Reuse keyvalues

CHANGELOG.md

@@ -19,6 +19,7 @@
 * [ENHANCEMENT] Switch open-telemetry/opentelemetry-collector to grafana/opentelemetry-collectorl fork, update it to 0.40.0 and add missing dependencies due to the change [#1142](https://github.com/grafana/tempo/pull/1142) (@tete17)
 * [ENHANCEMENT] Allow environment variables for Azure storage credentials [#1147](https://github.com/grafana/tempo/pull/1147) (@zalegrala)
 * [ENHANCEMENT] jsonnet: set rollingUpdate.maxSurge to 3 for distributor, frontend and queriers [#1164](https://github.com/grafana/tempo/pull/1164) (@kvrhdn)
+* [ENHANCEMENT] Reduce search data file sizes by optimizing contents [#1165](https://github.com/grafana/tempo/pull/1165) (@mdisibio)
 * [BUGFIX] Add process name to vulture traces to work around display issues [#1127](https://github.com/grafana/tempo/pull/1127) (@mdisibio)
 * [BUGFIX] Fixed issue where compaction sometimes dropped spans. [#1130](https://github.com/grafana/tempo/pull/1130) (@joe-elliott)
 * [BUGFIX] Ensure that the admin client jsonnet has correct S3 bucket property. (@hedss)

pkg/tempofb/SearchBlockHeader_util.go

@@ -63,7 +63,7 @@ func (s *SearchBlockHeaderMutable) Contains(k []byte, v []byte, _ *KeyValues) bo
 func (s *SearchBlockHeaderMutable) ToBytes() []byte {
 	b := flatbuffers.NewBuilder(1024)
 
-	tags := s.Tags.WriteToBuilder(b)
+	tags := WriteSearchDataMap(b, s.Tags, nil)
 
 	SearchBlockHeaderStart(b)
 	SearchBlockHeaderAddMinDurationNanos(b, s.MinDur)

pkg/tempofb/search_entry_mutable.go

@@ -0,0 +1,60 @@
+package tempofb
+
+import (
+	flatbuffers "github.com/google/flatbuffers/go"
+	"github.com/grafana/tempo/tempodb/encoding/common"
+)
+
+// SearchEntryMutable is a mutable form of the flatbuffer-compiled SearchEntry struct to make building and transporting easier.
+type SearchEntryMutable struct {
+	TraceID           common.ID
+	Tags              SearchDataMap
+	StartTimeUnixNano uint64
+	EndTimeUnixNano   uint64
+}
+
+// AddTag adds the unique tag name and value to the search data. No effect if the pair is already present.
+func (s *SearchEntryMutable) AddTag(k string, v string) {
+	if s.Tags == nil {
+		s.Tags = NewSearchDataMap()
+	}
+	s.Tags.Add(k, v)
+}
+
+// SetStartTimeUnixNano records the earliest of all timestamps passed to this function.
+func (s *SearchEntryMutable) SetStartTimeUnixNano(t uint64) {
+	if t > 0 && (s.StartTimeUnixNano == 0 || s.StartTimeUnixNano > t) {
+		s.StartTimeUnixNano = t
+	}
+}
+
+// SetEndTimeUnixNano records the latest of all timestamps passed to this function.
+func (s *SearchEntryMutable) SetEndTimeUnixNano(t uint64) {
+	if t > 0 && t > s.EndTimeUnixNano {
+		s.EndTimeUnixNano = t
+	}
+}
+
+func (s *SearchEntryMutable) ToBytes() []byte {
+	b := flatbuffers.NewBuilder(2048)
+	offset := s.WriteToBuilder(b, nil)
+	b.Finish(offset)
+	return b.FinishedBytes()
+}
+
+func (s *SearchEntryMutable) WriteToBuilder(b *flatbuffers.Builder, kvCache map[uint64]flatbuffers.UOffsetT) flatbuffers.UOffsetT {
+	if s.Tags == nil {
+		s.Tags = NewSearchDataMap()
+	}
+
+	idOffset := b.CreateByteString(s.TraceID)
+
+	tagOffset := WriteSearchDataMap(b, s.Tags, kvCache)
+
+	SearchEntryStart(b)
+	SearchEntryAddId(b, idOffset)
+	SearchEntryAddStartTimeUnixNano(b, s.StartTimeUnixNano)
+	SearchEntryAddEndTimeUnixNano(b, s.EndTimeUnixNano)
+	SearchEntryAddTags(b, tagOffset)
+	return SearchEntryEnd(b)
+}

