Datadog

Oodle ingests metrics, logs, and traces (APM) from the Datadog agent. This guide covers all supported signals and configuration methods.

The quickest way to get started is through the Datadog integration tile on the Integrations page — it generates the exact configuration for your instance.

Supported signals

Signal	Dual-ship env var	Single-ship env var
Metrics	`DD_ADDITIONAL_ENDPOINTS`	`DD_DD_URL` + `DD_API_KEY`
Logs	`DD_LOGS_CONFIG_ADDITIONAL_ENDPOINTS`	`DD_LOGS_CONFIG_DD_URL`
Traces	`DD_APM_ADDITIONAL_ENDPOINTS`	`DD_APM_DD_URL`

note

The endpoint URL and API key for your instance are shown in the Datadog tile on the Settings → Integrations page.

note

Wherever metrics are shipped to Oodle, set DD_USE_V3_API_SERIES_ENABLED=false (agent config key use_v3_api.series.enabled: false) so the Agent submits series on the v2 metrics intake that Oodle accepts. This setting is included in every metrics configuration below.

Dual-shipping (Datadog + Oodle)

Send data to both Datadog and Oodle simultaneously.

Datadog Helm chart

Add the following to your Helm values.yaml for the Datadog Helm chart:

datadog:
  env:
    # Metrics
    - name: DD_ADDITIONAL_ENDPOINTS
      value: '{"<oodle_metrics_endpoint>": ["<oodle_api_key>"]}'
    - name: DD_USE_V3_API_SERIES_ENABLED
      value: 'false'
    # Logs
    - name: DD_LOGS_CONFIG_ADDITIONAL_ENDPOINTS
      value: '[{"api_key": "<oodle_api_key>", "Host": "<oodle_logs_host>", "Port": 443, "is_reliable": false}]'
    - name: DD_LOGS_CONFIG_FORCE_USE_HTTP
      value: 'true'
    # APM / Traces
    - name: DD_APM_ADDITIONAL_ENDPOINTS
      value: '{"<oodle_traces_endpoint>": ["<oodle_api_key>"]}'

clusterAgent:
  env:
    # Metrics (cluster agent also needs this)
    - name: DD_ADDITIONAL_ENDPOINTS
      value: '{"<oodle_metrics_endpoint>": ["<oodle_api_key>"]}'
    - name: DD_USE_V3_API_SERIES_ENABLED
      value: 'false'

Then install or upgrade:

helm repo add --force-update datadog \
  https://helm.datadoghq.com && \
helm upgrade --install datadog-agent datadog/datadog \
  --values values.yaml

Datadog Operator

Apply the following DatadogAgent resource:

apiVersion: datadoghq.com/v2alpha1
kind: DatadogAgent
metadata:
  name: datadog

spec:
  override:
    nodeAgent:
      containers:
        agent:
          env:
            # Metrics
            - name: DD_ADDITIONAL_ENDPOINTS
              value: '{"<oodle_metrics_endpoint>": ["<oodle_api_key>"]}'
            - name: DD_USE_V3_API_SERIES_ENABLED
              value: 'false'
            # Logs
            - name: DD_LOGS_CONFIG_ADDITIONAL_ENDPOINTS
              value: '[{"api_key": "<oodle_api_key>", "Host": "<oodle_logs_host>", "Port": 443, "is_reliable": false}]'
            - name: DD_LOGS_CONFIG_FORCE_USE_HTTP
              value: 'true'
        trace-agent:
          env:
            # APM / Traces
            - name: DD_APM_ADDITIONAL_ENDPOINTS
              value: '{"<oodle_traces_endpoint>": ["<oodle_api_key>"]}'
    clusterAgent:
      env:
        # Metrics (cluster agent also needs this)
        - name: DD_ADDITIONAL_ENDPOINTS
          value: '{"<oodle_metrics_endpoint>": ["<oodle_api_key>"]}'
        - name: DD_USE_V3_API_SERIES_ENABLED
          value: 'false'

kubectl apply -f datadog-agent.yaml

Datadog Agent YAML

Add the following to your datadog.yaml:

# Metrics
additional_endpoints:
  "<oodle_metrics_endpoint>":
    - <oodle_api_key>
use_v3_api:
  series:
    enabled: false

# APM / Traces
apm_config:
  additional_endpoints:
    "<oodle_traces_endpoint>":
      - <oodle_api_key>

# Logs
logs_config:
  force_use_http: true
  additional_endpoints:
    - api_key: <oodle_api_key>
      Host: <oodle_logs_host>
      Port: 443
      is_reliable: false

Environment variables

If you are not using Helm, Operator, or YAML config, set these environment variables on both the agent and cluster-agent containers:

# Metrics
DD_ADDITIONAL_ENDPOINTS=\
'{"<oodle_metrics_endpoint>": ["<oodle_api_key>"]}'
DD_USE_V3_API_SERIES_ENABLED=false

# Logs
DD_LOGS_CONFIG_ADDITIONAL_ENDPOINTS=\
'[{"api_key": "<oodle_api_key>", "Host": "<oodle_logs_host>", "Port": 443, "is_reliable": false}]'
DD_LOGS_CONFIG_FORCE_USE_HTTP='true'

# APM / Traces
DD_APM_ADDITIONAL_ENDPOINTS=\
'{"<oodle_traces_endpoint>": ["<oodle_api_key>"]}'

Single-shipping (Oodle only)

Stop sending data to Datadog and send only to Oodle.

