[exporter/datadog]: Add datadog span operation name remapping config …

…option (#3444)

* [exporter/datadog]: add span remap option

* [exporter/datadog]: add documentation around span_name_remappings config option

* [exporter/datadog]: spelling and linting and other assorted small joys

* Update exporter/datadogexporter/translate_traces.go

Co-authored-by: Pablo Baeyens <pablo.baeyens@datadoghq.com>

* [exporter/datadog]: update span remapping to use a map

* [exporter/datadog]: linting span remapper

* [exporter/datadog]: add config formatting

* [exporter/datadog]: update span name remappings, make some details private

* [exporter/datadog]: make more details of denylister private

* [expoter/datadog]: also make traceconnection private

Co-authored-by: Pablo Baeyens <pablo.baeyens@datadoghq.com>

exporter/datadogexporter/config/config.go

@@ -108,6 +108,13 @@ type TracesConfig struct {
 	// all entries must be surrounded by double quotes and separated by commas.
 	// ignore_resources: ["(GET|POST) /healthcheck"]
 	IgnoreResources []string `mapstructure:"ignore_resources"`
+
+	// SpanNameRemappings is the map of datadog span names and preferred name to map to. This can be used to
+	// automatically map Datadog Span Operation Names to an updated value. All entries should be key/value pairs.
+	// span_name_remappings:
+	//   io.opentelemetry.javaagent.spring.client: spring.client
+	//   instrumentation::express.server: express
+	SpanNameRemappings map[string]string `mapstructure:"span_name_remappings"`
 }
 
 // TagsConfig defines the tag-related configuration
@@ -247,5 +254,16 @@ func (c *Config) Validate() error {
 			}
 		}
 	}
+
+	if c.Traces.SpanNameRemappings != nil {
+		for key, value := range c.Traces.SpanNameRemappings {
+			if value == "" {
+				return fmt.Errorf("'%s' is not valid value for span name remapping", value)
+			}
+			if key == "" {
+				return fmt.Errorf("'%s' is not valid key for span name remapping", key)
+			}
+		}
+	}
 	return nil
 }

exporter/datadogexporter/config/config_test.go

@@ -155,3 +155,12 @@ func TestIgnoreResourcesValidation(t *testing.T) {
 	require.NoError(t, noErr)
 	require.Error(t, err)
 }
+
+func TestSpanNameRemappingsValidation(t *testing.T) {
+	validCfg := Config{Traces: TracesConfig{SpanNameRemappings: map[string]string{"old.opentelemetryspan.name": "updated.name"}}}
+	invalidCfg := Config{Traces: TracesConfig{SpanNameRemappings: map[string]string{"oldname": ""}}}
+	noErr := validCfg.Validate()
+	err := invalidCfg.Validate()
+	require.NoError(t, noErr)
+	require.Error(t, err)
+}

exporter/datadogexporter/denylister.go

@@ -20,16 +20,16 @@ import (
 	"github.com/DataDog/datadog-agent/pkg/trace/exportable/pb"
 )
 
-// Denylister holds a list of regular expressions which will match resources
+// denylister holds a list of regular expressions which will match resources
 // on spans that should be dropped.
 // From: https://github.com/DataDog/datadog-agent/blob/a6872e436681ea2136cf8a67465e99fdb4450519/pkg/trace/filters/blacklister.go#L15-L19
-type Denylister struct {
+type denylister struct {
 	list []*regexp.Regexp
 }
 
-// Allows returns true if the Denylister permits this span.
+// allows returns true if the Denylister permits this span.
 // From: https://github.com/DataDog/datadog-agent/blob/a6872e436681ea2136cf8a67465e99fdb4450519/pkg/trace/filters/blacklister.go#L21-L29
-func (f *Denylister) Allows(span *pb.Span) bool {
+func (f *denylister) allows(span *pb.Span) bool {
 	for _, entry := range f.list {
 		if entry.MatchString(span.Resource) {
 			return false
@@ -38,11 +38,11 @@ func (f *Denylister) Allows(span *pb.Span) bool {
 	return true
 }
 
