dd-trace includes experimental support for OpenTelemetry metrics. It is designed as a drop-in replacement for the OpenTelemetry Metrics SDK and integrates with OTLP export infrastructure.
This feature is experimental. Enable it by setting DD_METRICS_OTEL_ENABLED=true.
Setup
DD_METRICS_OTEL_ENABLED=true node app.js
require('dd-trace').init()
const { metrics } = require('@opentelemetry/api')
const meter = metrics.getMeter('my-service', '1.0.0')
Instrument types
Counter
A monotonically increasing value. Use for counts of events.
const requestCounter = meter.createCounter('http.requests', {
description: 'Total HTTP requests',
unit: 'requests',
})
// Record a value with attributes
requestCounter.add(1, { method: 'GET', status: 200 })
Histogram
Records distributions of values. Use for latencies, sizes, and other measured values.
const durationHistogram = meter.createHistogram('http.duration', {
description: 'HTTP request duration',
unit: 'ms',
})
durationHistogram.record(145, { route: '/api/users' })
UpDownCounter
A counter that can increase or decrease. Use for values that go up and down, such as active connections.
const connectionCounter = meter.createUpDownCounter('active.connections', {
description: 'Active connections',
unit: 'connections',
})
connectionCounter.add(1) // New connection
connectionCounter.add(-1) // Connection closed
ObservableGauge
An asynchronous instrument for values observed at collection time. Use for current state values like CPU usage or memory.
const cpuGauge = meter.createObservableGauge('system.cpu.usage', {
description: 'CPU usage percentage',
unit: 'percent',
})
cpuGauge.addCallback((result) => {
const cpuUsage = process.cpuUsage()
result.observe(cpuUsage.system / 1_000_000, { core: '0' })
})
Full example
require('dd-trace').init()
const { metrics } = require('@opentelemetry/api')
const meter = metrics.getMeter('my-service', '1.0.0')
// Counters
const requestCounter = meter.createCounter('http.requests')
const errorCounter = meter.createCounter('http.errors')
// Histogram
const latencyHistogram = meter.createHistogram('http.latency', { unit: 'ms' })
// UpDownCounter
const activeRequests = meter.createUpDownCounter('http.active_requests')
app.use((req, res, next) => {
const start = Date.now()
activeRequests.add(1)
res.on('finish', () => {
const duration = Date.now() - start
const labels = { method: req.method, status: res.statusCode }
requestCounter.add(1, labels)
latencyHistogram.record(duration, labels)
activeRequests.add(-1)
if (res.statusCode >= 500) {
errorCounter.add(1, labels)
}
})
next()
})
Configuration
The following environment variables control OpenTelemetry metrics behavior:
| Variable | Default | Description |
|---|
DD_METRICS_OTEL_ENABLED | false | Enable OpenTelemetry metrics support |
OTEL_EXPORTER_OTLP_METRICS_ENDPOINT | http://localhost:4318/v1/metrics | OTLP endpoint for metrics. Falls back to OTEL_EXPORTER_OTLP_ENDPOINT + /v1/metrics |
OTEL_EXPORTER_OTLP_METRICS_HEADERS | {} | Headers for metrics requests (JSON format). Falls back to OTEL_EXPORTER_OTLP_HEADERS |
OTEL_EXPORTER_OTLP_METRICS_PROTOCOL | http/protobuf | OTLP protocol. Options: http/protobuf, http/json. Falls back to OTEL_EXPORTER_OTLP_PROTOCOL |
OTEL_EXPORTER_OTLP_METRICS_TIMEOUT | 10000 | Request timeout in ms. Falls back to OTEL_EXPORTER_OTLP_TIMEOUT |
OTEL_METRIC_EXPORT_INTERVAL | 10000 | Export interval in ms |
OTEL_EXPORTER_OTLP_METRICS_TEMPORALITY_PREFERENCE | DELTA | Aggregation temporality. Options: CUMULATIVE, DELTA, LOWMEMORY |
OTEL_BSP_MAX_QUEUE_SIZE | 2048 | Maximum metrics to queue before dropping |
