Merge

dbms/include/DB/Functions/FunctionsStringArray.h

@@ -238,7 +238,7 @@ public:
 class ExtractAllImpl
 {
 private:
-	const OptimizedRegularExpression * re = nullptr;
+	Regexps::Pointer re;
 	OptimizedRegularExpression::MatchVec matches;
 	size_t capture;
 
@@ -263,7 +263,7 @@ public:
 			throw Exception("Illegal column " + col->getName() + " of first argument of function " + getName() + ". Must be constant string.",
 				ErrorCodes::ILLEGAL_COLUMN);
 
-		re = &Regexps::get(col->getData());
+		re = Regexps::get(col->getData());
 		capture = re->getNumberOfSubpatterns() > 0 ? 1 : 0;
 
 		matches.resize(capture + 1);

dbms/include/DB/Functions/FunctionsStringSearch.h

@@ -3,6 +3,7 @@
 #include <Poco/Mutex.h>
 
 #include <statdaemons/OptimizedRegularExpression.h>
+#include <statdaemons/stdext.h>
 
 #include <DB/DataTypes/DataTypesNumberFixed.h>
 #include <DB/DataTypes/DataTypeString.h>
@@ -13,6 +14,9 @@
 #include <re2/re2.h>
 #include <re2/stringpiece.h>
 
+#include <mutex>
+#include <stack>
+
 namespace DB
 {
 
@@ -240,38 +244,80 @@ inline bool likePatternIsStrstr(const String & pattern, String & res)
 }
 
 
-struct Regexps
+namespace Regexps
 {
-	typedef std::map<String, OptimizedRegularExpression> KnownRegexps;
+	struct Holder;
+	struct Deleter;
 
-	static const OptimizedRegularExpression & get(const std::string & pattern)
+	using Regexp = OptimizedRegularExpressionImpl<false>;
+	using KnownRegexps = std::map<String, std::unique_ptr<Holder>>;
+	using Pointer = std::unique_ptr<Regexp, Deleter>;
+
+	///	Container for regular expressions with embedded mutex for safe addition and removal
+	struct Holder
 	{
-		/// В GCC thread safe statics.
-		static KnownRegexps known_regexps;
-		static Poco::FastMutex mutex;
-		Poco::ScopedLock<Poco::FastMutex> lock(mutex);
+		std::mutex mutex;
+		std::stack<std::unique_ptr<Regexp>> stack;
 
-		KnownRegexps::const_iterator it = known_regexps.find(pattern);
-		if (known_regexps.end() == it)
-			it = known_regexps.insert(std::make_pair(pattern, OptimizedRegularExpression(pattern))).first;
+		/**	Extracts and returns a pointer from the collection if it's not empty,
+		*	creates a new one by calling provided f() otherwise.
+		*/
+		template <typename Factory> Pointer get(Factory && f);
+	};
 
-		return it->second;
+	/**	Specialized deleter for std::unique_ptr.
+	*	Returns underlying pointer back to holder thus reclaiming its ownership.
+	*/
+	struct Deleter
+	{
+		Holder * holder;
+
+		Deleter(Holder * holder = nullptr) : holder{holder} {}
+
+		void operator()(Regexp * owning_ptr) const
+		{
+			std::lock_guard<std::mutex> lock{holder->mutex};
+			holder->stack.emplace(owning_ptr);
 		}
 
-	static const OptimizedRegularExpression & getLike(const std::string & pattern)
+	};
+
+	template <typename Factory>
+	inline Pointer Holder::get(Factory && f)
 	{
-		/// В GCC thread safe statics.
-		static KnownRegexps known_regexps;
-		static Poco::FastMutex mutex;
-		Poco::ScopedLock<Poco::FastMutex> lock(mutex);
+		std::lock_guard<std::mutex> lock{mutex};
 
-		KnownRegexps::const_iterator it = known_regexps.find(pattern);
-		if (known_regexps.end() == it)
- 			it = known_regexps.insert(std::make_pair(pattern, OptimizedRegularExpression(likePatternToRegexp(pattern)))).first;
+		if (stack.empty())
+			return { f(), this };
 
-		return it->second;
+		auto regexp = stack.top().release();
+		stack.pop();
+
+		return { regexp, this };
 	}
-};
+
+	template <bool like>
+	inline Regexp createRegexp(const std::string & pattern) { return pattern; }
+	template <>
+	inline Regexp createRegexp<true>(const std::string & pattern) { return likePatternToRegexp(pattern); }
+
+	template <bool like = false>
+	inline Pointer get(const std::string & pattern)
+	{
+		/// C++11 has thread-safe function-local statics on most modern compilers.
+		static KnownRegexps known_regexps;
+		static std::mutex mutex;
+		std::lock_guard<std::mutex> lock{mutex};
+
+		auto it = known_regexps.find(pattern);
+		if (known_regexps.end() == it)
+			it = known_regexps.emplace(pattern, stdext::make_unique<Holder>()).first;
+
+		return it->second->get([&pattern] {
+			return new Regexp{createRegexp<like>(pattern)};
+		});
+	}
+}
 
