Merged with master

.github/ISSUE_TEMPLATE/bug-report-or-unexpected-behaviour.md

@@ -1,16 +1,16 @@
 ---
-name: Bug report
+name: Bug report or unexpected behaviour
 about: Create a report to help us improve ClickHouse
 title: ''
-labels: bug, issue
+labels: bug
 assignees: ''
 
 ---
 
 (you don't have to strictly follow this form)
 
-**Describe the bug**
+**Describe the bug or unexpected behaviour**
-A clear and concise description of what the bug is.
+A clear and concise description of what works not as it is supposed to.
 
 **How to reproduce**
 * Which ClickHouse server version to use

.github/ISSUE_TEMPLATE/documentation-issue.md

@@ -0,0 +1,16 @@
+---
+name: Documentation issue
+about: Report something incorrect or missing in documentation
+title: ''
+labels: documentation
+assignees: BayoNet
+
+---
+
+(you don't have to strictly follow this form)
+
+**Describe the issue**
+A clear and concise description of what's wrong in documentation.
+
+**Additional context**
+Add any other context about the problem here.

.github/ISSUE_TEMPLATE/performance-issue.md

@@ -0,0 +1,27 @@
+---
+name: Performance issue
+about: Report something working slower than expected
+title: ''
+labels: performance
+assignees: ''
+
+---
+
+(you don't have to strictly follow this form)
+
+**Describe the situation**
+What exactly works slower than expected?
+
+**How to reproduce**
+* Which ClickHouse server version to use
+* Which interface to use, if matters
+* Non-default settings, if any
+* `CREATE TABLE` statements for all tables involved
+* Sample data for all these tables, use [clickhouse-obfuscator](https://github.com/yandex/ClickHouse/blob/master/dbms/programs/obfuscator/Obfuscator.cpp#L42-L80) if necessary
+* Queries to run that lead to slow performance
+
+**Expected performance**
+What are your performance expectation, why do you think they are realistic? Has it been working faster in older ClickHouse releases? Is it working faster in some specific other system?
+
+**Additional context**
+Add any other context about the problem here.

dbms/CMakeLists.txt

@@ -372,8 +372,8 @@ if (USE_PROTOBUF)
 endif ()
 
 if (USE_HDFS)
-    target_link_libraries (clickhouse_common_io PRIVATE ${HDFS3_LIBRARY})
+    target_link_libraries (clickhouse_common_io PUBLIC ${HDFS3_LIBRARY})
-    target_include_directories (clickhouse_common_io SYSTEM BEFORE PRIVATE ${HDFS3_INCLUDE_DIR})
+    target_include_directories (clickhouse_common_io SYSTEM BEFORE PUBLIC ${HDFS3_INCLUDE_DIR})
 endif()
 
 if (USE_BROTLI)

dbms/cmake/version.cmake

@@ -1,11 +1,11 @@
 # This strings autochanged from release_lib.sh:
-set(VERSION_REVISION 54425)
+set(VERSION_REVISION 54426)
 set(VERSION_MAJOR 19)
-set(VERSION_MINOR 14)
+set(VERSION_MINOR 15)
 set(VERSION_PATCH 1)
-set(VERSION_GITHASH adfc36917222bdb03eba069f0cad0f4f5b8f1c94)
+set(VERSION_GITHASH 6f1a8c37abe6ee4e7ee74c0b5cb9c05a87417b61)
-set(VERSION_DESCRIBE v19.14.1.1-prestable)
+set(VERSION_DESCRIBE v19.15.1.1-prestable)
-set(VERSION_STRING 19.14.1.1)
+set(VERSION_STRING 19.15.1.1)
 # end of autochange
 
 set(VERSION_EXTRA "" CACHE STRING "")

dbms/programs/client/Client.cpp

@@ -106,6 +106,7 @@ namespace ErrorCodes
     extern const int CANNOT_SET_SIGNAL_HANDLER;
     extern const int CANNOT_READLINE;
     extern const int SYSTEM_ERROR;
+    extern const int INVALID_USAGE_OF_INPUT;
 }
 
 
@@ -431,8 +432,14 @@ private:
             /// Load command history if present.
             if (config().has("history_file"))
                 history_file = config().getString("history_file");
-            else if (!home_path.empty())
+            else
-                history_file = home_path + "/.clickhouse-client-history";
+            {
+                auto history_file_from_env = getenv("CLICKHOUSE_HISTORY_FILE");
+                if (history_file_from_env)
+                    history_file = history_file_from_env;
+                else if (!home_path.empty())
+                    history_file = home_path + "/.clickhouse-client-history";
+            }
 
             if (!history_file.empty())
             {
@@ -837,9 +844,17 @@ private:
 
             connection->forceConnected(connection_parameters.timeouts);
 
-            /// INSERT query for which data transfer is needed (not an INSERT SELECT) is processed separately.
+            ASTPtr input_function;
-            if (insert && !insert->select)
+            if (insert && insert->select)
+                insert->tryFindInputFunction(input_function);
+
+            /// INSERT query for which data transfer is needed (not an INSERT SELECT or input()) is processed separately.
+            if (insert && (!insert->select || input_function))
+            {
+                if (input_function && insert->format.empty())
+                    throw Exception("FORMAT must be specified for function input()", ErrorCodes::INVALID_USAGE_OF_INPUT);
                 processInsertQuery();
+            }
             else
                 processOrdinaryQuery();
         }

dbms/programs/odbc-bridge/ODBCBridge.cpp

@@ -15,6 +15,7 @@
 #include <common/logger_useful.h>
 #include <ext/scope_guard.h>
 #include <ext/range.h>
+#include <Common/SensitiveDataMasker.h>
 
 namespace DB
 {
@@ -165,8 +166,7 @@ int ODBCBridge::main(const std::vector<std::string> & /*args*/)
 
     if (config().has("query_masking_rules"))
     {
-        context->setSensitiveDataMasker(std::make_unique<SensitiveDataMasker>(config(), "query_masking_rules"));
+        SensitiveDataMasker::setInstance(std::make_unique<SensitiveDataMasker>(config(), "query_masking_rules"));
-        setLoggerSensitiveDataMasker(logger(), context->getSensitiveDataMasker());
     }
 
     auto server = Poco::Net::HTTPServer(

dbms/programs/server/MySQLHandler.cpp

@@ -293,7 +293,8 @@ void MySQLHandler::comQuery(ReadBuffer & payload)
         should_replace = true;
     }
 
-    executeQuery(should_replace ? empty_select : payload, *out, true, connection_context, set_content_type, nullptr);
+    Context query_context = connection_context;
+    executeQuery(should_replace ? empty_select : payload, *out, true, query_context, set_content_type, nullptr);
 
     if (!with_output)
         packet_sender->sendPacket(OK_Packet(0x00, client_capability_flags, 0, 0, 0), true);

dbms/programs/server/Server.cpp

@@ -55,6 +55,7 @@
 #include "TCPHandlerFactory.h"
 #include "Common/config_version.h"
 #include "MySQLHandlerFactory.h"
+#include <Common/SensitiveDataMasker.h>
 
 
 #if defined(__linux__)
@@ -279,8 +280,6 @@ int Server::main(const std::vector<std::string> & /*args*/)
           */
         LOG_INFO(log, "Shutting down storages.");
 
-        // global_context is the owner of sensitive_data_masker, which will be destoyed after global_context->shutdown() call
-        setLoggerSensitiveDataMasker(logger(), nullptr);
         global_context->shutdown();
 
         LOG_DEBUG(log, "Shutted down storages.");
@@ -414,7 +413,7 @@ int Server::main(const std::vector<std::string> & /*args*/)
 
     if (config().has("query_masking_rules"))
     {
-        global_context->setSensitiveDataMasker(std::make_unique<SensitiveDataMasker>(config(), "query_masking_rules"));
+        SensitiveDataMasker::setInstance(std::make_unique<SensitiveDataMasker>(config(), "query_masking_rules"));
     }
 
     auto main_config_reloader = std::make_unique<ConfigReloader>(config_path,
@@ -426,10 +425,6 @@ int Server::main(const std::vector<std::string> & /*args*/)
         {
             setTextLog(global_context->getTextLog());
             buildLoggers(*config, logger());
-            if (auto masker = global_context->getSensitiveDataMasker())
-            {
-                setLoggerSensitiveDataMasker(logger(), masker);
-            }
             global_context->setClustersConfig(config);
             global_context->setMacros(std::make_unique<Macros>(*config, "macros"));
         },

dbms/programs/server/TCPHandler.cpp

@@ -203,6 +203,43 @@ void TCPHandler::runImpl()
                 state.maybe_compressed_in.reset(); /// For more accurate accounting by MemoryTracker.
             });
 
+            /// Send structure of columns to client for function input()
+            query_context->setInputInitializer([this] (Context & context, const StoragePtr & input_storage)
+            {
+                if (&context != &query_context.value())
+                    throw Exception("Unexpected context in Input initializer", ErrorCodes::LOGICAL_ERROR);
+
+                state.need_receive_data_for_input = true;
+
+                /// Send ColumnsDescription for input storage.
+                if (client_revision >= DBMS_MIN_REVISION_WITH_COLUMN_DEFAULTS_METADATA
+                    && query_context->getSettingsRef().input_format_defaults_for_omitted_fields)
+                {
+                    sendTableColumns(input_storage->getColumns());
+                }
+
+                /// Send block to the client - input storage structure.
+                state.input_header = input_storage->getSampleBlock();
+                sendData(state.input_header);
+            });
+
+            query_context->setInputBlocksReaderCallback([&connection_settings, this] (Context & context) -> Block
+            {
+                if (&context != &query_context.value())
+                    throw Exception("Unexpected context in InputBlocksReader", ErrorCodes::LOGICAL_ERROR);
+
+                size_t poll_interval;
+                int receive_timeout;
+                std::tie(poll_interval, receive_timeout) = getReadTimeouts(connection_settings);
+                if (!readDataNext(poll_interval, receive_timeout))
+                {
+                    state.block_in.reset();
+                    state.maybe_compressed_in.reset();
+                    return Block();
+                }
+                return state.block_for_input;
+            });
+
             customizeContext(*query_context);
 
             bool may_have_embedded_data = client_revision >= DBMS_MIN_REVISION_WITH_CLIENT_SUPPORT_EMBEDDED_DATA;
@@ -218,6 +255,13 @@ void TCPHandler::runImpl()
             /// Does the request require receive data from client?
             if (state.need_receive_data_for_insert)
                 processInsertQuery(connection_settings);
+            else if (state.need_receive_data_for_input)
+            {
+                /// It is special case for input(), all works for reading data from client will be done in callbacks.
+                /// state.io.in is NullAndDoCopyBlockInputStream so read it once.
+                state.io.in->read();
+                state.io.onFinish();
+            }
             else if (state.io.pipeline.initialized())
                 processOrdinaryQueryWithProcessors(query_context->getSettingsRef().max_threads);
             else
@@ -324,7 +368,50 @@ void TCPHandler::runImpl()
 }
 
 
-void TCPHandler::readData(const Settings & connection_settings)
+bool TCPHandler::readDataNext(const size_t & poll_interval, const int & receive_timeout)
+{
+    Stopwatch watch(CLOCK_MONOTONIC_COARSE);
+
+    /// We are waiting for a packet from the client. Thus, every `POLL_INTERVAL` seconds check whether we need to shut down.
+    while (true)
+    {
+        if (static_cast<ReadBufferFromPocoSocket &>(*in).poll(poll_interval))
+            break;
+
+        /// Do we need to shut down?
+        if (server.isCancelled())
+            return false;
+
+        /** Have we waited for data for too long?
+         *  If we periodically poll, the receive_timeout of the socket itself does not work.
+         *  Therefore, an additional check is added.
+         */
+        double elapsed = watch.elapsedSeconds();
+        if (elapsed > receive_timeout)
+        {
+            std::stringstream ss;
+            ss << "Timeout exceeded while receiving data from client.";
+            ss << " Waited for " << static_cast<size_t>(elapsed) << " seconds,";
+            ss << " timeout is " << receive_timeout << " seconds.";
+
+            throw Exception(ss.str(), ErrorCodes::SOCKET_TIMEOUT);
+        }
+    }
+
+    /// If client disconnected.
+    if (in->eof())
+        return false;
+
+    /// We accept and process data. And if they are over, then we leave.
+    if (!receivePacket())
+        return false;
+
+    sendLogs();
+    return true;
+}
+
+
+std::tuple<size_t, int> TCPHandler::getReadTimeouts(const Settings & connection_settings)
 {
     const auto receive_timeout = query_context->getSettingsRef().receive_timeout.value;
 
@@ -334,48 +421,21 @@ void TCPHandler::readData(const Settings & connection_settings)
     constexpr size_t min_poll_interval = 5000; // 5 ms
     size_t poll_interval = std::max(min_poll_interval, std::min(default_poll_interval, current_poll_interval));
 
+    return std::make_tuple(poll_interval, receive_timeout.totalSeconds());
+}
+
+
+void TCPHandler::readData(const Settings & connection_settings)
+{
+    size_t poll_interval;
+    int receive_timeout;
+
+    std::tie(poll_interval, receive_timeout) = getReadTimeouts(connection_settings);
     sendLogs();
 
     while (true)
-    {
+        if (!readDataNext(poll_interval, receive_timeout))
-        Stopwatch watch(CLOCK_MONOTONIC_COARSE);
-
-        /// We are waiting for a packet from the client. Thus, every `POLL_INTERVAL` seconds check whether we need to shut down.
-        while (true)
-        {
-            if (static_cast<ReadBufferFromPocoSocket &>(*in).poll(poll_interval))
-                break;
-
-            /// Do we need to shut down?
-            if (server.isCancelled())
-                return;
-
-            /** Have we waited for data for too long?
-             *  If we periodically poll, the receive_timeout of the socket itself does not work.
-             *  Therefore, an additional check is added.
-             */
-            double elapsed = watch.elapsedSeconds();
-            if (elapsed > receive_timeout.totalSeconds())
-            {
-                std::stringstream ss;
-                ss << "Timeout exceeded while receiving data from client.";
-                ss << " Waited for " << static_cast<size_t>(elapsed) << " seconds,";
-                ss << " timeout is " << receive_timeout.totalSeconds() << " seconds.";
-
-                throw Exception(ss.str(), ErrorCodes::SOCKET_TIMEOUT);
-            }
-        }
-
-        /// If client disconnected.
-        if (in->eof())
             return;
-
-        /// We accept and process data. And if they are over, then we leave.
-        if (!receivePacket())
-            break;
-
-        sendLogs();
-    }
 }
 
 
@@ -904,7 +964,7 @@ bool TCPHandler::receiveData()
     {
         /// If there is an insert request, then the data should be written directly to `state.io.out`.
         /// Otherwise, we write the blocks in the temporary `external_table_name` table.
-        if (!state.need_receive_data_for_insert)
+        if (!state.need_receive_data_for_insert && !state.need_receive_data_for_input)
         {
             StoragePtr storage;
             /// If such a table does not exist, create it.
@@ -918,7 +978,9 @@ bool TCPHandler::receiveData()
             /// The data will be written directly to the table.
             state.io.out = storage->write(ASTPtr(), *query_context);
         }
-        if (block)
+        if (state.need_receive_data_for_input)
+            state.block_for_input = block;
+        else
             state.io.out->write(block);
         return true;
     }
@@ -960,6 +1022,8 @@ void TCPHandler::initBlockInput()
         Block header;
         if (state.io.out)
             header = state.io.out->getHeader();
+        else if (state.need_receive_data_for_input)
+            header = state.input_header;
 
         last_block_in.header = header;
         last_block_in.compression = state.compression;

dbms/programs/server/TCPHandler.h

@@ -64,6 +64,13 @@ struct QueryState
     /// Request requires data from the client (INSERT, but not INSERT SELECT).
     bool need_receive_data_for_insert = false;
 
+    /// Request requires data from client for function input()
+    bool need_receive_data_for_input = false;
+    /// temporary place for incoming data block for input()
+    Block block_for_input;
+    /// sample block from StorageInput
+    Block input_header;
+
     /// To output progress, the difference after the previous sending of progress.
     Progress progress;
 
@@ -147,7 +154,9 @@ private:
     bool receivePacket();
     void receiveQuery();
     bool receiveData();
+    bool readDataNext(const size_t & poll_interval, const int & receive_timeout);
     void readData(const Settings & global_settings);
+    std::tuple<size_t, int> getReadTimeouts(const Settings & global_settings);
 
     [[noreturn]] void receiveUnexpectedData();
     [[noreturn]] void receiveUnexpectedQuery();

dbms/programs/server/config.d/query_masking_rules.xml

@@ -1,3 +1,5 @@
+<?xml version="1.0"?>
+<!-- Config for test server -->
 <yandex>
     <query_masking_rules>
         <!-- An example of query masking rule to remove profanity.
@@ -8,5 +10,10 @@
             <regexp>(?i:shit)</regexp>
             <replace>substance</replace>
         </rule>
+        <!-- This rule is used in tests. -->
+        <rule>
+            <regexp>TOPSECRET.TOPSECRET</regexp>
+            <replace>[hidden]</replace>
+        </rule>
     </query_masking_rules>
 </yandex>

dbms/programs/server/config.xml

@@ -447,7 +447,7 @@
     <query_masking_rules>
         <rule>
             <name>hide SSN</name>
-            <regexp>(^|\D)\d{3}-\d{2}-\d{4}($|\D)</regexp>
+            <regexp>\b\d{3}-\d{2}-\d{4}\b</regexp>
             <replace>000-00-0000</replace>
         </rule>
     </query_masking_rules>

dbms/src/AggregateFunctions/AggregateFunctionGroupUniqArray.h

@@ -10,6 +10,7 @@
 #include <Columns/ColumnArray.h>
 
 #include <Common/HashTable/HashSet.h>
+#include <Common/HashTable/HashTableKeyHolder.h>
 #include <Common/assert_cast.h>
 
 #include <AggregateFunctions/IAggregateFunction.h>
@@ -132,11 +133,6 @@ struct AggregateFunctionGroupUniqArrayGenericData
     Set value;
 };
 
-
-/// Helper function for deserialize and insert for the class AggregateFunctionGroupUniqArrayGeneric
-template <bool is_plain_column>
-static StringRef getSerializationImpl(const IColumn & column, size_t row_num, Arena & arena);
-
 template <bool is_plain_column>
 static void deserializeAndInsertImpl(StringRef str, IColumn & data_to);
 
@@ -154,9 +150,18 @@ class AggregateFunctionGroupUniqArrayGeneric
 
     using State = AggregateFunctionGroupUniqArrayGenericData;
 
-    static StringRef getSerialization(const IColumn & column, size_t row_num, Arena & arena)
+    static auto getKeyHolder(const IColumn & column, size_t row_num, Arena & arena)
     {
-        return getSerializationImpl<is_plain_column>(column, row_num, arena);
+        if constexpr (is_plain_column)
+        {
+            return ArenaKeyHolder{column.getDataAt(row_num), arena};
+        }
+        else
+        {
+            const char * begin = nullptr;
+            StringRef serialized = column.serializeValueIntoArena(row_num, arena, begin);
+            return SerializedKeyHolder{serialized, arena};
+        }
     }
 
     static void deserializeAndInsert(StringRef str, IColumn & data_to)
@@ -209,26 +214,13 @@ public:
     void add(AggregateDataPtr place, const IColumn ** columns, size_t row_num, Arena * arena) const override
     {
         auto & set = this->data(place).value;
+        if (limit_num_elems && set.size() >= max_elems)
+            return;
 
         bool inserted;
         State::Set::iterator it;
-
+        auto key_holder = getKeyHolder(*columns[0], row_num, *arena);
-        if (limit_num_elems && set.size() >= max_elems)
+        set.emplace(key_holder, it, inserted);
-            return;
-        StringRef str_serialized = getSerialization(*columns[0], row_num, *arena);
-
-        set.emplace(str_serialized, it, inserted);
-
-        if constexpr (!is_plain_column)
-        {
-            if (!inserted)
-                arena->rollback(str_serialized.size);
-        }
-        else
-        {
-            if (inserted)
-                it->getValueMutable().data = arena->insert(str_serialized.data, str_serialized.size);
-        }
     }
 
     void merge(AggregateDataPtr place, ConstAggregateDataPtr rhs, Arena * arena) const override
@@ -241,15 +233,11 @@ public:
         for (auto & rhs_elem : rhs_set)
         {
             if (limit_num_elems && cur_set.size() >= max_elems)
-                return ;
+                return;
-            cur_set.emplace(rhs_elem.getValue(), it, inserted);
+
-            if (inserted)
+            // We have to copy the keys to our arena.
-            {
+            assert(arena != nullptr);
-                if (it->getValue().size)
+            cur_set.emplace(ArenaKeyHolder{rhs_elem.getValue(), *arena}, it, inserted);
-                    it->getValueMutable().data = arena->insert(it->getValue().data, it->getValue().size);
-                else
-                    it->getValueMutable().data = nullptr;
-            }
         }
     }
 
@@ -271,20 +259,6 @@ public:
     const char * getHeaderFilePath() const override { return __FILE__; }
 };
 
-
-template <>
-inline StringRef getSerializationImpl<false>(const IColumn & column, size_t row_num, Arena & arena)
-{
-    const char * begin = nullptr;
-    return column.serializeValueIntoArena(row_num, arena, begin);
-}
-
-template <>
-inline StringRef getSerializationImpl<true>(const IColumn & column, size_t row_num, Arena &)
-{
-    return column.getDataAt(row_num);
-}
-
 template <>
 inline void deserializeAndInsertImpl<false>(StringRef str, IColumn & data_to)
 {

dbms/src/Client/ConnectionPoolWithFailover.cpp

@@ -31,8 +31,9 @@ namespace ErrorCodes
 ConnectionPoolWithFailover::ConnectionPoolWithFailover(
         ConnectionPoolPtrs nested_pools_,
         LoadBalancing load_balancing,
-        time_t decrease_error_period_)
+        time_t decrease_error_period_,
-    : Base(std::move(nested_pools_), decrease_error_period_, &Logger::get("ConnectionPoolWithFailover"))
+        size_t max_error_cap_)
+    : Base(std::move(nested_pools_), decrease_error_period_, max_error_cap_, &Logger::get("ConnectionPoolWithFailover"))
     , default_load_balancing(load_balancing)
 {
     const std::string & local_hostname = getFQDNOrHostName();
@@ -73,6 +74,31 @@ IConnectionPool::Entry ConnectionPoolWithFailover::get(const ConnectionTimeouts
     return Base::get(try_get_entry, get_priority);
 }
 
+ConnectionPoolWithFailover::Status ConnectionPoolWithFailover::getStatus() const
+{
+    const Base::PoolStates states = getPoolStates();
+    const Base::NestedPools pools = nested_pools;
+    assert(states.size() == pools.size());
+
+    ConnectionPoolWithFailover::Status result;
+    result.reserve(states.size());
+    const time_t since_last_error_decrease = time(nullptr) - last_error_decrease_time;
+
+    for (size_t i = 0; i < states.size(); ++i)
+    {
+        const auto rounds_to_zero_errors = states[i].error_count ? bitScanReverse(states[i].error_count) + 1 : 0;
+        const auto seconds_to_zero_errors = std::max(static_cast<time_t>(0), rounds_to_zero_errors * decrease_error_period - since_last_error_decrease);
+
+        result.emplace_back(NestedPoolStatus{
+            pools[i].get(),
+            states[i].error_count,
+            std::chrono::seconds{seconds_to_zero_errors}
+        });
+    }
+
+    return result;
+}
+
 std::vector<IConnectionPool::Entry> ConnectionPoolWithFailover::getMany(const ConnectionTimeouts & timeouts,
                                                                         const Settings * settings,
                                                                         PoolMode pool_mode)

dbms/src/Client/ConnectionPoolWithFailover.h

@@ -3,6 +3,9 @@
 #include <Common/PoolWithFailoverBase.h>
 #include <Client/ConnectionPool.h>
 
+#include <chrono>
+#include <vector>
+
 
 namespace DB
 {
@@ -34,7 +37,8 @@ public:
     ConnectionPoolWithFailover(
             ConnectionPoolPtrs nested_pools_,
             LoadBalancing load_balancing,
-            time_t decrease_error_period_ = DBMS_CONNECTION_POOL_WITH_FAILOVER_DEFAULT_DECREASE_ERROR_PERIOD);
+            time_t decrease_error_period_ = DBMS_CONNECTION_POOL_WITH_FAILOVER_DEFAULT_DECREASE_ERROR_PERIOD,
+            size_t max_error_cap = DBMS_CONNECTION_POOL_WITH_FAILOVER_MAX_ERROR_COUNT);
 
     using Entry = IConnectionPool::Entry;
 
@@ -64,6 +68,16 @@ public:
             PoolMode pool_mode,
             const QualifiedTableName & table_to_check);
 
+    struct NestedPoolStatus
+    {
+        const IConnectionPool * pool;
+        size_t error_count;
+        std::chrono::seconds estimated_recovery_time;
+    };
+
+    using Status = std::vector<NestedPoolStatus>;
+    Status getStatus() const;
+
 private:
     /// Get the values of relevant settings and call Base::getMany()
     std::vector<TryResult> getManyImpl(

dbms/src/Common/Arena.h

@@ -170,11 +170,14 @@ public:
 
