Merge branch 'master' into startup-sanity-checks

.clang-tidy

@@ -210,3 +210,6 @@ CheckOptions:
     value: false
   - key: performance-move-const-arg.CheckTriviallyCopyableMove
     value: false
+    # Workaround clang-tidy bug: https://github.com/llvm/llvm-project/issues/46097
+  - key: readability-identifier-naming.TypeTemplateParameterIgnoredRegexp
+    value: expr-type

.github/workflows/master.yml

@@ -360,6 +360,52 @@ jobs:
           docker kill "$(docker ps -q)" ||:
           docker rm -f "$(docker ps -a -q)" ||:
           sudo rm -fr "$TEMP_PATH" "$CACHES_PATH"
+  BuilderBinGCC:
+    needs: [DockerHubPush]
+    if: ${{ !contains(github.event.pull_request.labels.*.name, 'pr-documentation') && !contains(github.event.pull_request.labels.*.name, 'pr-doc-fix') }}
+    runs-on: [self-hosted, builder]
+    steps:
+      - name: Set envs
+        run: |
+          cat >> "$GITHUB_ENV" << 'EOF'
+          TEMP_PATH=${{runner.temp}}/build_check
+          IMAGES_PATH=${{runner.temp}}/images_path
+          REPO_COPY=${{runner.temp}}/build_check/ClickHouse
+          CACHES_PATH=${{runner.temp}}/../ccaches
+          CHECK_NAME=ClickHouse build check (actions)
+          BUILD_NAME=binary_gcc
+          EOF
+      - name: Download changed images
+        uses: actions/download-artifact@v2
+        with:
+          name: changed_images
+          path: ${{ env.IMAGES_PATH }}
+      - name: Clear repository
+        run: |
+          sudo rm -fr "$GITHUB_WORKSPACE" && mkdir "$GITHUB_WORKSPACE"
+      - name: Check out repository code
+        uses: actions/checkout@v2
+        with:
+          submodules: 'true'
+          fetch-depth: 0 # otherwise we will have no info about contributors
+      - name: Build
+        run: |
+          sudo rm -fr "$TEMP_PATH"
+          mkdir -p "$TEMP_PATH"
+          cp -r "$GITHUB_WORKSPACE" "$TEMP_PATH"
+          cd "$REPO_COPY/tests/ci" && python3 build_check.py "$CHECK_NAME" "$BUILD_NAME"
+      - name: Upload build URLs to artifacts
+        if: ${{ success() || failure() }}
+        uses: actions/upload-artifact@v2
+        with:
+          name: ${{ env.BUILD_NAME }}
+          path: ${{ env.TEMP_PATH }}/${{ env.BUILD_NAME }}.json
+      - name: Cleanup
+        if: always()
+        run: |
+          docker kill "$(docker ps -q)" ||:
+          docker rm -f "$(docker ps -a -q)" ||:
+          sudo rm -fr "$TEMP_PATH" "$CACHES_PATH"
   BuilderDebAsan:
     needs: [DockerHubPush]
     runs-on: [self-hosted, builder]
@@ -918,6 +964,7 @@ jobs:
       - BuilderDebRelease
       - BuilderDebAarch64
       - BuilderBinRelease
+      - BuilderBinGCC
       - BuilderDebAsan
       - BuilderDebTsan
       - BuilderDebUBsan
@@ -2608,6 +2655,40 @@ jobs:
           docker kill "$(docker ps -q)" ||:
           docker rm -f "$(docker ps -a -q)" ||:
           sudo rm -fr "$TEMP_PATH"
+  UnitTestsReleaseGCC:
+    needs: [BuilderBinGCC]
+    runs-on: [self-hosted, fuzzer-unit-tester]
+    steps:
+      - name: Set envs
+        run: |
+          cat >> "$GITHUB_ENV" << 'EOF'
+          TEMP_PATH=${{runner.temp}}/unit_tests_asan
+          REPORTS_PATH=${{runner.temp}}/reports_dir
+          CHECK_NAME=Unit tests (release-gcc, actions)
+          REPO_COPY=${{runner.temp}}/unit_tests_asan/ClickHouse
+          EOF
+      - name: Download json reports
+        uses: actions/download-artifact@v2
+        with:
+          path: ${{ env.REPORTS_PATH }}
+      - name: Clear repository
+        run: |
+          sudo rm -fr "$GITHUB_WORKSPACE" && mkdir "$GITHUB_WORKSPACE"
+      - name: Check out repository code
+        uses: actions/checkout@v2
+      - name: Unit test
+        run: |
+          sudo rm -fr "$TEMP_PATH"
+          mkdir -p "$TEMP_PATH"
+          cp -r "$GITHUB_WORKSPACE" "$TEMP_PATH"
+          cd "$REPO_COPY/tests/ci"
+          python3 unit_tests_check.py "$CHECK_NAME"
+      - name: Cleanup
+        if: always()
+        run: |
+          docker kill "$(docker ps -q)" ||:
+          docker rm -f "$(docker ps -a -q)" ||:
+          sudo rm -fr "$TEMP_PATH"
   UnitTestsTsan:
     needs: [BuilderDebTsan]
     runs-on: [self-hosted, fuzzer-unit-tester]

.github/workflows/pull_request.yml

@@ -370,6 +370,48 @@ jobs:
           docker kill "$(docker ps -q)" ||:
           docker rm -f "$(docker ps -a -q)" ||:
           sudo rm -fr "$TEMP_PATH" "$CACHES_PATH"
+  BuilderBinGCC:
+    needs: [DockerHubPush, FastTest]
+    runs-on: [self-hosted, builder]
+    steps:
+      - name: Set envs
+        run: |
+          cat >> "$GITHUB_ENV" << 'EOF'
+          TEMP_PATH=${{runner.temp}}/build_check
+          IMAGES_PATH=${{runner.temp}}/images_path
+          REPO_COPY=${{runner.temp}}/build_check/ClickHouse
+          CACHES_PATH=${{runner.temp}}/../ccaches
+          CHECK_NAME=ClickHouse build check (actions)
+          BUILD_NAME=binary_gcc
+          EOF
+      - name: Download changed images
+        uses: actions/download-artifact@v2
+        with:
+          name: changed_images
+          path: ${{ runner.temp }}/images_path
+      - name: Check out repository code
+        uses: actions/checkout@v2
+        with:
+          submodules: 'true'
+          fetch-depth: 0 # otherwise we will have no info about contributors
+      - name: Build
+        run: |
+          sudo rm -fr "$TEMP_PATH"
+          mkdir -p "$TEMP_PATH"
+          cp -r "$GITHUB_WORKSPACE" "$TEMP_PATH"
+          cd "$REPO_COPY/tests/ci" && python3 build_check.py "$CHECK_NAME" "$BUILD_NAME"
+      - name: Upload build URLs to artifacts
+        if: ${{ success() || failure() }}
+        uses: actions/upload-artifact@v2
+        with:
+          name: ${{ env.BUILD_NAME }}
+          path: ${{ runner.temp }}/build_check/${{ env.BUILD_NAME }}.json
+      - name: Cleanup
+        if: always()
+        run: |
+          docker kill "$(docker ps -q)" ||:
+          docker rm -f "$(docker ps -a -q)" ||:
+          sudo rm -fr "$TEMP_PATH" "$CACHES_PATH"
   BuilderDebAarch64:
     needs: [DockerHubPush, FastTest]
     runs-on: [self-hosted, builder]
@@ -963,6 +1005,7 @@ jobs:
       - BuilderDebRelease
       - BuilderDebAarch64
       - BuilderBinRelease
+      - BuilderBinGCC
       - BuilderDebAsan
       - BuilderDebTsan
       - BuilderDebUBsan
@@ -2808,6 +2851,40 @@ jobs:
           docker kill "$(docker ps -q)" ||:
           docker rm -f "$(docker ps -a -q)" ||:
           sudo rm -fr "$TEMP_PATH"
+  UnitTestsReleaseGCC:
+    needs: [BuilderBinGCC]
+    runs-on: [self-hosted, fuzzer-unit-tester]
+    steps:
+      - name: Set envs
+        run: |
+          cat >> "$GITHUB_ENV" << 'EOF'
+          TEMP_PATH=${{runner.temp}}/unit_tests_asan
+          REPORTS_PATH=${{runner.temp}}/reports_dir
+          CHECK_NAME=Unit tests (release-gcc, actions)
+          REPO_COPY=${{runner.temp}}/unit_tests_asan/ClickHouse
+          EOF
+      - name: Download json reports
+        uses: actions/download-artifact@v2
+        with:
+          path: ${{ env.REPORTS_PATH }}
+      - name: Clear repository
+        run: |
+          sudo rm -fr "$GITHUB_WORKSPACE" && mkdir "$GITHUB_WORKSPACE"
+      - name: Check out repository code
+        uses: actions/checkout@v2
+      - name: Unit test
+        run: |
+          sudo rm -fr "$TEMP_PATH"
+          mkdir -p "$TEMP_PATH"
+          cp -r "$GITHUB_WORKSPACE" "$TEMP_PATH"
+          cd "$REPO_COPY/tests/ci"
+          python3 unit_tests_check.py "$CHECK_NAME"
+      - name: Cleanup
+        if: always()
+        run: |
+          docker kill "$(docker ps -q)" ||:
+          docker rm -f "$(docker ps -a -q)" ||:
+          sudo rm -fr "$TEMP_PATH"
   UnitTestsTsan:
     needs: [BuilderDebTsan]
     runs-on: [self-hosted, fuzzer-unit-tester]

CHANGELOG.md

@@ -11,6 +11,7 @@
 
 * Make `arrayCompact` function behave as other higher-order functions: perform compaction not of lambda function results but on the original array. If you're using nontrivial lambda functions in arrayCompact you may restore old behaviour by wrapping `arrayCompact` arguments into `arrayMap`. Closes [#34010](https://github.com/ClickHouse/ClickHouse/issues/34010) [#18535](https://github.com/ClickHouse/ClickHouse/issues/18535) [#14778](https://github.com/ClickHouse/ClickHouse/issues/14778). [#34795](https://github.com/ClickHouse/ClickHouse/pull/34795) ([Alexandre Snarskii](https://github.com/snar)).
 * Change implementation specific behavior on overflow of function `toDatetime`. It will be saturated to the nearest min/max supported instant of datetime instead of wraparound. This change is highlighted as "backward incompatible" because someone may unintentionally rely on the old behavior. [#32898](https://github.com/ClickHouse/ClickHouse/pull/32898) ([HaiBo Li](https://github.com/marising)).
+* Make function `cast(value, 'IPv4')`, `cast(value, 'IPv6')` behave same as `toIPv4`, `toIPv6` functions. Changed behavior of incorrect IP address passed into functions `toIPv4`,` toIPv6`,  now if invalid IP address passes into this functions exception will be raised, before this function return default value. Added functions `IPv4StringToNumOrDefault`, `IPv4StringToNumOrNull`, `IPv6StringToNumOrDefault`, `IPv6StringOrNull` `toIPv4OrDefault`, `toIPv4OrNull`, `toIPv6OrDefault`, `toIPv6OrNull`. Functions `IPv4StringToNumOrDefault `, `toIPv4OrDefault `, `toIPv6OrDefault ` should be used if previous logic relied on `IPv4StringToNum`, `toIPv4`, `toIPv6` returning default value for invalid address. Added setting `cast_ipv4_ipv6_default_on_conversion_error`, if this setting enabled, then IP address conversion functions will behave as before. Closes [#22825](https://github.com/ClickHouse/ClickHouse/issues/22825). Closes [#5799](https://github.com/ClickHouse/ClickHouse/issues/5799). Closes [#35156](https://github.com/ClickHouse/ClickHouse/issues/35156). [#35240](https://github.com/ClickHouse/ClickHouse/pull/35240) ([Maksim Kita](https://github.com/kitaisreal)).
 
 #### New Feature
 
@@ -366,7 +367,7 @@
 
 #### Improvement
 
-* Now date time conversion functions that generates time before `1970-01-01 00:00:00` will be saturated to zero instead of overflow. [#29953](https://github.com/ClickHouse/ClickHouse/pull/29953) ([Amos Bird](https://github.com/amosbird)). It also fixes a bug in index analysis if date truncation function would yield result before the Unix epoch. 
+* Now date time conversion functions that generates time before `1970-01-01 00:00:00` will be saturated to zero instead of overflow. [#29953](https://github.com/ClickHouse/ClickHouse/pull/29953) ([Amos Bird](https://github.com/amosbird)). It also fixes a bug in index analysis if date truncation function would yield result before the Unix epoch.
 * Always display resource usage (total CPU usage, total RAM usage and max RAM usage per host) in client. [#33271](https://github.com/ClickHouse/ClickHouse/pull/33271) ([alexey-milovidov](https://github.com/alexey-milovidov)).
 * Improve `Bool` type serialization and deserialization, check the range of values. [#32984](https://github.com/ClickHouse/ClickHouse/pull/32984) ([Kruglov Pavel](https://github.com/Avogar)).
 * If an invalid setting is defined using the `SET` query or using the query parameters in the HTTP request, error message will contain suggestions that are similar to the invalid setting string (if any exists). [#32946](https://github.com/ClickHouse/ClickHouse/pull/32946) ([Antonio Andelic](https://github.com/antonio2368)).

CMakeLists.txt

@@ -261,12 +261,12 @@ endif ()
 # Add a section with the hash of the compiled machine code for integrity checks.
 # Only for official builds, because adding a section can be time consuming (rewrite of several GB).
 # And cross compiled binaries are not supported (since you cannot execute clickhouse hash-binary)
-if (OBJCOPY_PATH AND YANDEX_OFFICIAL_BUILD AND (NOT CMAKE_TOOLCHAIN_FILE))
+if (OBJCOPY_PATH AND CLICKHOUSE_OFFICIAL_BUILD AND (NOT CMAKE_TOOLCHAIN_FILE))
     set (USE_BINARY_HASH 1)
 endif ()
 
 # Allows to build stripped binary in a separate directory
-if (OBJCOPY_PATH AND READELF_PATH)
+if (OBJCOPY_PATH AND STRIP_PATH)
     option(INSTALL_STRIPPED_BINARIES "Build stripped binaries with debug info in separate directory" OFF)
     if (INSTALL_STRIPPED_BINARIES)
         set(STRIPPED_BINARIES_OUTPUT "stripped" CACHE STRING "A separate directory for stripped information")

base/glibc-compatibility/CMakeLists.txt

@@ -51,6 +51,6 @@ if (GLIBC_COMPATIBILITY)
 
     message (STATUS "Some symbols from glibc will be replaced for compatibility")
 
