fix(blooms): Exclude label filters where label name is part of the series labels. (#14661)

2026-03-13 09:33:58 +08:00 · 2024-10-31 15:51:15 +01:00
parent ebbbccfdaf
commit d1668f6a11
16 changed files with 278 additions and 112 deletions
--- a/pkg/bloomgateway/bloomgateway_test.go
+++ b/pkg/bloomgateway/bloomgateway_test.go
@@ -15,6 +15,7 @@ import (
 	"github.com/pkg/errors"
 	"github.com/prometheus/client_golang/prometheus"
 	"github.com/prometheus/common/model"
+	"github.com/prometheus/prometheus/model/labels"
 	"github.com/stretchr/testify/require"

 	"github.com/grafana/loki/v3/pkg/logproto"
@@ -41,7 +42,15 @@ func stringSlice[T fmt.Stringer](s []T) []string {

 func groupRefs(t *testing.T, chunkRefs []*logproto.ChunkRef) []*logproto.GroupedChunkRefs {
 	t.Helper()
-	return groupChunkRefs(chunkRefs, nil)
+	grouped := groupChunkRefs(nil, chunkRefs, nil)
+	// Put fake labels to the series
+	for _, g := range grouped {
+		g.Labels = &logproto.IndexSeries{
+			Labels: logproto.FromLabelsToLabelAdapters(labels.FromStrings("foo", fmt.Sprintf("%d", g.Fingerprint))),
+		}
+	}
+
+	return grouped
 }

 func newLimits() *validation.Overrides {
@@ -295,12 +304,18 @@ func TestBloomGateway_FilterChunkRefs(t *testing.T) {
 				{Fingerprint: 1000, Tenant: tenantID, Refs: []*logproto.ShortRef{
 					{From: 1696248000000, Through: 1696251600000, Checksum: 2},
 					{From: 1696244400000, Through: 1696248000000, Checksum: 4},
+				}, Labels: &logproto.IndexSeries{
+					Labels: logproto.FromLabelsToLabelAdapters(labels.FromStrings("foo", "1000")),
 				}},
 				{Fingerprint: 2000, Tenant: tenantID, Refs: []*logproto.ShortRef{
 					{From: 1696255200000, Through: 1696258800000, Checksum: 3},
+				}, Labels: &logproto.IndexSeries{
+					Labels: logproto.FromLabelsToLabelAdapters(labels.FromStrings("foo", "2000")),
 				}},
 				{Fingerprint: 3000, Tenant: tenantID, Refs: []*logproto.ShortRef{
 					{From: 1696240800000, Through: 1696244400000, Checksum: 1},
+				}, Labels: &logproto.IndexSeries{
+					Labels: logproto.FromLabelsToLabelAdapters(labels.FromStrings("foo", "3000")),
 				}},
 			},
 		}, res)
@@ -405,6 +420,7 @@ func TestBloomGateway_FilterChunkRefs(t *testing.T) {
 			// see MkBasicSeriesWithBlooms() in pkg/storage/bloom/v1/test_util.go
 			rnd := rand.Intn(len(inputChunkRefs))
 			fp := inputChunkRefs[rnd].Fingerprint
+			lbs := inputChunkRefs[rnd].Labels
 			chks := inputChunkRefs[rnd].Refs
 			key := fmt.Sprintf("%s:%04x", model.Fingerprint(fp), 0)

@@ -428,6 +444,7 @@ func TestBloomGateway_FilterChunkRefs(t *testing.T) {
 				ChunkRefs: []*logproto.GroupedChunkRefs{
 					{
 						Fingerprint: fp,
+						Labels:      lbs,
 						Refs:        chks,
 						Tenant:      tenantID,
 					},
--- a/pkg/bloomgateway/cache.go
+++ b/pkg/bloomgateway/cache.go
@@ -72,6 +72,7 @@ func (e extractor) Extract(start, end int64, r resultscache.Response, _, _ int64
 		if len(refs) > 0 {
 			chunkRefs = append(chunkRefs, &logproto.GroupedChunkRefs{
 				Fingerprint: chunkRef.Fingerprint,
+				Labels:      chunkRef.Labels,
 				Tenant:      chunkRef.Tenant,
 				Refs:        refs,
 			})
--- a/pkg/bloomgateway/client.go
+++ b/pkg/bloomgateway/client.go
@@ -322,6 +322,7 @@ func mergeSeries(input [][]*logproto.GroupedChunkRefs, buf []*logproto.GroupedCh
 			}
 			return &logproto.GroupedChunkRefs{
 				Fingerprint: a.Fingerprint,
+				Labels:      a.Labels,
 				Tenant:      a.Tenant,
 				Refs:        mergeChunkSets(a.Refs, b.Refs),
 			}
--- a/pkg/bloomgateway/multiplexing.go
+++ b/pkg/bloomgateway/multiplexing.go
@@ -174,6 +174,7 @@ func (it *requestIterator) Next() bool {
 	it.curr = v1.Request{
 		Recorder: it.recorder,
 		Fp:       model.Fingerprint(group.Fingerprint),
+		Labels:   logproto.FromLabelAdaptersToLabels(group.Labels.Labels),
 		Chks:     convertToChunkRefs(group.Refs),
 		Search:   it.search,
 		Response: it.channel,
--- a/pkg/bloomgateway/multiplexing_test.go
+++ b/pkg/bloomgateway/multiplexing_test.go
@@ -7,6 +7,7 @@ import (
 	"time"

 	"github.com/prometheus/common/model"
+	"github.com/prometheus/prometheus/model/labels"
 	"github.com/stretchr/testify/require"

 	v2 "github.com/grafana/loki/v3/pkg/iter/v2"
@@ -73,6 +74,8 @@ func TestTask_RequestIterator(t *testing.T) {
 			Refs: []*logproto.GroupedChunkRefs{
 				{Fingerprint: 100, Tenant: tenant, Refs: []*logproto.ShortRef{
 					{From: ts.Add(-3 * time.Hour), Through: ts.Add(-2 * time.Hour), Checksum: 100},
+				}, Labels: &logproto.IndexSeries{
+					Labels: logproto.FromLabelsToLabelAdapters(labels.FromStrings("foo", "100")),
 				}},
 			},
 		}
@@ -83,9 +86,13 @@ func TestTask_RequestIterator(t *testing.T) {
 			Refs: []*logproto.GroupedChunkRefs{
 				{Fingerprint: 100, Tenant: tenant, Refs: []*logproto.ShortRef{
 					{From: ts.Add(-1 * time.Hour), Through: ts, Checksum: 200},
+				}, Labels: &logproto.IndexSeries{
+					Labels: logproto.FromLabelsToLabelAdapters(labels.FromStrings("foo", "100")),
 				}},
 				{Fingerprint: 200, Tenant: tenant, Refs: []*logproto.ShortRef{
 					{From: ts.Add(-1 * time.Hour), Through: ts, Checksum: 300},
+				}, Labels: &logproto.IndexSeries{
+					Labels: logproto.FromLabelsToLabelAdapters(labels.FromStrings("foo", "200")),
 				}},
 			},
 		}
@@ -96,6 +103,8 @@ func TestTask_RequestIterator(t *testing.T) {
 			Refs: []*logproto.GroupedChunkRefs{
 				{Fingerprint: 200, Tenant: tenant, Refs: []*logproto.ShortRef{
 					{From: ts.Add(-1 * time.Hour), Through: ts, Checksum: 400},
+				}, Labels: &logproto.IndexSeries{
+					Labels: logproto.FromLabelsToLabelAdapters(labels.FromStrings("foo", "200")),
 				}},
 			},
 		}
--- a/pkg/bloomgateway/querier.go
+++ b/pkg/bloomgateway/querier.go
@@ -10,6 +10,7 @@ import (
 	"github.com/prometheus/client_golang/prometheus"
 	"github.com/prometheus/client_golang/prometheus/promauto"
 	"github.com/prometheus/common/model"
+	"github.com/prometheus/prometheus/model/labels"