     /** Rollback just performed allocation.
       * Must pass size not more that was just allocated.
+	  * Return the resulting head pointer, so that the caller can assert that
+	  * the allocation it intended to roll back was indeed the last one.
       */
-    void rollback(size_t size)
+    void * rollback(size_t size)
     {
         head->pos -= size;
         ASAN_POISON_MEMORY_REGION(head->pos, size + pad_right);
+        return head->pos;
     }
 
     /** Begin or expand allocation of contiguous piece of memory without alignment.

dbms/src/Common/ColumnsHashing.h

@@ -1,6 +1,8 @@
 #pragma once
 
 
+#include <Common/HashTable/HashTable.h>
+#include <Common/HashTable/HashTableKeyHolder.h>
 #include <Common/ColumnsHashingImpl.h>
 #include <Common/Arena.h>
 #include <Common/LRUCache.h>
@@ -57,7 +59,7 @@ struct HashMethodOneNumber
     using Base::getHash; /// (const Data & data, size_t row, Arena & pool) -> size_t
 
     /// Is used for default implementation in HashMethodBase.
-    FieldType getKey(size_t row, Arena &) const { return unalignedLoad<FieldType>(vec + row * sizeof(FieldType)); }
+    FieldType getKeyHolder(size_t row, Arena &) const { return unalignedLoad<FieldType>(vec + row * sizeof(FieldType)); }
 
     /// Get StringRef from value which can be inserted into column.
     static StringRef getValueRef(const Value & value)
@@ -86,24 +88,24 @@ struct HashMethodString
         chars = column_string.getChars().data();
     }
 
-    auto getKey(ssize_t row, Arena &) const
+    auto getKeyHolder(ssize_t row, [[maybe_unused]] Arena & pool) const
     {
-        return StringRef(chars + offsets[row - 1], offsets[row] - offsets[row - 1] - 1);
+        StringRef key(chars + offsets[row - 1], offsets[row] - offsets[row - 1] - 1);
+
+        if constexpr (place_string_to_arena)
+        {
+            return ArenaKeyHolder{key, pool};
+        }
+        else
+        {
+            return key;
+        }
     }
 
     static StringRef getValueRef(const Value & value) { return value.first; }
 
 protected:
     friend class columns_hashing_impl::HashMethodBase<Self, Value, Mapped, use_cache>;
-
-    static ALWAYS_INLINE void onNewKey([[maybe_unused]] StringRef & key, [[maybe_unused]] Arena & pool)
-    {
-        if constexpr (place_string_to_arena)
-        {
-            if (key.size)
-                key.data = pool.insert(key.data, key.size);
-        }
-    }
 };
 
 
@@ -126,17 +128,24 @@ struct HashMethodFixedString
         chars = &column_string.getChars();
     }
 
-    StringRef getKey(size_t row, Arena &) const { return StringRef(&(*chars)[row * n], n); }
+    auto getKeyHolder(size_t row, [[maybe_unused]] Arena & pool) const
+    {
+        StringRef key(&(*chars)[row * n], n);
+
+        if constexpr (place_string_to_arena)
+        {
+            return ArenaKeyHolder{key, pool};
+        }
+        else
+        {
+            return key;
+        }
+    }
 
     static StringRef getValueRef(const Value & value) { return value.first; }
 
 protected:
     friend class columns_hashing_impl::HashMethodBase<Self, Value, Mapped, use_cache>;
-    static ALWAYS_INLINE void onNewKey([[maybe_unused]] StringRef & key, [[maybe_unused]] Arena & pool)
-    {
-        if constexpr (place_string_to_arena)
-            key.data = pool.insert(key.data, key.size);
-    }
 };
 
 
@@ -316,10 +325,10 @@ struct HashMethodSingleLowCardinalityColumn : public SingleColumnMethod
         }
     }
 
-    /// Get the key from the key columns for insertion into the hash table.
+    /// Get the key holder from the key columns for insertion into the hash table.
-    ALWAYS_INLINE auto getKey(size_t row, Arena & pool) const
+    ALWAYS_INLINE auto getKeyHolder(size_t row, Arena & pool) const
     {
-        return Base::getKey(getIndexAt(row), pool);
+        return Base::getKeyHolder(getIndexAt(row), pool);
     }
 
     template <typename Data>
@@ -347,30 +356,23 @@ struct HashMethodSingleLowCardinalityColumn : public SingleColumnMethod
                 return EmplaceResult(false);
         }
 
-        auto key = getKey(row_, pool);
+        auto key_holder = getKeyHolder(row_, pool);
 
         bool inserted = false;
         typename Data::iterator it;
         if (saved_hash)
-            data.emplace(key, it, inserted, saved_hash[row]);
+            data.emplace(key_holder, it, inserted, saved_hash[row]);
         else
-            data.emplace(key, it, inserted);
+            data.emplace(key_holder, it, inserted);
 
         visit_cache[row] = VisitValue::Found;
 
-        if (inserted)
-        {
-            if constexpr (has_mapped)
-            {
-                new(&it->getSecond()) Mapped();
-                Base::onNewKey(it->getFirstMutable(), pool);
-            }
-            else
-                Base::onNewKey(*it, pool);
-        }
-
         if constexpr (has_mapped)
         {
+            if (inserted)
+            {
+                new (&it->getSecond()) Mapped();
+            }
             mapped_cache[row] = it->getSecond();
             return EmplaceResult(it->getSecond(), mapped_cache[row], inserted);
         }
@@ -407,13 +409,13 @@ struct HashMethodSingleLowCardinalityColumn : public SingleColumnMethod
                 return FindResult(visit_cache[row] == VisitValue::Found);
         }
 
-        auto key = getKey(row_, pool);
+        auto key_holder = getKeyHolder(row_, pool);
 
         typename Data::iterator it;
         if (saved_hash)
-            it = data.find(key, saved_hash[row]);
+            it = data.find(*key_holder, saved_hash[row]);
         else
-            it = data.find(key);
+            it = data.find(*key_holder);
 
         bool found = it != data.end();
         visit_cache[row] = found ? VisitValue::Found : VisitValue::NotFound;
@@ -493,7 +495,7 @@ struct HashMethodKeysFixed
         }
     }
 
-    ALWAYS_INLINE Key getKey(size_t row, Arena &) const
+    ALWAYS_INLINE Key getKeyHolder(size_t row, Arena &) const
     {
         if constexpr (has_nullable_keys)
         {
@@ -532,12 +534,12 @@ struct HashMethodSerialized
 protected:
     friend class columns_hashing_impl::HashMethodBase<Self, Value, Mapped, false>;
 
-    ALWAYS_INLINE StringRef getKey(size_t row, Arena & pool) const
+    ALWAYS_INLINE SerializedKeyHolder getKeyHolder(size_t row, Arena & pool) const
     {
-        return serializeKeysToPoolContiguous(row, keys_size, key_columns, pool);
+        return SerializedKeyHolder{
+            serializeKeysToPoolContiguous(row, keys_size, key_columns, pool),
+            pool};
     }
-
-    static ALWAYS_INLINE void onExistingKey(StringRef & key, Arena & pool) { pool.rollback(key.size); }
 };
 
 /// For the case when there is one string key.
@@ -554,7 +556,10 @@ struct HashMethodHashed
     HashMethodHashed(ColumnRawPtrs key_columns_, const Sizes &, const HashMethodContextPtr &)
         : key_columns(std::move(key_columns_)) {}
 
-    ALWAYS_INLINE Key getKey(size_t row, Arena &) const { return hash128(row, key_columns.size(), key_columns); }
+    ALWAYS_INLINE Key getKeyHolder(size_t row, Arena &) const
+    {
+        return hash128(row, key_columns.size(), key_columns);
+    }
 
     static ALWAYS_INLINE StringRef getValueRef(const Value & value)
     {

dbms/src/Common/ColumnsHashingImpl.h

@@ -2,6 +2,7 @@
 
 #include <Columns/IColumn.h>
 #include <Common/assert_cast.h>
+#include <Common/HashTable/HashTableKeyHolder.h>
 #include <Interpreters/AggregationCommon.h>
 
 
@@ -117,26 +118,22 @@ public:
     template <typename Data>
     ALWAYS_INLINE EmplaceResult emplaceKey(Data & data, size_t row, Arena & pool)
     {
-        auto key = static_cast<Derived &>(*this).getKey(row, pool);
+        auto key_holder = static_cast<Derived &>(*this).getKeyHolder(row, pool);
-        return emplaceKeyImpl(key, data, pool);
+        return emplaceImpl(key_holder, data);
     }
 
     template <typename Data>
     ALWAYS_INLINE FindResult findKey(Data & data, size_t row, Arena & pool)
     {
-        auto key = static_cast<Derived &>(*this).getKey(row, pool);
+        auto key_holder = static_cast<Derived &>(*this).getKeyHolder(row, pool);
-        auto res = findKeyImpl(key, data);
+        return findKeyImpl(keyHolderGetKey(key_holder), data);
-        static_cast<Derived &>(*this).onExistingKey(key, pool);
-        return res;
     }
 
     template <typename Data>
     ALWAYS_INLINE size_t getHash(const Data & data, size_t row, Arena & pool)
     {
-        auto key = static_cast<Derived &>(*this).getKey(row, pool);
+        auto key_holder = static_cast<Derived &>(*this).getKeyHolder(row, pool);
-        auto res = data.hash(key);
+        return data.hash(keyHolderGetKey(key_holder));
-        static_cast<Derived &>(*this).onExistingKey(key, pool);
-        return res;
     }
 
 protected:
@@ -157,20 +154,13 @@ protected:
         }
     }
 
-    template <typename Key>
+    template <typename Data, typename KeyHolder>
-    static ALWAYS_INLINE void onNewKey(Key & /*key*/, Arena & /*pool*/) {}
+    ALWAYS_INLINE EmplaceResult emplaceImpl(KeyHolder & key_holder, Data & data)
-    template <typename Key>
-    static ALWAYS_INLINE void onExistingKey(Key & /*key*/, Arena & /*pool*/) {}
-
-    template <typename Data, typename Key>
-    ALWAYS_INLINE EmplaceResult emplaceKeyImpl(Key key, Data & data, Arena & pool)
     {
         if constexpr (Cache::consecutive_keys_optimization)
         {
-            if (cache.found && cache.check(key))
+            if (cache.found && cache.check(keyHolderGetKey(key_holder)))
             {
-                static_cast<Derived &>(*this).onExistingKey(key, pool);
-
                 if constexpr (has_mapped)
                     return EmplaceResult(cache.value.second, cache.value.second, false);
                 else
@@ -180,7 +170,7 @@ protected:
 
         typename Data::iterator it;
         bool inserted = false;
-        data.emplace(key, it, inserted);
+        data.emplace(key_holder, it, inserted);
 
         [[maybe_unused]] Mapped * cached = nullptr;
         if constexpr (has_mapped)
@@ -191,13 +181,8 @@ protected:
             if constexpr (has_mapped)
             {
                 new(&it->getSecond()) Mapped();
-                static_cast<Derived &>(*this).onNewKey(it->getFirstMutable(), pool);
             }
-            else
-                static_cast<Derived &>(*this).onNewKey(it->getValueMutable(), pool);
         }
-        else
-            static_cast<Derived &>(*this).onExistingKey(key, pool);
 
         if constexpr (consecutive_keys_optimization)
         {

dbms/src/Common/ErrorCodes.cpp

@@ -449,6 +449,9 @@ namespace ErrorCodes
     extern const int READONLY_SETTING = 472;
     extern const int DEADLOCK_AVOIDED = 473;
     extern const int INVALID_TEMPLATE_FORMAT = 474;
+    extern const int INVALID_WITH_FILL_EXPRESSION = 475;
+    extern const int WITH_TIES_WITHOUT_ORDER_BY = 476;
+    extern const int INVALID_USAGE_OF_INPUT = 477;
 
     extern const int KEEPER_EXCEPTION = 999;
     extern const int POCO_EXCEPTION = 1000;

dbms/src/Common/HashTable/FixedHashTable.h

@@ -262,8 +262,9 @@ public:
     iterator end() { return iterator(this, buf + BUFFER_SIZE); }
 
 
-protected:
+public:
-    void ALWAYS_INLINE emplaceImpl(Key x, iterator & it, bool & inserted)
+    /// The last parameter is unused but exists for compatibility with HashTable interface.
+    void ALWAYS_INLINE emplace(Key x, iterator & it, bool & inserted, size_t /* hash */ = 0)
     {
         it = iterator(this, &buf[x]);
 
@@ -278,22 +279,16 @@ protected:
         ++m_size;
     }
 
-
-public:
     std::pair<iterator, bool> ALWAYS_INLINE insert(const value_type & x)
     {
         std::pair<iterator, bool> res;
-        emplaceImpl(Cell::getKey(x), res.first, res.second);
+        emplace(Cell::getKey(x), res.first, res.second);
         if (res.second)
             res.first.ptr->setMapped(x);
 
         return res;
     }
 
-
-    void ALWAYS_INLINE emplace(Key x, iterator & it, bool & inserted) { emplaceImpl(x, it, inserted); }
-    void ALWAYS_INLINE emplace(Key x, iterator & it, bool & inserted, size_t) { emplaceImpl(x, it, inserted); }
-
     iterator ALWAYS_INLINE find(Key x)
     {
         return !buf[x].isZero(*this) ? iterator(this, &buf[x]) : end();

dbms/src/Common/HashTable/HashTable.h

@@ -21,6 +21,7 @@
 #include <IO/VarInt.h>
 
 #include <Common/HashTable/HashTableAllocator.h>
+#include <Common/HashTable/HashTableKeyHolder.h>
 
 #ifdef DBMS_HASH_MAP_DEBUG_RESIZES
     #include <iostream>
@@ -630,6 +631,8 @@ protected:
 
 
     /// If the key is zero, insert it into a special place and return true.
+    /// We don't have to persist a zero key, because it's not actually inserted.
+    /// That's why we just take a Key by value, an not a key holder.
     bool ALWAYS_INLINE emplaceIfZero(Key x, iterator & it, bool & inserted, size_t hash_value)
     {
         /// If it is claimed that the zero key can not be inserted into the table.
@@ -655,17 +658,23 @@ protected:
         return false;
     }
 
-    void ALWAYS_INLINE emplaceNonZeroImpl(size_t place_value, Key x, iterator & it, bool & inserted, size_t hash_value)
+    template <typename KeyHolder>
+    void ALWAYS_INLINE emplaceNonZeroImpl(size_t place_value, KeyHolder && key_holder,
+                                          iterator & it, bool & inserted, size_t hash_value)
     {
         it = iterator(this, &buf[place_value]);
 
         if (!buf[place_value].isZero(*this))
         {
+            keyHolderDiscardKey(key_holder);
             inserted = false;
             return;
         }
 
-        new(&buf[place_value]) Cell(x, *this);
+        keyHolderPersistKey(key_holder);
+        const auto & key = keyHolderGetKey(key_holder);
+
+        new(&buf[place_value]) Cell(key, *this);
         buf[place_value].setHash(hash_value);
         inserted = true;
         ++m_size;
@@ -687,19 +696,21 @@ protected:
                 throw;
             }
 
-            it = find(x, hash_value);
+            it = find(keyHolderGetKey(key_holder), hash_value);
         }
     }
 
     /// Only for non-zero keys. Find the right place, insert the key there, if it does not already exist. Set iterator to the cell in output parameter.
-    void ALWAYS_INLINE emplaceNonZero(Key x, iterator & it, bool & inserted, size_t hash_value)
+    template <typename KeyHolder>
+    void ALWAYS_INLINE emplaceNonZero(KeyHolder && key_holder, iterator & it,
+                                      bool & inserted, size_t hash_value)
     {
-        size_t place_value = findCell(x, hash_value, grower.place(hash_value));
+        const auto & key = keyHolderGetKey(key_holder);
-        emplaceNonZeroImpl(place_value, x, it, inserted, hash_value);
+        size_t place_value = findCell(key, hash_value, grower.place(hash_value));
+        emplaceNonZeroImpl(place_value, key_holder, it, inserted, hash_value);
     }
 
 
-
 public:
     /// Insert a value. In the case of any more complex values, it is better to use the `emplace` function.
     std::pair<iterator, bool> ALWAYS_INLINE insert(const value_type & x)
@@ -708,7 +719,9 @@ public:
 
         size_t hash_value = hash(Cell::getKey(x));
         if (!emplaceIfZero(Cell::getKey(x), res.first, res.second, hash_value))
+        {
             emplaceNonZero(Cell::getKey(x), res.first, res.second, hash_value);
+        }
 
         if (res.second)
             res.first.ptr->setMapped(x);
@@ -739,19 +752,20 @@ public:
       * if (inserted)
       *     new(&it->second) Mapped(value);
       */
-    void ALWAYS_INLINE emplace(Key x, iterator & it, bool & inserted)
+    template <typename KeyHolder>
+    void ALWAYS_INLINE emplace(KeyHolder && key_holder, iterator & it, bool & inserted)
     {
-        size_t hash_value = hash(x);
+        const auto & key = keyHolderGetKey(key_holder);
-        if (!emplaceIfZero(x, it, inserted, hash_value))
+        emplace(key_holder, it, inserted, hash(key));
-            emplaceNonZero(x, it, inserted, hash_value);
     }
 
-
+    template <typename KeyHolder>
-    /// Same, but with a precalculated value of hash function.
+    void ALWAYS_INLINE emplace(KeyHolder && key_holder, iterator & it,
-    void ALWAYS_INLINE emplace(Key x, iterator & it, bool & inserted, size_t hash_value)
+                                  bool & inserted, size_t hash_value)
     {
-        if (!emplaceIfZero(x, it, inserted, hash_value))
+        const auto & key = keyHolderGetKey(key_holder);
-            emplaceNonZero(x, it, inserted, hash_value);
+        if (!emplaceIfZero(key, it, inserted, hash_value))
+            emplaceNonZero(key_holder, it, inserted, hash_value);
     }
 
     /// Copy the cell from another hash table. It is assumed that the cell is not zero, and also that there was no such key in the table yet.

dbms/src/Common/HashTable/HashTableKeyHolder.h

@@ -0,0 +1,130 @@
+#pragma once
+
+#include <Common/Arena.h>
+
+/**
+  * In some aggregation scenarios, when adding a key to the hash table, we
+  * start with a temporary key object, and if it turns out to be a new key,
+  * we must make it persistent (e.g. copy to an Arena) and use the resulting
+  * persistent object as hash table key. This happens only for StringRef keys,
+  * because other key types are stored by value, but StringRef is a pointer-like
+  * type: the actual data are stored elsewhere. Even for StringRef, we don't
+  * make a persistent copy of the key in each of the following cases:
+  * 1) the aggregation method doesn't use temporary keys, so they're persistent
+  *    from the start;
+  * 1) the key is already present in the hash table;
+  * 3) that particular key is stored by value, e.g. a short StringRef key in
+  *    StringHashMap.
+  *
+  * In the past, the caller was responsible for making the key persistent after
+  * in was inserted. emplace() returned whether the key is new or not, so the
+  * caller only stored new keys (this is case (2) from the above list). However,
+  * now we are adding a compound hash table for StringRef keys, so case (3)
+  * appears. The decision about persistence now depends on some properties of
+  * the key, and the logic of this decision is tied to the particular hash table
+  * implementation. This means that the hash table user now doesn't have enough
+  * data and logic to make this decision by itself.
+  *
+  * To support these new requirements, we now manage key persistence by passing
+  * a special key holder to emplace(), which has the functions to make the key
+  * persistent or to discard it. emplace() then calls these functions at the
+  * appropriate moments.
+  *
+  * This approach has the following benefits:
+  * - no extra runtime branches in the caller to make the key persistent.
+  * - no additional data is stored in the hash table itself, which is important
+  *   when it's used in aggregate function states.
+  * - no overhead when the key memory management isn't needed: we just pass the
+  *   bare key without any wrapper to emplace(), and the default callbacks do
+  *   nothing.
+  *
+  * This file defines the default key persistence functions, as well as two
+  * different key holders and corresponding functions for storing StringRef
+  * keys to Arena.
+  */
+
+/**
+  * Returns the key. Can return the temporary key initially.
+  * After the call to keyHolderPersistKey(), must return the persistent key.
+  */
+template <typename Key>
+inline Key & ALWAYS_INLINE keyHolderGetKey(Key && key) { return key; }
+
+/**
+  * Make the key persistent. keyHolderGetKey() must return the persistent key
+  * after this call.
+  */
+template <typename Key>
+inline void ALWAYS_INLINE keyHolderPersistKey(Key &&) {}
+
+/**
+  * Discard the key. Calling keyHolderGetKey() is ill-defined after this.
+  */
+template <typename Key>
+inline void ALWAYS_INLINE keyHolderDiscardKey(Key &&) {}
+
+namespace DB
+{
+
+/**
+  * ArenaKeyHolder is a key holder for hash tables that serializes a StringRef
+  * key to an Arena.
+  */
+struct ArenaKeyHolder
+{
+    StringRef key;
+    Arena & pool;
+
+};
+
+}
+
+inline StringRef & ALWAYS_INLINE keyHolderGetKey(DB::ArenaKeyHolder & holder)
+{
+    return holder.key;
+}
+
+inline void ALWAYS_INLINE keyHolderPersistKey(DB::ArenaKeyHolder & holder)
+{
+    // Hash table shouldn't ask us to persist a zero key
+    assert(holder.key.size > 0);
+    holder.key.data = holder.pool.insert(holder.key.data, holder.key.size);
+}
+
+inline void ALWAYS_INLINE keyHolderDiscardKey(DB::ArenaKeyHolder &)
+{
+}
+
+namespace DB
+{
+
+/**
+  * SerializedKeyHolder is a key holder for a StringRef key that is already
+  * serialized to an Arena. The key must be the last allocation in this Arena,
+  * and is discarded by rolling back the allocation.
+  */
+struct SerializedKeyHolder
+{
+    StringRef key;
+    Arena & pool;
+};
+
+}
+
+inline StringRef & ALWAYS_INLINE keyHolderGetKey(DB::SerializedKeyHolder & holder)
+{
+    return holder.key;
+}
+
+inline void ALWAYS_INLINE keyHolderPersistKey(DB::SerializedKeyHolder &)
+{
+}
+
+inline void ALWAYS_INLINE keyHolderDiscardKey(DB::SerializedKeyHolder & holder)
+{
+    [[maybe_unused]] void * new_head = holder.pool.rollback(holder.key.size);
+    assert(new_head == holder.key.data);
+    holder.key.data = nullptr;
+    holder.key.size = 0;
+}
+

dbms/src/Common/HashTable/TwoLevelHashTable.h

@@ -235,19 +235,22 @@ public:
       * if (inserted)
       *     new(&it->second) Mapped(value);
       */
-    void ALWAYS_INLINE emplace(Key x, iterator & it, bool & inserted)
+    template <typename KeyHolder>
+    void ALWAYS_INLINE emplace(KeyHolder && key_holder, iterator & it, bool & inserted)
     {
-        size_t hash_value = hash(x);
+        size_t hash_value = hash(keyHolderGetKey(key_holder));
-        emplace(x, it, inserted, hash_value);
+        emplace(key_holder, it, inserted, hash_value);
     }
 
 
     /// Same, but with a precalculated values of hash function.
-    void ALWAYS_INLINE emplace(Key x, iterator & it, bool & inserted, size_t hash_value)
+    template <typename KeyHolder>
+    void ALWAYS_INLINE emplace(KeyHolder && key_holder, iterator & it,
+                                  bool & inserted, size_t hash_value)
     {
         size_t buck = getBucketFromHash(hash_value);
         typename Impl::iterator impl_it;
-        impls[buck].emplace(x, impl_it, inserted, hash_value);
+        impls[buck].emplace(key_holder, impl_it, inserted, hash_value);
         it = iterator(this, buck, impl_it);
     }
 

dbms/src/Common/PoolWithFailoverBase.h

@@ -56,9 +56,11 @@ public:
     PoolWithFailoverBase(
             NestedPools nested_pools_,
             time_t decrease_error_period_,
+            size_t max_error_cap_,
             Logger * log_)
         : nested_pools(std::move(nested_pools_))
         , decrease_error_period(decrease_error_period_)
+        , max_error_cap(max_error_cap_)
         , shared_pool_states(nested_pools.size())
         , log(log_)
     {
@@ -120,12 +122,14 @@ protected:
 
     /// This function returns a copy of pool states to avoid race conditions when modifying shared pool states.
     PoolStates updatePoolStates();
+    PoolStates getPoolStates() const;
 
     NestedPools nested_pools;
 
     const time_t decrease_error_period;
+    const size_t max_error_cap;
 
-    std::mutex pool_states_mutex;
+    mutable std::mutex pool_states_mutex;
     PoolStates shared_pool_states;
     /// The time when error counts were last decreased.
     time_t last_error_decrease_time = 0;
@@ -193,7 +197,10 @@ PoolWithFailoverBase<TNestedPool>::getMany(
     {
         std::lock_guard lock(pool_states_mutex);
         for (const ShuffledPool & pool: shuffled_pools)
-            shared_pool_states[pool.index].error_count += pool.error_count;
+        {
+            auto & pool_state = shared_pool_states[pool.index];
+            pool_state.error_count = std::min(max_error_cap, pool_state.error_count + pool.error_count);
+        }
     });
 
     std::string fail_messages;
@@ -236,7 +243,7 @@ PoolWithFailoverBase<TNestedPool>::getMany(
                             << (shuffled_pool.error_count + 1) << ", reason: " << fail_message);
                 ProfileEvents::increment(ProfileEvents::DistributedConnectionFailTry);
 