-elseif (YANDEX_OFFICIAL_BUILD)
+elseif (CLICKHOUSE_OFFICIAL_BUILD)
     message (WARNING "Option GLIBC_COMPATIBILITY must be turned on for production builds.")
 endif ()

cmake/strip.sh

@@ -1,28 +0,0 @@
-#!/usr/bin/env bash
-
-BINARY_PATH=$1
-BINARY_NAME=$(basename "$BINARY_PATH")
-DESTINATION_STRIPPED_DIR=$2
-OBJCOPY_PATH=${3:objcopy}
-READELF_PATH=${4:readelf}
-
-BUILD_ID=$($READELF_PATH -n "$1" | sed -n '/Build ID/ { s/.*: //p; q; }')
-BUILD_ID_PREFIX=${BUILD_ID:0:2}
-BUILD_ID_SUFFIX=${BUILD_ID:2}
-
-DESTINATION_DEBUG_INFO_DIR="$DESTINATION_STRIPPED_DIR/lib/debug/.build-id"
-DESTINATION_STRIP_BINARY_DIR="$DESTINATION_STRIPPED_DIR/bin"
-
-mkdir -p "$DESTINATION_DEBUG_INFO_DIR/$BUILD_ID_PREFIX"
-mkdir -p "$DESTINATION_STRIP_BINARY_DIR"
-
-
-cp "$BINARY_PATH" "$DESTINATION_STRIP_BINARY_DIR/$BINARY_NAME"
-
-$OBJCOPY_PATH --only-keep-debug --compress-debug-sections "$DESTINATION_STRIP_BINARY_DIR/$BINARY_NAME" "$DESTINATION_DEBUG_INFO_DIR/$BUILD_ID_PREFIX/$BUILD_ID_SUFFIX.debug"
-chmod 0644 "$DESTINATION_DEBUG_INFO_DIR/$BUILD_ID_PREFIX/$BUILD_ID_SUFFIX.debug"
-chown 0:0 "$DESTINATION_DEBUG_INFO_DIR/$BUILD_ID_PREFIX/$BUILD_ID_SUFFIX.debug"
-
-strip --remove-section=.comment --remove-section=.note "$DESTINATION_STRIP_BINARY_DIR/$BINARY_NAME"
-
-$OBJCOPY_PATH --add-gnu-debuglink "$DESTINATION_DEBUG_INFO_DIR/$BUILD_ID_PREFIX/$BUILD_ID_SUFFIX.debug" "$DESTINATION_STRIP_BINARY_DIR/$BINARY_NAME"

cmake/strip_binary.cmake

@@ -11,16 +11,43 @@ macro(clickhouse_strip_binary)
        message(FATAL_ERROR "A binary path name must be provided for stripping binary")
    endif()
 
-
    if (NOT DEFINED STRIP_DESTINATION_DIR)
        message(FATAL_ERROR "Destination directory for stripped binary must be provided")
    endif()
 
    add_custom_command(TARGET ${STRIP_TARGET} POST_BUILD
-     COMMAND bash ${ClickHouse_SOURCE_DIR}/cmake/strip.sh ${STRIP_BINARY_PATH} ${STRIP_DESTINATION_DIR} ${OBJCOPY_PATH} ${READELF_PATH}
+       COMMAND mkdir -p "${STRIP_DESTINATION_DIR}/lib/debug/bin"
-     COMMENT "Stripping clickhouse binary" VERBATIM
+       COMMAND mkdir -p "${STRIP_DESTINATION_DIR}/bin"
+       COMMAND cp "${STRIP_BINARY_PATH}" "${STRIP_DESTINATION_DIR}/bin/${STRIP_TARGET}"
+       COMMAND "${OBJCOPY_PATH}" --only-keep-debug --compress-debug-sections "${STRIP_DESTINATION_DIR}/bin/${STRIP_TARGET}" "${STRIP_DESTINATION_DIR}/lib/debug/bin/${STRIP_TARGET}.debug"
+       COMMAND chmod 0644 "${STRIP_DESTINATION_DIR}/lib/debug/bin/${STRIP_TARGET}.debug"
+       COMMAND "${STRIP_PATH}" --remove-section=.comment --remove-section=.note "${STRIP_DESTINATION_DIR}/bin/${STRIP_TARGET}"
+       COMMAND "${OBJCOPY_PATH}" --add-gnu-debuglink "${STRIP_DESTINATION_DIR}/lib/debug/bin/${STRIP_TARGET}.debug" "${STRIP_DESTINATION_DIR}/bin/${STRIP_TARGET}"
+       COMMENT "Stripping clickhouse binary" VERBATIM
    )
 
    install(PROGRAMS ${STRIP_DESTINATION_DIR}/bin/${STRIP_TARGET} DESTINATION ${CMAKE_INSTALL_BINDIR} COMPONENT clickhouse)
-   install(DIRECTORY ${STRIP_DESTINATION_DIR}/lib/debug DESTINATION ${CMAKE_INSTALL_LIBDIR} COMPONENT clickhouse)
+   install(FILES ${STRIP_DESTINATION_DIR}/lib/debug/bin/${STRIP_TARGET}.debug DESTINATION ${CMAKE_INSTALL_LIBDIR}/debug/${CMAKE_INSTALL_FULL_BINDIR}/${STRIP_TARGET}.debug COMPONENT clickhouse)
+endmacro()
+
+
+macro(clickhouse_make_empty_debug_info_for_nfpm)
+   set(oneValueArgs TARGET DESTINATION_DIR)
+   cmake_parse_arguments(EMPTY_DEBUG "" "${oneValueArgs}" "" ${ARGN})
+
+   if (NOT DEFINED EMPTY_DEBUG_TARGET)
+       message(FATAL_ERROR "A target name must be provided for stripping binary")
+   endif()
+
+   if (NOT DEFINED EMPTY_DEBUG_DESTINATION_DIR)
+       message(FATAL_ERROR "Destination directory for empty debug must be provided")
+   endif()
+
+   add_custom_command(TARGET ${EMPTY_DEBUG_TARGET} POST_BUILD
+       COMMAND mkdir -p "${EMPTY_DEBUG_DESTINATION_DIR}/lib/debug"
+       COMMAND touch "${EMPTY_DEBUG_DESTINATION_DIR}/lib/debug/${EMPTY_DEBUG_TARGET}.debug"
+       COMMENT "Addiding empty debug info for NFPM" VERBATIM
+   )
+
+   install(FILES "${EMPTY_DEBUG_DESTINATION_DIR}/lib/debug/${EMPTY_DEBUG_TARGET}.debug" DESTINATION "${CMAKE_INSTALL_LIBDIR}/debug/${CMAKE_INSTALL_FULL_BINDIR}" COMPONENT clickhouse)
 endmacro()

cmake/tools.cmake

@@ -170,32 +170,32 @@ else ()
     message (FATAL_ERROR "Cannot find objcopy.")
 endif ()
 
-# Readelf (FIXME copypaste)
+# Strip (FIXME copypaste)
 
 if (COMPILER_GCC)
-    find_program (READELF_PATH NAMES "llvm-readelf" "llvm-readelf-13" "llvm-readelf-12" "llvm-readelf-11" "readelf")
+    find_program (STRIP_PATH NAMES "llvm-strip" "llvm-strip-13" "llvm-strip-12" "llvm-strip-11" "strip")
 else ()
-    find_program (READELF_PATH NAMES "llvm-readelf-${COMPILER_VERSION_MAJOR}" "llvm-readelf" "readelf")
+    find_program (STRIP_PATH NAMES "llvm-strip-${COMPILER_VERSION_MAJOR}" "llvm-strip" "strip")
 endif ()
 
-if (NOT READELF_PATH AND OS_DARWIN)
+if (NOT STRIP_PATH AND OS_DARWIN)
     find_program (BREW_PATH NAMES "brew")
     if (BREW_PATH)
         execute_process (COMMAND ${BREW_PATH} --prefix llvm ERROR_QUIET OUTPUT_STRIP_TRAILING_WHITESPACE OUTPUT_VARIABLE LLVM_PREFIX)
         if (LLVM_PREFIX)
-            find_program (READELF_PATH NAMES "llvm-readelf" PATHS "${LLVM_PREFIX}/bin" NO_DEFAULT_PATH)
+            find_program (STRIP_PATH NAMES "llvm-strip" PATHS "${LLVM_PREFIX}/bin" NO_DEFAULT_PATH)
         endif ()
-        if (NOT READELF_PATH)
+        if (NOT STRIP_PATH)
             execute_process (COMMAND ${BREW_PATH} --prefix binutils ERROR_QUIET OUTPUT_STRIP_TRAILING_WHITESPACE OUTPUT_VARIABLE BINUTILS_PREFIX)
             if (BINUTILS_PREFIX)
-                find_program (READELF_PATH NAMES "readelf" PATHS "${BINUTILS_PREFIX}/bin" NO_DEFAULT_PATH)
+                find_program (STRIP_PATH NAMES "strip" PATHS "${BINUTILS_PREFIX}/bin" NO_DEFAULT_PATH)
             endif ()
         endif ()
     endif ()
 endif ()
 
-if (READELF_PATH)
+if (STRIP_PATH)
-    message (STATUS "Using readelf: ${READELF_PATH}")
+    message (STATUS "Using strip: ${STRIP_PATH}")
 else ()
-    message (FATAL_ERROR "Cannot find readelf.")
+    message (FATAL_ERROR "Cannot find strip.")
 endif ()

cmake/version.cmake

@@ -18,6 +18,6 @@ set (VERSION_STRING_SHORT "${VERSION_MAJOR}.${VERSION_MINOR}")
 
 math (EXPR VERSION_INTEGER "${VERSION_PATCH} + ${VERSION_MINOR}*1000 + ${VERSION_MAJOR}*1000000")
 
-if(YANDEX_OFFICIAL_BUILD)
+if(CLICKHOUSE_OFFICIAL_BUILD)
     set(VERSION_OFFICIAL " (official build)")
 endif()

contrib/avro-cmake/CMakeLists.txt

@@ -69,9 +69,10 @@ endif ()
 target_compile_options(_avrocpp PRIVATE ${SUPPRESS_WARNINGS})
 
 # create a symlink to include headers with <avro/...>
+set(AVRO_INCLUDE_DIR "${CMAKE_CURRENT_BINARY_DIR}/include")
 ADD_CUSTOM_TARGET(avro_symlink_headers ALL
-    COMMAND ${CMAKE_COMMAND} -E make_directory "${AVROCPP_ROOT_DIR}/include"
+    COMMAND ${CMAKE_COMMAND} -E make_directory "${AVRO_INCLUDE_DIR}"
-    COMMAND ${CMAKE_COMMAND} -E create_symlink "${AVROCPP_ROOT_DIR}/api" "${AVROCPP_ROOT_DIR}/include/avro"
+    COMMAND ${CMAKE_COMMAND} -E create_symlink "${AVROCPP_ROOT_DIR}/api" "${AVRO_INCLUDE_DIR}/avro"
 )
 add_dependencies(_avrocpp avro_symlink_headers)
-target_include_directories(_avrocpp SYSTEM BEFORE PUBLIC "${AVROCPP_ROOT_DIR}/include")
+target_include_directories(_avrocpp SYSTEM BEFORE PUBLIC "${AVRO_INCLUDE_DIR}")

contrib/boost-cmake/CMakeLists.txt

@@ -27,7 +27,11 @@ target_include_directories (_boost_headers_only SYSTEM BEFORE INTERFACE ${LIBRAR
 
 # asio
 
-target_compile_definitions (_boost_headers_only INTERFACE BOOST_ASIO_STANDALONE=1)
+target_compile_definitions (_boost_headers_only INTERFACE
+  BOOST_ASIO_STANDALONE=1
+  # Avoid using of deprecated in c++ > 17 std::result_of
+  BOOST_ASIO_HAS_STD_INVOKE_RESULT=1
+)
 
 # iostreams
 

contrib/hyperscan

@@ -1 +1 @@
-Subproject commit e9f08df0213fc637aac0a5bbde9beeaeba2fe9fa
+Subproject commit 5edc68c5ac68d2d4f876159e9ee84def6d3dc87c

contrib/libcxx

@@ -1 +1 @@
-Subproject commit 61e60294b1de01483caa9f5d00f437c99b674de6
+Subproject commit 172b2ae074f6755145b91c53a95c8540c1468239

contrib/libcxx-cmake/CMakeLists.txt

@@ -18,12 +18,14 @@ set(SRCS
 "${LIBCXX_SOURCE_DIR}/src/filesystem/directory_iterator.cpp"
 "${LIBCXX_SOURCE_DIR}/src/filesystem/int128_builtins.cpp"
 "${LIBCXX_SOURCE_DIR}/src/filesystem/operations.cpp"
+"${LIBCXX_SOURCE_DIR}/src/format.cpp"
 "${LIBCXX_SOURCE_DIR}/src/functional.cpp"
 "${LIBCXX_SOURCE_DIR}/src/future.cpp"
 "${LIBCXX_SOURCE_DIR}/src/hash.cpp"
 "${LIBCXX_SOURCE_DIR}/src/ios.cpp"
 "${LIBCXX_SOURCE_DIR}/src/ios.instantiations.cpp"
 "${LIBCXX_SOURCE_DIR}/src/iostream.cpp"
+"${LIBCXX_SOURCE_DIR}/src/legacy_pointer_safety.cpp"
 "${LIBCXX_SOURCE_DIR}/src/locale.cpp"
 "${LIBCXX_SOURCE_DIR}/src/memory.cpp"
 "${LIBCXX_SOURCE_DIR}/src/mutex.cpp"
@@ -33,6 +35,9 @@ set(SRCS
 "${LIBCXX_SOURCE_DIR}/src/random.cpp"
 "${LIBCXX_SOURCE_DIR}/src/random_shuffle.cpp"
 "${LIBCXX_SOURCE_DIR}/src/regex.cpp"
+"${LIBCXX_SOURCE_DIR}/src/ryu/d2fixed.cpp"
+"${LIBCXX_SOURCE_DIR}/src/ryu/d2s.cpp"
+"${LIBCXX_SOURCE_DIR}/src/ryu/f2s.cpp"
 "${LIBCXX_SOURCE_DIR}/src/shared_mutex.cpp"
 "${LIBCXX_SOURCE_DIR}/src/stdexcept.cpp"
 "${LIBCXX_SOURCE_DIR}/src/string.cpp"
@@ -49,7 +54,9 @@ set(SRCS
 add_library(cxx ${SRCS})
 set_target_properties(cxx PROPERTIES FOLDER "contrib/libcxx-cmake")
 
-target_include_directories(cxx SYSTEM BEFORE PUBLIC $<BUILD_INTERFACE:${LIBCXX_SOURCE_DIR}/include>)
+target_include_directories(cxx SYSTEM BEFORE PUBLIC
+        $<BUILD_INTERFACE:${LIBCXX_SOURCE_DIR}/include>
+        $<BUILD_INTERFACE:${LIBCXX_SOURCE_DIR}>/src)
 target_compile_definitions(cxx PRIVATE -D_LIBCPP_BUILDING_LIBRARY -DLIBCXX_BUILDING_LIBCXXABI)
 