 
 /** like - использовать выражения LIKE, если true; использовать выражения re2, если false.
@@ -325,17 +371,17 @@ struct MatchImpl
 		}
 		else
 		{
-			const OptimizedRegularExpression & regexp = like ? Regexps::getLike(pattern) : Regexps::get(pattern);
+			const auto & regexp = Regexps::get<like>(pattern);
 			size_t size = offsets.size();
 			for (size_t i = 0; i < size; ++i)
-				res[i] = revert ^ regexp.match(reinterpret_cast<const char *>(&data[i != 0 ? offsets[i - 1] : 0]), (i != 0 ? offsets[i] - offsets[i - 1] : offsets[0]) - 1);
+				res[i] = revert ^ regexp->match(reinterpret_cast<const char *>(&data[i != 0 ? offsets[i - 1] : 0]), (i != 0 ? offsets[i] - offsets[i - 1] : offsets[0]) - 1);
 		}
 	}
 
 	static void constant(const std::string & data, const std::string & pattern, UInt8 & res)
 	{
-		const OptimizedRegularExpression & regexp = like ? Regexps::getLike(pattern) : Regexps::get(pattern);
-		res = revert ^ regexp.match(data);
+		const auto & regexp = Regexps::get<like>(pattern);
+		res = revert ^ regexp->match(data);
 	}
 };
 
@@ -349,9 +395,9 @@ struct ExtractImpl
 		res_data.reserve(data.size() / 5);
 		res_offsets.resize(offsets.size());
 
-		const OptimizedRegularExpression & regexp = Regexps::get(pattern);
+		const auto & regexp = Regexps::get(pattern);
 
-		unsigned capture = regexp.getNumberOfSubpatterns() > 0 ? 1 : 0;
+		unsigned capture = regexp->getNumberOfSubpatterns() > 0 ? 1 : 0;
 		OptimizedRegularExpression::MatchVec matches;
 		matches.reserve(capture + 1);
 		size_t prev_offset = 0;
@@ -361,10 +407,10 @@ struct ExtractImpl
 		{
 			size_t cur_offset = offsets[i];
 
-			unsigned count = regexp.match(reinterpret_cast<const char *>(&data[prev_offset]), cur_offset - prev_offset - 1, matches, capture + 1);
+			unsigned count = regexp->match(reinterpret_cast<const char *>(&data[prev_offset]), cur_offset - prev_offset - 1, matches, capture + 1);
 			if (count > capture && matches[capture].offset != std::string::npos)
 			{
-				const OptimizedRegularExpression::Match & match = matches[capture];
+				const auto & match = matches[capture];
 				res_data.resize(res_offset + match.length + 1);
 				memcpy(&res_data[res_offset], &data[prev_offset + match.offset], match.length);
 				res_offset += match.length;
@@ -462,7 +508,7 @@ struct ReplaceRegexpImpl
 
 				if (searcher.Match(input, start_pos, input.length(), re2::RE2::Anchor::UNANCHORED, matches, capture))
 				{
-					const re2::StringPiece & match = matches[0];
+					const auto & match = matches[0];
 					size_t char_to_copy = (match.data() - input.data()) - start_pos;
 
 					/// Копируем данные без изменения
@@ -544,7 +590,7 @@ struct ReplaceRegexpImpl
 
 				if (searcher.Match(input, start_pos, input.length(), re2::RE2::Anchor::UNANCHORED, matches, capture))
 				{
-					const re2::StringPiece & match = matches[0];
+					const auto & match = matches[0];
 					size_t char_to_copy = (match.data() - input.data()) - start_pos;
 
 					/// Копируем данные без изменения
@@ -618,7 +664,7 @@ struct ReplaceRegexpImpl
 
 			if (searcher.Match(input, start_pos, input.length(), re2::RE2::Anchor::UNANCHORED, matches, capture))
 			{
-				const re2::StringPiece & match = matches[0];
+				const auto & match = matches[0];
 				size_t char_to_copy = (match.data() - input.data()) - start_pos;
 
 				/// Копируем данные без изменения

dbms/include/DB/Storages/IStorage.h

@@ -191,7 +191,7 @@ public:
 	  * В этой функции нужно переименовать директорию с данными, если она есть.
 	  * Вызывается при заблокированной на запись структуре таблицы.
 	  */
-	virtual void rename(const String & new_path_to_db, const String & new_name)
+	virtual void rename(const String & new_path_to_db, const String & new_database_name, const String & new_table_name)
 	{
 		throw Exception("Method rename is not supported by storage " + getName(), ErrorCodes::NOT_IMPLEMENTED);
 	}

dbms/include/DB/Storages/MergeTree/MergeTreeData.h

@@ -10,6 +10,7 @@
 #include <DB/IO/ReadBufferFromString.h>
 #include <DB/IO/WriteBufferFromFile.h>
 #include <DB/Common/escapeForFileName.h>
+#include <DB/DataTypes/DataTypeString.h>
 #include <Poco/RWLock.h>
 
 
@@ -572,6 +573,18 @@ public:
 
 	const NamesAndTypesList & getColumnsList() const { return *columns; }
 
+	NameAndTypePair getColumn(const String &column_name) const
+	{
+		if (column_name == "_part") return NameAndTypePair("_part", new DataTypeString);
+		return getRealColumn(column_name);
+	}
+
+	bool hasColumn(const String &column_name) const
+	{
+		if (column_name == "_part") return true;
+		return hasRealColumn(column_name);
+	}
+
 	String getFullPath() const { return full_path; }
 
 	String getLogName() const { return log_name; }
@@ -615,8 +628,9 @@ public:
 	void replaceParts(const DataPartsVector & remove, const DataPartsVector & add, bool clear_without_timeout);
 
 	/** Переименовывает кусок в prefix_кусок и убирает его из рабочего набора.
+	  * Если restore_covered, добавляет в рабочий набор неактивные куски, слиянием которых получен удаляемый кусок.
 	  */
-	void renameAndDetachPart(DataPartPtr part, const String & prefix);
+	void renameAndDetachPart(DataPartPtr part, const String & prefix, bool restore_covered = false);
 