-                ++shuffled_pool.error_count;
+                shuffled_pool.error_count = std::min(max_error_cap, shuffled_pool.error_count + 1);
 
                 if (shuffled_pool.error_count >= max_tries)
                 {
@@ -297,7 +304,8 @@ void PoolWithFailoverBase<TNestedPool>::reportError(const Entry & entry)
         if (nested_pools[i]->contains(entry))
         {
             std::lock_guard lock(pool_states_mutex);
-            ++shared_pool_states[i].error_count;
+            auto & pool_state = shared_pool_states[i];
+            pool_state.error_count = std::min(max_error_cap, pool_state.error_count + 1);
             return;
         }
     }
@@ -373,3 +381,11 @@ PoolWithFailoverBase<TNestedPool>::updatePoolStates()
     }
     return result;
 }
+
+template <typename TNestedPool>
+typename PoolWithFailoverBase<TNestedPool>::PoolStates
+PoolWithFailoverBase<TNestedPool>::getPoolStates() const
+{
+    std::lock_guard lock(pool_states_mutex);
+    return shared_pool_states;
+}

dbms/src/Common/SensitiveDataMasker.cpp

@@ -24,6 +24,7 @@ namespace DB
 namespace ErrorCodes
 {
     extern const int CANNOT_COMPILE_REGEXP;
+    extern const int LOGICAL_ERROR;
     extern const int NO_ELEMENTS_IN_CONFIG;
     extern const int INVALID_CONFIG_PARAMETER;
 }
@@ -38,7 +39,9 @@ private:
     const RE2 regexp;
     const re2::StringPiece replacement;
 
+#ifndef NDEBUG
     mutable std::atomic<std::uint64_t> matches_count = 0;
+#endif
 
 public:
     //* TODO: option with hyperscan? https://software.intel.com/en-us/articles/why-and-how-to-replace-pcre-with-hyperscan
@@ -61,15 +64,37 @@ public:
     uint64_t apply(std::string & data) const
     {
         auto m = RE2::GlobalReplace(&data, regexp, replacement);
+#ifndef NDEBUG
         matches_count += m;
+#endif
         return m;
     }
 
     const std::string & getName() const { return name; }
     const std::string & getReplacementString() const { return replacement_string; }
+#ifndef NDEBUG
     uint64_t getMatchesCount() const { return matches_count; }
+#endif
+
 };
 
+std::unique_ptr<SensitiveDataMasker> SensitiveDataMasker::sensitive_data_masker = nullptr;
+
+void SensitiveDataMasker::setInstance(std::unique_ptr<SensitiveDataMasker> sensitive_data_masker_)
+{
+    if (!sensitive_data_masker_)
+        throw Exception("Logical error: the 'sensitive_data_masker' is not set", ErrorCodes::LOGICAL_ERROR);
+
+    if (sensitive_data_masker_->rulesCount() > 0)
+    {
+        sensitive_data_masker = std::move(sensitive_data_masker_);
+    }
+}
+
+SensitiveDataMasker * SensitiveDataMasker::getInstance()
+{
+    return sensitive_data_masker.get();
+}
 
 SensitiveDataMasker::SensitiveDataMasker(const Poco::Util::AbstractConfiguration & config, const std::string & config_prefix)
 {
@@ -90,7 +115,7 @@ SensitiveDataMasker::SensitiveDataMasker(const Poco::Util::AbstractConfiguration
             if (!used_names.insert(rule_name).second)
             {
                 throw Exception(
-                    "Query_masking_rules configuration contains more than one rule named '" + rule_name + "'.",
+                    "query_masking_rules configuration contains more than one rule named '" + rule_name + "'.",
                     ErrorCodes::INVALID_CONFIG_PARAMETER);
             }
 
@@ -99,7 +124,7 @@ SensitiveDataMasker::SensitiveDataMasker(const Poco::Util::AbstractConfiguration
             if (regexp.empty())
             {
                 throw Exception(
-                    "Query_masking_rules configuration, rule '" + rule_name + "' has no <regexp> node or <regexp> is empty.",
+                    "query_masking_rules configuration, rule '" + rule_name + "' has no <regexp> node or <regexp> is empty.",
                     ErrorCodes::NO_ELEMENTS_IN_CONFIG);
             }
 

dbms/src/Common/SensitiveDataMasker.h

@@ -4,7 +4,6 @@
 #include <vector>
 #include <cstdint>
 
-
 namespace Poco
 {
 namespace Util
@@ -13,6 +12,32 @@ namespace Util
 }
 }
 
+/// SensitiveDataMasker allows to remove sensitive data from queries using set of regexp-based rules
+
+/// It's used as a singelton via getInstance method
+
+/// Initially it's empty (nullptr) and after manual initialization
+/// (one-time, done by setInstance call) it takes the proper value which
+/// is stored in unique_ptr.
+
+/// It looks like the singelton is the best option here, as
+/// two users of that object (OwnSplitChannel & Interpreters/executeQuery)
+/// can't own/share that Masker properly without syncronization & locks,
+/// and we can't afford setting global locks for each logged line.
+
+/// I've considered singleton alternatives, but it's unclear who should own the object,
+/// and it introduce unnecessary complexity in implementation (passing references back and forward):
+///
+///  context can't own, as Context is destroyed before logger,
+///    and logger lives longer and logging can still happen after Context destruction.
+///    resetting masker in the logger at the moment of
+///    context destruction can't be done w/o synchronization / locks in a safe manner.
+///
+///  logger is Poco derived and i didn't want to brake it's interface,
+///    also logger can be dynamically reconfigured without server restart,
+///    and it actually recreates OwnSplitChannel when reconfiguration happen,
+///    so that makes it's quite tricky. So it a bad candidate for owning masker too.
+
 namespace DB
 {
 class SensitiveDataMasker
@@ -20,6 +45,7 @@ class SensitiveDataMasker
 private:
     class MaskingRule;
     std::vector<std::unique_ptr<MaskingRule>> all_masking_rules;
+    static std::unique_ptr<SensitiveDataMasker> sensitive_data_masker;
 
 public:
     SensitiveDataMasker(const Poco::Util::AbstractConfiguration & config, const std::string & config_prefix);
@@ -28,6 +54,11 @@ public:
     /// Returns the number of matched rules.
     size_t wipeSensitiveData(std::string & data) const;
 
+    /// setInstance is not thread-safe and should be called once in single-thread mode.
+    /// https://github.com/yandex/ClickHouse/pull/6810#discussion_r321183367
+    static void setInstance(std::unique_ptr<SensitiveDataMasker> sensitive_data_masker_);
+    static SensitiveDataMasker * getInstance();
+
     /// Used in tests.
     void addMaskingRule(const std::string & name, const std::string & regexp_string, const std::string & replacement_string);
 

dbms/src/Common/SharedBlockRowRef.h

@@ -0,0 +1,89 @@
+#pragma once
+
+#include <algorithm>
+#include <Core/Block.h>
+#include <Columns/IColumn.h>
+#include <boost/smart_ptr/intrusive_ptr.hpp>
+
+
+namespace DB
+{
+
+/// Allows you refer to the row in the block and hold the block ownership,
+///  and thus avoid creating a temporary row object.
+/// Do not use std::shared_ptr, since there is no need for a place for `weak_count` and `deleter`;
+///  does not use Poco::SharedPtr, since you need to allocate a block and `refcount` in one piece;
+///  does not use Poco::AutoPtr, since it does not have a `move` constructor and there are extra checks for nullptr;
+/// The reference counter is not atomic, since it is used from one thread.
+namespace detail
+{
+    struct SharedBlock : Block
+    {
+        int refcount = 0;
+
+        ColumnRawPtrs all_columns;
+        ColumnRawPtrs sort_columns;
+
+        SharedBlock(Block && block) : Block(std::move(block)) {}
+    };
+}
+
+inline void intrusive_ptr_add_ref(detail::SharedBlock * ptr)
+{
+    ++ptr->refcount;
+}
+
+inline void intrusive_ptr_release(detail::SharedBlock * ptr)
+{
+    if (0 == --ptr->refcount)
+        delete ptr;
+}
+
+using SharedBlockPtr = boost::intrusive_ptr<detail::SharedBlock>;
+
+struct SharedBlockRowRef
+{
+    ColumnRawPtrs * columns = nullptr;
+    size_t row_num;
+    SharedBlockPtr shared_block;
+
+    void swap(SharedBlockRowRef & other)
+    {
+        std::swap(columns, other.columns);
+        std::swap(row_num, other.row_num);
+        std::swap(shared_block, other.shared_block);
+    }
+
+    /// The number and types of columns must match.
+    bool operator==(const SharedBlockRowRef & other) const
+    {
+        size_t size = columns->size();
+        for (size_t i = 0; i < size; ++i)
+            if (0 != (*columns)[i]->compareAt(row_num, other.row_num, *(*other.columns)[i], 1))
+                return false;
+        return true;
+    }
+
+    bool operator!=(const SharedBlockRowRef & other) const
+    {
+        return !(*this == other);
+    }
+
+    void reset()
+    {
+        SharedBlockRowRef empty;
+        swap(empty);
+    }
+
+    bool empty() const { return columns == nullptr; }
+    size_t size() const { return empty() ? 0 : columns->size(); }
+
+    void set(SharedBlockPtr & shared_block_, ColumnRawPtrs * columns_, size_t row_num_)
+    {
+        shared_block = shared_block_;
+        columns = columns_;
+        row_num = row_num_;
+    }
+};
+
+}

dbms/src/Common/TraceCollector.cpp

@@ -49,15 +49,26 @@ TraceCollector::TraceCollector(std::shared_ptr<TraceLog> & trace_log_)
 #if !defined(__FreeBSD__)
     /** Increase pipe size to avoid slowdown during fine-grained trace collection.
       */
-    constexpr int max_pipe_capacity_to_set = 1048576;
     int pipe_size = fcntl(trace_pipe.fds_rw[1], F_GETPIPE_SZ);
     if (-1 == pipe_size)
-        throwFromErrno("Cannot get pipe capacity", ErrorCodes::CANNOT_FCNTL);
+    {
-    for (errno = 0; errno != EPERM && pipe_size < max_pipe_capacity_to_set; pipe_size *= 2)
+        if (errno == EINVAL)
-        if (-1 == fcntl(trace_pipe.fds_rw[1], F_SETPIPE_SZ, pipe_size * 2) && errno != EPERM)
+        {
-            throwFromErrno("Cannot increase pipe capacity to " + toString(pipe_size * 2), ErrorCodes::CANNOT_FCNTL);
+            LOG_INFO(log, "Cannot get pipe capacity, " << errnoToString(ErrorCodes::CANNOT_FCNTL) << ". Very old Linux kernels have no support for this fcntl.");
+            /// It will work nevertheless.
+        }
+        else
+            throwFromErrno("Cannot get pipe capacity", ErrorCodes::CANNOT_FCNTL);
+    }
+    else
+    {
+        constexpr int max_pipe_capacity_to_set = 1048576;
+        for (errno = 0; errno != EPERM && pipe_size < max_pipe_capacity_to_set; pipe_size *= 2)
+            if (-1 == fcntl(trace_pipe.fds_rw[1], F_SETPIPE_SZ, pipe_size * 2) && errno != EPERM)
+                throwFromErrno("Cannot increase pipe capacity to " + toString(pipe_size * 2), ErrorCodes::CANNOT_FCNTL);
 
-    LOG_TRACE(log, "Pipe capacity is " << formatReadableSizeWithBinarySuffix(std::min(pipe_size, max_pipe_capacity_to_set)));
+        LOG_TRACE(log, "Pipe capacity is " << formatReadableSizeWithBinarySuffix(std::min(pipe_size, max_pipe_capacity_to_set)));
+    }
 #endif
 
     thread = ThreadFromGlobalPool(&TraceCollector::run, this);

dbms/src/Common/parseGlobs.cpp

@@ -0,0 +1,76 @@
+#include <Common/parseGlobs.h>
+#include <re2/re2.h>
+#include <re2/stringpiece.h>
+#include <algorithm>
+#include <sstream>
+
+namespace DB
+{
+/* Transforms string from grep-wildcard-syntax ("{N..M}", "{a,b,c}" as in remote table function and "*", "?") to perl-regexp for using re2 library fo matching
+ * with such steps:
+ * 1) search intervals like {0..9} and enums like {abc,xyz,qwe} in {}, replace them by regexp with pipe (expr1|expr2|expr3),
+ * 2) search and replace "*" and "?".
+ * Before each search need to escape symbols that we would not search.
+ *
+ * There are few examples in unit tests.
+ */
+std::string makeRegexpPatternFromGlobs(const std::string & initial_str_with_globs)
+{
+    std::ostringstream oss_for_escaping;
+    /// Escaping only characters that not used in glob syntax
+    for (const auto & letter : initial_str_with_globs)
+    {
+        if ((letter == '[') || (letter == ']') || (letter == '|') || (letter == '+') || (letter == '-') || (letter == '(') || (letter == ')'))
+            oss_for_escaping << '\\';
+        oss_for_escaping << letter;
+    }
+    std::string escaped_with_globs = oss_for_escaping.str();
+
+    static const re2::RE2 enum_or_range(R"({([\d]+\.\.[\d]+|[^{}*,]+,[^{}*]*[^{}*,])})");    /// regexp for {expr1,expr2,expr3} or {M..N}, where M and N - non-negative integers, expr's should be without {}*,
+    re2::StringPiece input(escaped_with_globs);
+    re2::StringPiece matched;
+    std::ostringstream oss_for_replacing;
+    size_t current_index = 0;
+    while (RE2::FindAndConsume(&input, enum_or_range, &matched))
+    {
+        std::string buffer = matched.ToString();
+        oss_for_replacing << escaped_with_globs.substr(current_index, matched.data() - escaped_with_globs.data() - current_index - 1) << '(';
+
+        if (buffer.find(',') == std::string::npos)
+        {
+            size_t range_begin, range_end;
+            char point;
+            std::istringstream iss_range(buffer);
+            iss_range >> range_begin >> point >> point >> range_end;
+            oss_for_replacing << range_begin;
+            for (size_t i = range_begin + 1; i <= range_end; ++i)
+            {
+                oss_for_replacing << '|' << i;
+            }
+        }
+        else
+        {
+            std::replace(buffer.begin(), buffer.end(), ',', '|');
+            oss_for_replacing << buffer;
+        }
+        oss_for_replacing << ")";
+        current_index = input.data() - escaped_with_globs.data();
+    }
+    oss_for_replacing << escaped_with_globs.substr(current_index);
+    std::string almost_res = oss_for_replacing.str();
+    std::ostringstream oss_final_processing;
+    for (const auto & letter : almost_res)
+    {
+        if ((letter == '?') || (letter == '*'))
+        {
+            oss_final_processing << "[^/]";   /// '?' is any symbol except '/'
+            if (letter == '?')
+                continue;
+        }
+        if ((letter == '.') || (letter == '{') || (letter == '}'))
+            oss_final_processing << '\\';
+        oss_final_processing << letter;
+    }
+    return oss_final_processing.str();
+}
+}

dbms/src/Common/parseGlobs.h

@@ -0,0 +1,10 @@
+#pragma once
+#include <string>
+#include <vector>
+
+namespace DB
+{
+/* Parse globs in string and make a regexp for it.
+ */
+std::string makeRegexpPatternFromGlobs(const std::string & path);
+}

dbms/src/Common/tests/gtest_makeRegexpPatternFromGlobs.cpp

@@ -0,0 +1,20 @@
+#include <Common/parseGlobs.h>
+#include <re2/re2.h>
+#include <gtest/gtest.h>
+
+
+using namespace DB;
+
+
+TEST(Common, makeRegexpPatternFromGlobs)
+{
+    EXPECT_EQ(makeRegexpPatternFromGlobs("f{01..09}"), "f(1|2|3|4|5|6|7|8|9)");
+    EXPECT_EQ(makeRegexpPatternFromGlobs("f{01..9}"), "f(1|2|3|4|5|6|7|8|9)");
+    EXPECT_EQ(makeRegexpPatternFromGlobs("f{0001..0000009}"), "f(1|2|3|4|5|6|7|8|9)");
+    EXPECT_EQ(makeRegexpPatternFromGlobs("f{1..2}{1..2}"), "f(1|2)(1|2)");
+    EXPECT_EQ(makeRegexpPatternFromGlobs("f{1..1}{1..1}"), "f(1)(1)");
+    EXPECT_EQ(makeRegexpPatternFromGlobs("f{0..0}{0..0}"), "f(0)(0)");
+    EXPECT_EQ(makeRegexpPatternFromGlobs("file{1..5}"),"file(1|2|3|4|5)");
+    EXPECT_EQ(makeRegexpPatternFromGlobs("file{1,2,3}"),"file(1|2|3)");
+    EXPECT_EQ(makeRegexpPatternFromGlobs("{1,2,3}blabla{a.x,b.x,c.x}smth[]_else{aa,bb}?*"), "(1|2|3)blabla(a\\.x|b\\.x|c\\.x)smth\\[\\]_else(aa|bb)[^/][^/]*");
+}

dbms/src/Common/tests/gtest_sensitive_data_masker.cpp

@@ -59,7 +59,6 @@ TEST(Common, SensitiveDataMasker)
         "SELECT id FROM mysql('localhost:3308', 'database', 'table', 'root', '******') WHERE "
         "ssn='000-00-0000' or email='hidden@hidden.test'");
 
-
 #ifndef NDEBUG
     // simple benchmark
     auto start = std::chrono::high_resolution_clock::now();
@@ -86,7 +85,7 @@ TEST(Common, SensitiveDataMasker)
         maskerbad.addMaskingRule("bad regexp", "**", "");
         ADD_FAILURE() << "addMaskingRule() should throw an error" << std::endl;
     }
-    catch (DB::Exception & e)
+    catch (const DB::Exception & e)
     {
         EXPECT_EQ(
             std::string(e.what()),
@@ -170,7 +169,7 @@ TEST(Common, SensitiveDataMasker)
 
         ADD_FAILURE() << "XML should throw an error on bad XML" << std::endl;
     }
-    catch (DB::Exception & e)
+    catch (const DB::Exception & e)
     {
         EXPECT_EQ(
             std::string(e.what()),
@@ -178,7 +177,6 @@ TEST(Common, SensitiveDataMasker)
         EXPECT_EQ(e.code(), DB::ErrorCodes::INVALID_CONFIG_PARAMETER);
     }
 
-
     try
     {
         std::istringstream xml_isteam_bad(R"END(<?xml version="1.0"?>
@@ -193,7 +191,7 @@ TEST(Common, SensitiveDataMasker)
 
         ADD_FAILURE() << "XML should throw an error on bad XML" << std::endl;
     }
-    catch (DB::Exception & e)
+    catch (const DB::Exception & e)
     {
         EXPECT_EQ(
             std::string(e.what()),
@@ -201,7 +199,6 @@ TEST(Common, SensitiveDataMasker)
         EXPECT_EQ(e.code(), DB::ErrorCodes::NO_ELEMENTS_IN_CONFIG);
     }
 
-
     try
     {
         std::istringstream xml_isteam_bad(R"END(<?xml version="1.0"?>
@@ -216,7 +213,7 @@ TEST(Common, SensitiveDataMasker)
 
         ADD_FAILURE() << "XML should throw an error on bad XML" << std::endl;
     }
-    catch (DB::Exception & e)
+    catch (const DB::Exception & e)
     {
         EXPECT_EQ(
             std::string(e.message()),

dbms/src/Core/Defines.h

@@ -39,7 +39,9 @@
 #define DBMS_CONNECTION_POOL_WITH_FAILOVER_DEFAULT_MAX_TRIES 3
 /// each period reduces the error counter by 2 times
 /// too short a period can cause errors to disappear immediately after creation.
-#define DBMS_CONNECTION_POOL_WITH_FAILOVER_DEFAULT_DECREASE_ERROR_PERIOD (2 * DBMS_DEFAULT_SEND_TIMEOUT_SEC)
+#define DBMS_CONNECTION_POOL_WITH_FAILOVER_DEFAULT_DECREASE_ERROR_PERIOD 60
+/// replica error max cap, this is to prevent replica from accumulating too many errors and taking to long to recover.
+#define DBMS_CONNECTION_POOL_WITH_FAILOVER_MAX_ERROR_COUNT 1000
 
 #define DBMS_MIN_REVISION_WITH_CLIENT_INFO 54032
 #define DBMS_MIN_REVISION_WITH_SERVER_TIMEZONE 54058

dbms/src/Core/Names.h

@@ -16,4 +16,7 @@ using NameOrderedSet = std::set<std::string>;
 using NameToNameMap = std::unordered_map<std::string, std::string>;
 using NameToNameSetMap = std::unordered_map<std::string, NameSet>;
 
+using NameWithAlias = std::pair<std::string, std::string>;
+using NamesWithAliases = std::vector<NameWithAlias>;
+
 }

dbms/src/Core/Settings.h

@@ -348,6 +348,9 @@ struct Settings : public SettingsCollection<Settings>
     M(SettingBool, check_query_single_value_result, true, "Return check query result as single 1/0 value") \
     M(SettingBool, allow_drop_detached, false, "Allow ALTER TABLE ... DROP DETACHED PART[ITION] ... queries") \
     \
+    M(SettingSeconds, distributed_replica_error_half_life, DBMS_CONNECTION_POOL_WITH_FAILOVER_DEFAULT_DECREASE_ERROR_PERIOD, "Time period reduces replica error counter by 2 times.") \
+    M(SettingUInt64, distributed_replica_error_cap, DBMS_CONNECTION_POOL_WITH_FAILOVER_MAX_ERROR_COUNT, "Max number of errors per replica, prevents piling up increadible amount of errors if replica was offline for some time and allows it to be reconsidered in a shorter amount of time.") \
+    \
     M(SettingBool, allow_experimental_live_view, false, "Enable LIVE VIEW. Not mature enough.") \
     M(SettingSeconds, live_view_heartbeat_interval, DEFAULT_LIVE_VIEW_HEARTBEAT_INTERVAL_SEC, "The heartbeat interval in seconds to indicate live query is alive.") \
     M(SettingSeconds, temporary_live_view_timeout, DEFAULT_TEMPORARY_LIVE_VIEW_TIMEOUT_SEC, "Timeout after which temporary live view is deleted.") \

dbms/src/Core/SortDescription.h

@@ -4,13 +4,22 @@
 #include <memory>
 #include <cstddef>
 #include <string>
-
+#include <Core/Field.h>
 
 class Collator;
 
 namespace DB
 {
 
+struct FillColumnDescription
+{
+    /// All missed values in range [FROM, TO) will be filled
+    /// Range [FROM, TO) respects sorting direction
+    Field fill_from;        /// Fill value >= FILL_FROM
+    Field fill_to;          /// Fill value + STEP < FILL_TO
+    Field fill_step;        /// Default = 1 or -1 according to direction
+};
+
 /// Description of the sorting rule by one column.
 struct SortColumnDescription
 {
@@ -20,12 +29,23 @@ struct SortColumnDescription
     int nulls_direction;     /// 1 - NULLs and NaNs are greater, -1 - less.
                              /// To achieve NULLS LAST, set it equal to direction, to achieve NULLS FIRST, set it opposite.
     std::shared_ptr<Collator> collator; /// Collator for locale-specific comparison of strings
+    bool with_fill;
+    FillColumnDescription fill_description;
 
-    SortColumnDescription(size_t column_number_, int direction_, int nulls_direction_, const std::shared_ptr<Collator> & collator_ = nullptr)
-        : column_number(column_number_), direction(direction_), nulls_direction(nulls_direction_), collator(collator_) {}
 
-    SortColumnDescription(const std::string & column_name_, int direction_, int nulls_direction_, const std::shared_ptr<Collator> & collator_ = nullptr)
+    SortColumnDescription(
-        : column_name(column_name_), column_number(0), direction(direction_), nulls_direction(nulls_direction_), collator(collator_) {}
+            size_t column_number_, int direction_, int nulls_direction_,
+            const std::shared_ptr<Collator> & collator_ = nullptr, bool with_fill_ = false,
+            const FillColumnDescription & fill_description_ = {})
+            : column_number(column_number_), direction(direction_), nulls_direction(nulls_direction_), collator(collator_)
+            , with_fill(with_fill_), fill_description(fill_description_) {}
+
+    SortColumnDescription(
+            const std::string & column_name_, int direction_, int nulls_direction_,
+            const std::shared_ptr<Collator> & collator_ = nullptr, bool with_fill_ = false,
+            const FillColumnDescription & fill_description_ = {})
+            : column_name(column_name_), column_number(0), direction(direction_), nulls_direction(nulls_direction_)
+            , collator(collator_), with_fill(with_fill_), fill_description(fill_description_) {}
 
     bool operator == (const SortColumnDescription & other) const
     {

dbms/src/DataStreams/AggregatingSortedBlockInputStream.h

@@ -50,8 +50,8 @@ private:
     std::vector<ColumnAggregateFunction *> columns_to_aggregate;
     std::vector<SimpleAggregateDescription> columns_to_simple_aggregate;
 
-    RowRef current_key;        /// The current primary key.
+    SharedBlockRowRef current_key;        /// The current primary key.
-    RowRef next_key;           /// The primary key of the next row.
+    SharedBlockRowRef next_key;           /// The primary key of the next row.
 
