[processor/tailsampling] Remove internally used fields from samplingpolicy.TraceData (open-telemetry#44435)

#### Description

These fields are only used by the internals of the tail sampling
processor to track metrics or the decision that was made. They are not
needed for any samplers, and therefore should not be used by any sampler
extensions either.

Author: csmarchbanks

.chloggen/remove-extra-fields-in-trace-data.yaml

@@ -0,0 +1,27 @@
+# Use this changelog template to create an entry for release notes.
+
+# One of 'breaking', 'deprecation', 'new_component', 'enhancement', 'bug_fix'
+change_type: breaking
+
+# The name of the component, or a single word describing the area of concern, (e.g. receiver/filelog)
+component: processor/tail_sampling
+
+# A brief description of the change.  Surround your text with quotes ("") if it needs to start with a backtick (`).
+note: Remove only internally relevant fields from samplingpolicy.TraceData.
+
+# Mandatory: One or more tracking issues related to the change. You can use the PR number here if no issue exists.
+issues: [44435]
+
+# (Optional) One or more lines of additional information to render under the primary note.
+# These lines will be padded with 2 spaces and then inserted directly into the document.
+# Use pipe (|) for multiline entries.
+subtext:
+
+# If your change doesn't affect end users or the exported elements of any package,
+# you should instead start your pull request title with [chore] or use the "Skip Changelog" label.
+# Optional: The change log or logs in which this entry should be included.
+# e.g. '[user]' or '[user, api]'
+# Include 'user' if the change is relevant to end users.
+# Include 'api' if there is a change to a library API.
+# Default: '[user]'
+change_logs: [api]

processor/tailsamplingprocessor/pkg/samplingpolicy/samplingpolicy.go

@@ -5,24 +5,17 @@ package samplingpolicy // import "github.com/open-telemetry/opentelemetry-collec
 
 import (
 	"context"
-	"time"
 
 	"go.opentelemetry.io/collector/pdata/pcommon"
 	"go.opentelemetry.io/collector/pdata/ptrace"
 )
 
 // TraceData stores the sampling related trace data.
 type TraceData struct {
-	// Arrival time the first span for the trace was received.
-	ArrivalTime time.Time
-	// DecisionTime time when sampling decision was taken.
-	DecisionTime time.Time
 	// SpanCount track the number of spans on the trace.
 	SpanCount int64
 	// ReceivedBatches stores all the batches received for the trace.
 	ReceivedBatches ptrace.Traces
-	// FinalDecision.
-	FinalDecision Decision
 }
 
 // Decision gives the status of sampling decision.

processor/tailsamplingprocessor/processor.go

@@ -41,6 +41,17 @@ type policy struct {
 	attribute metric.MeasurementOption
 }
 
+// traceData is a wrapper around the publically used samplingpolicy.TraceData
+// that tracks information related to the decision making process but not
+// needed by any sampler implementations.
+type traceData struct {
+	samplingpolicy.TraceData
+
+	arrivalTime   time.Time
+	decisionTime  time.Time
+	finalDecision samplingpolicy.Decision
+}
+
 type tailSamplingSpanProcessor struct {
 	ctx context.Context
 
@@ -51,7 +62,7 @@ type tailSamplingSpanProcessor struct {
 	deleteTraceQueue   *list.List
 	nextConsumer       consumer.Traces
 	policies           []*policy
-	idToTrace          map[pcommon.TraceID]*samplingpolicy.TraceData
+	idToTrace          map[pcommon.TraceID]*traceData
 	tickerFrequency    time.Duration
 	decisionBatcher    idbatcher.Batcher
 	sampledIDCache     cache.Cache[bool]
@@ -103,7 +114,7 @@ func newTracesProcessor(ctx context.Context, set processor.Settings, nextConsume
 		sampledIDCache:     sampledDecisions,
 		nonSampledIDCache:  nonSampledDecisions,
 		logger:             set.Logger,
-		idToTrace:          make(map[pcommon.TraceID]*samplingpolicy.TraceData),
+		idToTrace:          make(map[pcommon.TraceID]*traceData),
 		deleteTraceQueue:   list.New(),
 		sampleOnFirstMatch: cfg.SampleOnFirstMatch,
 		blockOnOverflow:    cfg.BlockOnOverflow,
@@ -536,14 +547,14 @@ func (tsp *tailSamplingSpanProcessor) samplingPolicyOnTick() bool {
 			metrics.idNotFoundOnMapCount++
 			continue
 		}
-		trace.DecisionTime = time.Now()
+		trace.decisionTime = time.Now()
 
