Day 20 of #100DaysOfCloud, Focusing on integrating Grafana with Prometheus on a Kubernetes cluster 🎯📊

Cloud-native monitoring isn’t just about collecting metrics—it’s about visualizing them in a meaningful way. Today, I explored how to integrate Grafana with Prometheus using a Kubernetes lab from KillerKoda, a great free platform for hands-on learning.

This integration helps you build real-time dashboards to monitor pods, nodes, and workloads, which is critical in any DevOps or SRE role.

🔧 Step-by-Step: Grafana + Prometheus Integration using KillerKoda

🧪 You can follow along using KillerKoda’s interactive Kubernetes terminal at https://killerkoda.com

✅ 1. Install the Prometheus Operator

The Prometheus Operator makes it easier to deploy and manage Prometheus instances on Kubernetes.

This installs the Custom Resource Definitions (CRDs), RBAC, and deployments required for the operator.

✅ 2. Configure RBAC Permissions

Ensure the Prometheus operator has the right cluster role bindings and service accounts to access Kubernetes metrics:

Apply it with:

✅ 3. Deploy Prometheus using the Operator

Next, create a Prometheus instance using a custom resource:

Apply:

✅ 4. Expose Prometheus as a Service

You can expose Prometheus using a NodePort or LoadBalancer (if supported):

After forwarding the port in killercoda to 9090, you can access it locally from the access port in it and it will open the Prometheus UI

✅ 5. Create a ServiceMonitor

The ServiceMonitor tells Prometheus what services to scrape:

📈 Grafana Integration

Once Prometheus is running:

• Install Grafana (e.g., via Helm):

• Access Grafana UI and add Prometheus as a data source:

• Import dashboards or build custom ones using PromQL.

✅ From My Experience

In my cloud engineering experience , we set up Prometheus and Grafana for AKS clusters to monitor:

• Node CPU/memory usage

• Pod restarts and failure rates

• HTTP request latencies using Ingress metrics

This helped our SRE teams detect anomalies in real-time and reduce mean time to resolution (MTTR).

🧠 Takeaway:

Learning how to set up Prometheus + Grafana on Kubernetes helps you monitor critical systems effectively. Using KillerKoda makes it easier for beginners to try this in a safe, sandboxed environment.

#100DaysOfCloud #Kubernetes #DevOps #Grafana #Prometheus #CloudNative #Monitoring #KillerKoda #Azure #SRE #AKS