 	"github.com/grafana/loki/v3/pkg/logproto"
 	"github.com/grafana/loki/v3/pkg/querier/plan"
@@ -101,7 +102,7 @@ func convertToShortRef(ref *logproto.ChunkRef) *logproto.ShortRef {
 	return &logproto.ShortRef{From: ref.From, Through: ref.Through, Checksum: ref.Checksum}
 }

-func (bq *BloomQuerier) FilterChunkRefs(ctx context.Context, tenant string, from, through model.Time, chunkRefs []*logproto.ChunkRef, queryPlan plan.QueryPlan) ([]*logproto.ChunkRef, error) {
+func (bq *BloomQuerier) FilterChunkRefs(ctx context.Context, tenant string, from, through model.Time, series map[uint64]labels.Labels, chunkRefs []*logproto.ChunkRef, queryPlan plan.QueryPlan) ([]*logproto.ChunkRef, error) {
 	// Shortcut that does not require any filtering
 	if !bq.limits.BloomGatewayEnabled(tenant) || len(chunkRefs) == 0 || len(v1.ExtractTestableLabelMatchers(queryPlan.AST)) == 0 {
 		return chunkRefs, nil
@@ -112,7 +113,7 @@ func (bq *BloomQuerier) FilterChunkRefs(ctx context.Context, tenant string, from

 	grouped := groupedChunksRefPool.Get(len(chunkRefs))
 	defer groupedChunksRefPool.Put(grouped)
-	grouped = groupChunkRefs(chunkRefs, grouped)
+	grouped = groupChunkRefs(series, chunkRefs, grouped)

 	preFilterChunks := len(chunkRefs)
 	preFilterSeries := len(grouped)
@@ -225,7 +226,7 @@ func (bq *BloomQuerier) FilterChunkRefs(ctx context.Context, tenant string, from
 // groups them by fingerprint.
 // The second argument `grouped` can be used to pass a buffer to avoid allocations.
 // If it's nil, the returned slice will be allocated.
-func groupChunkRefs(chunkRefs []*logproto.ChunkRef, grouped []*logproto.GroupedChunkRefs) []*logproto.GroupedChunkRefs {
+func groupChunkRefs(series map[uint64]labels.Labels, chunkRefs []*logproto.ChunkRef, grouped []*logproto.GroupedChunkRefs) []*logproto.GroupedChunkRefs {
 	seen := make(map[uint64]int, len(grouped))
 	for _, chunkRef := range chunkRefs {
 		if idx, found := seen[chunkRef.Fingerprint]; found {
@@ -234,10 +235,14 @@ func groupChunkRefs(chunkRefs []*logproto.ChunkRef, grouped []*logproto.GroupedC
 			seen[chunkRef.Fingerprint] = len(grouped)
 			grouped = append(grouped, &logproto.GroupedChunkRefs{
 				Fingerprint: chunkRef.Fingerprint,
-				Tenant:      chunkRef.UserID,
-				Refs:        []*logproto.ShortRef{convertToShortRef(chunkRef)},
+				Labels: &logproto.IndexSeries{
+					Labels: logproto.FromLabelsToLabelAdapters(series[chunkRef.Fingerprint]),
+				},
+				Tenant: chunkRef.UserID,
+				Refs:   []*logproto.ShortRef{convertToShortRef(chunkRef)},
 			})
 		}
 	}
+
 	return grouped
 }
--- a/pkg/bloomgateway/querier_test.go
+++ b/pkg/bloomgateway/querier_test.go
@@ -2,6 +2,7 @@ package bloomgateway

 import (
 	"context"
+	"fmt"
 	"math/rand"
 	"sort"
 	"testing"
@@ -10,6 +11,7 @@ import (
 	"github.com/go-kit/log"
 	"github.com/pkg/errors"
 	"github.com/prometheus/common/model"
+	"github.com/prometheus/prometheus/model/labels"
 	"github.com/stretchr/testify/require"

 	"github.com/grafana/loki/v3/pkg/logproto"
@@ -79,7 +81,7 @@ func TestBloomQuerier(t *testing.T) {
 		}
 		expr, err := syntax.ParseExpr(`{foo="bar"}`)
 		require.NoError(t, err)
-		res, err := bq.FilterChunkRefs(ctx, tenant, from, through, chunkRefs, plan.QueryPlan{AST: expr})
+		res, err := bq.FilterChunkRefs(ctx, tenant, from, through, nil, chunkRefs, plan.QueryPlan{AST: expr})
 		require.NoError(t, err)
 		require.Equal(t, chunkRefs, res)
 		require.Equal(t, 0, c.callCount)
@@ -95,7 +97,7 @@ func TestBloomQuerier(t *testing.T) {
 		chunkRefs := []*logproto.ChunkRef{}
 		expr, err := syntax.ParseExpr(`{foo="bar"} | trace_id="exists"`)
 		require.NoError(t, err)
-		res, err := bq.FilterChunkRefs(ctx, tenant, from, through, chunkRefs, plan.QueryPlan{AST: expr})
+		res, err := bq.FilterChunkRefs(ctx, tenant, from, through, nil, chunkRefs, plan.QueryPlan{AST: expr})
 		require.NoError(t, err)
 		require.Equal(t, chunkRefs, res)
 		require.Equal(t, 0, c.callCount)
@@ -115,7 +117,7 @@ func TestBloomQuerier(t *testing.T) {
 		}
 		expr, err := syntax.ParseExpr(`{foo="bar"} | trace_id="exists"`)
 		require.NoError(t, err)
-		res, err := bq.FilterChunkRefs(ctx, tenant, from, through, chunkRefs, plan.QueryPlan{AST: expr})
+		res, err := bq.FilterChunkRefs(ctx, tenant, from, through, nil, chunkRefs, plan.QueryPlan{AST: expr})
 		require.Error(t, err)
 		require.Nil(t, res)
 	})
@@ -134,7 +136,7 @@ func TestBloomQuerier(t *testing.T) {
 		}
 		expr, err := syntax.ParseExpr(`{foo="bar"} | trace_id="exists"`)
 		require.NoError(t, err)
-		res, err := bq.FilterChunkRefs(ctx, tenant, from, through, chunkRefs, plan.QueryPlan{AST: expr})
+		res, err := bq.FilterChunkRefs(ctx, tenant, from, through, nil, chunkRefs, plan.QueryPlan{AST: expr})
 		require.NoError(t, err)
 		require.Equal(t, chunkRefs, res)
 		require.Equal(t, 2, c.callCount)
@@ -143,28 +145,44 @@ func TestBloomQuerier(t *testing.T) {
 }

 func TestGroupChunkRefs(t *testing.T) {
-	chunkRefs := []*logproto.ChunkRef{
-		{Fingerprint: 0x00, UserID: "tenant", From: mktime("2024-04-20 00:00"), Through: mktime("2024-04-20 00:59")},
-		{Fingerprint: 0x00, UserID: "tenant", From: mktime("2024-04-20 01:00"), Through: mktime("2024-04-20 01:59")},
-		{Fingerprint: 0x01, UserID: "tenant", From: mktime("2024-04-20 00:00"), Through: mktime("2024-04-20 00:59")},
-		{Fingerprint: 0x01, UserID: "tenant", From: mktime("2024-04-20 01:00"), Through: mktime("2024-04-20 01:59")},
-		{Fingerprint: 0x02, UserID: "tenant", From: mktime("2024-04-20 00:00"), Through: mktime("2024-04-20 00:59")},
-		{Fingerprint: 0x02, UserID: "tenant", From: mktime("2024-04-20 01:00"), Through: mktime("2024-04-20 01:59")},
+	series := []labels.Labels{
+		labels.FromStrings("app", "1"),
+		labels.FromStrings("app", "2"),
+		labels.FromStrings("app", "3"),
+	}
+	seriesMap := make(map[uint64]labels.Labels)
+	for _, s := range series {
+		seriesMap[s.Hash()] = s
 	}

