DuckDB 是近年来颇受关注的OLAP数据库，号称是OLAP领域的SQLite，以精巧简单，性能优异而著称。笔者前段时间在调研Doris的Pipeline的算子并行方案，而DuckDB基于论文《Morsel-Driven Parallelism: A NUMA-Aware Query Evaluation Framework for the Many-Core Age》实现SQL算子的高效并行化的Pipeline执行引擎，所以笔者花了一些时间进行了学习和总结，这里结合了Mark Raasveldt进行的分享和原始代码来一一剖析DuckDB在执行算子并行上的具体实现。

1. 基础知识

问题1：并行task的数目由什么决定 ？

Pipeline的核心是：Morsel-Driven，数据是拆分成了小部分的数据。所以并行Task的核心是：能够利用多线程来处理数据，每一个数据拆分为小部分，所以拆分并行的数目由Source决定。

DuckDB在GlobalSource上实现了一个虚函数MaxThread来决定task数目：

每一个算子的GlobalSource抽象了自己的并行度：

问题2：并行task的怎么样进行多线程同步：

  • 多线程的竞争只会发生在SinkOperator上，也就是Pipeline的尾端。
  • parallelism-aware的算法需要实现在Sink端
  • 其他的非Sink operators (比如：Hash Join Probe, Projection, Filter等)， 不需要感知多线程同步的问题

问题3：DuckDB的是如何抽象接口的：

Sink的Opeartor 定义了两种类型：GlobalState, LocalState

  1. GlobalState: 每个查询的Operator全局只有一个GlobalSinkState，记录全局部分的信息
class PhysicalOperator {public:unique_ptr sink_state;
  1. LocalState: 每个查询的PipelineExecutor都有一个LocalSinkState，都是局部私有
//! The Pipeline class represents an execution pipelineclass PipelineExecutor {private://! The local sink state (if any)unique_ptr local_sink_state;

后续会详细解析不同的sink之间的LocalState和GlobalState如何配合的，核心部分如下：

Sink ：处理LocalState的数据

Combine：合并LocalState到GlobalState之中

2. 核心算子的并行

这部分进行各个算子的源码剖析，笔者在源码的关键部分加上了中文注释，以方便大家的理解

Sort算子

  • Sink接口：这里需要注意的是DuckDB排序是进行了列转行的工作的，后续读取时需要行转列。Sink这部分相当于实现了部分数据的排序工作。
SinkResultType PhysicalOrder::Sink(ExecutionContext &context, GlobalSinkState &gstate_p, LocalSinkState &lstate_p,                                   DataChunk &input) const {auto &lstate = (OrderLocalSinkState &)lstate_p;              // keys 是排序的列block，payload是输出的排序后数据，这里调用LocalState的SinkChunk，进行数据的转行，local_sort_state.SinkChunk(keys, payload);// 数据达到内存阈值的时候进行基数排序处理，排序之后的结果存入LocalState的本地的SortedBlock中if (local_sort_state.SizeInBytes() >= gstate.memory_per_thread) {local_sort_state.Sort(global_sort_state, true);}return SinkResultType::NEED_MORE_INPUT;}
  • Combine接口： 加锁，拷贝sorted block到Global State
void PhysicalOrder::Combine(ExecutionContext &context, GlobalSinkState &gstate_p, LocalSinkState &lstate_p) const {auto &gstate = (OrderGlobalSinkState &)gstate_p;auto &lstate = (OrderLocalSinkState &)lstate_p;        // 排序剩余内存中不满的数据local_sort_state.Sort(*this, external || !local_sort_state.sorted_blocks.empty());// Append local state sorted data to this global statelock_guard append_guard(lock);for (auto &sb : local_sort_state.sorted_blocks) {sorted_blocks.push_back(move(sb));}}
  • MergeTask：启动核数相同的task来进行Merge (这里可以看出DuckDB对于多线程的使用是很激进的)， 这里是通过Event的机制实现的
void Schedule() override {auto &context = pipeline->GetClientContext();idx_t num_threads = ts.NumberOfThreads();vector<unique_ptr> merge_tasks;for (idx_t tnum = 0; tnum < num_threads; tnum++) {merge_tasks.push_back(make_unique(shared_from_this(), context, gstate));}SetTasks(move(merge_tasks));}class PhysicalOrderMergeTask : public ExecutorTask {public:TaskExecutionResult ExecuteTask(TaskExecutionMode mode) override {// Initialize merge sorted and iterate until doneauto &global_sort_state = state.global_sort_state;MergeSorter merge_sorter(global_sort_state, BufferManager::GetBufferManager(context));        // 加锁，获取两路，不断进行两路归并，最终完成全局排序。while (true) {{lock_guard pair_guard(state.lock);if (state.pair_idx == state.num_pairs) {break;}GetNextPartition();}MergePartition();}event->FinishTask();return TaskExecutionResult::TASK_FINISHED;}