     /** We support two different cursors - with Collation and without.
      *  Templates are used instead of polymorphic SortCursor and calls to virtual functions.

dbms/src/DataStreams/CollapsingSortedBlockInputStream.h

@@ -47,12 +47,12 @@ private:
     /// Read is finished.
     bool finished = false;
 
-    RowRef current_key;         /// The current primary key.
+    SharedBlockRowRef current_key;         /// The current primary key.
-    RowRef next_key;            /// The primary key of the next row.
+    SharedBlockRowRef next_key;            /// The primary key of the next row.
 
-    RowRef first_negative;        /// The first negative row for the current primary key.
+    SharedBlockRowRef first_negative;        /// The first negative row for the current primary key.
-    RowRef last_positive;         /// The last positive row for the current primary key.
+    SharedBlockRowRef last_positive;         /// The last positive row for the current primary key.
-    RowRef last_negative;         /// Last negative row. It is only stored if there is not one row is written to output.
+    SharedBlockRowRef last_negative;         /// Last negative row. It is only stored if there is not one row is written to output.
 
     size_t count_positive = 0;    /// The number of positive rows for the current primary key.
     size_t count_negative = 0;    /// The number of negative rows for the current primary key.

dbms/src/DataStreams/FillingBlockInputStream.cpp

@@ -0,0 +1,186 @@
+#include <DataStreams/FillingBlockInputStream.h>
+#include <Interpreters/convertFieldToType.h>
+#include <DataTypes/DataTypesNumber.h>
+
+namespace DB
+{
+
+namespace ErrorCodes
+{
+    extern const int INVALID_WITH_FILL_EXPRESSION;
+}
+
+FillingBlockInputStream::FillingBlockInputStream(
+        const BlockInputStreamPtr & input, const SortDescription & sort_description_)
+        : sort_description(sort_description_), filling_row(sort_description_), next_row(sort_description_)
+{
+    children.push_back(input);
+    header = children.at(0)->getHeader();
+
+    std::vector<bool> is_fill_column(header.columns());
+    for (const auto & elem : sort_description)
+        is_fill_column[header.getPositionByName(elem.column_name)] = true;
+
+    auto try_convert_fields = [](FillColumnDescription & descr, const DataTypePtr & type)
+    {
+        auto max_type = Field::Types::Null;
+        WhichDataType which(type);
+        DataTypePtr to_type;
+        if (isInteger(type) || which.isDateOrDateTime())
+        {
+            max_type = Field::Types::Int64;
+            to_type = std::make_shared<DataTypeInt64>();
+        }
+        else if (which.isFloat())
+        {
+            max_type = Field::Types::Float64;
+            to_type = std::make_shared<DataTypeFloat64>();
+        }
+
+        if (descr.fill_from.getType() > max_type || descr.fill_to.getType() > max_type
+            || descr.fill_step.getType() > max_type)
+            return false;
+        descr.fill_from = convertFieldToType(descr.fill_from, *to_type);
+        descr.fill_to = convertFieldToType(descr.fill_to, *to_type);
+        descr.fill_step = convertFieldToType(descr.fill_step, *to_type);
+
+        return true;
+    };
+
+    for (size_t i = 0; i < header.columns(); ++i)
+    {
+        if (is_fill_column[i])
+        {
+            size_t pos = fill_column_positions.size();
+            auto & descr = filling_row.getFillDescription(pos);
+            auto type = header.getByPosition(i).type;
+            if (!try_convert_fields(descr, type))
+                throw Exception("Incompatible types of WITH FILL expression values with column type "
+                    + type->getName(), ErrorCodes::INVALID_WITH_FILL_EXPRESSION);
+
+            if (type->isValueRepresentedByUnsignedInteger() &&
+                ((!descr.fill_from.isNull() && less(descr.fill_from, Field{0}, 1)) ||
+                    (!descr.fill_to.isNull() && less(descr.fill_to, Field{0}, 1))))
+            {
+                throw Exception("WITH FILL bound values cannot be negative for unsigned type "
+                    + type->getName(), ErrorCodes::INVALID_WITH_FILL_EXPRESSION);
+            }
+
+            fill_column_positions.push_back(i);
+        }
+        else
+            other_column_positions.push_back(i);
+    }
+}
+
+
+Block FillingBlockInputStream::readImpl()
+{
+    Columns old_fill_columns;
+    Columns old_other_columns;
+    MutableColumns res_fill_columns;
+    MutableColumns res_other_columns;
+
+    auto init_columns_by_positions = [](const Block & block, Columns & columns,
+        MutableColumns & mutable_columns, const Positions & positions)
+    {
+        for (size_t pos : positions)
+        {
+            auto column = block.getByPosition(pos).column;
+            columns.push_back(column);
+            mutable_columns.push_back(column->cloneEmpty()->assumeMutable());
+        }
+    };
+
+    auto block = children.back()->read();
+    if (!block)
+    {
+        init_columns_by_positions(header, old_fill_columns, res_fill_columns, fill_column_positions);
+        init_columns_by_positions(header, old_other_columns, res_other_columns, other_column_positions);
+
+        bool should_insert_first = next_row < filling_row;
+
+        bool generated = false;
+        for (size_t i = 0; i < filling_row.size(); ++i)
+            next_row[i] = filling_row.getFillDescription(i).fill_to;
+
+        if (should_insert_first && filling_row < next_row)
+            insertFromFillingRow(res_fill_columns, res_other_columns, filling_row);
+
+        while (filling_row.next(next_row))
+        {
+            generated = true;
+            insertFromFillingRow(res_fill_columns, res_other_columns, filling_row);
+        }
+
+        if (generated)
+            return createResultBlock(res_fill_columns, res_other_columns);
+
+        return block;
+    }
+
+    size_t rows = block.rows();
+    init_columns_by_positions(block, old_fill_columns, res_fill_columns, fill_column_positions);
+    init_columns_by_positions(block, old_other_columns, res_other_columns, other_column_positions);
+
+    if (first)
+    {
+        for (size_t i = 0; i < filling_row.size(); ++i)
+        {
+            auto current_value = (*old_fill_columns[i])[0];
+            const auto & fill_from = filling_row.getFillDescription(i).fill_from;
+            if (!fill_from.isNull() && !equals(current_value, fill_from))
+            {
+                filling_row.initFromDefaults(i);
+                if (less(fill_from, current_value, filling_row.getDirection(i)))
+                    insertFromFillingRow(res_fill_columns, res_other_columns, filling_row);
+                break;
+            }
+            filling_row[i] = current_value;
+        }
+        first = false;
+    }
+
+    for (size_t row_ind = 0; row_ind < rows; ++row_ind)
+    {
+        bool should_insert_first = next_row < filling_row;
+
+        for (size_t i = 0; i < filling_row.size(); ++i)
+        {
+            auto current_value = (*old_fill_columns[i])[row_ind];
+            const auto & fill_to = filling_row.getFillDescription(i).fill_to;
+
+            if (fill_to.isNull() || less(current_value, fill_to, filling_row.getDirection(i)))
+                next_row[i] = current_value;
+            else
+                next_row[i] = fill_to;
+        }
+
+        /// A case, when at previous step row was initialized from defaults 'fill_from' values
+        ///  and probably we need to insert it to block.
+        if (should_insert_first && filling_row < next_row)
+            insertFromFillingRow(res_fill_columns, res_other_columns, filling_row);
+
+        /// Insert generated filling row to block, while it is less than current row in block.
+        while (filling_row.next(next_row))
+            insertFromFillingRow(res_fill_columns, res_other_columns, filling_row);
+
+        copyRowFromColumns(res_fill_columns, old_fill_columns, row_ind);
+        copyRowFromColumns(res_other_columns, old_other_columns, row_ind);
+    }
+
+    return createResultBlock(res_fill_columns, res_other_columns);
+}
+
+Block FillingBlockInputStream::createResultBlock(MutableColumns & fill_columns, MutableColumns & other_columns) const
+{
+    MutableColumns result_columns(header.columns());
+    for (size_t i = 0; i < fill_columns.size(); ++i)
+        result_columns[fill_column_positions[i]] = std::move(fill_columns[i]);
+    for (size_t i = 0; i < other_columns.size(); ++i)
+        result_columns[other_column_positions[i]] = std::move(other_columns[i]);
+
+    return header.cloneWithColumns(std::move(result_columns));
+}
+
+}

dbms/src/DataStreams/FillingBlockInputStream.h

@@ -0,0 +1,39 @@
+#pragma once
+
+#include <DataStreams/IBlockInputStream.h>
+#include <Interpreters/FillingRow.h>
+
+namespace DB
+{
+
+/** Implements modifier WITH FILL of ORDER BY clause.
+ *  It fills gaps in data stream by rows with missing values in columns with set WITH FILL and deafults in other columns.
+ *  Optionally FROM, TO and STEP values can be specified.
+ */
+class FillingBlockInputStream : public IBlockInputStream
+{
+public:
+    FillingBlockInputStream(const BlockInputStreamPtr & input, const SortDescription & fill_description_);
+
+    String getName() const override { return "Filling"; }
+
+    Block getHeader() const override { return header; }
+
+protected:
+    Block readImpl() override;
+
+private:
+    Block createResultBlock(MutableColumns & fill_columns, MutableColumns & other_columns) const;
+
+    const SortDescription sort_description; /// Contains only rows with WITH FILL.
+    FillingRow filling_row; /// Current row, which is used to fill gaps.
+    FillingRow next_row; /// Row to which we need to generate filling rows.
+    Block header;
+
+    using Positions = std::vector<size_t>;
+    Positions fill_column_positions;
+    Positions other_column_positions;
+    bool first = true;
+};
+
+}

dbms/src/DataStreams/GraphiteRollupSortedBlockInputStream.cpp

@@ -321,7 +321,7 @@ void GraphiteRollupSortedBlockInputStream::finishCurrentGroup(MutableColumns & m
 }
 
 
-void GraphiteRollupSortedBlockInputStream::accumulateRow(RowRef & row)
+void GraphiteRollupSortedBlockInputStream::accumulateRow(SharedBlockRowRef & row)
 {
     const Graphite::AggregationPattern * aggregation_pattern = std::get<1>(current_rule);
     if (aggregate_state_created)

dbms/src/DataStreams/GraphiteRollupSortedBlockInputStream.h

@@ -204,7 +204,7 @@ private:
     StringRef current_group_path;
 
     /// Last row with maximum version for current primary key (time bucket).
-    RowRef current_subgroup_newest_row;
+    SharedBlockRowRef current_subgroup_newest_row;
 
     /// Time of last read row
     time_t current_time = 0;
@@ -236,7 +236,7 @@ private:
     void finishCurrentGroup(MutableColumns & merged_columns);
 
     /// Update the state of the aggregate function with the new `value`.
-    void accumulateRow(RowRef & row);
+    void accumulateRow(SharedBlockRowRef & row);
 };
 
 }

dbms/src/DataStreams/InputStreamFromASTInsertQuery.cpp

@@ -16,11 +16,12 @@ namespace DB
 namespace ErrorCodes
 {
     extern const int LOGICAL_ERROR;
+    extern const int INVALID_USAGE_OF_INPUT;
 }
 
 
 InputStreamFromASTInsertQuery::InputStreamFromASTInsertQuery(
-    const ASTPtr & ast, ReadBuffer * input_buffer_tail_part, const Block & header, const Context & context)
+    const ASTPtr & ast, ReadBuffer * input_buffer_tail_part, const Block & header, const Context & context, const ASTPtr & input_function)
 {
     const auto * ast_insert_query = ast->as<ASTInsertQuery>();
 
@@ -29,7 +30,11 @@ InputStreamFromASTInsertQuery::InputStreamFromASTInsertQuery(
 
     String format = ast_insert_query->format;
     if (format.empty())
+    {
+        if (input_function)
+            throw Exception("FORMAT must be specified for function input()", ErrorCodes::INVALID_USAGE_OF_INPUT);
         format = "Values";
+    }
 
     /// Data could be in parsed (ast_insert_query.data) and in not parsed yet (input_buffer_tail_part) part of query.
 
@@ -51,7 +56,7 @@ InputStreamFromASTInsertQuery::InputStreamFromASTInsertQuery(
 
     res_stream = context.getInputFormat(format, *input_buffer_contacenated, header, context.getSettings().max_insert_block_size);
 
-    if (context.getSettingsRef().input_format_defaults_for_omitted_fields && !ast_insert_query->table.empty())
+    if (context.getSettingsRef().input_format_defaults_for_omitted_fields && !ast_insert_query->table.empty() && !input_function)
     {
         StoragePtr storage = context.getTable(ast_insert_query->database, ast_insert_query->table);
         auto column_defaults = storage->getColumns().getDefaults();

dbms/src/DataStreams/InputStreamFromASTInsertQuery.h

@@ -19,7 +19,11 @@ class Context;
 class InputStreamFromASTInsertQuery : public IBlockInputStream
 {
 public:
-    InputStreamFromASTInsertQuery(const ASTPtr & ast, ReadBuffer * input_buffer_tail_part, const Block & header, const Context & context);
+    InputStreamFromASTInsertQuery(const ASTPtr & ast,
+        ReadBuffer * input_buffer_tail_part,
+        const Block & header,
+        const Context & context,
+        const ASTPtr & input_function);
 
     Block readImpl() override { return res_stream->read(); }
     void readPrefixImpl() override { return res_stream->readPrefix(); }

dbms/src/DataStreams/LimitBlockInputStream.cpp

@@ -6,8 +6,30 @@
 namespace DB
 {
 
-LimitBlockInputStream::LimitBlockInputStream(const BlockInputStreamPtr & input, UInt64 limit_, UInt64 offset_, bool always_read_till_end_, bool use_limit_as_total_rows_approx)
+/// gets pointers to all columns of block, which were used for ORDER BY
-    : limit(limit_), offset(offset_), always_read_till_end(always_read_till_end_)
+static ColumnRawPtrs extractSortColumns(const Block & block, const SortDescription & description)
+{
+    size_t size = description.size();
+    ColumnRawPtrs res;
+    res.reserve(size);
+
+    for (size_t i = 0; i < size; ++i)
+    {
+        const IColumn * column = !description[i].column_name.empty()
+                                    ? block.getByName(description[i].column_name).column.get()
+                                    : block.safeGetByPosition(description[i].column_number).column.get();
+        res.emplace_back(column);
+    }
+
+    return res;
+}
+
+
+LimitBlockInputStream::LimitBlockInputStream(
+    const BlockInputStreamPtr & input, UInt64 limit_, UInt64 offset_, bool always_read_till_end_,
+    bool use_limit_as_total_rows_approx, bool with_ties_, const SortDescription & description_)
+    : limit(limit_), offset(offset_), always_read_till_end(always_read_till_end_), with_ties(with_ties_)
+    , description(description_)
 {
     if (use_limit_as_total_rows_approx)
     {
@@ -17,13 +39,45 @@ LimitBlockInputStream::LimitBlockInputStream(const BlockInputStreamPtr & input,
     children.push_back(input);
 }
 
-
 Block LimitBlockInputStream::readImpl()
 {
     Block res;
     UInt64 rows = 0;
 
-    /// pos - how many rows were read, including the last read block
+    /// pos >= offset + limit and all rows in the end of previous block were equal
+    ///  to row at 'limit' position. So we check current block.
+    if (!ties_row_ref.empty() && pos >= offset + limit)
+    {
+        res = children.back()->read();
+        rows = res.rows();
+
+        if (!res)
+            return res;
+
+        SharedBlockPtr ptr = new detail::SharedBlock(std::move(res));
+        ptr->sort_columns = extractSortColumns(*ptr, description);
+
+        UInt64 len;
+        for (len = 0; len < rows; ++len)
+        {
+            SharedBlockRowRef current_row;
+            current_row.set(ptr, &ptr->sort_columns, len);
+
+            if (current_row != ties_row_ref)
+            {
+                ties_row_ref.reset();
+                break;
+            }
+        }
+
+        if (len < rows)
+        {
+            for (size_t i = 0; i < ptr->columns(); ++i)
+                ptr->safeGetByPosition(i).column = ptr->safeGetByPosition(i).column->cut(0, len);
+        }
+
+        return *ptr;
+    }
 
     if (pos >= offset + limit)
     {
@@ -46,9 +100,18 @@ Block LimitBlockInputStream::readImpl()
         pos += rows;
     } while (pos <= offset);
 
-    /// return the whole block
+    SharedBlockPtr ptr = new detail::SharedBlock(std::move(res));
+    if (with_ties)
+        ptr->sort_columns = extractSortColumns(*ptr, description);
+
+    /// give away the whole block
     if (pos >= offset + rows && pos <= offset + limit)
-        return res;
+    {
+        /// Save rowref for last row, because probalbly next block begins with the same row.
+        if (with_ties && pos == offset + limit)
+            ties_row_ref.set(ptr, &ptr->sort_columns, rows - 1);
+        return *ptr;
+    }
 
     /// give away a piece of the block
     UInt64 start = std::max(
@@ -60,13 +123,36 @@ Block LimitBlockInputStream::readImpl()
         static_cast<Int64>(pos) - static_cast<Int64>(offset),
         static_cast<Int64>(limit) + static_cast<Int64>(offset) - static_cast<Int64>(pos) + static_cast<Int64>(rows)));
 
-    for (size_t i = 0; i < res.columns(); ++i)
+
-        res.getByPosition(i).column = res.getByPosition(i).column->cut(start, length);
+    /// check if other rows in current block equals to last one in limit
+    if (with_ties)
+    {
+        ties_row_ref.set(ptr, &ptr->sort_columns, start + length - 1);
+
+        for (size_t i = ties_row_ref.row_num + 1; i < rows; ++i)
+        {
+            SharedBlockRowRef current_row;
+            current_row.set(ptr, &ptr->sort_columns, i);
+            if (current_row == ties_row_ref)
+                ++length;
+            else
+            {
+                ties_row_ref.reset();
+                break;
+            }
+        }
+    }
+
+    if (length == rows)
+        return *ptr;
+
+    for (size_t i = 0; i < ptr->columns(); ++i)
+        ptr->safeGetByPosition(i).column = ptr->safeGetByPosition(i).column->cut(start, length);
 
     // TODO: we should provide feedback to child-block, so it will know how many rows are actually consumed.
     //       It's crucial for streaming engines like Kafka.
 
-    return res;
+    return *ptr;
 }
 
 }

dbms/src/DataStreams/LimitBlockInputStream.h

@@ -1,6 +1,7 @@
 #pragma once
 
 #include <DataStreams/IBlockInputStream.h>
+#include <Common/SharedBlockRowRef.h>
 
 
 namespace DB
@@ -17,8 +18,13 @@ public:
       * If always_read_till_end = true - reads all the data to the end, but ignores them. This is necessary in rare cases:
       *  when otherwise, due to the cancellation of the request, we would not have received the data for GROUP BY WITH TOTALS from the remote server.
       * If use_limit_as_total_rows_approx = true, then addTotalRowsApprox is called to use the limit in progress & stats
+      * with_ties = true, when query has WITH TIES modifier. If so, description should be provided
+      * description lets us know which row we should check for equality
       */
-    LimitBlockInputStream(const BlockInputStreamPtr & input, UInt64 limit_, UInt64 offset_, bool always_read_till_end_ = false, bool use_limit_as_total_rows_approx = false);
+    LimitBlockInputStream(
+            const BlockInputStreamPtr & input, UInt64 limit_, UInt64 offset_,
+            bool always_read_till_end_ = false, bool use_limit_as_total_rows_approx = false,
+            bool with_ties_ = false, const SortDescription & description_ = {});
 
     String getName() const override { return "Limit"; }
 
@@ -32,6 +38,9 @@ private:
     UInt64 offset;
     UInt64 pos = 0;
     bool always_read_till_end;
+    bool with_ties;
+    const SortDescription description;
+    SharedBlockRowRef ties_row_ref;
 };
 
 }

dbms/src/DataStreams/MergingSortedBlockInputStream.h

@@ -5,6 +5,7 @@
 #include <boost/smart_ptr/intrusive_ptr.hpp>
 
 #include <common/logger_useful.h>
+#include <Common/SharedBlockRowRef.h>
 
 #include <Core/Row.h>
 #include <Core/SortDescription.h>
@@ -24,39 +25,6 @@ namespace ErrorCodes
 }
 
 
-/// Allows you refer to the row in the block and hold the block ownership,
-///  and thus avoid creating a temporary row object.
-/// Do not use std::shared_ptr, since there is no need for a place for `weak_count` and `deleter`;
-///  does not use Poco::SharedPtr, since you need to allocate a block and `refcount` in one piece;
-///  does not use Poco::AutoPtr, since it does not have a `move` constructor and there are extra checks for nullptr;
-/// The reference counter is not atomic, since it is used from one thread.
-namespace detail
-{
-struct SharedBlock : Block
-{
-    int refcount = 0;
-
-    ColumnRawPtrs all_columns;
-    ColumnRawPtrs sort_columns;
-
-    SharedBlock(Block && block) : Block(std::move(block)) {}
-};
-}
-
-using SharedBlockPtr = boost::intrusive_ptr<detail::SharedBlock>;
-
-inline void intrusive_ptr_add_ref(detail::SharedBlock * ptr)
-{
-    ++ptr->refcount;
-}
-
-inline void intrusive_ptr_release(detail::SharedBlock * ptr)
-{
-    if (0 == --ptr->refcount)
-        delete ptr;
-}
-
-
 /** Merges several sorted streams into one sorted stream.
   */
 class MergingSortedBlockInputStream : public IBlockInputStream
@@ -78,44 +46,6 @@ public:
     Block getHeader() const override { return header; }
 
 protected:
-    struct RowRef
-    {
-        ColumnRawPtrs * columns = nullptr;
-        size_t row_num = 0;
-        SharedBlockPtr shared_block;
-
-        void swap(RowRef & other)
-        {
-            std::swap(columns, other.columns);
-            std::swap(row_num, other.row_num);
-            std::swap(shared_block, other.shared_block);
-        }
-
-        /// The number and types of columns must match.
-        bool operator==(const RowRef & other) const
-        {
-            size_t size = columns->size();
-            for (size_t i = 0; i < size; ++i)
-                if (0 != (*columns)[i]->compareAt(row_num, other.row_num, *(*other.columns)[i], 1))
-                    return false;
-            return true;
-        }
-
-        bool operator!=(const RowRef & other) const
-        {
-            return !(*this == other);
-        }
-
-        void reset()
-        {
-            RowRef empty;
-            swap(empty);
-        }
-
-        bool empty() const { return columns == nullptr; }
-        size_t size() const { return empty() ? 0 : columns->size(); }
-    };
-
     /// Simple class, which allows to check stop condition during merge process
     /// in simple case it just compare amount of merged rows with max_block_size
     /// in `count_average` case it compares amount of merged rows with linear combination
@@ -148,7 +78,6 @@ protected:
         }
     };
 
-
     Block readImpl() override;
 
     void readSuffixImpl() override;
@@ -230,7 +159,7 @@ protected:
     }
 
     template <typename TSortCursor>
-    void setRowRef(RowRef & row_ref, TSortCursor & cursor)
+    void setRowRef(SharedBlockRowRef & row_ref, TSortCursor & cursor)
     {
         row_ref.row_num = cursor.impl->pos;
         row_ref.shared_block = source_blocks[cursor.impl->order];
@@ -238,7 +167,7 @@ protected:
     }
 
     template <typename TSortCursor>
-    void setPrimaryKeyRef(RowRef & row_ref, TSortCursor & cursor)
+    void setPrimaryKeyRef(SharedBlockRowRef & row_ref, TSortCursor & cursor)
     {
         row_ref.row_num = cursor.impl->pos;
         row_ref.shared_block = source_blocks[cursor.impl->order];

dbms/src/DataStreams/ReplacingSortedBlockInputStream.h

@@ -41,11 +41,11 @@ private:
     bool finished = false;
 
     /// Primary key of current row.
-    RowRef current_key;
+    SharedBlockRowRef current_key;
     /// Primary key of next row.
-    RowRef next_key;
+    SharedBlockRowRef next_key;
     /// Last row with maximum version for current primary key.
-    RowRef selected_row;
+    SharedBlockRowRef selected_row;
     /// The position (into current_row_sources) of the row with the highest version.
     size_t max_pos = 0;
 

dbms/src/DataStreams/SummingSortedBlockInputStream.h

@@ -129,8 +129,8 @@ private:
     std::vector<AggregateDescription> columns_to_aggregate;
     std::vector<MapDescription> maps_to_sum;
 
-    RowRef current_key;        /// The current primary key.
+    SharedBlockRowRef current_key;        /// The current primary key.
-    RowRef next_key;           /// The primary key of the next row.
+    SharedBlockRowRef next_key;           /// The primary key of the next row.
 
