milvus/internal/core/src/segcore/SegmentSealedImpl.cpp
cai.zhang fb089cda8b
enhance: Load raw data while scalar index doesn't have raw data (#28888)
issue: #28886

Signed-off-by: Cai Zhang <cai.zhang@zilliz.com>
2023-12-06 20:36:36 +08:00

1491 lines
57 KiB
C++

// Copyright (C) 2019-2020 Zilliz. All rights reserved.
//
// Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance
// with the License. You may obtain a copy of the License at
//
// http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software distributed under the License
// is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express
// or implied. See the License for the specific language governing permissions and limitations under the License
#include "SegmentSealedImpl.h"
#include <fcntl.h>
#include <fmt/core.h>
#include <algorithm>
#include <cstdint>
#include <filesystem>
#include <memory>
#include <string>
#include <string_view>
#include <unordered_map>
#include <vector>
#include "Utils.h"
#include "Types.h"
#include "common/Json.h"
#include "common/LoadInfo.h"
#include "common/EasyAssert.h"
#include "common/Array.h"
#include "google/protobuf/message_lite.h"
#include "mmap/Column.h"
#include "common/Consts.h"
#include "common/FieldMeta.h"
#include "common/Types.h"
#include "log/Log.h"
#include "pb/schema.pb.h"
#include "mmap/Types.h"
#include "query/ScalarIndex.h"
#include "query/SearchBruteForce.h"
#include "query/SearchOnSealed.h"
#include "storage/FieldData.h"
#include "storage/Util.h"
#include "storage/ThreadPools.h"
#include "storage/ChunkCacheSingleton.h"
#include "common/File.h"
#include "common/Tracer.h"
#include "index/VectorMemIndex.h"
namespace milvus::segcore {
static inline void
set_bit(BitsetType& bitset, FieldId field_id, bool flag = true) {
auto pos = field_id.get() - START_USER_FIELDID;
AssertInfo(pos >= 0, "invalid field id");
bitset[pos] = flag;
}
static inline bool
get_bit(const BitsetType& bitset, FieldId field_id) {
auto pos = field_id.get() - START_USER_FIELDID;
AssertInfo(pos >= 0, "invalid field id");
return bitset[pos];
}
void
SegmentSealedImpl::LoadIndex(const LoadIndexInfo& info) {
// print(info);
// NOTE: lock only when data is ready to avoid starvation
auto field_id = FieldId(info.field_id);
auto& field_meta = schema_->operator[](field_id);
if (field_meta.is_vector()) {
LoadVecIndex(info);
} else {
LoadScalarIndex(info);
}
}
void
SegmentSealedImpl::LoadVecIndex(const LoadIndexInfo& info) {
// NOTE: lock only when data is ready to avoid starvation
auto field_id = FieldId(info.field_id);
auto& field_meta = schema_->operator[](field_id);
AssertInfo(info.index_params.count("metric_type"),
"Can't get metric_type in index_params");
auto metric_type = info.index_params.at("metric_type");
auto row_count = info.index->Count();
AssertInfo(row_count > 0, "Index count is 0");
std::unique_lock lck(mutex_);
AssertInfo(
!get_bit(index_ready_bitset_, field_id),
"vector index has been exist at " + std::to_string(field_id.get()));
if (num_rows_.has_value()) {
AssertInfo(num_rows_.value() == row_count,
"field (" + std::to_string(field_id.get()) +
") data has different row count (" +
std::to_string(row_count) +
") than other column's row count (" +
std::to_string(num_rows_.value()) + ")");
}
if (get_bit(field_data_ready_bitset_, field_id)) {
fields_.erase(field_id);
set_bit(field_data_ready_bitset_, field_id, false);
} else if (get_bit(binlog_index_bitset_, field_id)) {
set_bit(binlog_index_bitset_, field_id, false);
vector_indexings_.drop_field_indexing(field_id);
}
update_row_count(row_count);
vector_indexings_.append_field_indexing(
field_id,
metric_type,
std::move(const_cast<LoadIndexInfo&>(info).index));
set_bit(index_ready_bitset_, field_id, true);
}
void
SegmentSealedImpl::LoadScalarIndex(const LoadIndexInfo& info) {
// NOTE: lock only when data is ready to avoid starvation
auto field_id = FieldId(info.field_id);
auto& field_meta = schema_->operator[](field_id);
auto row_count = info.index->Count();
AssertInfo(row_count > 0, "Index count is 0");
std::unique_lock lck(mutex_);
AssertInfo(
!get_bit(index_ready_bitset_, field_id),
"scalar index has been exist at " + std::to_string(field_id.get()));
if (num_rows_.has_value()) {
AssertInfo(num_rows_.value() == row_count,
"field (" + std::to_string(field_id.get()) +
") data has different row count (" +
std::to_string(row_count) +
") than other column's row count (" +
std::to_string(num_rows_.value()) + ")");
}
scalar_indexings_[field_id] =
std::move(const_cast<LoadIndexInfo&>(info).index);
// reverse pk from scalar index and set pks to offset
if (schema_->get_primary_field_id() == field_id) {
AssertInfo(field_id.get() != -1, "Primary key is -1");
AssertInfo(insert_record_.empty_pks(), "already exists");
switch (field_meta.get_data_type()) {
case DataType::INT64: {
auto int64_index = dynamic_cast<index::ScalarIndex<int64_t>*>(
scalar_indexings_[field_id].get());
for (int i = 0; i < row_count; ++i) {
insert_record_.insert_pk(int64_index->Reverse_Lookup(i), i);
}
insert_record_.seal_pks();
break;
}
case DataType::VARCHAR: {
auto string_index =
dynamic_cast<index::ScalarIndex<std::string>*>(
scalar_indexings_[field_id].get());
for (int i = 0; i < row_count; ++i) {
insert_record_.insert_pk(string_index->Reverse_Lookup(i),
i);
}
insert_record_.seal_pks();
break;
}
default: {
PanicInfo(DataTypeInvalid,
fmt::format("unsupported primary key type {}",
