liveBook · Manning

Unknown unknown

An unknown unknown is a relevant fact, condition, risk, knowledge gap, behavior, or failure possibility whose existence or possibility is not recognized.

How 2 Manning books explain it

Effective Platform Engineering

An unknown unknown is a relevant condition, behavior, or failure mode whose possibility has not previously been recognized.

How it is used
Guides observability practices toward discovering and resolving failures that monitoring has not anticipated.
How deep it goes
Explains discovery through telemetry, tracing, time-series analysis, and predictive models, but stops short of prescribing a specific technology.
Examples
Uses pod rescheduling, malformed API data, outages, overlapping dashboards, and outlier analysis as concrete failure and investigation examples.
What you can do after
Diagnose/debug with it by instrumenting systems, tracing transactions, correlating data, and analyzing anomalies to expose unfamiliar failure modes.

Read how Effective Platform Engineering explains unknown unknown →

Experimentation for Engineers

Factors affecting system performance that have not been considered because they are absent from the current model or knowledge.

How it is used
Frames experimentation as necessary when system complexity exceeds what domain knowledge and explicit models can capture.
How deep it goes
Defines unknown unknowns through signal, noise, and omitted factors, but does not develop a mitigation procedure.
Examples
Uses a Bing search change that generated more than $100 million in unexpected annual revenue.
What you can do after
Evaluate system changes with awareness that unmodeled factors may produce unexpectedly large positive or negative effects.

Read how Experimentation for Engineers explains unknown unknown →

Related concepts

These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.