 	/** Возвращает старые неактуальные куски, которые можно удалить. Одновременно удаляет их из списка кусков, но не с диска.
 	  */
@@ -639,7 +653,7 @@ public:
 	  * Сбрасывает кеши разжатых блоков и засечек.
 	  * Нужно вызывать под залоченным lockStructureForAlter().
 	  */
-	void setPath(const String & full_path);
+	void setPath(const String & full_path, bool move_data);
 
 	/* Проверить, что такой ALTER можно выполнить:
 	 *  - Есть все нужные столбцы.

dbms/include/DB/Storages/MergeTree/MergeTreeDataSelectExecutor.h

@@ -55,7 +55,8 @@ private:
 		size_t max_block_size,
 		bool use_uncompressed_cache,
 		ExpressionActionsPtr prewhere_actions,
-		const String & prewhere_column);
+		const String & prewhere_column,
+		bool add_virtual_column);
 
 	BlockInputStreams spreadMarkRangesAmongThreadsFinal(
 		RangesInDataParts parts,
@@ -64,7 +65,8 @@ private:
 		size_t max_block_size,
 		bool use_uncompressed_cache,
 		ExpressionActionsPtr prewhere_actions,
-		const String & prewhere_column);
+		const String & prewhere_column,
+		bool add_virtual_column);
 
 	/// Создать выражение "Sign == 1".
 	void createPositiveSignCondition(ExpressionActionsPtr & out_expression, String & out_column);

dbms/include/DB/Storages/MergeTree/ReplicatedMergeTreeBlockOutputStream.h

@@ -18,6 +18,7 @@ public:
 	void write(const Block & block) override
 	{
 		auto part_blocks = storage.writer.splitBlockIntoParts(block);
+
 		for (auto & current_block : part_blocks)
 		{
 			storage.data.delayInsertIfNeeded();

dbms/include/DB/Storages/StorageChunks.h

@@ -62,7 +62,7 @@ public:
 	}
 	
 	/// Переименование испортило бы целостность количества ссылок из таблиц ChunkRef.
-	void rename(const String & new_path_to_db, const String & new_name)
+	void rename(const String & new_path_to_db, const String & new_database_name, const String & new_table_name)
 	{
 		throw Exception("Table doesn't support renaming", ErrorCodes::NOT_IMPLEMENTED);
 	}

dbms/include/DB/Storages/StorageDistributed.h

@@ -56,7 +56,7 @@ public:
 		unsigned threads = 1);
 
 	void drop() override {}
-	void rename(const String & new_path_to_db, const String & new_name) { name = new_name; }
+	void rename(const String & new_path_to_db, const String & new_database_name, const String & new_table_name) { name = new_table_name; }
 	/// в подтаблицах добавлять и удалять столбы нужно вручную
 	/// структура подтаблиц не проверяется
 	void alter(const AlterCommands & params, const String & database_name, const String & table_name, Context & context);

dbms/include/DB/Storages/StorageLog.h

@@ -154,7 +154,7 @@ public:
 	BlockOutputStreamPtr write(
 		ASTPtr query);
 
-	void rename(const String & new_path_to_db, const String & new_name);
+	void rename(const String & new_path_to_db, const String & new_database_name, const String & new_table_name);
 
 protected:
 	String path;

dbms/include/DB/Storages/StorageMemory.h

@@ -83,7 +83,7 @@ public:
 		ASTPtr query);
 
 	void drop() override;
-	void rename(const String & new_path_to_db, const String & new_name) { name = new_name; }
+	void rename(const String & new_path_to_db, const String & new_database_name, const String & new_table_name) { name = new_table_name; }
 
 private:
 	String name;

dbms/include/DB/Storages/StorageMerge.h

@@ -45,7 +45,7 @@ public:
 		unsigned threads = 1);
 
 	void drop() override {}
-	void rename(const String & new_path_to_db, const String & new_name) { name = new_name; }
+	void rename(const String & new_path_to_db, const String & new_database_name, const String & new_table_name) { name = new_table_name; }
 	
 	/// в подтаблицах добавлять и удалять столбы нужно вручную
 	/// структура подтаблиц не проверяется

dbms/include/DB/Storages/StorageMergeTree.h

@@ -50,6 +50,16 @@ public:
 
 	const NamesAndTypesList & getColumnsList() const { return data.getColumnsList(); }
 
+	NameAndTypePair getColumn(const String &column_name) const
+	{
+		return data.getColumn(column_name);
+	}
+
+	bool hasColumn(const String &column_name) const
+	{
+		return data.hasColumn(column_name);
+	}
+
 	BlockInputStreams read(
 		const Names & column_names,
 		ASTPtr query,
@@ -69,7 +79,7 @@ public:
 
 	void drop() override;
 
-	void rename(const String & new_path_to_db, const String & new_name);
+	void rename(const String & new_path_to_db, const String & new_database_name, const String & new_table_name);
 
 	void alter(const AlterCommands & params, const String & database_name, const String & table_name, Context & context);
 

dbms/include/DB/Storages/StorageNull.h

@@ -42,7 +42,7 @@ public:
 		return new NullBlockOutputStream;
 	}
 
-	void rename(const String & new_path_to_db, const String & new_name) { name = new_name; }
+	void rename(const String & new_path_to_db, const String & new_database_name, const String & new_table_name) { name = new_table_name; }
 
 private:
 	String name;

dbms/include/DB/Storages/StorageReplicatedMergeTree.h

@@ -6,6 +6,7 @@
 #include <DB/Storages/MergeTree/MergeTreeDataWriter.h>
 #include <DB/Storages/MergeTree/MergeTreeDataSelectExecutor.h>
 #include <DB/Storages/MergeTree/ReplicatedMergeTreePartsExchange.h>
+#include <DB/DataTypes/DataTypesNumberFixed.h>
 #include <zkutil/ZooKeeper.h>
 #include <zkutil/LeaderElection.h>
 #include <statdaemons/threadpool.hpp>
@@ -50,6 +51,18 @@ public:
 
 	const NamesAndTypesList & getColumnsList() const override { return data.getColumnsList(); }
 