field_meta.get_data_type()));
}
}
}
set_bit(index_ready_bitset_, field_id, true);
update_row_count(row_count);
// release field column
fields_.erase(field_id);
set_bit(field_data_ready_bitset_, field_id, false);
lck.unlock();
}
void
SegmentSealedImpl::LoadFieldData(const LoadFieldDataInfo& load_info) {
// NOTE: lock only when data is ready to avoid starvation
// only one field for now, parallel load field data in golang
size_t num_rows = storage::GetNumRowsForLoadInfo(load_info);
for (auto& [id, info] : load_info.field_infos) {
AssertInfo(info.row_count > 0, "The row count of field data is 0");
auto field_id = FieldId(id);
auto insert_files = info.insert_files;
auto field_data_info =
FieldDataInfo(field_id.get(), num_rows, load_info.mmap_dir_path);
LOG_SEGCORE_INFO_ << "start to load field data " << id << " of segment "
<< this->id_;
auto parallel_degree = static_cast<uint64_t>(
DEFAULT_FIELD_MAX_MEMORY_LIMIT / FILE_SLICE_SIZE);
field_data_info.channel->set_capacity(parallel_degree * 2);
auto& pool =
ThreadPools::GetThreadPool(milvus::ThreadPoolPriority::MIDDLE);
auto load_future = pool.Submit(
LoadFieldDatasFromRemote, insert_files, field_data_info.channel);
LOG_SEGCORE_INFO_ << "finish submitting LoadFieldDatasFromRemote task "
"to thread pool, "
<< "segmentID:" << this->id_
<< ", fieldID:" << info.field_id;
if (!info.enable_mmap ||
SystemProperty::Instance().IsSystem(field_id)) {
LoadFieldData(field_id, field_data_info);
} else {
MapFieldData(field_id, field_data_info);
}
LOG_SEGCORE_INFO_ << "finish loading segment field, "
<< "segmentID:" << this->id_
<< ", fieldID:" << info.field_id;
}
}
void
SegmentSealedImpl::LoadFieldDataV2(const LoadFieldDataInfo& load_info) {
// NOTE: lock only when data is ready to avoid starvation
// only one field for now, parallel load field data in golang
size_t num_rows = storage::GetNumRowsForLoadInfo(load_info);
for (auto& [id, info] : load_info.field_infos) {
AssertInfo(info.row_count > 0, "The row count of field data is 0");
auto field_id = FieldId(id);
auto insert_files = info.insert_files;
auto field_data_info =
FieldDataInfo(field_id.get(), num_rows, load_info.mmap_dir_path);
auto parallel_degree = static_cast<uint64_t>(
DEFAULT_FIELD_MAX_MEMORY_LIMIT / FILE_SLICE_SIZE);
field_data_info.channel->set_capacity(parallel_degree * 2);
auto& pool =
ThreadPools::GetThreadPool(milvus::ThreadPoolPriority::MIDDLE);
// auto load_future = pool.Submit(
// LoadFieldDatasFromRemote, insert_files, field_data_info.channel);
auto res = milvus_storage::Space::Open(
load_info.url,
milvus_storage::Options{nullptr, load_info.storage_version});
AssertInfo(res.ok(),
fmt::format("init space failed: {}, error: {}",
load_info.url,
res.status().ToString()));
std::shared_ptr<milvus_storage::Space> space = std::move(res.value());
auto load_future = pool.Submit(
LoadFieldDatasFromRemote2, space, schema_, field_data_info);
LOG_SEGCORE_INFO_ << "finish submitting LoadFieldDatasFromRemote task "
"to thread pool, "
<< "segmentID:" << this->id_
<< ", fieldID:" << info.field_id;
if (load_info.mmap_dir_path.empty() ||
SystemProperty::Instance().IsSystem(field_id)) {
LoadFieldData(field_id, field_data_info);
} else {
MapFieldData(field_id, field_data_info);
}
LOG_SEGCORE_INFO_ << "finish loading segment field, "
<< "segmentID:" << this->id_
<< ", fieldID:" << info.field_id;
}
}
void
SegmentSealedImpl::LoadFieldData(FieldId field_id, FieldDataInfo& data) {
auto num_rows = data.row_count;
if (SystemProperty::Instance().IsSystem(field_id)) {
auto system_field_type =
SystemProperty::Instance().GetSystemFieldType(field_id);
if (system_field_type == SystemFieldType::Timestamp) {
std::vector<Timestamp> timestamps(num_rows);
int64_t offset = 0;
auto field_data = CollectFieldDataChannel(data.channel);
for (auto& data : field_data) {
int64_t row_count = data->get_num_rows();
std::copy_n(static_cast<const Timestamp*>(data->Data()),
row_count,
timestamps.data() + offset);
offset += row_count;
}
TimestampIndex index;
auto min_slice_length = num_rows < 4096 ? 1 : 4096;
auto meta = GenerateFakeSlices(
timestamps.data(), num_rows, min_slice_length);
index.set_length_meta(std::move(meta));
// todo ::opt to avoid copy timestamps from field data
index.build_with(timestamps.data(), num_rows);
// use special index
std::unique_lock lck(mutex_);
AssertInfo(insert_record_.timestamps_.empty(), "already exists");
insert_record_.timestamps_.fill_chunk_data(field_data);
insert_record_.timestamp_index_ = std::move(index);
AssertInfo(insert_record_.timestamps_.num_chunk() == 1,
"num chunk not equal to 1 for sealed segment");
} else {
AssertInfo(system_field_type == SystemFieldType::RowId,
"System field type of id column is not RowId");
auto field_data = CollectFieldDataChannel(data.channel);
// write data under lock
std::unique_lock lck(mutex_);
AssertInfo(insert_record_.row_ids_.empty(), "already exists");
insert_record_.row_ids_.fill_chunk_data(field_data);
AssertInfo(insert_record_.row_ids_.num_chunk() == 1,
"num chunk not equal to 1 for sealed segment");
}
++system_ready_count_;
} else {
// prepare data
auto& field_meta = (*schema_)[field_id];
auto data_type = field_meta.get_data_type();
// Don't allow raw data and index exist at the same time
// AssertInfo(!get_bit(index_ready_bitset_, field_id),
// "field data can't be loaded when indexing exists");
std::shared_ptr<ColumnBase> column{};
