Why do we need datadog Agent ? - harishgorla5/HINTechnologies_datadog GitHub Wiki

🐶 Datadog Agent & Helm vs YAML in Kubernetes

📌 Why Do We Need the Datadog Agent?

The Datadog Agent is a lightweight process that runs on each node of your Kubernetes (EKS) cluster. It collects and sends observability data (metrics, logs, traces) to Datadog.

✅ Features of the Datadog Agent

Feature	Description
Metrics Collection	Captures system-level metrics like CPU, memory, disk, container stats.
Log Collection	Gathers logs from containers, services, and host system.
APM (Application Performance Monitoring)	Captures distributed traces from microservices.
Process Monitoring	Tracks running processes and container lifecycles.
Autodiscovery	Automatically detects services like Apache, MySQL, etc.
Cluster Agent	Aggregates data from multiple agents and enhances performance.
Kubernetes Integration	Adds pod, service, and cluster-level observability using kubelet and labels.

Without the Datadog Agent, your Kubernetes cluster would not send any telemetry to Datadog — making it invisible in dashboards or monitors.

🎯 Why Use Helm Instead of Raw YAML?

Datadog provides both:

A Helm Chart for automated, configurable deployment.
YAML manifests, but they’re large and harder to maintain.

✅ Benefits of Using Helm

Feature	Helm	Raw YAML Files
Templating	Parameterized templates via `values.yaml`	Manually modify all fields
Upgrade-friendly	`helm upgrade` cleanly updates version/config	Must reapply full YAML set
Reusability	Can reuse same chart with different values	Must copy & edit YAML for each
Config management	All config stored in `values.yaml`	Spread across multiple YAMLs
Cleanup	`helm uninstall` removes all resources	Must delete each manually
Default Best Practices	Follows Datadog’s recommended setup	Must build from scratch

⚙️ Example: Install via Helm

helm install datadog-agent datadog/datadog   --set datadog.apiKeyExistingSecret=datadog-secret   --set datadog.site='us5.datadoghq.com'   --set datadog.logs.enabled=true   --set datadog.apm.portEnabled=true   --set datadog.processAgent.enabled=true   --set datadog.kubeStateMetricsEnabled=true   --set datadog.leaderElection=true   --set clusterAgent.enabled=true   --set clusterAgent.metricsProvider.enabled=true   --namespace default

👇 Equivalent YAML (Example Only - Would Require 100s of Lines)

apiVersion: apps/v1
kind: DaemonSet
metadata:
  name: datadog-agent
  namespace: default
spec:
  template:
    spec:
      containers:
      - name: agent
        image: gcr.io/datadog/agent:latest
        env:
        - name: DD_API_KEY
          valueFrom:
            secretKeyRef:
              name: datadog-secret
              key: api-key
        # Other configurations...

Writing full YAML for:

DaemonSet
Cluster Agent
RBAC
ConfigMaps
Secrets
Admission Controllers

…can be very complex and error-prone compared to using a Helm chart.

✅ Summary

Category	Recommendation
Install Agent	✅ Use Helm
Collect Metrics	✅ Enable in Helm Chart
Enable APM, Logs	✅ Use Helm flags
Manage Config	✅ Use `values.yaml`
Avoid Manual Errors	✅ Use Helm