+	NameAndTypePair getColumn(const String &column_name) const
+	{
+		if (column_name == "_replicated") return NameAndTypePair("_replicated", new DataTypeUInt8);
+		return data.getColumn(column_name);
+	}
+
+	bool hasColumn(const String &column_name) const
+	{
+		if (column_name == "_replicated") return true;
+		return data.hasColumn(column_name);
+	}
+
 	BlockInputStreams read(
 		const Names & column_names,
 		ASTPtr query,
@@ -68,6 +81,8 @@ public:
 	  */
 	void drop() override;
 
+	void rename(const String & new_path_to_db, const String & new_database_name, const String & new_table_name);
+
 	bool supportsIndexForIn() const override { return true; }
 
 	/// Добавить кусок в очередь кусков, чьи данные нужно проверить в фоновом потоке.

dbms/include/DB/Storages/StorageTinyLog.h

@@ -127,7 +127,7 @@ public:
 
 	void drop() override;
 	
-	void rename(const String & new_path_to_db, const String & new_name);
+	void rename(const String & new_path_to_db, const String & new_database_name, const String & new_table_name);
 
 private:
 	String path;

dbms/src/Common/VirtualColumnUtils.cpp

@@ -145,7 +145,7 @@ static ASTPtr buildWhereExpression(const ASTs & functions)
 BlockInputStreamPtr getVirtualColumnsBlocks(ASTPtr query, const Block & input, const Context & context)
 {
 	const ASTSelectQuery & select = typeid_cast<ASTSelectQuery & >(*query);
-	if (!select.where_expression)
+	if (!select.where_expression && !select.prewhere_expression)
 		return new OneBlockInputStream(input);
 
 	ASTPtr new_query = new ASTSelectQuery();
@@ -164,7 +164,10 @@ BlockInputStreamPtr getVirtualColumnsBlocks(ASTPtr query, const Block & input, c
 		select_list.children.push_back(new ASTIdentifier(StringRange(), columns[i]));
 
 	std::vector<ASTPtr> functions;
+	if (select.where_expression)
 		extractFunctions(select.where_expression, columns, functions);
+	if (select.prewhere_expression)
+		extractFunctions(select.prewhere_expression, columns, functions);
 	new_select.where_expression = buildWhereExpression(functions);
 
 	if (new_select.select_expression_list)

dbms/src/Interpreters/InterpreterRenameQuery.cpp

@@ -59,9 +59,9 @@ void InterpreterRenameQuery::execute()
 
 		context.assertTableDoesntExist(to_database_name, to_table_name);
 
-		/// Уведомляем таблицу о том, что она переименовается. Если таблица не поддерживает переименование - кинется исключение.
+		/// Уведомляем таблицу о том, что она переименовывается. Если таблица не поддерживает переименование - кинется исключение.
 
-		table->rename(path + "data/" + to_database_name_escaped + "/", to_table_name);
+		table->rename(path + "data/" + to_database_name_escaped + "/", to_database_name, to_table_name);
 
 		/// Пишем новый файл с метаданными.
 		{

dbms/src/Server/Server.cpp

@@ -5,6 +5,7 @@
 
 #include <Yandex/ApplicationServerExt.h>
 #include <statdaemons/ConfigProcessor.h>
+#include <statdaemons/stdext.h>
 
 #include <DB/Interpreters/loadMetadata.h>
 #include <DB/Storages/StorageSystemNumbers.h>
@@ -372,9 +373,7 @@ int Server::main(const std::vector<std::string> & args)
 		global_context->setDefaultReplicaName(config().getString("replica_name"));
 
 	std::string users_config_path = config().getString("users_config", config().getString("config-file", "config.xml"));
-	auto users_config_reloader = std::unique_ptr<UsersConfigReloader>{
-		new UsersConfigReloader(users_config_path, global_context.get())
-	};
+	auto users_config_reloader = stdext::make_unique<UsersConfigReloader>(users_config_path, global_context.get());
 
 	/// Максимальное количество одновременно выполняющихся запросов.
 	global_context->getProcessList().setMaxSize(config().getInt("max_concurrent_queries", 0));
@@ -411,7 +410,7 @@ int Server::main(const std::vector<std::string> & args)
 
 	{
 		const auto profile_events_transmitter = config().getBool("use_graphite", true)
-			? std::unique_ptr<ProfileEventsTransmitter>{new ProfileEventsTransmitter{}}
+			? stdext::make_unique<ProfileEventsTransmitter>()
 			: nullptr;
 
 		bool use_olap_server = config().getBool("use_olap_http_server", false);

dbms/src/Storages/MergeTree/MergeTreeData.cpp

@@ -317,12 +317,16 @@ void MergeTreeData::clearOldParts()
 	}
 }
 
-void MergeTreeData::setPath(const String & new_full_path)
+void MergeTreeData::setPath(const String & new_full_path, bool move_data)
 {
+	if (move_data)
+	{
 		Poco::File(full_path).renameTo(new_full_path);
-	full_path = new_full_path;
-
+		/// Если данные перемещать не нужно, значит их переместил кто-то другой. Расчитываем, что он еще и сбросил кеши.
 		context.resetCaches();
+	}
+
+	full_path = new_full_path;
 }
 
 void MergeTreeData::dropAllData()
@@ -716,14 +720,67 @@ void MergeTreeData::replaceParts(const DataPartsVector & remove, const DataParts
 	}
 }
 