if (datatype_is_variable(data_type)) {
int64_t field_data_size = 0;
switch (data_type) {
case milvus::DataType::STRING:
case milvus::DataType::VARCHAR: {
auto var_column =
std::make_shared<VariableColumn<std::string>>(
num_rows, field_meta);
storage::FieldDataPtr field_data;
while (data.channel->pop(field_data)) {
for (auto i = 0; i < field_data->get_num_rows(); i++) {
auto str = static_cast<const std::string*>(
field_data->RawValue(i));
auto str_size = str->size();
var_column->Append(str->data(), str_size);
field_data_size += str_size;
}
}
var_column->Seal();
LoadStringSkipIndex(field_id, 0, *var_column);
column = std::move(var_column);
break;
}
case milvus::DataType::JSON: {
auto var_column =
std::make_shared<VariableColumn<milvus::Json>>(
num_rows, field_meta);
storage::FieldDataPtr field_data;
while (data.channel->pop(field_data)) {
for (auto i = 0; i < field_data->get_num_rows(); i++) {
auto padded_string =
static_cast<const milvus::Json*>(
field_data->RawValue(i))
->data();
auto padded_string_size = padded_string.size();
var_column->Append(padded_string.data(),
padded_string_size);
field_data_size += padded_string_size;
}
}
var_column->Seal();
column = std::move(var_column);
break;
}
case milvus::DataType::ARRAY: {
auto var_column =
std::make_shared<ArrayColumn>(num_rows, field_meta);
storage::FieldDataPtr field_data;
while (data.channel->pop(field_data)) {
for (auto i = 0; i < field_data->get_num_rows(); i++) {
auto rawValue = field_data->RawValue(i);
auto array =
static_cast<const milvus::Array*>(rawValue);
var_column->Append(*array);
}
}
var_column->Seal();
column = std::move(var_column);
break;
}
default: {
PanicInfo(DataTypeInvalid,
fmt::format("unsupported data type", data_type));
}
}
// update average row data size
SegmentInternalInterface::set_field_avg_size(
field_id, num_rows, field_data_size);
} else {
column = std::make_shared<Column>(num_rows, field_meta);
storage::FieldDataPtr field_data;
while (data.channel->pop(field_data)) {
column->AppendBatch(field_data);
}
LoadPrimitiveSkipIndex(
field_id, 0, data_type, column->Span().data(), num_rows);
}
AssertInfo(column->NumRows() == num_rows,
fmt::format("data lost while loading column {}: loaded "
"num rows {} but expected {}",
data.field_id,
column->NumRows(),
num_rows));
{
std::unique_lock lck(mutex_);
fields_.emplace(field_id, column);
}
// set pks to offset
if (schema_->get_primary_field_id() == field_id) {
AssertInfo(field_id.get() != -1, "Primary key is -1");
AssertInfo(insert_record_.empty_pks(), "already exists");
insert_record_.insert_pks(data_type, column);
insert_record_.seal_pks();
}
bool use_temp_index = false;
{
// update num_rows to build temperate binlog index
std::unique_lock lck(mutex_);
update_row_count(num_rows);
}
if (generate_binlog_index(field_id)) {
std::unique_lock lck(mutex_);
fields_.erase(field_id);
set_bit(field_data_ready_bitset_, field_id, false);
use_temp_index = true;
}
if (!use_temp_index) {
std::unique_lock lck(mutex_);
set_bit(field_data_ready_bitset_, field_id, true);
}
}
{
std::unique_lock lck(mutex_);
update_row_count(num_rows);
}
}
void
SegmentSealedImpl::MapFieldData(const FieldId field_id, FieldDataInfo& data) {
auto filepath = std::filesystem::path(data.mmap_dir_path) /
std::to_string(get_segment_id()) /
std::to_string(field_id.get());
auto dir = filepath.parent_path();
std::filesystem::create_directories(dir);
auto file = File::Open(filepath.string(), O_CREAT | O_TRUNC | O_RDWR);
auto& field_meta = (*schema_)[field_id];
auto data_type = field_meta.get_data_type();
// write the field data to disk
size_t total_written{0};
auto data_size = 0;
std::vector<uint64_t> indices{};
std::vector<std::vector<uint64_t>> element_indices{};
storage::FieldDataPtr field_data;
while (data.channel->pop(field_data)) {
data_size += field_data->Size();
auto written =
WriteFieldData(file, data_type, field_data, element_indices);
if (written != field_data->Size()) {
break;
}
for (auto i = 0; i < field_data->get_num_rows(); i++) {
auto size = field_data->Size(i);
indices.emplace_back(total_written);
total_written += size;
}
}
AssertInfo(
total_written == data_size,
fmt::format(
"failed to write data file {}, written {} but total {}, err: {}",
filepath.c_str(),
total_written,
data_size,
strerror(errno)));
auto num_rows = data.row_count;
std::shared_ptr<ColumnBase> column{};
if (datatype_is_variable(data_type)) {
switch (data_type) {
case milvus::DataType::STRING:
case milvus::DataType::VARCHAR: {
auto var_column = std::make_shared<VariableColumn<std::string>>(
file, total_written, field_meta);
var_column->Seal(std::move(indices));
column = std::move(var_column);
break;
}
case milvus::DataType::JSON: {
auto var_column =
std::make_shared<VariableColumn<milvus::Json>>(
file, total_written, field_meta);
var_column->Seal(std::move(indices));
column = std::move(var_column);
break;
}
case milvus::DataType::ARRAY: {
auto arr_column = std::make_shared<ArrayColumn>(
file, total_written, field_meta);
arr_column->Seal(std::move(indices),
std::move(element_indices));
column = std::move(arr_column);
break;
}
default: {
PanicInfo(DataTypeInvalid,
fmt::format("unsupported data type {}", data_type));
}
}
} else {
column = std::make_shared<Column>(file, total_written, field_meta);
}
{
std::unique_lock lck(mutex_);
fields_.emplace(field_id, column);
}
auto ok = unlink(filepath.c_str());
AssertInfo(ok == 0,
