> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://mezmo.ferndocs.com/datadog-agent-v2-source/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://mezmo.ferndocs.com/_mcp/server.

# Datadog Agent V2

> This feature is experimental. Behavior and configuration options may change in future releases.

## Description

The Datadog Agent V2 source receives logs, metrics, and traces directly from a Datadog Agent. Unlike the original Datadog Agent source, this version provides separate output ports for each telemetry type, so you can route logs, metrics, and traces to different destinations in your pipeline.

## Outputs

* **Logs**: Log events from the Datadog Agent
* **Metrics**: Metric events from the Datadog Agent, including sketches (distribution metrics)
* **Traces**: Trace events from the Datadog Agent

Connect each output to your desired processor or destination. You can leave outputs unconnected if you don't need that telemetry type.

## Configuration

Add a source and choose **Datadog Agent V2**, optionally enter a **Title** and **Description**, then click **Save**.

Under **Access Key Management**:

1. Click **Create new key**.
2. Enter a title.
3. Set the type to **User Defined** and enter your Datadog API key.
4. Click **Create**.

### Datadog Agent Configuration

Configure the Datadog Agent to send data to your Mezmo pipeline URLs. The Datadog Agent V2 source uses Observability Pipelines Worker configuration options.

#### Requirements

The running agent version must be \[ >= 6.45.0 or >= 7.45.0 ]

#### Configuration Options

| Option                                              | Description                                                                                           |
| --------------------------------------------------- | ----------------------------------------------------------------------------------------------------- |
| `DD_API_KEY`                                        | The API key the agent is running with.                                                                |
| `DD_LOGS_ENABLED`                                   | Set to `true` to allow the agent to collect logs.                                                     |
| `DD_OBSERVABILITY_PIPELINES_WORKER_LOGS_ENABLED`    | Set to `true` to enable the agent to send logs to Mezmo.                                              |
| `DD_OBSERVABILITY_PIPELINES_WORKER_LOGS_URL`        | The Mezmo pipeline URL to send logs to. Use the format `<route-id>.v1.pipeline.mezmo.com:443`.        |
| `DD_OBSERVABILITY_PIPELINES_WORKER_METRICS_ENABLED` | Set to `true` to enable the agent to send metrics to Mezmo.                                           |
| `DD_OBSERVABILITY_PIPELINES_WORKER_METRICS_URL`     | The Mezmo pipeline URL to send metrics to. Use the format `https://pipeline.mezmo.com/v1/<route-id>`. |
| `DD_OBSERVABILITY_PIPELINES_WORKER_TRACES_ENABLED`  | Set to `true` to enable the agent to send traces to Mezmo.                                            |
| `DD_OBSERVABILITY_PIPELINES_WORKER_TRACES_URL`      | The Mezmo pipeline URL to send traces to. Use the format `https://pipeline.mezmo.com/v1/<route-id>`.  |

#### Running the Datadog Agent via Docker

Example pipeline publishing logs, metrics, and traces to Mezmo:

```bash
docker run -d --name datadog-agent \
--cgroupns host \
--pid host \
-e DD_API_KEY=<DATADOG_API_KEY> \
-e DD_LOGS_ENABLED=true \
-e DD_LOGS_CONFIG_CONTAINER_COLLECT_ALL=true \
-e DD_LOGS_CONFIG_DOCKER_CONTAINER_USE_FILE=true \
-e DD_CONTAINER_EXCLUDE="name:datadog-agent" \
-e DD_OBSERVABILITY_PIPELINES_WORKER_LOGS_ENABLED=true \
-e DD_OBSERVABILITY_PIPELINES_WORKER_LOGS_URL="<route-id>.v1.pipeline.mezmo.com:443" \
-e DD_OBSERVABILITY_PIPELINES_WORKER_METRICS_ENABLED=true \
-e DD_OBSERVABILITY_PIPELINES_WORKER_METRICS_URL="https://pipeline.mezmo.com/v1/<route-id>" \
-e DD_OBSERVABILITY_PIPELINES_WORKER_TRACES_ENABLED=true \
-e DD_OBSERVABILITY_PIPELINES_WORKER_TRACES_URL="https://pipeline.mezmo.com/v1/<route-id>" \
-v /var/run/docker.sock:/var/run/docker.sock:ro \
-v /var/lib/docker/containers:/var/lib/docker/containers:ro \
-v /opt/datadog-agent/run:/opt/datadog-agent/run:rw \
gcr.io/datadoghq/agent:latest
```

Note that `DD_OBSERVABILITY_PIPELINES_WORKER_LOGS_URL` does not include `https` as the protocol is already assumed by the agent.

#### Running the local Datadog Agent

Configure the agent using the `datadog.yaml` configuration file. See the [Datadog Agent configuration files documentation](https://docs.datadoghq.com/agent/guide/agent-configuration-files/?tab=agentv6v7) for file locations.

Example pipeline publishing logs, metrics, and traces to Mezmo:

```yaml
api_key: <DATADOG_API_KEY>
logs_enabled: true
logs_config:
    container_collect_all: true
observability_pipelines_worker:
    logs:
        enabled: true
        url: <route-id>.v1.pipeline.mezmo.com
    metrics:
        enabled: true
        url: pipeline.mezmo.com/v1/<route-id>
    traces:
        enabled: true
        url: pipeline.mezmo.com/v1/<route-id>
```

## Example Pipeline

A common setup routes logs to Mezmo for analysis while forwarding metrics and traces back to Datadog:

1. Add a **Datadog Agent V2** source.
2. Connect the **Logs** output to a **Mezmo Log Analysis** destination.
3. Connect the **Metrics** output to a **Datadog Metrics** destination.
4. Connect the **Traces** output to a **Datadog Traces** destination.