解读Prometheus监控Kubernetes的配置文件

Posted 2021-04-16 DevOps技术栈

来源：https://segmentfault.com/a/1190000013230914

前言

Prometheus 是一个开源和社区驱动的监控&报警&时序数据库的项目。来源于谷歌BorgMon项目。现在最常见的Kubernetes容器管理系统中，通常会搭配Prometheus进行监控。主要监控：

• Node:如主机CPU，内存，网络吞吐和带宽占用，磁盘I/O和磁盘使用等指标。node-exporter采集。

• 容器关键指标:集群中容器的CPU详细状况，内存详细状况，Network，FileSystem和Subcontainer等。通过cadvisor采集。

• Kubernetes集群上部署的应用：监控部署在Kubernetes集群上的应用。主要是pod，service，ingress和endpoint。通过black-box和kube-apiserver的接口采集。

prometheus自身提供了一些资源的自动发现功能，下面是我从官方github上截图，罗列了目前提供的资源发现：

上图可知prometheus自身提供了自动发现kubernetes的监控目标的功能。相应，配置文件官方也提供了一份，今天我们就解读一下该配置文件。

配置文件解读

首先直接上官方的配置文件：

# A scrape configuration for running Prometheus on a Kubernetes cluster. # This uses separate scrape configs for cluster components (i.e. API server, node) # and services to allow each to use different authentication configs. # # Kubernetes labels will be added as Prometheus labels on metrics via the # `labelmap` relabeling action. # # If you are using Kubernetes 1.7.2 or earlier, please take note of the comments # for the kubernetes-cadvisor job; you will need to edit or remove this job. # Scrape config for API servers. # # Kubernetes exposes API servers as endpoints to the default/kubernetes # service so this uses `endpoints` role and uses relabelling to only keep # the endpoints associated with the default/kubernetes service using the # default named port `https`. This works for single API server deployments as # well as HA API server deployments. scrape_configs: - job_name: 'kubernetes-apiservers'  kubernetes_sd_configs:  - role: endpoints  # Default to scraping over https. If required, just disable this or change to  # `http`.  scheme: https  # This TLS & bearer token file config is used to connect to the actual scrape  # endpoints for cluster components. This is separate to discovery auth  # configuration because discovery & scraping are two separate concerns in  # Prometheus. The discovery auth config is automatic if Prometheus runs inside  # the cluster. Otherwise, more config options have to be provided within the  # <kubernetes_sd_config>.  tls_config:    ca_file: /var/run/secrets/kubernetes.io/serviceaccount/ca.crt    # If your node certificates are self-signed or use a different CA to the    # master CA, then disable certificate verification below. Note that    # certificate verification is an integral part of a secure infrastructure    # so this should only be disabled in a controlled environment. You can    # disable certificate verification by uncommenting the line below.    #    # insecure_skip_verify: true  bearer_token_file: /var/run/secrets/kubernetes.io/serviceaccount/token  # Keep only the default/kubernetes service endpoints for the https port. This  # will add targets for each API server which Kubernetes adds an endpoint to  # the default/kubernetes service.  relabel_configs:  - source_labels: [__meta_kubernetes_namespace, __meta_kubernetes_service_name, __meta_kubernetes_endpoint_port_name]    action: keep    regex: default;kubernetes;https # Scrape config for nodes (kubelet). # # Rather than connecting directly to the node, the scrape is proxied though the # Kubernetes apiserver.  This means it will work if Prometheus is running out of # cluster, or can't connect to nodes for some other reason (e.g. because of # firewalling). - job_name: 'kubernetes-nodes'  # Default to scraping over https. If required, just disable this or change to  # `http`.  scheme: https  # This TLS & bearer token file config is used to connect to the actual scrape  # endpoints for cluster components. This is separate to discovery auth  # configuration because discovery & scraping are two separate concerns in  # Prometheus. The discovery auth config is automatic if Prometheus runs inside  # the cluster. Otherwise, more config options have to be provided within the  # <kubernetes_sd_config>.  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/${1}/proxy/metrics # Scrape config for Kubelet cAdvisor. # # This is required for Kubernetes 1.7.3 and later, where cAdvisor metrics # (those whose names begin with 'container_') have been removed from the # Kubelet metrics endpoint.  This job scrapes the cAdvisor endpoint to # retrieve those metrics. # # In Kubernetes 1.7.0-1.7.2, these metrics are only exposed on the cAdvisor # HTTP endpoint; use "replacement: /api/v1/nodes/${1}:4194/proxy/metrics" # in that case (and ensure cAdvisor's HTTP server hasn't been disabled with # the --cadvisor-port=0 Kubelet flag). # # This job is not necessary and should be removed in Kubernetes 1.6 and # earlier versions, or it will cause the metrics to be scraped twice. - job_name: 'kubernetes-cadvisor'  # Default to scraping over https. If required, just disable this or change to  # `http`.  scheme: https  # This TLS & bearer token file config is used to connect to the actual scrape  # endpoints for cluster components. This is separate to discovery auth  # configuration because discovery & scraping are two separate concerns in  # Prometheus. The discovery auth config is automatic if Prometheus runs inside  # the cluster. Otherwise, more config options have to be provided within the  # <kubernetes_sd_config>.  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/cadvisor # Scrape config for service endpoints. # # The relabeling allows the actual service scrape endpoint to be configured # via the following annotations: # # * `prometheus.io/scrape`: Only scrape services that have a value of `true` # * `prometheus.io/scheme`: If the metrics endpoint is secured then you will need # to set this to `https` & most likely set the `tls_config` of the scrape config. # * `prometheus.io/path`: If the metrics path is not `/metrics` override this. # * `prometheus.io/port`: If the metrics are exposed on a different port to the # service then set this appropriately. - job_name: 'kubernetes-service-endpoints'  kubernetes_sd_configs:  - role: endpoints  relabel_configs:  - source_labels: [__meta_kubernetes_service_annotation_prometheus_io_scrape]    action: keep    regex: true  - source_labels: [__meta_kubernetes_service_annotation_prometheus_io_scheme]    action: replace    target_label: __scheme__    regex: (https?)  - source_labels: [__meta_kubernetes_service_annotation_prometheus_io_path]    action: replace    target_label: __metrics_path__    regex: (.+)  - source_labels: [__address__, __meta_kubernetes_service_annotation_prometheus_io_port]    action: replace    target_label: __address__    regex: ([^:]+)(?::d+)?;(d+)    replacement: $1:$2  - action: labelmap    regex: __meta_kubernetes_service_label_(.+)  - source_labels: [__meta_kubernetes_namespace]    action: replace    target_label: kubernetes_namespace  - source_labels: [__meta_kubernetes_service_name]    action: replace    target_label: kubernetes_name # Example scrape config for probing services via the Blackbox Exporter. # # The relabeling allows the actual service scrape endpoint to be configured # via the following annotations: # # * `prometheus.io/probe`: Only probe services that have a value of `true` - 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.example.com: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 # Example scrape config for probing ingresses via the Blackbox Exporter. # # The relabeling allows the actual ingress scrape endpoint to be configured # via the following annotations: # # * `prometheus.io/probe`: Only probe services that have a value of `true` - 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.example.com: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 # Example scrape config for pods # # The relabeling allows the actual pod scrape endpoint to be configured via the # following annotations: # # * `prometheus.io/scrape`: Only scrape pods that have a value of `true` # * `prometheus.io/path`: If the metrics path is not `/metrics` override this. # * `prometheus.io/port`: Scrape the pod on the indicated port instead of the # pod's declared ports (default is a port-free target if none are declared). - 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