-	result := groupChunkRefs(chunkRefs, nil)
+	chunkRefs := []*logproto.ChunkRef{
+		{Fingerprint: series[0].Hash(), UserID: "tenant", From: mktime("2024-04-20 00:00"), Through: mktime("2024-04-20 00:59")},
+		{Fingerprint: series[0].Hash(), UserID: "tenant", From: mktime("2024-04-20 01:00"), Through: mktime("2024-04-20 01:59")},
+		{Fingerprint: series[1].Hash(), UserID: "tenant", From: mktime("2024-04-20 00:00"), Through: mktime("2024-04-20 00:59")},
+		{Fingerprint: series[1].Hash(), UserID: "tenant", From: mktime("2024-04-20 01:00"), Through: mktime("2024-04-20 01:59")},
+		{Fingerprint: series[2].Hash(), UserID: "tenant", From: mktime("2024-04-20 00:00"), Through: mktime("2024-04-20 00:59")},
+		{Fingerprint: series[2].Hash(), UserID: "tenant", From: mktime("2024-04-20 01:00"), Through: mktime("2024-04-20 01:59")},
+	}
+
+	result := groupChunkRefs(seriesMap, chunkRefs, nil)
 	require.Equal(t, []*logproto.GroupedChunkRefs{
-		{Fingerprint: 0x00, Tenant: "tenant", Refs: []*logproto.ShortRef{
+		{Fingerprint: series[0].Hash(), Tenant: "tenant", Refs: []*logproto.ShortRef{
 			{From: mktime("2024-04-20 00:00"), Through: mktime("2024-04-20 00:59")},
 			{From: mktime("2024-04-20 01:00"), Through: mktime("2024-04-20 01:59")},
+		}, Labels: &logproto.IndexSeries{
+			Labels: logproto.FromLabelsToLabelAdapters(series[0]),
 		}},
-		{Fingerprint: 0x01, Tenant: "tenant", Refs: []*logproto.ShortRef{
+		{Fingerprint: series[1].Hash(), Tenant: "tenant", Refs: []*logproto.ShortRef{
 			{From: mktime("2024-04-20 00:00"), Through: mktime("2024-04-20 00:59")},
 			{From: mktime("2024-04-20 01:00"), Through: mktime("2024-04-20 01:59")},
+		}, Labels: &logproto.IndexSeries{
+			Labels: logproto.FromLabelsToLabelAdapters(series[1]),
 		}},
-		{Fingerprint: 0x02, Tenant: "tenant", Refs: []*logproto.ShortRef{
+		{Fingerprint: series[2].Hash(), Tenant: "tenant", Refs: []*logproto.ShortRef{
 			{From: mktime("2024-04-20 00:00"), Through: mktime("2024-04-20 00:59")},
 			{From: mktime("2024-04-20 01:00"), Through: mktime("2024-04-20 01:59")},
+		}, Labels: &logproto.IndexSeries{
+			Labels: logproto.FromLabelsToLabelAdapters(series[2]),
 		}},
 	}, result)
 }
@@ -175,11 +193,15 @@ func BenchmarkGroupChunkRefs(b *testing.B) {
 	n := 1000  // num series
 	m := 10000 // num chunks per series
 	chunkRefs := make([]*logproto.ChunkRef, 0, n*m)
+	series := make(map[uint64]labels.Labels, n)

 	for i := 0; i < n; i++ {
+		s := labels.FromStrings("app", fmt.Sprintf("%d", i))
+		sFP := s.Hash()
+		series[sFP] = s
 		for j := 0; j < m; j++ {
 			chunkRefs = append(chunkRefs, &logproto.ChunkRef{
-				Fingerprint: uint64(n),
+				Fingerprint: sFP,
 				UserID:      "tenant",
 				From:        mktime("2024-04-20 00:00"),
 				Through:     mktime("2024-04-20 00:59"),
@@ -196,5 +218,5 @@ func BenchmarkGroupChunkRefs(b *testing.B) {
 	b.StartTimer()

 	groups := make([]*logproto.GroupedChunkRefs, 0, n)
-	groupChunkRefs(chunkRefs, groups)
+	groupChunkRefs(series, chunkRefs, groups)
 }
--- a/pkg/bloomgateway/util.go
+++ b/pkg/bloomgateway/util.go
@@ -110,6 +110,7 @@ func partitionSeriesByDay(from, through model.Time, seriesWithChunks []*logproto

 			res = append(res, &logproto.GroupedChunkRefs{
 				Fingerprint: series.Fingerprint,
+				Labels:      series.Labels,
 				Tenant:      series.Tenant,
 				Refs:        relevantChunks,
 			})
--- a/pkg/indexgateway/gateway.go
+++ b/pkg/indexgateway/gateway.go
@@ -58,7 +58,7 @@ type IndexClientWithRange struct {
 }

 type BloomQuerier interface {
-	FilterChunkRefs(ctx context.Context, tenant string, from, through model.Time, chunks []*logproto.ChunkRef, plan plan.QueryPlan) ([]*logproto.ChunkRef, error)
+	FilterChunkRefs(ctx context.Context, tenant string, from, through model.Time, series map[uint64]labels.Labels, chunks []*logproto.ChunkRef, plan plan.QueryPlan) ([]*logproto.ChunkRef, error)
 }

 type Gateway struct {
@@ -225,12 +225,16 @@ func (g *Gateway) GetChunkRef(ctx context.Context, req *logproto.GetChunkRefRequ
 		return nil, err
 	}

+	series := make(map[uint64]labels.Labels)
 	result = &logproto.GetChunkRefResponse{
 		Refs: make([]*logproto.ChunkRef, 0, len(chunks)),
 	}
 	for _, cs := range chunks {
 		for i := range cs {
 			result.Refs = append(result.Refs, &cs[i].ChunkRef)
+			if _, ok := series[cs[i].Fingerprint]; !ok {
+				series[cs[i].Fingerprint] = cs[i].Metric
+			}
 		}
 	}

@@ -257,7 +261,7 @@ func (g *Gateway) GetChunkRef(ctx context.Context, req *logproto.GetChunkRefRequ
 		return result, nil
 	}