fmt::format("failed to unlink mmap data file {}, err: {}",
filepath.c_str(),
strerror(errno)));
// set pks to offset
if (schema_->get_primary_field_id() == field_id) {
AssertInfo(field_id.get() != -1, "Primary key is -1");
AssertInfo(insert_record_.empty_pks(), "already exists");
insert_record_.insert_pks(data_type, column);
insert_record_.seal_pks();
}
std::unique_lock lck(mutex_);
set_bit(field_data_ready_bitset_, field_id, true);
}
void
SegmentSealedImpl::LoadDeletedRecord(const LoadDeletedRecordInfo& info) {
AssertInfo(info.row_count > 0, "The row count of deleted record is 0");
AssertInfo(info.primary_keys, "Deleted primary keys is null");
AssertInfo(info.timestamps, "Deleted timestamps is null");
// step 1: get pks and timestamps
auto field_id = schema_->get_primary_field_id().value_or(FieldId(-1));
AssertInfo(field_id.get() != -1, "Primary key is -1");
auto& field_meta = schema_->operator[](field_id);
int64_t size = info.row_count;
std::vector<PkType> pks(size);
ParsePksFromIDs(pks, field_meta.get_data_type(), *info.primary_keys);
auto timestamps = reinterpret_cast<const Timestamp*>(info.timestamps);
// step 2: fill pks and timestamps
deleted_record_.push(pks, timestamps);
}
void
SegmentSealedImpl::AddFieldDataInfoForSealed(
const LoadFieldDataInfo& field_data_info) {
// copy assignment
field_data_info_ = field_data_info;
}
// internal API: support scalar index only
int64_t
SegmentSealedImpl::num_chunk_index(FieldId field_id) const {
auto& field_meta = schema_->operator[](field_id);
if (field_meta.is_vector()) {
return int64_t(vector_indexings_.is_ready(field_id));
}
return scalar_indexings_.count(field_id);
}
int64_t
SegmentSealedImpl::num_chunk_data(FieldId field_id) const {
return get_bit(field_data_ready_bitset_, field_id) ? 1 : 0;
}
int64_t
SegmentSealedImpl::num_chunk() const {
return 1;
}
int64_t
SegmentSealedImpl::size_per_chunk() const {
return get_row_count();
}
SpanBase
SegmentSealedImpl::chunk_data_impl(FieldId field_id, int64_t chunk_id) const {
std::shared_lock lck(mutex_);
AssertInfo(get_bit(field_data_ready_bitset_, field_id),
"Can't get bitset element at " + std::to_string(field_id.get()));
auto& field_meta = schema_->operator[](field_id);
if (auto it = fields_.find(field_id); it != fields_.end()) {
auto& field_data = it->second;
return field_data->Span();
}
auto field_data = insert_record_.get_field_data_base(field_id);
AssertInfo(field_data->num_chunk() == 1,
"num chunk not equal to 1 for sealed segment");
return field_data->get_span_base(0);
}
const index::IndexBase*
SegmentSealedImpl::chunk_index_impl(FieldId field_id, int64_t chunk_id) const {
AssertInfo(scalar_indexings_.find(field_id) != scalar_indexings_.end(),
"Cannot find scalar_indexing with field_id: " +
std::to_string(field_id.get()));
auto ptr = scalar_indexings_.at(field_id).get();
return ptr;
}
int64_t
SegmentSealedImpl::GetMemoryUsageInBytes() const {
// TODO: add estimate for index
std::shared_lock lck(mutex_);
auto row_count = num_rows_.value_or(0);
return schema_->get_total_sizeof() * row_count;
}
int64_t
SegmentSealedImpl::get_row_count() const {
std::shared_lock lck(mutex_);
return num_rows_.value_or(0);
}
int64_t
SegmentSealedImpl::get_deleted_count() const {
std::shared_lock lck(mutex_);
return deleted_record_.size();
}
const Schema&
SegmentSealedImpl::get_schema() const {
return *schema_;
}
void
SegmentSealedImpl::mask_with_delete(BitsetType& bitset,
int64_t ins_barrier,
Timestamp timestamp) const {
auto del_barrier = get_barrier(get_deleted_record(), timestamp);
if (del_barrier == 0) {
return;
}
auto bitmap_holder = get_deleted_bitmap(
del_barrier, ins_barrier, deleted_record_, insert_record_, timestamp);
if (!bitmap_holder || !bitmap_holder->bitmap_ptr) {
return;
}
auto& delete_bitset = *bitmap_holder->bitmap_ptr;
AssertInfo(delete_bitset.size() == bitset.size(),
"Deleted bitmap size not equal to filtered bitmap size");
bitset |= delete_bitset;
}
void
SegmentSealedImpl::vector_search(SearchInfo& search_info,
const void* query_data,
int64_t query_count,
Timestamp timestamp,
const BitsetView& bitset,
SearchResult& output) const {
AssertInfo(is_system_field_ready(), "System field is not ready");
auto field_id = search_info.field_id_;
auto& field_meta = schema_->operator[](field_id);
AssertInfo(field_meta.is_vector(),
"The meta type of vector field is not vector type");
if (get_bit(binlog_index_bitset_, field_id)) {
AssertInfo(
vec_binlog_config_.find(field_id) != vec_binlog_config_.end(),
"The binlog params is not generate.");
auto binlog_search_info =
vec_binlog_config_.at(field_id)->GetSearchConf(search_info);
AssertInfo(vector_indexings_.is_ready(field_id),
"vector indexes isn't ready for field " +
std::to_string(field_id.get()));
query::SearchOnSealedIndex(*schema_,
vector_indexings_,
binlog_search_info,
query_data,
query_count,
bitset,
output);
milvus::tracer::AddEvent(
"finish_searching_vector_temperate_binlog_index");
} else if (get_bit(index_ready_bitset_, field_id)) {
AssertInfo(vector_indexings_.is_ready(field_id),
"vector indexes isn't ready for field " +
std::to_string(field_id.get()));
query::SearchOnSealedIndex(*schema_,
vector_indexings_,
search_info,
query_data,
query_count,
bitset,
output);
milvus::tracer::AddEvent("finish_searching_vector_index");
} else {
AssertInfo(
get_bit(field_data_ready_bitset_, field_id),