 # Enable capturing stack traces for all exceptions.

contrib/libcxxabi

@@ -1 +1 @@
-Subproject commit df8f1e727dbc9e2bedf2282096fa189dc3fe0076
+Subproject commit 6eb7cc7a7bdd779e6734d1b9fb451df2274462d7

contrib/libcxxabi-cmake/CMakeLists.txt

@@ -1,24 +1,24 @@
 set(LIBCXXABI_SOURCE_DIR "${ClickHouse_SOURCE_DIR}/contrib/libcxxabi")
 
 set(SRCS
-"${LIBCXXABI_SOURCE_DIR}/src/stdlib_stdexcept.cpp"
-"${LIBCXXABI_SOURCE_DIR}/src/cxa_virtual.cpp"
-"${LIBCXXABI_SOURCE_DIR}/src/cxa_thread_atexit.cpp"
-"${LIBCXXABI_SOURCE_DIR}/src/fallback_malloc.cpp"
-"${LIBCXXABI_SOURCE_DIR}/src/cxa_guard.cpp"
-"${LIBCXXABI_SOURCE_DIR}/src/cxa_default_handlers.cpp"
-"${LIBCXXABI_SOURCE_DIR}/src/cxa_personality.cpp"
-"${LIBCXXABI_SOURCE_DIR}/src/stdlib_exception.cpp"
 "${LIBCXXABI_SOURCE_DIR}/src/abort_message.cpp"
+"${LIBCXXABI_SOURCE_DIR}/src/cxa_aux_runtime.cpp"
+"${LIBCXXABI_SOURCE_DIR}/src/cxa_default_handlers.cpp"
 "${LIBCXXABI_SOURCE_DIR}/src/cxa_demangle.cpp"
 "${LIBCXXABI_SOURCE_DIR}/src/cxa_exception.cpp"
-"${LIBCXXABI_SOURCE_DIR}/src/cxa_handlers.cpp"
 "${LIBCXXABI_SOURCE_DIR}/src/cxa_exception_storage.cpp"
-"${LIBCXXABI_SOURCE_DIR}/src/private_typeinfo.cpp"
+"${LIBCXXABI_SOURCE_DIR}/src/cxa_guard.cpp"
-"${LIBCXXABI_SOURCE_DIR}/src/stdlib_typeinfo.cpp"
+"${LIBCXXABI_SOURCE_DIR}/src/cxa_handlers.cpp"
-"${LIBCXXABI_SOURCE_DIR}/src/cxa_aux_runtime.cpp"
+"${LIBCXXABI_SOURCE_DIR}/src/cxa_personality.cpp"
+"${LIBCXXABI_SOURCE_DIR}/src/cxa_thread_atexit.cpp"
 "${LIBCXXABI_SOURCE_DIR}/src/cxa_vector.cpp"
+"${LIBCXXABI_SOURCE_DIR}/src/cxa_virtual.cpp"
+"${LIBCXXABI_SOURCE_DIR}/src/fallback_malloc.cpp"
+"${LIBCXXABI_SOURCE_DIR}/src/private_typeinfo.cpp"
+"${LIBCXXABI_SOURCE_DIR}/src/stdlib_exception.cpp"
 "${LIBCXXABI_SOURCE_DIR}/src/stdlib_new_delete.cpp"
+"${LIBCXXABI_SOURCE_DIR}/src/stdlib_stdexcept.cpp"
+"${LIBCXXABI_SOURCE_DIR}/src/stdlib_typeinfo.cpp"
 )
 
 add_library(cxxabi ${SRCS})
@@ -30,6 +30,7 @@ target_compile_options(cxxabi PRIVATE -w)
 target_include_directories(cxxabi SYSTEM BEFORE
     PUBLIC $<BUILD_INTERFACE:${LIBCXXABI_SOURCE_DIR}/include>
     PRIVATE $<BUILD_INTERFACE:${LIBCXXABI_SOURCE_DIR}/../libcxx/include>
+    PRIVATE $<BUILD_INTERFACE:${LIBCXXABI_SOURCE_DIR}/../libcxx/src>
 )
 target_compile_definitions(cxxabi PRIVATE -D_LIBCPP_BUILDING_LIBRARY)
 target_compile_options(cxxabi PRIVATE -nostdinc++ -fno-sanitize=undefined -Wno-macro-redefined) # If we don't disable UBSan, infinite recursion happens in dynamic_cast.

contrib/libxml2

@@ -1 +1 @@
-Subproject commit 18890f471c420411aa3c989e104d090966ec9dbf
+Subproject commit a075d256fd9ff15590b86d981b75a50ead124fca

contrib/llvm-cmake/CMakeLists.txt

@@ -1,12 +1,9 @@
-# During cross-compilation in our CI we have to use llvm-tblgen and other building tools
+if (APPLE OR NOT ARCH_AMD64 OR SANITIZE STREQUAL "undefined")
-# tools to be build for host architecture and everything else for target architecture (e.g. AArch64)
+   set (ENABLE_EMBEDDED_COMPILER_DEFAULT OFF)
-# Possible workaround is to use llvm-tblgen from some package...
-# But lets just enable LLVM for native builds
-if (CMAKE_CROSSCOMPILING OR SANITIZE STREQUAL "undefined")
-    set (ENABLE_EMBEDDED_COMPILER_DEFAULT OFF)
 else()
-    set (ENABLE_EMBEDDED_COMPILER_DEFAULT ON)
+   set (ENABLE_EMBEDDED_COMPILER_DEFAULT ON)
 endif()
+
 option (ENABLE_EMBEDDED_COMPILER "Enable support for 'compile_expressions' option for query execution" ${ENABLE_EMBEDDED_COMPILER_DEFAULT})
 
 if (NOT ENABLE_EMBEDDED_COMPILER)

contrib/replxx

@@ -1 +1 @@
-Subproject commit 9460e5e0fc10f78f460af26a6bd928798cac864d
+Subproject commit 6f0b6f151ae2a044625ae93acd19ca365fcea64d

docker/docs/check/Dockerfile

@@ -1,4 +1,3 @@
-# rebuild in #33610
 # docker build -t clickhouse/docs-check .
 ARG FROM_TAG=latest
 FROM clickhouse/docs-builder:$FROM_TAG

docker/packager/packager

@@ -163,6 +163,7 @@ def parse_env_variables(
         cmake_flags.append("-DCMAKE_INSTALL_PREFIX=/usr")
         cmake_flags.append("-DCMAKE_INSTALL_SYSCONFDIR=/etc")
         cmake_flags.append("-DCMAKE_INSTALL_LOCALSTATEDIR=/var")
+        cmake_flags.append("-DBUILD_STANDALONE_KEEPER=ON")
         if is_release_build(build_type, package_type, sanitizer, split_binary):
             cmake_flags.append("-DINSTALL_STRIPPED_BINARIES=ON")
 
@@ -244,7 +245,7 @@ def parse_env_variables(
         result.append(f"AUTHOR='{author}'")
 
     if official:
-        cmake_flags.append("-DYANDEX_OFFICIAL_BUILD=1")
+        cmake_flags.append("-DCLICKHOUSE_OFFICIAL_BUILD=1")
 
     result.append('CMAKE_FLAGS="' + " ".join(cmake_flags) + '"')
 

docker/test/fasttest/run.sh

@@ -267,6 +267,7 @@ function run_tests
     local test_opts=(
         --hung-check
         --fast-tests-only
+        --no-random-settings
         --no-long
         --testname
         --shard

docker/test/fuzzer/run-fuzzer.sh

@@ -13,7 +13,7 @@ script_dir="$( cd "$( dirname "${BASH_SOURCE[0]}" )" >/dev/null 2>&1 && pwd )"
 echo "$script_dir"
 repo_dir=ch
 BINARY_TO_DOWNLOAD=${BINARY_TO_DOWNLOAD:="clang-13_debug_none_bundled_unsplitted_disable_False_binary"}
-BINARY_URL_TO_DOWNLOAD=${BINARY_URL_TO_DOWNLOAD:="https://clickhouse-builds.s3.yandex.net/$PR_TO_TEST/$SHA_TO_TEST/clickhouse_build_check/$BINARY_TO_DOWNLOAD/clickhouse"}
+BINARY_URL_TO_DOWNLOAD=${BINARY_URL_TO_DOWNLOAD:="https://clickhouse-builds.s3.amazonaws.com/$PR_TO_TEST/$SHA_TO_TEST/clickhouse_build_check/$BINARY_TO_DOWNLOAD/clickhouse"}
 
 function clone
 {

docker/test/keeper-jepsen/run.sh

@@ -2,7 +2,7 @@
 set -euo pipefail
 
 
-CLICKHOUSE_PACKAGE=${CLICKHOUSE_PACKAGE:="https://clickhouse-builds.s3.yandex.net/$PR_TO_TEST/$SHA_TO_TEST/clickhouse_build_check/clang-13_relwithdebuginfo_none_bundled_unsplitted_disable_False_binary/clickhouse"}
+CLICKHOUSE_PACKAGE=${CLICKHOUSE_PACKAGE:="https://clickhouse-builds.s3.amazonaws.com/$PR_TO_TEST/$SHA_TO_TEST/clickhouse_build_check/clang-13_relwithdebuginfo_none_bundled_unsplitted_disable_False_binary/clickhouse"}
 CLICKHOUSE_REPO_PATH=${CLICKHOUSE_REPO_PATH:=""}
 
 
@@ -10,7 +10,7 @@ if [ -z "$CLICKHOUSE_REPO_PATH" ]; then
     CLICKHOUSE_REPO_PATH=ch
     rm -rf ch ||:
     mkdir ch ||:
-    wget -nv -nd -c "https://clickhouse-test-reports.s3.yandex.net/$PR_TO_TEST/$SHA_TO_TEST/repo/clickhouse_no_subs.tar.gz"
+    wget -nv -nd -c "https://clickhouse-test-reports.s3.amazonaws.com/$PR_TO_TEST/$SHA_TO_TEST/repo/clickhouse_no_subs.tar.gz"
     tar -C ch --strip-components=1 -xf clickhouse_no_subs.tar.gz
     ls -lath ||:
 fi

docker/test/performance-comparison/compare.sh

@@ -1294,15 +1294,15 @@ create table ci_checks engine File(TSVWithNamesAndTypes, 'ci-checks.tsv')
         select '' test_name,
             '$(sed -n 's/.*<!--message: \(.*\)-->/\1/p' report.html)' test_status,
             0 test_duration_ms,
-            'https://clickhouse-test-reports.s3.yandex.net/$PR_TO_TEST/$SHA_TO_TEST/performance_comparison/report.html#fail1' report_url
+            'https://clickhouse-test-reports.s3.amazonaws.com/$PR_TO_TEST/$SHA_TO_TEST/performance_comparison/report.html#fail1' report_url
         union all
             select test || ' #' || toString(query_index), 'slower' test_status, 0 test_duration_ms,
-                'https://clickhouse-test-reports.s3.yandex.net/$PR_TO_TEST/$SHA_TO_TEST/performance_comparison/report.html#changes-in-performance.'
+                'https://clickhouse-test-reports.s3.amazonaws.com/$PR_TO_TEST/$SHA_TO_TEST/performance_comparison/report.html#changes-in-performance.'
                     || test || '.' || toString(query_index) report_url
             from queries where changed_fail != 0 and diff > 0
         union all
             select test || ' #' || toString(query_index), 'unstable' test_status, 0 test_duration_ms,
-                'https://clickhouse-test-reports.s3.yandex.net/$PR_TO_TEST/$SHA_TO_TEST/performance_comparison/report.html#unstable-queries.'
+                'https://clickhouse-test-reports.s3.amazonaws.com/$PR_TO_TEST/$SHA_TO_TEST/performance_comparison/report.html#unstable-queries.'
                     || test || '.' || toString(query_index) report_url
             from queries where unstable_fail != 0
     )

docker/test/performance-comparison/download.sh

@@ -16,26 +16,17 @@ right_sha=$4
 datasets=${CHPC_DATASETS-"hits1 hits10 hits100 values"}
 
 declare -A dataset_paths
-if [[ $S3_URL == *"s3.amazonaws.com"* ]]; then
+dataset_paths["hits10"]="https://clickhouse-private-datasets.s3.amazonaws.com/hits_10m_single/partitions/hits_10m_single.tar"
-    dataset_paths["hits10"]="https://clickhouse-private-datasets.s3.amazonaws.com/hits_10m_single/partitions/hits_10m_single.tar"
+dataset_paths["hits100"]="https://clickhouse-private-datasets.s3.amazonaws.com/hits_100m_single/partitions/hits_100m_single.tar"
-    dataset_paths["hits100"]="https://clickhouse-private-datasets.s3.amazonaws.com/hits_100m_single/partitions/hits_100m_single.tar"
+dataset_paths["hits1"]="https://clickhouse-datasets.s3.amazonaws.com/hits/partitions/hits_v1.tar"
-    dataset_paths["hits1"]="https://clickhouse-datasets.s3.amazonaws.com/hits/partitions/hits_v1.tar"
+dataset_paths["values"]="https://clickhouse-datasets.s3.amazonaws.com/values_with_expressions/partitions/test_values.tar"
-    dataset_paths["values"]="https://clickhouse-datasets.s3.amazonaws.com/values_with_expressions/partitions/test_values.tar"
-else
-    dataset_paths["hits10"]="https://s3.mds.yandex.net/clickhouse-private-datasets/hits_10m_single/partitions/hits_10m_single.tar"
-    dataset_paths["hits100"]="https://s3.mds.yandex.net/clickhouse-private-datasets/hits_100m_single/partitions/hits_100m_single.tar"
-    dataset_paths["hits1"]="https://clickhouse-datasets.s3.yandex.net/hits/partitions/hits_v1.tar"
-    dataset_paths["values"]="https://clickhouse-datasets.s3.yandex.net/values_with_expressions/partitions/test_values.tar"
-fi
 
 
 function download
 {
     # Historically there were various paths for the performance test package.
     # Test all of them.
-    declare -a urls_to_try=("https://s3.amazonaws.com/clickhouse-builds/$left_pr/$left_sha/performance/performance.tgz"
+    declare -a urls_to_try=("https://s3.amazonaws.com/clickhouse-builds/$left_pr/$left_sha/performance/performance.tgz")
-                            "https://clickhouse-builds.s3.yandex.net/$left_pr/$left_sha/clickhouse_build_check/performance/performance.tgz"
-                           )
 
     for path in "${urls_to_try[@]}"
     do

docker/test/performance-comparison/entrypoint.sh

@@ -4,7 +4,7 @@ set -ex
 CHPC_CHECK_START_TIMESTAMP="$(date +%s)"
 export CHPC_CHECK_START_TIMESTAMP
 
-S3_URL=${S3_URL:="https://clickhouse-builds.s3.yandex.net"}
+S3_URL=${S3_URL:="https://clickhouse-builds.s3.amazonaws.com"}
 
 COMMON_BUILD_PREFIX="/clickhouse_build_check"
 if [[ $S3_URL == *"s3.amazonaws.com"* ]]; then
@@ -64,9 +64,7 @@ function find_reference_sha
         # Historically there were various path for the performance test package,
         # test all of them.
         unset found
-        declare -a urls_to_try=("https://s3.amazonaws.com/clickhouse-builds/0/$REF_SHA/performance/performance.tgz"
+        declare -a urls_to_try=("https://s3.amazonaws.com/clickhouse-builds/0/$REF_SHA/performance/performance.tgz")
-                                "https://clickhouse-builds.s3.yandex.net/0/$REF_SHA/clickhouse_build_check/performance/performance.tgz"
-                               )
         for path in "${urls_to_try[@]}"
         do
             if curl_with_retry "$path"

docker/test/stateful/Dockerfile

@@ -11,7 +11,7 @@ RUN apt-get update -y \
 
 COPY s3downloader /s3downloader
 
-ENV S3_URL="https://clickhouse-datasets.s3.yandex.net"
+ENV S3_URL="https://clickhouse-datasets.s3.amazonaws.com"
 ENV DATASETS="hits visits"
 ENV EXPORT_S3_STORAGE_POLICIES=1
 

docker/test/stateful/s3downloader

@@ -10,7 +10,7 @@ import requests
 import tempfile
 
 
-DEFAULT_URL = 'https://clickhouse-datasets.s3.yandex.net'
+DEFAULT_URL = 'https://clickhouse-datasets.s3.amazonaws.com'
 
 AVAILABLE_DATASETS = {
     'hits': 'hits_v1.tar',

docker/test/stateless/setup_minio.sh

@@ -41,6 +41,7 @@ sleep 5
 ./mc admin user add clickminio test testtest
 ./mc admin policy set clickminio readwrite user=test
 ./mc mb clickminio/test
+./mc policy set public clickminio/test
 
 
 # Upload data to Minio. By default after unpacking all tests will in

docker/test/stress/Dockerfile

@@ -29,7 +29,7 @@ COPY ./download_previous_release /download_previous_release
 COPY run.sh /
 
 ENV DATASETS="hits visits"
-ENV S3_URL="https://clickhouse-datasets.s3.yandex.net"
+ENV S3_URL="https://clickhouse-datasets.s3.amazonaws.com"
 ENV EXPORT_S3_STORAGE_POLICIES=1
 
