[API Testing]
Telemetry for Testers: Logs and Traces
We stay pragmatic: reading JSON logs, understanding trace spans, and knowing when to stop digging and pull an engineer. Labs use OpenTelemetry-flavoured examples with anonymised payloads so you can practise correlation without exposing real customer data.
Included focus areas
- Structured logging fields testers should insist on
- Trace span naming that survives refactors
- Correlation IDs across async workers
- Dashboards for reproduction breadcrumbs
- Privacy redaction checklist
- When to export traces vs screenshots
- Incident comms templates for partial information
Outcomes you can demonstrate
- Reproduce a sample defect using only logs and traces provided
- Draft a span naming proposal for one service boundary
- Identify three missing fields you would lobby engineering to add
Facilitator
Riley Ng
API testing coach focused on payment gateways and logistics integrations.
FAQ
Vendor specific?
Examples skew OpenTelemetry-neutral; one lab uses a hosted trace UI you can trial for fourteen days.
Coding?
Light JSON editing only; no service instrumentation authoring is required.
Limitations?
We do not operate a production observability tenant for you; bring your own org if you want to map lessons directly.
Participant notes
Correlation lab made our client’s ghost bug reproducible in a day. Still noisy in prod, but I know what to ask for now.
Sat in on the final lab; language was accessible without dumbing down the tracing concepts.