"Field Data is not loaded: " + std::to_string(field_id.get()));
AssertInfo(num_rows_.has_value(), "Can't get row count value");
auto row_count = num_rows_.value();
auto vec_data = fields_.at(field_id);
query::SearchOnSealed(*schema_,
vec_data->Data(),
search_info,
query_data,
query_count,
row_count,
bitset,
output);
milvus::tracer::AddEvent("finish_searching_vector_data");
}
}
std::tuple<std::string, int64_t>
SegmentSealedImpl::GetFieldDataPath(FieldId field_id, int64_t offset) const {
auto offset_in_binlog = offset;
auto data_path = std::string();
auto it = field_data_info_.field_infos.find(field_id.get());
AssertInfo(it != field_data_info_.field_infos.end(),
fmt::format("cannot find binlog file for field: {}, seg: {}",
field_id.get(),
id_));
auto field_info = it->second;
for (auto i = 0; i < field_info.insert_files.size(); i++) {
if (offset_in_binlog < field_info.entries_nums[i]) {
data_path = field_info.insert_files[i];
break;
} else {
offset_in_binlog -= field_info.entries_nums[i];
}
}
return {data_path, offset_in_binlog};
}
std::tuple<std::string, std::shared_ptr<ColumnBase>> static ReadFromChunkCache(
const storage::ChunkCachePtr& cc, const std::string& data_path) {
auto column = cc->Read(data_path);
cc->Prefetch(data_path);
return {data_path, column};
}
std::unique_ptr<DataArray>
SegmentSealedImpl::get_vector(FieldId field_id,
const int64_t* ids,
int64_t count) const {
auto& field_meta = schema_->operator[](field_id);
AssertInfo(field_meta.is_vector(), "vector field is not vector type");
if (!get_bit(index_ready_bitset_, field_id) &&
!get_bit(binlog_index_bitset_, field_id)) {
return fill_with_empty(field_id, count);
}
AssertInfo(vector_indexings_.is_ready(field_id),
"vector index is not ready");
auto field_indexing = vector_indexings_.get_field_indexing(field_id);
auto vec_index =
dynamic_cast<index::VectorIndex*>(field_indexing->indexing_.get());
AssertInfo(vec_index, "invalid vector indexing");
auto index_type = vec_index->GetIndexType();
auto metric_type = vec_index->GetMetricType();
auto has_raw_data = vec_index->HasRawData();
if (has_raw_data) {
// If index has raw data, get vector from memory.
auto ids_ds = GenIdsDataset(count, ids);
auto vector = vec_index->GetVector(ids_ds);
return segcore::CreateVectorDataArrayFrom(
vector.data(), count, field_meta);
} else {
// If index doesn't have raw data, get vector from chunk cache.
auto cc = storage::ChunkCacheSingleton::GetInstance().GetChunkCache();
// group by data_path
auto id_to_data_path =
std::unordered_map<std::int64_t,
std::tuple<std::string, int64_t>>{};
auto path_to_column =
std::unordered_map<std::string, std::shared_ptr<ColumnBase>>{};
for (auto i = 0; i < count; i++) {
const auto& tuple = GetFieldDataPath(field_id, ids[i]);
id_to_data_path.emplace(ids[i], tuple);
path_to_column.emplace(std::get<0>(tuple), nullptr);
}
// read and prefetch
auto& pool =
ThreadPools::GetThreadPool(milvus::ThreadPoolPriority::HIGH);
std::vector<
std::future<std::tuple<std::string, std::shared_ptr<ColumnBase>>>>
futures;
futures.reserve(path_to_column.size());
for (const auto& iter : path_to_column) {
const auto& data_path = iter.first;
futures.emplace_back(
pool.Submit(ReadFromChunkCache, cc, data_path));
}
for (int i = 0; i < futures.size(); ++i) {
const auto& [data_path, column] = futures[i].get();
path_to_column[data_path] = column;
}
// assign to data array
auto row_bytes = field_meta.get_sizeof();
auto buf = std::vector<char>(count * row_bytes);
for (auto i = 0; i < count; i++) {
AssertInfo(id_to_data_path.count(ids[i]) != 0, "id not found");
const auto& [data_path, offset_in_binlog] =
id_to_data_path.at(ids[i]);
AssertInfo(path_to_column.count(data_path) != 0,
"column not found");
const auto& column = path_to_column.at(data_path);
AssertInfo(offset_in_binlog * row_bytes < column->ByteSize(),
fmt::format("column idx out of range, idx: {}, size: {}",
offset_in_binlog * row_bytes,
column->ByteSize()));
auto vector = &column->Data()[offset_in_binlog * row_bytes];
std::memcpy(buf.data() + i * row_bytes, vector, row_bytes);
}
return segcore::CreateVectorDataArrayFrom(
buf.data(), count, field_meta);
}
}
void
SegmentSealedImpl::DropFieldData(const FieldId field_id) {
if (SystemProperty::Instance().IsSystem(field_id)) {
auto system_field_type =
SystemProperty::Instance().GetSystemFieldType(field_id);
std::unique_lock lck(mutex_);
--system_ready_count_;
if (system_field_type == SystemFieldType::RowId) {
insert_record_.row_ids_.clear();
} else if (system_field_type == SystemFieldType::Timestamp) {
insert_record_.timestamps_.clear();
}
lck.unlock();
} else {
auto& field_meta = schema_->operator[](field_id);
std::unique_lock lck(mutex_);
if (get_bit(field_data_ready_bitset_, field_id)) {
set_bit(field_data_ready_bitset_, field_id, false);
insert_record_.drop_field_data(field_id);
}
if (get_bit(binlog_index_bitset_, field_id)) {
set_bit(binlog_index_bitset_, field_id, false);
vector_indexings_.drop_field_indexing(field_id);
}
lck.unlock();
}
}
void
SegmentSealedImpl::DropIndex(const FieldId field_id) {
AssertInfo(!SystemProperty::Instance().IsSystem(field_id),
"Field id:" + std::to_string(field_id.get()) +
" isn't one of system type when drop index");
auto& field_meta = schema_->operator[](field_id);
AssertInfo(field_meta.is_vector(),
"Field meta of offset:" + std::to_string(field_id.get()) +
" is not vector type");
std::unique_lock lck(mutex_);