当然该配置文件，是在prometheus部署在k8s中生效的,即in-cluster模式。

kubernetes-apiservers

该项主要是让prometheus程序可以访问kube-apiserver，进而进行服务发现。看一下服务发现的代码可以看出，主要服务发现：node，service，ingress，pod。

    switch d.role {    case "endpoints":        var wg sync.WaitGroup        for _, namespace := range namespaces {            elw := cache.NewListWatchFromClient(rclient, "endpoints", namespace, nil)            slw := cache.NewListWatchFromClient(rclient, "services", namespace, nil)            plw := cache.NewListWatchFromClient(rclient, "pods", namespace, nil)            eps := NewEndpoints(                log.With(d.logger, "role", "endpoint"),                cache.NewSharedInformer(slw, &apiv1.Service{}, resyncPeriod),                cache.NewSharedInformer(elw, &apiv1.Endpoints{}, resyncPeriod),                cache.NewSharedInformer(plw, &apiv1.Pod{}, resyncPeriod),            )            go eps.endpointsInf.Run(ctx.Done())            go eps.serviceInf.Run(ctx.Done())            go eps.podInf.Run(ctx.Done())            for !eps.serviceInf.HasSynced() {                time.Sleep(100 * time.Millisecond)            }            for !eps.endpointsInf.HasSynced() {                time.Sleep(100 * time.Millisecond)            }            for !eps.podInf.HasSynced() {                time.Sleep(100 * time.Millisecond)            }            wg.Add(1)            go func() {                defer wg.Done()                eps.Run(ctx, ch)            }()        }        wg.Wait()    case "pod":        var wg sync.WaitGroup        for _, namespace := range namespaces {            plw := cache.NewListWatchFromClient(rclient, "pods", namespace, nil)            pod := NewPod(                log.With(d.logger, "role", "pod"),                cache.NewSharedInformer(plw, &apiv1.Pod{}, resyncPeriod),            )            go pod.informer.Run(ctx.Done())            for !pod.informer.HasSynced() {                time.Sleep(100 * time.Millisecond)            }            wg.Add(1)            go func() {                defer wg.Done()                pod.Run(ctx, ch)            }()        }        wg.Wait()    case "service":        var wg sync.WaitGroup        for _, namespace := range namespaces {            slw := cache.NewListWatchFromClient(rclient, "services", namespace, nil)            svc := NewService(                log.With(d.logger, "role", "service"),                cache.NewSharedInformer(slw, &apiv1.Service{}, resyncPeriod),            )            go svc.informer.Run(ctx.Done())            for !svc.informer.HasSynced() {                time.Sleep(100 * time.Millisecond)            }            wg.Add(1)            go func() {                defer wg.Done()                svc.Run(ctx, ch)            }()        }        wg.Wait()    case "ingress":        var wg sync.WaitGroup        for _, namespace := range namespaces {            ilw := cache.NewListWatchFromClient(reclient, "ingresses", namespace, nil)            ingress := NewIngress(                log.With(d.logger, "role", "ingress"),                cache.NewSharedInformer(ilw, &extensionsv1beta1.Ingress{}, resyncPeriod),            )            go ingress.informer.Run(ctx.Done())            for !ingress.informer.HasSynced() {                time.Sleep(100 * time.Millisecond)            }            wg.Add(1)            go func() {                defer wg.Done()                ingress.Run(ctx, ch)            }()        }        wg.Wait()    case "node":        nlw := cache.NewListWatchFromClient(rclient, "nodes", api.NamespaceAll, nil)        node := NewNode(            log.With(d.logger, "role", "node"),            cache.NewSharedInformer(nlw, &apiv1.Node{}, resyncPeriod),        )        go node.informer.Run(ctx.Done())        for !node.informer.HasSynced() {            time.Sleep(100 * time.Millisecond)        }        node.Run(ctx, ch)    default:        level.Error(d.logger).Log("msg", "unknown Kubernetes discovery kind", "role", d.role)    }   