 CMD ["/bin/bash", "/run.sh"]

docs/en/engines/table-engines/integrations/hive.md

@@ -137,7 +137,7 @@ CREATE TABLE test.test_orc
     `f_array_array_float` Array(Array(Float32)),
     `day` String
 )
-ENGINE = Hive('thrift://202.168.117.26:9083', 'test', 'test_orc')
+ENGINE = Hive('thrift://localhost:9083', 'test', 'test_orc')
 PARTITION BY day
 
 ```

docs/en/operations/server-configuration-parameters/settings.md

@@ -1616,3 +1616,14 @@ Possible values:
 
 Default value: `10000`.
 
+## global_memory_usage_overcommit_max_wait_microseconds {#global_memory_usage_overcommit_max_wait_microseconds}
+
+Sets maximum waiting time for global overcommit tracker.
+
+Possible values:
+
+-   Positive integer.
+
+Default value: `0`.
+
+

docs/en/operations/settings/memory-overcommit.md

@@ -0,0 +1,31 @@
+# Memory overcommit
+
+Memory overcommit is an experimental technique intended to allow to set more flexible memory limits for queries.
+
+The idea of this technique is to introduce settings which can represent guaranteed amount of memory a query can use.
+When memory overcommit is enabled and the memory limit is reached ClickHouse will select the most overcommitted query and try to free memory by killing this query.
+
+When memory limit is reached any query will wait some time during atempt to allocate new memory.
+If timeout is passed and memory is freed, the query continues execution. Otherwise an exception will be thrown and the query is killed.
+
+Selection of query to stop or kill is performed by either global or user overcommit trackers depending on what memory limit is reached.
+
+## User overcommit tracker
+
+User overcommit tracker finds a query with the biggest overcommit ratio in the user's query list.
+Overcommit ratio is computed as number of allocated bytes divided by value of `max_guaranteed_memory_usage` setting.
+
+Waiting timeout is set by `memory_usage_overcommit_max_wait_microseconds` setting.
+
+**Example**
+
+```sql
+SELECT number FROM numbers(1000) GROUP BY number SETTINGS max_guaranteed_memory_usage=4000, memory_usage_overcommit_max_wait_microseconds=500
+```
+
+## Global overcommit tracker
+
+Global overcommit tracker finds a query with the biggest overcommit ratio in the list of all queries.
+In this case overcommit ratio is computed as number of allocated bytes divided by value of `max_guaranteed_memory_usage_for_user` setting.
+
+Waiting timeout is set by `global_memory_usage_overcommit_max_wait_microseconds` parameter in the configuration file.

docs/en/operations/settings/settings.md

@@ -4220,10 +4220,36 @@ Possible values:
 -   0 — Disabled.
 -   1 — Enabled. The wait time equal shutdown_wait_unfinished config.
 
-Default value: 0.
+Default value: `0`.
 
 ## shutdown_wait_unfinished
 
 The waiting time in seconds for currently handled connections when shutdown server.
 
-Default Value: 5.
+Default Value: `5`.
+
+## max_guaranteed_memory_usage
+
+Maximum guaranteed memory usage for processing of single query.
+It represents soft limit in case when hard limit is reached on user level.
+Zero means unlimited.
+Read more about [memory overcommit](memory-overcommit.md).
+
+Default value: `0`.
+
+## memory_usage_overcommit_max_wait_microseconds
+
+Maximum time thread will wait for memory to be freed in the case of memory overcommit on a user level.
+If the timeout is reached and memory is not freed, an exception is thrown.
+Read more about [memory overcommit](memory-overcommit.md).
+
+Default value: `0`.
+
+## max_guaranteed_memory_usage_for_user
+
+Maximum guaranteed memory usage for processing all concurrently running queries for the user.
+It represents soft limit in case when hard limit is reached on global level.
+Zero means unlimited.
+Read more about [memory overcommit](memory-overcommit.md).
+
+Default value: `0`.

docs/tools/requirements.txt

@@ -10,7 +10,7 @@ cssmin==0.2.0
 future==0.18.2
 htmlmin==0.1.12
 idna==2.10
-Jinja2>=3.0.3
+Jinja2==3.0.3
 jinja2-highlight==0.6.1
 jsmin==3.0.0
 livereload==2.6.3

docs/zh/engines/table-engines/integrations/hive.md

@@ -140,7 +140,7 @@ CREATE TABLE test.test_orc
     `f_array_array_float` Array(Array(Float32)),
     `day` String
 )
-ENGINE = Hive('thrift://202.168.117.26:9083', 'test', 'test_orc')
+ENGINE = Hive('thrift://localhost:9083', 'test', 'test_orc')
 PARTITION BY day
 