vector_indexings_.drop_field_indexing(field_id);
set_bit(index_ready_bitset_, field_id, false);
}
void
SegmentSealedImpl::check_search(const query::Plan* plan) const {
AssertInfo(plan, "Search plan is null");
AssertInfo(plan->extra_info_opt_.has_value(),
"Extra info of search plan doesn't have value");
if (!is_system_field_ready()) {
PanicInfo(
FieldNotLoaded,
"failed to load row ID or timestamp, potential missing bin logs or "
"empty segments. Segment ID = " +
std::to_string(this->id_));
}
auto& request_fields = plan->extra_info_opt_.value().involved_fields_;
auto field_ready_bitset =
field_data_ready_bitset_ | index_ready_bitset_ | binlog_index_bitset_;
AssertInfo(request_fields.size() == field_ready_bitset.size(),
"Request fields size not equal to field ready bitset size when "
"check search");
auto absent_fields = request_fields - field_ready_bitset;
if (absent_fields.any()) {
auto field_id =
FieldId(absent_fields.find_first() + START_USER_FIELDID);
auto& field_meta = schema_->operator[](field_id);
PanicInfo(
FieldNotLoaded,
"User Field(" + field_meta.get_name().get() + ") is not loaded");
}
}
SegmentSealedImpl::SegmentSealedImpl(SchemaPtr schema,
IndexMetaPtr index_meta,
const SegcoreConfig& segcore_config,
int64_t segment_id)
: segcore_config_(segcore_config),
field_data_ready_bitset_(schema->size()),
index_ready_bitset_(schema->size()),
binlog_index_bitset_(schema->size()),
scalar_indexings_(schema->size()),
insert_record_(*schema, MAX_ROW_COUNT),
schema_(schema),
id_(segment_id),
col_index_meta_(index_meta) {
}
SegmentSealedImpl::~SegmentSealedImpl() {
auto cc = storage::ChunkCacheSingleton::GetInstance().GetChunkCache();
if (cc == nullptr) {
return;
}
// munmap and remove binlog from chunk cache
for (const auto& iter : field_data_info_.field_infos) {
for (const auto& binlog : iter.second.insert_files) {
cc->Remove(binlog);
}
}
}
void
SegmentSealedImpl::bulk_subscript(SystemFieldType system_type,
const int64_t* seg_offsets,
int64_t count,
void* output) const {
AssertInfo(is_system_field_ready(),
"System field isn't ready when do bulk_insert");
switch (system_type) {
case SystemFieldType::Timestamp:
AssertInfo(
insert_record_.timestamps_.num_chunk() == 1,
"num chunk of timestamp not equal to 1 for sealed segment");
bulk_subscript_impl<Timestamp>(
this->insert_record_.timestamps_.get_chunk_data(0),
seg_offsets,
count,
static_cast<Timestamp*>(output));
break;
case SystemFieldType::RowId:
AssertInfo(insert_record_.row_ids_.num_chunk() == 1,
"num chunk of rowID not equal to 1 for sealed segment");
bulk_subscript_impl<int64_t>(
this->insert_record_.row_ids_.get_chunk_data(0),
seg_offsets,
count,
static_cast<int64_t*>(output));
break;
default:
PanicInfo(DataTypeInvalid,
fmt::format("unknown subscript fields", system_type));
}
}
template <typename S, typename T>
void
SegmentSealedImpl::bulk_subscript_impl(const void* src_raw,
const int64_t* seg_offsets,
int64_t count,
T* dst) {
static_assert(IsScalar<T>);
auto src = static_cast<const S*>(src_raw);
for (int64_t i = 0; i < count; ++i) {
auto offset = seg_offsets[i];
dst[i] = src[offset];
}
}
template <typename S, typename T>
void
SegmentSealedImpl::bulk_subscript_impl(const ColumnBase* column,
const int64_t* seg_offsets,
int64_t count,
void* dst_raw) {
auto field = reinterpret_cast<const VariableColumn<S>*>(column);
auto dst = reinterpret_cast<T*>(dst_raw);
for (int64_t i = 0; i < count; ++i) {
auto offset = seg_offsets[i];
dst[i] = std::move(T(field->RawAt(offset)));
}
}
template <typename S, typename T>
void
SegmentSealedImpl::bulk_subscript_ptr_impl(
const ColumnBase* column,
const int64_t* seg_offsets,
int64_t count,
google::protobuf::RepeatedPtrField<T>* dst) {
auto field = reinterpret_cast<const VariableColumn<S>*>(column);
for (int64_t i = 0; i < count; ++i) {
auto offset = seg_offsets[i];
dst->at(i) = std::move(T(field->RawAt(offset)));
}
}
template <typename T>
void
SegmentSealedImpl::bulk_subscript_array_impl(
const ColumnBase* column,
const int64_t* seg_offsets,
int64_t count,
google::protobuf::RepeatedPtrField<T>* dst) {
auto field = reinterpret_cast<const ArrayColumn*>(column);
for (int64_t i = 0; i < count; ++i) {
auto offset = seg_offsets[i];
dst->at(i) = std::move(field->RawAt(offset));
}
}
// for vector
void
SegmentSealedImpl::bulk_subscript_impl(int64_t element_sizeof,
const void* src_raw,
const int64_t* seg_offsets,
int64_t count,
void* dst_raw) {
auto column = reinterpret_cast<const char*>(src_raw);
auto dst_vec = reinterpret_cast<char*>(dst_raw);
for (int64_t i = 0; i < count; ++i) {
auto offset = seg_offsets[i];
auto src = column + element_sizeof * offset;
auto dst = dst_vec + i * element_sizeof;
memcpy(dst, src, element_sizeof);
}
}
std::unique_ptr<DataArray>
SegmentSealedImpl::fill_with_empty(FieldId field_id, int64_t count) const {
auto& field_meta = schema_->operator[](field_id);
if (datatype_is_vector(field_meta.get_data_type())) {
return CreateVectorDataArray(count, field_meta);
}
return CreateScalarDataArray(count, field_meta);
}
std::unique_ptr<DataArray>
SegmentSealedImpl::get_raw_data(FieldId field_id,
const FieldMeta& field_meta,
const int64_t* seg_offsets,
int64_t count) const {
// DO NOT directly access the column by map like: `fields_.at(field_id)->Data()`,
// we have to clone the shared pointer,
// to make sure it won't get released if segment released
auto column = fields_.at(field_id);
