How do you stream a large file or dataset as a response without loading it fully into memory in NestJS?

0-2 years experience

1. 1We need to let users download a CSV export of 10,000 rows. How would you implement the endpoint in NestJS so that the file is streamed to the client without loading the whole CSV into memory?
2. 2If you use the built‑in res object to pipe a read stream, what steps must you take to set the correct headers and ensure the response ends properly?
3. 3What would happen if you accidentally returned the entire buffer from the controller instead of streaming it?

2-5 years experience

1. 1Your team added a new feature that streams video chunks from S3 through a NestJS gateway, but after deployment some users see incomplete downloads. Walk me through how you'd debug the streaming pipeline.
2. 2Explain the trade‑offs between using NestJS's StreamableFile class versus manually piping a Node.js fs.createReadStream in a controller.
3. 3Suppose you need to add rate limiting to a large file download endpoint without breaking the stream. How would you approach that?

5-8 years experience

1. 1Design a reusable NestJS interceptor that can stream any large JSON dataset from a database cursor while handling back‑pressure and error propagation.
2. 2At scale, you have multiple instances behind a load balancer serving large file streams. What architectural considerations (e.g., connection limits, memory usage, CDN) would you address?
3. 3How would you modify the streaming implementation to support range requests for resumable downloads, and what impact does that have on NestJS middleware?

8+ years experience

1. 1Our monolith is being broken into microservices, and the file‑streaming service must serve both internal services and external clients. Describe the end‑to‑end architecture, including protocol choices, contract versioning, and how you’d ensure backward compatibility.
2. 2If the organization wants to migrate from on‑prem file storage to a cloud object store, what changes would you make to the NestJS streaming layer to minimize impact across teams?
3. 3Discuss how you would set up observability (metrics, tracing) for streaming endpoints to detect slow consumers and prevent resource exhaustion across the fleet.