MENTAL MODEL · M4579

Cache Invalidation

SystemsHigh supportSystems Theory

Included

Updated 2026-08-02

Loading revision record…

INTRODUCTION

English translation pending.

CORE DEFINITION

Cache invalidation is the discipline of deciding when a stored copy of data must be discarded or refreshed because the original changed. It is often called one of the two hard problems in computer science, alongside naming things, because any strategy trades freshness against speed and complexity. Write-through, write-behind, and time-based expiry are common policies, and each leaves a window in which readers can see stale values.

SCAFFOLDING EFFECT

psychology

Reduce cognitive load

- Name the trade: state whether you are buying speed at the cost of temporary inconsistency - Bound the staleness: decide how long a stale copy may survive and enforce that limit - Invalidate on change: update or drop the copy the moment the source moves rather than hoping

anchor

Anchor fast decisions

A cache serves stale data whenever the source changes without the copy being told. The faster the cache, the longer the interval during which readers may act on outdated values. Because every policy must choose when to refresh, consistency and latency cannot both be maximized, and the residual risk has to be managed rather than eliminated.

MINIMUM ACTION

In progress 0/1

Practice this model in one real situation:

Choose a write-through or write-invalidate policy and document it explicitly. 2. Handle concurrent invalidation so two writers cannot leave a stale copy behind. 3. Set a maximum staleness bound and monitor how often it is exceeded. 4. Test the system by changing source data and checking what readers actually see.

Check to track your progress (stored locally)

Learning progress0%

account_treeGenealogyexpand_more

menu_bookReferencesexpand_more

Source support: Explicit

link
en.wikipedia.orghttps://en.wikipedia.org/wiki/Cache_invalidationZH · Explicit
verified

RELATED MODELS