Enterprise / Security & Governance

Log Exports

Browse docs

--- title: "Enterprise Log Exports" description: "Export request logs and telemetry to OTLP, Datadog, webhook, or S3 destinations." icon: "download" ---

Overview

Enterprise supports automated export of request logs and telemetry to external observability systems. The exporter runs as a background loop, batching records and shipping them on a configurable interval with support for sampling, header redaction, and custom labels.

Export Destinations

Destination	Auth Modes	Use Case
`otlp`	none, bearer	OpenTelemetry-compatible collectors
`datadog`	api_key	Datadog observability platform
`webhook`	none, bearer, api_key	Custom HTTP endpoints
`s3`	aws_iam	S3-style object storage

Configuration

YAML

yaml

observability_exports:
  enabled: true
  destination: "webhook"  # otlp, datadog, webhook, or s3
  endpoint: "https://observability.example.com/aurora"
  format: "json"
  auth_mode: "none"
  sample_rate: 1.0
  redact_headers: true
  labels:
    environment: "production"
    cluster: "aurora-enterprise"

Environment Variables

env

OBSERVABILITY_EXPORTS_ENABLED=true
OBSERVABILITY_EXPORTS_DESTINATION=datadog
OBSERVABILITY_EXPORTS_ENDPOINT=https://api.datadoghq.com
OBSERVABILITY_EXPORTS_FORMAT=datadog_json
OBSERVABILITY_EXPORTS_AUTH_MODE=api_key
OBSERVABILITY_EXPORTS_SAMPLE_RATE=1.0
OBSERVABILITY_EXPORTS_REDACT_HEADERS=true

Configuration Fields

Field	Type	Default	Description
`destination`	string	â€”	`otlp`, `datadog`, `webhook`, or `s3`
`endpoint`	string	â€”	Target URL for the export destination
`format`	string	`json`	Payload format: `json`, `otlp_json`, `datadog_json`, `s3_json`
`auth_mode`	string	`none`	`none`, `bearer`, `api_key`, or `aws_iam`
`sample_rate`	float	`1.0`	Fraction of requests to export (0.0 - 1.0)
`redact_headers`	bool	`true`	Strip sensitive headers before export
`labels`	map	â€”	Custom key-value labels attached to each record

Datadog Example

yaml

observability_exports:
  enabled: true
  destination: "datadog"
  endpoint: "https://api.datadoghq.com"
  format: "datadog_json"
  auth_mode: "api_key"
  sample_rate: 0.5
  redact_headers: true
  labels:
    env: "production"

Requires DD_API_KEY or DATADOG_API_KEY in environment or config for the auth header.

OTLP Example

yaml

observability_exports:
  enabled: true
  destination: "otlp"
  endpoint: "https://otlp-collector.example.com:4318"
  format: "otlp_json"
  auth_mode: "bearer"
  sample_rate: 1.0

S3 Example

yaml

observability_exports:
  enabled: true
  destination: "s3"
  endpoint: "https://s3.us-east-1.amazonaws.com/aurora-logs"
  format: "s3_json"
  auth_mode: "aws_iam"
  sample_rate: 1.0

Export Behavior

The exporter buffers records in memory and flushes on a periodic interval
Sample rate is applied per-record using a hash of the request ID for consistency
Header redaction strips Authorization, X-Api-Key, Cookie, and similar sensitive headers
Failed exports are retried with backoff; persistent failures are logged
Labels are attached to every exported record for downstream filtering

Data Format

Exported records include:

json

{
  "timestamp": "2026-06-13T12:00:00Z",
  "request_id": "req_abc123",
  "model": "gpt-4o",
  "provider": "openai-primary",
  "user_path": "/team/alpha",
  "input_tokens": 150,
  "output_tokens": 75,
  "total_tokens": 225,
  "latency_ms": 1200,
  "status_code": 200,
  "error": null,
  "labels": {
    "environment": "production"
  }
}

Admin API

For endpoint reference see the Admin API section.