kubernetes-nodes

发现node以后，通过/api/v1/nodes/${1}/proxy/metrics来获取node的metrics。

kubernetes-cadvisor

cadvisor已经被集成在kubelet中，所以发现了node就相当于发现了cadvisor。通过 /api/v1/nodes/${1}/proxy/metrics/cadvisor采集容器指标。

kubernetes-services和kubernetes-ingresses

该两种资源监控方式差不多，都是需要安装black-box，然后类似于探针去定时访问，根据返回的http状态码来判定service和ingress的服务可用性。
PS：不过我自己在这里和官方的稍微有点区别，

- target_label: __address__
      replacement: blackbox-exporter.example.com:9115

官方大致是需要我们要创建black-box 的ingress从外部访问，这样从效率和安全性都不是最合适的。所以我一般都是直接内部dns访问。如下

- target_label: __address__      replacement: blackbox-exporter.kube-system:9115

当然看源码可以发现，并不是所有的service和ingress都会健康监测，如果需要将服务进行健康监测，那么你部署应用的yaml文件加一些注解。例如：
对于service和ingress：
需要加注解：prometheus.io/scrape: 'true'

apiVersion: v1kind: Servicemetadata:  annotations:    prometheus.io/scrape: 'true'  name: prometheus-node-exporter  namespace: kube-system  labels:    app: prometheus    component: node-exporterspec:  clusterIP: None  ports:    - name: prometheus-node-exporter      port: 9100      protocol: TCP  selector:    app: prometheus    component: node-exporter  type: ClusterIP

kubernetes-pods

对于pod的监测也是需要加注解：

• prometheus.io/scrape，为true则会将pod作为监控目标。

• prometheus.io/path，默认为/metrics

• prometheus.io/port , 端口

所以看到此处可以看出，该job并不是监控pod的指标，pod已经通过前面的cadvisor采集。此处是对pod中应用的监控。写过exporter的人应该对这个概念非常清楚。通俗讲，就是你pod中的应用提供了prometheus的监控功能，加上对应的注解，那么该应用的metrics会定时被采集走。

kubernetes-service-endpoints

对于服务的终端节点，也需要加注解：

• prometheus.io/scrape，为true则会将pod作为监控目标。

• prometheus.io/path，默认为/metrics

• prometheus.io/port , 端口

• prometheus.io/scheme 默认http，如果为了安全设置了https，此处需要改为https

这个基本上同上的。采集service-endpoints的metrics。

个人认为：如果某些部署应用只有pod没有service，那么这种情况只能在pod上加注解，通过kubernetes-pods采集metrics。如果有service，那么就无需在pod加注解了，直接在service上加即可。毕竟service-endpoints最终也会落到pod上。

总结

配置项总结

• kubernetes-service-endpoints和kubernetes-pods采集应用中metrics，当然并不是所有的都提供了metrics接口。

• kubernetes-ingresses 和kubernetes-services 健康监测服务和ingress健康的状态

• kubernetes-cadvisor 和 kubernetes-nodes，通过发现node，监控node 和容器的cpu等指标

自动发现源码

参考client-go和prometheus自动发现k8s，这种监听k8s集群中资源的变化，使用informer实现，不要轮询kube-apiserver接口。

参考

该配置文件需要部署一些组件来支持prometheus对k8s的监控，例如black-exporter。因为要自动发现，获取集群的一些信息，所以也要做rbac的授权。具体参考：https://github.com/giantswarm/kubernetes-prometheus

人生舞台的大幕随时都可能拉开，关键是你愿意表演，还是选择躲避。