Overview

This document provides configuration steps for using the Datadog Agent to send logs to a Datadog CloudPrem deployment. Unlike the Datadog SaaS platform, CloudPrem requires specific Agent configurations to ensure logs are enriched with necessary host-level tags and sent to the correct endpoint. This guide covers how to set these configurations for the most common deployment methods.

Key requirements

To send logs with the Datadog Agent to CloudPrem, you must configure two environment variables:

DD_LOGS_CONFIG_LOGS_DD_URL

Set this to your CloudPrem indexer endpoint, usually http://<RELEASE_NAME>-indexer.<NAMESPACE_NAME>.svc.cluster.local:7280. This tells the Agent where to send the logs

DD_LOGS_CONFIG_EXPECTED_TAGS_DURATION

(Optional) This is an optional but highly recommended variable. Set it to a large value, like “100000” (approximately 5 years). This ensures the Agent adds host-level tags to every log it sends. The Datadog SaaS platform automatically enriches logs with these tags after ingestion, but CloudPrem requires the Agent to add them upfront.

Send Kubernetes logs with the Datadog Operator

To deploy the Agent on Kubernetes using the Datadog Operator, follow the Getting Started with Datadog Operator guide. When you reach Step 3, use the following datadog-agent.yaml configuration instead of the example provided in the guide.

apiVersion: datadoghq.com/v2alpha1
kind: DatadogAgent
metadata:
  name: datadog
spec:
  global:
    clusterName: <CLUSTER_NAME>
    site: datadoghq.com
    credentials:
      apiSecret:
        secretName: datadog-secret
        keyName: api-key
    env:
      - name: DD_LOGS_CONFIG_LOGS_DD_URL
        value: http://<RELEASE_NAME>-indexer.<NAMESPACE_NAME>.svc.cluster.local:7280
      - name: DD_LOGS_CONFIG_EXPECTED_TAGS_DURATION
        value: "100000"

  features:
    logCollection:
      enabled: true
      containerCollectAll: true

    otlp:
      receiver:
        protocols:
          grpc:
            enabled: true
            endpoint: 0.0.0.0:4417

    prometheusScrape:
      enabled: true
      enableServiceEndpoints: true

Configuration options

Endpoint configuration

The Datadog Agent can be configured to send logs to CloudPrem using different endpoints:

DD_LOGS_CONFIG_LOGS_DD_URL=http://<RELEASE_NAME>-indexer.<NAMESPACE_NAME>.svc.cluster.local:7280

DD_LOGS_CONFIG_LOGS_DD_URL=https://cloudprem-internal.your-domain.com

Additional Agent configuration

You can also configure additional features to send cluster metadata to Datadog:

features:
  prometheusScrape:
    enabled: true
    enableServiceEndpoints: true

features:
  otlp:
    receiver:
      protocols:
        grpc:
          enabled: true
          endpoint: 0.0.0.0:4417

Alternative deployment methods

If you are not using the Datadog Operator, you can deploy the Agent using one of these common methods:

Helm chart deployment

Run the following command to deploy the Agent using the Helm chart, setting the log-specific environment variables directly.

helm install datadog-agent datadog/datadog \
  --set datadog.apiKey=<YOUR_API_KEY> \
  --set datadog.logs.enabled=true \
  --set datadog.logs.containerCollectAll=true \
  --set datadog.logsConfigContainerCollectAll=true \
  --set agents.containers.agent.env[0].name=DD_LOGS_CONFIG_LOGS_DD_URL \
  --set agents.containers.agent.env[0].value=http://<RELEASE_NAME>-indexer.<NAMESPACE_NAME>.svc.cluster.local:7280

DaemonSet deployment

For custom deployments, set the environment variable in your DaemonSet:

apiVersion: apps/v1
kind: DaemonSet
metadata:
  name: datadog-agent
spec:
  template:
    spec:
      containers:
      - name: agent
        image: gcr.io/datadoghq/agent:latest
        env:
        - name: DD_API_KEY
          value: <YOUR_API_KEY>
        - name: DD_LOGS_ENABLED
          value: "true"
        - name: DD_LOGS_CONFIG_CONTAINER_COLLECT_ALL
          value: "true"
        - name: DD_LOGS_CONFIG_LOGS_DD_URL
          value: "http://<RELEASE_NAME>-indexer.<NAMESPACE_NAME>.svc.cluster.local:7280"

Verification

After the Agent is deployed, you can verify that logs are being sent and received correctly.

Check Agent status

Use kubectl exec to check the Agent’s status and confirm it is configured to send logs.

# Check Agent status and logs configuration
kubectl exec -it <datadog-agent-pod> -- agent status | grep -A 10 "Logs Agent"

# Check Agent logs for CloudPrem connection
kubectl logs <datadog-agent-pod> | grep -i cloudprem

Check logs are indexed in CloudPrem

Run this command to query the CloudPrem searcher and verify that it’s indexing the JSON logs.

kubectl exec -it <RELEASE_NAME>-searcher-0 -n <NAMESPACE_NAME> -- curl 'http://localhost:7280/api/v1/datadog/search?query='

Troubleshooting

Agent not sending logs:

• Verify the DD_LOGS_CONFIG_LOGS_DD_URL environment variable is set correctly
• Check Agent pod logs: kubectl logs <datadog-agent-pod>
• Ensure log collection is enabled: DD_LOGS_ENABLED=true

CloudPrem not receiving logs:

• Check CloudPrem indexer logs: kubectl logs -n <NAMESPACE_NAME> -l app=<RELEASE_NAME>-indexer
• Verify network connectivity between Agent and CloudPrem indexer
• Confirm CloudPrem service is running: kubectl get pods -n <NAMESPACE_NAME>

Further reading

추가 유용한 문서, 링크 및 기사:

Observability Pipelines IntegrationDOCUMENTATION REST API IntegrationDOCUMENTATION Datadog Operator GuideDOCUMENTATION