Metrics

Export Traefik metrics to Prometheus, Datadog, StatsD, InfluxDB, and OpenTelemetry for performance monitoring and analysis.

Metrics

Traefik provides comprehensive metrics for monitoring performance, traffic patterns, and system health. Metrics are available in multiple formats to integrate with your existing monitoring infrastructure.

Supported Backends

Traefik supports these metrics backends:

• Prometheus - Pull-based scraping with histogram support
• Datadog - Push metrics to Datadog Agent via StatsD or UDS
• StatsD - Standard StatsD protocol
• InfluxDB 2.x - Time-series database with tags
• OpenTelemetry - OTLP protocol for metrics

Quick Start

Enable Prometheus metrics:

YAML

metrics:
  prometheus: {}

Access metrics at http://localhost:8080/metrics (on the Traefik entrypoint).

Common Configuration

Add Internals

addInternals · boolean · default: "false"

Enable metrics for internal resources like ping@internal.

YAML

metrics:
  addInternals: true
  prometheus: {}

Prometheus

Prometheus is a pull-based monitoring system that scrapes metrics from HTTP endpoints.

Basic Configuration

YAML

metrics:
  prometheus:
    buckets:
      - 0.1
      - 0.3
      - 1.2
      - 5.0
    addEntryPointsLabels: true
    addRoutersLabels: true
    addServicesLabels: true

Configuration Options

buckets · array · default: "[0.1, 0.3, 1.2, 5.0]"

Histogram buckets for latency metrics (in seconds).

addEntryPointsLabels · boolean · default: "true"

Enable metrics on entrypoints.

addRoutersLabels · boolean · default: "false"

Enable metrics on routers (high cardinality - use with caution).

addServicesLabels · boolean · default: "true"

Enable metrics on services.

Custom EntryPoint

Expose metrics on a dedicated entrypoint:

YAML

entryPoints:
  web:
    address: ':80'
  metrics:
    address: ':8082'

metrics:
  prometheus:
    entryPoint: metrics

Manual Routing

manualRouting · boolean · default: "false"

Disable the default internal router to create custom routing for prometheus@internal service.

YAML

metrics:
  prometheus:
    manualRouting: true

Header Labels

Add custom labels from request headers:

headerLabels · object · default: "{}"

Map of label names to request header names.

YAML

metrics:
  prometheus:
    headerLabels:
      useragent: User-Agent
      country: X-Country-Code

The Host header is promoted to Request.Host in Go and is not available in the header map. Use X-Forwarded-Host instead.

Prometheus Scrape Configuration

scrape_configs:
  - job_name: 'traefik'
    static_configs:
      - targets: ['traefik:8080']

Datadog

Send metrics to Datadog Agent via StatsD protocol or Unix Domain Socket.

Basic Configuration

YAML

metrics:
  datadog:
    address: 127.0.0.1:8125
    pushInterval: 10s
    addEntryPointsLabels: true
    addServicesLabels: true
    prefix: traefik

Configuration Options

address · string · default: "127.0.0.1:8125" · required

Address of the Datadog Agent. Supports UDP and Unix Domain Sockets.

Supported address formats:

• 127.0.0.1:8125 - UDP socket
• unix:///path/to/datadog.socket - Unix socket (auto-detect type)
• unixgram:///path/to/datadog.socket - SOCK_DGRAM socket
• unixstream:///path/to/datadog.socket - SOCK_STREAM socket

pushInterval · duration · default: "10s"

Interval between metric pushes to Datadog Agent.

prefix · string · default: "traefik"

Prefix for all metrics.

Unix Domain Socket Example

YAML

metrics:
  datadog:
    address: unix:///var/run/datadog/dsd.socket

StatsD

Send metrics using the StatsD protocol.

Basic Configuration

YAML

metrics:
  statsD:
    address: localhost:8125
    pushInterval: 10s
    prefix: traefik
    addEntryPointsLabels: true
    addServicesLabels: true

Configuration Options

address · string · default: "localhost:8125" · required

Address of the StatsD server.

pushInterval · duration · default: "10s"

Interval between metric pushes.

prefix · string · default: "traefik"

Prefix for all metrics.

InfluxDB 2.x

Send metrics to InfluxDB 2.x time-series database.

Basic Configuration

YAML

metrics:
  influxDB2:
    address: http://localhost:8086
    token: mytoken
    org: myorg
    bucket: traefik
    pushInterval: 10s
    additionalLabels:
      environment: production
      datacenter: us-east-1

Configuration Options

address · string · default: "http://localhost:8086" · required

Address of the InfluxDB 2.x instance.

token · string · required

Authentication token. Can be a token value or file path.

org · string · required

Organization name where metrics are stored.

bucket · string · required

Bucket name where metrics are stored.

pushInterval · duration · default: "10s"

Interval between metric pushes.

additionalLabels · object · default: "{}"

Additional InfluxDB tags on all metrics.

