Skip to main content

Deploy Datadog Operator and DatadogAgent on SleakOps

Install the Datadog Operator and deploy a DatadogAgent custom resource on a SleakOps EKS cluster to enable APM, log collection, and Kubernetes event monitoring.

info

This guide was tested with DatadogOperator v2.8.0 and DatadogAgent v7.63.3.

Prerequisites​

In the cluster:

  • Create the datadog namespace
  • Create the datadog-secret Secret containing your Datadog API key:
kubectl create secret generic datadog-secret \
  --from-literal=api-key=<YOUR_DATADOG_API_KEY> \
  -n datadog

Optional but recommended: Create a dedicated NodePool for Datadog (e.g., datadog-app). If you use a dedicated NodePool, workloads you want to monitor must also run on that NodePool.

Step 1 — Deploy the Datadog Operator​

helm upgrade -i datadog datadog/datadog-operator \
  --namespace datadog \
  --create-namespace \
  --set "tolerations[0].key=karpenter.sh/nodepool" \
  --set "tolerations[0].operator=Equal" \
  --set "tolerations[0].value=<NODEPOOL-NAME>"

Replace <NODEPOOL-NAME> with your NodePool name (e.g., datadog-app-amd).

Step 2 — Prepare Workloads for Instrumentation​

Add these annotations/labels to the Deployment spec of each workload you want Datadog to instrument:

spec:
  template:
    metadata:
      labels:
        admission.datadoghq.com/enabled: "true"
      annotations:
        admission.datadoghq.com/js-lib.version: v5.45.0

In the workload's Var Group, add:

DD_LOGS_ENABLED=true
DD_LOGS_INJECTION=true
DD_ENV=<your-env>
DD_SERVICE=<your-service-name>

Step 3 — Deploy the DatadogAgent​

Create datadogagent.yaml:

apiVersion: datadoghq.com/v2alpha1
kind: DatadogAgent
metadata:
  name: datadog
  namespace: datadog
spec:
  global:
    clusterName: <CLUSTER-NAME>
    site: datadoghq.com    # see https://docs.datadoghq.com/getting_started/site/
    credentials:
      apiSecret:
        secretName: datadog-secret
        keyName: api-key
  features:
    admissionController:
      enabled: true
      mutateUnlabelled: false
    apm:
      enabled: true
      hostPortConfig:
        enabled: true
        hostPort: 8126
      instrumentation:
        enabled: false
    logCollection:
      enabled: true
      containerCollectAll: true
    eventCollection:
      collectKubernetesEvents: true
  override:
    clusterAgent:
      tolerations:
        - key: karpenter.sh/nodepool
          operator: Equal
          value: <NODEPOOL-NAME>
    nodeAgent:
      env:
        - name: DD_LOGS_CONFIG_AUTO_MULTI_LINE_DETECTION
          value: "true"
        - name: DD_PROFILING_ENABLED
          value: "true"
        - name: DD_TRACE_ENABLED
          value: "true"
        - name: DD_LOGS_ENABLED
          value: "true"
        - name: DD_LOGS_INJECTION
          value: "true"
        - name: DD_CONTAINER_EXCLUDE
          value: "name:.*"
        - name: DD_CONTAINER_INCLUDE
          value: "name:<container-name>"    # match spec.template.spec.containers.name
      tolerations:
        - key: karpenter.sh/nodepool
          operator: Equal
          value: <NODEPOOL-NAME>

Apply it:

kubectl apply -f datadogagent.yaml -n datadog

Step 4 — Verify the Deployment​

kubectl get datadogagent -n datadog
kubectl get pods -n datadog

All pods should reach Running state within a few minutes.

tip

This is a reference configuration that can be adjusted for your environment — expand DD_CONTAINER_INCLUDE to cover more workloads or modify any feature flags as needed.