Datadog Helm chart

datadog:
  apiKey: <oodle_api_key>
  dd_url: <oodle_metrics_endpoint>
  logs:
    enabled: true
  apm:
    enabled: true
  env:
    # Metrics
    - name: DD_USE_V3_API_SERIES_ENABLED
      value: 'false'
    # Logs
    - name: DD_LOGS_CONFIG_DD_URL
      value: '<oodle_logs_base_url>'
    - name: DD_LOGS_CONFIG_FORCE_USE_HTTP
      value: 'true'
    # APM / Traces
    - name: DD_APM_DD_URL
      value: '<oodle_traces_endpoint>'

Then install or upgrade:

helm repo add --force-update datadog \
  https://helm.datadoghq.com && \
helm upgrade --install datadog-agent datadog/datadog \
  --values values.yaml

Datadog Operator

apiVersion: datadoghq.com/v2alpha1
kind: DatadogAgent
metadata:
  name: datadog

spec:
  global:
    endpoint:
      url: <oodle_metrics_endpoint>
    credentials:
      apiKey: <oodle_api_key>

  override:
    nodeAgent:
      containers:
        agent:
          env:
            # Metrics
            - name: DD_USE_V3_API_SERIES_ENABLED
              value: 'false'
            # Logs
            - name: DD_LOGS_CONFIG_DD_URL
              value: '<oodle_logs_base_url>'
            - name: DD_LOGS_CONFIG_FORCE_USE_HTTP
              value: 'true'
        trace-agent:
          env:
            # APM / Traces
            - name: DD_APM_DD_URL
              value: '<oodle_traces_endpoint>'

kubectl apply -f datadog-agent.yaml

Datadog Agent YAML

# Metrics
dd_url: <oodle_metrics_endpoint>
api_key: <oodle_api_key>
use_v3_api:
  series:
    enabled: false

# APM / Traces
apm_config:
  apm_dd_url: <oodle_traces_endpoint>

# Logs
logs_config:
  logs_dd_url: <oodle_logs_base_url>
  force_use_http: true

Environment variables

# Metrics
DD_DD_URL=<oodle_metrics_endpoint>
DD_API_KEY=<oodle_api_key>
DD_USE_V3_API_SERIES_ENABLED=false

# Logs
DD_LOGS_CONFIG_DD_URL='<oodle_logs_base_url>'
DD_LOGS_CONFIG_FORCE_USE_HTTP='true'

# APM / Traces
DD_APM_DD_URL='<oodle_traces_endpoint>'

Amazon ECS

Run the Datadog Agent in Amazon ECS and send its metrics and traces to Oodle with the same environment variables as on other platforms. You set them on the Agent container: the sidecar in each task on Fargate, or the Agent daemon task on EC2.

tip

The Datadog integration tile has an Amazon ECS Fargate tab. It generates the configuration below with the endpoints and the API key of your instance. For logs from ECS, and for routers such as FireLens, see the Amazon ECS guide.

Fargate with APM: dual-shipping

This setup keeps your Datadog configuration and adds Oodle as a second destination. The application code does not change.

Store the values that hold the Oodle API key as SSM SecureString parameters. If you put them in the environment of the task definition, anyone who can read the task definition can read the key.

aws ssm put-parameter --overwrite --type SecureString \
  --name /datadog/oodle-additional-endpoints \
  --value '{"<oodle_metrics_endpoint>":["<oodle_api_key>"]}'

aws ssm put-parameter --overwrite --type SecureString \
  --name /datadog/oodle-apm-additional-endpoints \
  --value '{"<oodle_traces_endpoint>":["<oodle_api_key>"]}'

Give the task execution role ssm:GetParameters on arn:aws:ssm:<region>:<account-id>:parameter/datadog/*. If the parameters use a customer-managed KMS key, also give it kms:Decrypt.

Add one Agent container to each task definition. The /datadog/api-key parameter holds your Datadog API key. Set DD_SITE to the site of your Datadog account.

