k8s监控-Prometheus

HoweWWW

已于 2024-09-17 20:31:12 修改

阅读量1.4k

点赞数 15

CC 4.0 BY-SA版权

文章标签： kubernetes prometheus 容器

于 2024-09-17 20:27:01 首次发布

本文链接：https://blog.youkuaiyun.com/HoweWWW/article/details/142318251

2.1 下载部署Prometheus所需资源

一 Prometheus简介

Prometheus是一个开源的服务监控系统和时序数据库

其提供了通用的数据模型和快捷数据采集、存储和查询接口

它的核心组件Prometheus服务器定期从静态配置的监控目标或者基于服务发现自动配置的目标中进行拉取数据

新拉取到啊的数据大于配置的内存缓存区时，数据就会持久化到存储设备当中

1.1 Prometheus架构

1.1.1 组件功能：

监控代理程序：如node_exporter：收集主机的指标数据，如平均负载、CPU、内存、磁盘、网络等等多个维度的指标数据。
kubelet（cAdvisor）：收集容器指标数据，也是K8S的核心指标收集，每个容器的相关指标数据包括：CPU使用率、限额、文件系统读写限额、内存使用率和限额、网络报文发送、接收、丢弃速率等等。
API Server：收集API Server的性能指标数据，包括控制队列的性能、请求速率和延迟时长等等
etcd：收集etcd存储集群的相关指标数据
kube-state-metrics：该组件可以派生出k8s相关的多个指标数据，主要是资源类型相关的计数器和元数据信息，包括制定类型的对象总数、资源限额、容器状态以及Pod资源标签系列等。

每个被监控的主机都可以通过专用的exporter程序提供输出监控数据的接口，并等待Prometheus服务器周期性的进行数据抓取
如果存在告警规则，则抓取到数据之后会根据规则进行计算，满足告警条件则会生成告警，并发送到Alertmanager完成告警的汇总和分发
当被监控的目标有主动推送数据的需求时，可以以Pushgateway组件进行接收并临时存储数据，然后等待Prometheus服务器完成数据的采集
任何被监控的目标都需要事先纳入到监控系统中才能进行时序数据采集、存储、告警和展示
监控目标可以通过配置信息以静态形式指定，也可以让Prometheus通过服务发现的机制进行动态管理

二在k8s中部署Prometheus

2.1 下载部署Prometheus所需资源

#在helm中添加Prometheus仓库（网络巨好才做）
[root@k8s-master helm]# helm repo add prometheus-community https://prometheus-community.github.io/helm-charts
"prometheus-community" has been added to your repositories

#下载Prometheus项目
[root@k8s-master helm]# helm pull prometheus-community/kube-prometheus-stack
[root@k8s-master helm]# ls
kube-prometheus-stack-62.6.0.tgz

================================================================

2.2 部署步骤

根据所有项目中的values.yaml中指定的image路径下载容器镜像并上传至harbor仓库

[root@k8s-master ~]# mkdir prometheus
[root@k8s-master ~]# cd prometheus/

[root@k8s-master prometheus]# ls
grafana-11.2.0.tar                kube-state-metrics-2.13.0.tar  nginx-exporter-1.3.0-debian-12-r2.tar  prometheus-62.6.0.tar
kube-prometheus-stack-62.6.0.tgz  nginx-18.1.11.tgz              node-exporter-1.8.2.tar

-----------------------------------------------------------------------------------
[root@k8s-master prometheus]# tar zxf kube-prometheus-stack-62.6.0.tgz 
[root@k8s-master prometheus]# cd kube-prometheus-stack/
[root@k8s-master kube-prometheus-stack]# ls
Chart.lock  charts  Chart.yaml  CONTRIBUTING.md  README.md  templates  values.yaml

#修改到本地harbor仓库
[root@k8s-master kube-prometheus-stack]# vim values.yaml
 227   imageRegistry: "reg.exam.com"