 ```

docs/zh/operations/system-tables/functions.md

@@ -15,7 +15,7 @@
 ```
 ┌─name─────────────────────┬─is_aggregate─┬─case_insensitive─┬─alias_to─┐
 │ sumburConsistentHash     │            0 │                0 │          │
-│ yandexConsistentHash     │            0 │                0 │          │
+│ kostikConsistentHash     │            0 │                0 │          │
 │ demangle                 │            0 │                0 │          │
 │ addressToLine            │            0 │                0 │          │
 │ JSONExtractRaw           │            0 │                0 │          │

packages/clickhouse-common-static-dbg.yaml

@@ -21,8 +21,12 @@ description: |
     This package contains the debugging symbols for clickhouse-common.
 
 contents:
-- src: root/usr/lib/debug
+- src: root/usr/lib/debug/usr/bin/clickhouse.debug
-  dst: /usr/lib/debug
+  dst: /usr/lib/debug/usr/bin/clickhouse.debug
+- src: root/usr/lib/debug/usr/bin/clickhouse-odbc-bridge.debug
+  dst: /usr/lib/debug/usr/bin/clickhouse-odbc-bridge.debug
+- src: root/usr/lib/debug/usr/bin/clickhouse-library-bridge.debug
+  dst: /usr/lib/debug/usr/bin/clickhouse-library-bridge.debug
 # docs
 - src: ../AUTHORS
   dst: /usr/share/doc/clickhouse-common-static-dbg/AUTHORS

packages/clickhouse-keeper-dbg.yaml

@@ -0,0 +1,28 @@
+# package sources should be placed in ${PWD}/root
+# nfpm should run from the same directory with a config
+name: "clickhouse-keeper-dbg"
+arch: "${DEB_ARCH}" # amd64, arm64
+platform: "linux"
+version: "${CLICKHOUSE_VERSION_STRING}"
+vendor: "ClickHouse Inc."
+homepage: "https://clickhouse.com"
+license: "Apache"
+section: "database"
+priority: "optional"
+maintainer: "ClickHouse Dev Team <packages+linux@clickhouse.com>"
+description: |
+  debugging symbols for clickhouse-keeper
+    This package contains the debugging symbols for clickhouse-keeper.
+
+contents:
+- src: root/usr/lib/debug/usr/bin/clickhouse-keeper.debug
+  dst: /usr/lib/debug/usr/bin/clickhouse-keeper.debug
+# docs
+- src: ../AUTHORS
+  dst: /usr/share/doc/clickhouse-keeper-dbg/AUTHORS
+- src: ../CHANGELOG.md
+  dst: /usr/share/doc/clickhouse-keeper-dbg/CHANGELOG.md
+- src: ../LICENSE
+  dst: /usr/share/doc/clickhouse-keeper-dbg/LICENSE
+- src: ../README.md
+  dst: /usr/share/doc/clickhouse-keeper-dbg/README.md

packages/clickhouse-keeper.yaml

@@ -0,0 +1,40 @@
+# package sources should be placed in ${PWD}/root
+# nfpm should run from the same directory with a config
+name: "clickhouse-keeper"
+arch: "${DEB_ARCH}" # amd64, arm64
+platform: "linux"
+version: "${CLICKHOUSE_VERSION_STRING}"
+vendor: "ClickHouse Inc."
+homepage: "https://clickhouse.com"
+license: "Apache"
+section: "database"
+priority: "optional"
+
+conflicts:
+- clickhouse-server
+depends:
+- adduser
+suggests:
+- clickhouse-keeper-dbg
+
+maintainer: "ClickHouse Dev Team <packages+linux@clickhouse.com>"
+description: |
+  Static clickhouse-keeper binary
+    A stand-alone clickhouse-keeper package
+
+
+contents:
+- src: root/etc/clickhouse-keeper
+  dst: /etc/clickhouse-keeper
+  type: config
+- src: root/usr/bin/clickhouse-keeper
+  dst: /usr/bin/clickhouse-keeper
+# docs
+- src: ../AUTHORS
+  dst: /usr/share/doc/clickhouse-keeper/AUTHORS
+- src: ../CHANGELOG.md
+  dst: /usr/share/doc/clickhouse-keeper/CHANGELOG.md
+- src: ../LICENSE
+  dst: /usr/share/doc/clickhouse-keeper/LICENSE
+- src: ../README.md
+  dst: /usr/share/doc/clickhouse-keeper/README.md

programs/CMakeLists.txt

@@ -473,18 +473,11 @@ else ()
     if (INSTALL_STRIPPED_BINARIES)
         clickhouse_strip_binary(TARGET clickhouse DESTINATION_DIR ${CMAKE_CURRENT_BINARY_DIR}/${STRIPPED_BINARIES_OUTPUT} BINARY_PATH clickhouse)
     else()
+        clickhouse_make_empty_debug_info_for_nfpm(TARGET clickhouse DESTINATION_DIR ${CMAKE_CURRENT_BINARY_DIR}/${STRIPPED_BINARIES_OUTPUT})
         install (TARGETS clickhouse RUNTIME DESTINATION ${CMAKE_INSTALL_BINDIR} COMPONENT clickhouse)
     endif()
 endif()
 
-if (NOT INSTALL_STRIPPED_BINARIES)
-    # Install dunny debug directory
-    # TODO: move logic to every place where clickhouse_strip_binary is used
-    add_custom_command(TARGET clickhouse POST_BUILD COMMAND echo > .empty )
-    install(FILES "${CMAKE_CURRENT_BINARY_DIR}/.empty" DESTINATION ${CMAKE_INSTALL_LIBDIR}/debug/.empty)
-endif()
-
-
 if (ENABLE_TESTS)
     set (CLICKHOUSE_UNIT_TESTS_TARGETS unit_tests_dbms)
     add_custom_target (clickhouse-tests ALL DEPENDS ${CLICKHOUSE_UNIT_TESTS_TARGETS})

programs/keeper/CMakeLists.txt

@@ -71,17 +71,11 @@ if (BUILD_STANDALONE_KEEPER)
         ${CMAKE_CURRENT_SOURCE_DIR}/../../src/Compression/CompressedReadBuffer.cpp
         ${CMAKE_CURRENT_SOURCE_DIR}/../../src/Compression/CompressedReadBufferFromFile.cpp
         ${CMAKE_CURRENT_SOURCE_DIR}/../../src/Compression/CompressedWriteBuffer.cpp
-        ${CMAKE_CURRENT_SOURCE_DIR}/../../src/Compression/CompressionCodecDelta.cpp
-        ${CMAKE_CURRENT_SOURCE_DIR}/../../src/Compression/CompressionCodecDoubleDelta.cpp
-        ${CMAKE_CURRENT_SOURCE_DIR}/../../src/Compression/CompressionCodecEncrypted.cpp
-        ${CMAKE_CURRENT_SOURCE_DIR}/../../src/Compression/CompressionCodecGorilla.cpp
         ${CMAKE_CURRENT_SOURCE_DIR}/../../src/Compression/CompressionCodecLZ4.cpp
         ${CMAKE_CURRENT_SOURCE_DIR}/../../src/Compression/CompressionCodecMultiple.cpp
         ${CMAKE_CURRENT_SOURCE_DIR}/../../src/Compression/CompressionCodecNone.cpp
-        ${CMAKE_CURRENT_SOURCE_DIR}/../../src/Compression/CompressionCodecT64.cpp
         ${CMAKE_CURRENT_SOURCE_DIR}/../../src/Compression/CompressionCodecZSTD.cpp
         ${CMAKE_CURRENT_SOURCE_DIR}/../../src/Compression/CompressionFactory.cpp
-        ${CMAKE_CURRENT_SOURCE_DIR}/../../src/Compression/getCompressionCodecForFile.cpp
         ${CMAKE_CURRENT_SOURCE_DIR}/../../src/Compression/ICompressionCodec.cpp
         ${CMAKE_CURRENT_SOURCE_DIR}/../../src/Compression/LZ4_decompress_faster.cpp
 
@@ -137,5 +131,10 @@ if (BUILD_STANDALONE_KEEPER)
     add_dependencies(clickhouse-keeper clickhouse_keeper_configs)
     set_target_properties(clickhouse-keeper PROPERTIES RUNTIME_OUTPUT_DIRECTORY ../)
 
-    install(TARGETS clickhouse-keeper RUNTIME DESTINATION ${CMAKE_INSTALL_BINDIR} COMPONENT clickhouse)
+    if (INSTALL_STRIPPED_BINARIES)
+        clickhouse_strip_binary(TARGET clickhouse-keeper DESTINATION_DIR ${CMAKE_CURRENT_BINARY_DIR}/../${STRIPPED_BINARIES_OUTPUT} BINARY_PATH ../clickhouse-keeper)
+    else()
+        clickhouse_make_empty_debug_info_for_nfpm(TARGET clickhouse-keeper DESTINATION_DIR ${CMAKE_CURRENT_BINARY_DIR}/../${STRIPPED_BINARIES_OUTPUT})
+        install(TARGETS clickhouse-keeper RUNTIME DESTINATION ${CMAKE_INSTALL_BINDIR} COMPONENT clickhouse)
+    endif()
 endif()

programs/library-bridge/CMakeLists.txt

@@ -27,5 +27,6 @@ set_target_properties(clickhouse-library-bridge PROPERTIES RUNTIME_OUTPUT_DIRECT
 if (INSTALL_STRIPPED_BINARIES)
     clickhouse_strip_binary(TARGET clickhouse-library-bridge DESTINATION_DIR ${CMAKE_CURRENT_BINARY_DIR}/../${STRIPPED_BINARIES_OUTPUT} BINARY_PATH ../clickhouse-library-bridge)
 else()
+    clickhouse_make_empty_debug_info_for_nfpm(TARGET clickhouse-library-bridge DESTINATION_DIR ${CMAKE_CURRENT_BINARY_DIR}/../${STRIPPED_BINARIES_OUTPUT})
     install(TARGETS clickhouse-library-bridge RUNTIME DESTINATION ${CMAKE_INSTALL_BINDIR} COMPONENT clickhouse)
 endif()

programs/odbc-bridge/CMakeLists.txt

@@ -42,6 +42,7 @@ endif()
 if (INSTALL_STRIPPED_BINARIES)
     clickhouse_strip_binary(TARGET clickhouse-odbc-bridge DESTINATION_DIR ${CMAKE_CURRENT_BINARY_DIR}/../${STRIPPED_BINARIES_OUTPUT} BINARY_PATH ../clickhouse-odbc-bridge)
 else()
+    clickhouse_make_empty_debug_info_for_nfpm(TARGET clickhouse-odbc-bridge DESTINATION_DIR ${CMAKE_CURRENT_BINARY_DIR}/../${STRIPPED_BINARIES_OUTPUT})
     install(TARGETS clickhouse-odbc-bridge RUNTIME DESTINATION ${CMAKE_INSTALL_BINDIR} COMPONENT clickhouse)
 endif()
 

programs/server/Server.cpp

@@ -966,6 +966,36 @@ int Server::main(const std::vector<std::string> & /*args*/)
         fs::create_directories(path / "metadata_dropped/");
     }
 
+#if USE_ROCKSDB
+    /// Initialize merge tree metadata cache
+    if (config().has("merge_tree_metadata_cache"))
+    {
+        fs::create_directories(path / "rocksdb/");
+        size_t size = config().getUInt64("merge_tree_metadata_cache.lru_cache_size", 256 << 20);
+        bool continue_if_corrupted = config().getBool("merge_tree_metadata_cache.continue_if_corrupted", false);
+        try
+        {
+            LOG_DEBUG(
+                log, "Initiailizing merge tree metadata cache lru_cache_size:{} continue_if_corrupted:{}", size, continue_if_corrupted);
+            global_context->initializeMergeTreeMetadataCache(path_str + "/" + "rocksdb", size);
+        }
+        catch (...)
+        {
+            if (continue_if_corrupted)
+            {
+                /// Rename rocksdb directory and reinitialize merge tree metadata cache
+                time_t now = time(nullptr);
+                fs::rename(path / "rocksdb", path / ("rocksdb.old." + std::to_string(now)));
+                global_context->initializeMergeTreeMetadataCache(path_str + "/" + "rocksdb", size);
+            }
+            else
+            {
+                throw;
+            }
+        }
+    }
+#endif
+
     if (config().has("interserver_http_port") && config().has("interserver_https_port"))
         throw Exception("Both http and https interserver ports are specified", ErrorCodes::EXCESSIVE_ELEMENT_IN_CONFIG);
 

programs/server/config.xml

@@ -1294,4 +1294,10 @@
         </tables>
     </rocksdb>
     -->
+
+    <!-- Uncomment if enable merge tree metadata cache -->
+    <merge_tree_metadata_cache>
+        <lru_cache_size>268435456</lru_cache_size>
+        <continue_if_corrupted>true</continue_if_corrupted>
+    </merge_tree_metadata_cache>
 </clickhouse>

src/Access/Common/QuotaDefs.h

@@ -13,7 +13,7 @@ enum class QuotaType
 {
     QUERIES,        /// Number of queries.
     QUERY_SELECTS,  /// Number of select queries.
-    QUERY_INSERTS,  /// Number of inserts queries.
+    QUERY_INSERTS,  /// Number of insert queries.
     ERRORS,         /// Number of queries with exceptions.
     RESULT_ROWS,    /// Number of rows returned as result.
     RESULT_BYTES,   /// Number of bytes returned as result.

src/AggregateFunctions/AggregateFunctionSumMap.cpp

@@ -67,7 +67,7 @@ auto parseArguments(const std::string & name, const DataTypes & arguments)
         values_types.push_back(array_type->getNestedType());
     }
 
-    return std::tuple{std::move(keys_type), std::move(values_types), tuple_argument};
+    return std::tuple<DataTypePtr, DataTypes, bool>{std::move(keys_type), std::move(values_types), tuple_argument};
 }
 
 // This function instantiates a particular overload of the sumMap family of

src/CMakeLists.txt

@@ -494,6 +494,11 @@ endif()
 
 target_link_libraries (clickhouse_common_io PUBLIC ch_contrib::fast_float)
 
+if (USE_ORC)
+    dbms_target_link_libraries(PUBLIC ${ORC_LIBRARIES})
+    dbms_target_include_directories(SYSTEM BEFORE PUBLIC ${ORC_INCLUDE_DIR} "${CMAKE_BINARY_DIR}/contrib/orc/c++/include")
+endif ()
+
 if (TARGET ch_contrib::rocksdb)
     dbms_target_link_libraries(PUBLIC ch_contrib::rocksdb)
 endif()
@@ -573,10 +578,6 @@ if (ENABLE_TESTS)
         target_link_libraries(unit_tests_dbms PRIVATE ch_contrib::simdjson)
     endif()
 
-    if(TARGET ch_contrib::rapidjson)
-        target_include_directories(unit_tests_dbms PRIVATE ch_contrib::rapidjson)
-    endif()
-
     if (TARGET ch_contrib::yaml_cpp)
         target_link_libraries(unit_tests_dbms PRIVATE ch_contrib::yaml_cpp)
     endif()

src/Client/ConnectionPoolWithFailover.cpp

@@ -29,15 +29,15 @@ ConnectionPoolWithFailover::ConnectionPoolWithFailover(
         time_t decrease_error_period_,
         size_t max_error_cap_)
     : Base(std::move(nested_pools_), decrease_error_period_, max_error_cap_, &Poco::Logger::get("ConnectionPoolWithFailover"))
-    , default_load_balancing(load_balancing)
+    , get_priority_load_balancing(load_balancing)
 {
     const std::string & local_hostname = getFQDNOrHostName();
 
-    hostname_differences.resize(nested_pools.size());
+    get_priority_load_balancing.hostname_differences.resize(nested_pools.size());
     for (size_t i = 0; i < nested_pools.size(); ++i)
     {
         ConnectionPool & connection_pool = dynamic_cast<ConnectionPool &>(*nested_pools[i]);
-        hostname_differences[i] = getHostNameDifference(local_hostname, connection_pool.getHost());
+        get_priority_load_balancing.hostname_differences[i] = getHostNameDifference(local_hostname, connection_pool.getHost());
     }
 }
 
@@ -51,36 +51,15 @@ IConnectionPool::Entry ConnectionPoolWithFailover::get(const ConnectionTimeouts
     };
 
     size_t offset = 0;
+    LoadBalancing load_balancing = get_priority_load_balancing.load_balancing;
     if (settings)
-        offset = settings->load_balancing_first_offset % nested_pools.size();
-    GetPriorityFunc get_priority;
-    switch (settings ? LoadBalancing(settings->load_balancing) : default_load_balancing)
     {
-    case LoadBalancing::NEAREST_HOSTNAME:
+        offset = settings->load_balancing_first_offset % nested_pools.size();
-        get_priority = [&](size_t i) { return hostname_differences[i]; };
+        load_balancing = LoadBalancing(settings->load_balancing);
-        break;
-    case LoadBalancing::IN_ORDER:
-        get_priority = [](size_t i) { return i; };
-        break;
-    case LoadBalancing::RANDOM:
-        break;
-    case LoadBalancing::FIRST_OR_RANDOM:
-        get_priority = [offset](size_t i) -> size_t { return i != offset; };
-        break;
-    case LoadBalancing::ROUND_ROBIN:
-        if (last_used >= nested_pools.size())
-            last_used = 0;
-        ++last_used;
-        /* Consider nested_pools.size() equals to 5
-         * last_used = 1 -> get_priority: 0 1 2 3 4
-         * last_used = 2 -> get_priority: 4 0 1 2 3
-         * last_used = 3 -> get_priority: 4 3 0 1 2
-         * ...
-         * */
-        get_priority = [&](size_t i) { ++i; return i < last_used ? nested_pools.size() - i : i - last_used; };
-        break;
     }
 
+    GetPriorityFunc get_priority = get_priority_load_balancing.getPriorityFunc(load_balancing, offset, nested_pools.size());
+
     UInt64 max_ignored_errors = settings ? settings->distributed_replica_max_ignored_errors.value : 0;
     bool fallback_to_stale_replicas = settings ? settings->fallback_to_stale_replicas_for_distributed_queries.value : true;
 
@@ -173,38 +152,14 @@ std::vector<ConnectionPoolWithFailover::TryResult> ConnectionPoolWithFailover::g
 ConnectionPoolWithFailover::Base::GetPriorityFunc ConnectionPoolWithFailover::makeGetPriorityFunc(const Settings * settings)
 {
     size_t offset = 0;
+    LoadBalancing load_balancing = get_priority_load_balancing.load_balancing;
     if (settings)
-        offset = settings->load_balancing_first_offset % nested_pools.size();
-
-    GetPriorityFunc get_priority;
-    switch (settings ? LoadBalancing(settings->load_balancing) : default_load_balancing)
     {
-        case LoadBalancing::NEAREST_HOSTNAME:
+        offset = settings->load_balancing_first_offset % nested_pools.size();
-            get_priority = [&](size_t i) { return hostname_differences[i]; };
+        load_balancing = LoadBalancing(settings->load_balancing);
-            break;
-        case LoadBalancing::IN_ORDER:
-            get_priority = [](size_t i) { return i; };
-            break;
-        case LoadBalancing::RANDOM:
-            break;
-        case LoadBalancing::FIRST_OR_RANDOM:
-            get_priority = [offset](size_t i) -> size_t { return i != offset; };
-            break;
-        case LoadBalancing::ROUND_ROBIN:
-            if (last_used >= nested_pools.size())
-                last_used = 0;
-            ++last_used;
-            /* Consider nested_pools.size() equals to 5
-             * last_used = 1 -> get_priority: 0 1 2 3 4
-             * last_used = 2 -> get_priority: 5 0 1 2 3
-             * last_used = 3 -> get_priority: 5 4 0 1 2
-             * ...
-             * */
-            get_priority = [&](size_t i) { ++i; return i < last_used ? nested_pools.size() - i : i - last_used; };
-            break;
     }
 
-    return get_priority;
+    return get_priority_load_balancing.getPriorityFunc(load_balancing, offset, nested_pools.size());
 }
 
 std::vector<ConnectionPoolWithFailover::TryResult> ConnectionPoolWithFailover::getManyImpl(

src/Client/ConnectionPoolWithFailover.h

@@ -1,6 +1,7 @@
 #pragma once
 
 #include <Common/PoolWithFailoverBase.h>
+#include <Common/GetPriorityForLoadBalancing.h>
 #include <Client/ConnectionPool.h>
 
 #include <chrono>
@@ -109,9 +110,7 @@ private:
 