pkg/tempofb/search_page_builder.go

@@ -0,0 +1,66 @@
+package tempofb
+
+import flatbuffers "github.com/google/flatbuffers/go"
+
+type SearchPageBuilder struct {
+	builder     *flatbuffers.Builder
+	allTags     SearchDataMap
+	pageEntries []flatbuffers.UOffsetT
+	kvcache     map[uint64]flatbuffers.UOffsetT
+}
+
+func NewSearchPageBuilder() *SearchPageBuilder {
+	return &SearchPageBuilder{
+		builder: flatbuffers.NewBuilder(1024),
+		allTags: NewSearchDataMap(),
+		kvcache: map[uint64]flatbuffers.UOffsetT{},
+	}
+}
+
+func (b *SearchPageBuilder) AddData(data *SearchEntryMutable) int {
+	if data.Tags != nil {
+		data.Tags.Range(func(k, v string) {
+			b.allTags.Add(k, v)
+		})
+	}
+
+	oldOffset := b.builder.Offset()
+	offset := data.WriteToBuilder(b.builder, b.kvcache)
+	b.pageEntries = append(b.pageEntries, offset)
+
+	// bytes written
+	return int(offset - oldOffset)
+}
+
+func (b *SearchPageBuilder) Finish() []byte {
+	// At this point all individual entries have been written
+	// to the fb builder. Now we need to wrap them up in the final
+	// batch object.
+
+	// Create vector
+	SearchPageStartEntriesVector(b.builder, len(b.pageEntries))
+	for _, entry := range b.pageEntries {
+		b.builder.PrependUOffsetT(entry)
+	}
+	entryVector := b.builder.EndVector(len(b.pageEntries))
+
+	// Create batch-level tags
+	tagOffset := WriteSearchDataMap(b.builder, b.allTags, b.kvcache)
+
+	// Write final batch object
+	SearchPageStart(b.builder)
+	SearchPageAddEntries(b.builder, entryVector)
+	SearchPageAddTags(b.builder, tagOffset)
+	batch := SearchPageEnd(b.builder)
+	b.builder.Finish(batch)
+	buf := b.builder.FinishedBytes()
+
+	return buf
+}
+
+func (b *SearchPageBuilder) Reset() {
+	b.builder.Reset()
+	b.pageEntries = b.pageEntries[:0]
+	b.allTags = NewSearchDataMap()
+	b.kvcache = map[uint64]flatbuffers.UOffsetT{}
+}

pkg/tempofb/searchdata_util.go

@@ -4,123 +4,8 @@ import (
 	"bytes"
 
 	flatbuffers "github.com/google/flatbuffers/go"
-	"github.com/grafana/tempo/tempodb/encoding/common"
 )
 
