Use cases

Queryable observability

Keep logs, traces, metrics, and operational events in their raw shape so you can ask new questions later.

Queryable observability

Use RawTree for observability when the important question is not known ahead of time. Send traces, logs, metrics, deploy events, and service events into RawTree, then query the raw telemetry with SQL or ask an agent to inspect it.

RawTree is not an all-in-one incident platform. It is the queryable event store underneath: a place to keep high-cardinality telemetry, correlate signals, and preserve details that would be flattened away by fixed dashboards.

Where it fits

Queryable observability is useful when you need to investigate across raw signals:

  • Which users were affected by a latency spike?
  • Did errors start after a deploy, config change, or traffic shift?
  • Which services, regions, tenants, or sandboxes share the same failure pattern?
  • What did the original log or span attributes say before a dashboard aggregated them?

Use RawTree alongside your existing telemetry stack when you want an ad hoc SQL layer over the events themselves.

OpenTelemetry

Use OpenTelemetry (OTel) as a standard way to move traces, logs, and metrics into RawTree from SDKs, collectors, and instrumented services. Point an OpenTelemetry SDK or Collector at RawTree and it writes each signal to queryable tables by default.

Start with native OTLP in the Transforms guide, then use the specific OpenTelemetry guide when you need a concrete integration.

What to store

Start with the signals that help you reconstruct a timeline:

  • Traces for request paths, spans, durations, and attributes.
  • Logs for application messages, exceptions, and structured context.
  • Metrics for sampled measurements and rollups.
  • Deploy, release, feature-flag, and infrastructure events for correlation.

Because RawTree keeps events queryable in SQL, you can join operational context with application telemetry instead of only reading each stream in isolation.

Start with

  1. Send one service or Collector pipeline to RawTree with OTLP.
  2. Query recent rows from traces, logs, or metrics to confirm ingestion.
  3. Add service, environment, tenant, user, or sandbox attributes that make investigations easier.
  4. Save the SQL queries that answer recurring debugging questions.

Examples