KUBERNETES MONITORING PART2

Step 1: Deploying Prometheus for Kubernetes Monitoring

1. Access the Monitoring Directory:

• Navigate to the “monitoring” directory where all the Prometheus deployment files are located.

 

2. Create Prometheus Resources:

• Execute the following command:

$ kubectl create -f prometheus

• This action sets up the Prometheus pod, services, and deployment within the “monitoring” Namespace.

 

3. Access Prometheus Web Interface:

• Use the following syntax in a web browser:

http://<public-ip-of-worker-node>:32300

• You’ll be able to view the Prometheus interface.

 

Remember to replace <public-ip-of-worker-node> with the actual public IP address of your worker node.

When reviewing the output of Kubernetes queries, it’s evident that the information is presented in a non-graphical format, requiring careful examination to understand the deployment status and identify any potential issues with pods or deployments.

Enter Grafana! Grafana offers an intuitive dashboard interface that transforms raw metrics data into visual representations. With Grafana, you can monitor the health and performance of your Kubernetes cluster through charts, graphs, and other visual elements. It simplifies the process of identifying trends, anomalies, and bottlenecks, making it an essential tool for efficient cluster management. 

Step2: Deploying Grafana for Kubernetes Monitoring

In this step, we’ll deploy Grafana to enhance our Kubernetes monitoring capabilities. Follow these instructions:

1. Deployment Files:

• Locate the Grafana deployment files in the “monitoring” directory.
• Execute the following command to create Grafana resources:

kubectl create -f grafana/

• This command sets up Grafana pods, services, and deployments within the “monitoring” namespace.

 

2. Verify Grafana Deployment:

• Query the “monitoring” namespace to check Grafana’s status:

kubectl get all -n monitoring

• You should see output similar to the following:

NAME                                      READY   STATUS    RESTARTS   AGE
pod/alertmanager-54f9b89f9c-smhct         1/1     Running   0          28h
pod/grafana-5844bc979c-t9srg              1/1     Running   0          2d1h
pod/prometheus-deployment-f9fb8d846-592pm 1/1     Running   0          2d1h

NAME                                      TYPE      CLUSTER-IP       EXTERNAL-IP   PORT(S)           AGE
service/alertmanager                      NodePort  10.96.109.24      <none>        9093:31000/TCP    28h
service/grafana                           NodePort  10.105.161.41    <none>        3000:32500/TCP    2d1h
service/prometheus-service                NodePort  10.107.155.121   <none>        8080:32300/TCP    2d1h

NAME                                      READY   UP-TO-DATE   AVAILABLE   AGE
deployment.apps/alertmanager              1/1     1            1           28h
deployment.apps/grafana                   1/1     1            1           2d1h
deployment.apps/prometheus-deployment     1/1     1            1           2d1h

NAME                                      DESIRED   CURRENT   READY   AGE
replicaset.apps/alertmanager-54f9b89f9c    1         1         1       28h
replicaset.apps/grafana-5844bc979c         1         1         1       2d1h
replicaset.apps/prometheus-deployment-f9fb8d846 1         1         1       2d1h

 

3. Namespace Check:

• Confirm that Alert Manager, Prometheus, and Grafana are running in the “monitoring” namespace.
• Additionally, check if “kube-metrics” and “metrics-server” are operational in the “kube-system” namespace:

kubectl get all -n kube-system

 

4. Access Grafana:

• Grafana is exposed on port 32500.
• To access Grafana via a web browser, use the following syntax:

http://<public-ip-of-worker-node>:32500