Software Diagnostics and Observability Institute

The Seven-Layer Observability Design Model is a conceptual framework for designing observability in modern software systems, modeled on the OSI networking model. It does not claim that observability literally has the same layers as networking. Instead, it uses the OSI-style layered thinking to separate concerns, clarify responsibilities, and make observability systems easier to reason about, build, debug, and evolve.

Layer 1: Signal Source Layer - the lowest layer of the model. It corresponds to where observable events originate.
Layer 2: Instrumentation Layer - turns raw system behavior into explicit observability signals.
Layer 3: Collection Layer - gathers observability signals from different sources.
Layer 4: Transport Layer - moves observability data from collectors to processing and storage systems.
Layer 5: Storage and Representation Layer - stores observability data in forms suitable for retrieval, querying, correlation, and analysis.
Layer 6: Correlation and Context Layer - connects isolated signals into meaningful diagnostic structures.
Layer 7: Interpretation and Action Layer - converts observability data into understanding and response.
Meta-observability: Observability of Observability - A production observability stack should itself be observable across the same seven layers.