add validation test for v3 metrics for all its endpoints (#20094)

add unit test for v3 metrics for all its exposed endpoints

Bonus:
  - support OpenMetrics encoding
  - adds boot time for prometheus
  - continueOnError is better to serve as
    much metrics as possible.

cmd/metrics-router.go

@@ -38,7 +38,8 @@ const (
 
 // Standard env prometheus auth type
 const (
-	EnvPrometheusAuthType = "MINIO_PROMETHEUS_AUTH_TYPE"
+	EnvPrometheusAuthType    = "MINIO_PROMETHEUS_AUTH_TYPE"
+	EnvPrometheusOpenMetrics = "MINIO_PROMETHEUS_OPEN_METRICS"
 )
 
 type prometheusAuthType string
@@ -58,14 +59,15 @@ func registerMetricsRouter(router *mux.Router) {
 	if authType == prometheusPublic {
 		auth = NoAuthMiddleware
 	}
+
 	metricsRouter.Handle(prometheusMetricsPathLegacy, auth(metricsHandler()))
 	metricsRouter.Handle(prometheusMetricsV2ClusterPath, auth(metricsServerHandler()))
 	metricsRouter.Handle(prometheusMetricsV2BucketPath, auth(metricsBucketHandler()))
 	metricsRouter.Handle(prometheusMetricsV2NodePath, auth(metricsNodeHandler()))
 	metricsRouter.Handle(prometheusMetricsV2ResourcePath, auth(metricsResourceHandler()))
 
-	// Metrics v3!
+	// Metrics v3
-	metricsV3Server := newMetricsV3Server(authType)
+	metricsV3Server := newMetricsV3Server(auth)
 
 	// Register metrics v3 handler. It also accepts an optional query
 	// parameter `?list` - see handler for details.

cmd/metrics-v3-handler.go

@@ -23,9 +23,12 @@ import (
 	"net/http"
 	"slices"
 	"strings"
+	"sync"
 
+	"github.com/minio/minio/internal/config"
 	"github.com/minio/minio/internal/mcontext"
 	"github.com/minio/mux"
+	"github.com/minio/pkg/v3/env"
 	"github.com/prometheus/client_golang/prometheus"
 	"github.com/prometheus/client_golang/prometheus/promhttp"
 )
@@ -33,41 +36,39 @@ import (
 type promLogger struct{}
 
 func (p promLogger) Println(v ...interface{}) {
-	s := make([]string, 0, len(v))
+	metricsLogIf(GlobalContext, fmt.Errorf("metrics handler error: %v", v))
-	for _, val := range v {
-		s = append(s, fmt.Sprintf("%v", val))
-	}
-	err := fmt.Errorf("metrics handler error: %v", strings.Join(s, " "))
-	metricsLogIf(GlobalContext, err)
 }
 
 type metricsV3Server struct {
 	registry *prometheus.Registry
 	opts     promhttp.HandlerOpts
-	authFn   func(http.Handler) http.Handler
+	auth     func(http.Handler) http.Handler
 
 	metricsData *metricsV3Collection
 }
 
-func newMetricsV3Server(authType prometheusAuthType) *metricsV3Server {
+var (
+	globalMetricsV3CollectorPaths []collectorPath
+	globalMetricsV3Once           sync.Once
+)
+
+func newMetricsV3Server(auth func(h http.Handler) http.Handler) *metricsV3Server {
 	registry := prometheus.NewRegistry()
-	authFn := AuthMiddleware
-	if authType == prometheusPublic {
-		authFn = NoAuthMiddleware
-	}
-
 	metricGroups := newMetricGroups(registry)
-
+	globalMetricsV3Once.Do(func() {
+		globalMetricsV3CollectorPaths = metricGroups.collectorPaths
+	})
 	return &metricsV3Server{
 		registry: registry,
 		opts: promhttp.HandlerOpts{
 			ErrorLog:            promLogger{},
-			ErrorHandling:       promhttp.HTTPErrorOnError,
+			ErrorHandling:       promhttp.ContinueOnError,
 			Registry:            registry,
 			MaxRequestsInFlight: 2,
+			EnableOpenMetrics:   env.Get(EnvPrometheusOpenMetrics, config.EnableOff) == config.EnableOn,
+			ProcessStartTime:    globalBootTime,
 		},
-		authFn: authFn,
+		auth:        auth,
-
 		metricsData: metricGroups,
 	}
 }
@@ -221,7 +222,7 @@ func (h *metricsV3Server) ServeHTTP(w http.ResponseWriter, r *http.Request) {
 	pathComponents := mux.Vars(r)["pathComps"]
 	isListingRequest := r.Form.Has("list")
 