The export status is visible in the dashboard under Settings -> Observability Exports.

← All docs

Enterprise / Security & Governance

Log Exports

Browse docs

--- title: "Enterprise Log Exports" description: "Export request logs and telemetry to OTLP, Datadog, webhook, or S3 destinations." icon: "download" ---

Overview

Export Destinations

Destination	Auth Modes	Use Case
`otlp`	none, bearer	OpenTelemetry-compatible collectors
`datadog`	api_key	Datadog observability platform
`webhook`	none, bearer, api_key	Custom HTTP endpoints
`s3`	aws_iam	S3-style object storage

Configuration

YAML

yaml

observability_exports:
  enabled: true
  destination: "webhook"  # otlp, datadog, webhook, or s3
  endpoint: "https://observability.example.com/aurora"
  format: "json"
  auth_mode: "none"
  sample_rate: 1.0
  redact_headers: true
  labels:
    environment: "production"
    cluster: "aurora-enterprise"

Environment Variables

env

OBSERVABILITY_EXPORTS_ENABLED=true
OBSERVABILITY_EXPORTS_DESTINATION=datadog
OBSERVABILITY_EXPORTS_ENDPOINT=https://api.datadoghq.com
OBSERVABILITY_EXPORTS_FORMAT=datadog_json
OBSERVABILITY_EXPORTS_AUTH_MODE=api_key
OBSERVABILITY_EXPORTS_SAMPLE_RATE=1.0
OBSERVABILITY_EXPORTS_REDACT_HEADERS=true

Configuration Fields

Field	Type	Default	Description
`destination`	string	â€”	`otlp`, `datadog`, `webhook`, or `s3`
`endpoint`	string	â€”	Target URL for the export destination
`format`	string	`json`	Payload format: `json`, `otlp_json`, `datadog_json`, `s3_json`
`auth_mode`	string	`none`	`none`, `bearer`, `api_key`, or `aws_iam`
`sample_rate`	float	`1.0`	Fraction of requests to export (0.0 - 1.0)
`redact_headers`	bool	`true`	Strip sensitive headers before export
`labels`	map	â€”	Custom key-value labels attached to each record

Datadog Example

yaml

observability_exports:
  enabled: true
  destination: "datadog"
  endpoint: "https://api.datadoghq.com"
  format: "datadog_json"
  auth_mode: "api_key"
  sample_rate: 0.5
  redact_headers: true
  labels:
    env: "production"

Requires DD_API_KEY or DATADOG_API_KEY in environment or config for the auth header.

OTLP Example

yaml

observability_exports:
  enabled: true
  destination: "otlp"
  endpoint: "https://otlp-collector.example.com:4318"
  format: "otlp_json"
  auth_mode: "bearer"
  sample_rate: 1.0

S3 Example

yaml

observability_exports:
  enabled: true
  destination: "s3"
  endpoint: "https://s3.us-east-1.amazonaws.com/aurora-logs"
  format: "s3_json"
  auth_mode: "aws_iam"
  sample_rate: 1.0

Export Behavior

The exporter buffers records in memory and flushes on a periodic interval
Sample rate is applied per-record using a hash of the request ID for consistency
Header redaction strips Authorization, X-Api-Key, Cookie, and similar sensitive headers
Failed exports are retried with backoff; persistent failures are logged
Labels are attached to every exported record for downstream filtering

Data Format

Exported records include:

json

{
  "timestamp": "2026-06-13T12:00:00Z",
  "request_id": "req_abc123",
  "model": "gpt-4o",
  "provider": "openai-primary",
  "user_path": "/team/alpha",
  "input_tokens": 150,
  "output_tokens": 75,
  "total_tokens": 225,
  "latency_ms": 1200,
  "status_code": 200,
  "error": null,
  "labels": {
    "environment": "production"
  }
}

Admin API

For endpoint reference see the Admin API section.

The export status is visible in the dashboard under Settings -> Observability Exports.