auto ret = fill_with_empty(field_id, count);
switch (field_meta.get_data_type()) {
case DataType::VARCHAR:
case DataType::STRING: {
bulk_subscript_ptr_impl<std::string>(
column.get(),
seg_offsets,
count,
ret->mutable_scalars()->mutable_string_data()->mutable_data());
break;
}
case DataType::JSON: {
bulk_subscript_ptr_impl<Json, std::string>(
column.get(),
seg_offsets,
count,
ret->mutable_scalars()->mutable_json_data()->mutable_data());
break;
}
case DataType::ARRAY: {
bulk_subscript_array_impl(
column.get(),
seg_offsets,
count,
ret->mutable_scalars()->mutable_array_data()->mutable_data());
break;
}
case DataType::BOOL: {
bulk_subscript_impl<bool>(column->Data(),
seg_offsets,
count,
ret->mutable_scalars()
->mutable_bool_data()
->mutable_data()
->mutable_data());
break;
}
case DataType::INT8: {
bulk_subscript_impl<int8_t>(column->Data(),
seg_offsets,
count,
ret->mutable_scalars()
->mutable_int_data()
->mutable_data()
->mutable_data());
break;
}
case DataType::INT16: {
bulk_subscript_impl<int16_t>(column->Data(),
seg_offsets,
count,
ret->mutable_scalars()
->mutable_int_data()
->mutable_data()
->mutable_data());
break;
}
case DataType::INT32: {
bulk_subscript_impl<int32_t>(column->Data(),
seg_offsets,
count,
ret->mutable_scalars()
->mutable_int_data()
->mutable_data()
->mutable_data());
break;
}
case DataType::INT64: {
bulk_subscript_impl<int64_t>(column->Data(),
seg_offsets,
count,
ret->mutable_scalars()
->mutable_long_data()
->mutable_data()
->mutable_data());
break;
}
case DataType::FLOAT: {
bulk_subscript_impl<float>(column->Data(),
seg_offsets,
count,
ret->mutable_scalars()
->mutable_float_data()
->mutable_data()
->mutable_data());
break;
}
case DataType::DOUBLE: {
bulk_subscript_impl<double>(column->Data(),
seg_offsets,
count,
ret->mutable_scalars()
->mutable_double_data()
->mutable_data()
->mutable_data());
break;
}
case DataType::VECTOR_FLOAT: {
bulk_subscript_impl(field_meta.get_sizeof(),
column->Data(),
seg_offsets,
count,
ret->mutable_vectors()
->mutable_float_vector()
->mutable_data()
->mutable_data());
break;
}
case DataType::VECTOR_FLOAT16: {
bulk_subscript_impl(
field_meta.get_sizeof(),
column->Data(),
seg_offsets,
count,
ret->mutable_vectors()->mutable_float16_vector()->data());
break;
}
case DataType::VECTOR_BINARY: {
bulk_subscript_impl(
field_meta.get_sizeof(),
column->Data(),
seg_offsets,
count,
ret->mutable_vectors()->mutable_binary_vector()->data());
break;
}
default: {
PanicInfo(DataTypeInvalid,
fmt::format("unsupported data type {}",
field_meta.get_data_type()));
}
}
return ret;
}
std::unique_ptr<DataArray>
SegmentSealedImpl::bulk_subscript(FieldId field_id,
const int64_t* seg_offsets,
int64_t count) const {
auto& field_meta = schema_->operator[](field_id);
// if count == 0, return empty data array
if (count == 0) {
return fill_with_empty(field_id, count);
}
if (HasIndex(field_id)) {
// if field has load scalar index, reverse raw data from index
if (!datatype_is_vector(field_meta.get_data_type())) {
AssertInfo(num_chunk() == 1,
"num chunk not equal to 1 for sealed segment");
auto index = chunk_index_impl(field_id, 0);
if (index->HasRawData()) {
return ReverseDataFromIndex(
index, seg_offsets, count, field_meta);
}
return get_raw_data(field_id, field_meta, seg_offsets, count);
}
return get_vector(field_id, seg_offsets, count);
}
Assert(get_bit(field_data_ready_bitset_, field_id));
return get_raw_data(field_id, field_meta, seg_offsets, count);
}
bool
SegmentSealedImpl::HasIndex(FieldId field_id) const {
std::shared_lock lck(mutex_);
return get_bit(index_ready_bitset_, field_id) |
get_bit(binlog_index_bitset_, field_id);
}
bool
SegmentSealedImpl::HasFieldData(FieldId field_id) const {
std::shared_lock lck(mutex_);
if (SystemProperty::Instance().IsSystem(field_id)) {
return is_system_field_ready();
} else {
return get_bit(field_data_ready_bitset_, field_id);
}
}
bool
SegmentSealedImpl::HasRawData(int64_t field_id) const {
std::shared_lock lck(mutex_);
auto fieldID = FieldId(field_id);
const auto& field_meta = schema_->operator[](fieldID);
if (datatype_is_vector(field_meta.get_data_type())) {
if (get_bit(index_ready_bitset_, fieldID) |
get_bit(binlog_index_bitset_, fieldID)) {
AssertInfo(vector_indexings_.is_ready(fieldID),
"vector index is not ready");
auto field_indexing = vector_indexings_.get_field_indexing(fieldID);
auto vec_index = dynamic_cast<index::VectorIndex*>(
field_indexing->indexing_.get());
return vec_index->HasRawData();
}
} else {
auto scalar_index = scalar_indexings_.find(fieldID);
if (scalar_index != scalar_indexings_.end()) {
return scalar_index->second->HasRawData();
}
}
return true;
}
std::pair<std::unique_ptr<IdArray>, std::vector<SegOffset>>
SegmentSealedImpl::search_ids(const IdArray& id_array,
Timestamp timestamp) const {
auto field_id = schema_->get_primary_field_id().value_or(FieldId(-1));
AssertInfo(field_id.get() != -1, "Primary key is -1");
auto& field_meta = schema_->operator[](field_id);
auto data_type = field_meta.get_data_type();
auto ids_size = GetSizeOfIdArray(id_array);
std::vector<PkType> pks(ids_size);
ParsePksFromIDs(pks, data_type, id_array);
auto res_id_arr = std::make_unique<IdArray>();
std::vector<SegOffset> res_offsets;
res_offsets.reserve(pks.size());
for (auto& pk : pks) {
auto segOffsets = insert_record_.search_pk(pk, timestamp);
for (auto offset : segOffsets) {
switch (data_type) {
case DataType::INT64: {
