Telemetry signals
Logs, metrics, traces, profiles, and events are complementary evidence types rather than a fixed maturity ladder.
Read the page as a transparent path: establish context, inspect the evidence, then test the published stance.
Choose evidence for the question
OpenTelemetry documents traces, metrics, logs, and baggage as distinct signals and context mechanisms [1]. Profiles add sampled evidence about where a program spends resources. These forms complement each other, but they differ in cardinality, cost, timing, and investigative value.
| Signal | Strong fit | Common limitation |
|---|---|---|
| Metrics | Trends, thresholds, efficient aggregation | Detail is lost during aggregation |
| Logs | Discrete events and rich local detail | Inconsistent structure and high volume |
| Traces | Request paths and distributed latency | Sampling and propagation gaps |
| Profiles | Code-level CPU and memory behavior | Correlation and production overhead |
The right mix follows the operational questions, risk, and budget. Treat “logs, metrics, and traces” as a useful vocabulary, not a completeness certificate.
Source record
Bibliography
OpenTelemetry Authors, ‘Signals.’ [Online]. Available: https://opentelemetry.io/docs/concepts/signals/. Accessed: Aug. 9, 2026.
Known limits
Unresolved unknowns
- Terminology for events and profiles varies across products.
Public review
Discuss and correct
Comments are public. Proposed factual corrections must be incorporated into version-controlled research before they are treated as published.
Local preview does not have Giscus configured. Native comments and Giscus are never presented as competing primary systems on the same page.