set search_list(ef) default size and throw exception if search list(e…

…f) < topk (#923)

Signed-off-by: cqy123456 <[email protected]>

knowhere/index/vector_index/IndexDiskANNConfig.cpp

@@ -220,13 +220,13 @@ void
 from_json(const Config& config, DiskANNQueryConfig& query_conf) {
     CheckNumericParamAndSet<uint64_t>(config, kK, kKMinValue, kKMaxValue, query_conf.k);
     auto search_list_threshold = query_conf.k < kKThreshold ? kKThreshold : query_conf.k;
-    if (config.contains(kSearchListSize) && config[kSearchListSize].get<uint32_t>() >= search_list_threshold) {
+    if (config.contains(kSearchListSize)) {
         // The search_list_size should be no less than the k.
         CheckNumericParamAndSet<uint32_t>(config, kSearchListSize, query_conf.k,
                                         std::max(kSearchListSizeMaxValue, static_cast<uint32_t>(10 * query_conf.k)),
                                         query_conf.search_list_size);
     } else {
-        // if search_list_size not set (==0), not in json string or smaller than k, modify the value.
+        // if search_list_size not set (==0), not in json string, modify the value.
         query_conf.search_list_size = search_list_threshold;
     }
     CheckNumericParamAndSet<uint32_t>(config, kBeamwidth, kBeamwidthMinValue, kBeamwidthMaxValue, query_conf.beamwidth);

knowhere/index/vector_index/IndexDiskANNConfig.h

@@ -86,7 +86,7 @@ struct DiskANNQueryConfig {
     uint64_t k;
     // A list of search_list sizes to perform searches with. Larger parameters will result in slower latencies, but
     // higher accuracies. Must be at least the value of k.
-    uint32_t search_list_size = 0;
+    uint32_t search_list_size = 128;
     // The beamwidth to be used for search. This is the maximum number of IO requests each query will issue per
     // iteration of search code. Larger beamwidth will result in fewer IO round-trips per query but might result in
     // slightly higher total number of IO requests to SSD per query. For the highest query throughput with a fixed SSD

knowhere/index/vector_index/IndexHNSW.cpp

@@ -29,6 +29,21 @@
 #include "index/vector_index/helpers/RangeUtil.h"
 
 namespace knowhere {
+namespace {
+    inline int64_t
+    CheckAndGetEfValue(const Config& config) {
+        auto topk_val = GetMetaTopk(config);
+        if (CheckKeyInConfig(config, indexparam::EF)) {
+            auto ef_val = GetIndexParamEf(config);
+            if (ef_val < topk_val) {
+                KNOWHERE_THROW_MSG("ef is smaller than topk in hnsw.");
+            }
+            return ef_val;
+        } else {
+            return std::max(knowhere::DEFAULT_HNSW_EF, topk_val);
+        }
+    }
+}  // namespace
 
 BinarySet
 IndexHNSW::Serialize(const Config& config) {
@@ -252,7 +267,8 @@ IndexHNSW::QueryImpl(int64_t n, const float* xq, int64_t k, float* distances, in
         feder = std::make_unique<feder::hnsw::FederResult>();
     }
 
-    size_t ef = GetIndexParamEf(config);
+    size_t ef = CheckAndGetEfValue(config);
+
     hnswlib::SearchParam param{ef};
     bool transform = (index_->metric_type_ == 1);  // InnerProduct: 1
 
@@ -293,7 +309,7 @@ IndexHNSW::QueryByRangeImpl(int64_t n, const float* xq, float*& distances, int64
         feder = std::make_unique<feder::hnsw::FederResult>();
     }
 
-    size_t ef = GetIndexParamEf(config);
+    size_t ef = CheckAndGetEfValue(config);
     hnswlib::SearchParam param{ef};
 
     float radius = GetMetaRadius(config);

knowhere/index/vector_index/helpers/IndexParameter.h

@@ -176,7 +176,7 @@ DEFINE_CONFIG_SETTER(SetIndexParamEfConstruction, indexparam::EFCONSTRUCTION, in
 DEFINE_CONFIG_GETTER_WITH_DEFAULT_VALUE(GetIndexParamHNSWM, indexparam::HNSW_M, DEFAULT_HNSW_M, int64_t)
 DEFINE_CONFIG_SETTER(SetIndexParamHNSWM, indexparam::HNSW_M, int64_t)
 
-DEFINE_CONFIG_GETTER_WITH_DEFAULT_VALUE(GetIndexParamEf, indexparam::EF, DEFAULT_HNSW_EF, int64_t)
+DEFINE_CONFIG_GETTER(GetIndexParamEf, indexparam::EF, int64_t)
 DEFINE_CONFIG_SETTER(SetIndexParamEf, indexparam::EF, int64_t)
 
 DEFINE_CONFIG_GETTER(GetIndexParamOverviewLevels, indexparam::OVERVIEW_LEVELS, int64_t)

unittest/test_diskann.cpp

@@ -477,7 +477,7 @@ TEST_P(DiskANNTest, search_without_search_list_size) {
     cfg.clear();
     knowhere::DiskANNQueryConfig::Set(cfg, tmp_config);
     search_list_size = knowhere::DiskANNQueryConfig::Get(cfg).search_list_size;
-    EXPECT_EQ(search_list_size, 16);
+    EXPECT_EQ(search_list_size, 128);
 }
 
 TEST_P(DiskANNTest, knn_search_test) {