ทำความเข้าใจพื้นฐานประสิทธิภาพฐานข้อมูล

Your database is the heart of your app — when it's slow, everything is. Tuning database performance keeps the whole experience fast and reliable.

Database performance isn't one number. It's responsiveness (how fast queries return), throughput (queries per second), and how well it uses server resources.

Response time is the full round-trip: from sending a query to receiving all its data. Lower means a snappier app — a slow page often hides a slow query.

Throughput measures operations per second — the database's capacity under load. High throughput is what keeps an e-commerce site fast with many users at once.

Databases lean hard on server resources: CPU for query logic, RAM for caching, disk I/O for storage, and network for client traffic. Use them well.

A CPU bottleneck pins the processor near 100% from inefficient queries. A memory bottleneck forces slow disk swapping when RAM runs out. Both crush speed.

Disk I/O bottlenecks are a top performance killer. Disk access is orders of magnitude slower than memory, so hitting it too often drags queries down.

A network bottleneck adds latency between app and database, while concurrency issues cause locks — one operation stalling while it waits for another.

Get a quick feel for query speed by timing execution — for example, psql's \timing. Run the simple queries below to try it.

Which of the following are considered key metrics for evaluating database performance?

Recap: database performance rests on response time, throughput, and resource use — watch for CPU, memory, disk I/O, network, and concurrency bottlenecks.