-// NewDenylister creates a new Denylister based on the given list of
+// newDenylister creates a new Denylister based on the given list of
 // regular expressions.
 // From: https://github.com/DataDog/datadog-agent/blob/a6872e436681ea2136cf8a67465e99fdb4450519/pkg/trace/filters/blacklister.go#L41-L45
-func NewDenylister(exprs []string) *Denylister {
-	return &Denylister{list: compileRules(exprs)}
+func newDenylister(exprs []string) *denylister {
+	return &denylister{list: compileRules(exprs)}
 }
 
 // compileRules compiles as many rules as possible from the list of expressions.

exporter/datadogexporter/denylister_test.go

@@ -68,16 +68,16 @@ func TestDenylister(t *testing.T) {
 	for _, test := range tests {
 		span := testSpan()
 		span.Resource = test.resource
-		filter := NewDenylister(test.filter)
+		filter := newDenylister(test.filter)
 
-		assert.Equal(t, test.expectation, filter.Allows(span))
+		assert.Equal(t, test.expectation, filter.allows(span))
 	}
 }
 
 func TestCompileRules(t *testing.T) {
-	filter := NewDenylister([]string{"\n{6}"})
+	filter := newDenylister([]string{"\n{6}"})
 	for i := 0; i < 100; i++ {
 		span := testSpan()
-		assert.True(t, filter.Allows(span))
+		assert.True(t, filter.allows(span))
 	}
 }

exporter/datadogexporter/example/config.yaml

@@ -120,7 +120,18 @@ exporters:
       ## A blacklist of regular expressions can be provided to disable certain traces based on their resource name
       ## all entries must be surrounded by double quotes and separated by commas.
       #
-      # ignore_resources: ["(GET|POST) /healthcheck"]      
+      # ignore_resources: ["(GET|POST) /healthcheck"] 
+
+      ## @param span_name_remappings - map of key/value pairs - optional
+      ## A map of Datadog span operation name keys and preferred name valuues to update those names to. This can be used to
+      ## automatically map Datadog Span Operation Names to an updated value, and is useful when a user wants to
+      ## shorten or modify span names to something more user friendly in the case of instrumentation libraries with
+      ## particularly verbose names.
+      #
+      # span_name_remappings: 
+      #   io.opentelemetry.javaagent.spring.client: spring.client
+      #   instrumentation::express.server: express
+
 
 service:
   pipelines:

exporter/datadogexporter/trace_connection.go

@@ -29,13 +29,13 @@ import (
 	"github.com/open-telemetry/opentelemetry-collector-contrib/exporter/datadogexporter/utils"
 )
 
-// TraceEdgeConnection is used to send data to trace edge
-type TraceEdgeConnection interface {
+// traceEdgeConnection is used to send data to trace edge
+type traceEdgeConnection interface {
 	SendTraces(ctx context.Context, trace *pb.TracePayload, maxRetries int) error
 	SendStats(ctx context.Context, stats *stats.Payload, maxRetries int) error
 }
 
-type traceEdgeConnection struct {
+type traceEdgeConnectionImpl struct {
 	traceURL           string
 	statsURL           string
 	apiKey             string
@@ -49,10 +49,10 @@ const (
 	traceEdgeRetryInterval time.Duration = 10 * time.Second
 )
 
-// createTraceEdgeConnection returns a new TraceEdgeConnection
-func createTraceEdgeConnection(rootURL, apiKey string, buildInfo component.BuildInfo) TraceEdgeConnection {
+// createTraceEdgeConnection returns a new traceEdgeConnection
+func createTraceEdgeConnection(rootURL, apiKey string, buildInfo component.BuildInfo) traceEdgeConnection {
 