----------------------------------------------------------------------------------
#导入镜像
[root@k8s-master prometheus]# docker load -i prometheus-62.6.0.tar 
Loaded image: quay.io/prometheus/prometheus:v2.54.1
Loaded image: quay.io/thanos/thanos:v0.36.1
Loaded image: quay.io/prometheus/alertmanager:v0.27.0
Loaded image: quay.io/prometheus-operator/admission-webhook:v0.76.1
Loaded image: registry.k8s.io/ingress-nginx/kube-webhook-certgen:v20221220-controller-v1.5.1-58-g787ea74b6
Loaded image: quay.io/prometheus-operator/prometheus-operator:v0.76.1
Loaded image: quay.io/prometheus-operator/prometheus-config-reloader:v0.76.1

#打包并上传镜像
[root@k8s-master prometheus]# docker tag quay.io/prometheus/prometheus:v2.54.1 reg.exam.com/prometheus/prometheus:v2.54.1
[root@k8s-master prometheus]# docker push reg.exam.com/prometheus/prometheus:v2.54.1

[root@k8s-master prometheus]# docker tag quay.io/thanos/thanos:v0.36.1 reg.exam.com/thanos/thanos:v0.36.1
[root@k8s-master prometheus]# docker push reg.exam.com/thanos/thanos:v0.36.1

[root@k8s-master prometheus]# docker tag quay.io/prometheus/alertmanager:v0.27.0 reg.exam.com/prometheus/alertmanager:v0.27.0
[root@k8s-master prometheus]# docker push reg.exam.com/prometheus/alertmanager:v0.27.0

[root@k8s-master prometheus]# docker tag quay.io/prometheus-operator/admission-webhook:v0.76.1 reg.exam.com/prometheus-operator/admission-webhook:v0.76.1
[root@k8s-master prometheus]# docker push reg.exam.com/prometheus-operator/admission-webhook:v0.76.1

[root@k8s-master prometheus]# docker tag registry.k8s.io/ingress-nginx/kube-webhook-certgen:v20221220-controller-v1.5.1-58-g787ea74b6 reg.exam.com/ingress-nginx/kube-webhook-certgen:v20221220-controller-v1.5.1-58-g787ea74b6
[root@k8s-master prometheus]# docker push reg.exam.com/ingress-nginx/kube-webhook-certgen:v20221220-controller-v1.5.1-58-g787ea74b6

[root@k8s-master prometheus]# docker tag quay.io/prometheus-operator/prometheus-operator:v0.76.1 reg.exam.com/prometheus-operator/prometheus-operator:v0.76.1
[root@k8s-master prometheus]# docker push reg.exam.com/prometheus-operator/prometheus-operator:v0.76.1

[root@k8s-master prometheus]# docker tag quay.io/prometheus-operator/prometheus-config-reloader:v0.76.1 reg.exam.com/prometheus-operator/prometheus-config-reloader:v0.76.1
[root@k8s-master prometheus]# docker push reg.exam.com/prometheus-operator/prometheus-config-reloader:v0.76.1

#更改仓库地址
[root@k8s-master prometheus]# cd kube-prometheus-stack/charts/grafana/

[root@k8s-master grafana]# pwd
/root/prometheus/kube-prometheus-stack/charts/grafana

[root@k8s-master grafana]# vim values.yaml 
3   	imageRegistry: "reg.exam.com"
418     tag: "latest"

-----------------------------------------------------------------------------

#导入grafana镜像包
[root@k8s-master prometheus]# docker load -i grafana-11.2.0.tar 
Loaded image: grafana/grafana:11.2.0
Loaded image: quay.io/kiwigrid/k8s-sidecar:1.27.4
Loaded image: grafana/grafana-image-renderer:latest
Loaded image: bats/bats:v1.4.1

#打包上传到harbor仓库
[root@k8s-master prometheus]# docker tag grafana/grafana:11.2.0 reg.exam.com/grafana/grafana:11.2.0
[root@k8s-master prometheus]# docker push reg.exam.com/grafana/grafana:11.2.0

[root@k8s-master prometheus]# docker tag quay.io/kiwigrid/k8s-sidecar:1.27.4 reg.exam.com/kiwigrid/k8s-sidecar:1.27.4
[root@k8s-master prometheus]# docker push reg.exam.com/kiwigrid/k8s-sidecar:1.27.4

