Push down label_name filter on Series API (#2163)

This will filter label names on the ingesters directly.

api/ingester/v1/ingester.proto

@@ -26,6 +26,7 @@ message ProfileTypesResponse {
 
 message SeriesRequest {
   repeated string matchers = 1;
+  repeated string label_names = 2;
 }
 
 message SeriesResponse {

docs/sources/configure-server/reference-configuration-parameters/index.md

@@ -221,11 +221,11 @@ limits:
   # Maximum size of a profile in bytes. This is based off the uncompressed size.
   # 0 to disable.
   # CLI flag: -validation.max-profile-size-bytes
-  [max_profile_size_bytes: <int> | default = 2097152]
+  [max_profile_size_bytes: <int> | default = 4194304]
 
   # Maximum number of samples in a profile. 0 to disable.
   # CLI flag: -validation.max-profile-stacktrace-samples
-  [max_profile_stacktrace_samples: <int> | default = 2000]
+  [max_profile_stacktrace_samples: <int> | default = 4000]
 
   # Maximum number of labels in a profile sample. 0 to disable.
   # CLI flag: -validation.max-profile-stacktrace-sample-labels

go.work.sum

@@ -420,6 +420,7 @@ github.com/go-kit/kit v0.12.0 h1:e4o3o3IsBfAKQh5Qbbiqyfu97Ku7jrO/JbohvztANh4=
 github.com/go-kit/kit v0.12.0/go.mod h1:lHd+EkCZPIwYItmGDDRdhinkzX2A1sj+M9biaEaizzs=
 github.com/go-logr/logr v0.1.0/go.mod h1:ixOQHD9gLJUVQQ2ZOR7zLEifBX6tGkNJF4QyIY7sIas=
 github.com/go-ole/go-ole v1.2.6 h1:/Fpf6oFPoeFik9ty7siob0G6Ke8QvQEuVcuChpwXzpY=
+github.com/go-ole/go-ole v1.2.6/go.mod h1:pprOEPIfldk/42T2oK7lQ4v4JSDwmV0As9GaiUsvbm0=
 github.com/go-openapi/runtime v0.25.0 h1:7yQTCdRbWhX8vnIjdzU8S00tBYf7Sg71EBeorlPHvhc=
 github.com/go-openapi/runtime v0.25.0/go.mod h1:Ux6fikcHXyyob6LNWxtE96hWwjBPYF0DXgVFuMTneOs=
 github.com/go-playground/assert/v2 v2.0.1 h1:MsBgLAaY856+nPRTKrp3/OZK38U/wa0CcBYNjji3q3A=
@@ -739,6 +740,7 @@ github.com/shurcooL/vfsgen v0.0.0-20200824052919-0d455de96546/go.mod h1:TrYk7fJV
 github.com/sirupsen/logrus v1.7.0/go.mod h1:yWOB1SBYBC5VeMP7gHvWumXLIWorT60ONWic61uBYv0=
 github.com/slok/go-http-metrics v0.9.0 h1:o7A7j2DHs7Bnz8aGMotQLele28vf/Cl+O9cwJ6HyGk4=
 github.com/slok/go-http-metrics v0.9.0/go.mod h1:VCio4Xl8m11JM/0Sl9265RdKyiMypzMo3w1M8xcZGtk=
+github.com/spaolacci/murmur3 v1.1.0/go.mod h1:JwIasOWyU6f++ZhiEuf87xNszmSA2myDM2Kzu9HwQUA=
 github.com/spf13/afero v1.8.2/go.mod h1:CtAatgMJh6bJEIs48Ay/FOnkljP3WeGUG0MC1RfAqwo=
 github.com/spf13/afero v1.9.5 h1:stMpOSZFs//0Lv29HduCmli3GUfpFoF3Y1Q/aXj/wVM=
 github.com/spf13/afero v1.9.5/go.mod h1:UBogFpq8E9Hx+xc5CNTTEpTnuHVmXDwZcZcE1eb/UhQ=