-	return &traceEdgeConnection{
+	return &traceEdgeConnectionImpl{
 		traceURL:  rootURL + "/api/v0.2/traces",
 		statsURL:  rootURL + "/api/v0.2/stats",
 		buildInfo: buildInfo,
@@ -69,7 +69,7 @@ type Payload struct {
 }
 
 // SendTraces serializes a trace payload to protobuf and sends it to Trace Edge
-func (con *traceEdgeConnection) SendTraces(ctx context.Context, trace *pb.TracePayload, maxRetries int) error {
+func (con *traceEdgeConnectionImpl) SendTraces(ctx context.Context, trace *pb.TracePayload, maxRetries int) error {
 	binary, marshallErr := proto.Marshal(trace)
 	if marshallErr != nil {
 		return fmt.Errorf("failed to serialize trace payload to protobuf: %w", marshallErr)
@@ -108,7 +108,7 @@ func (con *traceEdgeConnection) SendTraces(ctx context.Context, trace *pb.TraceP
 }
 
 // SendStats serializes a stats payload to json and sends it to Trace Edge
-func (con *traceEdgeConnection) SendStats(ctx context.Context, sts *stats.Payload, maxRetries int) error {
+func (con *traceEdgeConnectionImpl) SendStats(ctx context.Context, sts *stats.Payload, maxRetries int) error {
 	var b bytes.Buffer
 	err := stats.EncodePayload(&b, sts)
 	if err != nil {
@@ -145,7 +145,7 @@ func (con *traceEdgeConnection) SendStats(ctx context.Context, sts *stats.Payloa
 }
 
 // sendPayloadToTraceEdge sends a payload to Trace Edge
-func (con *traceEdgeConnection) sendPayloadToTraceEdge(ctx context.Context, apiKey string, payload *Payload, url string) (bool, error) {
+func (con *traceEdgeConnectionImpl) sendPayloadToTraceEdge(ctx context.Context, apiKey string, payload *Payload, url string) (bool, error) {
 
 	// Create the request to be sent to the API
 	req, err := http.NewRequestWithContext(ctx, "POST", url, bytes.NewBuffer(payload.Bytes))

exporter/datadogexporter/traces_exporter.go

@@ -35,10 +35,10 @@ type traceExporter struct {
 	params         component.ExporterCreateSettings
 	cfg            *config.Config
 	ctx            context.Context
-	edgeConnection TraceEdgeConnection
+	edgeConnection traceEdgeConnection
 	obfuscator     *obfuscate.Obfuscator
 	client         *datadog.Client
-	denylister     *Denylister
+	denylister     *denylister
 }
 
 var (
@@ -69,7 +69,7 @@ func newTracesExporter(ctx context.Context, params component.ExporterCreateSetti
 	obfuscator := obfuscate.NewObfuscator(obfuscatorConfig)
 
 	// a denylist for dropping ignored resources
-	denylister := NewDenylister(cfg.Traces.IgnoreResources)
+	denylister := newDenylister(cfg.Traces.IgnoreResources)
 
 	exporter := &traceExporter{
 		params:         params,

exporter/datadogexporter/translate_traces.go

@@ -64,7 +64,7 @@ const (
 )
 
 // converts Traces into an array of datadog trace payloads grouped by env
-func convertToDatadogTd(td pdata.Traces, fallbackHost string, cfg *config.Config, blk *Denylister, buildInfo component.BuildInfo) ([]*pb.TracePayload, []datadog.Metric) {
+func convertToDatadogTd(td pdata.Traces, fallbackHost string, cfg *config.Config, blk *denylister, buildInfo component.BuildInfo) ([]*pb.TracePayload, []datadog.Metric) {
 	// TODO:
 	// do we apply other global tags, like version+service, to every span or only root spans of a service
 	// should globalTags['service'] take precedence over a trace's resource.service.name? I don't believe so, need to confirm
@@ -76,14 +76,19 @@ func convertToDatadogTd(td pdata.Traces, fallbackHost string, cfg *config.Config
 	seenHosts := make(map[string]struct{})
 	var series []datadog.Metric
 	pushTime := pdata.TimestampFromTime(time.Now())
+
+	spanNameMap := cfg.Traces.SpanNameRemappings
+
 	for i := 0; i < resourceSpans.Len(); i++ {
 		rs := resourceSpans.At(i)
 		host, ok := metadata.HostnameFromAttributes(rs.Resource().Attributes())
 		if !ok {
 			host = fallbackHost
 		}
+
 		seenHosts[host] = struct{}{}
-		payload := resourceSpansToDatadogSpans(rs, host, cfg, blk)
+		payload := resourceSpansToDatadogSpans(rs, host, cfg, blk, spanNameMap)
+
 		traces = append(traces, &payload)
 	}
 
