OpenShift采用Blackbox_exporter进行服务状态监控

本文主要是针对prometheus的blackbox_exporter，对集群中的服务进行状态的监控

因为OpenShift 3.11版本自己带的promethues修改起来有一些问题，所以自己安装和部署了自己的prometheus和grafana

1.BlackBox安装

blackbox的详细情况可以参考

https://github.com/prometheus/blackbox_exporter

安装过程如下

git clone https://github.com/prometheus/blackbox_exporter.git

在https://github.com/prometheus/blackbox_exporter/releases

下载最新版本的blackbox_exporter-0.14.0.linux-amd64.tar.gz , 解压放到相应的目录下，然后构建镜像

docker build -t blackbox_exporter .

blackbox可以直接通过docker run来运行在本地，比如

docker run -d -p 9115:9115 --name blackbox blackbox_exporter

但问题是我们需要针对ocp内部的服务进行监控，因此最好还是作为pod运行在ocp的内部，这样通过服务就可以获取固定的url.

2.Blackbox部署成为Pod

下面三条语句，比较简单了...

oc new-project blackbox
oc import-image blackbox:latest --from=registry.example.com/openshift/blackbox_exporter:latest  --insecure --confirm

oc new-app blackbox:latest --name=blackbox

完成后看到有个blackbox pod在运行

 3.部署prometheus

Prometheus部署使用的是

https://github.com/openshift/origin/tree/master/examples/prometheus

下载相关版本镜像，推送到registry.example.com,然后

oc new-project prometheus
oc new-app -f prometheus.yaml -p NAMESPACE=prometheus

因为原本prometheus的配置是在内部的，为了便于修改，直接放到宿主机上。

将需要监控的url写入到prometheus.yml的配置

配置文件如下

[root@master prometheus]# cat prometheus.yml
rule_files:
  - '*.rules'
.....
- job_name: blackbox
  metrics_path: /probe
  params:
    module: [http_2xx]
  static_configs:
  - targets:
      - http://ericapp1.ericproject1:8080
      - http://ericapp2.ericproject1:8080
  relabel_configs:
  - source_labels: [__address__]
    target_label: __param_target
  - source_labels: [__param_target]
    target_label: instance
  - target_label: __address__
    replacement: blackbox.blackbox:9115 # Blackbox exporter.

alerting:
  alertmanagers:
  - scheme: http
    static_configs:
    - targets:
      - "localhost:9093"

主要看blackbox这段.

oc create configmap prom-config --from-file=prometheus.yml

oc set volume sts/prometheus  --add --overwrite  --name=config-volume   --mount-path=/etc/prometheus/myconfig  --source='{"configMap": { "name": "prom-config"}}'

然后进入console修改prometheus的yaml

 一切完成，进入prometheus的主页面，看到相关的指标

 

 4.部署grafana

可以参考

https://github.com/openshift/origin/tree/master/examples/grafana

https://labs.consol.de/development/2018/01/19/openshift_application_monitoring.html

oc new-project grafana

oc new-app -f grafana.yaml -p NAMESPACE=grafana

oc policy add-role-to-user view system:serviceaccount:grafana:grafana-ocp -n prometheus

登录grafana界面，添加数据源，这个数据源这里折腾了一段时间

 因为是走的https,所以必须需要token,这个token值为

[root@master prometheus]# oc sa get-token grafana-ocp
eyJhbGciOiJSUzI1NiIsImtpZCI6IiJ9.eyJpc3MiOiJrdWJlcm5ldGVzL3NlcnZpY2VhY2NvdW50Iiwia3ViZXJuZXRlcy5pby9zZXJ2aWNlYWNjb3VudC9uYW1lc3BhY2UiOiJncmFmYW5hIiwia3ViZXJuZXRlcy5pby9zZXJ2aWNlYWNjb3VudC9zZWNyZXQubmFtZSI6ImdyYWZhbmEtb2NwLXRva2VuLXhodDl2Iiwia3ViZXJuZXRlcy5pby9zZXJ2aWNlYWNjb3VudC9zZXJ2aWNlLWFjY291bnQubmFtZSI6ImdyYWZhbmEtb2NwIiwia3ViZXJuZXRlcy5pby9zZXJ2aWNlYWNjb3VudC9zZXJ2aWNlLWFjY291bnQudWlkIjoiM2IyYmU5YjQtNThkMS0xMWU5LThlMDQtMDgwMDI3ZGM5OTFhIiwic3ViIjoic3lzdGVtOnNlcnZpY2VhY2NvdW50OmdyYWZhbmE6Z3JhZmFuYS1vY3AifQ.NAf03C7v42aBNrRbwLcIosKDFgv3WyBA2-jZ73RqmX-OyNkshzgkvO6RZDGU7Oi9x27Hu7S69CW-2nZBDqSuCKP_fOyr7pKkOZTHfmgnXQdOeqnG8WSGM5HMbMsLbrAgOMZSehG5jFos-FaJb6jYJwQx485u0ViwWcSN58y7m4unIyrXgvzvjCoFijIjoCcYOvJklZVAsToVIFnevOcXEdJhtoitFiG2Agg4_KSiEUOhhKzyvKCIOLfSZN1UGsicUx_vFmqix8ARmOegILJNdRHksBAXRdReZuaJlJ1-FJnh1j9-QVVSSgw2DP6JrMEhS9gV38VWiWhoizSZLqsgCw

 Save Test后一定需要出现is working啊.

New Dashboard

 singlestat->edit->metrics,修改options

修改完最后

 好啦，搞定！