{
  "name": "datadog-agent",
  "image": "public.ecr.aws/datadog/agent:latest",
  "essential": true,
  "environment": [
    { "name": "ECS_FARGATE", "value": "true" },
    { "name": "DD_SITE", "value": "datadoghq.com" },
    { "name": "DD_USE_V3_API_SERIES_ENABLED", "value": "false" },
    { "name": "DD_APM_ENABLED", "value": "true" }
  ],
  "secrets": [
    { "name": "DD_API_KEY", "valueFrom": "arn:aws:ssm:<region>:<account-id>:parameter/datadog/api-key" },
    { "name": "DD_ADDITIONAL_ENDPOINTS", "valueFrom": "arn:aws:ssm:<region>:<account-id>:parameter/datadog/oodle-additional-endpoints" },
    { "name": "DD_APM_ADDITIONAL_ENDPOINTS", "valueFrom": "arn:aws:ssm:<region>:<account-id>:parameter/datadog/oodle-apm-additional-endpoints" }
  ],
  "portMappings": [
    { "containerPort": 8126, "protocol": "tcp" },
    { "containerPort": 8125, "protocol": "udp" }
  ],
  "healthCheck": {
    "command": ["CMD-SHELL", "agent health"],
    "interval": 10,
    "timeout": 5,
    "retries": 3,
    "startPeriod": 15
  }
}

Add the service identity to each app container. This is Datadog Unified Service Tagging. The environment variables tag the spans, and the Docker labels tag the container metrics.

{
  "environment": [
    { "name": "DD_SERVICE", "value": "checkout" },
    { "name": "DD_ENV", "value": "production" },
    { "name": "DD_VERSION", "value": "1.4.2" }
  ],
  "dockerLabels": {
    "com.datadoghq.tags.service": "checkout",
    "com.datadoghq.tags.env": "production",
    "com.datadoghq.tags.version": "1.4.2"
  },
  "dependsOn": [{ "containerName": "datadog-agent", "condition": "HEALTHY" }]
}

Do not set DD_AGENT_HOST. The containers in an awsvpc task share one network namespace, so the tracer reaches the Agent on its default address, localhost:8126. The dependsOn entry starts the app after the Agent accepts traces.

Deploy the new task definitions. Then make sure that the stats and the spans arrive. Use your own env and service name:
```
oodle metrics query --query 'sum by (service) (rate(dd_trace_stats_hits{env="production"}[5m]))'
oodle traces list --service checkout --start -5m --end now
```
The first command shows one series per service. The second command must return traces. The spans also show in the Traces Explorer, and the stats fill the Service Catalog.

On Fargate the Agent does not collect the logs of the other containers in the task. To send logs, use FireLens, as the Amazon ECS guide shows.

Fargate: single-shipping (Oodle only)

Store the Oodle API key in /datadog/api-key, and give the Agent the Oodle URLs instead of additional endpoints. You do not need DD_SITE or the two additional-endpoint secrets.

"environment": [
  { "name": "ECS_FARGATE", "value": "true" },
  { "name": "DD_DD_URL", "value": "<oodle_metrics_endpoint>" },
  { "name": "DD_USE_V3_API_SERIES_ENABLED", "value": "false" },
  { "name": "DD_APM_ENABLED", "value": "true" },
  { "name": "DD_APM_DD_URL", "value": "<oodle_traces_endpoint>" }
],
"secrets": [
  { "name": "DD_API_KEY", "valueFrom": "arn:aws:ssm:<region>:<account-id>:parameter/datadog/api-key" }
]

The app containers need the same fields as in step 4 above.

EC2 and the Datadog Terraform module

On the EC2 launch type, the Agent runs as a daemon task. The official terraform-aws-ecs-datadog module (ecs_fargate / ecs_ec2) has a dd_environment input for the same variables. A runnable example for both launch types is in the oodle-onboarding datadog/ecs demo.

note

For single-shipping on the EC2 launch type, also set DD_ORCHESTRATOR_EXPLORER_ENABLED=false. Orchestrator Explorer ships to the site-based process intake of Datadog, which DD_DD_URL cannot redirect to Oodle.

Support

If you need assistance or have any questions, please reach out to us through:

Email at [email protected]

Supported signals​

Dual-shipping (Datadog + Oodle)​

Datadog Helm chart​

Datadog Operator​

Datadog Agent YAML​

Environment variables​

Single-shipping (Oodle only)​

Datadog Helm chart​

Datadog Operator​

Datadog Agent YAML​

Environment variables​

Amazon ECS​

Fargate with APM: dual-shipping​

Fargate: single-shipping (Oodle only)​

EC2 and the Datadog Terraform module​

Support

Supported signals

Dual-shipping (Datadog + Oodle)

Datadog Helm chart

Datadog Operator

Datadog Agent YAML

Environment variables

Single-shipping (Oodle only)

Datadog Helm chart

Datadog Operator

Datadog Agent YAML

Environment variables

Amazon ECS

Fargate with APM: dual-shipping

Fargate: single-shipping (Oodle only)

EC2 and the Datadog Terraform module