If You have 20 pods running a microservices app, and your manager asks for a live dashboard showing CPU/memory usage of each pod. How do you make this available in Grafana?

If I have 20 pods running a microservices application and my manager asks for a live dashboard showing the CPU and memory usage of each pod, I would use Grafana connected to Prometheus to create that view.

First, I make sure that Prometheus is already deployed and properly scraping metrics from the Kubernetes cluster — specifically from the kubelet and cAdvisor endpoints, which expose pod-level resource usage data.

Then, in Grafana, I connect Prometheus as the data source if it’s not already linked.

After that, I either create a custom dashboard or import a pre-built Kubernetes monitoring dashboard from Grafana’s community library. These dashboards typically include visual panels that show real-time CPU and memory usage at the pod level.

To focus on just those 20 pods from my microservices app, I can filter the dashboard using labels, namespaces, or deployment names — whichever uniquely identifies those pods. This gives the manager a clean, focused view of only the pods related to that application.

The dashboard will continuously update in real time, and I can also configure it to show trends over time, alerts for resource spikes, or any anomalies in pod behavior.

So in summary, I rely on Prometheus for collecting the metrics and Grafana for visualizing them — filtered specifically for the 20 pods that matter to the team.