Prometheus Metrics in Kotlin: Integration with Micrometer and Ktor

Prometheus Metrics in Kotlin: From Basics to Production Solutions

Modern backend applications require constant monitoring. Prometheus is the de facto standard for metrics collection in the microservices world. In this article, we will explore how to integrate Prometheus into Kotlin applications using the Micrometer library and the Ktor framework. You will learn how to create custom counters, histograms, and timers, as well as configure an endpoint for metrics collection.

1. Why Micrometer?

Micrometer is a monitoring facade, similar to SLF4J for logging. It supports many backends: Prometheus, Graphite, Datadog, InfluxDB, and others. By using Micrometer, you are not tied to a specific monitoring system — you simply change the dependency.

For Prometheus, Micrometer provides a special registry PrometheusMeterRegistry, which exports metrics in a text format understandable by Prometheus.

2. Adding Dependencies

Add the following dependencies to your build.gradle.kts:

dependencies {    // Micrometer + Prometheus    implementation("io.micrometer:micrometer-core:1.12.5")    implementation("io.micrometer:micrometer-registry-prometheus:1.12.5")

    // For Ktor    implementation("io.ktor:ktor-server-core:2.3.9")    implementation("io.ktor:ktor-server-netty:2.3.9")    implementation("io.ktor:ktor-server-metrics-micrometer:2.3.9")}

If you are using Spring Boot, simply add spring-boot-starter-actuator and micrometer-registry-prometheus — Spring Boot will automatically configure the /actuator/prometheus endpoint.

3. Basic Setup in Ktor

Let's create a simple Ktor application with Prometheus metrics. We will need:

• PrometheusMeterRegistry — the metrics registry
• KtorMetricsPlugin — a plugin for automatic HTTP metrics collection
• A /metrics endpoint for Prometheus to scrape

import io.ktor.server.application.*import io.ktor.server.engine.*import io.ktor.server.netty.*import io.ktor.server.response.*import io.ktor.server.routing.*import io.micrometer.prometheus.PrometheusConfigimport io.micrometer.prometheus.PrometheusMeterRegistryimport io.ktor.server.plugins.metrics.micrometer.*

fun main() {    val prometheusRegistry = PrometheusMeterRegistry(PrometheusConfig.DEFAULT)

    embeddedServer(Netty, port = 8080) {        install(MicrometerMetrics) {            registry = prometheusRegistry            // Enable default metrics collection (JVM, CPU, memory)            meterBinders = listOf(                JvmGcMetrics(),                JvmMemoryMetrics(),                ProcessorMetrics()            )        }

        routing {            // Endpoint for Prometheus            get("/metrics") {                call.respondText(                    contentType = ContentType.parse("text/plain; version=0.0.4"),                    text = prometheusRegistry.scrape()                )            }

            // Example business endpoint            get("/hello") {                call.respondText("Hello, Kotlin!")            }        }    }.start(wait = true)}

4. Custom Metrics

In addition to automatic metrics, you can create your own. Let's look at three main types:

• Counter — a counter (e.g., number of requests)
• Timer — execution time of an operation
• Gauge — a current value (e.g., number of active users)

import io.micrometer.core.instrument.Counterimport io.micrometer.core.instrument.Timerimport io.micrometer.core.instrument.Gauge

// Create metricsval requestCounter: Counter = Counter    .builder(