enhance: add vector reserve to improve memory allocation in segcore (#45757)

This commit optimizes std::vector usage across segcore by adding
reserve() calls where the size is known in advance, reducing memory
reallocations during push_back operations.

Changes:
- TimestampIndex.cpp: Reserve space for prefix_sums and
timestamp_barriers
- SegmentGrowingImpl.cpp: Reserve space for binlog info vectors
- ChunkedSegmentSealedImpl.cpp: Reserve space for futures and field data
vectors
- storagev2translator/GroupChunkTranslator.cpp: Reserve space for
metadata vectors

This improves performance by avoiding multiple memory reallocations when
the vector size is predictable.

issue: #45679

---------

Signed-off-by: Buqian Zheng <[email protected]>

Author: zhengbuqian

internal/core/src/segcore/ChunkedSegmentSealedImpl.cpp

@@ -391,6 +391,7 @@ ChunkedSegmentSealedImpl::load_column_group_data_internal(
         // warmup will be disabled only when all columns are not in load list
         bool merged_in_load_list = false;
         std::vector<FieldId> milvus_field_ids;
+        milvus_field_ids.reserve(field_id_list.size());
         for (int i = 0; i < field_id_list.size(); ++i) {
             milvus_field_ids.push_back(FieldId(field_id_list.Get(i)));
             merged_in_load_list = merged_in_load_list ||
@@ -2672,6 +2673,7 @@ ChunkedSegmentSealedImpl::Load(milvus::tracer::TraceContext& trace_ctx) {
     // Step 2: Load indexes in parallel using thread pool
     auto& pool = ThreadPools::GetThreadPool(milvus::ThreadPoolPriority::LOW);
     std::vector<std::future<void>> load_index_futures;
+    load_index_futures.reserve(field_id_to_index_info.size());
 
     for (const auto& pair : field_id_to_index_info) {
         auto field_id = pair.first;
@@ -2704,6 +2706,7 @@ ChunkedSegmentSealedImpl::Load(milvus::tracer::TraceContext& trace_ctx) {
 
     // Wait for all index loading to complete and collect exceptions
     std::vector<std::exception_ptr> index_exceptions;
+    index_exceptions.reserve(load_index_futures.size());
     for (auto& future : load_index_futures) {
         try {
             future.get();
@@ -2754,6 +2757,10 @@ ChunkedSegmentSealedImpl::Load(milvus::tracer::TraceContext& trace_ctx) {
 
         // Calculate total row count and collect binlog paths
         int64_t total_entries = 0;
+        auto binlog_count = field_binlog.binlogs().size();
+        field_binlog_info.insert_files.reserve(binlog_count);
+        field_binlog_info.entries_nums.reserve(binlog_count);
+        field_binlog_info.memory_sizes.reserve(binlog_count);
         for (const auto& binlog : field_binlog.binlogs()) {
             field_binlog_info.insert_files.push_back(binlog.log_path());
             field_binlog_info.entries_nums.push_back(binlog.entries_num());
@@ -2774,6 +2781,7 @@ ChunkedSegmentSealedImpl::Load(milvus::tracer::TraceContext& trace_ctx) {
                  field_data_to_load.size(),
                  id_);
         std::vector<std::future<void>> load_field_futures;
+        load_field_futures.reserve(field_data_to_load.size());
 
         for (const auto& [field_id, load_field_data_info] :
              field_data_to_load) {
@@ -2787,6 +2795,7 @@ ChunkedSegmentSealedImpl::Load(milvus::tracer::TraceContext& trace_ctx) {
 
         // Wait for all field data loading to complete and collect exceptions
         std::vector<std::exception_ptr> field_exceptions;
+        field_exceptions.reserve(load_field_futures.size());
         for (auto& future : load_field_futures) {
             try {
                 future.get();

internal/core/src/segcore/SegmentGrowingImpl.cpp

@@ -1344,6 +1344,10 @@ SegmentGrowingImpl::Load(milvus::tracer::TraceContext& trace_ctx) {
 
         // Process each binlog
         int64_t total_row_count = 0;
+        auto binlog_count = field_binlog.binlogs().size();
+        binlog_info.entries_nums.reserve(binlog_count);
+        binlog_info.insert_files.reserve(binlog_count);
+        binlog_info.memory_sizes.reserve(binlog_count);
         for (const auto& binlog : field_binlog.binlogs()) {
             binlog_info.entries_nums.push_back(binlog.entries_num());
             binlog_info.insert_files.push_back(binlog.log_path());
@@ -1353,6 +1357,7 @@ SegmentGrowingImpl::Load(milvus::tracer::TraceContext& trace_ctx) {
         binlog_info.row_count = total_row_count;
 
