gometrics

A lightweight, zero-dependency metrics collection library for Go.

Features

• Thread-safe — built for concurrent Go applications
• Zero dependencies — uses only Go standard library
• Simple API — one-line metric tracking
• Prometheus compatible — industry-standard format
• Three metric types:
  • Counter — tracks totals (only increases)
  • Gauge — tracks current values (increases/decreases)
  • Timer — measures durations

Installation

go get github.com/eyop23/gometrics

Quick Start

package main

import (
    "net/http"
    "time"
    "github.com/eyop23/gometrics/pkg/metrics"
)

func handler(w http.ResponseWriter, r *http.Request) {
    // Track requests
    metrics.Inc("http.requests")
    
    // Track active connections
    metrics.GaugeInc("active.connections")
    defer metrics.GaugeDec("active.connections")
    
    // Track response time
    metrics.TrackFunc("api.response.time", func() {
        // your business logic here
        time.Sleep(time.Millisecond * 100)
    })
    
    w.Write([]byte("Hello World"))
}

func main() {
    http.HandleFunc("/", handler)
    
    // Start metrics server
    go metrics.Serve(":9090")
    
    // Start application
    http.ListenAndServe(":8080", nil)
}

Usage

Counters (only increase)

metrics.Inc("user.signups")           // increment by 1
metrics.IncBy("bytes.sent", 1024)     // increment by specific amount

Gauges (increase or decrease)

metrics.Set("queue.size", 42)         // set to specific value
metrics.GaugeInc("active.users")      // increment by 1
metrics.GaugeDec("active.users")      // decrement by 1

Timers (measure duration)

// Manual timing
start := time.Now()
doWork()
metrics.Track("work.duration", time.Since(start))

// Automatic timing
metrics.TrackFunc("database.query", func() {
    db.Query("SELECT * FROM users")
})

Viewing Metrics

Start your application and visit:

http://localhost:9090/metrics

Example output:

http.requests 1543
active.connections 12
user.signups 89
api.response.time_avg_ms 120
api.response.time_count 1543

Integration with Monitoring Tools

This library outputs metrics in Prometheus format. You can:

• Point Prometheus at the /metrics endpoint
• Create Grafana dashboards from the data
• Set up alerts based on thresholds

License