-void MergeTreeData::renameAndDetachPart(DataPartPtr part, const String & prefix)
+void MergeTreeData::renameAndDetachPart(DataPartPtr part, const String & prefix, bool restore_covered)
 {
+	LOG_INFO(log, "Renaming " << part->name << " to " << prefix << part->name << " and detaching it.");
+
 	Poco::ScopedLock<Poco::FastMutex> lock(data_parts_mutex);
 	Poco::ScopedLock<Poco::FastMutex> lock_all(all_data_parts_mutex);
+
 	if (!all_data_parts.erase(part))
 		throw Exception("No such data part", ErrorCodes::NO_SUCH_DATA_PART);
+
 	data_parts.erase(part);
 	part->renameAddPrefix(prefix);
+
+	if (restore_covered)
+	{
+		auto it = all_data_parts.lower_bound(part);
+		Strings restored;
+		bool error = false;
+
+		UInt64 pos = part->left;
+
+		if (it != all_data_parts.begin())
+		{
+			--it;
+			if (part->contains(**it))
+			{
+				if ((*it)->left != part->left)
+					error = true;
+				data_parts.insert(*it);
+				pos = (*it)->right + 1;
+				restored.push_back((*it)->name);
+			}
+			else
+				error = true;
+			++it;
+		}
+		else
+			error = true;
+
+		for (; it != all_data_parts.end() && part->contains(**it); ++it)
+		{
+			if ((*it)->left < pos)
+				continue;
+			if ((*it)->left > pos)
+				error = true;
+			data_parts.insert(*it);
+			pos = (*it)->right + 1;
+			restored.push_back((*it)->name);
+		}
+
+		if (pos != part->right + 1)
+			error = true;
+
+		for (const String & name : restored)
+		{
+			LOG_INFO(log, "Activated part " << name);
+		}
+
+		if (error)
+			LOG_ERROR(log, "The set of parts restored in place of " << part->name << " looks incomplete. There might or might not be a data loss.");
+	}
 }
 
 MergeTreeData::DataParts MergeTreeData::getDataParts()

dbms/src/Storages/MergeTree/MergeTreeDataMerger.cpp

@@ -251,6 +251,10 @@ MergeTreeData::DataPartPtr MergeTreeDataMerger::mergeParts(
 {
 	LOG_DEBUG(log, "Merging " << parts.size() << " parts: from " << parts.front()->name << " to " << parts.back()->name << " into " << merged_name);
 
+	String merged_dir = data.getFullPath() + merged_name;
+	if (Poco::File(merged_dir).exists())
+		throw Exception("Directory " + merged_dir + " already exists", ErrorCodes::DIRECTORY_ALREADY_EXISTS);
+
 	NameSet union_columns_set;
 	for (const MergeTreeData::DataPartPtr & part : parts)
 	{

dbms/src/Storages/MergeTree/MergeTreeDataSelectExecutor.cpp

@@ -6,6 +6,7 @@
 #include <DB/DataStreams/ConcatBlockInputStream.h>
 #include <DB/DataStreams/CollapsingFinalBlockInputStream.h>
 #include <DB/DataTypes/DataTypesNumberFixed.h>
+#include <DB/Common/VirtualColumnUtils.h>
 
 namespace DB
 {
@@ -19,6 +20,19 @@ MergeTreeDataSelectExecutor::MergeTreeDataSelectExecutor(MergeTreeData & data_)
 
 }
 
+/// Построить блок состоящий только из возможных значений виртуальных столбцов
+static Block getBlockWithVirtualColumns(const MergeTreeData::DataPartsVector & parts)
+{
+	Block res;
+	ColumnWithNameAndType _part(new ColumnString, new DataTypeString, "_part");
+
+	for (const auto & part : parts)
+		_part.column->insert(part->name);
+
+	res.insert(_part);
+	return res;
+}
+
 BlockInputStreams MergeTreeDataSelectExecutor::read(
 	const Names & column_names_to_return,
 	ASTPtr query,
@@ -27,30 +41,60 @@ BlockInputStreams MergeTreeDataSelectExecutor::read(
 	size_t max_block_size,
 	unsigned threads)
 {
-	data.check(column_names_to_return);
+	MergeTreeData::DataPartsVector parts;
+
+	{
+		auto parts_set = data.getDataParts();
+		parts.assign(parts_set.begin(), parts_set.end());
+	}
+
+	/// Если в запросе есть ограничения на виртуальный столбец _part, выберем только подходящие под него куски.
+	Names virt_column_names, real_column_names;
+	for (const auto & it : column_names_to_return)
+		if (it != "_part")
+			real_column_names.push_back(it);
+		else
+			virt_column_names.push_back(it);
+
+	Block virtual_columns_block = getBlockWithVirtualColumns(parts);
+	BlockInputStreamPtr virtual_columns;
+
+	/// Если запрошен хотя бы один виртуальный столбец, пробуем индексировать
+	if (!virt_column_names.empty())
+		virtual_columns = VirtualColumnUtils::getVirtualColumnsBlocks(query->clone(), virtual_columns_block, data.context);
+	else /// Иначе, считаем допустимыми все возможные значения
+		virtual_columns = new OneBlockInputStream(virtual_columns_block);
+
+	std::multiset<String> values = VirtualColumnUtils::extractSingleValueFromBlocks<String>(virtual_columns, "_part");
+
+	data.check(real_column_names);
 	processed_stage = QueryProcessingStage::FetchColumns;
 
 	PKCondition key_condition(query, data.context, data.getColumnsList(), data.getSortDescription());
 	PKCondition date_condition(query, data.context, data.getColumnsList(), SortDescription(1, SortColumnDescription(data.date_column_name, 1)));
 