-	chunkRefs, err := g.bloomQuerier.FilterChunkRefs(ctx, instanceID, req.From, req.Through, result.Refs, req.Plan)
+	chunkRefs, err := g.bloomQuerier.FilterChunkRefs(ctx, instanceID, req.From, req.Through, series, result.Refs, req.Plan)
 	if err != nil {
 		return nil, err
 	}
--- a/pkg/logproto/bloomgateway.pb.go
+++ b/pkg/logproto/bloomgateway.pb.go
@@ -178,6 +178,9 @@ type GroupedChunkRefs struct {
 	Fingerprint uint64      `protobuf:"varint,1,opt,name=fingerprint,proto3" json:"fingerprint,omitempty"`
 	Tenant      string      `protobuf:"bytes,2,opt,name=tenant,proto3" json:"tenant,omitempty"`
 	Refs        []*ShortRef `protobuf:"bytes,3,rep,name=refs,proto3" json:"refs,omitempty"`
+	// Labels are only populated on FilterChunkRefRequest. They are not returned on FilterChunkRefResponse
+	// TODO(salvacorts): Consider two different messages for FilterChunkRefRequest and FilterChunkRefResponse
+	Labels *IndexSeries `protobuf:"bytes,4,opt,name=labels,proto3" json:"labels,omitempty"`
 }

 func (m *GroupedChunkRefs) Reset()      { *m = GroupedChunkRefs{} }
@@ -233,6 +236,13 @@ func (m *GroupedChunkRefs) GetRefs() []*ShortRef {
 	return nil
 }

+func (m *GroupedChunkRefs) GetLabels() *IndexSeries {
+	if m != nil {
+		return m.Labels
+	}
+	return nil
+}
+
 func init() {
 	proto.RegisterType((*FilterChunkRefRequest)(nil), "logproto.FilterChunkRefRequest")
 	proto.RegisterType((*FilterChunkRefResponse)(nil), "logproto.FilterChunkRefResponse")
@@ -243,39 +253,41 @@ func init() {
 func init() { proto.RegisterFile("pkg/logproto/bloomgateway.proto", fileDescriptor_a50b5dd1dbcd1415) }

