How do I configure PodMonitor to scrape metrics from labeled pods?

In this guide, we will configure a PodMonitor in Kubernetes to scrape metrics from pods that have specific labels using Pulumi. The PodMonitor custom resource is part of the Prometheus Operator, which allows you to define how Prometheus should scrape metrics from pods.

Detailed Explanation

1. Install Pulumi and the necessary plugins: Ensure you have Pulumi installed and the Kubernetes plugin set up.
2. Define the Kubernetes provider: Configure the provider to connect to your Kubernetes cluster.
3. Create a Namespace: Define a namespace where the resources will be deployed.
4. Deploy a Pod with Specific Labels: Create a pod that has the labels Prometheus will use to identify it.
5. Create a PodMonitor: Define the PodMonitor resource to specify how Prometheus should scrape metrics from the pods with the defined labels.

Key Points

• PodMonitor: Custom resource to configure Prometheus scraping from pods.
• Labels: Used to select which pods Prometheus should scrape.
• Namespace: Logical grouping of resources in Kubernetes.

Concluding Summary

We configured a PodMonitor to scrape metrics from pods with specific labels in a Kubernetes cluster using Pulumi. This setup ensures that Prometheus can automatically discover and scrape metrics from the desired pods.

import * as pulumi from "@pulumi/pulumi";
import * as k8s from "@pulumi/kubernetes";

// Create a Kubernetes namespace
const namespace = new k8s.core.v1.Namespace("monitoring", {
    metadata: { name: "monitoring" },
});

// Deploy a pod with specific labels
const appPod = new k8s.core.v1.Pod("appPod", {
    metadata: {
        namespace: namespace.metadata.name,
        labels: { app: "myapp" },
    },
    spec: {
        containers: [{
            name: "app-container",
            image: "nginx",
            ports: [{ containerPort: 80 }],
        }],
    },
}, { dependsOn: [namespace] });

// Create a PodMonitor to scrape metrics from pods with the label "app=myapp"
const podMonitor = new k8s.apiextensions.CustomResource("podMonitor", {
    apiVersion: "monitoring.coreos.com/v1",
    kind: "PodMonitor",
    metadata: {
        namespace: namespace.metadata.name,
        name: "myapp-pod-monitor",
    },
    spec: {
        selector: {
            matchLabels: { app: "myapp" },
        },
        namespaceSelector: {
            matchNames: [namespace.metadata.name],
        },
        podMetricsEndpoints: [{
            port: "metrics",
        }],
    },
}, { dependsOn: [namespace, appPod] });

In this code:

• A namespace monitoring is created.
• A pod labeled app=myapp is deployed in the monitoring namespace.
• A PodMonitor is created to scrape metrics from pods with the label app=myapp in the monitoring namespace.