聚合算子(这里分析的是Prefetch Agg Operator算子)

  • Sink接口：和Sort算子一样，这里拆分为Group Chunk和Aggregate Input Chunk，可以理解为代表聚合时的key与value列。注意此时Sink接口上的聚合是在LocalSinkState上完成的。
SinkResultType PhysicalPerfectHashAggregate::Sink(ExecutionContext &context, GlobalSinkState &state,                                                  LocalSinkState &lstate_p, DataChunk &input) const {lstate.ht->AddChunk(group_chunk, aggregate_input_chunk);}void PerfectAggregateHashTable::AddChunk(DataChunk &groups, DataChunk &payload) {auto address_data = FlatVector::GetData(addresses);memset(address_data, 0, groups.size() * sizeof(uintptr_t));D_ASSERT(groups.ColumnCount() == group_minima.size());// 计算group key列对应的entry的位置idx_t current_shift = total_required_bits;for (idx_t i = 0; i < groups.ColumnCount(); i++) {current_shift -= required_bits[i];ComputeGroupLocation(groups.data[i], group_minima[i], address_data, current_shift, groups.size());}// 通过data加上面的entry位置 + tuple的偏移量，计算出对应的内存地址，并进行initidx_t needs_init = 0;for (idx_t i = 0; i < groups.size(); i++) {D_ASSERT(address_data[i] < total_groups);const auto group = address_data[i];address_data[i] = uintptr_t(data) + address_data[i] * tuple_size;}RowOperations::InitializeStates(layout, addresses, sel, needs_init);// after finding the group location we update the aggregatesidx_t payload_idx = 0;auto &aggregates = layout.GetAggregates();for (idx_t aggr_idx = 0; aggr_idx < aggregates.size(); aggr_idx++) {auto &aggregate = aggregates[aggr_idx];auto input_count = (idx_t)aggregate.child_count;                // 进行聚合的Update操作RowOperations::UpdateStates(aggregate, addresses, payload, payload_idx, payload.size());}}
  • Combine接口： 加锁，merge local hash table 与 global hash table
void PhysicalPerfectHashAggregate::Combine(ExecutionContext &context, GlobalSinkState &gstate_p,                                           LocalSinkState &lstate_p) const {auto &lstate = (PerfectHashAggregateLocalState &)lstate_p;auto &gstate = (PerfectHashAggregateGlobalState &)gstate_p;lock_guard l(gstate.lock);gstate.ht->Combine(*lstate.ht);}
        // local state的地址vectorVector source_addresses(LogicalType::POINTER);       // global state的地址vectorVector target_addresses(LogicalType::POINTER);auto source_addresses_ptr = FlatVector::GetData(source_addresses);auto target_addresses_ptr = FlatVector::GetData(target_addresses);// 遍历所有hash table的表，然后进行合并对应能够合并的keydata_ptr_t source_ptr = other.data;data_ptr_t target_ptr = data;idx_t combine_count = 0;idx_t reinit_count = 0;const auto &reinit_sel = *FlatVector::IncrementalSelectionVector();for (idx_t i = 0; i < total_groups; i++) {auto has_entry_source = other.group_is_set[i];// we only have any work to do if the source has an entry for this groupif (has_entry_source) {auto has_entry_target = group_is_set[i];if (has_entry_target) {// both source and target have an entry: need to combinesource_addresses_ptr[combine_count] = source_ptr;target_addresses_ptr[combine_count] = target_ptr;combine_count++;if (combine_count == STANDARD_VECTOR_SIZE) {RowOperations::CombineStates(layout, source_addresses, target_addresses, combine_count);combine_count = 0;}} else {group_is_set[i] = true;// only source has an entry for this group: we can just memcpy it overmemcpy(target_ptr, source_ptr, tuple_size);// we clear this entry in the other HT as we "consume" the entry hereother.group_is_set[i] = false;}}source_ptr += tuple_size;target_ptr += tuple_size;}        // 做对应的merge操作RowOperations::CombineStates(layout, source_addresses, target_addresses, combine_count);