[root@k8s-master prometheus]# docker tag grafana/grafana-image-renderer:latest reg.exam.com/grafana/grafana-image-renderer:latest
[root@k8s-master prometheus]# docker push reg.exam.com/grafana/grafana-image-renderer:latest

[root@k8s-master prometheus]# docker tag bats/bats:v1.4.1 reg.exam.com/bats/bats:v1.4.1
[root@k8s-master prometheus]# docker push reg.exam.com/bats/bats:v1.4.1 

#修改配置文件中的仓库地址
[root@k8s-master kube-state-metrics]# pwd
/root/prometheus/kube-prometheus-stack/charts/kube-state-metrics
[root@k8s-master kube-state-metrics]# ls
Chart.yaml  README.md  templates  values.yaml
[root@k8s-master kube-state-metrics]# vim values.yaml 
  4   registry: reg.exam.com
 29   imageRegistry: "reg.exam.com"

-------------------------------------------------------------------------------------

#导入镜像
[root@k8s-master prometheus]# docker load -i kube-state-metrics-2.13.0.tar 
Loaded image: registry.k8s.io/kube-state-metrics/kube-state-metrics:v2.13.0
Loaded image: quay.io/brancz/kube-rbac-proxy:v0.18.0

#打包上传
[root@k8s-master prometheus]# docker tag registry.k8s.io/kube-state-metrics/kube-state-metrics:v2.13.0 reg.exam.com/kube-state-metrics/kube-state-metrics:v2.13.0
[root@k8s-master prometheus]# docker push reg.exam.com/kube-state-metrics/kube-state-metrics:v2.13.0

[root@k8s-master prometheus]# docker tag quay.io/brancz/kube-rbac-proxy:v0.18.0 reg.exam.com/brancz/kube-rbac-proxy:v0.18.0
[root@k8s-master prometheus]# docker push reg.exam.com/brancz/kube-rbac-proxy:v0.18.0

#修改node监控配置文件仓库地址
[root@k8s-master prometheus-node-exporter]# pwd
/root/prometheus/kube-prometheus-stack/charts/prometheus-node-exporter
[root@k8s-master prometheus-node-exporter]# ls
Chart.yaml  ci  README.md  templates  values.yaml

[root@k8s-master prometheus-node-exporter]# vim values.yaml 
  5   registry: reg.exam.com
 36   imageRegistry: "reg.exam.com"
 
------------------------------------------------------------------------------

#导入node镜像
[root@k8s-master prometheus]# docker load -i node-exporter-1.8.2.tar 
Loaded image: quay.io/prometheus/node-exporter:v1.8.2
Loaded image: quay.io/brancz/kube-rbac-proxy:v0.18.0	#已上传

#打包上传
[root@k8s-master prometheus]# docker tag quay.io/prometheus/node-exporter:v1.8.2 reg.exam.com/prometheus/node-exporter:v1.8.2
[root@k8s-master prometheus]# docker push reg.exam.com/prometheus/node-exporter:v1.8.2

[root@k8s-master prometheus]# docker tag quay.io/brancz/kube-rbac-proxy:v0.18.0 reg.exam.com/brancz/kube-rbac-proxy:v0.18.0 
[root@k8s-master prometheus]# docker push reg.exam.com/brancz/kube-rbac-proxy:v0.18.0

================================================================================

创建命名空间

[root@k8s-master prometheus]# kubectl create namespace kube-prometheus-stack
namespace/kube-prometheus-stack created

[root@k8s-master prometheus]# kubectl get namespaces 
NAME                    STATUS   AGE
default                 Active   154m
kube-node-lease         Active   154m
kube-prometheus-stack   Active   8s
kube-public             Active   154m
kube-system             Active   154m
metallb-system          Active   54m

利用helm安装Prometheus ！注意，在安装过程中千万别ctrl+c！

[root@k8s-master prometheus]# cd kube-prometheus-stack/
[root@k8s-master kube-prometheus-stack]# 

# . 代表当前位置/root/prometheus/kube-prometheus-stack
[root@k8s-master kube-prometheus-stack]# helm -n kube-prometheus-stack install kube-prometheus-stack .
NAME: kube-prometheus-stack
LAST DEPLOYED: Thu Sep 12 20:54:37 2024
NAMESPACE: kube-prometheus-stack
STATUS: deployed
REVISION: 1
NOTES:
kube-prometheus-stack has been installed. Check its status by running:
  kubectl --namespace kube-prometheus-stack get pods -l "release=kube-prometheus-stack"

Visit https://github.com/prometheus-operator/kube-prometheus for instructions on how to create & configure Alertmanager and Prometheus instances using the Operator.