res_id_arr->mutable_int_id()->add_data(
std::get<int64_t>(pk));
break;
}
case DataType::VARCHAR: {
res_id_arr->mutable_str_id()->add_data(
std::get<std::string>(std::move(pk)));
break;
}
default: {
PanicInfo(DataTypeInvalid,
fmt::format("unsupported type {}", data_type));
}
}
res_offsets.push_back(offset);
}
}
return {std::move(res_id_arr), std::move(res_offsets)};
}
SegcoreError
SegmentSealedImpl::Delete(int64_t reserved_offset, // deprecated
int64_t size,
const IdArray* ids,
const Timestamp* timestamps_raw) {
auto field_id = schema_->get_primary_field_id().value_or(FieldId(-1));
AssertInfo(field_id.get() != -1, "Primary key is -1");
auto& field_meta = schema_->operator[](field_id);
std::vector<PkType> pks(size);
ParsePksFromIDs(pks, field_meta.get_data_type(), *ids);
// filter out the deletions that the primary key not exists
std::vector<std::tuple<Timestamp, PkType>> ordering(size);
for (int i = 0; i < size; i++) {
ordering[i] = std::make_tuple(timestamps_raw[i], pks[i]);
}
auto end =
std::remove_if(ordering.begin(),
ordering.end(),
[&](const std::tuple<Timestamp, PkType>& record) {
return !insert_record_.contain(std::get<1>(record));
});
size = end - ordering.begin();
ordering.resize(size);
if (size == 0) {
return SegcoreError::success();
}
// step 1: sort timestamp
std::sort(ordering.begin(), ordering.end());
std::vector<PkType> sort_pks(size);
std::vector<Timestamp> sort_timestamps(size);
for (int i = 0; i < size; i++) {
auto [t, pk] = ordering[i];
sort_timestamps[i] = t;
sort_pks[i] = pk;
}
deleted_record_.push(sort_pks, sort_timestamps.data());
return SegcoreError::success();
}
std::string
SegmentSealedImpl::debug() const {
std::string log_str;
log_str += "Sealed\n";
log_str += "\n";
return log_str;
}
void
SegmentSealedImpl::LoadSegmentMeta(
const proto::segcore::LoadSegmentMeta& segment_meta) {
std::unique_lock lck(mutex_);
std::vector<int64_t> slice_lengths;
for (auto& info : segment_meta.metas()) {
slice_lengths.push_back(info.row_count());
}
insert_record_.timestamp_index_.set_length_meta(std::move(slice_lengths));
PanicInfo(NotImplemented, "unimplemented");
}
int64_t
SegmentSealedImpl::get_active_count(Timestamp ts) const {
// TODO optimize here to reduce expr search range
return this->get_row_count();
}
void
SegmentSealedImpl::mask_with_timestamps(BitsetType& bitset_chunk,
Timestamp timestamp) const {
// TODO change the
AssertInfo(insert_record_.timestamps_.num_chunk() == 1,
"num chunk not equal to 1 for sealed segment");
const auto& timestamps_data = insert_record_.timestamps_.get_chunk(0);
AssertInfo(timestamps_data.size() == get_row_count(),
fmt::format("Timestamp size not equal to row count: {}, {}",
timestamps_data.size(),
get_row_count()));
auto range = insert_record_.timestamp_index_.get_active_range(timestamp);
// range == (size_, size_) and size_ is this->timestamps_.size().
// it means these data are all useful, we don't need to update bitset_chunk.
// It can be thought of as an OR operation with another bitmask that is all 0s, but it is not necessary to do so.
if (range.first == range.second && range.first == timestamps_data.size()) {
// just skip
return;
}
// range == (0, 0). it means these data can not be used, directly set bitset_chunk to all 1s.
// It can be thought of as an OR operation with another bitmask that is all 1s.
if (range.first == range.second && range.first == 0) {
bitset_chunk.set();
return;
}
auto mask = TimestampIndex::GenerateBitset(
timestamp, range, timestamps_data.data(), timestamps_data.size());
bitset_chunk |= mask;
}
bool
SegmentSealedImpl::generate_binlog_index(const FieldId field_id) {
if (col_index_meta_ == nullptr)
return false;
auto& field_meta = schema_->operator[](field_id);
if (field_meta.is_vector() &&
field_meta.get_data_type() == DataType::VECTOR_FLOAT &&
segcore_config_.get_enable_interim_segment_index()) {
try {
auto& field_index_meta =
col_index_meta_->GetFieldIndexMeta(field_id);
auto& index_params = field_index_meta.GetIndexParams();
if (index_params.find(knowhere::meta::INDEX_TYPE) ==
index_params.end() ||
index_params.at(knowhere::meta::INDEX_TYPE) ==
knowhere::IndexEnum::INDEX_FAISS_IDMAP) {
return false;
}
// get binlog data and meta
auto row_count = num_rows_.value();
auto dim = field_meta.get_dim();
auto vec_data = fields_.at(field_id);
auto dataset =
knowhere::GenDataSet(row_count, dim, (void*)vec_data->Data());
dataset->SetIsOwner(false);
// generate index params
auto field_binlog_config = std::unique_ptr<VecIndexConfig>(
new VecIndexConfig(row_count,
field_index_meta,
segcore_config_,
SegmentType::Sealed));
auto build_config = field_binlog_config->GetBuildBaseParams();
build_config[knowhere::meta::DIM] = std::to_string(dim);
build_config[knowhere::meta::NUM_BUILD_THREAD] = std::to_string(1);
auto index_metric = field_binlog_config->GetMetricType();
index::IndexBasePtr vec_index =
std::make_unique<index::VectorMemIndex<float>>(
field_binlog_config->GetIndexType(),
index_metric,
knowhere::Version::GetCurrentVersion().VersionNumber());
vec_index->BuildWithDataset(dataset, build_config);
vector_indexings_.append_field_indexing(
field_id, index_metric, std::move(vec_index));
vec_binlog_config_[field_id] = std::move(field_binlog_config);
set_bit(binlog_index_bitset_, field_id, true);
return true;
} catch (std::exception& e) {
return false;
}
} else {
return false;
}
}
} // namespace milvus::segcore