Observability

Observability is a cross-cutting concern: logs, metrics, and distributed traces let you run data products in production with the same expectations as other tier-one services. Signals are available at product scope and, for administration, at organization scope (for example, aggregating logs or usage across everything owned by a branch of the org tree).

• What you can expect — Correlation with request and stream context, OpenTelemetry-style integration where the platform supports it, and the ability to triage ingest/outlet path latency and error rates per product and version.
• Org-wide views — Hierarchical organizations are the primary anchor for quota and ownership as well as roll-up of telemetry when you need to reason about an entire line of business (see Organizations).

For incident and support flows, link identity from Authentication and authorization to trace and audit trails where your deployment enables them.

See also — Control plane · Applications · Data product lifecycle