Skip to main content

tinkr trace

Retrieve distributed traces for a service and display the slowest or most error-prone requests.

tinkr trace <service> [options]

Arguments​

ArgumentDescription
serviceService name as configured in the active profile

Options​

FlagDefaultDescription
--since TEXT1hLook-back window — e.g. 30m, 2h, 24h
--limit INT20Maximum number of traces to return
--tags TEXT—Filter by tag key=value pairs (e.g. http.status_code=500)
--jsonoffEmit raw JSON

Examples​

# Last 20 traces
tinkr trace payments-api

# Last 30 minutes, up to 50 traces
tinkr trace payments-api --since 30m --limit 50

# Only failed traces
tinkr trace payments-api --tags http.status_code=500

# JSON output
tinkr trace payments-api --since 1h --json

Output​

Traces for payments-api (last 1h, 20 results)

  TRACE ID              OPERATION               DURATION   SPANS  STATUS
  7f3a2b1c4d5e6f78      POST /v1/charges        4.2s       12     ERROR
  1a2b3c4d5e6f7a8b      POST /v1/charges        1.1s        8     OK
  9e8d7c6b5a4f3e2d      GET  /v1/customers       98ms        3     OK
  ...

  Slowest trace: 7f3a2b1c4d5e6f78

  ┌ POST /v1/charges (payments-api)  4.2s
  ├── validate_card                   12ms
  ├── db.query customers              45ms
  ├── stripe.charge (external)        3.8s  ← slow
  └── db.insert payment_events        22ms

Tracing backends​

BackendTracing system
GrafanaGrafana Tempo
CloudWatchAWS X-Ray
GCPGoogle Cloud Trace
AzureApplication Insights
DatadogDatadog APM
ElasticElastic APM
OTelOpenSearch (OTel Collector)

Tracing requirements​

Tracing must be enabled in your application. Tinkr reads from the tracing backend — it does not instrument your code.

  • Grafana: Configure your app to send spans to Tempo
  • CloudWatch: Enable X-Ray tracing in your ECS task definition / Lambda configuration
  • GCP: Enable Cloud Trace API; use the OpenTelemetry or Cloud Trace SDK
  • Azure: Enable Application Insights; SDK auto-instruments most frameworks
  • Datadog: Install the Datadog APM library; traces flow through the Datadog agent
  • Elastic: Use the Elastic APM agent; traces go to the APM server
  • OTel: Send spans to the OTel Collector → OpenSearch backend

See also​