-// SearchEntryMutable is a mutable form of the flatbuffer-compiled SearchEntry struct to make building and transporting easier.
-type SearchEntryMutable struct {
-	TraceID           common.ID
-	Tags              SearchDataMap
-	StartTimeUnixNano uint64
-	EndTimeUnixNano   uint64
-}
-
-// AddTag adds the unique tag name and value to the search data. No effect if the pair is already present.
-func (s *SearchEntryMutable) AddTag(k string, v string) {
-	if s.Tags == nil {
-		s.Tags = NewSearchDataMap()
-	}
-	s.Tags.Add(k, v)
-}
-
-// SetStartTimeUnixNano records the earliest of all timestamps passed to this function.
-func (s *SearchEntryMutable) SetStartTimeUnixNano(t uint64) {
-	if t > 0 && (s.StartTimeUnixNano == 0 || s.StartTimeUnixNano > t) {
-		s.StartTimeUnixNano = t
-	}
-}
-
-// SetEndTimeUnixNano records the latest of all timestamps passed to this function.
-func (s *SearchEntryMutable) SetEndTimeUnixNano(t uint64) {
-	if t > 0 && t > s.EndTimeUnixNano {
-		s.EndTimeUnixNano = t
-	}
-}
-
-func (s *SearchEntryMutable) ToBytes() []byte {
-	b := flatbuffers.NewBuilder(2048)
-	offset := s.WriteToBuilder(b)
-	b.Finish(offset)
-	return b.FinishedBytes()
-}
-
-func (s *SearchEntryMutable) WriteToBuilder(b *flatbuffers.Builder) flatbuffers.UOffsetT {
-	if s.Tags == nil {
-		s.Tags = NewSearchDataMap()
-	}
-
-	idOffset := b.CreateByteString(s.TraceID)
-
-	tagOffset := s.Tags.WriteToBuilder(b)
-
-	SearchEntryStart(b)
-	SearchEntryAddId(b, idOffset)
-	SearchEntryAddStartTimeUnixNano(b, s.StartTimeUnixNano)
-	SearchEntryAddEndTimeUnixNano(b, s.EndTimeUnixNano)
-	SearchEntryAddTags(b, tagOffset)
-	return SearchEntryEnd(b)
-}
-
-type SearchPageBuilder struct {
-	builder     *flatbuffers.Builder
-	allTags     SearchDataMap
-	pageEntries []flatbuffers.UOffsetT
-}
-
-func NewSearchPageBuilder() *SearchPageBuilder {
-	return &SearchPageBuilder{
-		builder: flatbuffers.NewBuilder(1024),
-		allTags: NewSearchDataMap(),
-	}
-}
-
-func (b *SearchPageBuilder) AddData(data *SearchEntryMutable) int {
-	if data.Tags != nil {
-		data.Tags.Range(func(k, v string) {
-			b.allTags.Add(k, v)
-		})
-	}
-
-	oldOffset := b.builder.Offset()
-	offset := data.WriteToBuilder(b.builder)
-	b.pageEntries = append(b.pageEntries, offset)
-
-	// bytes written
-	return int(offset - oldOffset)
-}
-
-func (b *SearchPageBuilder) Finish() []byte {
-	// At this point all individual entries have been written
-	// to the fb builder. Now we need to wrap them up in the final
-	// batch object.
-
-	// Create vector
-	SearchPageStartEntriesVector(b.builder, len(b.pageEntries))
-	for _, entry := range b.pageEntries {
-		b.builder.PrependUOffsetT(entry)
-	}
-	entryVector := b.builder.EndVector(len(b.pageEntries))
-
-	// Create batch-level tags
-	tagOffset := b.allTags.WriteToBuilder(b.builder)
-
-	// Write final batch object
-	SearchPageStart(b.builder)
-	SearchPageAddEntries(b.builder, entryVector)
-	SearchPageAddTags(b.builder, tagOffset)
-	batch := SearchPageEnd(b.builder)
-	b.builder.Finish(batch)
-	buf := b.builder.FinishedBytes()
-
-	return buf
-}
-
-func (b *SearchPageBuilder) Reset() {
-	b.builder.Reset()
-	b.pageEntries = b.pageEntries[:0]
-	b.allTags = NewSearchDataMap()
-}
-
 // Get searches the entry and returns the first value found for the given key.
 func (s *SearchEntry) Get(k string) string {
 	kv := &KeyValues{}