Recache

Recache is the process of refreshing or rebuilding data stored in a cache to reflect the most recent state of the underlying data source. The aim is to prevent stale content and to maintain quick access by serving up-to-date information from fast storage.

Recaching is typically triggered by one or more mechanisms: time-based expiration (TTL) that marks entries as valid for a fixed period; explicit invalidation when the data changes; and event-driven or proactive cache warming, which refreshes content ahead of demand or immediately after a change.

Recache occurs across layers such as application caches (in-memory or distributed like Redis or Memcached), database query caches, and content delivery networks (CDNs) that refresh assets from the origin. Common strategies include lazy recaching (refresh on cache miss after expiration) and eager or proactive recaching (preemptively refreshing items after an update or according to a schedule).

Managing recaching involves trade-offs between data freshness, cache hit rate, and backend load. Aggressive recaching reduces latency and staleness but can cause bursts of cache traffic and higher origin or database load. Cache stampede prevention techniques, versioning, and careful TTL design help mitigate these risks. Different systems adopt varying consistency models, often favoring eventual consistency with TTL or invalidation semantics.

See also: cache invalidation, TTL, cache warming, cache coherence.