Where are system metrics and indicators collected?

System metrics and indicators are usually collected from the following places:

1. Inside the application — using built-in monitoring libraries (e.g., Prometheus client for Go). Here, performance metrics, request counts, errors, etc., are collected.

2. At the infrastructure level — metrics from servers, containers, network equipment (CPU, memory, network).

3. In logging and tracing systems — for behavior and error analysis.

4. In specialized monitoring services — which aggregate and visualize data (Grafana, Datadog).

Example in Go using Prometheus:

import (
    "github.com/prometheus/client_golang/prometheus"
    "github.com/prometheus/client_golang/prometheus/promhttp"
    "net/http"
)

var requests = prometheus.NewCounter(
    prometheus.CounterOpts{
        Name: "http_requests_total",
        Help: "Number of HTTP requests",
    },
)

func init() {
    prometheus.MustRegister(requests)
}

func handler(w http.ResponseWriter, r *http.Request) {
    requests.Inc()
    w.Write([]byte("Hello, metrics!"))
}

func main() {
    http.HandleFunc("/", handler)
    http.Handle("/metrics", promhttp.Handler())
    http.ListenAndServe(":8080", nil)
}