     Row current_row;
     bool current_row_is_zero = true;    /// Are all summed columns zero (or empty)? It is updated incrementally.

dbms/src/DataStreams/VersionedCollapsingSortedBlockInputStream.cpp

@@ -47,7 +47,7 @@ void VersionedCollapsingSortedBlockInputStream::insertGap(size_t gap_size)
     }
 }
 
-void VersionedCollapsingSortedBlockInputStream::insertRow(size_t skip_rows, const RowRef & row, MutableColumns & merged_columns)
+void VersionedCollapsingSortedBlockInputStream::insertRow(size_t skip_rows, const SharedBlockRowRef & row, MutableColumns & merged_columns)
 {
     const auto & columns = row.shared_block->all_columns;
     for (size_t i = 0; i < num_columns; ++i)
@@ -111,7 +111,7 @@ void VersionedCollapsingSortedBlockInputStream::merge(MutableColumns & merged_co
         SortCursor current = queue.top();
         size_t current_block_granularity = current->rows;
 
-        RowRef next_key;
+        SharedBlockRowRef next_key;
 
         Int8 sign = assert_cast<const ColumnInt8 &>(*current->all_columns[sign_column_number]).getData()[current->pos];
 

dbms/src/DataStreams/VersionedCollapsingSortedBlockInputStream.h

@@ -197,7 +197,7 @@ private:
     Int8 sign_in_queue = 0;
     const size_t max_rows_in_queue;
     /// Rows with the same primary key and sign.
-    FixedSizeDequeWithGaps<RowRef> current_keys;
+    FixedSizeDequeWithGaps<SharedBlockRowRef> current_keys;
 
     size_t blocks_written = 0;
 
@@ -207,7 +207,7 @@ private:
     void merge(MutableColumns & merged_columns, std::priority_queue<SortCursor> & queue);
 
     /// Output to result row for the current primary key.
-    void insertRow(size_t skip_rows, const RowRef & row, MutableColumns & merged_columns);
+    void insertRow(size_t skip_rows, const SharedBlockRowRef & row, MutableColumns & merged_columns);
 
     void insertGap(size_t gap_size);
 };

dbms/src/Functions/FunctionsBitmap.h

@@ -348,10 +348,10 @@ private:
             const UInt32 range_start = is_column_const[1] ? (*container1)[0] : (*container1)[i];
             const UInt32 range_end = is_column_const[2] ? (*container2)[0] : (*container2)[i];
 
-            auto bd2 = new AggregateFunctionGroupBitmapData<T>();
+            col_to->insertDefault();
-            bd0.rbs.rb_range(range_start, range_end, bd2->rbs);
+            AggregateFunctionGroupBitmapData<T> & bd2
-
+                = *reinterpret_cast<AggregateFunctionGroupBitmapData<T> *>(col_to->getData()[i]);
-            col_to->insertFrom(reinterpret_cast<ConstAggregateDataPtr>(bd2));
+            bd0.rbs.rb_range(range_start, range_end, bd2.rbs);
         }
         block.getByPosition(result).column = std::move(col_to);
     }

dbms/src/Functions/array/arrayElement.cpp

@@ -858,7 +858,7 @@ void FunctionArrayElement::perform(Block & block, const ColumnNumbers & argument
             builder.initSink(input_rows_count);
 
         if (index == 0u)
-            throw Exception("Array indices is 1-based", ErrorCodes::ZERO_ARRAY_OR_TUPLE_INDEX);
+            throw Exception("Array indices are 1-based", ErrorCodes::ZERO_ARRAY_OR_TUPLE_INDEX);
 
         if (!(executeNumberConst<UInt8>(block, arguments, result, index, builder)
             || executeNumberConst<UInt16>(block, arguments, result, index, builder)

dbms/src/IO/BufferBase.h

@@ -88,21 +88,12 @@ public:
     }
 
     /** How many bytes have been read/written, counting those that are still in the buffer. */
-    size_t count() const
+    size_t count() const { return bytes + offset(); }
-    {
-        return bytes + offset();
-    }
 
     /** Check that there is more bytes in buffer after cursor. */
-    bool ALWAYS_INLINE hasPendingData() const
+    bool ALWAYS_INLINE hasPendingData() const { return available() > 0; }
-    {
-        return pos != working_buffer.end();
-    }
 
-    bool isPadded() const
+    bool isPadded() const { return padded; }
-    {
-        return padded;
-    }
 
 protected:
     /// Read/write position.

dbms/src/IO/DelimitedReadBuffer.h

@@ -1,57 +0,0 @@
-#pragma once
-
-#include <IO/ReadBuffer.h>
-#include <Common/typeid_cast.h>
-
-namespace DB
-{
-/// Consistently reads from one sub-buffer in a circle, and delimits its output with a character.
-/// Owns sub-buffer.
-class DelimitedReadBuffer : public ReadBuffer
-{
-public:
-    DelimitedReadBuffer(std::unique_ptr<ReadBuffer> buffer_, char delimiter_) : ReadBuffer(nullptr, 0), buffer(std::move(buffer_)), delimiter(delimiter_)
-    {
-        // TODO: check that `buffer_` is not nullptr.
-    }
-
-    template <class BufferType>
-    BufferType * subBufferAs()
-    {
-        return typeid_cast<BufferType *>(buffer.get());
-    }
-
-    void reset()
-    {
-        BufferBase::set(nullptr, 0, 0);
-    }
-
-protected:
-    // XXX: don't know how to guarantee that the next call to this method is done after we read all previous data.
-    bool nextImpl() override
-    {
-        if (put_delimiter)
-        {
-            BufferBase::set(&delimiter, 1, 0);
-            put_delimiter = false;
-        }
-        else
-        {
-            if (!buffer->next())
-                return false;
-
-            BufferBase::set(buffer->position(), buffer->available(), 0);
-            put_delimiter = (delimiter != 0);
-        }
-
-        return true;
-    }
-
-private:
-    std::unique_ptr<ReadBuffer> buffer; // FIXME: should be `const`, but `ReadBuffer` doesn't allow
-    char delimiter; // FIXME: should be `const`, but `ReadBuffer` doesn't allow
-
-    bool put_delimiter = false;
-};
-
-}

dbms/src/IO/HDFSCommon.cpp

@@ -1,4 +1,5 @@
 #include <IO/HDFSCommon.h>
+#include <Poco/URI.h>
 
 #if USE_HDFS
 #include <Common/Exception.h>
@@ -11,8 +12,9 @@ extern const int BAD_ARGUMENTS;
 extern const int NETWORK_ERROR;
 }
 
-HDFSBuilderPtr createHDFSBuilder(const Poco::URI & uri)
+HDFSBuilderPtr createHDFSBuilder(const std::string & uri_str)
 {
+    const Poco::URI uri(uri_str);
     auto & host = uri.getHost();
     auto port = uri.getPort();
     auto & path = uri.getPath();

dbms/src/IO/HDFSCommon.h

@@ -1,7 +1,6 @@
 #include <Common/config.h>
 #include <memory>
 #include <type_traits>
-#include <Poco/URI.h>
 
 #if USE_HDFS
 #include <hdfs/hdfs.h>
@@ -27,12 +26,32 @@ struct HDFSFsDeleter
 
 }
 
+struct HDFSFileInfo
+{
+    hdfsFileInfo * file_info;
+    int length;
+
+    HDFSFileInfo()
+        : file_info(nullptr)
+        , length(0)
+    {
+    }
+    HDFSFileInfo(const HDFSFileInfo & other) = delete;
+    HDFSFileInfo(HDFSFileInfo && other) = default;
+    HDFSFileInfo & operator=(const HDFSFileInfo & other) = delete;
+    HDFSFileInfo & operator=(HDFSFileInfo && other) = default;
+
+    ~HDFSFileInfo()
+    {
+        hdfsFreeFileInfo(file_info, length);
+    }
+};
 using HDFSBuilderPtr = std::unique_ptr<hdfsBuilder, detail::HDFSBuilderDeleter>;
 using HDFSFSPtr = std::unique_ptr<std::remove_pointer_t<hdfsFS>, detail::HDFSFsDeleter>;
 
 // set read/connect timeout, default value in libhdfs3 is about 1 hour, and too large
 /// TODO Allow to tune from query Settings.
-HDFSBuilderPtr createHDFSBuilder(const Poco::URI & hdfs_uri);
+HDFSBuilderPtr createHDFSBuilder(const std::string & hdfs_uri);
 HDFSFSPtr createHDFSFS(hdfsBuilder * builder);
 }
 #endif

dbms/src/IO/ReadBufferFromHDFS.cpp

@@ -2,7 +2,6 @@
 
 #if USE_HDFS
 #include <IO/HDFSCommon.h>
-#include <Poco/URI.h>
 #include <hdfs/hdfs.h>
 
 
@@ -16,7 +15,7 @@ namespace ErrorCodes
 
 struct ReadBufferFromHDFS::ReadBufferFromHDFSImpl
 {
-    Poco::URI hdfs_uri;
+    std::string hdfs_uri;
     hdfsFile fin;
     HDFSBuilderPtr builder;
     HDFSFSPtr fs;
@@ -26,8 +25,8 @@ struct ReadBufferFromHDFS::ReadBufferFromHDFSImpl
         , builder(createHDFSBuilder(hdfs_uri))
         , fs(createHDFSFS(builder.get()))
     {
-
+        const size_t begin_of_path = hdfs_uri.find('/', hdfs_uri.find("//") + 2);
-        auto & path = hdfs_uri.getPath();
+        const std::string path = hdfs_uri.substr(begin_of_path);
         fin = hdfsOpenFile(fs.get(), path.c_str(), O_RDONLY, 0, 0, 0);
 
         if (fin == nullptr)
@@ -39,7 +38,7 @@ struct ReadBufferFromHDFS::ReadBufferFromHDFSImpl
     {
         int bytes_read = hdfsRead(fs.get(), fin, start, size);
         if (bytes_read < 0)
-            throw Exception("Fail to read HDFS file: " + hdfs_uri.toString() + " " + std::string(hdfsGetLastError()),
+            throw Exception("Fail to read HDFS file: " + hdfs_uri + " " + std::string(hdfsGetLastError()),
                 ErrorCodes::NETWORK_ERROR);
         return bytes_read;
     }

dbms/src/IO/WriteBufferFromHDFS.cpp

@@ -2,7 +2,6 @@
 
 #if USE_HDFS
 
-#include <Poco/URI.h>
 #include <IO/WriteBufferFromHDFS.h>
 #include <IO/HDFSCommon.h>
 #include <hdfs/hdfs.h>
@@ -21,7 +20,7 @@ extern const int CANNOT_FSYNC;
 
 struct WriteBufferFromHDFS::WriteBufferFromHDFSImpl
 {
-    Poco::URI hdfs_uri;
+    std::string hdfs_uri;
     hdfsFile fout;
     HDFSBuilderPtr builder;
     HDFSFSPtr fs;
@@ -31,7 +30,11 @@ struct WriteBufferFromHDFS::WriteBufferFromHDFSImpl
         , builder(createHDFSBuilder(hdfs_uri))
         , fs(createHDFSFS(builder.get()))
     {
-        auto & path = hdfs_uri.getPath();
+        const size_t begin_of_path = hdfs_uri.find('/', hdfs_uri.find("//") + 2);
+        const std::string path = hdfs_uri.substr(begin_of_path);
+        if (path.find("*?{") != std::string::npos)
+            throw Exception("URI '" + hdfs_uri + "' contains globs, so the table is in readonly mode", ErrorCodes::CANNOT_OPEN_FILE);
+
         fout = hdfsOpenFile(fs.get(), path.c_str(), O_WRONLY, 0, 0, 0);
 
         if (fout == nullptr)
@@ -52,7 +55,7 @@ struct WriteBufferFromHDFS::WriteBufferFromHDFSImpl
     {
         int bytes_written = hdfsWrite(fs.get(), fout, start, size);
         if (bytes_written < 0)
-            throw Exception("Fail to write HDFS file: " + hdfs_uri.toString() + " " + std::string(hdfsGetLastError()),
+            throw Exception("Fail to write HDFS file: " + hdfs_uri + " " + std::string(hdfsGetLastError()),
                 ErrorCodes::NETWORK_ERROR);
         return bytes_written;
     }
@@ -61,7 +64,7 @@ struct WriteBufferFromHDFS::WriteBufferFromHDFSImpl
     {
         int result = hdfsSync(fs.get(), fout);
         if (result < 0)
-            throwFromErrno("Cannot HDFS sync" + hdfs_uri.toString() + " " + std::string(hdfsGetLastError()),
+            throwFromErrno("Cannot HDFS sync" + hdfs_uri + " " + std::string(hdfsGetLastError()),
                 ErrorCodes::CANNOT_FSYNC);
     }
 };

dbms/src/Interpreters/AnalyzedJoin.cpp

@@ -93,14 +93,14 @@ NameSet AnalyzedJoin::getOriginalColumnsSet() const
     return out;
 }
 
-std::unordered_map<String, String> AnalyzedJoin::getOriginalColumnsMap(const NameSet & required_columns) const
+NamesWithAliases AnalyzedJoin::getNamesWithAliases(const NameSet & required_columns) const
 {
-    std::unordered_map<String, String> out;
+    NamesWithAliases out;
     for (const auto & column : required_columns)
     {
         auto it = original_names.find(column);
         if (it != original_names.end())
-            out.insert(*it);
+            out.emplace_back(it->second, it->first); /// {original_name, name}
     }
     return out;
 }
@@ -129,15 +129,15 @@ Names AnalyzedJoin::requiredJoinedNames() const
     return Names(required_columns_set.begin(), required_columns_set.end());
 }
 
-void AnalyzedJoin::appendRequiredColumns(const Block & sample, NameSet & required_columns) const
+NamesWithAliases AnalyzedJoin::getRequiredColumns(const Block & sample, const Names & action_required_columns) const
 {
-    for (auto & column : key_names_right)
+    NameSet required_columns(action_required_columns.begin(), action_required_columns.end());
+
+    for (auto & column : requiredJoinedNames())
         if (!sample.has(column))
             required_columns.insert(column);
 
-    for (auto & column : columns_added_by_join)
+    return getNamesWithAliases(required_columns);
-        if (!sample.has(column.name))
-            required_columns.insert(column.name);
 }
 
 void AnalyzedJoin::addJoinedColumn(const NameAndTypePair & joined_column)

dbms/src/Interpreters/AnalyzedJoin.h

@@ -64,12 +64,12 @@ public:
 
     NameSet getQualifiedColumnsSet() const;
     NameSet getOriginalColumnsSet() const;
-    std::unordered_map<String, String> getOriginalColumnsMap(const NameSet & required_columns) const;
+    NamesWithAliases getNamesWithAliases(const NameSet & required_columns) const;
+    NamesWithAliases getRequiredColumns(const Block & sample, const Names & action_columns) const;
 
     void deduplicateAndQualifyColumnNames(const NameSet & left_table_columns, const String & right_table_prefix);
     size_t rightKeyInclusion(const String & name) const;
 
-    void appendRequiredColumns(const Block & sample, NameSet & required_columns) const;
     void addJoinedColumn(const NameAndTypePair & joined_column);
     void addJoinedColumnsAndCorrectNullability(Block & sample_block) const;
 

dbms/src/Interpreters/Cluster.cpp

@@ -347,7 +347,8 @@ Cluster::Cluster(const Poco::Util::AbstractConfiguration & config, const Setting
             }
 
             ConnectionPoolWithFailoverPtr shard_pool = std::make_shared<ConnectionPoolWithFailover>(
-                        all_replicas_pools, settings.load_balancing);
+                        all_replicas_pools, settings.load_balancing,
+                        settings.distributed_replica_error_half_life.totalSeconds(), settings.distributed_replica_error_cap);
 
             if (weight)
                 slot_to_shard.insert(std::end(slot_to_shard), weight, shards_info.size());
@@ -398,7 +399,8 @@ Cluster::Cluster(const Settings & settings, const std::vector<std::vector<String
         }
 
         ConnectionPoolWithFailoverPtr shard_pool = std::make_shared<ConnectionPoolWithFailover>(
-                all_replicas, settings.load_balancing);
+                all_replicas, settings.load_balancing,
+                settings.distributed_replica_error_half_life.totalSeconds(), settings.distributed_replica_error_cap);
 
         slot_to_shard.insert(std::end(slot_to_shard), default_weight, shards_info.size());
         shards_info.push_back({{}, current_shard_num, default_weight, std::move(shard_local_addresses), std::move(shard_pool),

dbms/src/Interpreters/ClusterProxy/executeQuery.cpp

@@ -30,7 +30,7 @@ Context removeUserRestrictionsFromSettings(const Context & context, const Settin
     /// Set as unchanged to avoid sending to remote server.
     new_settings.max_concurrent_queries_for_user.changed = false;
     new_settings.max_memory_usage_for_user.changed = false;
-    new_settings.max_memory_usage_for_all_queries = false;
+    new_settings.max_memory_usage_for_all_queries.changed = false;
 
     Context new_context(context);
     new_context.setSettings(new_settings);

dbms/src/Interpreters/Context.cpp

@@ -143,8 +143,6 @@ struct ContextShared
     std::unique_ptr<DDLWorker> ddl_worker;                  /// Process ddl commands from zk.
     /// Rules for selecting the compression settings, depending on the size of the part.
     mutable std::unique_ptr<CompressionCodecSelector> compression_codec_selector;
-    /// Allows to remove sensitive data from queries using set of regexp-based rules
-    std::unique_ptr<SensitiveDataMasker> sensitive_data_masker;
     std::optional<MergeTreeSettings> merge_tree_settings; /// Settings of MergeTree* engines.
     size_t max_table_size_to_drop = 50000000000lu;          /// Protects MergeTree tables from accidental DROP (50GB by default)
     size_t max_partition_size_to_drop = 50000000000lu;      /// Protects MergeTree partitions from accidental DROP (50GB by default)
@@ -287,8 +285,6 @@ struct ContextShared
 
         /// Stop trace collector if any
         trace_collector.reset();
-
-        sensitive_data_masker.reset();
     }
 
     bool hasTraceCollector()
@@ -538,23 +534,6 @@ String Context::getUserFilesPath() const
     return shared->user_files_path;
 }
 
-void Context::setSensitiveDataMasker(std::unique_ptr<SensitiveDataMasker> sensitive_data_masker)
-{
-    if (!sensitive_data_masker)
-        throw Exception("Logical error: the 'sensitive_data_masker' is not set", ErrorCodes::LOGICAL_ERROR);
-
-    if (sensitive_data_masker->rulesCount() > 0)
-    {
-        auto lock = getLock();
-        shared->sensitive_data_masker = std::move(sensitive_data_masker);
-    }
-}
-
-SensitiveDataMasker * Context::getSensitiveDataMasker() const
-{
-    return shared->sensitive_data_masker.get();
-}
-
 void Context::setPath(const String & path)
 {
     auto lock = getLock();
@@ -1222,8 +1201,8 @@ void Context::setCurrentQueryId(const String & query_id)
             } words;
         } random;
 
-        random.words.a = thread_local_rng();
+        random.words.a = thread_local_rng(); //-V656
-        random.words.b = thread_local_rng();
+        random.words.b = thread_local_rng(); //-V656
 
         /// Use protected constructor.
         struct qUUID : Poco::UUID
@@ -2053,6 +2032,51 @@ void Context::initializeExternalTablesIfSet()
 }
 
 
+void Context::setInputInitializer(InputInitializer && initializer)
+{
+    if (input_initializer_callback)
+        throw Exception("Input initializer is already set", ErrorCodes::LOGICAL_ERROR);
+
+    input_initializer_callback = std::move(initializer);
+}
+
+
+void Context::initializeInput(const StoragePtr & input_storage)
+{
+    if (!input_initializer_callback)
+        throw Exception("Input initializer is not set", ErrorCodes::LOGICAL_ERROR);
+
+    input_initializer_callback(*this, input_storage);
+    /// Reset callback
+    input_initializer_callback = {};
+}
+
+
+void Context::setInputBlocksReaderCallback(InputBlocksReader && reader)
+{
+    if (input_blocks_reader)
+        throw Exception("Input blocks reader is already set", ErrorCodes::LOGICAL_ERROR);
+
+    input_blocks_reader = std::move(reader);
+}
+
+
+InputBlocksReader Context::getInputBlocksReaderCallback() const
+{
+    return input_blocks_reader;
+}
+
+
+void Context::resetInputCallbacks()
+{
+    if (input_initializer_callback)
+        input_initializer_callback = {};
+
+    if (input_blocks_reader)
+        input_blocks_reader = {};
+}
+
+
 SessionCleaner::~SessionCleaner()
 {
     try

dbms/src/Interpreters/Context.h

@@ -12,7 +12,6 @@
 #include <Common/MultiVersion.h>
 #include <Common/ThreadPool.h>
 #include "config_core.h"
-#include <Common/SensitiveDataMasker.h>
 #include <Storages/IStorage_fwd.h>
 #include <atomic>
 #include <chrono>
@@ -101,6 +100,11 @@ using TableAndCreateASTs = std::map<String, TableAndCreateAST>;
 /// Callback for external tables initializer
 using ExternalTablesInitializer = std::function<void(Context &)>;
 
+/// Callback for initialize input()
+using InputInitializer = std::function<void(Context &, const StoragePtr &)>;
+/// Callback for reading blocks of data from client for function input()
+using InputBlocksReader = std::function<Block(Context &)>;
+
 /// An empty interface for an arbitrary object that may be attached by a shared pointer
 /// to query context, when using ClickHouse as a library.
 struct IHostContext
@@ -125,6 +129,9 @@ private:
     ClientInfo client_info;
     ExternalTablesInitializer external_tables_initializer_callback;
 
+    InputInitializer input_initializer_callback;
+    InputBlocksReader input_blocks_reader;
+
     std::shared_ptr<QuotaForIntervals> quota;           /// Current quota. By default - empty quota, that have no limits.
     String current_database;
     Settings settings;                                  /// Setting for query execution.
@@ -178,9 +185,6 @@ public:
     String getFlagsPath() const;
     String getUserFilesPath() const;
 
-    void setSensitiveDataMasker(std::unique_ptr<SensitiveDataMasker> sensitive_data_masker);
-    SensitiveDataMasker * getSensitiveDataMasker() const;
-
     void setPath(const String & path);
     void setTemporaryPath(const String & path);
     void setFlagsPath(const String & path);
@@ -217,6 +221,17 @@ public:
     /// This method is called in executeQuery() and will call the external tables initializer.
     void initializeExternalTablesIfSet();
 
+    /// When input() is present we have to send columns structure to client
+    void setInputInitializer(InputInitializer && initializer);
+    /// This method is called in StorageInput::read while executing query
+    void initializeInput(const StoragePtr & input_storage);
+
+    /// Callback for read data blocks from client one by one for function input()
+    void setInputBlocksReaderCallback(InputBlocksReader && reader);
+    /// Get callback for reading data for input()
+    InputBlocksReader getInputBlocksReaderCallback() const;
+    void resetInputCallbacks();
+
     ClientInfo & getClientInfo() { return client_info; }
     const ClientInfo & getClientInfo() const { return client_info; }
 

dbms/src/Interpreters/DatabaseAndTableWithAlias.cpp

@@ -1,6 +1,5 @@
 #include <Interpreters/DatabaseAndTableWithAlias.h>
 #include <Interpreters/IdentifierSemantic.h>
-#include <Interpreters/AnalyzedJoin.h> /// for getNamesAndTypeListFromTableExpression
 #include <Interpreters/Context.h>
 #include <Common/typeid_cast.h>
 
@@ -15,6 +14,8 @@ namespace DB
 
 NameSet removeDuplicateColumns(NamesAndTypesList & columns);
 
+struct ASTTableExpression;
+NamesAndTypesList getNamesAndTypeListFromTableExpression(const ASTTableExpression & table_expression, const Context & context);
 
 DatabaseAndTableWithAlias::DatabaseAndTableWithAlias(const ASTIdentifier & identifier, const String & current_database)
 {

dbms/src/Interpreters/ExpressionActions.h

@@ -20,9 +20,6 @@ namespace ErrorCodes
     extern const int LOGICAL_ERROR;
 }
 
-using NameWithAlias = std::pair<std::string, std::string>;
-using NamesWithAliases = std::vector<NameWithAlias>;
-
 class AnalyzedJoin;
 
 class IPreparedFunction;

dbms/src/Interpreters/ExpressionAnalyzer.cpp

@@ -29,6 +29,7 @@
 #include <Interpreters/PredicateExpressionsOptimizer.h>
 #include <Interpreters/ExternalDictionaries.h>
 #include <Interpreters/Set.h>
+#include <Interpreters/AnalyzedJoin.h>
 
 #include <AggregateFunctions/AggregateFunctionFactory.h>
 #include <AggregateFunctions/parseAggregateFunctionParameters.h>
@@ -417,8 +418,7 @@ bool SelectQueryExpressionAnalyzer::appendJoin(ExpressionActionsChain & chain, b
     if (!ast_join)
         return false;
 
-    SubqueryForSet & subquery_for_set = getSubqueryForJoin(*ast_join);
+    makeTableJoin(*ast_join);
-    syntax->analyzed_join->setHashJoin(subquery_for_set.join);
 
     initChain(chain, sourceColumns());
     ExpressionActionsChain::Step & step = chain.steps.back();
@@ -456,7 +456,15 @@ static JoinPtr tryGetStorageJoin(const ASTTablesInSelectQueryElement & join_elem
     return {};
 }
 
