feat(dogstatsd): add cardinality common field

Signed-off-by: Wassim DHIF <[email protected]>
DataDog · Jan 19, 2025 · 29786f5 · 29786f5
1 parent cb0d65b
commit 29786f5
Show file tree

Hide file tree

Showing 8 changed files with 74 additions and 8 deletions.
diff --git a/comp/dogstatsd/server/enrich.go b/comp/dogstatsd/server/enrich.go
@@ -37,14 +37,15 @@ type enrichConfig struct {
 }
 
 // extractTagsMetadata returns tags (client tags + host tag) and information needed to query tagger (origins, cardinality).
-func extractTagsMetadata(tags []string, originFromUDS string, localData origindetection.LocalData, externalData origindetection.ExternalData, conf enrichConfig) ([]string, string, taggertypes.OriginInfo, metrics.MetricSource) {
+func extractTagsMetadata(tags []string, originFromUDS string, localData origindetection.LocalData, externalData origindetection.ExternalData, cardinality string, conf enrichConfig) ([]string, string, taggertypes.OriginInfo, metrics.MetricSource) {
 	host := conf.defaultHostname
 	metricSource := metrics.MetricSourceDogstatsd
 	origin := taggertypes.OriginInfo{
 		ContainerIDFromSocket: originFromUDS,
 		LocalData:             localData,
 		ExternalData:          externalData,
 		ProductOrigin:         origindetection.ProductOriginDogStatsD,
+		Cardinality:           cardinality,
 	}
 
 	n := 0
@@ -55,7 +56,7 @@ func extractTagsMetadata(tags []string, originFromUDS string, localData originde
 		} else if strings.HasPrefix(tag, entityIDTagPrefix) {
 			origin.LocalData.PodUID = tag[len(entityIDTagPrefix):]
 			continue
-		} else if strings.HasPrefix(tag, CardinalityTagPrefix) {
+		} else if strings.HasPrefix(tag, CardinalityTagPrefix) && origin.Cardinality == "" {
 			origin.Cardinality = tag[len(CardinalityTagPrefix):]
 			continue
 		} else if strings.HasPrefix(tag, jmxCheckNamePrefix) {
@@ -112,7 +113,7 @@ func tsToFloatForSamples(ts time.Time) float64 {
 
 func enrichMetricSample(dest []metrics.MetricSample, ddSample dogstatsdMetricSample, origin string, listenerID string, conf enrichConfig) []metrics.MetricSample {
 	metricName := ddSample.name
-	tags, hostnameFromTags, extractedOrigin, metricSource := extractTagsMetadata(ddSample.tags, origin, ddSample.localData, ddSample.externalData, conf)
+	tags, hostnameFromTags, extractedOrigin, metricSource := extractTagsMetadata(ddSample.tags, origin, ddSample.localData, ddSample.externalData, ddSample.cardinality, conf)
 
 	if !isExcluded(metricName, conf.metricPrefix, conf.metricPrefixBlacklist) {
 		metricName = conf.metricPrefix + metricName
@@ -192,7 +193,7 @@ func enrichEventAlertType(dogstatsdAlertType alertType) metricsevent.AlertType {
 }
 
 func enrichEvent(event dogstatsdEvent, origin string, conf enrichConfig) *metricsevent.Event {
-	tags, hostnameFromTags, extractedOrigin, _ := extractTagsMetadata(event.tags, origin, event.localData, event.externalData, conf)
+	tags, hostnameFromTags, extractedOrigin, _ := extractTagsMetadata(event.tags, origin, event.localData, event.externalData, event.cardinality, conf)
 
 	enrichedEvent := &metricsevent.Event{
 		Title:          event.title,
@@ -229,7 +230,7 @@ func enrichServiceCheckStatus(status serviceCheckStatus) servicecheck.ServiceChe
 }
 
 func enrichServiceCheck(serviceCheck dogstatsdServiceCheck, origin string, conf enrichConfig) *servicecheck.ServiceCheck {
-	tags, hostnameFromTags, extractedOrigin, _ := extractTagsMetadata(serviceCheck.tags, origin, serviceCheck.localData, serviceCheck.externalData, conf)
+	tags, hostnameFromTags, extractedOrigin, _ := extractTagsMetadata(serviceCheck.tags, origin, serviceCheck.localData, serviceCheck.externalData, serviceCheck.cardinality, conf)
 
 	enrichedServiceCheck := &servicecheck.ServiceCheck{
 		CheckName:  serviceCheck.name,