 var fileDescriptor_a50b5dd1dbcd1415 = []byte{
-	// 504 bytes of a gzipped FileDescriptorProto
+	// 532 bytes of a gzipped FileDescriptorProto
 	0x1f, 0x8b, 0x08, 0x00, 0x00, 0x00, 0x00, 0x00, 0x02, 0xff, 0xcc, 0x53, 0xcd, 0x6e, 0xd3, 0x40,
-	0x10, 0xf6, 0x36, 0x21, 0x24, 0x1b, 0xfe, 0xb4, 0x82, 0xca, 0x0a, 0xd2, 0xc6, 0xca, 0x81, 0xfa,
-	0xe4, 0x95, 0x52, 0x21, 0x71, 0xe1, 0x92, 0x4a, 0x54, 0xdc, 0x60, 0x41, 0x1c, 0x90, 0x38, 0x38,
-	0xee, 0xfa, 0x47, 0xb6, 0x77, 0xdc, 0xf5, 0x1a, 0xd4, 0x1b, 0x8f, 0xc0, 0x63, 0xf0, 0x04, 0x3c,
-	0x43, 0x8f, 0x11, 0xa7, 0x8a, 0x43, 0x45, 0x9c, 0x0b, 0xc7, 0x3e, 0x02, 0xb2, 0x1d, 0x37, 0x4d,
-	0x05, 0xaa, 0xc4, 0x89, 0x93, 0x77, 0x76, 0xbe, 0x19, 0x7f, 0xf3, 0x7d, 0x3b, 0x78, 0x9c, 0xc5,
-	0x01, 0x4b, 0x20, 0xc8, 0x14, 0x68, 0x60, 0xf3, 0x04, 0x20, 0x0d, 0x5c, 0x2d, 0x3e, 0xb9, 0x27,
-	0x4e, 0x7d, 0x45, 0xfa, 0x6d, 0x72, 0xf4, 0x30, 0x80, 0x00, 0x1a, 0x5c, 0x75, 0x6a, 0xf2, 0xa3,
-	0xc7, 0x5b, 0x0d, 0xda, 0x43, 0x93, 0x9c, 0x7c, 0xdf, 0xc1, 0x8f, 0x5e, 0x44, 0x89, 0x16, 0xea,
-	0x20, 0x2c, 0x64, 0xcc, 0x85, 0xcf, 0xc5, 0x71, 0x21, 0x72, 0x4d, 0x0e, 0x70, 0xd7, 0x57, 0x90,
-	0x9a, 0xc8, 0x42, 0x76, 0x67, 0xc6, 0x4e, 0xcf, 0xc7, 0xc6, 0x8f, 0xf3, 0xf1, 0x5e, 0x10, 0xe9,
-	0xb0, 0x98, 0x3b, 0x1e, 0xa4, 0x2c, 0x53, 0x90, 0x0a, 0x1d, 0x8a, 0x22, 0x67, 0x1e, 0xa4, 0x29,
-	0x48, 0x96, 0xc2, 0x91, 0x48, 0x9c, 0xb7, 0x51, 0x2a, 0x78, 0x5d, 0x4c, 0x5e, 0xe2, 0xdb, 0x3a,
-	0x54, 0x50, 0x04, 0xa1, 0xb9, 0xf3, 0x6f, 0x7d, 0xda, 0x7a, 0xe2, 0xe0, 0xae, 0x12, 0x7e, 0x6e,
-	0x76, 0xac, 0x8e, 0x3d, 0x9c, 0x8e, 0x9c, 0xcb, 0x41, 0x0e, 0x15, 0x14, 0x99, 0x38, 0x6a, 0xf9,
-	0xe7, 0xbc, 0xc6, 0x11, 0x17, 0x77, 0xb3, 0xc4, 0x95, 0xe6, 0x2d, 0x0b, 0xd9, 0xc3, 0xe9, 0xbd,
-	0x0d, 0xfe, 0x55, 0xe2, 0xca, 0xd9, 0xf3, 0x35, 0x8f, 0xa7, 0x57, 0x78, 0x04, 0xca, 0xf5, 0x5d,
-	0xe9, 0xb2, 0x04, 0xe2, 0x88, 0x7d, 0xdc, 0x67, 0x95, 0x6e, 0xc7, 0x85, 0x50, 0x91, 0x50, 0xac,
-	0x6a, 0xe5, 0xbc, 0x2e, 0x84, 0x3a, 0xa9, 0xca, 0x79, 0xdd, 0x9a, 0xec, 0xe2, 0xde, 0x3c, 0x01,
-	0x2f, 0xce, 0xcd, 0x9e, 0xd5, 0xb1, 0x07, 0x7c, 0x1d, 0x4d, 0x38, 0xde, 0xbd, 0xae, 0x69, 0x9e,
-	0x81, 0xcc, 0x05, 0x79, 0x86, 0x07, 0x5e, 0xcb, 0xd3, 0x44, 0x37, 0x4e, 0xb2, 0x01, 0x4f, 0xbe,
-	0x21, 0xdc, 0x7f, 0x13, 0x82, 0xd2, 0x5c, 0xf8, 0xff, 0x9d, 0x37, 0x23, 0xdc, 0xf7, 0x42, 0xe1,
-	0xc5, 0x79, 0x91, 0x9a, 0x1d, 0x0b, 0xd9, 0x77, 0xf9, 0x65, 0x3c, 0xd1, 0xf8, 0xc1, 0xf5, 0xb9,
-	0x88, 0x85, 0x87, 0x7e, 0x24, 0x03, 0xa1, 0x32, 0x15, 0x49, 0x5d, 0x8f, 0xd1, 0xe5, 0x57, 0xaf,
-	0x2a, 0x69, 0xb5, 0x90, 0xae, 0xd4, 0x35, 0xb7, 0x01, 0x5f, 0x47, 0xe4, 0xc9, 0xd6, 0x2b, 0x20,
-	0x1b, 0xed, 0x5a, 0x6d, 0x1a, 0xf7, 0xa7, 0x3e, 0xbe, 0x33, 0xab, 0x56, 0xe5, 0xb0, 0x59, 0x15,
-	0xf2, 0x0e, 0xdf, 0xdf, 0xb6, 0x24, 0x27, 0xe3, 0x4d, 0xf1, 0x1f, 0x37, 0x60, 0x64, 0xfd, 0x1d,
-	0xd0, 0xd8, 0x39, 0x31, 0x66, 0x1f, 0x16, 0x4b, 0x6a, 0x9c, 0x2d, 0xa9, 0x71, 0xb1, 0xa4, 0xe8,
-	0x73, 0x49, 0xd1, 0xd7, 0x92, 0xa2, 0xd3, 0x92, 0xa2, 0x45, 0x49, 0xd1, 0xcf, 0x92, 0xa2, 0x5f,
-	0x25, 0x35, 0x2e, 0x4a, 0x8a, 0xbe, 0xac, 0xa8, 0xb1, 0x58, 0x51, 0xe3, 0x6c, 0x45, 0x8d, 0xf7,
-	0x7b, 0x37, 0xbc, 0xba, 0xf6, 0xbf, 0xf3, 0x5e, 0xfd, 0xd9, 0xff, 0x1d, 0x00, 0x00, 0xff, 0xff,
-	0x30, 0x14, 0xc6, 0xc9, 0x05, 0x04, 0x00, 0x00,
+	0x10, 0xf6, 0x36, 0x21, 0x24, 0x1b, 0xfe, 0xb4, 0xa2, 0x95, 0x15, 0xa4, 0x8d, 0x95, 0x03, 0xcd,
+	0x05, 0x5b, 0x4a, 0x85, 0xc4, 0x85, 0x4b, 0x2a, 0x51, 0xf5, 0x06, 0x5b, 0xc4, 0x01, 0x89, 0x83,
+	0xe3, 0x8c, 0x7f, 0x14, 0x7b, 0xd7, 0xdd, 0x5d, 0x03, 0xbd, 0xf1, 0x08, 0xbc, 0x03, 0x17, 0x9e,
+	0x80, 0x67, 0xe8, 0x31, 0xe2, 0x54, 0x71, 0xa8, 0x88, 0x73, 0xe1, 0xd8, 0x47, 0x40, 0x76, 0xe2,
+	0xba, 0xa9, 0x40, 0x95, 0x38, 0x71, 0xf2, 0xee, 0xce, 0x37, 0xe3, 0x6f, 0xbe, 0x6f, 0x06, 0xf7,
+	0xd3, 0x59, 0xe0, 0xc4, 0x22, 0x48, 0xa5, 0xd0, 0xc2, 0x99, 0xc4, 0x42, 0x24, 0x81, 0xab, 0xe1,
+	0x83, 0x7b, 0x62, 0x97, 0x4f, 0xa4, 0x5d, 0x05, 0x7b, 0x0f, 0x03, 0x11, 0x88, 0x15, 0xae, 0x38,
+	0xad, 0xe2, 0xbd, 0x47, 0x1b, 0x05, 0xaa, 0xc3, 0x2a, 0x38, 0xf8, 0xbe, 0x85, 0xb7, 0x5f, 0x44,
+	0xb1, 0x06, 0xb9, 0x1f, 0x66, 0x7c, 0xc6, 0xc0, 0x67, 0x70, 0x9c, 0x81, 0xd2, 0x64, 0x1f, 0x37,
+	0x7d, 0x29, 0x12, 0x13, 0x59, 0x68, 0xd8, 0x18, 0x3b, 0xa7, 0xe7, 0x7d, 0xe3, 0xc7, 0x79, 0x7f,
+	0x37, 0x88, 0x74, 0x98, 0x4d, 0x6c, 0x4f, 0x24, 0x4e, 0x2a, 0x45, 0x02, 0x3a, 0x84, 0x4c, 0x39,
+	0x9e, 0x48, 0x12, 0xc1, 0x9d, 0x44, 0x4c, 0x21, 0xb6, 0x5f, 0x47, 0x09, 0xb0, 0x32, 0x99, 0x1c,
+	0xe2, 0xdb, 0x3a, 0x94, 0x22, 0x0b, 0x42, 0x73, 0xeb, 0xdf, 0xea, 0x54, 0xf9, 0xc4, 0xc6, 0x4d,
+	0x09, 0xbe, 0x32, 0x1b, 0x56, 0x63, 0xd8, 0x1d, 0xf5, 0xec, 0xcb, 0x46, 0x0e, 0xa4, 0xc8, 0x52,
+	0x98, 0x56, 0xfc, 0x15, 0x2b, 0x71, 0xc4, 0xc5, 0xcd, 0x34, 0x76, 0xb9, 0x79, 0xcb, 0x42, 0xc3,
+	0xee, 0xe8, 0x5e, 0x8d, 0x7f, 0x19, 0xbb, 0x7c, 0xfc, 0x7c, 0xcd, 0xe3, 0xe9, 0x15, 0x1e, 0x81,
+	0x74, 0x7d, 0x97, 0xbb, 0x4e, 0x2c, 0x66, 0x91, 0xf3, 0x7e, 0xcf, 0x29, 0x74, 0x3b, 0xce, 0x40,
+	0x46, 0x20, 0x9d, 0xa2, 0x94, 0xfd, 0x2a, 0x03, 0x79, 0x52, 0xa4, 0xb3, 0xb2, 0x34, 0xd9, 0xc1,
+	0xad, 0x49, 0x2c, 0xbc, 0x99, 0x32, 0x5b, 0x56, 0x63, 0xd8, 0x61, 0xeb, 0xdb, 0x80, 0xe1, 0x9d,
+	0xeb, 0x9a, 0xaa, 0x54, 0x70, 0x05, 0xe4, 0x19, 0xee, 0x78, 0x15, 0x4f, 0x13, 0xdd, 0xd8, 0x49,
+	0x0d, 0x1e, 0x7c, 0x43, 0xb8, 0x7d, 0x14, 0x0a, 0xa9, 0x19, 0xf8, 0xff, 0x9d, 0x37, 0x3d, 0xdc,
+	0xf6, 0x42, 0xf0, 0x66, 0x2a, 0x4b, 0xcc, 0x86, 0x85, 0x86, 0x77, 0xd9, 0xe5, 0x7d, 0xf0, 0x05,
+	0xe1, 0x07, 0xd7, 0x1b, 0x23, 0x16, 0xee, 0xfa, 0x11, 0x0f, 0x40, 0xa6, 0x32, 0xe2, 0xba, 0xec,
+	0xa3, 0xc9, 0xae, 0x3e, 0x15, 0xda, 0x6a, 0xe0, 0x2e, 0xd7, 0x25, 0xb9, 0x0e, 0x5b, 0xdf, 0xc8,
+	0xe3, 0x8d, 0x31, 0x20, 0xb5, 0x78, 0x95, 0x38, 0x6b, 0xfb, 0x9f, 0xe0, 0x56, 0xec, 0x4e, 0x20,
+	0x56, 0x66, 0xb3, 0x1c, 0x80, 0xed, 0x1a, 0x79, 0xc8, 0xa7, 0xf0, 0xf1, 0xa8, 0xf0, 0x55, 0xb1,
+	0x35, 0x68, 0xe4, 0xe3, 0x3b, 0xe3, 0x62, 0xb5, 0x0e, 0x56, 0xab, 0x45, 0xde, 0xe0, 0xfb, 0x9b,
+	0x16, 0x2a, 0xd2, 0xaf, 0x2b, 0xfc, 0x71, 0x63, 0x7a, 0xd6, 0xdf, 0x01, 0x2b, 0xfb, 0x07, 0xc6,
+	0xf8, 0xdd, 0x7c, 0x41, 0x8d, 0xb3, 0x05, 0x35, 0x2e, 0x16, 0x14, 0x7d, 0xca, 0x29, 0xfa, 0x9a,
+	0x53, 0x74, 0x9a, 0x53, 0x34, 0xcf, 0x29, 0xfa, 0x99, 0x53, 0xf4, 0x2b, 0xa7, 0xc6, 0x45, 0x4e,
+	0xd1, 0xe7, 0x25, 0x35, 0xe6, 0x4b, 0x6a, 0x9c, 0x2d, 0xa9, 0xf1, 0x76, 0xf7, 0x86, 0x29, 0xad,
+	0xfe, 0x3b, 0x69, 0x95, 0x9f, 0xbd, 0xdf, 0x01, 0x00, 0x00, 0xff, 0xff, 0xcf, 0x05, 0x31, 0x08,
+	0x35, 0x04, 0x00, 0x00,
 }

