Realtime, Streams & GraphQL — Chapter Quiz

Question 1

1

To stream a response, what kind of value should the handler return?

AA stringBA BufferCA Readable stream (or call reply.send(stream))DA Promise of an array

Return any Node Readable stream — or pass it to reply.send(...). Fastify pipes it to the socket and respects backpressure.

Question 2

2

What content type identifies Server-Sent Events?

Atext/event-streamBtext/plainCapplication/sseDapplication/json-stream

SSE uses Content-Type: text/event-stream. Each event is a line-delimited block of data: ... lines terminated by a blank line.

Question 3

3

@fastify/websocket handles WebSocket upgrades on routes registered with { websocket: true }.

TTrueFFalse

Register the plugin, then declare a route with the websocket option. Fastify negotiates the upgrade and hands you a connection socket pair.

Question 4

4

What is Mercurius?

AA GraphQL server adapter for FastifyBA schema migration toolCA REST-to-GraphQL gateway onlyDA test runner

Mercurius is the Fastify-native GraphQL server with subscriptions, federation, and JIT — written by the Fastify team.

Question 5

5

The Node-native HTTP client recommended for outbound calls from Fastify (built by the same team) is called ____.

undici is the high-performance HTTP/1.1 client used internally by node:fetch. Use its Pool or Agent for connection reuse.

Question 6

6

Which library is most commonly used for background job queues in Node/Fastify apps?

ABullMQ (on Redis)BCeleryCSidekiqDMongo Queue

BullMQ is the de facto Node job queue, built on Redis streams. Celery and Sidekiq are Python/Ruby ecosystems.

Question 7

7

What does OpenTelemetry add to a Fastify app?

ADistributed tracing, metrics, and structured spans across servicesBA schema languageCA new loggerDCSRF protection

OpenTelemetry standardizes traces, metrics, and (eventually) logs across services and exporters (Jaeger, Tempo, Datadog, etc.).

Question 8

8

Which patterns help a Fastify app scale beyond a single process?

Select all that apply.

ANode cluster (or PM2 cluster mode)BA reverse proxy load balancing to multiple containersCSplitting into smaller services that communicate over HTTP / queueDReducing keep-alive to 0 so connections close quickly

Cluster, horizontal scaling behind a load balancer, and decomposing into services are all valid. Killing keep-alive makes things slower, not more scalable.

Test what you learned.