Project versions:

current - October-10-2019

Kubernetes: v1.16.1 [current v1.16.1]

Node exporter: v0.18.1 [current v0.18.1]

Kube-state-metrics: v1.8.0 [current v1.8.0]

Prometheus: v2.13.0 [current v2.13.0]

Grafana: v5.0.0-5.0.4 [current v6.4.2] *explained below

All private

Namespace: monprivate

Folder: manifests_monprivate

All exposed

Namespace: monexposed

Folder: manifests_monexposed

Running:

git clone https://github.com/andre-lx/grafana-prometheus-exporters.git
cd grafana-prometheus-exporters/manifests_monXXXXX/

kubectl apply -f .
kubectl apply -f prometheus/kube-state-metrics/
kubectl apply -f prometheus/node-exporter/
kubectl apply -f prometheus/
kubectl apply -f grafana/

Grafana Kubernetes APP

Install

This manifests will also install the official grafana kubernetes plugin (not updated for a while). In order to use my fork (with the updated dashboards and metrics to match the new versions of the exporters), install git and git clone my fork (pending pull request):

kubectl --namespace=monXXXXX exec -it pod/grafana-core-xxxxxx -- /bin/bash
cd /var/lib/grafana/plugins/
mv grafana-kubernetes-app/ grafana-kubernetes-app.old

apt-get update 
apt-get install git

git clone https://github.com/andre-lx/kubernetes-app.git grafana-kubernetes-app

Open grafana -> Plugin config -> Dashboards -> Re-import all

The official plugin works in the most recent grafana versions, but the custom pannels created for version 5.0.0 starts to fail in some functionalities for versions 5.0.4+.

Config services

Prometheus:

Cluster config:

Do not use the deploy button present in the plugin, since you already have all the needed exporters on the system

Accessing

monprivate:

Grafana:
kubectl port-forward --address=0.0.0.0 pod/grafana-core-xxxx 5000:3000 --namespace=monprivate

Prometheus:
kubectl port-forward --address=0.0.0.0 pod/prometheus-core-xxxx 5001:9090 --namespace=monprivate

Access using localhost:5000 and localhost:5001

monexposed:

For node exporters:

http://node_ip or master_node_ip:check port on get

For grafana, prometheus and kube-state-metrics exporter:

http://any_node_ip:check port on get

Implementation result:

[root@k8s-master ~]# kubectl get all --namespace=monprivate
NAME                                                           
pod/grafana-core-6754fb48f8-zdgtg         
pod/kube-state-metrics-55fdc45f47-btwrx  
pod/prometheus-core-bf84cb6bb-5hfnk       
pod/prometheus-node-exporter-ntqm6      
pod/prometheus-node-exporter-asw22 
pod/prometheus-node-exporter-asd2g 
pod/prometheus-node-exporter-ertzs 

NAME                               TYPE        CLUSTER-IP       EXTERNAL-IP   PORT(S)    
service/grafana                    ClusterIP   10.101.192.15    <none>        3000/TCP   
service/kube-state-metrics         ClusterIP   10.110.250.218   <none>        8080/TCP   
service/prometheus                 ClusterIP   10.99.61.63      <none>        9090/TCP   
service/prometheus-node-exporter   ClusterIP   10.106.215.249   <none>        9100/TCP   

NAME                                      DESIRED   CURRENT   READY   UP-TO-DATE   AVAILABLE   NODE SELECTOR   
daemonset.apps/prometheus-node-exporter   4         4         4       4            4           <none>          

NAME                                 READY   UP-TO-DATE   AVAILABLE   
deployment.apps/grafana-core         1/1     1            1           
deployment.apps/kube-state-metrics   1/1     1            1           
deployment.apps/prometheus-core      1/1     1            1           

NAME                                            DESIRED   CURRENT   READY  
replicaset.apps/grafana-core-6754fb48f8         1         1         1       
replicaset.apps/kube-state-metrics-55fdc45f47   1         1         1       
replicaset.apps/prometheus-core-bf84cb6bb       1         1         1       

[root@k8s-master ~]# kubectl get all --namespace=monexposed
NAME                                      
pod/grafana-core-6754fb48f8-5cjbk         
pod/kube-state-metrics-55fdc45f47-zmskt   
pod/prometheus-core-bf84cb6bb-pj2k6       
pod/prometheus-node-exporter-24mjd        
pod/prometheus-node-exporter-bcgsz 
pod/prometheus-node-exporter-bfdsz 
pod/prometheus-node-exporter-dajqa

NAME                               TYPE       CLUSTER-IP      EXTERNAL-IP   PORT(S)          
service/grafana                    NodePort   10.102.38.125   <none>        3000:32348/TCP   
service/kube-state-metrics         NodePort   10.109.35.231   <none>        8080:32137/TCP   
service/prometheus                 NodePort   10.105.164.24   <none>        9090:30975/TCP   
service/prometheus-node-exporter   NodePort   10.105.120.38   <none>        9100:30958/TCP   

NAME                                      DESIRED   CURRENT   READY   UP-TO-DATE   AVAILABLE   NODE SELECTOR   
daemonset.apps/prometheus-node-exporter   4         4         4       4            4           <none>          

NAME                                 READY   UP-TO-DATE   AVAILABLE   
deployment.apps/grafana-core         1/1     1            1           
deployment.apps/kube-state-metrics   1/1     1            1           
deployment.apps/prometheus-core      1/1     1            1           

NAME                                            DESIRED   CURRENT   READY   
replicaset.apps/grafana-core-6754fb48f8         1         1         1       
replicaset.apps/kube-state-metrics-55fdc45f47   1         1         1       
replicaset.apps/prometheus-core-bf84cb6bb       1         1         1       

---

---

---

Credits:

This is an original fork from: giantswarm/prometheus