查看所有pod是否运行
[root@k8s-master kube-prometheus-stack]# kubectl --namespace kube-prometheus-stack get pods
NAME                                                        READY   STATUS    RESTARTS   AGE
alertmanager-kube-prometheus-stack-alertmanager-0           2/2     Running   0          23m
kube-prometheus-stack-grafana-548c8fb6c4-29qdc              3/3     Running   0          23m
kube-prometheus-stack-kube-state-metrics-6688476957-n26gn   1/1     Running   0          23m
kube-prometheus-stack-operator-587f4b669b-8ztmk             1/1     Running   0          23m
kube-prometheus-stack-prometheus-node-exporter-j6j4t        1/1     Running   0          23m
kube-prometheus-stack-prometheus-node-exporter-pccpc        1/1     Running   0          23m
kube-prometheus-stack-prometheus-node-exporter-t77b8        1/1     Running   0          23m
prometheus-kube-prometheus-stack-prometheus-0               2/2     Running   0          23m

查看svc
[root@k8s-master kube-prometheus-stack]# kubectl -n kube-prometheus-stack get svc
NAME                                             TYPE        CLUSTER-IP       EXTERNAL-IP   PORT(S)                      AGE
alertmanager-operated                            ClusterIP   None             <none>        9093/TCP,9094/TCP,9094/UDP   23m
kube-prometheus-stack-alertmanager               ClusterIP   10.104.96.90     <none>        9093/TCP,8080/TCP            23m
kube-prometheus-stack-grafana                    ClusterIP   10.103.122.224   <none>        80/TCP                       23m
kube-prometheus-stack-kube-state-metrics         ClusterIP   10.104.185.222   <none>        8080/TCP                     23m
kube-prometheus-stack-operator                   ClusterIP   10.98.25.116     <none>        443/TCP                      23m
kube-prometheus-stack-prometheus                 ClusterIP   10.102.144.68    <none>        9090/TCP,8080/TCP            23m
kube-prometheus-stack-prometheus-node-exporter   ClusterIP   10.98.117.125    <none>        9100/TCP                     23m
prometheus-operated                              ClusterIP   None             <none>        9090/TCP                     23m

修改暴漏方式
[root@k8s-master kube-prometheus-stack]# kubectl -n kube-prometheus-stack edit svc kube-prometheus-stack-grafana 
39   type: LoadBalancer

各个svc的作用
alertmanager-operated 			告警管理
kube-prometheus-stack-grafana 	 展示prometheus采集到的指标
kube-prometheus-stack-prometheus-node-exporter 收集节点级别的指标的工具
kube-prometheus-stack-prometheus 主程序

各个svc的作用

alertmanager-operated 告警管理

kube-prometheus-stack-grafana 展示prometheus采集到的指标

kube-prometheus-stack-prometheus-node-exporter 收集节点级别的指标的工具

kube-prometheus-stack-prometheus 主程

2.3 登陆grafana

查看grafana密码
[root@k8s-master helm]# kubectl -n kube-prometheus-stack get secrets kube-prometheus-stack-grafana -o yaml
apiVersion: v1
data:
  admin-password: cHJvbS1vcGVyYXRvcg==
  admin-user: YWRtaW4=
  ldap-toml: ""
kind: Secret
metadata:
  annotations:
    meta.helm.sh/release-name: kube-prometheus-stack
    meta.helm.sh/release-namespace: kube-prometheus-stack
  creationTimestamp: "2024-09-12T12:54:47Z"
  labels:
    app.kubernetes.io/instance: kube-prometheus-stack
    app.kubernetes.io/managed-by: Helm
    app.kubernetes.io/name: grafana
    app.kubernetes.io/version: 11.2.0
    helm.sh/chart: grafana-8.5.1
  name: kube-prometheus-stack-grafana
  namespace: kube-prometheus-stack
  resourceVersion: "16943"
  uid: d19640ae-4b79-4013-ba03-e039fc98b493
type: Opaque