OpenTelemetry

Export metrics using the OpenTelemetry Protocol (OTLP).

YAML

metrics:
  otlp:
    http:
      endpoint: https://otel-collector:4318/v1/metrics
    grpc:
      endpoint: otel-collector:4317

See the metrics overview documentation for OpenTelemetry semantic conventions.

Available Metrics

Global Metrics

Metric	Type	Labels	Description
traefik_config_reloads_total	Counter	-	Total configuration reloads
traefik_config_last_reload_success	Gauge	-	Timestamp of last successful reload
traefik_open_connections	Gauge	entrypoint, protocol	Current open connections
traefik_tls_certs_not_after	Gauge	-	TLS certificate expiration timestamp

EntryPoint Metrics

Metric	Type	Labels	Description
traefik_entrypoint_requests_total	Counter	code, method, protocol, entrypoint	Total HTTP requests
traefik_entrypoint_requests_tls_total	Counter	tls_version, tls_cipher, entrypoint	Total HTTPS requests
traefik_entrypoint_request_duration_seconds	Histogram	code, method, protocol, entrypoint	Request duration
traefik_entrypoint_requests_bytes_total	Counter	code, method, protocol, entrypoint	Total request bytes
traefik_entrypoint_responses_bytes_total	Counter	code, method, protocol, entrypoint	Total response bytes

Router Metrics

High Cardinality Warning

Router metrics can create high cardinality in your metrics system. Enable only when necessary.

Metric	Type	Labels	Description
traefik_router_requests_total	Counter	code, method, protocol, router, service	Total requests per router
traefik_router_requests_tls_total	Counter	tls_version, tls_cipher, router, service	Total HTTPS requests per router
traefik_router_request_duration_seconds	Histogram	code, method, protocol, router, service	Request duration per router
traefik_router_requests_bytes_total	Counter	code, method, protocol, router, service	Request bytes per router
traefik_router_responses_bytes_total	Counter	code, method, protocol, router, service	Response bytes per router

Service Metrics

Metric	Type	Labels	Description
traefik_service_requests_total	Counter	code, method, protocol, service	Total requests per service
traefik_service_requests_tls_total	Counter	tls_version, tls_cipher, service	Total HTTPS requests per service
traefik_service_request_duration_seconds	Histogram	code, method, protocol, service	Request duration per service
traefik_service_retries_total	Counter	service	Total request retries
traefik_service_server_up	Gauge	service, url	Server health status (0=down, 1=up)
traefik_service_requests_bytes_total	Counter	code, method, protocol, service	Request bytes per service
traefik_service_responses_bytes_total	Counter	code, method, protocol, service	Response bytes per service

Metric Labels

Label	Description	Example
code	HTTP status code	200, 404, 500
entrypoint	EntryPoint name	web, websecure
method	HTTP method	GET, POST
protocol	Protocol	http, https
router	Router name	my-router@docker
service	Service name	my-service@docker
tls_version	TLS version	1.2, 1.3
tls_cipher	TLS cipher suite	TLS_AES_128_GCM_SHA256
url	Service backend URL	http://backend:8080

HTTP method values outside the standard HTTP/1.1 methods or HTTP/2 PRI verb are labeled as EXTENSION_METHOD.

OpenTelemetry Semantic Conventions

Traefik follows OpenTelemetry semantic conventions v1.23.1:

HTTP Server Metrics

Metric	Type	Description
http.server.request.duration	Histogram	Duration of HTTP server requests

Labels: error.type, http.request.method, http.response.status_code, network.protocol.name, server.address, server.port, url.scheme

HTTP Client Metrics

Metric	Type	Description
http.client.request.duration	Histogram	Duration of HTTP client requests

Labels: error.type, http.request.method, http.response.status_code, network.protocol.name, server.address, server.port, url.scheme

Official Grafana Dashboards

Traefik provides official Grafana dashboards:

• On-Premises Dashboard - Dashboard ID: 17346
• Kubernetes Dashboard - Dashboard ID: 17347

Complete Examples

Production Prometheus Setup

YAML

entryPoints:
  web:
    address: ':80'
  websecure:
    address: ':443'
  metrics:
    address: ':8082'

metrics:
  prometheus:
    entryPoint: metrics
    buckets:
      - 0.1
      - 0.3
      - 1.2
      - 5.0
      - 10.0
    addEntryPointsLabels: true
    addRoutersLabels: false
    addServicesLabels: true

Multi-Backend Setup

YAML

metrics:
  prometheus:
    addServicesLabels: true
  datadog:
    address: 127.0.0.1:8125
    pushInterval: 10s

You can enable multiple metrics backends simultaneously.