 func (this *FilterChunkRefRequest) Equal(that interface{}) bool {
@@ -416,6 +428,9 @@ func (this *GroupedChunkRefs) Equal(that interface{}) bool {
 			return false
 		}
 	}
+	if !this.Labels.Equal(that1.Labels) {
+		return false
+	}
 	return true
 }
 func (this *FilterChunkRefRequest) GoString() string {
@@ -462,13 +477,16 @@ func (this *GroupedChunkRefs) GoString() string {
 	if this == nil {
 		return "nil"
 	}
-	s := make([]string, 0, 7)
+	s := make([]string, 0, 8)
 	s = append(s, "&logproto.GroupedChunkRefs{")
 	s = append(s, "Fingerprint: "+fmt.Sprintf("%#v", this.Fingerprint)+",\n")
 	s = append(s, "Tenant: "+fmt.Sprintf("%#v", this.Tenant)+",\n")
 	if this.Refs != nil {
 		s = append(s, "Refs: "+fmt.Sprintf("%#v", this.Refs)+",\n")
 	}
+	if this.Labels != nil {
+		s = append(s, "Labels: "+fmt.Sprintf("%#v", this.Labels)+",\n")
+	}
 	s = append(s, "}")
 	return strings.Join(s, "")
 }
@@ -722,6 +740,18 @@ func (m *GroupedChunkRefs) MarshalToSizedBuffer(dAtA []byte) (int, error) {
 	_ = i
 	var l int
 	_ = l
+	if m.Labels != nil {
+		{
+			size, err := m.Labels.MarshalToSizedBuffer(dAtA[:i])
+			if err != nil {
+				return 0, err
+			}
+			i -= size
+			i = encodeVarintBloomgateway(dAtA, i, uint64(size))
+		}
+		i--
+		dAtA[i] = 0x22
+	}
 	if len(m.Refs) > 0 {
 		for iNdEx := len(m.Refs) - 1; iNdEx >= 0; iNdEx-- {
 			{
@@ -843,6 +873,10 @@ func (m *GroupedChunkRefs) Size() (n int) {
 			n += 1 + l + sovBloomgateway(uint64(l))
 		}
 	}
+	if m.Labels != nil {
+		l = m.Labels.Size()
+		n += 1 + l + sovBloomgateway(uint64(l))
+	}
 	return n
 }

@@ -911,6 +945,7 @@ func (this *GroupedChunkRefs) String() string {
 		`Fingerprint:` + fmt.Sprintf("%v", this.Fingerprint) + `,`,
 		`Tenant:` + fmt.Sprintf("%v", this.Tenant) + `,`,
 		`Refs:` + repeatedStringForRefs + `,`,
+		`Labels:` + strings.Replace(fmt.Sprintf("%v", this.Labels), "IndexSeries", "IndexSeries", 1) + `,`,
 		`}`,
 	}, "")
 	return s
@@ -1424,6 +1459,42 @@ func (m *GroupedChunkRefs) Unmarshal(dAtA []byte) error {
 				return err
 			}
 			iNdEx = postIndex
+		case 4:
+			if wireType != 2 {
+				return fmt.Errorf("proto: wrong wireType = %d for field Labels", wireType)
+			}
+			var msglen int
+			for shift := uint(0); ; shift += 7 {
+				if shift >= 64 {
+					return ErrIntOverflowBloomgateway
+				}
+				if iNdEx >= l {
+					return io.ErrUnexpectedEOF
+				}
+				b := dAtA[iNdEx]
+				iNdEx++
+				msglen |= int(b&0x7F) << shift
+				if b < 0x80 {
+					break
+				}
+			}
+			if msglen < 0 {
+				return ErrInvalidLengthBloomgateway
+			}
+			postIndex := iNdEx + msglen
+			if postIndex < 0 {
+				return ErrInvalidLengthBloomgateway
+			}
+			if postIndex > l {
+				return io.ErrUnexpectedEOF
+			}
+			if m.Labels == nil {
+				m.Labels = &IndexSeries{}
+			}
+			if err := m.Labels.Unmarshal(dAtA[iNdEx:postIndex]); err != nil {
+				return err
+			}
+			iNdEx = postIndex
 		default:
 			iNdEx = preIndex
 			skippy, err := skipBloomgateway(dAtA[iNdEx:])
--- a/pkg/logproto/bloomgateway.proto
+++ b/pkg/logproto/bloomgateway.proto
@@ -45,6 +45,9 @@ message GroupedChunkRefs {
  uint64 fingerprint = 1;
  string tenant = 2;
  repeated ShortRef refs = 3;
+  // Labels are only populated on FilterChunkRefRequest. They are not returned on FilterChunkRefResponse
+  // TODO(salvacorts): Consider two different messages for FilterChunkRefRequest and FilterChunkRefResponse
+  IndexSeries labels = 4;
 }

 service BloomGateway {
--- a/pkg/logproto/compat.go
+++ b/pkg/logproto/compat.go
@@ -391,6 +391,7 @@ func (m *FilterChunkRefRequest) WithStartEndForCache(start, end time.Time) resul
 		if len(refs) > 0 {
 			chunkRefs = append(chunkRefs, &GroupedChunkRefs{
 				Fingerprint: chunkRef.Fingerprint,
+				Labels:      chunkRef.Labels,
 				Tenant:      chunkRef.Tenant,
 				Refs:        refs,
 			})
--- a/pkg/storage/bloom/v1/bloom_tester.go
+++ b/pkg/storage/bloom/v1/bloom_tester.go
@@ -4,28 +4,32 @@ import (
 	"fmt"
 	"unsafe"