@@ -123,7 +128,7 @@ func aggregateTracePayloadsByEnv(tracePayloads []*pb.TracePayload) []*pb.TracePa
 }
 
 // converts a Trace's resource spans into a trace payload
-func resourceSpansToDatadogSpans(rs pdata.ResourceSpans, hostname string, cfg *config.Config, blk *Denylister) pb.TracePayload {
+func resourceSpansToDatadogSpans(rs pdata.ResourceSpans, hostname string, cfg *config.Config, blk *denylister, spanNameMap map[string]string) pb.TracePayload {
 	// get env tag
 	env := utils.NormalizeTag(cfg.Env)
 
@@ -156,7 +161,7 @@ func resourceSpansToDatadogSpans(rs pdata.ResourceSpans, hostname string, cfg *c
 		extractInstrumentationLibraryTags(ils.InstrumentationLibrary(), datadogTags)
 		spans := ils.Spans()
 		for j := 0; j < spans.Len(); j++ {
-			span := spanToDatadogSpan(spans.At(j), resourceServiceName, datadogTags, cfg)
+			span := spanToDatadogSpan(spans.At(j), resourceServiceName, datadogTags, cfg, spanNameMap)
 			var apiTrace *pb.APITrace
 			var ok bool
 
@@ -193,7 +198,7 @@ func resourceSpansToDatadogSpans(rs pdata.ResourceSpans, hostname string, cfg *c
 		// Root span is used to carry some trace-level metadata, such as sampling rate and priority.
 		rootSpan := utils.GetRoot(apiTrace)
 
-		if !blk.Allows(rootSpan) {
+		if !blk.allows(rootSpan) {
 			// drop trace by not adding to payload if it's root span matches denylist
 			continue
 		}
@@ -215,6 +220,7 @@ func spanToDatadogSpan(s pdata.Span,
 	serviceName string,
 	datadogTags map[string]string,
 	cfg *config.Config,
+	spanNameMap map[string]string,
 ) *pb.Span {
 
 	tags := aggregateSpanTags(s, datadogTags)
@@ -276,7 +282,7 @@ func spanToDatadogSpan(s pdata.Span,
 	span := &pb.Span{
 		TraceID:  decodeAPMTraceID(s.TraceID().Bytes()),
 		SpanID:   decodeAPMSpanID(s.SpanID().Bytes()),
-		Name:     getDatadogSpanName(s, tags),
+		Name:     remapDatadogSpanName(getDatadogSpanName(s, tags), spanNameMap),
 		Resource: resourceName,
 		Service:  normalizedServiceName,
 		Start:    int64(startTime),
@@ -631,3 +637,13 @@ func eventsToString(evts pdata.SpanEventSlice) string {
 	eventArrayBytes, _ := json.Marshal(&eventArray)
 	return string(eventArrayBytes)
 }
+
+// remapDatadogSpanName allows users to map their datadog span operation names to
+// another string as they see fit.
+func remapDatadogSpanName(name string, spanNameMap map[string]string) string {
+	if updatedSpanName := spanNameMap[name]; updatedSpanName != "" {
+		return updatedSpanName
+	}
+
+	return name
+}