When we discussed the REST principles defined by Roy Fielding, we mentioned that caching was a rather sensitive topic. In the end, our consumers will expect up-to-date results when executing a query. However, from a statistical point of view, data exposed in the web is more likely to be read rather than updated or deleted.
So, it is reasonable that some resources exposed by a public URL become a subject of millions of requests, considering taking off part of the load from the server to a cache. The HTTP protocol allows us to cache some responses for a given period of time. For instance, when multiple requests are received in a short period of time, querying for all items in the catalog of a given group, such as /catalog/v2, our service can utilize special HTTP headers that would force the HTTP server to cache the response for a defined time period. This would prevent redundant requests to the underlying database server.
Caching at the HTTP server level is achieved via special response...