-SubqueryForSet & SelectQueryExpressionAnalyzer::getSubqueryForJoin(const ASTTablesInSelectQueryElement & join_element)
+static ExpressionActionsPtr createJoinedBlockActions(const Context & context, const AnalyzedJoin & analyzed_join)
+{
+    ASTPtr expression_list = analyzed_join.rightKeysList();
+    auto syntax_result = SyntaxAnalyzer(context).analyze(expression_list,
+                                                         analyzed_join.columnsFromJoinedTable(), analyzed_join.requiredJoinedNames());
+    return ExpressionAnalyzer(expression_list, syntax_result, context).getActions(true, false);
+}
+
+void SelectQueryExpressionAnalyzer::makeTableJoin(const ASTTablesInSelectQueryElement & join_element)
 {
     /// Two JOINs are not supported with the same subquery, but different USINGs.
     auto join_hash = join_element.getTreeHash();
@@ -469,64 +477,45 @@ SubqueryForSet & SelectQueryExpressionAnalyzer::getSubqueryForJoin(const ASTTabl
         subquery_for_set.join = tryGetStorageJoin(join_element, context);
 
     if (!subquery_for_set.join)
-        makeHashJoin(join_element, subquery_for_set);
+    {
+        /// Actions which need to be calculated on joined block.
+        ExpressionActionsPtr joined_block_actions = createJoinedBlockActions(context, analyzedJoin());
 
-    return subquery_for_set;
+        if (!subquery_for_set.source)
+            makeSubqueryForJoin(join_element, joined_block_actions, subquery_for_set);
+
+        /// Test actions on sample block (early error detection)
+        Block sample_block = subquery_for_set.renamedSampleBlock();
+        joined_block_actions->execute(sample_block);
+
+        /// TODO You do not need to set this up when JOIN is only needed on remote servers.
+        subquery_for_set.join = analyzedJoin().makeHashJoin(sample_block, settings.size_limits_for_join);
+        subquery_for_set.joined_block_actions = joined_block_actions;
+    }
+
+    syntax->analyzed_join->setHashJoin(subquery_for_set.join);
 }
 
-void SelectQueryExpressionAnalyzer::makeHashJoin(const ASTTablesInSelectQueryElement & join_element,
+void SelectQueryExpressionAnalyzer::makeSubqueryForJoin(const ASTTablesInSelectQueryElement & join_element,
-                                                 SubqueryForSet & subquery_for_set) const
+                                                        const ExpressionActionsPtr & joined_block_actions,
+                                                        SubqueryForSet & subquery_for_set) const
 {
-    /// Actions which need to be calculated on joined block.
-    ExpressionActionsPtr joined_block_actions = createJoinedBlockActions();
-
     /** For GLOBAL JOINs (in the case, for example, of the push method for executing GLOBAL subqueries), the following occurs
         * - in the addExternalStorage function, the JOIN (SELECT ...) subquery is replaced with JOIN _data1,
         *   in the subquery_for_set object this subquery is exposed as source and the temporary table _data1 as the `table`.
         * - this function shows the expression JOIN _data1.
         */
-    if (!subquery_for_set.source)
-    {
-        ASTPtr table;
 
-        auto & table_to_join = join_element.table_expression->as<ASTTableExpression &>();
+    NamesWithAliases required_columns_with_aliases =
-        if (table_to_join.subquery)
+        analyzedJoin().getRequiredColumns(joined_block_actions->getSampleBlock(), joined_block_actions->getRequiredColumns());
-            table = table_to_join.subquery;
-        else if (table_to_join.table_function)
-            table = table_to_join.table_function;
-        else if (table_to_join.database_and_table_name)
-            table = table_to_join.database_and_table_name;
 
-        Names action_columns = joined_block_actions->getRequiredColumns();
+    Names original_columns;
-        NameSet required_columns(action_columns.begin(), action_columns.end());
+    for (auto & pr : required_columns_with_aliases)
+        original_columns.push_back(pr.first);
 
-        analyzedJoin().appendRequiredColumns(joined_block_actions->getSampleBlock(), required_columns);
+    auto interpreter = interpretSubquery(join_element.table_expression, context, subquery_depth, original_columns);
 
-        auto original_map = analyzedJoin().getOriginalColumnsMap(required_columns);
+    subquery_for_set.makeSource(interpreter, std::move(required_columns_with_aliases));
-        Names original_columns;
-        for (auto & pr : original_map)
-            original_columns.push_back(pr.second);
-
-        auto interpreter = interpretSubquery(table, context, subquery_depth, original_columns);
-
-        subquery_for_set.makeSource(interpreter, original_map);
-    }
-
-    Block sample_block = subquery_for_set.renamedSampleBlock();
-    joined_block_actions->execute(sample_block);
-
-    /// TODO You do not need to set this up when JOIN is only needed on remote servers.
-    subquery_for_set.join = analyzedJoin().makeHashJoin(sample_block, settings.size_limits_for_join);
-    subquery_for_set.joined_block_actions = joined_block_actions;
-}
-
-ExpressionActionsPtr SelectQueryExpressionAnalyzer::createJoinedBlockActions() const
-{
-    ASTPtr expression_list = analyzedJoin().rightKeysList();
-    Names required_columns = analyzedJoin().requiredJoinedNames();
-
-    auto syntax_result = SyntaxAnalyzer(context).analyze(expression_list, analyzedJoin().columnsFromJoinedTable(), required_columns);
-    return ExpressionAnalyzer(expression_list, syntax_result, context).getActions(true, false);
 }
 
 bool SelectQueryExpressionAnalyzer::appendPrewhere(

dbms/src/Interpreters/ExpressionAnalyzer.h

@@ -26,9 +26,6 @@ class ASTExpressionList;
 class ASTSelectQuery;
 struct ASTTablesInSelectQueryElement;
 
-struct SyntaxAnalyzerResult;
-using SyntaxAnalyzerResultPtr = std::shared_ptr<const SyntaxAnalyzerResult>;
-
 /// ExpressionAnalyzer sources, intermediates and results. It splits data and logic, allows to test them separately.
 struct ExpressionAnalyzerData
 {
@@ -222,9 +219,9 @@ private:
       */
     void tryMakeSetForIndexFromSubquery(const ASTPtr & subquery_or_table_name);
 
-    SubqueryForSet & getSubqueryForJoin(const ASTTablesInSelectQueryElement & join_element);
+    void makeTableJoin(const ASTTablesInSelectQueryElement & join_element);
-    ExpressionActionsPtr createJoinedBlockActions() const;
+    void makeSubqueryForJoin(const ASTTablesInSelectQueryElement & join_element, const ExpressionActionsPtr & joined_block_actions,
-    void makeHashJoin(const ASTTablesInSelectQueryElement & join_element, SubqueryForSet & subquery_for_set) const;
+                             SubqueryForSet & subquery_for_set) const;
 
     const ASTSelectQuery * getAggregatingQuery() const;
 };

dbms/src/Interpreters/FillingRow.cpp

@@ -0,0 +1,127 @@
+#include <Interpreters/FillingRow.h>
+
+namespace DB
+{
+
+bool less(const Field & lhs, const Field & rhs, int direction)
+{
+    if (direction == -1)
+        return applyVisitor(FieldVisitorAccurateLess(), rhs, lhs);
+
+    return applyVisitor(FieldVisitorAccurateLess(), lhs, rhs);
+}
+
+bool equals(const Field & lhs, const Field & rhs)
+{
+    return applyVisitor(FieldVisitorAccurateEquals(), lhs, rhs);
+}
+
+
+FillingRow::FillingRow(const SortDescription & description_) : description(description_)
+{
+    row.resize(description.size());
+}
+
+bool FillingRow::operator<(const FillingRow & other) const
+{
+    for (size_t i = 0; i < size(); ++i)
+    {
+        if (row[i].isNull() || other[i].isNull() || equals(row[i], other[i]))
+            continue;
+        return less(row[i], other[i], getDirection(i));
+    }
+    return false;
+}
+
+bool FillingRow::operator==(const FillingRow & other) const
+{
+    for (size_t i = 0; i < size(); ++i)
+        if (!equals(row[i], other[i]))
+            return false;
+    return true;
+}
+
+bool FillingRow::next(const FillingRow & to_row)
+{
+    size_t pos = 0;
+
+    /// Find position we need to increment for generating next row.
+    for (; pos < row.size(); ++pos)
+        if (!row[pos].isNull() && !to_row[pos].isNull() && !equals(row[pos], to_row[pos]))
+            break;
+
+    if (pos == row.size() || less(to_row[pos], row[pos], getDirection(pos)))
+        return false;
+
+    /// If we have any 'fill_to' value at position greater than 'pos',
+    ///  we need to generate rows up to 'fill_to' value.
+    for (size_t i = row.size() - 1; i > pos; --i)
+    {
+        if (getFillDescription(i).fill_to.isNull() || row[i].isNull())
+            continue;
+
+        auto next_value = row[i];
+        applyVisitor(FieldVisitorSum(getFillDescription(i).fill_step), next_value);
+        if (less(next_value, getFillDescription(i).fill_to, getDirection(i)))
+        {
+            row[i] = next_value;
+            initFromDefaults(i + 1);
+            return true;
+        }
+    }
+
+    auto next_value = row[pos];
+    applyVisitor(FieldVisitorSum(getFillDescription(pos).fill_step), next_value);
+
+    if (less(to_row[pos], next_value, getDirection(pos)))
+        return false;
+
+    row[pos] = next_value;
+    if (equals(row[pos], to_row[pos]))
+    {
+        bool is_less = false;
+        for (size_t i = pos + 1; i < size(); ++i)
+        {
+            const auto & fill_from = getFillDescription(i).fill_from;
+            if (!fill_from.isNull())
+                row[i] = fill_from;
+            else
+                row[i] = to_row[i];
+            is_less |= less(row[i], to_row[i], getDirection(i));
+        }
+
+        return is_less;
+    }
+
+    initFromDefaults(pos + 1);
+    return true;
+}
+
+void FillingRow::initFromDefaults(size_t from_pos)
+{
+    for (size_t i = from_pos; i < row.size(); ++i)
+        row[i] = getFillDescription(i).fill_from;
+}
+
+
+void insertFromFillingRow(MutableColumns & filling_columns, MutableColumns & other_columns, const FillingRow & filling_row)
+{
+    for (size_t i = 0; i < filling_columns.size(); ++i)
+    {
+        if (filling_row[i].isNull())
+            filling_columns[i]->insertDefault();
+        else
+            filling_columns[i]->insert(filling_row[i]);
+    }
+
+    for (size_t i = 0; i < other_columns.size(); ++i)
+        other_columns[i]->insertDefault();
+}
+
+void copyRowFromColumns(MutableColumns & dest, const Columns & source, size_t row_num)
+{
+    for (size_t i = 0; i < source.size(); ++i)
+        dest[i]->insertFrom(*source[i], row_num);
+}
+
+}

dbms/src/Interpreters/FillingRow.h

@@ -0,0 +1,44 @@
+#pragma once
+#include <Core/SortDescription.h>
+#include <Columns/IColumn.h>
+#include <Common/FieldVisitors.h>
+
+namespace DB
+{
+
+/// Compares fields in terms of sorting order, considering direction.
+bool less(const Field & lhs, const Field & rhs, int direction);
+bool equals(const Field & lhs, const Field & rhs);
+
+/** Helps to implement modifier WITH FILL for ORDER BY clause.
+ *  Stores row as array of fields and provides functions to generate next row for filling gaps and for comparing rows.
+ *  Used in FillingBlockInputStream and in FillingTransform.
+ */
+class FillingRow
+{
+public:
+    FillingRow(const SortDescription & sort_description);
+
+    /// Generates next row according to fill 'from', 'to' and 'step' values.
+    bool next(const FillingRow & to_row);
+
+    void initFromDefaults(size_t from_pos = 0);
+
+    Field & operator[](size_t ind) { return row[ind]; }
+    const Field & operator[](size_t ind) const { return row[ind]; }
+    size_t size() const { return row.size(); }
+    bool operator<(const FillingRow & other) const;
+    bool operator==(const FillingRow & other) const;
+
+    int getDirection(size_t ind) const { return description[ind].direction; }
+    FillColumnDescription & getFillDescription(size_t ind) { return description[ind].fill_description; }
+
+private:
+    std::vector<Field> row;
+    SortDescription description;
+};
+
+void insertFromFillingRow(MutableColumns & filling_columns, MutableColumns & other_columns, const FillingRow & filling_row);
+void copyRowFromColumns(MutableColumns & dest, const Columns & source, size_t row_num);
+
+}

dbms/src/Interpreters/InterpreterInsertQuery.cpp

@@ -155,7 +155,7 @@ BlockIO InterpreterInsertQuery::execute()
     }
     else if (query.data && !query.has_tail) /// can execute without additional data
     {
-        res.in = std::make_shared<InputStreamFromASTInsertQuery>(query_ptr, nullptr, query_sample_block, context);
+        res.in = std::make_shared<InputStreamFromASTInsertQuery>(query_ptr, nullptr, query_sample_block, context, nullptr);
         res.in = std::make_shared<NullAndDoCopyBlockInputStream>(res.in, res.out);
         res.out = nullptr;
     }

dbms/src/Interpreters/InterpreterSelectQuery.cpp

@@ -24,6 +24,7 @@
 #include <DataStreams/ConvertColumnLowCardinalityToFullBlockInputStream.h>
 #include <DataStreams/ConvertingBlockInputStream.h>
 #include <DataStreams/ReverseBlockInputStream.h>
+#include <DataStreams/FillingBlockInputStream.h>
 
 #include <Parsers/ASTFunction.h>
 #include <Parsers/ASTIdentifier.h>
@@ -44,6 +45,7 @@
 #include <Interpreters/DatabaseAndTableWithAlias.h>
 #include <Interpreters/JoinToSubqueryTransformVisitor.h>
 #include <Interpreters/CrossToInnerJoinVisitor.h>
+#include <Interpreters/AnalyzedJoin.h>
 
 #include <Storages/MergeTree/MergeTreeData.h>
 #include <Storages/MergeTree/MergeTreeWhereOptimizer.h>
@@ -57,6 +59,7 @@
 #include <Core/Field.h>
 #include <Core/Types.h>
 #include <Columns/Collator.h>
+#include <Common/FieldVisitors.h>
 #include <Common/typeid_cast.h>
 #include <Common/checkStackSize.h>
 #include <Parsers/queryToString.h>
@@ -81,6 +84,7 @@
 #include <Processors/Transforms/CreatingSetsTransform.h>
 #include <Processors/Transforms/RollupTransform.h>
 #include <Processors/Transforms/CubeTransform.h>
+#include <Processors/Transforms/FillingTransform.h>
 #include <Processors/LimitTransform.h>
 #include <Processors/Transforms/FinishSortingTransform.h>
 #include <DataTypes/DataTypeAggregateFunction.h>
@@ -103,6 +107,7 @@ namespace ErrorCodes
     extern const int PARAMETER_OUT_OF_BOUND;
     extern const int ARGUMENT_OUT_OF_BOUND;
     extern const int INVALID_LIMIT_EXPRESSION;
+    extern const int INVALID_WITH_FILL_EXPRESSION;
 }
 
 namespace
@@ -414,8 +419,6 @@ QueryPipeline InterpreterSelectQuery::executeWithProcessors()
 
 Block InterpreterSelectQuery::getSampleBlockImpl()
 {
-    FilterInfoPtr filter_info;
-
     /// Need to create sets before analyzeExpressions(). Otherwise some sets for index won't be created.
     query_analyzer->makeSetsForIndex(getSelectQuery().where());
     query_analyzer->makeSetsForIndex(getSelectQuery().prewhere());
@@ -427,8 +430,9 @@ Block InterpreterSelectQuery::getSampleBlockImpl()
             options.to_stage,
             context,
             storage,
-            true,
+            true,  // only_types
-            filter_info);
+            {}     // filter_info
+        );
 
     if (options.to_stage == QueryProcessingStage::Enum::FetchColumns)
     {
@@ -682,8 +686,62 @@ InterpreterSelectQuery::analyzeExpressions(
     return res;
 }
 
+static Field getWithFillFieldValue(const ASTPtr & node, const Context & context)
+{
+    const auto & [field, type] = evaluateConstantExpression(node, context);
 
-static SortDescription getSortDescription(const ASTSelectQuery & query)
+    if (!isColumnedAsNumber(type))
+        throw Exception("Illegal type " + type->getName() + " of WITH FILL expression, must be numeric type", ErrorCodes::INVALID_WITH_FILL_EXPRESSION);
+
+    return field;
+}
+
+static FillColumnDescription getWithFillDescription(const ASTOrderByElement & order_by_elem, const Context & context)
+{
+    FillColumnDescription descr;
+    if (order_by_elem.fill_from)
+        descr.fill_from = getWithFillFieldValue(order_by_elem.fill_from, context);
+    if (order_by_elem.fill_to)
+        descr.fill_to = getWithFillFieldValue(order_by_elem.fill_to, context);
+    if (order_by_elem.fill_step)
+        descr.fill_step = getWithFillFieldValue(order_by_elem.fill_step, context);
+    else
+        descr.fill_step = order_by_elem.direction;
+
+    if (applyVisitor(FieldVisitorAccurateEquals(), descr.fill_step, Field{0}))
+        throw Exception("WITH FILL STEP value cannot be zero", ErrorCodes::INVALID_WITH_FILL_EXPRESSION);
+
+    if (order_by_elem.direction == 1)
+    {
+        if (applyVisitor(FieldVisitorAccurateLess(), descr.fill_step, Field{0}))
+            throw Exception("WITH FILL STEP value cannot be negative for sorting in ascending direction",
+                ErrorCodes::INVALID_WITH_FILL_EXPRESSION);
+
+        if (!descr.fill_from.isNull() && !descr.fill_to.isNull() &&
+            applyVisitor(FieldVisitorAccurateLess(), descr.fill_to, descr.fill_from))
+        {
+            throw Exception("WITH FILL TO value cannot be less than FROM value for sorting in ascending direction",
+                ErrorCodes::INVALID_WITH_FILL_EXPRESSION);
+        }
+    }
+    else
+    {
+        if (applyVisitor(FieldVisitorAccurateLess(), Field{0}, descr.fill_step))
+            throw Exception("WITH FILL STEP value cannot be positive for sorting in descending direction",
+                ErrorCodes::INVALID_WITH_FILL_EXPRESSION);
+
+        if (!descr.fill_from.isNull() && !descr.fill_to.isNull() &&
+            applyVisitor(FieldVisitorAccurateLess(), descr.fill_from, descr.fill_to))
+        {
+            throw Exception("WITH FILL FROM value cannot be less than TO value for sorting in descending direction",
+                ErrorCodes::INVALID_WITH_FILL_EXPRESSION);
+        }
+    }
+
+    return descr;
+}
+
+static SortDescription getSortDescription(const ASTSelectQuery & query, const Context & context)
 {
     SortDescription order_descr;
     order_descr.reserve(query.orderBy()->children.size());
@@ -696,13 +754,19 @@ static SortDescription getSortDescription(const ASTSelectQuery & query)
         if (order_by_elem.collation)
             collator = std::make_shared<Collator>(order_by_elem.collation->as<ASTLiteral &>().value.get<String>());
 
-        order_descr.emplace_back(name, order_by_elem.direction, order_by_elem.nulls_direction, collator);
+        if (order_by_elem.with_fill)
+        {
+            FillColumnDescription fill_desc = getWithFillDescription(order_by_elem, context);
+            order_descr.emplace_back(name, order_by_elem.direction,
+                order_by_elem.nulls_direction, collator, true, fill_desc);
+        }
+        else
+            order_descr.emplace_back(name, order_by_elem.direction, order_by_elem.nulls_direction, collator);
     }
 
     return order_descr;
 }
 
-
 static UInt64 getLimitUIntValue(const ASTPtr & node, const Context & context)
 {
     const auto & [field, type] = evaluateConstantExpression(node, context);
@@ -737,7 +801,7 @@ static std::pair<UInt64, UInt64> getLimitLengthAndOffset(const ASTSelectQuery &
 static UInt64 getLimitForSorting(const ASTSelectQuery & query, const Context & context)
 {
     /// Partial sort can be done if there is LIMIT but no DISTINCT or LIMIT BY.
-    if (!query.distinct && !query.limitBy())
+    if (!query.distinct && !query.limitBy() && !query.limit_with_ties)
     {
         auto [limit_length, limit_offset] = getLimitLengthAndOffset(query, context);
         return limit_length + limit_offset;
@@ -752,7 +816,7 @@ static SortingInfoPtr optimizeReadInOrder(const MergeTreeData & merge_tree, cons
     if (!merge_tree.hasSortingKey())
         return {};
 
-    auto order_descr = getSortDescription(query);
+    auto order_descr = getSortDescription(query, context);
     SortDescription prefix_order_descr;
     int read_direction = order_descr.at(0).direction;
 
@@ -927,6 +991,21 @@ void InterpreterSelectQuery::executeImpl(TPipeline & pipeline, const BlockInputS
                 pipeline.streams.back() = std::make_shared<FilterBlockInputStream>(
                     pipeline.streams.back(), expressions.prewhere_info->prewhere_actions,
                     expressions.prewhere_info->prewhere_column_name, expressions.prewhere_info->remove_prewhere_column);
+
+            // To remove additional columns in dry run
+            // For example, sample column which can be removed in this stage
+            if (expressions.prewhere_info->remove_columns_actions)
+            {
+                if constexpr (pipeline_with_processors)
+                {
+                    pipeline.addSimpleTransform([&](const Block & header)
+                    {
+                        return std::make_shared<ExpressionTransform>(header, expressions.prewhere_info->remove_columns_actions);
+                    });
+                }
+                else
+                    pipeline.streams.back() = std::make_shared<ExpressionBlockInputStream>(pipeline.streams.back(), expressions.prewhere_info->remove_columns_actions);
+            }
         }
     }
     else
@@ -1174,7 +1253,7 @@ void InterpreterSelectQuery::executeImpl(TPipeline & pipeline, const BlockInputS
             /** Optimization - if there are several sources and there is LIMIT, then first apply the preliminary LIMIT,
               * limiting the number of rows in each up to `offset + limit`.
               */
-            if (query.limitLength() && pipeline.hasMoreThanOneStream() && !query.distinct && !expressions.has_limit_by && !settings.extremes)
+            if (query.limitLength() && !query.limit_with_ties && pipeline.hasMoreThanOneStream() && !query.distinct && !expressions.has_limit_by && !settings.extremes)
             {
                 executePreLimit(pipeline);
             }
@@ -1207,6 +1286,8 @@ void InterpreterSelectQuery::executeImpl(TPipeline & pipeline, const BlockInputS
                 executeLimitBy(pipeline);
             }
 
+            executeWithFill(pipeline);
+
             /** We must do projection after DISTINCT because projection may remove some columns.
               */
             executeProjection(pipeline, expressions.final_projection);
@@ -1223,7 +1304,6 @@ void InterpreterSelectQuery::executeImpl(TPipeline & pipeline, const BlockInputS
         executeSubqueriesInSetsAndJoins(pipeline, expressions.subqueries_for_sets);
 }
 
-
 template <typename TPipeline>
 void InterpreterSelectQuery::executeFetchColumns(
         QueryProcessingStage::Enum processing_stage, TPipeline & pipeline,
@@ -1421,11 +1501,12 @@ void InterpreterSelectQuery::executeFetchColumns(
 
     auto [limit_length, limit_offset] = getLimitLengthAndOffset(query, context);
 
-    /** Optimization - if not specified DISTINCT, WHERE, GROUP, HAVING, ORDER, LIMIT BY but LIMIT is specified, and limit + offset < max_block_size,
+    /** Optimization - if not specified DISTINCT, WHERE, GROUP, HAVING, ORDER, LIMIT BY, WITH TIES but LIMIT is specified, and limit + offset < max_block_size,
      *  then as the block size we will use limit + offset (not to read more from the table than requested),
      *  and also set the number of threads to 1.
      */
     if (!query.distinct
+        && !query.limit_with_ties
         && !query.prewhere()
         && !query.where()
         && !query.groupBy()
@@ -1496,12 +1577,22 @@ void InterpreterSelectQuery::executeFetchColumns(
             streams = {std::make_shared<NullBlockInputStream>(storage->getSampleBlockForColumns(required_columns))};
 
             if (query_info.prewhere_info)
+            {
                 streams.back() = std::make_shared<FilterBlockInputStream>(
                     streams.back(),
                     prewhere_info->prewhere_actions,
                     prewhere_info->prewhere_column_name,
                     prewhere_info->remove_prewhere_column);
 
+                // To remove additional columns
+                // In some cases, we did not read any marks so that the pipeline.streams is empty
+                // Thus, some columns in prewhere are not removed as expected
+                // This leads to mismatched header in distributed table
+                if (query_info.prewhere_info->remove_columns_actions)
+                {
+                    streams.back() = std::make_shared<ExpressionBlockInputStream>(streams.back(), query_info.prewhere_info->remove_columns_actions);
+                }
+            }
         }
 
         for (auto & stream : streams)
@@ -2011,7 +2102,7 @@ void InterpreterSelectQuery::executeExpression(QueryPipeline & pipeline, const E
 void InterpreterSelectQuery::executeOrder(Pipeline & pipeline, SortingInfoPtr sorting_info)
 {
     auto & query = getSelectQuery();
-    SortDescription order_descr = getSortDescription(query);
+    SortDescription order_descr = getSortDescription(query, context);
     const Settings & settings = context.getSettingsRef();
     UInt64 limit = getLimitForSorting(query, context);
 
