fix: empty search groupby result causing crash(milvus-io#43137) (milvus-io#43214)

related: milvus-io#43137

Signed-off-by: MrPresent-Han <chun.han@gmail.com>
Co-authored-by: MrPresent-Han <chun.han@gmail.com>

Author: MrPresent-Han

internal/core/src/segcore/ReduceUtils.cpp

@@ -25,12 +25,12 @@ AssembleGroupByValues(
     const std::vector<GroupByValueType>& group_by_vals,
     milvus::query::Plan* plan) {
     auto group_by_field_id = plan->plan_node_->search_info_.group_by_field_id_;
-    if (group_by_field_id.has_value() && group_by_vals.size() > 0) {
-        auto group_by_values_field =
-            std::make_unique<milvus::proto::schema::ScalarField>();
+    if (group_by_field_id.has_value()) {
         auto valid_data =
             std::make_unique<google::protobuf::RepeatedField<bool>>();
         valid_data->Resize(group_by_vals.size(), true);
+        auto group_by_values_field =
+            std::make_unique<milvus::proto::schema::ScalarField>();
         auto group_by_field =
             plan->schema_->operator[](group_by_field_id.value());
         auto group_by_data_type = group_by_field.get_data_type();
@@ -134,7 +134,6 @@ AssembleGroupByValues(
         group_by_field_value->mutable_valid_data()->MergeFrom(*valid_data);
         group_by_field_value->mutable_scalars()->MergeFrom(
             *group_by_values_field.get());
-        return;
     }
 }
 

internal/proxy/search_reduce_util.go

@@ -118,7 +118,7 @@ func reduceAdvanceGroupBy(ctx context.Context, subSearchResultData []*schemapb.S
 
 	gpFieldBuilder, err := typeutil.NewFieldDataBuilder(subSearchResultData[0].GetGroupByFieldValue().GetType(), true, int(limit))
 	if err != nil {
-		return ret, err
+		return ret, merr.WrapErrServiceInternal("failed to construct group by field data builder, this is abnormal as segcore should always set up a group by field, no matter data status, check code on qn", err.Error())
 	}
 	// reducing nq * topk results
 	for nqIdx := int64(0); nqIdx < nq; nqIdx++ {
@@ -222,9 +222,10 @@ func reduceSearchResultDataWithGroupBy(ctx context.Context, subSearchResultData
 		totalResCount += subSearchNqOffset[i][nq-1]
 		subSearchGroupByValIterator[i] = typeutil.GetDataIterator(subSearchResultData[i].GetGroupByFieldValue())
 	}
+
 	gpFieldBuilder, err := typeutil.NewFieldDataBuilder(subSearchResultData[0].GetGroupByFieldValue().GetType(), true, int(limit))
 	if err != nil {
-		return ret, err
+		return ret, merr.WrapErrServiceInternal("failed to construct group by field data builder, this is abnormal as segcore should always set up a group by field, no matter data status, check code on qn", err.Error())
 	}
 
 	var realTopK int64 = -1

internal/proxy/search_reduce_util_test.go

@@ -61,6 +61,29 @@ func (struts *SearchReduceUtilTestSuite) TestReduceSearchResult() {
 	}
 }
 
+func (struts *SearchReduceUtilTestSuite) TestReduceSearchResultWithEmtpyGroupData() {
+	nq := int64(1)
+	topk := int64(1)
+	emptyData := &schemapb.SearchResultData{
+		NumQueries:       nq,
+		TopK:             topk,
+		FieldsData:       make([]*schemapb.FieldData, 0),
+		Scores:           make([]float32, 0),
+		Ids:              &schemapb.IDs{},
+		Topks:            make([]int64, 0),
+		OutputFields:     make([]string, 0),
+		AllSearchCount:   0,
+		Distances:        make([]float32, 0),
+		Recalls:          make([]float32, 0),
+		PrimaryFieldName: "",
+	}
+	results, err := reduceSearchResultDataWithGroupBy(context.Background(), []*schemapb.SearchResultData{emptyData},
+		nq, topk, "L2", schemapb.DataType_Int64, 0, 1)
+	struts.Error(err)
+	struts.ErrorContains(err, "failed to construct group by field data builder")
+	struts.Nil(results.Results.GetGroupByFieldValue())
+}
+
 func TestSearchReduceUtilTestSuite(t *testing.T) {
 	suite.Run(t, new(SearchReduceUtilTestSuite))
 }

internal/querynodev2/delegator/delegator.go

@@ -368,7 +368,7 @@ func (sd *shardDelegator) search(ctx context.Context, req *querypb.SearchRequest
 		return nil, err
 	}
 
-	log.Debug("Delegator search done")
+	log.Debug("Delegator search done", zap.Int("results", len(results)))
 
 	return results, nil
 }

internal/querynodev2/segments/result.go

@@ -57,6 +57,7 @@ func ReduceSearchResults(ctx context.Context, results []*internalpb.SearchResult
 	})
 
 	if len(results) == 1 {
+		log.Debug("Shortcut return ReduceSearchResults", zap.Any("result info", info))
 		return results[0], nil
 	}
 

internal/querynodev2/segments/search_reduce.go

@@ -10,6 +10,7 @@ import (
 	"github.com/milvus-io/milvus-proto/go-api/v2/schemapb"
 	"github.com/milvus-io/milvus/internal/util/reduce"
 	"github.com/milvus-io/milvus/pkg/v2/log"
+	"github.com/milvus-io/milvus/pkg/v2/util/merr"
 	"github.com/milvus-io/milvus/pkg/v2/util/paramtable"
 	"github.com/milvus-io/milvus/pkg/v2/util/typeutil"
 )
@@ -107,6 +108,7 @@ func (sbr *SearchGroupByReduce) ReduceSearchResultData(ctx context.Context, sear
 	log := log.Ctx(ctx)
 
 	if len(searchResultData) == 0 {
+		log.Debug("Shortcut return SearchGroupByReduce, directly return empty result", zap.Any("result info", info))
 		return &schemapb.SearchResultData{
 			NumQueries: info.GetNq(),
 			TopK:       info.GetTopK(),
@@ -137,7 +139,7 @@ func (sbr *SearchGroupByReduce) ReduceSearchResultData(ctx context.Context, sear
 	}
 	gpFieldBuilder, err := typeutil.NewFieldDataBuilder(searchResultData[0].GetGroupByFieldValue().GetType(), true, int(info.GetTopK()))
 	if err != nil {
-		return nil, err
+		return ret, merr.WrapErrServiceInternal("failed to construct group by field data builder, this is abnormal as segcore should always set up a group by field, no matter data status, check code on qn", err.Error())
 	}
 
 	var filteredCount int64

pkg/util/typeutil/schema.go

@@ -1458,7 +1458,7 @@ func GetPK(data *schemapb.IDs, idx int64) interface{} {
 }
 
 func GetDataIterator(field *schemapb.FieldData) func(int) any {
-	if field.ValidData != nil {
+	if field.GetValidData() != nil {
 		// unpack valid data
 		idxs := make([]int, len(field.ValidData))
 		validCnt := 0