Caching

Kind: pattern

Layer: Performance Core

Record: architecture:caching

Severity: contextual

Scope: data access, computation, API

Canonical: Ontology

A design pattern that stores the result of an expensive read or computation under a key derived from every input it depends on, and serves it again while that key matches.

Listed in Architecture principles, after Partitioning and before Statelessness.

Requires

• Invalidation Policy

Reinforces

• Latency Reduction
• Scalability

Enables

• Reduced Load

Conflicts with

• Cache Poisoning by Design

In tension with

• Consistency
• Always-Fresh Reads

Tensions

• Caching / Consistency
• Caching / Always-Fresh Reads

Violated by

• Repeated Stable Computation

Refactored by

• Context-Keyed Cache
• Cache Invalidation on Write

Severity

• contextual

Category

• Scalability / Performance / Optimization

Enforced by

• rules/eslint/require-cache-coverage.eslint.rule.ts

Refactors

• Latency
• Time Complexity
• Bottleneck Analysis

Reinforced by

• CDN / Edge Caching

Linked from

• Computation and resource
• Anti-patterns
• Scalability / Performance / Optimization
• Transactions / State / Concurrency
• Scalability / Performance / Optimization
• Severity levels
• The resolutions

in this codebase

• diagramDigest