Join算子

  • Sink接口：和Sort算子一样，注意此时Sink接口上的hash 表是在LocalSinkState上完成的。
SinkResultType PhysicalHashJoin::Sink(ExecutionContext &context, GlobalSinkState &gstate_p, LocalSinkState &lstate_p,                                      DataChunk &input) const {auto &gstate = (HashJoinGlobalSinkState &)gstate_p;auto &lstate = (HashJoinLocalSinkState &)lstate_p;lstate.join_keys.Reset();lstate.build_executor.Execute(input, lstate.join_keys);// build the HTauto &ht = *lstate.hash_table;if (!right_projection_map.empty()) {// there is a projection map: fill the build chunk with the projected columnslstate.build_chunk.Reset();lstate.build_chunk.SetCardinality(input);for (idx_t i = 0; i < right_projection_map.size(); i++) {lstate.build_chunk.data[i].Reference(input.data[right_projection_map[i]]);}                // 构建local state的hash 表ht.Build(lstate.join_keys, lstate.build_chunk)return SinkResultType::NEED_MORE_INPUT;}
  • Combine接口： 加锁，拷贝local state的hash表到global state
void PhysicalHashJoin::Combine(ExecutionContext &context, GlobalSinkState &gstate_p, LocalSinkState &lstate_p) const {auto &gstate = (HashJoinGlobalSinkState &)gstate_p;auto &lstate = (HashJoinLocalSinkState &)lstate_p;if (lstate.hash_table) {lock_guard local_ht_lock(gstate.lock);gstate.local_hash_tables.push_back(move(lstate.hash_table));}}
  • MergeTask：启动核数相同的task来进行Hash table的Merge (这里可以看出DuckDB对于多线程的使用是很激进的)， 每个任务merge一部分Block(DuckDB之中的行数据，落盘使用）
void Schedule() override {auto &context = pipeline->GetClientContext();vector<unique_ptr> finalize_tasks;auto &ht = *sink.hash_table;const auto &block_collection = ht.GetBlockCollection();const auto &blocks = block_collection.blocks;const auto num_blocks = blocks.size();if (block_collection.count < PARALLEL_CONSTRUCT_THRESHOLD && !context.config.verify_parallelism) {// Single-threaded finalizefinalize_tasks.push_back(    make_unique(shared_from_this(), context, sink, 0, num_blocks, false));} else {// Parallel finalizeidx_t num_threads = TaskScheduler::GetScheduler(context).NumberOfThreads();auto blocks_per_thread = MaxValue((num_blocks + num_threads - 1) / num_threads, 1);idx_t block_idx = 0;for (idx_t thread_idx = 0; thread_idx < num_threads; thread_idx++) {auto block_idx_start = block_idx;auto block_idx_end = MinValue(block_idx_start + blocks_per_thread, num_blocks);finalize_tasks.push_back(make_unique(shared_from_this(), context, sink,                                                           block_idx_start, block_idx_end, true));block_idx = block_idx_end;if (block_idx == num_blocks) {break;}}}SetTasks(move(finalize_tasks));}template static inline void InsertHashesLoop(atomic pointers[], const hash_t indices[], const idx_t count,                                    const data_ptr_t key_locations[], const idx_t pointer_offset) {for (idx_t i = 0; i < count; i++) {auto index = indices[i];if (PARALLEL) {data_ptr_t head;do {head = pointers[index];Store(head, key_locations[i] + pointer_offset);} while (!std::atomic_compare_exchange_weak(&pointers[index], &head, key_locations[i]));} else {// set prev in current key to the value (NOTE: this will be nullptr if there is none)Store(pointers[index], key_locations[i] + pointer_offset);// set pointer to current tuplepointers[index] = key_locations[i];}}}
  • 并行扫描hash表，进行outer数据的处理：
void PhysicalHashJoin::GetData(ExecutionContext &context, DataChunk &chunk, GlobalSourceState &gstate_p,                               LocalSourceState &lstate_p) const {auto &sink = (HashJoinGlobalSinkState &)*sink_state;auto &gstate = (HashJoinGlobalSourceState &)gstate_p;auto &lstate = (HashJoinLocalSourceState &)lstate_p;sink.scanned_data = true;if (!sink.external) {if (IsRightOuterJoin(join_type)) {{lock_guard guard(gstate.lock);                                // 拆解扫描部分hash表的数据lstate.ScanFullOuter(sink, gstate);}                        // 扫描hash表读取数据sink.hash_table->GatherFullOuter(chunk, lstate.addresses, lstate.full_outer_found_entries);}return;}}void HashJoinLocalSourceState::ScanFullOuter(HashJoinGlobalSinkState &sink, HashJoinGlobalSourceState &gstate) {auto &fo_ss = gstate.full_outer_scan;idx_t scan_index_before = fo_ss.scan_index;full_outer_found_entries = sink.hash_table->ScanFullOuter(fo_ss, addresses);idx_t scanned = fo_ss.scan_index - scan_index_before;full_outer_in_progress = scanned;}

小结

  • DuckDB在多线程同步，核心就是在Combine的时候：加锁，并发是通过原子变量的方式实现并发写入hash表的操作
  • 通过local/global 拆分私有内存和公共内存，并发的基础是在私有内存上进行运算，同步的部分主要在公有内存的更新

3. Spill To Disk的实现

DuckDB并没有如笔者预期的实现异步IO， 所以任意的执行线程是有可能Stall在系统的I/O调度上的，我想大概率是DuckDB本身的定位对于高并发场景的支持不是那么敏感所导致的。这里他们也作为了后续TODO的计划之一。

4. 参考资料

DuckDB源码

Push-Based Execution in DuckDB