Skip to content
actor-ts

OTel metrics adapter

OpenTelemetry is the vendor-neutral standard for observability data. The OtelMetricsAdapter mirrors framework metrics into an OTel Meter — pipe through OTel’s exporters (OTLP, Datadog, Grafana Cloud, Honeycomb, New Relic, etc.) without touching the framework’s metrics API.

import { metrics as otelMetrics } from '@opentelemetry/api';
import { MeterProvider } from '@opentelemetry/sdk-metrics';
import { OTLPMetricExporter } from '@opentelemetry/exporter-metrics-otlp-http';
import { PeriodicExportingMetricReader } from '@opentelemetry/sdk-metrics';
import { ActorSystem, MetricsExtensionId, OtelMetricsAdapter } from 'actor-ts';


// 1. Set up OTel SDK
const meterProvider = new MeterProvider({
  readers: [new PeriodicExportingMetricReader({
    exporter: new OTLPMetricExporter({
      url: 'https://otel-collector.example.com/v1/metrics',
    }),
    exportIntervalMillis: 10_000,
  })],
});
otelMetrics.setGlobalMeterProvider(meterProvider);


// 2. Wire the adapter
const system = ActorSystem.create('my-app');
const metrics = system.extension(MetricsExtensionId);


const adapter = new OtelMetricsAdapter({
  source: metrics.registry,
  meter:  otelMetrics.getMeter('actor-ts'),
});


adapter.start();

The OTel SDK handles exporting; the framework just emits to its registry as usual.

Configuration

Section titled “Configuration”
interface OtelMetricsAdapterSettings {
  source:           MetricsRegistry;
  meter:            Meter;                 // from @opentelemetry/api
  syncIntervalMs?:  number;                // default 5000
  prefix?:          string;
}

Same shape as the prom-client adapter — a source registry, a target SDK component, configurable sync interval.

Type mapping

Section titled “Type mapping”
Framework typeOTel instrument
CounterCounter
GaugeObservableGauge
HistogramHistogram
TimerHistogram

Labels become OTel attributes. Bucket boundaries propagate for histograms.

Peer dependencies

Section titled “Peer dependencies”
Terminal window
npm install @opentelemetry/api @opentelemetry/sdk-metrics
# Plus an exporter:
npm install @opentelemetry/exporter-metrics-otlp-http
# or @opentelemetry/exporter-prometheus, etc.

The framework doesn’t bundle OTel packages — bring your own SDK + exporter combo.

When to use OTel vs Prometheus exporter

Section titled “When to use OTel vs Prometheus exporter”
Use OTel when…Use Prometheus exporter when…
Targeting Datadog / New Relic / Grafana Cloud / HoneycombTargeting self-hosted Prometheus
Mixing metrics, traces, logs into one OTel pipelineMetrics-only setups
Pushing metrics out (no scrape)Pull-based scraping

OTel is the flexible path — your backend can change without rewiring instrumentation. Prometheus exporter is the simpler path when your backend is just Prometheus.

Combining with tracing

Section titled “Combining with tracing”
const tracer = otelMetrics.getMeter('actor-ts');
// + the tracing adapter:
const tracerAdapter = new OtelTracerAdapter({
  tracer: otelTrace.getTracer('actor-ts'),
});

Both adapters share the OTel SDK setup — metrics and traces flow through the same OTel exporter pipeline, correlated by the OTel context.

See OTel tracing adapter.

Where to next

Section titled “Where to next”