Question 1

Vertical vs horizontal scaling for Node apps?

Accepted Answer

Vertical: bigger machine; horizontal: more processes/instances behind a load balancer — Node usually scales out. Single process limited by one event loop. Combine with clustering/K8s replicas. Statelessness enables horizontal scale.

Question 2

What must be externalized to scale Node horizontally?

Accepted Answer

Sessions, uploads, job queues, rate-limit counters, and any in-memory singleton state. Sticky sessions are a crutch. Use Redis/DB/object storage. Design for multi-instance from day one.

Question 3

How do queues help scale Node workloads?

Accepted Answer

Absorb spikes, retry failures, and move slow work off HTTP request threads. Choose at-least-once semantics consciously. Idempotent consumers. Observe lag as a first-class metric.

Question 4

Caching strategies in Node services?

Accepted Answer

HTTP cache headers, CDN, Redis for hot keys, and in-process LRU cautiously (per instance). Cache invalidation and stampedes matter. Don't cache personalized data publicly. Set TTLs and jitter.

Question 5

Scenario: one noisy neighbor tenant saturates a shared Node API.

Accepted Answer

Per-tenant rate limits, isolation pools, fair scheduling, and possibly dedicated cells for large tenants. Quotas at gateway. Backpressure. Multi-tenant isolation is a senior/lead topic.

Question 6

Lead: capacity plan a Node fleet for Black Friday.

Accepted Answer

Load test with production-like data, autoscale policies, DB headroom, cache warmups, and degradations (feature flags). Practice failovers. Pre-scale. Define brownout modes (disable noncritical). Communicate SLOs.

Scaling Node.js Interview Questions

Other Node.js interview topics