     GetPriorityFunc makeGetPriorityFunc(const Settings * settings);
 
-    std::vector<size_t> hostname_differences; /// Distances from name of this host to the names of hosts of pools.
+    GetPriorityForLoadBalancing get_priority_load_balancing;
-    size_t last_used = 0; /// Last used for round_robin policy.
-    LoadBalancing default_load_balancing;
 };
 
 using ConnectionPoolWithFailoverPtr = std::shared_ptr<ConnectionPoolWithFailover>;

src/Columns/ColumnObject.cpp

@@ -521,7 +521,7 @@ ColumnObject::ColumnObject(bool is_nullable_)
 {
 }
 
-ColumnObject::ColumnObject(SubcolumnsTree && subcolumns_, bool is_nullable_)
+ColumnObject::ColumnObject(Subcolumns && subcolumns_, bool is_nullable_)
     : is_nullable(is_nullable_)
     , subcolumns(std::move(subcolumns_))
     , num_rows(subcolumns.empty() ? 0 : (*subcolumns.begin())->data.size())
@@ -696,7 +696,7 @@ const ColumnObject::Subcolumn & ColumnObject::getSubcolumn(const PathInData & ke
 ColumnObject::Subcolumn & ColumnObject::getSubcolumn(const PathInData & key)
 {
     if (const auto * node = subcolumns.findLeaf(key))
-        return const_cast<SubcolumnsTree::Node *>(node)->data;
+        return const_cast<Subcolumns::Node *>(node)->data;
 
     throw Exception(ErrorCodes::ILLEGAL_COLUMN, "There is no subcolumn {} in ColumnObject", key.getPath());
 }
@@ -794,7 +794,7 @@ bool ColumnObject::isFinalized() const
 void ColumnObject::finalize()
 {
     size_t old_size = size();
-    SubcolumnsTree new_subcolumns;
+    Subcolumns new_subcolumns;
     for (auto && entry : subcolumns)
     {
         const auto & least_common_type = entry->data.getLeastCommonType();

src/Columns/ColumnObject.h

@@ -138,20 +138,20 @@ public:
         size_t num_of_defaults_in_prefix = 0;
     };
 
-    using SubcolumnsTree = SubcolumnsTree<Subcolumn>;
+    using Subcolumns = SubcolumnsTree<Subcolumn>;
 
 private:
     /// If true then all subcolumns are nullable.
     const bool is_nullable;
 
-    SubcolumnsTree subcolumns;
+    Subcolumns subcolumns;
     size_t num_rows;
 
 public:
     static constexpr auto COLUMN_NAME_DUMMY = "_dummy";
 
     explicit ColumnObject(bool is_nullable_);
-    ColumnObject(SubcolumnsTree && subcolumns_, bool is_nullable_);
+    ColumnObject(Subcolumns && subcolumns_, bool is_nullable_);
 
     /// Checks that all subcolumns have consistent sizes.
     void checkConsistency() const;
@@ -173,8 +173,8 @@ public:
     /// It cares about consistency of sizes of Nested arrays.
     void addNestedSubcolumn(const PathInData & key, const FieldInfo & field_info, size_t new_size);
 
-    const SubcolumnsTree & getSubcolumns() const { return subcolumns; }
+    const Subcolumns & getSubcolumns() const { return subcolumns; }
-    SubcolumnsTree & getSubcolumns() { return subcolumns; }
+    Subcolumns & getSubcolumns() { return subcolumns; }
     PathsInData getKeys() const;
 
     /// Finalizes all subcolumns.

src/Columns/MaskOperations.cpp

@@ -83,11 +83,20 @@ size_t extractMaskNumericImpl(
     const PaddedPODArray<UInt8> * null_bytemap,
     PaddedPODArray<UInt8> * nulls)
 {
+    if constexpr (!column_is_short)
+    {
+        if (data.size() != mask.size())
+            throw Exception(ErrorCodes::LOGICAL_ERROR, "The size of a full data column is not equal to the size of a mask");
+    }
+
     size_t ones_count = 0;
     size_t data_index = 0;
-    size_t mask_size = mask.size();
 
-    for (size_t i = 0; i != mask_size; ++i)
+    size_t mask_size = mask.size();
+    size_t data_size = data.size();
+
+    size_t i = 0;
+    for (; i != mask_size && data_index != data_size; ++i)
     {
         // Change mask only where value is 1.
         if (!mask[i])
@@ -120,6 +129,13 @@ size_t extractMaskNumericImpl(
 
         mask[i] = value;
     }
+
+    if constexpr (column_is_short)
+    {
+        if (data_index != data_size)
+            throw Exception(ErrorCodes::LOGICAL_ERROR, "The size of a short column is not equal to the number of ones in a mask");
+    }
+
     return ones_count;
 }
 

src/Common/FiberStack.h

@@ -31,8 +31,8 @@ public:
     /// probably it worth to try to increase stack size for coroutines.
     ///
     /// Current value is just enough for all tests in our CI. It's not selected in some special
-    /// way. We will have 40 pages with 4KB page size.
+    /// way. We will have 80 pages with 4KB page size.
-    static constexpr size_t default_stack_size = 192 * 1024; /// 64KB was not enough for tests
+    static constexpr size_t default_stack_size = 320 * 1024; /// 64KB was not enough for tests
 
     explicit FiberStack(size_t stack_size_ = default_stack_size) : stack_size(stack_size_)
     {

src/Common/FileSegment.cpp

@@ -437,6 +437,7 @@ String FileSegment::stateToString(FileSegment::State state)
         case FileSegment::State::SKIP_CACHE:
             return "SKIP_CACHE";
     }
+    __builtin_unreachable();
 }
 
 String FileSegmentsHolder::toString()

src/Common/GetPriorityForLoadBalancing.cpp

@@ -0,0 +1,49 @@
+#include <Common/GetPriorityForLoadBalancing.h>
+
+namespace DB
+{
+
+namespace ErrorCodes
+{
+    extern const int LOGICAL_ERROR;
+}
+
+std::function<size_t(size_t index)> GetPriorityForLoadBalancing::getPriorityFunc(LoadBalancing load_balance, size_t offset, size_t pool_size) const
+{
+    std::function<size_t(size_t index)> get_priority;
+    switch (load_balance)
+    {
+        case LoadBalancing::NEAREST_HOSTNAME:
+            if (hostname_differences.empty())
+                throw Exception(ErrorCodes::LOGICAL_ERROR, "It's a bug: hostname_differences is not initialized");
+            get_priority = [&](size_t i) { return hostname_differences[i]; };
+            break;
+        case LoadBalancing::IN_ORDER:
+            get_priority = [](size_t i) { return i; };
+            break;
+        case LoadBalancing::RANDOM:
+            break;
+        case LoadBalancing::FIRST_OR_RANDOM:
+            get_priority = [offset](size_t i) -> size_t { return i != offset; };
+            break;
+        case LoadBalancing::ROUND_ROBIN:
+            if (last_used >= pool_size)
+                last_used = 0;
+            ++last_used;
+            /* Consider pool_size equals to 5
+             * last_used = 1 -> get_priority: 0 1 2 3 4
+             * last_used = 2 -> get_priority: 4 0 1 2 3
+             * last_used = 3 -> get_priority: 4 3 0 1 2
+             * ...
+             * */
+            get_priority = [&](size_t i)
+            {
+                ++i;
+                return i < last_used ? pool_size - i : i - last_used;
+            };
+            break;
+    }
+    return get_priority;
+}
+
+}

src/Common/GetPriorityForLoadBalancing.h

@@ -0,0 +1,34 @@
+#pragma once
+
+#include <Core/SettingsEnums.h>
+
+namespace DB
+{
+
+class GetPriorityForLoadBalancing
+{
+public:
+    GetPriorityForLoadBalancing(LoadBalancing load_balancing_) : load_balancing(load_balancing_) {}
+    GetPriorityForLoadBalancing(){}
+
+    bool operator == (const GetPriorityForLoadBalancing & other) const
+    {
+        return load_balancing == other.load_balancing && hostname_differences == other.hostname_differences;
+    }
+
+    bool operator != (const GetPriorityForLoadBalancing & other) const
+    {
+        return !(*this == other);
+    }
+
+    std::function<size_t(size_t index)> getPriorityFunc(LoadBalancing load_balance, size_t offset, size_t pool_size) const;
+
+    std::vector<size_t> hostname_differences; /// Distances from name of this host to the names of hosts of pools.
+
+    LoadBalancing load_balancing = LoadBalancing::RANDOM;
+
+private:
+    mutable size_t last_used = 0; /// Last used for round_robin policy.
+};
+
+}

src/Common/OvercommitTracker.cpp

@@ -23,6 +23,12 @@ void OvercommitTracker::setMaxWaitTime(UInt64 wait_time)
 
 bool OvercommitTracker::needToStopQuery(MemoryTracker * tracker)
 {
+    // NOTE: Do not change the order of locks
+    //
+    // global_mutex must be acquired before overcommit_m, because
+    // method OvercommitTracker::unsubscribe(MemoryTracker *) is
+    // always called with already acquired global_mutex in
+    // ProcessListEntry::~ProcessListEntry().
     std::unique_lock<std::mutex> global_lock(global_mutex);
     std::unique_lock<std::mutex> lk(overcommit_m);
 
@@ -76,7 +82,7 @@ void UserOvercommitTracker::pickQueryToExcludeImpl()
     MemoryTracker * query_tracker = nullptr;
     OvercommitRatio current_ratio{0, 0};
     // At this moment query list must be read only.
-    // BlockQueryIfMemoryLimit is used in ProcessList to guarantee this.
+    // This is guaranteed by locking global_mutex in OvercommitTracker::needToStopQuery.
     auto & queries = user_process_list->queries;
     LOG_DEBUG(logger, "Trying to choose query to stop from {} queries", queries.size());
     for (auto const & query : queries)
@@ -111,9 +117,9 @@ void GlobalOvercommitTracker::pickQueryToExcludeImpl()
     MemoryTracker * query_tracker = nullptr;
     OvercommitRatio current_ratio{0, 0};
     // At this moment query list must be read only.
-    // BlockQueryIfMemoryLimit is used in ProcessList to guarantee this.
+    // This is guaranteed by locking global_mutex in OvercommitTracker::needToStopQuery.
-    LOG_DEBUG(logger, "Trying to choose query to stop");
+    LOG_DEBUG(logger, "Trying to choose query to stop from {} queries", process_list->size());
-    process_list->processEachQueryStatus([&](DB::QueryStatus const & query)
+    for (auto const & query : process_list->processes)
     {
         if (query.isKilled())
             return;
@@ -134,7 +140,7 @@ void GlobalOvercommitTracker::pickQueryToExcludeImpl()
             query_tracker = memory_tracker;
             current_ratio   = ratio;
         }
-    });
+    }
     LOG_DEBUG(logger, "Selected to stop query with overcommit ratio {}/{}",
         current_ratio.committed, current_ratio.soft_limit);
     picked_tracker = query_tracker;

src/Common/OvercommitTracker.h

@@ -43,8 +43,6 @@ class MemoryTracker;
 // is killed to free memory.
 struct OvercommitTracker : boost::noncopyable
 {
-    explicit OvercommitTracker(std::mutex & global_mutex_);
-
     void setMaxWaitTime(UInt64 wait_time);
 
     bool needToStopQuery(MemoryTracker * tracker);
@@ -54,8 +52,12 @@ struct OvercommitTracker : boost::noncopyable
     virtual ~OvercommitTracker() = default;
 
 protected:
+    explicit OvercommitTracker(std::mutex & global_mutex_);
+
     virtual void pickQueryToExcludeImpl() = 0;
 
+    // This mutex is used to disallow concurrent access
+    // to picked_tracker and cancelation_state variables.
     mutable std::mutex overcommit_m;
     mutable std::condition_variable cv;
 
@@ -87,6 +89,11 @@ private:
         }
     }
 
+    // Global mutex which is used in ProcessList to synchronize
+    // insertion and deletion of queries.
+    // OvercommitTracker::pickQueryToExcludeImpl() implementations
+    // require this mutex to be locked, because they read list (or sublist)
+    // of queries.
     std::mutex & global_mutex;
 };
 

src/Common/ProfileEvents.cpp

@@ -9,6 +9,7 @@
     M(SelectQuery, "Same as Query, but only for SELECT queries.") \
     M(InsertQuery, "Same as Query, but only for INSERT queries.") \
     M(AsyncInsertQuery, "Same as InsertQuery, but only for asynchronous INSERT queries.") \
+    M(AsyncInsertBytes, "Data size in bytes of asynchronous INSERT queries.") \
     M(FailedQuery, "Number of failed queries.") \
     M(FailedSelectQuery, "Same as FailedQuery, but only for SELECT queries.") \
     M(FailedInsertQuery, "Same as FailedQuery, but only for INSERT queries.") \
@@ -284,6 +285,13 @@
     \
     M(MainConfigLoads, "Number of times the main configuration was reloaded.") \
     \
+    M(MergeTreeMetadataCacheGet, "Number of rocksdb reads(used for merge tree metadata cache)") \
+    M(MergeTreeMetadataCachePut, "Number of rocksdb puts(used for merge tree metadata cache)") \
+    M(MergeTreeMetadataCacheDelete, "Number of rocksdb deletes(used for merge tree metadata cache)") \
+    M(MergeTreeMetadataCacheSeek, "Number of rocksdb seeks(used for merge tree metadata cache)") \
+    M(MergeTreeMetadataCacheHit, "Number of times the read of meta file was done from MergeTree metadata cache") \
+    M(MergeTreeMetadataCacheMiss, "Number of times the read of meta file was not done from MergeTree metadata cache") \
+    \
     M(ScalarSubqueriesGlobalCacheHit, "Number of times a read from a scalar subquery was done using the global cache") \
     M(ScalarSubqueriesLocalCacheHit, "Number of times a read from a scalar subquery was done using the local cache") \
     M(ScalarSubqueriesCacheMiss, "Number of times a read from a scalar subquery was not cached and had to be calculated completely")

src/Common/ZooKeeper/CMakeLists.txt

@@ -22,7 +22,6 @@ target_link_libraries (clickhouse_common_zookeeper_no_log
     PRIVATE
         string_utils
 )
-
 if (ENABLE_EXAMPLES)
     add_subdirectory(examples)
 endif()

src/Common/ZooKeeper/ZooKeeper.cpp

@@ -5,15 +5,15 @@
 
 #include <functional>
 #include <filesystem>
-#include <pcg-random/pcg_random.hpp>
 
-#include <base/logger_useful.h>
 #include <base/find_symbols.h>
-#include <Common/randomSeed.h>
+#include <base/getFQDNOrHostName.h>
 #include <Common/StringUtils/StringUtils.h>
 #include <Common/Exception.h>
+#include <Common/isLocalAddress.h>
 
 #include <Poco/Net/NetException.h>
+#include <Poco/Net/DNS.h>
 
 
 #define ZOOKEEPER_CONNECTION_TIMEOUT_MS 1000
@@ -48,7 +48,7 @@ static void check(Coordination::Error code, const std::string & path)
 
 
 void ZooKeeper::init(const std::string & implementation_, const Strings & hosts_, const std::string & identity_,
-                     int32_t session_timeout_ms_, int32_t operation_timeout_ms_, const std::string & chroot_)
+                     int32_t session_timeout_ms_, int32_t operation_timeout_ms_, const std::string & chroot_, const GetPriorityForLoadBalancing & get_priority_load_balancing_)
 {
     log = &Poco::Logger::get("ZooKeeper");
     hosts = hosts_;
@@ -57,6 +57,7 @@ void ZooKeeper::init(const std::string & implementation_, const Strings & hosts_
     operation_timeout_ms = operation_timeout_ms_;
     chroot = chroot_;
     implementation = implementation_;
+    get_priority_load_balancing = get_priority_load_balancing_;
 
     if (implementation == "zookeeper")
     {
@@ -66,14 +67,13 @@ void ZooKeeper::init(const std::string & implementation_, const Strings & hosts_
         Coordination::ZooKeeper::Nodes nodes;
         nodes.reserve(hosts.size());
 
-        Strings shuffled_hosts = hosts;
         /// Shuffle the hosts to distribute the load among ZooKeeper nodes.
-        pcg64 generator(randomSeed());
+        std::vector<ShuffleHost> shuffled_hosts = shuffleHosts();
-        std::shuffle(shuffled_hosts.begin(), shuffled_hosts.end(), generator);
 
         bool dns_error = false;
-        for (auto & host_string : shuffled_hosts)
+        for (auto & host : shuffled_hosts)
         {
+            auto & host_string = host.host;
             try
             {
                 bool secure = bool(startsWith(host_string, "secure://"));
@@ -81,6 +81,7 @@ void ZooKeeper::init(const std::string & implementation_, const Strings & hosts_
                 if (secure)
                     host_string.erase(0, strlen("secure://"));
 
+                LOG_TEST(log, "Adding ZooKeeper host {} ({})", host_string, Poco::Net::SocketAddress{host_string}.toString());
                 nodes.emplace_back(Coordination::ZooKeeper::Node{Poco::Net::SocketAddress{host_string}, secure});
             }
             catch (const Poco::Net::HostNotFoundException & e)
@@ -154,23 +155,47 @@ void ZooKeeper::init(const std::string & implementation_, const Strings & hosts_
     }
 }
 
+std::vector<ShuffleHost> ZooKeeper::shuffleHosts() const
+{
+    std::function<size_t(size_t index)> get_priority = get_priority_load_balancing.getPriorityFunc(get_priority_load_balancing.load_balancing, 0, hosts.size());
+    std::vector<ShuffleHost> shuffle_hosts;
+    for (size_t i = 0; i < hosts.size(); ++i)
+    {
+        ShuffleHost shuffle_host;
+        shuffle_host.host = hosts[i];
+        if (get_priority)
+            shuffle_host.priority = get_priority(i);
+        shuffle_host.randomize();
+        shuffle_hosts.emplace_back(shuffle_host);
+    }
+
+    std::sort(
+        shuffle_hosts.begin(), shuffle_hosts.end(),
+        [](const ShuffleHost & lhs, const ShuffleHost & rhs)
+        {
+            return ShuffleHost::compare(lhs, rhs);
+        });
+
+    return shuffle_hosts;
+}
+
 ZooKeeper::ZooKeeper(const std::string & hosts_string, const std::string & identity_, int32_t session_timeout_ms_,
                      int32_t operation_timeout_ms_, const std::string & chroot_, const std::string & implementation_,
