How to Integrate Datadog with Kubernetes for Real-Time Monitoring (Step by Step)

How to Integrate Datadog with Kubernetes for Real-Time Monitoring

We’re building a Datadog Kubernetes integration to monitor the performance of our applications in real-time, which is crucial for maintaining application availability and performance.

Prerequisites

• Kubernetes 1.20+
• Datadog Account
• kubectl installed on your machine
• Helm 3+ installed
• Access to your Kubernetes cluster

Step 1: Signing up for Datadog

First things first, if you don’t have a Datadog account yet, you need to create one. This is mandatory because, well, Datadog won’t let you monitor anything if you don’t.

# Go to Datadog and sign up
open https://www.datadoghq.com/

After you create an account, you’ll need to get your Datadog API key. You can find it in the API section of your Datadog dashboard.

Step 2: Install the Datadog Agent using Helm

Now, it’s time to install the Datadog Agent in your Kubernetes cluster using Helm. If you’re not using Helm, you might need to switch to it — managing Kubernetes apps manually is seriously painful.

# Add the Datadog Helm repository
helm repo add datadog https://helm.datadoghq.com

# Update your Helm repo
helm repo update

# Install the Datadog Agent with the required configuration
helm install datadog datadog/datadog \
 --set api.key= \
 --set agents.image.tag= \
 --set cluster.agent.enabled=true

Replace `` with the API key you got earlier and `` with the Datadog Agent version you want to install (like `7.30.0`). You can check the latest version from Datadog’s official documentation.

You may run into permission issues if you’re not allowed to create resources in the default namespace. In that case, specify the namespace with: --namespace in the Helm install command.

Step 3: Verifying the Installation

After the Helm installation, check if the Datadog Agent pods are running in your cluster.

# Get all pods in the default namespace
kubectl get pods --namespace default

Look for pods with names that start with `datadog-agent`. If they are in the `Running` status, congratulations! If you see an `Error` status, check the logs for more information:

# Get logs of a specific agent pod
kubectl logs  --namespace default

Errors here can include image pull errors or permission issues, usually straightforward to resolve. Just make sure your service account has the right permissions if there’s an RBAC error.

Step 4: Enable Kubernetes Integration in Datadog

This step is all about connecting your Kubernetes cluster to your Datadog account. Without this, you won’t see any metrics from Kubernetes in your Datadog dashboard.

# Modify the Helm command with the following options
helm install datadog datadog/datadog \
 --set api.key= \
 --set apm.enabled=true \
 --set kubelet.enabled=true \
 --set logs.enabled=true \
 --set clusterAgent.enabled=true

This command enables APM, Kubelet, and logs. Each of these features provides essential insights into your cluster’s health and performance. If you run into issues with APM not showing data, make sure your services are instrumented properly.

The Gotchas

Here’s where things get tricky. Tutorials often gloss over these pain points. You’ve been warned:

• Resource Limits: Without defining resource requests and limits for the agent, you’ll encounter OOM issues as your cluster scales. The agent can demand resources as it processes logs and metrics.
• Network Policies: If you’re using network policies, ensure you allow traffic from the Datadog Agent to the Datadog backend. Missing outbound rules can result in data not being sent.
• Service Discovery: Make sure your service discovery tags are correctly set up, or else the agents may fail to report the relevant service metrics.
• Avoiding Helm Errors: Helm errors can pop up for a variety of reasons—from incorrect chart versions to outdated API versions. Always use the `–debug` flag to get detailed output.

Full Code: Putting It All Together

Once you get everything running, your commands will look something like this:

# Sign up at Datadog
open https://www.datadoghq.com/

# Add Helm repo
helm repo add datadog https://helm.datadoghq.com
helm repo update

# Install the Datadog Agent
helm install datadog datadog/datadog \
 --set api.key= \
 --set agents.image.tag= \
 --set cluster.agent.enabled=true \
 --set apm.enabled=true \
 --set kubelet.enabled=true \
 --set logs.enabled=true

While this may seem like a lot, once you get your environment set up, most of this stuff becomes second nature. I remember the first time I tried to get Datadog integrated—let’s just say I ended up with more than a few broken Helm installs.

What’s Next?

After deploying Datadog with your Kubernetes integration, your next concrete step is to set up monitors and alerts based on the logs and metrics you’re collecting. This will take your monitoring game to the next level—ensure you don’t sleep through an application outage.

FAQ

• Q1: Can I monitor multiple clusters with a single Datadog account?
  A1: Yes, as long as you configure each cluster with a unique Datadog Agent and API key.
• Q2: Do I need a paid plan for full features?
  A2: Some features like APM and log monitoring may require a paid Datadog plan, so check what service level you’ll need.
• Q3: What kind of metrics should I monitor?
  A3: Focus on CPU, memory usage, and pod status. Also, keep an eye on network traffic and storage usage for a holistic view.

Data Sources

For full details and any other questions regarding Datadog, visit the official Datadog documentation and the Kubernetes section of their site here.

Last updated May 06, 2026. Data sourced from official docs and community benchmarks.

🕒 Published: May 6, 2026