To send Datadog Agent logs to the Observability Pipelines Worker, update your Agent configuration file with the following:

observability_pipelines_worker:
  logs:
    enabled: true
    url: "http://<OPW_HOST>:8282"

<OPW_HOST> is the host IP address or the load balancer URL associated with the Observability Pipelines Worker.

For CloudFormation installs, use the LoadBalancerDNS CloudFormation output for the URL.

For Kubernetes installs, you can use the internal DNS record of the Observability Pipelines Worker service. For example: http://opw-observability-pipelines-worker.default.svc.cluster.local:<PORT>.

Optionally, to ensure host tags are included in your logs, add the following to your Agent configuration file:

logs_config:
  expected_tags_duration: "1000000m"

expected_tags_duration ensures the Agent attaches host-level tags (such as EC2 instance tags) directly to each log event sent to Observability Pipelines. Without this setting, these tags are applied at the Datadog intake level and are not present in the log payload when it passes through Observability Pipelines. Setting the duration to 1000000m (approximately 1.5 years) covers the standard lifecycle of a node. Note that tags may increase the size of your log.

After you restart the Agent, your observability data should be going to the Worker, processed by the pipeline, and delivered to Datadog.

To send Datadog Agent logs to the Observability Pipelines Worker, update your Datadog Helm chart datadog-values.yaml with the following environment variables. See Agent Environment Variables for more information.

datadog:
  env:
    - name: DD_OBSERVABILITY_PIPELINES_WORKER_LOGS_ENABLED
      value: true
    - name: DD_OBSERVABILITY_PIPELINES_WORKER_LOGS_URL
      value: "http://<OPW_HOST>:8282"

<OPW_HOST> is the IP/URL of the host (or load balancer) associated with the Observability Pipelines Worker.

For Kubernetes installs, you can use the internal DNS record of the Observability Pipelines Worker service. For example: http://opw-observability-pipelines-worker.default.svc.cluster.local:<PORT>.

Optionally, to ensure host tags are included in your logs, add the following to your Agent configuration file:

    - name: DD_LOGS_CONFIG_EXPECTED_TAGS_DURATION
      value: "1000000m"

DD_LOGS_CONFIG_EXPECTED_TAGS_DURATION ensures the Agent attaches host-level tags (such as EC2 instance tags) directly to each log event sent to Observability Pipelines. Without this setting, these tags are applied at the Datadog intake level and are not present in the log payload when it passes through Observability Pipelines. Setting the duration to 1000000m (approximately 1.5 years) covers the standard lifecycle of a node. Note that tags may increase the size of your log.

Datadog Agent 메트릭을 Observability Pipelines Worker로 전송하려면 Agent 구성 파일을 다음 내용으로 업데이트합니다.

observability_pipelines_worker:
  metrics:
    enabled: true
    url: "http://<OPW_HOST>:8383"

<OPW_HOST>는 Observability Pipelines Worker와 연결된 호스트 IP 주소 또는 로드 밸런서 URL입니다.

• CloudFormation 설치의 경우 LoadBalancerDNS CloudFormation 출력을 URL로 사용합니다.
• Kubernetes 설치의 경우 Observability Pipelines Worker 서비스의 내부 DNS 레코드를 사용할 수 있습니다. 예: http://opw-observability-pipelines-worker.default.svc.cluster.local:<PORT>.

참고: Worker가 포트 8282에서 로그를 수신 대기 중인 경우, 메트릭에는 8383과 같은 다른 포트를 사용해야 합니다.

Agent를 재시작하면 관측 가능성 데이터가 Worker로 전송되고, 파이프라인에서 처리된 후 Datadog으로 전달됩니다.

Datadog Agent 메트릭을 Observability Pipelines Worker로 전송하려면 Datadog Helm 차트 datadog-values.yaml을 다음 환경 변수로 업데이트합니다. 자세한 내용은 Agent 환경 변수를 참조하세요.

datadog:
  env:
    - name: DD_OBSERVABILITY_PIPELINES_WORKER_METRICS_ENABLED
      value: true
    - name: DD_OBSERVABILITY_PIPELINES_WORKER_METRICS_URL
      value: "http://<OPW_HOST>:8383"

<OPW_HOST>는 Observability Pipelines Worker와 연결된 호스트 IP 주소 또는 로드 밸런서 URL입니다.

Kubernetes 설치의 경우 Observability Pipelines Worker 서비스의 내부 DNS 레코드를 사용할 수 있습니다. 예: http://opw-observability-pipelines-worker.default.svc.cluster.local:<PORT>.

참고: Worker가 포트 8282에서 로그를 수신 대기 중인 경우, 메트릭에는 8383과 같은 다른 포트를 사용해야 합니다.