Layer 3 · Go deeper · 6 min

Go deeper

Size it, build it, then make the calls an interviewer will push you on.

Build-up · 4 steps

From a staircase of queries to two

The same 50-restaurant page, traced twice. Drawn to show order, not to scale.

Step 1 of 4

SELECT … LIMIT 50 returns 50 restaurants with their dishes not yet loaded. So far, one round trip.

4 interview moments

The interviewer pushes on the hard parts. Pick an approach and see how it holds up.

Deep dive 1 of 4
Interviewer
“A Django page lists 50 restaurants, each with its zone and its dishes. How do you load them?”
Your move. Pick an approach:
Deep dive 2 of 4
Interviewer
“A GraphQL query asks for 50 restaurants and each one’s brand. The brand resolver runs one query per call. What do you do?”
Your move. Pick an approach:
Deep dive 3 of 4
Interviewer
“The feed service calls the user service once per author, 50 calls a page. How do you fix it?”
Your move. Pick an approach:
Deep dive 4 of 4
Interviewer
“An order page lazy-loads its 3 delivery addresses, so 4 queries. A reviewer says fix the N+1. Do you?”
Your move. Pick an approach:

Build-up · 4 steps

A GraphQL query asks for 50 restaurants and each one’s brand. The brand resolver hardly changes; the batching happens underneath it.

Step 1 of 4

The brand resolver runs once per restaurant. Each call now returns a promise from brandLoader.load(id) instead of sending a query.