@@ -2083,7 +2174,7 @@ void InterpreterSelectQuery::executeOrder(Pipeline & pipeline, SortingInfoPtr so
 void InterpreterSelectQuery::executeOrder(QueryPipeline & pipeline, SortingInfoPtr sorting_info)
 {
     auto & query = getSelectQuery();
-    SortDescription order_descr = getSortDescription(query);
+    SortDescription order_descr = getSortDescription(query, context);
     UInt64 limit = getLimitForSorting(query, context);
 
     const Settings & settings = context.getSettingsRef();
@@ -2164,7 +2255,7 @@ void InterpreterSelectQuery::executeOrder(QueryPipeline & pipeline, SortingInfoP
 void InterpreterSelectQuery::executeMergeSorted(Pipeline & pipeline)
 {
     auto & query = getSelectQuery();
-    SortDescription order_descr = getSortDescription(query);
+    SortDescription order_descr = getSortDescription(query, context);
     UInt64 limit = getLimitForSorting(query, context);
 
     const Settings & settings = context.getSettingsRef();
@@ -2191,7 +2282,7 @@ void InterpreterSelectQuery::executeMergeSorted(Pipeline & pipeline)
 void InterpreterSelectQuery::executeMergeSorted(QueryPipeline & pipeline)
 {
     auto & query = getSelectQuery();
-    SortDescription order_descr = getSortDescription(query);
+    SortDescription order_descr = getSortDescription(query, context);
     UInt64 limit = getLimitForSorting(query, context);
 
     const Settings & settings = context.getSettingsRef();
@@ -2238,7 +2329,7 @@ void InterpreterSelectQuery::executeDistinct(Pipeline & pipeline, bool before_or
         UInt64 limit_for_distinct = 0;
 
         /// If after this stage of DISTINCT ORDER BY is not executed, then you can get no more than limit_length + limit_offset of different rows.
-        if (!query.orderBy() || !before_order)
+        if ((!query.orderBy() || !before_order) && !query.limit_with_ties)
             limit_for_distinct = limit_length + limit_offset;
 
         pipeline.transform([&](auto & stream)
@@ -2307,9 +2398,16 @@ void InterpreterSelectQuery::executePreLimit(Pipeline & pipeline)
     if (query.limitLength())
     {
         auto [limit_length, limit_offset] = getLimitLengthAndOffset(query, context);
+        SortDescription sort_descr;
+        if (query.limit_with_ties)
+        {
+            if (!query.orderBy())
+                throw Exception("LIMIT WITH TIES without ORDER BY", ErrorCodes::LOGICAL_ERROR);
+            sort_descr = getSortDescription(query, context);
+        }
         pipeline.transform([&, limit = limit_length + limit_offset](auto & stream)
         {
-            stream = std::make_shared<LimitBlockInputStream>(stream, limit, 0, false);
+            stream = std::make_shared<LimitBlockInputStream>(stream, limit, 0, false, false, query.limit_with_ties, sort_descr);
         });
     }
 }
@@ -2421,17 +2519,73 @@ void InterpreterSelectQuery::executeLimit(Pipeline & pipeline)
         if (!query.group_by_with_totals && hasWithTotalsInAnySubqueryInFromClause(query))
             always_read_till_end = true;
 
+        SortDescription order_descr;
+        if (query.limit_with_ties)
+        {
+            if (!query.orderBy())
+                throw Exception("LIMIT WITH TIES without ORDER BY", ErrorCodes::LOGICAL_ERROR);
+            order_descr = getSortDescription(query, context);
+        }
+
         UInt64 limit_length;
         UInt64 limit_offset;
         std::tie(limit_length, limit_offset) = getLimitLengthAndOffset(query, context);
 
         pipeline.transform([&](auto & stream)
         {
-            stream = std::make_shared<LimitBlockInputStream>(stream, limit_length, limit_offset, always_read_till_end);
+            stream = std::make_shared<LimitBlockInputStream>(stream, limit_length, limit_offset, always_read_till_end, false, query.limit_with_ties, order_descr);
         });
     }
 }
 
+
+void InterpreterSelectQuery::executeWithFill(Pipeline & pipeline)
+{
+    auto & query = getSelectQuery();
+    if (query.orderBy())
+    {
+        SortDescription order_descr = getSortDescription(query, context);
+        SortDescription fill_descr;
+        for (auto & desc : order_descr)
+        {
+            if (desc.with_fill)
+                fill_descr.push_back(desc);
+        }
+
+        if (fill_descr.empty())
+            return;
+
+        pipeline.transform([&](auto & stream)
+        {
+            stream = std::make_shared<FillingBlockInputStream>(stream, fill_descr);
+        });
+    }
+}
+
+void InterpreterSelectQuery::executeWithFill(QueryPipeline & pipeline)
+{
+    auto & query = getSelectQuery();
+    if (query.orderBy())
+    {
+        SortDescription order_descr = getSortDescription(query, context);
+        SortDescription fill_descr;
+        for (auto & desc : order_descr)
+        {
+            if (desc.with_fill)
+                fill_descr.push_back(desc);
+        }
+
+        if (fill_descr.empty())
+            return;
+
+        pipeline.addSimpleTransform([&](const Block & header)
+        {
+            return std::make_shared<FillingTransform>(header, fill_descr);
+        });
+    }
+}
+
+
 void InterpreterSelectQuery::executeLimit(QueryPipeline & pipeline)
 {
     auto & query = getSelectQuery();
@@ -2459,13 +2613,21 @@ void InterpreterSelectQuery::executeLimit(QueryPipeline & pipeline)
         UInt64 limit_offset;
         std::tie(limit_length, limit_offset) = getLimitLengthAndOffset(query, context);
 
+        SortDescription order_descr;
+        if (query.limit_with_ties)
+        {
+            if (!query.orderBy())
+                throw Exception("LIMIT WITH TIES without ORDER BY", ErrorCodes::LOGICAL_ERROR);
+            order_descr = getSortDescription(query, context);
+        }
+
         pipeline.addSimpleTransform([&](const Block & header, QueryPipeline::StreamType stream_type) -> ProcessorPtr
         {
             if (stream_type != QueryPipeline::StreamType::Main)
                 return nullptr;
 
             return std::make_shared<LimitTransform>(
-                    header, limit_length, limit_offset, always_read_till_end);
+                    header, limit_length, limit_offset, always_read_till_end, query.limit_with_ties, order_descr);
         });
     }
 }

dbms/src/Interpreters/InterpreterSelectQuery.h

@@ -204,6 +204,7 @@ private:
     void executeHaving(Pipeline & pipeline, const ExpressionActionsPtr & expression);
     void executeExpression(Pipeline & pipeline, const ExpressionActionsPtr & expression);
     void executeOrder(Pipeline & pipeline, SortingInfoPtr sorting_info);
+    void executeWithFill(Pipeline & pipeline);
     void executeMergeSorted(Pipeline & pipeline);
     void executePreLimit(Pipeline & pipeline);
     void executeUnion(Pipeline & pipeline, Block header); /// If header is not empty, convert streams structure to it.
@@ -221,6 +222,7 @@ private:
     void executeHaving(QueryPipeline & pipeline, const ExpressionActionsPtr & expression);
     void executeExpression(QueryPipeline & pipeline, const ExpressionActionsPtr & expression);
     void executeOrder(QueryPipeline & pipeline, SortingInfoPtr sorting_info);
+    void executeWithFill(QueryPipeline & pipeline);
     void executeMergeSorted(QueryPipeline & pipeline);
     void executePreLimit(QueryPipeline & pipeline);
     void executeLimitBy(QueryPipeline & pipeline);

dbms/src/Interpreters/QueryNormalizer.cpp

@@ -3,7 +3,6 @@
 #include <Interpreters/QueryNormalizer.h>
 #include <Interpreters/IdentifierSemantic.h>
 #include <Interpreters/Context.h>
-#include <Interpreters/AnalyzedJoin.h>
 #include <Parsers/ASTFunction.h>
 #include <Parsers/ASTIdentifier.h>
 #include <Parsers/ASTSelectQuery.h>

dbms/src/Interpreters/SubqueryForSet.cpp

@@ -1,5 +1,4 @@
 #include <Interpreters/SubqueryForSet.h>
-#include <Interpreters/AnalyzedJoin.h>
 #include <Interpreters/InterpreterSelectWithUnionQuery.h>
 #include <DataStreams/LazyBlockInputStream.h>
 
@@ -7,26 +6,14 @@ namespace DB
 {
 
 void SubqueryForSet::makeSource(std::shared_ptr<InterpreterSelectWithUnionQuery> & interpreter,
-                                const std::unordered_map<String, String> & name_to_origin)
+                                NamesWithAliases && joined_block_aliases_)
 {
+    joined_block_aliases = std::move(joined_block_aliases_);
     source = std::make_shared<LazyBlockInputStream>(interpreter->getSampleBlock(),
                                                     [interpreter]() mutable { return interpreter->execute().in; });
 
-    for (const auto & names : name_to_origin)
-        joined_block_aliases.emplace_back(names.second, names.first);
-
     sample_block = source->getHeader();
-    for (const auto & name_with_alias : joined_block_aliases)
+    renameColumns(sample_block);
-    {
-        if (sample_block.has(name_with_alias.first))
-        {
-            auto pos = sample_block.getPositionByName(name_with_alias.first);
-            auto column = sample_block.getByPosition(pos);
-            sample_block.erase(pos);
-            column.name = name_with_alias.second;
-            sample_block.insert(std::move(column));
-        }
-    }
 }
 
 void SubqueryForSet::renameColumns(Block & block)

dbms/src/Interpreters/SubqueryForSet.h

@@ -31,7 +31,7 @@ struct SubqueryForSet
     StoragePtr table;
 
     void makeSource(std::shared_ptr<InterpreterSelectWithUnionQuery> & interpreter,
-                    const std::unordered_map<String, String> & name_to_origin);
+                    NamesWithAliases && joined_block_aliases_);
 
     Block renamedSampleBlock() const { return sample_block; }
     void renameColumns(Block & block);

dbms/src/Interpreters/SyntaxAnalyzer.cpp

@@ -17,6 +17,7 @@
 #include <Interpreters/OptimizeIfWithConstantConditionVisitor.h>
 #include <Interpreters/RequiredSourceColumnsVisitor.h>
 #include <Interpreters/GetAggregatesVisitor.h>
+#include <Interpreters/AnalyzedJoin.h>
 #include <Interpreters/ExpressionActions.h> /// getSmallestColumn()
 
 #include <Parsers/ASTExpressionList.h>

dbms/src/Interpreters/SyntaxAnalyzer.h

@@ -1,7 +1,7 @@
 #pragma once
 
+#include <Core/NamesAndTypes.h>
 #include <Interpreters/Aliases.h>
-#include <Interpreters/AnalyzedJoin.h>
 #include <Interpreters/SelectQueryOptions.h>
 #include <Storages/IStorage_fwd.h>
 
@@ -11,6 +11,9 @@ namespace DB
 NameSet removeDuplicateColumns(NamesAndTypesList & columns);
 
 class ASTFunction;
+class AnalyzedJoin;
+class Context;
+struct SelectQueryOptions;
 
 struct SyntaxAnalyzerResult
 {

dbms/src/Interpreters/executeQuery.cpp

@@ -21,6 +21,8 @@
 #include <Parsers/parseQuery.h>
 #include <Parsers/queryToString.h>
 
+#include <Storages/StorageInput.h>
+
 #include <Interpreters/Quota.h>
 #include <Interpreters/InterpreterFactory.h>
 #include <Interpreters/ProcessList.h>
@@ -31,6 +33,7 @@
 #include <Common/ProfileEvents.h>
 
 #include <Interpreters/DNSCacheUpdater.h>
+#include <Common/SensitiveDataMasker.h>
 
 #include <Processors/Transforms/LimitsCheckingTransform.h>
 #include <Processors/Transforms/MaterializingTransform.h>
@@ -76,7 +79,7 @@ static String prepareQueryForLogging(const String & query, Context & context)
 
     // wiping sensitive data before cropping query by log_queries_cut_to_length,
     // otherwise something like credit card without last digit can go to log
-    if (auto masker = context.getSensitiveDataMasker())
+    if (auto masker = SensitiveDataMasker::getInstance())
     {
         auto matches = masker->wipeSensitiveData(res);
         if (matches > 0)
@@ -178,7 +181,8 @@ static std::tuple<ASTPtr, BlockIO> executeQueryImpl(
     bool internal,
     QueryProcessingStage::Enum stage,
     bool has_query_tail,
-    bool allow_processors = true)
+    ReadBuffer * istr,
+    bool allow_processors)
 {
     time_t current_time = time(nullptr);
 
@@ -273,6 +277,28 @@ static std::tuple<ASTPtr, BlockIO> executeQueryImpl(
         /// Load external tables if they were provided
         context.initializeExternalTablesIfSet();
 
+        auto * insert_query = ast->as<ASTInsertQuery>();
+        if (insert_query && insert_query->select)
+        {
+            /// Prepare Input storage before executing interpreter if we already got a buffer with data.
+            if (istr)
+            {
+                ASTPtr input_function;
+                insert_query->tryFindInputFunction(input_function);
+                if (input_function)
+                {
+                    StoragePtr storage = context.executeTableFunction(input_function);
+                    auto & input_storage = dynamic_cast<StorageInput &>(*storage);
+                    BlockInputStreamPtr input_stream = std::make_shared<InputStreamFromASTInsertQuery>(ast, istr,
+                        input_storage.getSampleBlock(), context, input_function);
+                    input_storage.setInputStream(input_stream);
+                }
+            }
+        }
+        else
+            /// reset Input callbacks if query is not INSERT SELECT
+            context.resetInputCallbacks();
+
         auto interpreter = InterpreterFactory::get(ast, context, stage);
         bool use_processors = settings.experimental_use_processors && allow_processors && interpreter->canExecuteWithProcessors();
 
@@ -527,7 +553,8 @@ BlockIO executeQuery(
     bool allow_processors)
 {
     BlockIO streams;
-    std::tie(std::ignore, streams) = executeQueryImpl(query.data(), query.data() + query.size(), context, internal, stage, !may_have_embedded_data, allow_processors);
+    std::tie(std::ignore, streams) = executeQueryImpl(query.data(), query.data() + query.size(), context,
+        internal, stage, !may_have_embedded_data, nullptr, allow_processors);
     return streams;
 }
 
@@ -578,7 +605,7 @@ void executeQuery(
     ASTPtr ast;
     BlockIO streams;
 
-    std::tie(ast, streams) = executeQueryImpl(begin, end, context, false, QueryProcessingStage::Complete, may_have_tail);
+    std::tie(ast, streams) = executeQueryImpl(begin, end, context, false, QueryProcessingStage::Complete, may_have_tail, &istr, true);
 
     auto & pipeline = streams.pipeline;
 
@@ -586,7 +613,7 @@ void executeQuery(
     {
         if (streams.out)
         {
-            InputStreamFromASTInsertQuery in(ast, &istr, streams.out->getHeader(), context);
+            InputStreamFromASTInsertQuery in(ast, &istr, streams.out->getHeader(), context, nullptr);
             copyData(in, *streams.out);
         }
 

dbms/src/Interpreters/interpretSubquery.cpp

@@ -18,6 +18,19 @@ namespace DB
 std::shared_ptr<InterpreterSelectWithUnionQuery> interpretSubquery(
     const ASTPtr & table_expression, const Context & context, size_t subquery_depth, const Names & required_source_columns)
 {
+    if (auto * expr = table_expression->as<ASTTableExpression>())
+    {
+        ASTPtr table;
+        if (expr->subquery)
+            table = expr->subquery;
+        else if (expr->table_function)
+            table = expr->table_function;
+        else if (expr->database_and_table_name)
+            table = expr->database_and_table_name;
+
+        return interpretSubquery(table, context, subquery_depth, required_source_columns);
+    }
+
     /// Subquery or table name. The name of the table is similar to the subquery `SELECT * FROM t`.
     const auto * subquery = table_expression->as<ASTSubquery>();
     const auto * function = table_expression->as<ASTFunction>();

dbms/src/Parsers/ASTInsertQuery.cpp

@@ -1,10 +1,17 @@
 #include <iomanip>
 #include <Parsers/ASTInsertQuery.h>
+#include <Parsers/ASTFunction.h>
 
 
 namespace DB
 {
 
+namespace ErrorCodes
+{
+    extern const int INVALID_USAGE_OF_INPUT;
+}
+
+
 void ASTInsertQuery::formatImpl(const FormatSettings & settings, FormatState & state, FormatStateStacked frame) const
 {
     frame.need_parens = false;
@@ -50,4 +57,29 @@ void ASTInsertQuery::formatImpl(const FormatSettings & settings, FormatState & s
     }
 }
 
+
+void tryFindInputFunctionImpl(const ASTPtr & ast, ASTPtr & input_function)
+{
+    if (!ast)
+        return;
+    for (const auto & child : ast->children)
+        tryFindInputFunctionImpl(child, input_function);
+
+    if (const auto * table_function_ast = ast->as<ASTFunction>())
+    {
+        if (table_function_ast->name == "input")
+        {
+            if (input_function)
+                throw Exception("You can use 'input()' function only once per request.", ErrorCodes::INVALID_USAGE_OF_INPUT);
+            input_function = ast;
+        }
+    }
+}
+
+
+void ASTInsertQuery::tryFindInputFunction(ASTPtr & input_function) const
+{
+    tryFindInputFunctionImpl(select, input_function);
+}
+
 }

dbms/src/Parsers/ASTInsertQuery.h

@@ -30,6 +30,9 @@ public:
     /// Query has additional data, which will be sent later
     bool has_tail = false;
 
+    /// Try to find table function input() in SELECT part
+    void tryFindInputFunction(ASTPtr & input_function) const;
+
     /** Get the text that identifies this element. */
     String getID(char delim) const override { return "InsertQuery" + (delim + database) + delim + table; }
 

dbms/src/Parsers/ASTOrderByElement.cpp

@@ -25,6 +25,26 @@ void ASTOrderByElement::formatImpl(const FormatSettings & settings, FormatState
         settings.ostr << (settings.hilite ? hilite_keyword : "") << " COLLATE " << (settings.hilite ? hilite_none : "");
         collation->formatImpl(settings, state, frame);
     }
+
+    if (with_fill)
+    {
+        settings.ostr << (settings.hilite ? hilite_keyword : "") << " WITH FILL " << (settings.hilite ? hilite_none : "");
+        if (fill_from)
+        {
+            settings.ostr << (settings.hilite ? hilite_keyword : "") << " FROM " << (settings.hilite ? hilite_none : "");
+            fill_from->formatImpl(settings, state, frame);
+        }
+        if (fill_to)
+        {
+            settings.ostr << (settings.hilite ? hilite_keyword : "") << " TO " << (settings.hilite ? hilite_none : "");
+            fill_to->formatImpl(settings, state, frame);
+        }
+        if (fill_step)
+        {
+            settings.ostr << (settings.hilite ? hilite_keyword : "") << " STEP " << (settings.hilite ? hilite_none : "");
+            fill_step->formatImpl(settings, state, frame);
+        }
+    }
 }
 
 }

dbms/src/Parsers/ASTOrderByElement.h

@@ -18,12 +18,22 @@ public:
     /** Collation for locale-specific string comparison. If empty, then sorting done by bytes. */
     ASTPtr collation;
 
+    bool with_fill;
+    ASTPtr fill_from;
+    ASTPtr fill_to;
+    ASTPtr fill_step;
+
     ASTOrderByElement(
-        const int direction_, const int nulls_direction_, const bool nulls_direction_was_explicitly_specified_, ASTPtr & collation_)
+        const int direction_, const int nulls_direction_, const bool nulls_direction_was_explicitly_specified_,
+        ASTPtr & collation_, const bool with_fill_, ASTPtr & fill_from_, ASTPtr & fill_to_, ASTPtr & fill_step_)
         : direction(direction_)
         , nulls_direction(nulls_direction_)
         , nulls_direction_was_explicitly_specified(nulls_direction_was_explicitly_specified_)
         , collation(collation_)
+        , with_fill(with_fill_)
+        , fill_from(fill_from_)
+        , fill_to(fill_to_)
+        , fill_step(fill_step_)
     {
     }
 

dbms/src/Parsers/ASTSelectQuery.cpp

@@ -148,6 +148,8 @@ void ASTSelectQuery::formatImpl(const FormatSettings & s, FormatState & state, F
             s.ostr << ", ";
         }
         limitLength()->formatImpl(s, state, frame);
+        if (limit_with_ties)
+            s.ostr << (s.hilite ? hilite_keyword : "") << s.nl_or_ws << indent_str << " WITH TIES" << (s.hilite ? hilite_none : "");
     }
 
     if (settings())

dbms/src/Parsers/ASTSelectQuery.h

@@ -42,6 +42,7 @@ public:
     bool group_by_with_totals = false;
     bool group_by_with_rollup = false;
     bool group_by_with_cube = false;
+    bool limit_with_ties = false;
 
     ASTPtr & refSelect()    { return getExpression(Expression::SELECT); }
     ASTPtr & refTables()    { return getExpression(Expression::TABLES); }

dbms/src/Parsers/ExpressionElementParsers.cpp

@@ -1360,7 +1360,12 @@ bool ParserOrderByElement::parseImpl(Pos & pos, ASTPtr & node, Expected & expect
     ParserKeyword first("FIRST");
     ParserKeyword last("LAST");
     ParserKeyword collate("COLLATE");
+    ParserKeyword with_fill("WITH FILL");
+    ParserKeyword from("FROM");
+    ParserKeyword to("TO");
+    ParserKeyword step("STEP");
     ParserStringLiteral collate_locale_parser;
+    ParserExpressionWithOptionalAlias exp_parser(false);
 
     ASTPtr expr_elem;
     if (!elem_p.parse(pos, expr_elem, expected))
@@ -1395,7 +1400,27 @@ bool ParserOrderByElement::parseImpl(Pos & pos, ASTPtr & node, Expected & expect
             return false;
     }
 
-    node = std::make_shared<ASTOrderByElement>(direction, nulls_direction, nulls_direction_was_explicitly_specified, locale_node);
+    /// WITH FILL [FROM x] [TO y] [STEP z]
+    bool has_with_fill = false;
+    ASTPtr fill_from;
+    ASTPtr fill_to;
+    ASTPtr fill_step;
+    if (with_fill.ignore(pos))
+    {
+        has_with_fill = true;
+        if (from.ignore(pos) && !exp_parser.parse(pos, fill_from, expected))
+            return false;
+
+        if (to.ignore(pos) && !exp_parser.parse(pos, fill_to, expected))
+            return false;
+
+        if (step.ignore(pos) && !exp_parser.parse(pos, fill_step, expected))
+            return false;
+    }
+
+    node = std::make_shared<ASTOrderByElement>(
+            direction, nulls_direction, nulls_direction_was_explicitly_specified, locale_node,
+            has_with_fill, fill_from, fill_to, fill_step);
     node->children.push_back(expr_elem);
     if (locale_node)
         node->children.push_back(locale_node);

dbms/src/Parsers/ExpressionElementParsers.h

@@ -289,6 +289,7 @@ protected:
 /** Element of ORDER BY expression - same as expression element, but in addition, ASC[ENDING] | DESC[ENDING] could be specified
   *  and optionally, NULLS LAST|FIRST
   *  and optionally, COLLATE 'locale'.
+  *  and optionally, WITH FILL [FROM x] [TO y] [STEP z]
   */
 class ParserOrderByElement : public IParserBase
 {

dbms/src/Parsers/ParserInsertQuery.cpp

@@ -8,7 +8,7 @@
 #include <Parsers/ParserSelectWithUnionQuery.h>
 #include <Parsers/ParserInsertQuery.h>
 #include <Parsers/ParserSetQuery.h>
-#include <Parsers/ASTFunction.h>
+#include <Common/typeid_cast.h>
 
 
 namespace DB
@@ -97,6 +97,10 @@ bool ParserInsertQuery::parseImpl(Pos & pos, ASTPtr & node, Expected & expected)
         pos = before_select;
         ParserSelectWithUnionQuery select_p;
         select_p.parse(pos, select, expected);
+
+        /// FORMAT section is expected if we have input() in SELECT part
+        if (s_format.ignore(pos, expected) && !name_p.parse(pos, format, expected))
+            return false;
     }
     else
     {

dbms/src/Parsers/ParserSelectQuery.cpp

@@ -17,6 +17,7 @@ namespace ErrorCodes
 {
     extern const int SYNTAX_ERROR;
     extern const int TOP_AND_LIMIT_TOGETHER;
+    extern const int WITH_TIES_WITHOUT_ORDER_BY;
 }
 
 
@@ -41,6 +42,7 @@ bool ParserSelectQuery::parseImpl(Pos & pos, ASTPtr & node, Expected & expected)
     ParserKeyword s_rollup("ROLLUP");
     ParserKeyword s_cube("CUBE");
     ParserKeyword s_top("TOP");
+    ParserKeyword s_with_ties("WITH TIES");
     ParserKeyword s_offset("OFFSET");
 