+	"github.com/go-kit/log/level"
+	"github.com/prometheus/prometheus/model/labels"
+
 	"github.com/grafana/loki/v3/pkg/storage/bloom/v1/filter"
+	util_log "github.com/grafana/loki/v3/pkg/util/log"
 )

 type BloomTest interface {
-	Matches(bloom filter.Checker) bool
-	MatchesWithPrefixBuf(bloom filter.Checker, buf []byte, prefixLen int) bool
+	Matches(series labels.Labels, bloom filter.Checker) bool
+	MatchesWithPrefixBuf(series labels.Labels, bloom filter.Checker, buf []byte, prefixLen int) bool
 }

 type BloomTests []BloomTest

-func (b BloomTests) Matches(bloom filter.Checker) bool {
+func (b BloomTests) Matches(series labels.Labels, bloom filter.Checker) bool {
 	for _, test := range b {
-		if !test.Matches(bloom) {
+		if !test.Matches(series, bloom) {
 			return false
 		}
 	}
 	return true
 }

-func (b BloomTests) MatchesWithPrefixBuf(bloom filter.Checker, buf []byte, prefixLen int) bool {
+func (b BloomTests) MatchesWithPrefixBuf(series labels.Labels, bloom filter.Checker, buf []byte, prefixLen int) bool {
 	for _, test := range b {
-		if !test.MatchesWithPrefixBuf(bloom, buf, prefixLen) {
+		if !test.MatchesWithPrefixBuf(series, bloom, buf, prefixLen) {
 			return false
 		}
 	}
@@ -37,12 +41,12 @@ type matchAllTest struct{}
 var MatchAll = matchAllTest{}

 // Matches implements BloomTest
-func (n matchAllTest) Matches(_ filter.Checker) bool {
+func (n matchAllTest) Matches(_ labels.Labels, _ filter.Checker) bool {
 	return true
 }

 // MatchesWithPrefixBuf implements BloomTest
-func (n matchAllTest) MatchesWithPrefixBuf(_ filter.Checker, _ []byte, _ int) bool {
+func (n matchAllTest) MatchesWithPrefixBuf(_ labels.Labels, _ filter.Checker, _ []byte, _ int) bool {
 	return true
 }

@@ -72,13 +76,13 @@ func newOrTest(left, right BloomTest) orTest {
 }

 // Matches implements BloomTest
-func (o orTest) Matches(bloom filter.Checker) bool {
-	return o.left.Matches(bloom) || o.right.Matches(bloom)
+func (o orTest) Matches(series labels.Labels, bloom filter.Checker) bool {
+	return o.left.Matches(series, bloom) || o.right.Matches(series, bloom)
 }

 // MatchesWithPrefixBuf implements BloomTest
-func (o orTest) MatchesWithPrefixBuf(bloom filter.Checker, buf []byte, prefixLen int) bool {
-	return o.left.MatchesWithPrefixBuf(bloom, buf, prefixLen) || o.right.MatchesWithPrefixBuf(bloom, buf, prefixLen)
+func (o orTest) MatchesWithPrefixBuf(series labels.Labels, bloom filter.Checker, buf []byte, prefixLen int) bool {
+	return o.left.MatchesWithPrefixBuf(series, bloom, buf, prefixLen) || o.right.MatchesWithPrefixBuf(series, bloom, buf, prefixLen)
 }

 type andTest struct {
@@ -93,13 +97,13 @@ func newAndTest(left, right BloomTest) andTest {
 }

 // Matches implements BloomTest
-func (a andTest) Matches(bloom filter.Checker) bool {
-	return a.left.Matches(bloom) && a.right.Matches(bloom)
+func (a andTest) Matches(series labels.Labels, bloom filter.Checker) bool {
+	return a.left.Matches(series, bloom) && a.right.Matches(series, bloom)
 }

 // MatchesWithPrefixBuf implements BloomTest
-func (a andTest) MatchesWithPrefixBuf(bloom filter.Checker, buf []byte, prefixLen int) bool {
-	return a.left.MatchesWithPrefixBuf(bloom, buf, prefixLen) && a.right.MatchesWithPrefixBuf(bloom, buf, prefixLen)
+func (a andTest) MatchesWithPrefixBuf(series labels.Labels, bloom filter.Checker, buf []byte, prefixLen int) bool {
+	return a.left.MatchesWithPrefixBuf(series, bloom, buf, prefixLen) && a.right.MatchesWithPrefixBuf(series, bloom, buf, prefixLen)
 }

 func LabelMatchersToBloomTest(matchers ...LabelMatcher) BloomTest {
@@ -144,7 +148,7 @@ func newStringMatcherTest(matcher PlainLabelMatcher) stringMatcherTest {
 	return stringMatcherTest{matcher: matcher}
 }

-func (sm stringMatcherTest) Matches(bloom filter.Checker) bool {
+func (sm stringMatcherTest) Matches(series labels.Labels, bloom filter.Checker) bool {
 	// TODO(rfratto): reintroduce the use of a shared tokenizer here to avoid
 	// desyncing between how tokens are passed during building vs passed during
 	// querying.
@@ -155,44 +159,38 @@ func (sm stringMatcherTest) Matches(bloom filter.Checker) bool {
 	// 2. It should be possible to test for just the key

 	var (
-		combined = fmt.Sprintf("%s=%s", sm.matcher.Key, sm.matcher.Value)
-
-		rawKey      = unsafe.Slice(unsafe.StringData(sm.matcher.Key), len(sm.matcher.Key))
+		combined    = fmt.Sprintf("%s=%s", sm.matcher.Key, sm.matcher.Value)
 		rawCombined = unsafe.Slice(unsafe.StringData(combined), len(combined))
 	)

-	if !bloom.Test(rawKey) {
-		// The structured metadata key wasn't indexed. However, sm.matcher might be
-		// checking against a label which *does* exist, so we can't safely filter
-		// out this chunk.
-		//
-		// TODO(rfratto): The negative test here is a bit confusing, and the key
-		// presence test should likely be done higher up within FuseQuerier.
-		return true
-	}
-
-	return bloom.Test(rawCombined)
+	return sm.match(series, bloom, rawCombined)
 }

-func (sm stringMatcherTest) MatchesWithPrefixBuf(bloom filter.Checker, buf []byte, prefixLen int) bool {
+func (sm stringMatcherTest) MatchesWithPrefixBuf(series labels.Labels, bloom filter.Checker, buf []byte, prefixLen int) bool {
 	var (
-		combined = fmt.Sprintf("%s=%s", sm.matcher.Key, sm.matcher.Value)
-
-		prefixedKey      = appendToBuf(buf, prefixLen, sm.matcher.Key)
+		combined         = fmt.Sprintf("%s=%s", sm.matcher.Key, sm.matcher.Value)
 		prefixedCombined = appendToBuf(buf, prefixLen, combined)
 	)

-	if !bloom.Test(prefixedKey) {
-		// The structured metadata key wasn't indexed for a prefix. However,
-		// sm.matcher might be checking against a label which *does* exist, so we
-		// can't safely filter out this chunk.
-		//
-		// TODO(rfratto): The negative test here is a bit confusing, and the key
-		// presence test should likely be done higher up within FuseQuerier.
+	return sm.match(series, bloom, prefixedCombined)
+}
+
+// match returns true if the series matches the matcher or is in the bloom filter.
+func (sm stringMatcherTest) match(series labels.Labels, bloom filter.Checker, combined []byte) bool {
+	// If we don't have the series labels, we cannot disambiguate which labels come from the series in which case
+	// we may filter out chunks for queries like `{env="prod"} | env="prod"` if env=prod is not structured metadata
+	if len(series) == 0 {
+		level.Warn(util_log.Logger).Log("msg", "series has no labels, cannot filter out chunks")
 		return true
 	}