diff --git a/comp/dogstatsd/server/enrich_bench_test.go b/comp/dogstatsd/server/enrich_bench_test.go
@@ -35,7 +35,7 @@ func BenchmarkExtractTagsMetadata(b *testing.B) {
 			sb.ResetTimer()
 
 			for n := 0; n < sb.N; n++ {
-				tags, _, _, _ = extractTagsMetadata(baseTags, "", origindetection.LocalData{}, origindetection.ExternalData{}, conf)
+				tags, _, _, _ = extractTagsMetadata(baseTags, "", origindetection.LocalData{}, origindetection.ExternalData{}, "", conf)
 			}
 		})
 	}

diff --git a/comp/dogstatsd/server/enrich_test.go b/comp/dogstatsd/server/enrich_test.go
@@ -1115,6 +1115,7 @@ func TestEnrichTags(t *testing.T) {
 		originFromMsg []byte
 		localData     origindetection.LocalData
 		externalData  origindetection.ExternalData
+		cardinality   string
 		conf          enrichConfig
 	}
 	tests := []struct {
@@ -1463,12 +1464,46 @@ func TestEnrichTags(t *testing.T) {
 			},
 			wantedMetricSource: metrics.MetricSourceDogstatsd,
 		},
+		{
+			name: "cardinality field as none",
+			args: args{
+				cardinality: types.NoneCardinalityString,
+			},
+			wantedTags: nil,
+			wantedOrigin: taggertypes.OriginInfo{
+				Cardinality: "none",
+			},
+			wantedMetricSource: metrics.MetricSourceDogstatsd,
+		},
+		{
+			name: "cardinality field as high",
+			args: args{
+				cardinality: types.HighCardinalityString,
+			},
+			wantedTags: nil,
+			wantedOrigin: taggertypes.OriginInfo{
+				Cardinality: "high",
+			},
+			wantedMetricSource: metrics.MetricSourceDogstatsd,
+		},
+		{
+			name: "cardinality field with dd.internal.card",
+			args: args{
+				tags:        []string{"env:prod", "dd.internal.card:high"},
+				cardinality: types.NoneCardinalityString,
+			},
+			wantedTags: []string{"env:prod", "dd.internal.card:high"},
+			wantedOrigin: taggertypes.OriginInfo{
+				Cardinality: "none",
+			},
+			wantedMetricSource: metrics.MetricSourceDogstatsd,
+		},
 	}
 	for _, tt := range tests {
 		tt.wantedOrigin.ProductOrigin = origindetection.ProductOriginDogStatsD
 
 		t.Run(tt.name, func(t *testing.T) {
-			tags, host, origin, metricSource := extractTagsMetadata(tt.args.tags, tt.args.originFromUDS, tt.args.localData, tt.args.externalData, tt.args.conf)
+			tags, host, origin, metricSource := extractTagsMetadata(tt.args.tags, tt.args.originFromUDS, tt.args.localData, tt.args.externalData, tt.args.cardinality, tt.args.conf)
 			assert.Equal(t, tt.wantedTags, tags)
 			assert.Equal(t, tt.wantedHost, host)
 			assert.Equal(t, tt.wantedOrigin, origin)
@@ -1516,7 +1551,7 @@ func TestEnrichTagsWithJMXCheckName(t *testing.T) {
 	}
 	for _, tt := range tests {
 		t.Run(tt.name, func(t *testing.T) {
-			tags, _, _, metricSource := extractTagsMetadata(tt.tags, "", origindetection.LocalData{}, origindetection.ExternalData{}, enrichConfig{})
+			tags, _, _, metricSource := extractTagsMetadata(tt.tags, "", origindetection.LocalData{}, origindetection.ExternalData{}, "", enrichConfig{})
 			assert.Equal(t, tt.wantedTags, tags)
 			assert.Equal(t, tt.wantedMetricSource, metricSource)
 			assert.NotContains(t, tags, tt.jmxCheckName)

diff --git a/comp/dogstatsd/server/parse.go b/comp/dogstatsd/server/parse.go
@@ -48,6 +48,9 @@ var (
 
 	// externalDataPrefix is the prefix for a common field which contains the external data for Origin Detection.
 	externalDataPrefix = []byte("e:")
+
+	// cardinalityPrefix is the prefix for a common field which contains the cardinality for Origin Detection.
+	cardinalityPrefix = []byte("card:")
 )
 
