Introduction

This repo is a summary for my recent work.

My work is to monitorer k8s metrics using prometheus and auto-scaling resources, especially the storage resources.

I have wasted so much time on setting up the environment and related tools. And this is a summary I made.

So:

1. This is a summary for Kubernetes setup and installing kube-prometheus quickly（Actually, in just one step）.

2. Apart from that, this repo also includes storage auto-scaling quick setup yml.

3. In the future, I will include more summary in this repo.

Environment

Ubuntu18.04 is OK. No try on other system.

Step

Step1: setup Kubernetes

./k8s_setup.sh

After that, the K8s is set up. Run kubectl get po to test that.

---

Step2: setup kube-prometheus

cd my-kube-prometheus/manifests
kubectl apply -f .
kuebctl apply -f .
cd ../..

Step3: setup custom-metrics

cd custom-metrics-adapter/deploy/manifests/
k apply -f .
cd ../../..

---

Step4: setup Kube-prometheus

 k apply -f my-kube-prometheus/manifests/

Note: Some resources may not create correctly, just because of the wrong setup sequence. So running the upper command twice will solve problem.

And how to access the kube-prometheus? The following:

Accessing the dashboards

Prometheus, Grafana和Alertmanager dashboards可以通过kubectl port-forward快速访问: (1) Prometheus

kubectl --namespace monitoring port-forward svc/prometheus-k8s 9090 然后可以通过 http://localhost:9090访问Prometheus。

(2) Grafana

kubectl --namespace monitoring port-forward svc/grafana 3000 然后可以通过 http://localhost:3000访问Grafana，默认的用户名和密码是 admin:admin

(3) Alert Manager

kubectl --namespace monitoring port-forward svc/alertmanager-main 9093 然后可以通过 http://localhost:9093来访问。

---

Step3: setup csi-related resources

./lvm/csi_setup.sh

So, csi-related resources are set up.

The directory example show an example of creating pv dynamicly.