Building Custom Prometheus Metrics in PHP: Counters, Gauges, and Histograms

A developer tutorial demonstrates how to implement three core Prometheus metric types — counters, gauges, and histograms — using a minimal PHP application without any external framework. Counters track cumulative events such as total orders created, while gauges represent values that can rise or fall, like the current queue size. Histograms capture the distribution of continuous measurements, most commonly request durations, by sorting observations into predefined buckets. The PHP app runs on the built-in server and exposes all three metric types via a /metrics endpoint in the standard Prometheus text format. Together, these metric types cover the vast majority of real-world observability needs, enabling tools like Prometheus to calculate approximate percentiles at query time using histogram_quantile.