Server-Side Caching vs Client-Side Caching

A bug put a wrong price into an API response, and caches stored the response. How you recover depends entirely on where the copies are. If the copy is in your Redis, you delete one key, and every user sees the correct price. If the copy is in a million browsers, you cannot touch it. Each browser keeps serving its own copy 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 therefore who can remove it. Server-side caching keeps the copy on machines you run, and one copy serves everybody