-	MergeTreeData::DataPartsVector parts;
-
-	/// Выберем куски, в которых могут быть данные, удовлетворяющие date_condition.
+	/// Выберем куски, в которых могут быть данные, удовлетворяющие date_condition, и которые подходят под условие на _part.
 	{
-		MergeTreeData::DataParts data_parts = data.getDataParts();
+		auto prev_parts = parts;
+		parts.clear();
 
-		for (MergeTreeData::DataParts::iterator it = data_parts.begin(); it != data_parts.end(); ++it)
+		for (const auto & part : prev_parts)
 		{
-			Field left = static_cast<UInt64>((*it)->left_date);
-			Field right = static_cast<UInt64>((*it)->right_date);
+			if (values.find(part->name) == values.end())
+				continue;
 
-			if (date_condition.mayBeTrueInRange(&left, &right))
-				parts.push_back(*it);
+			Field left = static_cast<UInt64>(part->left_date);
+			Field right = static_cast<UInt64>(part->right_date);
+
+			if (!date_condition.mayBeTrueInRange(&left, &right))
+				continue;
+
+			parts.push_back(part);
 		}
 	}
 
 	/// Семплирование.
-	Names column_names_to_read = column_names_to_return;
+	Names column_names_to_read = real_column_names;
 	UInt64 sampling_column_value_limit = 0;
 	typedef Poco::SharedPtr<ASTFunction> ASTFunctionPtr;
 	ASTFunctionPtr filter_function;
@@ -185,7 +229,8 @@ BlockInputStreams MergeTreeDataSelectExecutor::read(
 			max_block_size,
 			settings.use_uncompressed_cache,
 			prewhere_actions,
-			prewhere_column);
+			prewhere_column,
+			!virt_column_names.empty());
 	}
 	else
 	{
@@ -196,7 +241,8 @@ BlockInputStreams MergeTreeDataSelectExecutor::read(
 			max_block_size,
 			settings.use_uncompressed_cache,
 			prewhere_actions,
-			prewhere_column);
+			prewhere_column,
+			!virt_column_names.empty());
 	}
 
 	if (select.sample_size)
@@ -220,7 +266,8 @@ BlockInputStreams MergeTreeDataSelectExecutor::spreadMarkRangesAmongThreads(
 	size_t max_block_size,
 	bool use_uncompressed_cache,
 	ExpressionActionsPtr prewhere_actions,
-	const String & prewhere_column)
+	const String & prewhere_column,
+	bool add_virtual_column)
 {
 	/// На всякий случай перемешаем куски.
 	std::random_shuffle(parts.begin(), parts.end());
@@ -282,6 +329,9 @@ BlockInputStreams MergeTreeDataSelectExecutor::spreadMarkRangesAmongThreads(
 						data.getFullPath() + part.data_part->name + '/', max_block_size, column_names, data,
 						part.data_part, part.ranges, use_uncompressed_cache,
 						prewhere_actions, prewhere_column));
+					if (add_virtual_column)
+						streams.back() = new AddingConstColumnBlockInputStream<String>(
+							streams.back(), new DataTypeString, part.data_part->name, "_part");
 					need_marks -= marks_in_part;
 					parts.pop_back();
 					sum_marks_in_parts.pop_back();
@@ -312,6 +362,9 @@ BlockInputStreams MergeTreeDataSelectExecutor::spreadMarkRangesAmongThreads(
 					data.getFullPath() + part.data_part->name + '/', max_block_size, column_names, data,
 					part.data_part, ranges_to_get_from_part, use_uncompressed_cache,
 					prewhere_actions, prewhere_column));
+				if (add_virtual_column)
+					streams.back() = new AddingConstColumnBlockInputStream<String>(
+						streams.back(), new DataTypeString, part.data_part->name, "_part");
 			}
 
 			if (streams.size() == 1)
@@ -334,7 +387,8 @@ BlockInputStreams MergeTreeDataSelectExecutor::spreadMarkRangesAmongThreadsFinal
 	size_t max_block_size,
 	bool use_uncompressed_cache,
 	ExpressionActionsPtr prewhere_actions,
-	const String & prewhere_column)
+	const String & prewhere_column,
+	bool add_virtual_column)
 {
 	size_t sum_marks = 0;
 	for (size_t i = 0; i < parts.size(); ++i)
@@ -358,6 +412,9 @@ BlockInputStreams MergeTreeDataSelectExecutor::spreadMarkRangesAmongThreadsFinal
 			data.getFullPath() + part.data_part->name + '/', max_block_size, column_names, data,
 			part.data_part, part.ranges, use_uncompressed_cache,
 			prewhere_actions, prewhere_column);
+		if (add_virtual_column)
+			source_stream = new AddingConstColumnBlockInputStream<String>(
+				source_stream, new DataTypeString, part.data_part->name, "_part");
 
 		to_collapse.push_back(new ExpressionBlockInputStream(source_stream, data.getPrimaryExpression()));
 	}

dbms/src/Storages/StorageLog.cpp

@@ -523,15 +523,15 @@ size_t StorageLog::marksCount()
 }
 
 
-void StorageLog::rename(const String & new_path_to_db, const String & new_name)
+void StorageLog::rename(const String & new_path_to_db, const String & new_database_name, const String & new_table_name)
 {
 	Poco::ScopedWriteRWLock lock(rwlock);
 