查看密码
[root@k8s-master helm]# echo -n "cHJvbS1vcGVyYXRvcg==" | base64 -d
prom-operator		#密码

prom-operator[root@k8s-master helm]# echo "YWRtaW4=" | base64 -d
admin				#用户

[root@k8s-master helm]# kubectl -n kube-prometheus-stack get svc
NAME                                             TYPE           CLUSTER-IP       EXTERNAL-IP     PORT(S)                      AGE
alertmanager-operated                            ClusterIP      None             <none>          9093/TCP,9094/TCP,9094/UDP   29m
kube-prometheus-stack-alertmanager               ClusterIP      10.104.96.90     <none>          9093/TCP,8080/TCP            29m
kube-prometheus-stack-grafana                    LoadBalancer   10.103.122.224   172.25.250.50   80:31471/TCP                 29m
kube-prometheus-stack-kube-state-metrics         ClusterIP      10.104.185.222   <none>          8080/TCP                     29m
kube-prometheus-stack-operator                   ClusterIP      10.98.25.116     <none>          443/TCP                      29m
kube-prometheus-stack-prometheus                 ClusterIP      10.102.144.68    <none>          9090/TCP,8080/TCP            29m
kube-prometheus-stack-prometheus-node-exporter   ClusterIP      10.98.117.125    <none>          9100/TCP                     29m
prometheus-operated                              ClusterIP      None             <none>          9090/TCP                     29m

#用分配的IP在网页查看

2.4 导入面板

官方监控模板：Grafana dashboards | Grafana Labs

面板不行就换

2.5 访问Prometheus主程序

[root@k8s-master helm]# kubectl -n kube-prometheus-stack edit svc kube-prometheus-stack-prometheus
48   type: LoadBalancer

[root@k8s-master helm]# kubectl -n kube-prometheus-stack get svc kube-prometheus-stack-prometheus
NAME                               TYPE           CLUSTER-IP      EXTERNAL-IP     PORT(S)                         AGE
kube-prometheus-stack-prometheus   LoadBalancer   10.102.144.68   172.25.250.51   9090:30607/TCP,8080:32132/TCP   43m

网页登录172.25.250.51:9090

三监控使用示例

3.1 建立监控项目

[root@k8s-master ~]# mkdir test
[root@k8s-master ~]# cd test/
[root@k8s-master test]# ls
nginx-18.1.11.tgz  nginx-exporter-1.3.0-debian-12-r2.tar

[root@k8s-master test]# tar zxf nginx-18.1.11.tgz 
[root@k8s-master test]# cd nginx/

修改项目开启监控
[root@k8s-master nginx]# vim values.yaml 
 925 metrics:
 926   ## @param metrics.enabled Start a Prometheus exporter sidecar container
 927   ##
 928   enabled: true		#改为true
...
1015   serviceMonitor:
1016     ## @param metrics.serviceMonitor.enabled Creates a Prometheus Operator ServiceMonitor (also requires `metrics.enable     d` to be `true`)
1017     ##
1018     enabled: true			#改为true
1019     ## @param metrics.serviceMonitor.namespace Namespace in which Prometheus is running
1020     ##
1021     namespace: "kube-prometheus-stack"		#更改命名空间
1022     ## @param metrics.serviceMonitor.jobLabel The name of the label on the target service to use as the job name in prom     etheus.
1023     ##
...
1046     labels: 
1047       release: kube-prometheus-stack		#添加指定监控标签

#查看标签
[root@k8s-master nginx]# kubectl -n kube-prometheus-stack get servicemonitors.monitoring.coreos.com --show-labels

安装项目，在安装之前一定要上传镜像到仓库中

[root@k8s-master nginx]# ls
nginx-1.27.1-debian-12-r2.tar

#第一个nginx
[root@k8s-master nginx]# docker load -i nginx-1.27.1-debian-12-r2.tar
30f5b1069b7f: Loading layer [==================================================>]  190.1MB/190.1MB
Loaded image: bitnami/nginx:1.27.1-debian-12-r2