     ParserNotEmptyExpressionList exp_list(false);
@@ -76,7 +78,7 @@ bool ParserSelectQuery::parseImpl(Pos & pos, ASTPtr & node, Expected & expected)
         }
     }
 
-    /// SELECT [DISTINCT] [TOP N] expr list
+    /// SELECT [DISTINCT] [TOP N [WITH TIES]] expr list
     {
         if (!s_select.ignore(pos, expected))
             return false;
@@ -100,6 +102,9 @@ bool ParserSelectQuery::parseImpl(Pos & pos, ASTPtr & node, Expected & expected)
                 if (!num.parse(pos, limit_length, expected))
                     return false;
             }
+
+            if (s_with_ties.ignore(pos, expected))
+                select_query->limit_with_ties = true;
         }
 
         if (!exp_list_for_select_clause.parse(pos, select_expression_list, expected))
@@ -197,12 +202,18 @@ bool ParserSelectQuery::parseImpl(Pos & pos, ASTPtr & node, Expected & expected)
             limit_offset = limit_length;
             if (!exp_elem.parse(pos, limit_length, expected))
                 return false;
+
+            if (s_with_ties.ignore(pos, expected))
+                select_query->limit_with_ties = true;
         }
         else if (s_offset.ignore(pos, expected))
         {
             if (!exp_elem.parse(pos, limit_offset, expected))
                 return false;
         }
+        else if (s_with_ties.ignore(pos, expected))
+            select_query->limit_with_ties = true;
+
         if (s_by.ignore(pos, expected))
         {
             limit_by_length = limit_length;
@@ -215,7 +226,7 @@ bool ParserSelectQuery::parseImpl(Pos & pos, ASTPtr & node, Expected & expected)
         }
     }
 
-    /// LIMIT length | LIMIT offset, length
+    /// LIMIT length [WITH TIES] | LIMIT offset, length [WITH TIES]
     if (s_limit.ignore(pos, expected))
     {
         if (!limit_by_length|| limit_length)
@@ -237,8 +248,15 @@ bool ParserSelectQuery::parseImpl(Pos & pos, ASTPtr & node, Expected & expected)
             if (!exp_elem.parse(pos, limit_offset, expected))
                 return false;
         }
+
+        if (s_with_ties.ignore(pos, expected))
+            select_query->limit_with_ties = true;
     }
 
+    /// WITH TIES was used without ORDER BY
+    if (!order_expression_list && select_query->limit_with_ties)
+        throw Exception("Can not use WITH TIES without ORDER BY", ErrorCodes::WITH_TIES_WITHOUT_ORDER_BY);
+
     /// SETTINGS key1 = value1, key2 = value2, ...
     if (s_settings.ignore(pos, expected))
     {

dbms/src/Processors/LimitTransform.cpp

@@ -6,19 +6,26 @@ namespace DB
 
 LimitTransform::LimitTransform(
     const Block & header_, size_t limit_, size_t offset_,
-    bool always_read_till_end_)
+    bool always_read_till_end_, bool with_ties_,
+    const SortDescription & description_)
     : IProcessor({header_}, {header_})
     , input(inputs.front()), output(outputs.front())
     , limit(limit_), offset(offset_)
     , always_read_till_end(always_read_till_end_)
+    , with_ties(with_ties_), description(description_)
 {
+    for (const auto & desc : description)
+    {
+        if (!desc.column_name.empty())
+            sort_column_positions.push_back(header_.getPositionByName(desc.column_name));
+        else
+            sort_column_positions.push_back(desc.column_number);
+    }
 }
 
 
 LimitTransform::Status LimitTransform::prepare()
 {
-
-
     /// Check can output.
     bool output_finished = false;
     if (output.isFinished())
@@ -46,7 +53,7 @@ LimitTransform::Status LimitTransform::prepare()
     }
 
     /// Check if we are done with pushing.
-    bool pushing_is_finished = rows_read >= offset + limit;
+    bool pushing_is_finished = (rows_read >= offset + limit) && ties_row_ref.empty();
     if (pushing_is_finished)
     {
         if (!always_read_till_end)
@@ -116,6 +123,13 @@ LimitTransform::Status LimitTransform::prepare()
         if (output.hasData())
             return Status::PortFull;
 
+        if (with_ties && rows_read == offset + limit)
+        {
+            SharedChunkPtr shared_chunk = new detail::SharedChunk(current_chunk.clone());
+            shared_chunk->sort_columns = extractSortColumns(shared_chunk->getColumns());
+            ties_row_ref.set(shared_chunk, &shared_chunk->sort_columns, shared_chunk->getNumRows() - 1);
+        }
+
         output.push(std::move(current_chunk));
         has_block = false;
 
@@ -132,8 +146,39 @@ LimitTransform::Status LimitTransform::prepare()
 
 void LimitTransform::work()
 {
-    size_t num_rows = current_chunk.getNumRows();
+    SharedChunkPtr shared_chunk = new detail::SharedChunk(std::move(current_chunk));
-    size_t num_columns = current_chunk.getNumColumns();
+    shared_chunk->sort_columns = extractSortColumns(shared_chunk->getColumns());
+
+    size_t num_rows = shared_chunk->getNumRows();
+    size_t num_columns = shared_chunk->getNumColumns();
+
+    if (!ties_row_ref.empty() && rows_read >= offset + limit)
+    {
+        UInt64 len;
+        for (len = 0; len < num_rows; ++len)
+        {
+            SharedChunkRowRef current_row;
+            current_row.set(shared_chunk, &shared_chunk->sort_columns, len);
+
+            if (current_row != ties_row_ref)
+            {
+                ties_row_ref.reset();
+                break;
+            }
+        }
+
+        auto columns = shared_chunk->detachColumns();
+
+        if (len < num_rows)
+        {
+            for (size_t i = 0; i < num_columns; ++i)
+                columns[i] = columns[i]->cut(0, len);
+        }
+
+        current_chunk.setColumns(std::move(columns), len);
+        block_processed = true;
+        return;
+    }
 
     /// return a piece of the block
     size_t start = std::max(
@@ -145,7 +190,33 @@ void LimitTransform::work()
         static_cast<Int64>(rows_read) - static_cast<Int64>(offset),
         static_cast<Int64>(limit) + static_cast<Int64>(offset) - static_cast<Int64>(rows_read) + static_cast<Int64>(num_rows)));
 
-    auto columns = current_chunk.detachColumns();
+    /// check if other rows in current block equals to last one in limit
+    if (with_ties)
+    {
+        ties_row_ref.set(shared_chunk, &shared_chunk->sort_columns, start + length - 1);
+        SharedChunkRowRef current_row;
+
+        for (size_t i = ties_row_ref.row_num + 1; i < num_rows; ++i)
+        {
+            current_row.set(shared_chunk, &shared_chunk->sort_columns, i);
+            if (current_row == ties_row_ref)
+                ++length;
+            else
+            {
+                ties_row_ref.reset();
+                break;
+            }
+        }
+    }
+
+    if (length == num_rows)
+    {
+        current_chunk = std::move(*shared_chunk);
+        block_processed = true;
+        return;
+    }
+
+    auto columns = shared_chunk->detachColumns();
 
     for (size_t i = 0; i < num_columns; ++i)
         columns[i] = columns[i]->cut(start, length);
@@ -155,5 +226,15 @@ void LimitTransform::work()
     block_processed = true;
 }
 
+ColumnRawPtrs LimitTransform::extractSortColumns(const Columns & columns)
+{
+    ColumnRawPtrs res;
+    res.reserve(description.size());
+    for (size_t pos : sort_column_positions)
+        res.push_back(columns[pos].get());
+
+    return res;
+}
+
 }
 

dbms/src/Processors/LimitTransform.h

@@ -1,7 +1,8 @@
 #pragma once
 
 #include <Processors/IProcessor.h>
-
+#include <Processors/SharedChunk.h>
+#include <Core/SortDescription.h>
 
 namespace DB
 {
@@ -23,10 +24,18 @@ private:
 
     UInt64 rows_before_limit_at_least = 0;
 
+    bool with_ties;
+    const SortDescription description;
+    SharedChunkRowRef ties_row_ref;
+
+    std::vector<size_t> sort_column_positions;
+    ColumnRawPtrs extractSortColumns(const Columns & columns);
+
 public:
     LimitTransform(
         const Block & header_, size_t limit_, size_t offset_,
-        bool always_read_till_end_ = false);
+        bool always_read_till_end_ = false, bool with_ties_ = false,
+        const SortDescription & description_ = {});
 
     String getName() const override { return "Limit"; }
 

dbms/src/Processors/SharedChunk.h

@@ -0,0 +1,91 @@
+#pragma once
+
+#include <algorithm>
+#include <Processors/Chunk.h>
+#include <Columns/IColumn.h>
+#include <boost/smart_ptr/intrusive_ptr.hpp>
+
+
+namespace DB
+{
+
+/// Allows you refer to the row in the block and hold the block ownership,
+///  and thus avoid creating a temporary row object.
+/// Do not use std::shared_ptr, since there is no need for a place for `weak_count` and `deleter`;
+///  does not use Poco::SharedPtr, since you need to allocate a block and `refcount` in one piece;
+///  does not use Poco::AutoPtr, since it does not have a `move` constructor and there are extra checks for nullptr;
+/// The reference counter is not atomic, since it is used from one thread.
+namespace detail
+{
+struct SharedChunk : Chunk
+{
+    int refcount = 0;
+
+    ColumnRawPtrs all_columns;
+    ColumnRawPtrs sort_columns;
+
+    SharedChunk(Chunk && chunk) : Chunk(std::move(chunk)) {}
+};
+
+}
+
+inline void intrusive_ptr_add_ref(detail::SharedChunk * ptr)
+{
+    ++ptr->refcount;
+}
+
+inline void intrusive_ptr_release(detail::SharedChunk * ptr)
+{
+    if (0 == --ptr->refcount)
+        delete ptr;
+}
+
+using SharedChunkPtr = boost::intrusive_ptr<detail::SharedChunk>;
+
+
+struct SharedChunkRowRef
+{
+    ColumnRawPtrs * columns = nullptr;
+    size_t row_num;
+    SharedChunkPtr shared_block;
+
+    void swap(SharedChunkRowRef & other)
+    {
+        std::swap(columns, other.columns);
+        std::swap(row_num, other.row_num);
+        std::swap(shared_block, other.shared_block);
+    }
+
+    /// The number and types of columns must match.
+    bool operator==(const SharedChunkRowRef & other) const
+    {
+        size_t size = columns->size();
+        for (size_t i = 0; i < size; ++i)
+            if (0 != (*columns)[i]->compareAt(row_num, other.row_num, *(*other.columns)[i], 1))
+                return false;
+        return true;
+    }
+
+    bool operator!=(const SharedChunkRowRef & other) const
+    {
+        return !(*this == other);
+    }
+
+    void reset()
+    {
+        SharedChunkRowRef empty;
+        swap(empty);
+    }
+
+    bool empty() const { return columns == nullptr; }
+    size_t size() const { return empty() ? 0 : columns->size(); }
+
+    void set(SharedChunkPtr & shared_block_, ColumnRawPtrs * columns_, size_t row_num_)
+    {
+        shared_block = shared_block_;
+        columns = columns_;
+        row_num = row_num_;
+    }
+};
+
+}

dbms/src/Processors/Transforms/FillingTransform.cpp

@@ -0,0 +1,201 @@
+#include <Processors/Transforms/FillingTransform.h>
+#include <Interpreters/convertFieldToType.h>
+#include <DataTypes/DataTypesNumber.h>
+
+namespace DB
+{
+
+namespace ErrorCodes
+{
+    extern const int INVALID_WITH_FILL_EXPRESSION;
+}
+
+
+FillingTransform::FillingTransform(
+        const Block & header_, const SortDescription & sort_description_)
+        : ISimpleTransform(header_, header_, true)
+        , sort_description(sort_description_)
+        , filling_row(sort_description_)
+        , next_row(sort_description_)
+{
+    std::vector<bool> is_fill_column(header_.columns());
+    for (const auto & elem : sort_description)
+        is_fill_column[header_.getPositionByName(elem.column_name)] = true;
+
+    auto try_convert_fields = [](FillColumnDescription & descr, const DataTypePtr & type)
+    {
+        auto max_type = Field::Types::Null;
+        WhichDataType which(type);
+        DataTypePtr to_type;
+        if (isInteger(type) || which.isDateOrDateTime())
+        {
+            max_type = Field::Types::Int64;
+            to_type = std::make_shared<DataTypeInt64>();
+        }
+        else if (which.isFloat())
+        {
+            max_type = Field::Types::Float64;
+            to_type = std::make_shared<DataTypeFloat64>();
+        }
+
+        if (descr.fill_from.getType() > max_type || descr.fill_to.getType() > max_type
+            || descr.fill_step.getType() > max_type)
+            return false;
+
+        descr.fill_from = convertFieldToType(descr.fill_from, *to_type);
+        descr.fill_to = convertFieldToType(descr.fill_to, *to_type);
+        descr.fill_step = convertFieldToType(descr.fill_step, *to_type);
+
+        return true;
+    };
+
+    for (size_t i = 0; i < header_.columns(); ++i)
+    {
+        if (is_fill_column[i])
+        {
+            size_t pos = fill_column_positions.size();
+            auto & descr = filling_row.getFillDescription(pos);
+            auto type = header_.getByPosition(i).type;
+            if (!try_convert_fields(descr, type))
+                throw Exception("Incompatible types of WITH FILL expression values with column type "
+                    + type->getName(), ErrorCodes::INVALID_WITH_FILL_EXPRESSION);
+
+            if (type->isValueRepresentedByUnsignedInteger() &&
+                ((!descr.fill_from.isNull() && less(descr.fill_from, Field{0}, 1)) ||
+                    (!descr.fill_to.isNull() && less(descr.fill_to, Field{0}, 1))))
+            {
+                throw Exception("WITH FILL bound values cannot be negative for unsigned type "
+                    + type->getName(), ErrorCodes::INVALID_WITH_FILL_EXPRESSION);
+            }
+
+            fill_column_positions.push_back(i);
+        }
+        else
+            other_column_positions.push_back(i);
+    }
+}
+
+IProcessor::Status FillingTransform::prepare()
+{
+    if (input.isFinished() && !output.isFinished() && !has_input && !generate_suffix)
+    {
+        should_insert_first = next_row < filling_row;
+
+        for (size_t i = 0; i < filling_row.size(); ++i)
+            next_row[i] = filling_row.getFillDescription(i).fill_to;
+
+        if (filling_row < next_row)
+        {
+            generate_suffix = true;
+            return Status::Ready;
+        }
+    }
+
+    return ISimpleTransform::prepare();
+}
+
+
+void FillingTransform::transform(Chunk & chunk)
+{
+    Columns old_fill_columns;
+    Columns old_other_columns;
+    MutableColumns res_fill_columns;
+    MutableColumns res_other_columns;
+
+    auto init_columns_by_positions = [](const Columns & old_columns, Columns & new_columns,
+        MutableColumns & new_mutable_columns, const Positions & positions)
+    {
+        for (size_t pos : positions)
+        {
+            new_columns.push_back(old_columns[pos]);
+            new_mutable_columns.push_back(old_columns[pos]->cloneEmpty()->assumeMutable());
+        }
+    };
+
+    if (generate_suffix)
+    {
+        const auto & empty_columns = inputs.front().getHeader().getColumns();
+        init_columns_by_positions(empty_columns, old_fill_columns, res_fill_columns, fill_column_positions);
+        init_columns_by_positions(empty_columns, old_other_columns, res_other_columns, other_column_positions);
+
+        if (should_insert_first && filling_row < next_row)
+            insertFromFillingRow(res_fill_columns, res_other_columns, filling_row);
+
+        while (filling_row.next(next_row))
+            insertFromFillingRow(res_fill_columns, res_other_columns, filling_row);
+
+        setResultColumns(chunk, res_fill_columns, res_other_columns);
+        return;
+    }
+
+    size_t num_rows = chunk.getNumRows();
+    auto old_columns = chunk.detachColumns();
+
+    init_columns_by_positions(old_columns, old_fill_columns, res_fill_columns, fill_column_positions);
+    init_columns_by_positions(old_columns, old_other_columns, res_other_columns, other_column_positions);
+
+    if (first)
+    {
+        for (size_t i = 0; i < filling_row.size(); ++i)
+        {
+            auto current_value = (*old_fill_columns[i])[0];
+            const auto & fill_from = filling_row.getFillDescription(i).fill_from;
+
+            if (!fill_from.isNull() && !equals(current_value, fill_from))
+            {
+                filling_row.initFromDefaults(i);
+                if (less(fill_from, current_value, filling_row.getDirection(i)))
+                    insertFromFillingRow(res_fill_columns, res_other_columns, filling_row);
+                break;
+            }
+            filling_row[i] = current_value;
+        }
+        first = false;
+    }
+
+    for (size_t row_ind = 0; row_ind < num_rows; ++row_ind)
+    {
+        should_insert_first = next_row < filling_row;
+
+        for (size_t i = 0; i < filling_row.size(); ++i)
+        {
+            auto current_value = (*old_fill_columns[i])[row_ind];
+            const auto & fill_to = filling_row.getFillDescription(i).fill_to;
+
+            if (fill_to.isNull() || less(current_value, fill_to, filling_row.getDirection(i)))
+                next_row[i] = current_value;
+            else
+                next_row[i] = fill_to;
+        }
+
+        /// A case, when at previous step row was initialized from defaults 'fill_from' values
+        ///  and probably we need to insert it to block.
+        if (should_insert_first && filling_row < next_row)
+            insertFromFillingRow(res_fill_columns, res_other_columns, filling_row);
+
+        /// Insert generated filling row to block, while it is less than current row in block.
+        while (filling_row.next(next_row))
+            insertFromFillingRow(res_fill_columns, res_other_columns, filling_row);
+
+        copyRowFromColumns(res_fill_columns, old_fill_columns, row_ind);
+        copyRowFromColumns(res_other_columns, old_other_columns, row_ind);
+    }
+
+    setResultColumns(chunk, res_fill_columns, res_other_columns);
+}
+
+void FillingTransform::setResultColumns(Chunk & chunk, MutableColumns & fill_columns, MutableColumns & other_columns) const
+{
+    MutableColumns result_columns(fill_columns.size() + other_columns.size());
+    /// fill_columns always non-empty.
+    size_t num_rows = fill_columns[0]->size();
+
+    for (size_t i = 0; i < fill_columns.size(); ++i)
+        result_columns[fill_column_positions[i]] = std::move(fill_columns[i]);
+    for (size_t i = 0; i < other_columns.size(); ++i)
+        result_columns[other_column_positions[i]] = std::move(other_columns[i]);
+
+    chunk.setColumns(std::move(result_columns), num_rows);
+}
+
+}

dbms/src/Processors/Transforms/FillingTransform.h

@@ -0,0 +1,42 @@
+#pragma once
+#include <Processors/ISimpleTransform.h>
+#include <Core/SortDescription.h>
+#include <Interpreters/FillingRow.h>
+
+namespace DB
+{
+
+/** Implements modifier WITH FILL of ORDER BY clause.
+ *  It fills gaps in data stream by rows with missing values in columns with set WITH FILL and deafult values in other columns.
+ *  Optionally FROM, TO and STEP values can be specified.
+ */
+class FillingTransform : public ISimpleTransform
+{
+public:
+    FillingTransform(const Block & header_, const SortDescription & fill_description_);
+
+    String getName() const override { return "FillingTransform"; }
+
+    Status prepare() override;
+
+protected:
+    void transform(Chunk & Chunk) override;
+
+private:
+    void setResultColumns(Chunk & chunk, MutableColumns & fill_columns, MutableColumns & other_columns) const;
+
+    const SortDescription sort_description; /// Contains only rows with WITH FILL.
+    FillingRow filling_row; /// Current row, which is used to fill gaps.
+    FillingRow next_row; /// Row to which we need to generate filling rows.
+
+    using Positions = std::vector<size_t>;
+    Positions fill_column_positions;
+    Positions other_column_positions;
+    bool first = true;
+    bool generate_suffix = false;
+
+    /// Determines should we insert filling row before start generating next rows.
+    bool should_insert_first = false;
+};
+
+}

dbms/src/Processors/Transforms/MergingSortedTransform.h

@@ -2,46 +2,13 @@
 #include <Processors/IProcessor.h>
 #include <Core/SortDescription.h>
 #include <Core/SortCursor.h>
+#include <Processors/SharedChunk.h>
 
 #include <queue>
 
 namespace DB
 {
 
-/// Allows you refer to the row in the block and hold the block ownership,
-///  and thus avoid creating a temporary row object.
-/// Do not use std::shared_ptr, since there is no need for a place for `weak_count` and `deleter`;
-///  does not use Poco::SharedPtr, since you need to allocate a block and `refcount` in one piece;
-///  does not use Poco::AutoPtr, since it does not have a `move` constructor and there are extra checks for nullptr;
-/// The reference counter is not atomic, since it is used from one thread.
-namespace detail
-{
-struct SharedChunk : Chunk
-{
-    int refcount = 0;
-
-    ColumnRawPtrs all_columns;
-    ColumnRawPtrs sort_columns;
-
-    SharedChunk(Chunk && chunk) : Chunk(std::move(chunk)) {}
-};
-
-}
-
-using SharedChunkPtr = boost::intrusive_ptr<detail::SharedChunk>;
-
-
-inline void intrusive_ptr_add_ref(detail::SharedChunk * ptr)
-{
-    ++ptr->refcount;
-}
-
-inline void intrusive_ptr_release(detail::SharedChunk * ptr)
-{
-    if (0 == --ptr->refcount)
-        delete ptr;
-}
-
 class MergingSortedTransform : public IProcessor
 {
 public:

dbms/src/Storages/Distributed/DirectoryMonitor.cpp

@@ -189,7 +189,9 @@ ConnectionPoolPtr StorageDistributedDirectoryMonitor::createPool(const std::stri
 
     auto pools = createPoolsForAddresses(name, pool_factory);
 
-    return pools.size() == 1 ? pools.front() : std::make_shared<ConnectionPoolWithFailover>(pools, LoadBalancing::RANDOM);
+    const auto settings = storage.global_context.getSettings();
+    return pools.size() == 1 ? pools.front() : std::make_shared<ConnectionPoolWithFailover>(pools, LoadBalancing::RANDOM,
+        settings.distributed_replica_error_half_life.totalSeconds(), settings.distributed_replica_error_cap);
 }
 
 

dbms/src/Storages/Kafka/KafkaBlockInputStream.cpp

@@ -28,10 +28,7 @@ KafkaBlockInputStream::~KafkaBlockInputStream()
         return;
 
     if (broken)
-    {
+        buffer->unsubscribe();
-        buffer->subBufferAs<ReadBufferFromKafkaConsumer>()->unsubscribe();
-        buffer->reset();
-    }
 
     storage.pushReadBuffer(buffer);
 }
@@ -50,23 +47,22 @@ void KafkaBlockInputStream::readPrefixImpl()
     if (!buffer)
         return;
 
-    buffer->subBufferAs<ReadBufferFromKafkaConsumer>()->subscribe(storage.getTopics());
+    buffer->subscribe(storage.getTopics());
 
     const auto & limits_ = getLimits();
-    const size_t poll_timeout = buffer->subBufferAs<ReadBufferFromKafkaConsumer>()->pollTimeout();
+    const size_t poll_timeout = buffer->pollTimeout();
     size_t rows_portion_size = poll_timeout ? std::min<size_t>(max_block_size, limits_.max_execution_time.totalMilliseconds() / poll_timeout) : max_block_size;
     rows_portion_size = std::max(rows_portion_size, 1ul);
 
     auto non_virtual_header = storage.getSampleBlockNonMaterialized(); /// FIXME: add materialized columns support
     auto read_callback = [this]
     {
-        const auto * sub_buffer = buffer->subBufferAs<ReadBufferFromKafkaConsumer>();
+        virtual_columns[0]->insert(buffer->currentTopic());     // "topic"
-        virtual_columns[0]->insert(sub_buffer->currentTopic());     // "topic"
+        virtual_columns[1]->insert(buffer->currentKey());       // "key"
-        virtual_columns[1]->insert(sub_buffer->currentKey());       // "key"
+        virtual_columns[2]->insert(buffer->currentOffset());    // "offset"
-        virtual_columns[2]->insert(sub_buffer->currentOffset());    // "offset"
+        virtual_columns[3]->insert(buffer->currentPartition()); // "partition"
-        virtual_columns[3]->insert(sub_buffer->currentPartition()); // "partition"
 
-        auto timestamp = sub_buffer->currentTimestamp();
+        auto timestamp = buffer->currentTimestamp();
         if (timestamp)
             virtual_columns[4]->insert(std::chrono::duration_cast<std::chrono::seconds>(timestamp->get_timestamp()).count()); // "timestamp"
     };
@@ -106,7 +102,7 @@ void KafkaBlockInputStream::readSuffixImpl()
     if (!buffer)
         return;
 
-    buffer->subBufferAs<ReadBufferFromKafkaConsumer>()->commit();
+    buffer->commit();
 
     broken = false;
 }