ClickHouse/dbms/src/Storages/StorageMergeTree.h
2018-02-22 13:03:35 +03:00

155 lines
5.6 KiB
C++

#pragma once
#include <ext/shared_ptr_helper.h>
#include <Storages/MergeTree/MergeTreeData.h>
#include <Storages/MergeTree/MergeTreeDataSelectExecutor.h>
#include <Storages/MergeTree/MergeTreeDataWriter.h>
#include <Storages/MergeTree/MergeTreeDataMerger.h>
#include <Storages/MergeTree/DiskSpaceMonitor.h>
#include <Storages/MergeTree/BackgroundProcessingPool.h>
#include <Common/SimpleIncrement.h>
namespace DB
{
/** See the description of the data structure in MergeTreeData.
*/
class StorageMergeTree : public ext::shared_ptr_helper<StorageMergeTree>, public IStorage
{
friend class MergeTreeBlockOutputStream;
public:
void startup() override;
void shutdown() override;
~StorageMergeTree() override;
std::string getName() const override
{
return data.merging_params.getModeName() + "MergeTree";
}
std::string getTableName() const override { return table_name; }
bool supportsSampling() const override { return data.supportsSampling(); }
bool supportsFinal() const override { return data.supportsFinal(); }
bool supportsPrewhere() const override { return data.supportsPrewhere(); }
const NamesAndTypesList & getColumnsListImpl() const override { return data.getColumnsListNonMaterialized(); }
NameAndTypePair getColumn(const String & column_name) const override
{
return data.getColumn(column_name);
}
bool hasColumn(const String & column_name) const override
{
return data.hasColumn(column_name);
}
BlockInputStreams read(
const Names & column_names,
const SelectQueryInfo & query_info,
const Context & context,
QueryProcessingStage::Enum & processed_stage,
size_t max_block_size,
unsigned num_streams) override;
BlockOutputStreamPtr write(const ASTPtr & query, const Settings & settings) override;
/** Perform the next step in combining the parts.
*/
bool optimize(const ASTPtr & query, const ASTPtr & partition, bool final, bool deduplicate, const Context & context) override;
void dropPartition(const ASTPtr & query, const ASTPtr & partition, bool detach, const Context & context) override;
void clearColumnInPartition(const ASTPtr & partition, const Field & column_name, const Context & context) override;
void attachPartition(const ASTPtr & partition, bool part, const Context & context) override;
void freezePartition(const ASTPtr & partition, const String & with_name, const Context & context) override;
void drop() override;
void rename(const String & new_path_to_db, const String & new_database_name, const String & new_table_name) override;
void alter(const AlterCommands & params, const String & database_name, const String & table_name, const Context & context) override;
bool supportsIndexForIn() const override { return true; }
bool mayBenefitFromIndexForIn(const ASTPtr & left_in_operand) const override { return data.mayBenefitFromIndexForIn(left_in_operand); }
bool checkTableCanBeDropped() const override;
MergeTreeData & getData() { return data; }
const MergeTreeData & getData() const { return data; }
String getDataPath() const override { return full_path; }
private:
String path;
String database_name;
String table_name;
String full_path;
Context & context;
BackgroundProcessingPool & background_pool;
MergeTreeData data;
MergeTreeDataSelectExecutor reader;
MergeTreeDataWriter writer;
MergeTreeDataMerger merger;
/// For block numbers.
SimpleIncrement increment{0};
/// For clearOldParts, clearOldTemporaryDirectories.
StopwatchWithLock time_after_previous_cleanup;
MergeTreeData::DataParts currently_merging;
std::mutex currently_merging_mutex;
Logger * log;
std::atomic<bool> shutdown_called {false};
BackgroundProcessingPool::TaskHandle merge_task_handle;
friend struct CurrentlyMergingPartsTagger;
/** Determines what parts should be merged and merges it.
* If aggressive - when selects parts don't takes into account their ratio size and novelty (used for OPTIMIZE query).
* Returns true if merge is finished successfully.
*/
bool merge(size_t aio_threshold, bool aggressive, const String & partition_id, bool final, bool deduplicate,
String * out_disable_reason = nullptr);
bool mergeTask();
protected:
/** Attach the table with the appropriate name, along the appropriate path (with / at the end),
* (correctness of names and paths are not checked)
* consisting of the specified columns.
*
* primary_expr_ast - expression for sorting;
* date_column_name - if not empty, the name of the column with the date used for partitioning by month;
otherwise, partition_expr_ast is used as the partitioning expression;
*/
StorageMergeTree(
const String & path_,
const String & database_name_,
const String & table_name_,
const NamesAndTypesList & columns_,
const NamesAndTypesList & materialized_columns_,
const NamesAndTypesList & alias_columns_,
const ColumnDefaults & column_defaults_,
bool attach,
Context & context_,
const ASTPtr & primary_expr_ast_,
const ASTPtr & secondary_sorting_expr_list_,
const String & date_column_name,
const ASTPtr & partition_expr_ast_,
const ASTPtr & sampling_expression_, /// nullptr, if sampling is not supported.
const MergeTreeData::MergingParams & merging_params_,
const MergeTreeSettings & settings_,
bool has_force_restore_data_flag);
};
}