Shahin

Cache

Advanced
Updated Jul 11, 2026 Shahin editorial 2 min read
PronunciationEN/kæʃ/

A cache is a stored ready-made copy of data or a response, so that next time it can be served fast from that copy instead of being rebuilt.

Caching sits at several layers of a page's journey: the browser cache that keeps static files on the user's device, the network and CDN cache between user and server, and the server-side cache that holds the result of an expensive query. Its effect on performance is large — a cached response can cut time-to-first-byte from seconds to milliseconds — but that same cache also complicates data analysis. The key point for any measurement tool is this: on a visit that used the cache, the DNS and connection stages return zero, so pooling these visits with fresh-connection visits into one average makes the number falsely optimistic. This is why network metrics must be read with a declared population: “fresh connections only.” Cache-control mechanics (things like ETag and Cache-Control headers) decide when the browser uses the stored copy and when it fetches a fresh one.