Datadog Backend

Uses the Datadog Logs API v2 for logs, Metrics API v1 for time series, and APM Traces API v2 for distributed tracing.

TINKR_BACKEND=datadog

---

Authentication

Datadog requires two keys:

• API key (DD_API_KEY) — identifies your organization; used for all API calls
• Application key (DD_APP_KEY) — grants read access to your Datadog account data

Both must be present for Tinkr to function.

Store them in your cloud secrets manager (AWS Secrets Manager, GCP Secret Manager, or Azure Key Vault) and inject as environment variables. Never hardcode them.

---

Environment variables

Variable	Required	Description
DD_API_KEY	Yes	Datadog API key
DD_APP_KEY	Yes	Datadog application key
DD_SITE	No	Datadog site (default: datadoghq.com; EU: datadoghq.eu)

---

Profile configuration

~/.tinkr/config.toml

[profiles.datadog-prod]
backend = "datadog"
site    = "datadoghq.com"

~/.tinkr/.env

DD_API_KEY=xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
DD_APP_KEY=xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

---

Log query

Tinkr calls the Datadog Logs API v2 (POST /api/v2/logs/events/search):

{
  "filter": {
    "query": "service:payments-api status:error",
    "from": "now-1h",
    "to": "now"
  },
  "sort": "-timestamp",
  "page": { "limit": 100 }
}

The service name is matched using the service tag. Ensure your application sets the DD_SERVICE environment variable or service tag.

---

Metrics

Tinkr calls the Datadog Metrics API v1 (GET /api/v1/query):

q=avg:trace.web.request{service:payments-api}.as_count()&from=...&to=...

Common Datadog metrics:

Metric	Description
trace.web.request	HTTP request count
trace.web.request.duration	Request duration
trace.web.request.errors	Request error count
aws.ecs.cpuutilization	ECS container CPU
aws.ecs.memory_utilization	ECS container memory
system.cpu.user	Host CPU
system.mem.used	Host memory

---

Distributed tracing (APM)

Tinkr calls the Datadog APM Traces API v2 (POST /api/v2/spans/events/search):

{
  "filter": {
    "query": "@service:payments-api",
    "from": "now-1h",
    "to": "now"
  }
}

APM must be enabled in your application. Use the Datadog APM library (ddtrace) or the OpenTelemetry SDK with the Datadog exporter.

# Python
pip install ddtrace
ddtrace-run python app.py

# Node.js
DD_TRACE_ENABLED=true node app.js

Traces are automatically correlated with logs when DD_LOGS_INJECTION=true is set.

---

Datadog site

Region	DD_SITE
US1 (default)	datadoghq.com
EU	datadoghq.eu
US3	us3.datadoghq.com
US5	us5.datadoghq.com
AP1	ap1.datadoghq.com
Gov	ddog-gov.com

---

Local development

export TINKR_BACKEND=datadog
export DD_API_KEY=xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
export DD_APP_KEY=xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
uv run tinkr-server