kube-prometheus

官方Readme

Readme

本项目使用方法

git clone https://github.com/kssamwang/kube-prometheus.git
cd kube-prometheus

安装

# Create the namespace and CRDs, and then wait for them to be availble before creating the remaining resources
kubectl create -f manifests/setup

# Wait until the "servicemonitors" CRD is created. The message "No resources found" means success in this context.
until kubectl get servicemonitors --all-namespaces ; do date; sleep 1; echo ""; done

kubectl create -f manifests/

移除

kubectl delete --ignore-not-found=true -f manifests/ -f manifests/setup

k8s安装配置Prometheus

1 下载kube-prometheus包安装

选择版本v0.10.0

git clone -b v0.10.0 https://github.com/prometheus-operator/kube-prometheus.git
cd kube-prometheus

安装

# Create the namespace and CRDs, and then wait for them to be availble before creating the remaining resources
kubectl create -f manifests/setup

# Wait until the "servicemonitors" CRD is created. The message "No resources found" means success in this context.
until kubectl get servicemonitors --all-namespaces ; do date; sleep 1; echo ""; done

kubectl create -f manifests/

移除

kubectl delete --ignore-not-found=true -f manifests/ -f manifests/setup

2 查看监控组件信息

查看CRD类型。

kubectl get crd | grep coreos

查看特定CRD类型下实例。

kubectl get prometheuses -n monitoring
kubectl get servicemonitors -n monitoring

查看创建的service。

kubectl get svc -n monitoring

查询monitoring命名空间所有组件的状态。

kubectl get po -n monitoring

查询非Running状态pod的方法，可以获得容器启动失败的原因。

kubectl describe  po prometheus-adapter-7858d4ddfd-55lnq  -n monitoring

3 解决组件所需镜像 ImagePullBackOff / ErrImageNeverPull

参考：https://blog.csdn.net/qq_45439217/article/details/123477846

3.1 prometheus-adapter

拉取可以替代的镜像到本地，打上标签

docker pull lbbi/prometheus-adapter:v0.9.1
docker tag docker.io/lbbi/prometheus-adapter:v0.9.1 k8s.gcr.io/prometheus-adapter:v0.9.1

vi manifests/prometheusAdapter-deployment.yaml

文件内容：

***              修改前               ***
 
        image: k8s.gcr.io/prometheus-adapter/prometheus-adapter:v0.9.1
        name: prometheus-adapter
 
***              修改后               ***
 
        # image: k8s.gcr.io/prometheus-adapter/prometheus-adapter:v0.9.1
        image: k8s.gcr.io/prometheus-adapter:v0.9.1   #  容器镜像标签，写错拉取本地镜像失败。
        imagePullPolicy: Never                        #  imagePullPolicy: Always 总是网络拉取镜像, 是k8s默认的拉取方式。
                                                      #  imagePullPolicy: Never 从不远程拉取镜像，只读取本地镜像。
                                                      #  imagePullPolicy: IfNotPresent 优先拉取本地镜像。
        name: prometheus-adapter 

重新执行该组件pod的yaml文件。

kubectl replace -f manifests/prometheusAdapter-deployment.yaml

3.2 kube-state-metrics

拉取可以替代的镜像到本地，打上标签

docker pull bitnami/kube-state-metrics:2.3.0

vi manifests/kubeStateMetrics-deployment.yaml

文件内容：

***              修改前               ***
 
        image: k8s.gcr.io/kube-state-metrics/kube-state-metrics:2.3.0
        name: kube-state-metrics
 
***              修改后               ***
 
        # image: k8s.gcr.io/kube-state-metrics/kube-state-metrics:2.3.0
        image: docker.io/bitnami/kube-state-metrics:2.3.0
        imagePullPolicy: IfNotPresent
        name: kube-state-metrics

重新执行该组件pod的yaml文件。

kubectl replace -f manifests/kubeStateMetrics-deployment.yaml