-	buckets := []string{}
+	var buckets []string
 	if strings.HasPrefix(pathComponents, "/bucket/") {
 		// bucket specific metrics, extract the bucket name from the path.
 		// it's the last part of the path. e.g. /bucket/api/<bucket-name>
@@ -246,5 +247,5 @@ func (h *metricsV3Server) ServeHTTP(w http.ResponseWriter, r *http.Request) {
 	})
 
 	// Add authentication
-	h.authFn(tracedHandler).ServeHTTP(w, r)
+	h.auth(tracedHandler).ServeHTTP(w, r)
 }

cmd/metrics-v3-types.go

@@ -35,8 +35,8 @@ type collectorPath string
 // converted to snake-case (by replaced '/' and '-' with '_') and prefixed with
 // `minio_`.
 func (cp collectorPath) metricPrefix() string {
-	s := strings.TrimPrefix(string(cp), "/")
+	s := strings.TrimPrefix(string(cp), SlashSeparator)
-	s = strings.ReplaceAll(s, "/", "_")
+	s = strings.ReplaceAll(s, SlashSeparator, "_")
 	s = strings.ReplaceAll(s, "-", "_")
 	return "minio_" + s
 }
@@ -56,8 +56,8 @@ func (cp collectorPath) isDescendantOf(arg string) bool {
 	if len(arg) >= len(descendant) {
 		return false
 	}
-	if !strings.HasSuffix(arg, "/") {
+	if !strings.HasSuffix(arg, SlashSeparator) {
-		arg += "/"
+		arg += SlashSeparator
 	}
 	return strings.HasPrefix(descendant, arg)
 }
@@ -72,10 +72,11 @@ const (
 	GaugeMT
 	// HistogramMT - represents a histogram metric.
 	HistogramMT
-	// rangeL - represents a range label.
-	rangeL = "range"
 )
 
+// rangeL - represents a range label.
+const rangeL = "range"
+
 func (mt MetricType) String() string {
 	switch mt {
 	case CounterMT:

cmd/server_test.go

@@ -35,6 +35,7 @@ import (
 	"time"
 
 	"github.com/dustin/go-humanize"
+	jwtgo "github.com/golang-jwt/jwt/v4"
 	"github.com/minio/minio-go/v7/pkg/set"
 	xhttp "github.com/minio/minio/internal/http"
 	"github.com/minio/pkg/v3/policy"
@@ -122,6 +123,9 @@ func runAllTests(suite *TestSuiteCommon, c *check) {
 	suite.TestObjectMultipartListError(c)
 	suite.TestObjectValidMD5(c)
 	suite.TestObjectMultipart(c)
+	suite.TestMetricsV3Handler(c)
+	suite.TestBucketSQSNotificationWebHook(c)
+	suite.TestBucketSQSNotificationAMQP(c)
 	suite.TearDownSuite(c)
 }
 
@@ -189,6 +193,36 @@ func (s *TestSuiteCommon) TearDownSuite(c *check) {
 	s.testServer.Stop()
 }
 
+const (
+	defaultPrometheusJWTExpiry = 100 * 365 * 24 * time.Hour
+)
+
+func (s *TestSuiteCommon) TestMetricsV3Handler(c *check) {
+	jwt := jwtgo.NewWithClaims(jwtgo.SigningMethodHS512, jwtgo.StandardClaims{
+		ExpiresAt: time.Now().UTC().Add(defaultPrometheusJWTExpiry).Unix(),
+		Subject:   s.accessKey,
+		Issuer:    "prometheus",
+	})
+
+	token, err := jwt.SignedString([]byte(s.secretKey))
+	c.Assert(err, nil)
+
+	for _, cpath := range globalMetricsV3CollectorPaths {
+		request, err := newTestSignedRequest(http.MethodGet, s.endPoint+minioReservedBucketPath+metricsV3Path+string(cpath),
+			0, nil, s.accessKey, s.secretKey, s.signer)
+		c.Assert(err, nil)
+
+		request.Header.Set("Authorization", "Bearer "+token)
+
+		// execute the request.
+		response, err := s.client.Do(request)
+		c.Assert(err, nil)
+
+		// assert the http response status code.
+		c.Assert(response.StatusCode, http.StatusOK)
+	}
+}
+
 func (s *TestSuiteCommon) TestBucketSQSNotificationWebHook(c *check) {
 	// Sample bucket notification.
 	bucketNotificationBuf := `<NotificationConfiguration><QueueConfiguration><Event>s3:ObjectCreated:Put</Event><Filter><S3Key><FilterRule><Name>prefix</Name><Value>images/</Value></FilterRule></S3Key></Filter><Id>1</Id><Queue>arn:minio:sqs:us-east-1:444455556666:webhook</Queue></QueueConfiguration></NotificationConfiguration>`