         // Set child field ids
+        binlog_info.child_field_ids.reserve(field_binlog.child_fields().size());
         for (const auto& child_field : field_binlog.child_fields()) {
             binlog_info.child_field_ids.push_back(child_field);
         }

internal/core/src/segcore/TimestampIndex.cpp

@@ -23,9 +23,11 @@ TimestampIndex::build_with(const Timestamp* timestamps, int64_t size) {
     auto num_slice = lengths_.size();
     Assert(num_slice > 0);
     std::vector<int64_t> prefix_sums;
+    prefix_sums.reserve(num_slice + 1);
     int offset = 0;
     prefix_sums.push_back(offset);
     std::vector<Timestamp> timestamp_barriers;
+    timestamp_barriers.reserve(num_slice + 1);
     Timestamp last_max_v = 0;
     for (int slice_id = 0; slice_id < num_slice; ++slice_id) {
         auto length = lengths_[slice_id];

internal/core/src/segcore/storagev2translator/GroupChunkTranslator.cpp

@@ -89,6 +89,8 @@ GroupChunkTranslator::GroupChunkTranslator(
                   .GetArrowFileSystem();
 
     // Get row group metadata from files
+    parquet_file_metadata_.reserve(insert_files_.size());
+    row_group_meta_list_.reserve(insert_files_.size());
     for (const auto& file : insert_files_) {
         auto reader = std::make_shared<milvus_storage::FileRowGroupReader>(
             fs,
@@ -114,11 +116,16 @@ GroupChunkTranslator::GroupChunkTranslator(
     file_row_group_prefix_sum_.reserve(row_group_meta_list_.size() + 1);
     file_row_group_prefix_sum_.push_back(
         0);  // Base case: 0 row groups before first file
+    size_t total_row_groups = 0;
     for (const auto& file_metas : row_group_meta_list_) {
+        total_row_groups += file_metas.size();
         file_row_group_prefix_sum_.push_back(file_row_group_prefix_sum_.back() +
                                              file_metas.size());
     }
 
+    meta_.num_rows_until_chunk_.reserve(total_row_groups + 1);
+    meta_.chunk_memory_size_.reserve(total_row_groups);
+
     meta_.num_rows_until_chunk_.push_back(0);
     for (const auto& row_group_meta : row_group_meta_list_) {
         for (int i = 0; i < row_group_meta.size(); ++i) {

internal/core/thirdparty/tantivy/tantivy-wrapper.h

@@ -355,6 +355,7 @@ struct TantivyIndexWrapper {
                         int64_t offset_begin) {
         assert(!finished_);
         std::vector<const char*> views;
+        views.reserve(len);
         for (uintptr_t i = 0; i < len; i++) {
             views.push_back(array[i].c_str());
         }
@@ -435,6 +436,7 @@ struct TantivyIndexWrapper {
 
         if constexpr (std::is_same_v<T, std::string>) {
             std::vector<const char*> views;
+            views.reserve(len);
             for (uintptr_t i = 0; i < len; i++) {
                 views.push_back(array[i].c_str());
             }
@@ -621,6 +623,7 @@ struct TantivyIndexWrapper {
 
         if constexpr (std::is_same_v<T, std::string>) {
             std::vector<const char*> views;
+            views.reserve(len);
             for (uintptr_t i = 0; i < len; i++) {
                 views.push_back(array[i].c_str());
             }
@@ -709,6 +712,7 @@ struct TantivyIndexWrapper {
                 } else {
                     // smaller integer should be converted first
                     std::vector<int64_t> buf(len);
+                    buf.reserve(len);
                     for (uintptr_t i = 0; i < len; ++i) {
                         buf[i] = static_cast<int64_t>(terms[i]);
                     }
@@ -726,6 +730,7 @@ struct TantivyIndexWrapper {
                         bitset);
                 } else {
                     std::vector<double> buf(len);
+                    buf.reserve(len);
                     for (uintptr_t i = 0; i < len; ++i) {
                         buf[i] = static_cast<double>(terms[i]);
                     }
@@ -736,6 +741,7 @@ struct TantivyIndexWrapper {
 
             if constexpr (std::is_same_v<T, std::string>) {
                 std::vector<const char*> views;
+                views.reserve(len);
                 for (uintptr_t i = 0; i < len; i++) {
                     views.push_back(terms[i].c_str());
                 }