skip to content

Caching & Persisted Queries

13 roadmaps32 questions

GraphQL gives up ordinary HTTP caching the moment every operation goes through one POST endpoint, and this is what you get back. Interviewers ask for the whole ladder, edge to client store.

part ofGraphQLoverview, primer and where to startread it →
on this pageshow

explore

questions

page 2 of 2

In a normalized client cache, what happens when two operations write the same entity with different values?

level: seniorimportance: nice to knowfreq 24%

basics

~20 s

The writes merge field by field into one shared entry, and for any field both carry, the write that lands last wins. Nothing compares freshness, so a slow response can overwrite a newer value and visibly change screens that never issued it.

open as a page

In a persisted document registry, how do you decide when an old operation entry is safe to delete?

level: seniorimportance: nice to knowfreq 23%

basics

~20 s

By client usage, never by age. An entry is safe to remove once no client build still in the field can send its identifier — proven from per-entry last-seen data and the releases you still support.

open as a page

showing 31–32 of 32