Known known

As explained in Effective Platform Engineering

A known-known is a failure or operating condition whose possibility and specific behavior are already understood.

A known-known is a condition that can go wrong and whose specific behavior is already understood. Observability can therefore define the healthy boundary, query for violations, alert reliably, and sometimes restore the system automatically. section 5.1.3

Treat known-knowns as explicit operational contracts:

• Measure the signals that identify the known healthy state. section 5.1.3
• Query and alert when a defined boundary is violated. section 5.1.3
• Develop self-healing strategies when the failure and remedy are predictable. section 5.1.3

Platform operations

A Kubernetes cluster sustaining CPU activity above 90% signals that scaling is not working as expected. Load testing may establish that a source-control project API handles only ten simultaneous requests. Telemetry and alerts can expose both limits before they disrupt users. section 5.1.3

Figure 5.3 Knowing what to look for when querying observability data is an evolving process.

Use the figure to distinguish what observability teams can query for now from conditions they must learn to recognize over time. section 5.1.3

How Effective Platform Engineering covers it

How deep it goes

Explains thresholds, capacity limits, telemetry, alerts, and self-healing, while contrasting conditions not yet fully known.

Examples

Uses a 90% Kubernetes CPU threshold, a ten-request API capacity limit, and a platform-adoption success target.

What it assumes

Assumes familiarity with healthy system states, metrics, APIs, Kubernetes, and operational alerting.

What you can do after

Diagnose known violations with telemetry and alerts, then apply self-healing strategies where remedies are predictable.

How it connects

Connects observability with telemetry, monitoring, alerting, load testing, Kubernetes, APIs, and self-healing.

liveBookmanning.com

Read Effective Platform Engineering on liveBook Get the book at manning.com

Related concepts

• self-healing mechanismuses · A self-healing mechanism automatically detects failures or deviations from a desired state and takes corrective action to restore or maintain system operation.
• known unknowncontrasts with · A known unknown is a potentially relevant fact or condition whose existence is recognized but whose behavior or details are not understood.

This explainer was generated by Manning's knowledge graph from the text of Effective Platform Engineering. The section links point to the parts of the book it draws on.