-                     std::shared_ptr<DB::ZooKeeperLog> zk_log_)
+                     std::shared_ptr<DB::ZooKeeperLog> zk_log_, const GetPriorityForLoadBalancing & get_priority_load_balancing_)
 {
     zk_log = std::move(zk_log_);
     Strings hosts_strings;
     splitInto<','>(hosts_strings, hosts_string);
 
-    init(implementation_, hosts_strings, identity_, session_timeout_ms_, operation_timeout_ms_, chroot_);
+    init(implementation_, hosts_strings, identity_, session_timeout_ms_, operation_timeout_ms_, chroot_, get_priority_load_balancing_);
 }
 
 ZooKeeper::ZooKeeper(const Strings & hosts_, const std::string & identity_, int32_t session_timeout_ms_,
                      int32_t operation_timeout_ms_, const std::string & chroot_, const std::string & implementation_,
-                     std::shared_ptr<DB::ZooKeeperLog> zk_log_)
+                     std::shared_ptr<DB::ZooKeeperLog> zk_log_, const GetPriorityForLoadBalancing & get_priority_load_balancing_)
 {
     zk_log = std::move(zk_log_);
-    init(implementation_, hosts_, identity_, session_timeout_ms_, operation_timeout_ms_, chroot_);
+    init(implementation_, hosts_, identity_, session_timeout_ms_, operation_timeout_ms_, chroot_, get_priority_load_balancing_);
 }
 
 struct ZooKeeperArgs
@@ -213,6 +238,15 @@ struct ZooKeeperArgs
             {
                 implementation = config.getString(config_name + "." + key);
             }
+            else if (key == "zookeeper_load_balancing")
+            {
+                String load_balancing_str = config.getString(config_name + "." + key);
+                /// Use magic_enum to avoid dependency from dbms (`SettingFieldLoadBalancingTraits::fromString(...)`)
+                auto load_balancing = magic_enum::enum_cast<DB::LoadBalancing>(Poco::toUpper(load_balancing_str));
+                if (!load_balancing)
+                    throw DB::Exception(DB::ErrorCodes::BAD_ARGUMENTS, "Unknown load balancing: {}", load_balancing_str);
+                get_priority_load_balancing.load_balancing = *load_balancing;
+            }
             else
                 throw KeeperException(std::string("Unknown key ") + key + " in config file", Coordination::Error::ZBADARGUMENTS);
         }
@@ -224,6 +258,15 @@ struct ZooKeeperArgs
             if (chroot.back() == '/')
                 chroot.pop_back();
         }
+
+        /// init get_priority_load_balancing
+        get_priority_load_balancing.hostname_differences.resize(hosts.size());
+        const String & local_hostname = getFQDNOrHostName();
+        for (size_t i = 0; i < hosts.size(); ++i)
+        {
+            const String & node_host = hosts[i].substr(0, hosts[i].find_last_of(':'));
+            get_priority_load_balancing.hostname_differences[i] = DB::getHostNameDifference(local_hostname, node_host);
+        }
     }
 
     Strings hosts;
@@ -232,13 +275,14 @@ struct ZooKeeperArgs
     int operation_timeout_ms;
     std::string chroot;
     std::string implementation;
+    GetPriorityForLoadBalancing get_priority_load_balancing;
 };
 
 ZooKeeper::ZooKeeper(const Poco::Util::AbstractConfiguration & config, const std::string & config_name, std::shared_ptr<DB::ZooKeeperLog> zk_log_)
     : zk_log(std::move(zk_log_))
 {
     ZooKeeperArgs args(config, config_name);
-    init(args.implementation, args.hosts, args.identity, args.session_timeout_ms, args.operation_timeout_ms, args.chroot);
+    init(args.implementation, args.hosts, args.identity, args.session_timeout_ms, args.operation_timeout_ms, args.chroot, args.get_priority_load_balancing);
 }
 
 bool ZooKeeper::configChanged(const Poco::Util::AbstractConfiguration & config, const std::string & config_name) const
@@ -249,8 +293,11 @@ bool ZooKeeper::configChanged(const Poco::Util::AbstractConfiguration & config,
     if (args.implementation == implementation && implementation == "testkeeper")
         return false;
 
-    return std::tie(args.implementation, args.hosts, args.identity, args.session_timeout_ms, args.operation_timeout_ms, args.chroot)
+    if (args.get_priority_load_balancing != get_priority_load_balancing)
-        != std::tie(implementation, hosts, identity, session_timeout_ms, operation_timeout_ms, chroot);
+        return true;
+
+    return std::tie(args.implementation, args.hosts, args.identity, args.session_timeout_ms, args.operation_timeout_ms, args.chroot, args.get_priority_load_balancing)
+        != std::tie(implementation, hosts, identity, session_timeout_ms, operation_timeout_ms, chroot, args.get_priority_load_balancing);
 }
 
 
@@ -757,7 +804,7 @@ bool ZooKeeper::waitForDisappear(const std::string & path, const WaitCondition &
 
 ZooKeeperPtr ZooKeeper::startNewSession() const
 {
-    return std::make_shared<ZooKeeper>(hosts, identity, session_timeout_ms, operation_timeout_ms, chroot, implementation, zk_log);
+    return std::make_shared<ZooKeeper>(hosts, identity, session_timeout_ms, operation_timeout_ms, chroot, implementation, zk_log, get_priority_load_balancing);
 }
 
 

src/Common/ZooKeeper/ZooKeeper.h

@@ -13,7 +13,10 @@
 #include <Common/Stopwatch.h>
 #include <Common/ZooKeeper/IKeeper.h>
 #include <Common/ZooKeeper/ZooKeeperConstants.h>
+#include <Common/GetPriorityForLoadBalancing.h>
+#include <Common/thread_local_rng.h>
 #include <unistd.h>
+#include <random>
 
 
 namespace ProfileEvents
@@ -37,6 +40,25 @@ namespace zkutil
 /// Preferred size of multi() command (in number of ops)
 constexpr size_t MULTI_BATCH_SIZE = 100;
 
+struct ShuffleHost
+{
+    String host;
+    Int64 priority = 0;
+    UInt32 random = 0;
+
+    void randomize()
+    {
+        random = thread_local_rng();
+    }
+
+    static bool compare(const ShuffleHost & lhs, const ShuffleHost & rhs)
+    {
+        return std::forward_as_tuple(lhs.priority, lhs.random)
+               < std::forward_as_tuple(rhs.priority, rhs.random);
+    }
+};
+
+using GetPriorityForLoadBalancing = DB::GetPriorityForLoadBalancing;
 
 /// ZooKeeper session. The interface is substantially different from the usual libzookeeper API.
 ///
@@ -58,14 +80,16 @@ public:
               int32_t operation_timeout_ms_ = Coordination::DEFAULT_OPERATION_TIMEOUT_MS,
               const std::string & chroot_ = "",
               const std::string & implementation_ = "zookeeper",
-              std::shared_ptr<DB::ZooKeeperLog> zk_log_ = nullptr);
+              std::shared_ptr<DB::ZooKeeperLog> zk_log_ = nullptr,
+              const GetPriorityForLoadBalancing & get_priority_load_balancing_ = {});
 
     explicit ZooKeeper(const Strings & hosts_, const std::string & identity_ = "",
               int32_t session_timeout_ms_ = Coordination::DEFAULT_SESSION_TIMEOUT_MS,
               int32_t operation_timeout_ms_ = Coordination::DEFAULT_OPERATION_TIMEOUT_MS,
               const std::string & chroot_ = "",
               const std::string & implementation_ = "zookeeper",
-              std::shared_ptr<DB::ZooKeeperLog> zk_log_ = nullptr);
+              std::shared_ptr<DB::ZooKeeperLog> zk_log_ = nullptr,
+              const GetPriorityForLoadBalancing & get_priority_load_balancing_ = {});
 
     /** Config of the form:
         <zookeeper>
@@ -91,6 +115,8 @@ public:
     */
     ZooKeeper(const Poco::Util::AbstractConfiguration & config, const std::string & config_name, std::shared_ptr<DB::ZooKeeperLog> zk_log_);
 
+    std::vector<ShuffleHost> shuffleHosts() const;
+
     /// Creates a new session with the same parameters. This method can be used for reconnecting
     /// after the session has expired.
     /// This object remains unchanged, and the new session is returned.
@@ -284,7 +310,7 @@ private:
     friend class EphemeralNodeHolder;
 
     void init(const std::string & implementation_, const Strings & hosts_, const std::string & identity_,
-              int32_t session_timeout_ms_, int32_t operation_timeout_ms_, const std::string & chroot_);
+              int32_t session_timeout_ms_, int32_t operation_timeout_ms_, const std::string & chroot_, const GetPriorityForLoadBalancing & get_priority_load_balancing_);
 
     /// The following methods don't any throw exceptions but return error codes.
     Coordination::Error createImpl(const std::string & path, const std::string & data, int32_t mode, std::string & path_created);
@@ -311,6 +337,8 @@ private:
     Poco::Logger * log = nullptr;
     std::shared_ptr<DB::ZooKeeperLog> zk_log;
 
+    GetPriorityForLoadBalancing get_priority_load_balancing;
+
     AtomicStopwatch session_uptime;
 };
 

src/Common/ZooKeeper/ZooKeeperImpl.cpp

@@ -451,7 +451,7 @@ void ZooKeeper::connect(
     }
     else
     {
-        LOG_TEST(log, "Connected to ZooKeeper at {} with session_id {}", socket.peerAddress().toString(), session_id);
+        LOG_TEST(log, "Connected to ZooKeeper at {} with session_id {}{}", socket.peerAddress().toString(), session_id, fail_reasons.str());
     }
 }
 

src/Common/formatIPv6.h

@@ -11,7 +11,7 @@
 constexpr size_t IPV4_BINARY_LENGTH = 4;
 constexpr size_t IPV6_BINARY_LENGTH = 16;
 constexpr size_t IPV4_MAX_TEXT_LENGTH = 15;     /// Does not count tail zero byte.
-constexpr size_t IPV6_MAX_TEXT_LENGTH = 39;
+constexpr size_t IPV6_MAX_TEXT_LENGTH = 45;     /// Does not count tail zero byte.
 
 namespace DB
 {

src/Common/isLocalAddress.cpp

@@ -124,6 +124,7 @@ bool isLocalAddress(const Poco::Net::SocketAddress & address, UInt16 clickhouse_
 
 size_t getHostNameDifference(const std::string & local_hostname, const std::string & host)
 {
+    /// FIXME should we replace it with Levenstein distance? (we already have it in NamePrompter)
     size_t hostname_difference = 0;
     for (size_t i = 0; i < std::min(local_hostname.length(), host.length()); ++i)
         if (local_hostname[i] != host[i])

src/Common/tests/gtest_global_context.cpp

@@ -1,7 +1,18 @@
 #include "gtest_global_context.h"
 
 const ContextHolder & getContext()
+{
+    return getMutableContext();
+}
+
+ContextHolder & getMutableContext()
 {
     static ContextHolder holder;
     return holder;
 }
+
+void destroyContext()
+{
+    auto & holder = getMutableContext();
+    return holder.destroy();
+}

src/Common/tests/gtest_global_context.h

@@ -16,6 +16,17 @@ struct ContextHolder
     }
 
     ContextHolder(ContextHolder &&) = default;
+
+    void destroy()
+    {
+        context->shutdown();
+        context.reset();
+        shared_context.reset();
+    }
 };
 
 const ContextHolder & getContext();
+
+ContextHolder & getMutableContext();
+
+void destroyContext();

src/Compression/CompressionFactory.cpp

@@ -165,25 +165,36 @@ void registerCodecNone(CompressionCodecFactory & factory);
 void registerCodecLZ4(CompressionCodecFactory & factory);
 void registerCodecLZ4HC(CompressionCodecFactory & factory);
 void registerCodecZSTD(CompressionCodecFactory & factory);
+void registerCodecMultiple(CompressionCodecFactory & factory);
+
+
+/// Keeper use only general-purpose codecs, so we don't need these special codecs
+/// in standalone build
+#ifndef KEEPER_STANDALONE_BUILD
+
 void registerCodecDelta(CompressionCodecFactory & factory);
 void registerCodecT64(CompressionCodecFactory & factory);
 void registerCodecDoubleDelta(CompressionCodecFactory & factory);
 void registerCodecGorilla(CompressionCodecFactory & factory);
 void registerCodecEncrypted(CompressionCodecFactory & factory);
-void registerCodecMultiple(CompressionCodecFactory & factory);
+
+#endif
 
 CompressionCodecFactory::CompressionCodecFactory()
 {
-    registerCodecLZ4(*this);
     registerCodecNone(*this);
+    registerCodecLZ4(*this);
     registerCodecZSTD(*this);
     registerCodecLZ4HC(*this);
+    registerCodecMultiple(*this);
+
+#ifndef KEEPER_STANDALONE_BUILD
     registerCodecDelta(*this);
     registerCodecT64(*this);
     registerCodecDoubleDelta(*this);
     registerCodecGorilla(*this);
     registerCodecEncrypted(*this);
-    registerCodecMultiple(*this);
+#endif
 
     default_codec = get("LZ4", {});
 }

src/Core/Block.cpp

@@ -13,6 +13,7 @@
 
 #include <iterator>
 #include <base/sort.h>
+#include <boost/algorithm/string.hpp>
 
 
 namespace DB
@@ -269,8 +270,18 @@ const ColumnWithTypeAndName & Block::safeGetByPosition(size_t position) const
 }
 
 
-const ColumnWithTypeAndName * Block::findByName(const std::string & name) const
+const ColumnWithTypeAndName * Block::findByName(const std::string & name, bool case_insensitive) const
 {
+    if (case_insensitive)
+    {
+        auto found = std::find_if(data.begin(), data.end(), [&](const auto & column) { return boost::iequals(column.name, name); });
+        if (found == data.end())
+        {
+            return nullptr;
+        }
+        return &*found;
+    }
+
     auto it = index_by_name.find(name);
     if (index_by_name.end() == it)
     {
@@ -280,19 +291,23 @@ const ColumnWithTypeAndName * Block::findByName(const std::string & name) const
 }
 
 
-const ColumnWithTypeAndName & Block::getByName(const std::string & name) const
+const ColumnWithTypeAndName & Block::getByName(const std::string & name, bool case_insensitive) const
 {
-    const auto * result = findByName(name);
+    const auto * result = findByName(name, case_insensitive);
     if (!result)
-        throw Exception("Not found column " + name + " in block. There are only columns: " + dumpNames()
+        throw Exception(
-            , ErrorCodes::NOT_FOUND_COLUMN_IN_BLOCK);
+            "Not found column " + name + " in block. There are only columns: " + dumpNames(), ErrorCodes::NOT_FOUND_COLUMN_IN_BLOCK);
 
     return *result;
 }
 
 
-bool Block::has(const std::string & name) const
+bool Block::has(const std::string & name, bool case_insensitive) const
 {
+    if (case_insensitive)
+        return std::find_if(data.begin(), data.end(), [&](const auto & column) { return boost::iequals(column.name, name); })
+            != data.end();
+
     return index_by_name.end() != index_by_name.find(name);
 }
 
@@ -301,8 +316,8 @@ size_t Block::getPositionByName(const std::string & name) const
 {
     auto it = index_by_name.find(name);
     if (index_by_name.end() == it)
-        throw Exception("Not found column " + name + " in block. There are only columns: " + dumpNames()
+        throw Exception(
-            , ErrorCodes::NOT_FOUND_COLUMN_IN_BLOCK);
+            "Not found column " + name + " in block. There are only columns: " + dumpNames(), ErrorCodes::NOT_FOUND_COLUMN_IN_BLOCK);
 
     return it->second;
 }

src/Core/Block.h

@@ -60,21 +60,21 @@ public:
     ColumnWithTypeAndName & safeGetByPosition(size_t position);
     const ColumnWithTypeAndName & safeGetByPosition(size_t position) const;
 
-    ColumnWithTypeAndName* findByName(const std::string & name)
+    ColumnWithTypeAndName* findByName(const std::string & name, bool case_insensitive = false)
     {
         return const_cast<ColumnWithTypeAndName *>(
-            const_cast<const Block *>(this)->findByName(name));
+            const_cast<const Block *>(this)->findByName(name, case_insensitive));
     }
 
-    const ColumnWithTypeAndName * findByName(const std::string & name) const;
+    const ColumnWithTypeAndName * findByName(const std::string & name, bool case_insensitive = false) const;
 
-    ColumnWithTypeAndName & getByName(const std::string & name)
+    ColumnWithTypeAndName & getByName(const std::string & name, bool case_insensitive = false)
     {
         return const_cast<ColumnWithTypeAndName &>(
-            const_cast<const Block *>(this)->getByName(name));
+            const_cast<const Block *>(this)->getByName(name, case_insensitive));
     }
 
-    const ColumnWithTypeAndName & getByName(const std::string & name) const;
+    const ColumnWithTypeAndName & getByName(const std::string & name, bool case_insensitive = false) const;
 
     Container::iterator begin() { return data.begin(); }
     Container::iterator end() { return data.end(); }
@@ -83,7 +83,7 @@ public:
     Container::const_iterator cbegin() const { return data.cbegin(); }
     Container::const_iterator cend() const { return data.cend(); }
 
-    bool has(const std::string & name) const;
+    bool has(const std::string & name, bool case_insensitive = false) const;
 
     size_t getPositionByName(const std::string & name) const;
 

src/Core/Settings.h

@@ -616,11 +616,13 @@ class IColumn;
     M(Bool, input_format_tsv_empty_as_default, false, "Treat empty fields in TSV input as default values.", 0) \
     M(Bool, input_format_tsv_enum_as_number, false, "Treat inserted enum values in TSV formats as enum indices \\N", 0) \
     M(Bool, input_format_null_as_default, true, "For text input formats initialize null fields with default values if data type of this field is not nullable", 0) \
-    M(Bool, input_format_use_lowercase_column_name, false, "Use lowercase column name while reading input formats", 0) \
