Allow selectively exporting all Go runtime metrics (#7990)

Signed-off-by: Ryan Brewster <rpb@anthropic.com>

plugin/metrics/README.md

@@ -55,7 +55,9 @@ This plugin can only be used once per Server Block.
 ## Syntax
 
 ~~~
-prometheus [ADDRESS]
+prometheus [ADDRESS] {
+    runtime_metrics
+}
 ~~~
 
 For each zone that you want to see metrics for.
@@ -63,6 +65,12 @@ For each zone that you want to see metrics for.
 It optionally takes a bind address to which the metrics are exported; the default
 listens on `localhost:9153`. The metrics path is fixed to `/metrics`.
 
+* `runtime_metrics` exports the full Go [runtime/metrics](https://pkg.go.dev/runtime/metrics)
+  set — notably `go_cpu_classes_*` for CPU attribution (GC mark/assist/pause vs user code)
+  and `go_sched_latencies_seconds` for goroutine scheduling delay. Adds roughly 100 scalars
+  and 8 histograms. This is a process-wide latch: enabling it in any server block enables it
+  for all, and it stays enabled across reloads until restart.
+
 ## Examples
 
 Use an alternative listening address:

plugin/metrics/setup.go

@@ -3,6 +3,7 @@ package metrics
 import (
 	"net"
 	"runtime"
+	"sync"
 
 	"github.com/coredns/caddy"
 	"github.com/coredns/coredns/core/dnsserver"
@@ -10,11 +11,19 @@ import (
 	"github.com/coredns/coredns/plugin"
 	"github.com/coredns/coredns/plugin/metrics/vars"
 	"github.com/coredns/coredns/plugin/pkg/uniq"
+
+	"github.com/prometheus/client_golang/prometheus"
+	"github.com/prometheus/client_golang/prometheus/collectors"
 )
 
 var (
 	u        = uniq.New()
 	registry = newReg()
+
+	// There is one Go runtime per process, so this is a latch: the first server
+	// block to enable runtime_metrics swaps the collector for everyone, and the
+	// swap persists across reloads until process restart.
+	runtimeMetricsOnce sync.Once
 )
 
 func init() { plugin.Register("prometheus", setup) }
@@ -96,9 +105,18 @@ func parse(c *caddy.Controller) (*Metrics, error) {
 			return met, c.ArgErr()
 		}
 
-		// Parse TLS block if present
 		for c.NextBlock() {
 			switch c.Val() {
+			case "runtime_metrics":
+				if len(c.RemainingArgs()) != 0 {
+					return nil, c.ArgErr()
+				}
+				runtimeMetricsOnce.Do(func() {
+					prometheus.Unregister(collectors.NewGoCollector())
+					prometheus.MustRegister(collectors.NewGoCollector(
+						collectors.WithGoCollectorRuntimeMetrics(collectors.MetricsAll),
+					))
+				})
 			case "tls":
 				if met.tlsConfigPath != "" {
 					return nil, c.Err("tls block already specified")

plugin/metrics/setup_test.go

@@ -15,8 +15,17 @@ func TestPrometheusParse(t *testing.T) {
 		// oks
 		{`prometheus`, false, "localhost:9153"},
 		{`prometheus localhost:53`, false, "localhost:53"},
+		{`prometheus {
+			runtime_metrics
+		}`, false, "localhost:9153"},
+		{`prometheus localhost:53 {
+			runtime_metrics
+		}`, false, "localhost:53"},
 		// fails
 		{`prometheus {}`, true, ""},
+		{`prometheus {
+			runtime_metrics extra_arg
+		}`, true, ""},
 		{`prometheus /foo`, true, ""},
 		{`prometheus a b c`, true, ""},
 	}