Server-Side Caching vs Client-Side Caching

A bug puts a wrong price into an API response, and the response is cached. How the team fixes it depends on where the cached copies are.

If the copy is in the team's own Redis cache, they delete one key, and every user sees the correct price immediately. If the copies are in a million users' browsers, the team cannot delete them. Each browser keeps showing its own copy until it checks with the server again, or until that copy expires.

Both kinds of caching keep a copy of a response, so the work behind it is not repeated. The difference is **who holds the copy, and so who can remove it