 // parser parses dogstatsd messages
@@ -174,6 +177,7 @@ func (p *parser) parseMetricSample(message []byte) (dogstatsdMetricSample, error
 	var tags []string
 	var localData origindetection.LocalData
 	var externalData origindetection.ExternalData
+	var cardinality string
 	var optionalField []byte
 	var timestamp time.Time
 	for message != nil {
@@ -207,6 +211,9 @@ func (p *parser) parseMetricSample(message []byte) (dogstatsdMetricSample, error
 		// external data
 		case p.dsdOriginEnabled && bytes.HasPrefix(optionalField, externalDataPrefix):
 			externalData = p.parseExternalData(optionalField[len(externalDataPrefix):])
+		// cardinality
+		case p.dsdOriginEnabled && bytes.HasPrefix(optionalField, cardinalityPrefix):
+			cardinality = string(optionalField[len(cardinalityPrefix):])
 		}
 	}
 
@@ -220,6 +227,7 @@ func (p *parser) parseMetricSample(message []byte) (dogstatsdMetricSample, error
 		tags:         tags,
 		localData:    localData,
 		externalData: externalData,
+		cardinality:  cardinality,
 		ts:           timestamp,
 	}, nil
 }

diff --git a/comp/dogstatsd/server/parse_events.go b/comp/dogstatsd/server/parse_events.go
@@ -43,6 +43,8 @@ type dogstatsdEvent struct {
 	localData origindetection.LocalData
 	// externalData is used for Origin Detection
 	externalData origindetection.ExternalData
+	// cardinality is used for Origin Detection
+	cardinality string
 }
 
 type eventHeader struct {
@@ -170,6 +172,8 @@ func (p *parser) applyEventOptionalField(event dogstatsdEvent, optionalField []b
 		newEvent.localData = p.parseLocalData(optionalField[len(localDataPrefix):])
 	case p.dsdOriginEnabled && bytes.HasPrefix(optionalField, externalDataPrefix):
 		newEvent.externalData = p.parseExternalData(optionalField[len(externalDataPrefix):])
+	case p.dsdOriginEnabled && bytes.HasPrefix(optionalField, cardinalityPrefix):
+		newEvent.cardinality = string(optionalField[len(cardinalityPrefix):])
 	}
 	if err != nil {
 		return event, err

diff --git a/comp/dogstatsd/server/parse_metrics.go b/comp/dogstatsd/server/parse_metrics.go
@@ -51,6 +51,8 @@ type dogstatsdMetricSample struct {
 	localData origindetection.LocalData
 	// externalData is used for Origin Detection
 	externalData origindetection.ExternalData
+	// cardinality is used for Origin Detection
+	cardinality string
 	// timestamp read in the message if any
 	ts time.Time
 }

diff --git a/comp/dogstatsd/server/parse_service_checks.go b/comp/dogstatsd/server/parse_service_checks.go
@@ -34,6 +34,8 @@ type dogstatsdServiceCheck struct {
 	localData origindetection.LocalData
 	// externalData is used for Origin Detection
 	externalData origindetection.ExternalData
+	// cardinality is used for Origin Detection
+	cardinality string
 }
 
 var (
@@ -104,6 +106,8 @@ func (p *parser) applyServiceCheckOptionalField(serviceCheck dogstatsdServiceChe
 		newServiceCheck.localData = p.parseLocalData(optionalField[len(localDataPrefix):])
 	case p.dsdOriginEnabled && bytes.HasPrefix(optionalField, externalDataPrefix):
 		newServiceCheck.externalData = p.parseExternalData(optionalField[len(externalDataPrefix):])
+	case p.dsdOriginEnabled && bytes.HasPrefix(optionalField, cardinalityPrefix):
+		newServiceCheck.cardinality = string(optionalField[len(cardinalityPrefix):])
 	}
 	if err != nil {
 		return serviceCheck, err

diff --git a/releasenotes/notes/dogstatsd_cardinality_field-564b3d846700335c.yaml b/releasenotes/notes/dogstatsd_cardinality_field-564b3d846700335c.yaml
@@ -0,0 +1,12 @@
+# Each section from every release note are combined when the
+# CHANGELOG.rst is rendered. So the text needs to be worded so that
+# it does not depend on any information only available in another
+# section. This may mean repeating some details, but each section
+# must be readable independently of the other.
+#
+# Each section note must be formatted as reStructuredText.
+---
+features:
+  - |
+    Add new `card:` common field to DogStatsD Datagram specification to allow
+    customer to specify the cardinality of the metric. This field is optional.