pkg/phlaredb/head.go

@@ -877,9 +877,22 @@ func (h *Head) Series(ctx context.Context, req *connect.Request[ingestv1.SeriesR
 	if err != nil {
 		return nil, err
 	}
+
+	// build up map of label names
+	labelNameMap := make(map[string]struct{}, len(req.Msg.LabelNames))
+	for _, labelName := range req.Msg.LabelNames {
+		labelNameMap[labelName] = struct{}{}
+	}
+
 	response := &ingestv1.SeriesResponse{}
 	uniqu := map[model.Fingerprint]struct{}{}
 	if err := h.forMatchingSelectors(selectors, func(lbs phlaremodel.Labels, fp model.Fingerprint) error {
+		if len(req.Msg.LabelNames) > 0 {
+			lbs = lbs.WithLabels(req.Msg.LabelNames...)
+			fp = model.Fingerprint(lbs.Hash())
+
+		}
+
 		if _, ok := uniqu[fp]; ok {
 			return nil
 		}
@@ -889,6 +902,7 @@ func (h *Head) Series(ctx context.Context, req *connect.Request[ingestv1.SeriesR
 	}); err != nil {
 		return nil, err
 	}
+
 	sort.Slice(response.LabelsSet, func(i, j int) bool {
 		return phlaremodel.CompareLabelPairs(response.LabelsSet[i].Labels, response.LabelsSet[j].Labels) < 0
 	})

pkg/phlaredb/head_test.go

@@ -307,18 +307,30 @@ func TestHeadSeries(t *testing.T) {
 	require.NoError(t, head.Ingest(context.Background(), newProfileFoo(), uuid.New(), fooLabels...))
 	require.NoError(t, head.Ingest(context.Background(), newProfileBar(), uuid.New(), barLabels...))
 
-	expected := phlaremodel.NewLabelsBuilder(nil).
+	lblBuilder := phlaremodel.NewLabelsBuilder(nil).
 		Set("namespace", "phlare").
 		Set("job", "foo").
 		Set("__period_type__", "type").
 		Set("__period_unit__", "unit").
 		Set("__type__", "type").
 		Set("__unit__", "unit").
-		Set("__profile_type__", ":type:unit:type:unit").
-		Labels()
+		Set("__profile_type__", ":type:unit:type:unit")
+	expected := lblBuilder.Labels()
 	res, err := head.Series(context.Background(), connect.NewRequest(&ingestv1.SeriesRequest{Matchers: []string{`{job="foo"}`}}))
 	require.NoError(t, err)
 	require.Equal(t, []*typesv1.Labels{{Labels: expected}}, res.Msg.LabelsSet)
+
+	// Test we can filter labelNames
+	res, err = head.Series(context.Background(), connect.NewRequest(&ingestv1.SeriesRequest{LabelNames: []string{"job", "not-existing"}}))
+	require.NoError(t, err)
+	lblBuilder.Reset(nil)
+	jobFoo := lblBuilder.Set("job", "foo").Labels()
+	lblBuilder.Reset(nil)
+	jobBar := lblBuilder.Set("job", "bar").Labels()
+	require.Len(t, res.Msg.LabelsSet, 2)
+	require.Contains(t, res.Msg.LabelsSet, &typesv1.Labels{Labels: jobFoo})
+	require.Contains(t, res.Msg.LabelsSet, &typesv1.Labels{Labels: jobBar})
+
 }
 
 func TestHeadProfileTypes(t *testing.T) {

pkg/querier/querier.go

@@ -229,11 +229,13 @@ func (q *Querier) Series(ctx context.Context, req *connect.Request[querierv1.Ser
 
 	responses, err := forAllIngesters(ctx, q.ingesterQuerier, func(childCtx context.Context, ic IngesterQueryClient) ([]*typesv1.Labels, error) {
 		res, err := ic.Series(childCtx, connect.NewRequest(&ingestv1.SeriesRequest{
-			Matchers: req.Msg.Matchers,
+			Matchers:   req.Msg.Matchers,
+			LabelNames: req.Msg.LabelNames,
 		}))
 		if err != nil {
 			return nil, err
 		}
+		// TODO: Remove this once we have the ingesters returning the filtered response, has been rolled out (f25).
 		filterSeriesByLabelNames(res.Msg.LabelsSet, labelNameMap)
 
 		return res.Msg.LabelsSet, nil