-	return bloom.Test(prefixedCombined)
+	// It's in the series if the key is set and has the same value.
+	// By checking val != "" we handle `{env="prod"} | user=""`.
+	val := series.Get(sm.matcher.Key)
+	inSeries := val != "" && val == sm.matcher.Value
+
+	inBloom := bloom.Test(combined)
+	return inSeries || inBloom
 }

 // appendToBuf is the equivalent of append(buf[:prefixLen], str). len(buf) must
--- a/pkg/storage/bloom/v1/bloom_tester_test.go
+++ b/pkg/storage/bloom/v1/bloom_tester_test.go
@@ -3,6 +3,7 @@ package v1
 import (
 	"testing"

+	"github.com/prometheus/prometheus/model/labels"
 	"github.com/stretchr/testify/require"

 	"github.com/grafana/loki/v3/pkg/logql/syntax"
@@ -20,9 +21,11 @@ func TestLabelMatchersToBloomTest(t *testing.T) {
 			tokenizer,
 			push.LabelAdapter{Name: "trace_id", Value: "exists_1"},
 			push.LabelAdapter{Name: "trace_id", Value: "exists_2"},
+			push.LabelAdapter{Name: "app", Value: "other"},
 		)
 	)

+	series := labels.FromStrings("env", "prod", "app", "fake")
 	tt := []struct {
 		name  string
 		query string
@@ -54,15 +57,40 @@ func TestLabelMatchersToBloomTest(t *testing.T) {
 			match: false,
 		},
 		{
-			name:  "ignore non-indexed key",
+			name:  "filter non-indexed key",
 			query: `{app="fake"} | noexist="noexist"`,
+			match: false,
+		},
+		{
+			name:  "filter non-indexed key with empty value",
+			query: `{app="fake"} | noexist=""`,
+			match: false,
+		},
+		{
+			name:  "ignore label from series",
+			query: `{app="fake"} | env="prod"`,
 			match: true,
 		},
 		{
-			name:  "ignore non-indexed key with empty value",
-			query: `{app="fake"} | noexist=""`,
+			name:  "filter label from series",
+			query: `{app="fake"} | env="dev"`, // env is set to prod in the series
+			match: false,
+		},
+		{
+			name:  "ignore label from series and structured metadata",
+			query: `{app="fake"} | app="other"`,
 			match: true,
 		},
+		{
+			name:  "filter series label with non-existing value",
+			query: `{app="fake"} | app="noexist"`,
+			match: false,
+		},
+		{
+			name:  "ignore label from series with empty value",
+			query: `{app="fake"} | app=""`,
+			match: false,
+		},
 		{
 			name:  "ignore unsupported operator",
 			query: `{app="fake"} | trace_id=~".*noexist.*"`,
@@ -99,8 +127,8 @@ func TestLabelMatchersToBloomTest(t *testing.T) {
 			bloomTest := LabelMatchersToBloomTest(matchers...)

 			// .Matches and .MatchesWithPrefixBuf should both have the same result.
-			require.Equal(t, tc.match, bloomTest.Matches(bloom))
-			require.Equal(t, tc.match, bloomTest.MatchesWithPrefixBuf(bloom, []byte(prefix), len(prefix)))
+			require.Equal(t, tc.match, bloomTest.Matches(series, bloom))
+			require.Equal(t, tc.match, bloomTest.MatchesWithPrefixBuf(series, bloom, []byte(prefix), len(prefix)))
 		})
 	}
 }
--- a/pkg/storage/bloom/v1/fuse.go
+++ b/pkg/storage/bloom/v1/fuse.go
@@ -7,6 +7,7 @@ import (
 	"github.com/go-kit/log"
 	"github.com/go-kit/log/level"
 	"github.com/prometheus/common/model"
+	"github.com/prometheus/prometheus/model/labels"
 	"go.uber.org/atomic"

 	iter "github.com/grafana/loki/v3/pkg/iter/v2"
@@ -15,6 +16,7 @@ import (

 type Request struct {
 	Fp       model.Fingerprint
+	Labels   labels.Labels
 	Chks     ChunkRefs
 	Search   BloomTest
 	Response chan<- Output
@@ -252,6 +254,7 @@ func (fq *FusedQuerier) Run() error {
 		nextBatch := fq.inputs.At()

 		fp := nextBatch[0].Fp
+		lbs := nextBatch[0].Labels

 		// advance the series iterator to the next fingerprint
 		if err := fq.bq.Seek(fp); err != nil {
@@ -276,13 +279,13 @@ func (fq *FusedQuerier) Run() error {
 			continue
 		}

-		fq.runSeries(schema, series, nextBatch)
+		fq.runSeries(schema, lbs, series, nextBatch)
 	}

 	return nil
 }

-func (fq *FusedQuerier) runSeries(_ Schema, series *SeriesWithMeta, reqs []Request) {
+func (fq *FusedQuerier) runSeries(_ Schema, lbs labels.Labels, series *SeriesWithMeta, reqs []Request) {
 	// For a given chunk|series to be removed, it must fail to match all blooms.
 	// Because iterating/loading blooms can be expensive, we iterate blooms one at a time, collecting
 	// the removals (failures) for each (bloom, chunk) pair.
@@ -372,7 +375,7 @@ func (fq *FusedQuerier) runSeries(_ Schema, series *SeriesWithMeta, reqs []Reque
 			// shortcut: series level removal
 			// we can skip testing chunk keys individually if the bloom doesn't match
 			// the query.
-			if !req.Search.Matches(bloom) {
+			if !req.Search.Matches(lbs, bloom) {
 				// Nothing else needs to be done for this (bloom, request);
 				// check the next input request
 				continue
@@ -387,7 +390,7 @@ func (fq *FusedQuerier) runSeries(_ Schema, series *SeriesWithMeta, reqs []Reque
 				// TODO(rfratto): reuse buffer between multiple calls to
 				// prefixForChunkRef and MatchesWithPrefixBuf to avoid allocations.
 				tokenBuf := prefixForChunkRef(chk)
-				if matched := req.Search.MatchesWithPrefixBuf(bloom, tokenBuf, len(tokenBuf)); matched {
+				if matched := req.Search.MatchesWithPrefixBuf(lbs, bloom, tokenBuf, len(tokenBuf)); matched {
 					inputs[j].found[k] = true
 				}
 			}
--- a/pkg/storage/bloom/v1/fuse_test.go
+++ b/pkg/storage/bloom/v1/fuse_test.go
@@ -10,6 +10,7 @@ import (
 	"github.com/go-kit/log"
 	"github.com/grafana/dskit/concurrency"
 	"github.com/prometheus/common/model"
+	"github.com/prometheus/prometheus/model/labels"
 	"github.com/stretchr/testify/require"

 	"github.com/grafana/loki/v3/pkg/compression"
@@ -27,12 +28,12 @@ var BloomPagePool = mempool.New("test", []mempool.Bucket{
 type singleKeyTest []byte

 // Matches implements BloomTest.
-func (s singleKeyTest) Matches(bloom filter.Checker) bool {
+func (s singleKeyTest) Matches(_ labels.Labels, bloom filter.Checker) bool {
 	return bloom.Test(s)
 }

 // MatchesWithPrefixBuf implements BloomTest.
-func (s singleKeyTest) MatchesWithPrefixBuf(bloom filter.Checker, buf []byte, prefixLen int) bool {
+func (s singleKeyTest) MatchesWithPrefixBuf(_ labels.Labels, bloom filter.Checker, buf []byte, prefixLen int) bool {
 	return bloom.Test(append(buf[:prefixLen], s...))
 }