[root@k8s-master nginx]# docker tag bitnami/nginx:1.27.1-debian-12-r2 reg.exam.com/bitnami/nginx:1.27.1-debian-12-r2
[root@k8s-master nginx]# docker push reg.exam.com/bitnami/nginx:1.27.1-debian-12-r2

#第二个nginx
[root@k8s-master nginx]# docker load -i nginx-exporter-1.3.0-debian-12-r2.tar 
016ff07f0ae3: Loading layer [==================================================>]  149.3MB/149.3MB
Loaded image: bitnami/nginx-exporter:1.3.0-debian-12-r2

[root@k8s-master nginx]# docker tag bitnami/nginx-exporter:1.3.0-debian-12-r2 reg.exam.com/bitnami/nginx-exporter:1.3.0-debian-12-r2
[root@k8s-master nginx]# docker push reg.exam.com/bitnami/nginx-exporter:1.3.0-debian-12-r2


#安装chart包
[root@k8s-master nginx]# helm install howe .
NAME: howe
LAST DEPLOYED: Thu Sep 12 21:52:15 2024
NAMESPACE: default
STATUS: deployed
REVISION: 1
TEST SUITE: None
NOTES:
CHART NAME: nginx
CHART VERSION: 18.1.11
APP VERSION: 1.27.1

[root@k8s-master nginx]# kubectl get pods
NAME                          READY   STATUS    RESTARTS   AGE
howe-nginx-54c97cb888-x5hhh   2/2     Running   0          21s

[root@k8s-master nginx]# kubectl get svc
NAME         TYPE           CLUSTER-IP      EXTERNAL-IP     PORT(S)                                     AGE
kubernetes   ClusterIP      10.96.0.1       <none>          443/TCP                                     4h56m
test-nginx   LoadBalancer   10.102.161.61   172.25.250.52   80:30614/TCP,443:31390/TCP,9113:32254/TCP   30s

[root@k8s-master nginx]# curl 172.25.250.52
<!DOCTYPE html>
<html>
<head>
<title>Welcome to nginx!</title>
<style>
html { color-scheme: light dark; }
body { width: 35em; margin: 0 auto;
font-family: Tahoma, Verdana, Arial, sans-serif; }
</style>
</head>
<body>
<h1>Welcome to nginx!</h1>
<p>If you see this page, the nginx web server is successfully installed and
working. Further configuration is required.</p>

<p>For online documentation and support please refer to
<a href="http://nginx.org/">nginx.org</a>.<br/>
Commercial support is available at
<a href="http://nginx.com/">nginx.com</a>.</p>

<p><em>Thank you for using nginx.</em></p>
</body>
</html>


压力测试：
[root@k8s-master nginx]# ab -c 5 -n 100 http://172.25.250.52/index.html
This is ApacheBench, Version 2.3 <$Revision: 1903618 $>
Copyright 1996 Adam Twiss, Zeus Technology Ltd, http://www.zeustech.net/
Licensed to The Apache Software Foundation, http://www.apache.org/

Benchmarking 172.25.250.52 (be patient).....done


Server Software:        nginx
Server Hostname:        172.25.250.52
Server Port:            80

Document Path:          /index.html
Document Length:        615 bytes

Concurrency Level:      5
Time taken for tests:   0.033 seconds
Complete requests:      100
Failed requests:        0
Total transferred:      87000 bytes
HTML transferred:       61500 bytes
Requests per second:    2991.15 [#/sec] (mean)
Time per request:       1.672 [ms] (mean)
Time per request:       0.334 [ms] (mean, across all concurrent requests)
Transfer rate:          2541.31 [Kbytes/sec] received

Connection Times (ms)
              min  mean[+/-sd] median   max
Connect:        0    1   0.1      0       1
Processing:     0    1   0.9      1       6
Waiting:        0    1   0.8      1       6
Total:          1    1   0.9      1       6
ERROR: The median and mean for the initial connection time are more than twice the standard
       deviation apart. These results are NOT reliable.

Percentage of the requests served within a certain time (ms)
  50%      1
  66%      1
  75%      1
  80%      2
  90%      2
  95%      3
  98%      6
  99%      6
 100%      6 (longest request)