{addr="127.0.0.1:80",le="+Inf"} 30nginx_http_upstream_header_time_count{addr="10.12.13.14:80"} 6nginx_http_upstream_header_time_count{addr="127.0.0.1:80"} 30nginx_http_upstream_header_time_sum{addr="10.12.13.14:80"} 12.015nginx_http_upstream_header_time_sum{addr="127.0.0.1:80"} 0.03# HELP nginx_http_upstream_requests Number of HTTP upstream requests# TYPE nginx_http_upstream_requests counternginx_http_upstream_requests{addr="10.12.13.14:80",status="504"} 6nginx_http_upstream_requests{addr="127.0.0.1:80",status="200"} 30# HELP nginx_http_upstream_response_time HTTP upstream response time# TYPE nginx_http_upstream_response_time histogramnginx_http_upstream_response_time_bucket{addr="10.12.13.14:80",le="03.000"} 6nginx_http_upstream_response_time_bucket{addr="10.12.13.14:80",le="04.000"} 6nginx_http_upstream_response_time_bucket{addr="10.12.13.14:80",le="05.000"} 6nginx_http_upstream_response_time_bucket{addr="10.12.13.14:80",le="10.000"} 6nginx_http_upstream_response_time_bucket{addr="10.12.13.14:80",le="+Inf"} 6nginx_http_upstream_response_time_bucket{addr="127.0.0.1:80",le="00.005"} 30nginx_http_upstream_response_time_bucket{addr="127.0.0.1:80",le="00.010"} 30nginx_http_upstream_response_time_bucket{addr="127.0.0.1:80",le="00.020"} 30nginx_http_upstream_response_time_bucket{addr="127.0.0.1:80",le="00.030"} 30nginx_http_upstream_response_time_bucket{addr="127.0.0.1:80",le="00.050"} 30nginx_http_upstream_response_time_bucket{addr="127.0.0.1:80",le="00.075"} 30nginx_http_upstream_response_time_bucket{addr="127.0.0.1:80",le="00.100"} 30nginx_http_upstream_response_time_bucket{addr="127.0.0.1:80",le="00.200"} 30nginx_http_upstream_response_time_bucket{addr="127.0.0.1:80",le="00.300"} 30nginx_http_upstream_response_time_bucket{addr="127.0.0.1:80",le="00.400"} 30nginx_http_upstream_response_time_bucket{addr="127.0.0.1:80",le="00.500"} 30nginx_http_upstream_response_time_bucket{addr="127.0.0.1:80",le="00.750"} 30nginx_http_upstream_response_time_bucket{addr="127.0.0.1:80",le="01.000"} 30nginx_http_upstream_response_time_bucket{addr="127.0.0.1:80",le="01.500"} 30nginx_http_upstream_response_time_bucket{addr="127.0.0.1:80",le="02.000"} 30nginx_http_upstream_response_time_bucket{addr="127.0.0.1:80",le="03.000"} 30nginx_http_upstream_response_time_bucket{addr="127.0.0.1:80",le="04.000"} 30nginx_http_upstream_response_time_bucket{addr="127.0.0.1:80",le="05.000"} 30nginx_http_upstream_response_time_bucket{addr="127.0.0.1:80",le="10.000"} 30nginx_http_upstream_response_time_bucket{addr="127.0.0.1:80",le="+Inf"} 30nginx_http_upstream_response_time_count{addr="10.12.13.14:80"} 6nginx_http_upstream_response_time_count{addr="127.0.0.1:80"} 30nginx_http_upstream_response_time_sum{addr="10.12.13.14:80"} 12.015nginx_http_upstream_response_time_sum{addr="127.0.0.1:80"} 0.03# HELP nginx_metric_errors_total Number of nginx-lua-prometheus errors# TYPE nginx_metric_errors_total counternginx_metric_errors_total 0

3. 验证

访问Prometheus，验证数据采集，打开Status下的Service Discovery，active的数目等于Pod数据即表示采集正常

打开Graph页面，选择我们刚才推送的数据指标名称，点击Execute，即可查看到采集上来的数据。prometheus查询语法请参考prometheus查询语句示例。

然后我们在grafana中创建一个简单的图表显示nginx的数据。

最后

使用Prometheus接入应用监控的方式非常简单，整个操作的流程非常kubernetes，这也是目前非常多的开源软件和Kubernetes集成的一种方式与趋势，在开发者习惯了之后，会越来越感受到这种方式的便利。

下一篇会讲到使用邮件或者钉钉报警的相关教程。

如有问题欢迎在下面留言交流。希望大家多多关注和点赞，谢谢！