-		decision := tsp.makeDecision(id, trace, metrics)
+		decision := tsp.makeDecision(id, &trace.TraceData, metrics)
 		globalTracesSampledByDecision[decision]++
 
 		// Sampled or not, remove the batches
 		allSpans := trace.ReceivedBatches
-		trace.FinalDecision = decision
+		trace.finalDecision = decision
 		trace.ReceivedBatches = ptrace.NewTraces()
 
 		if decision == samplingpolicy.Sampled {
@@ -691,10 +702,12 @@ func (tsp *tailSamplingSpanProcessor) processTrace(id pcommon.TraceID, rss ptrac
 
 	actualData, ok := tsp.idToTrace[id]
 	if !ok {
-		actualData = &samplingpolicy.TraceData{
-			ArrivalTime:     currTime,
-			SpanCount:       spanCount,
-			ReceivedBatches: ptrace.NewTraces(),
+		actualData = &traceData{
+			arrivalTime: currTime,
+			TraceData: samplingpolicy.TraceData{
+				SpanCount:       spanCount,
+				ReceivedBatches: ptrace.NewTraces(),
+			},
 		}
 
 		tsp.idToTrace[id] = actualData
@@ -709,7 +722,7 @@ func (tsp *tailSamplingSpanProcessor) processTrace(id pcommon.TraceID, rss ptrac
 		actualData.SpanCount += spanCount
 	}
 
-	finalDecision := actualData.FinalDecision
+	finalDecision := actualData.finalDecision
 
 	if finalDecision == samplingpolicy.Unspecified {
 		// If the final decision hasn't been made, add the new spans to the
@@ -729,8 +742,8 @@ func (tsp *tailSamplingSpanProcessor) processTrace(id pcommon.TraceID, rss ptrac
 		tsp.logger.Warn("Unexpected sampling decision", zap.Int("decision", int(finalDecision)))
 	}
 
-	if !actualData.DecisionTime.IsZero() {
-		tsp.telemetry.ProcessorTailSamplingSamplingLateSpanAge.Record(tsp.ctx, int64(time.Since(actualData.DecisionTime)/time.Second))
+	if !actualData.decisionTime.IsZero() {
+		tsp.telemetry.ProcessorTailSamplingSamplingLateSpanAge.Record(tsp.ctx, int64(time.Since(actualData.decisionTime)/time.Second))
 	}
 }
 
@@ -767,7 +780,7 @@ func (tsp *tailSamplingSpanProcessor) dropTrace(traceID pcommon.TraceID, deletio
 	}
 
 	delete(tsp.idToTrace, traceID)
-	tsp.telemetry.ProcessorTailSamplingSamplingTraceRemovalAge.Record(tsp.ctx, int64(deletionTime.Sub(trace.ArrivalTime)/time.Second))
+	tsp.telemetry.ProcessorTailSamplingSamplingTraceRemovalAge.Record(tsp.ctx, int64(deletionTime.Sub(trace.arrivalTime)/time.Second))
 }
 
 // forwardSpans sends the trace data to the next consumer. it is different from

processor/tailsamplingprocessor/processor_benchmarks_test.go

@@ -39,7 +39,6 @@ func BenchmarkSampling(b *testing.B) {
 
 	for _, batch := range batches {
 		sampleBatches = append(sampleBatches, &samplingpolicy.TraceData{
-			ArrivalTime:     time.Now(),
 			SpanCount:       0,
 			ReceivedBatches: batch,
 		})