 	/// Переименовываем директорию с данными.
-	Poco::File(path + escapeForFileName(name)).renameTo(new_path_to_db + escapeForFileName(new_name));
+	Poco::File(path + escapeForFileName(name)).renameTo(new_path_to_db + escapeForFileName(new_table_name));
 
 	path = new_path_to_db;
-	name = new_name;
+	name = new_table_name;
 
 	for (Files_t::iterator it = files.begin(); it != files.end(); ++it)
 	{

dbms/src/Storages/StorageMergeTree.cpp

@@ -89,14 +89,14 @@ void StorageMergeTree::drop()
 	data.dropAllData();
 }
 
-void StorageMergeTree::rename(const String & new_path_to_db, const String & new_name)
+void StorageMergeTree::rename(const String & new_path_to_db, const String & new_database_name, const String & new_table_name)
 {
-	std::string new_full_path = new_path_to_db + escapeForFileName(new_name) + '/';
+	std::string new_full_path = new_path_to_db + escapeForFileName(new_table_name) + '/';
 
-	data.setPath(new_full_path);
+	data.setPath(new_full_path, true);
 
 	path = new_path_to_db;
-	name = new_name;
+	name = new_table_name;
 	full_path = new_full_path;
 
 	increment.setPath(full_path + "increment.txt");

dbms/src/Storages/StorageReplicatedMergeTree.cpp

@@ -6,6 +6,7 @@
 #include <DB/IO/WriteBufferFromOStream.h>
 #include <DB/IO/ReadBufferFromString.h>
 #include <DB/Interpreters/InterpreterAlterQuery.h>
+#include <DB/Common/VirtualColumnUtils.h>
 #include <time.h>
 
 namespace DB
@@ -461,7 +462,7 @@ void StorageReplicatedMergeTree::checkParts(bool skip_sanity_checks)
 	for (MergeTreeData::DataPartPtr part : unexpected_parts)
 	{
 		LOG_ERROR(log, "Renaming unexpected part " << part->name << " to ignored_" + part->name);
-		data.renameAndDetachPart(part, "ignored_");
+		data.renameAndDetachPart(part, "ignored_", true);
 	}
 }
 
@@ -511,6 +512,12 @@ void StorageReplicatedMergeTree::checkPartAndAddToZooKeeper(MergeTreeData::DataP
 		checksums.checkEqual(part->checksums, true);
 	}
 
+	if (zookeeper->exists(replica_path + "/parts/" + part->name))
+	{
+		LOG_ERROR(log, "checkPartAndAddToZooKeeper: node " << replica_path + "/parts/" + part->name << " already exists");
+		return;
+	}
+
 	ops.push_back(new zkutil::Op::Check(
 		zookeeper_path + "/columns",
 		expected_columns_version));
@@ -533,6 +540,8 @@ void StorageReplicatedMergeTree::checkPartAndAddToZooKeeper(MergeTreeData::DataP
 
 void StorageReplicatedMergeTree::clearOldParts()
 {
+	auto table_lock = lockStructure(false);
+
 	MergeTreeData::DataPartsVector parts = data.grabOldParts();
 	size_t count = parts.size();
 
@@ -545,6 +554,8 @@ void StorageReplicatedMergeTree::clearOldParts()
 		{
 			MergeTreeData::DataPartPtr part = parts.back();
 
+			LOG_DEBUG(log, "Removing " << part->name);
+
 			zkutil::Ops ops;
 			ops.push_back(new zkutil::Op::Remove(replica_path + "/parts/" + part->name + "/columns", -1));
 			ops.push_back(new zkutil::Op::Remove(replica_path + "/parts/" + part->name + "/checksums", -1));
@@ -832,7 +843,7 @@ bool StorageReplicatedMergeTree::executeLogEntry(const LogEntry & entry, Backgro
 				}
 			}
 
-			auto table_lock = lockStructure(true);
+			auto table_lock = lockStructure(false);
 
 			MergeTreeData::Transaction transaction;
 			MergeTreeData::DataPartPtr part = merger.mergeParts(parts, entry.new_part_name, &transaction);
@@ -841,6 +852,10 @@ bool StorageReplicatedMergeTree::executeLogEntry(const LogEntry & entry, Backgro
 			checkPartAndAddToZooKeeper(part, ops);
 
 			zookeeper->multi(ops);
+
+			/** При ZCONNECTIONLOSS или ZOPERATIONTIMEOUT можем зря откатить локальные изменения кусков.
+			  * Это не проблема, потому что в таком случае слияние останется в очереди, и мы попробуем снова.
+			  */
 			transaction.commit();
 			merge_selecting_event.set();
 
@@ -1194,6 +1209,8 @@ void StorageReplicatedMergeTree::alterThread()
 					changed = true;
 			}
 
+			MergeTreeData::DataParts parts;
+
 			/// Если описание столбцов изменилось, обновим структуру таблицы локально.
 			if (changed)
 			{
@@ -1207,6 +1224,9 @@ void StorageReplicatedMergeTree::alterThread()
 						unreplicated_data->setColumnsList(columns);
 					columns_version = stat.version;
 					LOG_INFO(log, "Applied changes to table.");
+
+					/// Нужно получить список кусков под блокировкой таблицы, чтобы избежать race condition с мерджем.
+					parts = data.getDataParts();
 				}
 				else
 				{
@@ -1222,7 +1242,10 @@ void StorageReplicatedMergeTree::alterThread()
 
 				int changed_parts = 0;
 
-				auto parts = data.getDataParts();
+				if (!changed)
+					parts = data.getDataParts();
+
+				auto table_lock = lockStructure(false);
 
 				for (const MergeTreeData::DataPartPtr & part : parts)
 				{
@@ -1379,7 +1402,7 @@ void StorageReplicatedMergeTree::partCheckThread()
 					ActiveDataPartSet::parsePartName(part_name, part_info);
 
