Node performance

Node performance interviews judge diagnosis skill: event-loop health, CPU vs I/O classification, backpressure, profiling, and memory leak prevention—not premature micro-optimizations.

Subtopics: CPU-bound vs I/O-bound work, backpressure, profiling, and memory leaks.

On interviews: given slow endpoints or rising memory, name the measurement you would take first and the fix that matches the bottleneck class.

Common pitfalls: asyncifying CPU loops, unbounded caches and listeners, ignoring event-loop delay metrics, and scaling horizontally before fixing blocking work.

The trade-off is balancing simplicity, performance, safety, and operability — name which axis you optimized and what cost you accepted.

Checklist:

• Classify bottlenecks as CPU, I/O, or memory retention.
• Use streams and queues with explicit flow control.
• Profile with event-loop delay, CPU, and heap tools.
• Hunt retained references from caches, timers, and closures.