 					/** Будем проверять только куски, не полученные в результате слияния.
-					  * Для кусков, полученных в результате слияния такая проверка была бы некорректной,
+					  * Для кусков, полученных в результате слияния, такая проверка была бы некорректной,
 					  *  потому что слитого куска может еще ни у кого не быть.
 					  */
 					if (part_info.left == part_info.right)
@@ -1461,6 +1484,8 @@ void StorageReplicatedMergeTree::partCheckThread()
 			/// У нас есть этот кусок, и он активен.
 			else if (part->name == part_name)
 			{
+				auto table_lock = lockStructure(false);
+
 				/// Если кусок есть в ZooKeeper, сверим его данные с его чексуммами, а их с ZooKeeper.
 				if (zookeeper->exists(replica_path + "/parts/" + part_name))
 				{
@@ -1781,11 +1806,60 @@ BlockInputStreams StorageReplicatedMergeTree::read(
 		size_t max_block_size,
 		unsigned threads)
 {
-	BlockInputStreams res = reader.read(column_names, query, settings, processed_stage, max_block_size, threads);
+	Names virt_column_names;
+	Names real_column_names;
+	for (const auto & it : column_names)
+		if (it == "_replicated")
+			virt_column_names.push_back(it);
+		else
+			real_column_names.push_back(it);
 
-	if (unreplicated_reader)
+	Block virtual_columns_block;
+	ColumnUInt8 * column = new ColumnUInt8(2);
+	ColumnPtr column_ptr = column;
+	column->getData()[0] = 0;
+	column->getData()[1] = 1;
+	virtual_columns_block.insert(ColumnWithNameAndType(column_ptr, new DataTypeUInt8, "_replicated"));
+
+	BlockInputStreamPtr virtual_columns;
+
+	/// Если запрошен хотя бы один виртуальный столбец, пробуем индексировать
+	if (!virt_column_names.empty())
+		virtual_columns = VirtualColumnUtils::getVirtualColumnsBlocks(query->clone(), virtual_columns_block, context);
+	else /// Иначе, считаем допустимыми все возможные значения
+		virtual_columns = new OneBlockInputStream(virtual_columns_block);
+
+	std::multiset<UInt8> values = VirtualColumnUtils::extractSingleValueFromBlocks<UInt8>(virtual_columns, "_replicated");
+
+	BlockInputStreams res;
+
+	if (values.count(1))
 	{
-		BlockInputStreams res2 = unreplicated_reader->read(column_names, query, settings, processed_stage, max_block_size, threads);
+		res = reader.read(real_column_names, query, settings, processed_stage, max_block_size, threads);
+
+		for (auto & virtual_column : virt_column_names)
+		{
+			if (virtual_column == "_replicated")
+			{
+				for (auto & stream : res)
+					stream = new AddingConstColumnBlockInputStream<UInt8>(stream, new DataTypeUInt8, 1, "_replicated");
+			}
+		}
+	}
+
+	if (unreplicated_reader && values.count(0))
+	{
+		BlockInputStreams res2 = unreplicated_reader->read(real_column_names, query, settings, processed_stage, max_block_size, threads);
+
+		for (auto & virtual_column : virt_column_names)
+		{
+			if (virtual_column == "_replicated")
+			{
+				for (auto & stream : res2)
+					stream = new AddingConstColumnBlockInputStream<UInt8>(stream, new DataTypeUInt8, 0, "_replicated");
+			}
+		}
+
 		res.insert(res.begin(), res2.begin(), res2.end());
 	}
 
@@ -1935,6 +2009,21 @@ void StorageReplicatedMergeTree::drop()
 	data.dropAllData();
 }
 
+void StorageReplicatedMergeTree::rename(const String & new_path_to_db, const String & new_database_name, const String & new_table_name)
+{
+	std::string new_full_path = new_path_to_db + escapeForFileName(new_table_name) + '/';
+
+	data.setPath(new_full_path, true);
+	if (unreplicated_data)
+		unreplicated_data->setPath(new_full_path + "unreplicated/", false);
+
+	database_name = new_database_name;
+	table_name = new_table_name;
+	full_path = new_full_path;
+
+	/// TODO: Можно обновить названия логгеров.
+}
+
 void StorageReplicatedMergeTree::LogEntry::writeText(WriteBuffer & out) const
 {
 	writeString("format version: 1\n", out);

dbms/src/Storages/StorageTinyLog.cpp

@@ -353,13 +353,13 @@ void StorageTinyLog::addFile(const String & column_name, const IDataType & type,
 }
 
 
-void StorageTinyLog::rename(const String & new_path_to_db, const String & new_name)
+void StorageTinyLog::rename(const String & new_path_to_db, const String & new_database_name, const String & new_table_name)
 {
 	/// Переименовываем директорию с данными.
-	Poco::File(path + escapeForFileName(name)).renameTo(new_path_to_db + escapeForFileName(new_name));
+	Poco::File(path + escapeForFileName(name)).renameTo(new_path_to_db + escapeForFileName(new_table_name));
 
 	path = new_path_to_db;
-	name = new_name;
+	name = new_table_name;
 
 	for (Files_t::iterator it = files.begin(); it != files.end(); ++it)
 		it->second.data_file = Poco::File(path + escapeForFileName(name) + '/' + Poco::Path(it->second.data_file.path()).getFileName());

libs/libzkutil/src/ZooKeeper.cpp

@@ -366,6 +366,9 @@ int32_t ZooKeeper::trySet(const std::string & path, const std::string & data,
 
 int32_t ZooKeeper::multiImpl(const Ops & ops_, OpResultsPtr * out_results_)
 {
+	if (ops_.empty())
+		return ZOK;
+
 	size_t count = ops_.size();
 	OpResultsPtr out_results(new OpResults(count));