     M(Bool, input_format_arrow_import_nested, false, "Allow to insert array of structs into Nested table in Arrow input format.", 0) \
+    M(Bool, input_format_arrow_case_insensitive_column_matching, false, "Ignore case when matching Arrow columns with CH columns.", 0) \
     M(Bool, input_format_orc_import_nested, false, "Allow to insert array of structs into Nested table in ORC input format.", 0) \
     M(Int64, input_format_orc_row_batch_size, 100'000, "Batch size when reading ORC stripes.", 0) \
+    M(Bool, input_format_orc_case_insensitive_column_matching, false, "Ignore case when matching ORC columns with CH columns.", 0) \
     M(Bool, input_format_parquet_import_nested, false, "Allow to insert array of structs into Nested table in Parquet input format.", 0) \
+    M(Bool, input_format_parquet_case_insensitive_column_matching, false, "Ignore case when matching Parquet columns with CH columns.", 0) \
     M(Bool, input_format_allow_seeks, true, "Allow seeks while reading in ORC/Parquet/Arrow input formats", 0) \
     M(Bool, input_format_orc_allow_missing_columns, false, "Allow missing columns while reading ORC input formats", 0) \
     M(Bool, input_format_parquet_allow_missing_columns, false, "Allow missing columns while reading Parquet input formats", 0) \

src/Core/SettingsEnums.cpp

@@ -149,4 +149,5 @@ IMPLEMENT_SETTING_ENUM(MsgPackUUIDRepresentation , ErrorCodes::BAD_ARGUMENTS,
                         {"str", FormatSettings::MsgPackUUIDRepresentation::STR},
                         {"ext", FormatSettings::MsgPackUUIDRepresentation::EXT}})
 
+
 }

src/Core/SettingsEnums.h

@@ -187,5 +187,4 @@ DECLARE_SETTING_ENUM_WITH_RENAME(EnumComparingMode, FormatSettings::EnumComparin
 DECLARE_SETTING_ENUM_WITH_RENAME(EscapingRule, FormatSettings::EscapingRule)
 
 DECLARE_SETTING_ENUM_WITH_RENAME(MsgPackUUIDRepresentation, FormatSettings::MsgPackUUIDRepresentation)
-
 }

src/DataTypes/DataTypeFactory.cpp

@@ -33,7 +33,7 @@ DataTypePtr DataTypeFactory::get(const String & full_name) const
     /// Value 315 is known to cause stack overflow in some test configurations (debug build, sanitizers)
     /// let's make the threshold significantly lower.
     /// It is impractical for user to have complex data types with this depth.
-    static constexpr size_t data_type_max_parse_depth = 150;
+    static constexpr size_t data_type_max_parse_depth = 200;
 
     ParserDataType parser;
     ASTPtr ast = parseQuery(parser, full_name.data(), full_name.data() + full_name.size(), "data type", 0, data_type_max_parse_depth);

src/DataTypes/NestedUtils.cpp

@@ -15,6 +15,8 @@
 
 #include <Parsers/IAST.h>
 
+#include <boost/algorithm/string/case_conv.hpp>
+
 
 namespace DB
 {
@@ -227,14 +229,17 @@ void validateArraySizes(const Block & block)
 }
 
 
-std::unordered_set<String> getAllTableNames(const Block & block)
+std::unordered_set<String> getAllTableNames(const Block & block, bool to_lower_case)
 {
     std::unordered_set<String> nested_table_names;
-    for (auto & name : block.getNames())
+    for (const auto & name : block.getNames())
     {
         auto nested_table_name = Nested::extractTableName(name);
+        if (to_lower_case)
+            boost::to_lower(nested_table_name);
+
         if (!nested_table_name.empty())
-            nested_table_names.insert(nested_table_name);
+            nested_table_names.insert(std::move(nested_table_name));
     }
     return nested_table_names;
 }

src/DataTypes/NestedUtils.h

@@ -32,7 +32,7 @@ namespace Nested
     void validateArraySizes(const Block & block);
 
     /// Get all nested tables names from a block.
-    std::unordered_set<String> getAllTableNames(const Block & block);
+    std::unordered_set<String> getAllTableNames(const Block & block, bool to_lower_case = false);
 }
 
 }

src/DataTypes/Serializations/SerializationObject.cpp

@@ -63,12 +63,12 @@ private:
     size_t num_dimensions_to_keep;
 };
 
-using Node = typename ColumnObject::SubcolumnsTree::Node;
+using Node = typename ColumnObject::Subcolumns::Node;
 
 /// Finds a subcolumn from the same Nested type as @entry and inserts
 /// an array with default values with consistent sizes as in Nested type.
 bool tryInsertDefaultFromNested(
-    const std::shared_ptr<Node> & entry, const ColumnObject::SubcolumnsTree & subcolumns)
+    const std::shared_ptr<Node> & entry, const ColumnObject::Subcolumns & subcolumns)
 {
     if (!entry->path.hasNested())
         return false;
@@ -198,7 +198,7 @@ void SerializationObject<Parser>::deserializeWholeText(IColumn & column, ReadBuf
 template <typename Parser>
 void SerializationObject<Parser>::deserializeTextEscaped(IColumn & column, ReadBuffer & istr, const FormatSettings &) const
 {
-    deserializeTextImpl(column, [&](String & s) { readEscapedStringInto(s, istr); });
+    deserializeTextImpl(column, [&](String & s) { readEscapedString(s, istr); });
 }
 
 template <typename Parser>

src/Databases/DatabaseReplicated.cpp

@@ -88,6 +88,9 @@ DatabaseReplicated::DatabaseReplicated(
     /// If zookeeper chroot prefix is used, path should start with '/', because chroot concatenates without it.
     if (zookeeper_path.front() != '/')
         zookeeper_path = "/" + zookeeper_path;
+
+    if (!db_settings.collection_name.value.empty())
+        fillClusterAuthInfo(db_settings.collection_name.value, context_->getConfigRef());
 }
 
 String DatabaseReplicated::getFullReplicaName() const
@@ -191,22 +194,36 @@ ClusterPtr DatabaseReplicated::getClusterImpl() const
         shards.back().emplace_back(unescapeForFileName(host_port));
     }
 
-    String username = db_settings.cluster_username;
-    String password = db_settings.cluster_password;
     UInt16 default_port = getContext()->getTCPPort();
-    bool secure = db_settings.cluster_secure_connection;
 
     bool treat_local_as_remote = false;
     bool treat_local_port_as_remote = getContext()->getApplicationType() == Context::ApplicationType::LOCAL;
     return std::make_shared<Cluster>(
         getContext()->getSettingsRef(),
         shards,
-        username,
+        cluster_auth_info.cluster_username,
-        password,
+        cluster_auth_info.cluster_password,
         default_port,
         treat_local_as_remote,
         treat_local_port_as_remote,
-        secure);
+        cluster_auth_info.cluster_secure_connection,
+        /*priority=*/1,
+        database_name,
+        cluster_auth_info.cluster_secret);
+}
+
+
+void DatabaseReplicated::fillClusterAuthInfo(String collection_name, const Poco::Util::AbstractConfiguration & config_ref)
+{
+    const auto & config_prefix = fmt::format("named_collections.{}", collection_name);
+
+    if (!config_ref.has(config_prefix))
+            throw Exception(ErrorCodes::BAD_ARGUMENTS, "There is no collection named `{}` in config", collection_name);
+
+    cluster_auth_info.cluster_username = config_ref.getString(config_prefix + ".cluster_username", "");
+    cluster_auth_info.cluster_password = config_ref.getString(config_prefix + ".cluster_password", "");
+    cluster_auth_info.cluster_secret = config_ref.getString(config_prefix + ".cluster_secret", "");
+    cluster_auth_info.cluster_secure_connection = config_ref.getBool(config_prefix + ".cluster_secure_connection", false);
 }
 
 void DatabaseReplicated::tryConnectToZooKeeperAndInitDatabase(bool force_attach)

src/Databases/DatabaseReplicated.h

@@ -75,6 +75,16 @@ private:
     bool createDatabaseNodesInZooKeeper(const ZooKeeperPtr & current_zookeeper);
     void createReplicaNodesInZooKeeper(const ZooKeeperPtr & current_zookeeper);
 
+    struct
+    {
+        String cluster_username{"default"};
+        String cluster_password;
+        String cluster_secret;
+        bool cluster_secure_connection{false};
+    } cluster_auth_info;
+
+    void fillClusterAuthInfo(String collection_name, const Poco::Util::AbstractConfiguration & config);
+
     void checkQueryValid(const ASTPtr & query, ContextPtr query_context) const;
 
     void recoverLostReplica(const ZooKeeperPtr & current_zookeeper, UInt32 our_log_ptr, UInt32 max_log_ptr);

src/Databases/DatabaseReplicatedSettings.h

@@ -8,12 +8,11 @@ namespace DB
 class ASTStorage;
 
 #define LIST_OF_DATABASE_REPLICATED_SETTINGS(M) \
-    M(Float, max_broken_tables_ratio, 0.5, "Do not recover replica automatically if the ratio of staled tables to all tables is greater", 0) \
+    M(Float,  max_broken_tables_ratio, 0.5, "Do not recover replica automatically if the ratio of staled tables to all tables is greater", 0) \
     M(UInt64, max_replication_lag_to_enqueue, 10, "Replica will throw exception on attempt to execute query if its replication lag greater", 0) \
     M(UInt64, wait_entry_commited_timeout_sec, 3600, "Replicas will try to cancel query if timeout exceed, but initiator host has not executed it yet", 0) \
-    M(String, cluster_username, "default", "Username to use when connecting to hosts of cluster", 0) \
+    M(String, collection_name, "", "A name of a collection defined in server's config where all info for cluster authentication is defined", 0) \
-    M(String, cluster_password, "", "Password to use when connecting to hosts of cluster", 0) \
+
-    M(Bool, cluster_secure_connection, false, "Enable TLS when connecting to hosts of cluster", 0) \
 
 DECLARE_SETTINGS_TRAITS(DatabaseReplicatedSettingsTraits, LIST_OF_DATABASE_REPLICATED_SETTINGS)
 

src/Disks/IO/CachedReadBufferFromRemoteFS.h

@@ -96,6 +96,7 @@ private:
             case ReadType::REMOTE_FS_READ_AND_PUT_IN_CACHE:
                 return "REMOTE_FS_READ_AND_PUT_IN_CACHE";
         }
+        __builtin_unreachable();
     }
     size_t first_offset = 0;
 };

src/Disks/S3/DiskS3.cpp

@@ -20,6 +20,7 @@
 #include <Common/getRandomASCIIString.h>
 
 #include <Interpreters/Context.h>
+#include <Interpreters/threadPoolCallbackRunner.h>
 #include <IO/ReadBufferFromS3.h>
 #include <IO/ReadBufferFromString.h>
 #include <IO/ReadHelpers.h>
@@ -264,32 +265,6 @@ std::unique_ptr<WriteBufferFromFileBase> DiskS3::writeFile(const String & path,
     LOG_TRACE(log, "{} to file by path: {}. S3 path: {}",
               mode == WriteMode::Rewrite ? "Write" : "Append", backQuote(metadata_disk->getPath() + path), remote_fs_root_path + blob_name);
 
-    ScheduleFunc schedule = [pool = &getThreadPoolWriter(), thread_group = CurrentThread::getGroup()](auto callback)
-    {
-        pool->scheduleOrThrow([callback = std::move(callback), thread_group]()
-        {
-            if (thread_group)
-                CurrentThread::attachTo(thread_group);
-
-            SCOPE_EXIT_SAFE(
-                if (thread_group)
-                    CurrentThread::detachQueryIfNotDetached();
-
-                /// After we detached from the thread_group, parent for memory_tracker inside ThreadStatus will be reset to it's parent.
-                /// Typically, it may be changes from Process to User.
-                /// Usually it could be ok, because thread pool task is executed before user-level memory tracker is destroyed.
-                /// However, thread could stay alive inside the thread pool, and it's ThreadStatus as well.
-                /// When, finally, we destroy the thread (and the ThreadStatus),
-                /// it can use memory tracker in the ~ThreadStatus in order to alloc/free untracked_memory,\
-                /// and by this time user-level memory tracker may be already destroyed.
-                ///
-                /// As a work-around, reset memory tracker to total, which is always alive.
-                CurrentThread::get().memory_tracker.setParent(&total_memory_tracker);
-            );
-            callback();
-        });
-    };
-
     auto s3_buffer = std::make_unique<WriteBufferFromS3>(
         settings->client,
         bucket,
@@ -299,7 +274,7 @@ std::unique_ptr<WriteBufferFromFileBase> DiskS3::writeFile(const String & path,
         settings->s3_upload_part_size_multiply_parts_count_threshold,
         settings->s3_max_single_part_upload_size,
         std::move(object_metadata),
-        buf_size, std::move(schedule));
+        buf_size, threadPoolCallbackRunner(getThreadPoolWriter()));
 
     auto create_metadata_callback = [this, path, blob_name, mode] (size_t count)
     {

src/Formats/FormatFactory.cpp

@@ -89,10 +89,10 @@ FormatSettings getFormatSettings(ContextPtr context, const Settings & settings)
     format_settings.json.quote_64bit_integers = settings.output_format_json_quote_64bit_integers;
     format_settings.json.quote_denormals = settings.output_format_json_quote_denormals;
     format_settings.null_as_default = settings.input_format_null_as_default;
-    format_settings.use_lowercase_column_name = settings.input_format_use_lowercase_column_name;
     format_settings.decimal_trailing_zeros = settings.output_format_decimal_trailing_zeros;
     format_settings.parquet.row_group_size = settings.output_format_parquet_row_group_size;
     format_settings.parquet.import_nested = settings.input_format_parquet_import_nested;
+    format_settings.parquet.case_insensitive_column_matching = settings.input_format_parquet_case_insensitive_column_matching;
     format_settings.parquet.allow_missing_columns = settings.input_format_parquet_allow_missing_columns;
     format_settings.pretty.charset = settings.output_format_pretty_grid_charset.toString() == "ASCII" ? FormatSettings::Pretty::Charset::ASCII : FormatSettings::Pretty::Charset::UTF8;
     format_settings.pretty.color = settings.output_format_pretty_color;
@@ -123,9 +123,11 @@ FormatSettings getFormatSettings(ContextPtr context, const Settings & settings)
     format_settings.arrow.low_cardinality_as_dictionary = settings.output_format_arrow_low_cardinality_as_dictionary;
     format_settings.arrow.import_nested = settings.input_format_arrow_import_nested;
     format_settings.arrow.allow_missing_columns = settings.input_format_arrow_allow_missing_columns;
+    format_settings.arrow.case_insensitive_column_matching = settings.input_format_arrow_case_insensitive_column_matching;
     format_settings.orc.import_nested = settings.input_format_orc_import_nested;
     format_settings.orc.allow_missing_columns = settings.input_format_orc_allow_missing_columns;
     format_settings.orc.row_batch_size = settings.input_format_orc_row_batch_size;
+    format_settings.orc.case_insensitive_column_matching = settings.input_format_orc_case_insensitive_column_matching;
     format_settings.defaults_for_omitted_fields = settings.input_format_defaults_for_omitted_fields;
     format_settings.capn_proto.enum_comparing_mode = settings.format_capn_proto_enum_comparising_mode;
     format_settings.seekable_read = settings.input_format_allow_seeks;

src/Formats/FormatSettings.h

@@ -32,7 +32,6 @@ struct FormatSettings
     bool null_as_default = true;
     bool decimal_trailing_zeros = false;
     bool defaults_for_omitted_fields = true;
-    bool use_lowercase_column_name = false;
 
     bool seekable_read = true;
     UInt64 max_rows_to_read_for_schema_inference = 100;
@@ -75,6 +74,7 @@ struct FormatSettings
         bool low_cardinality_as_dictionary = false;
         bool import_nested = false;
         bool allow_missing_columns = false;
+        bool case_insensitive_column_matching = false;
     } arrow;
 
     struct
@@ -137,6 +137,7 @@ struct FormatSettings
         UInt64 row_group_size = 1000000;
         bool import_nested = false;
         bool allow_missing_columns = false;
+        bool case_insensitive_column_matching = false;
     } parquet;
 
     struct Pretty
@@ -217,6 +218,7 @@ struct FormatSettings
         bool import_nested = false;
         bool allow_missing_columns = false;
         int64_t row_batch_size = 100'000;
+        bool case_insensitive_column_matching = false;
     } orc;
 
     /// For capnProto format we should determine how to

src/Formats/registerFormats.cpp

@@ -13,6 +13,7 @@ void registerFileSegmentationEngineCSV(FormatFactory & factory);
 void registerFileSegmentationEngineJSONEachRow(FormatFactory & factory);
 void registerFileSegmentationEngineRegexp(FormatFactory & factory);
 void registerFileSegmentationEngineJSONAsString(FormatFactory & factory);
+void registerFileSegmentationEngineJSONAsObject(FormatFactory & factory);
 void registerFileSegmentationEngineJSONCompactEachRow(FormatFactory & factory);
 
 /// Formats for both input/output.
@@ -103,6 +104,7 @@ void registerProtobufSchemaReader(FormatFactory & factory);
 void registerProtobufListSchemaReader(FormatFactory & factory);
 void registerLineAsStringSchemaReader(FormatFactory & factory);
 void registerJSONAsStringSchemaReader(FormatFactory & factory);
+void registerJSONAsObjectSchemaReader(FormatFactory & factory);
 void registerRawBLOBSchemaReader(FormatFactory & factory);
 void registerMsgPackSchemaReader(FormatFactory & factory);
 void registerCapnProtoSchemaReader(FormatFactory & factory);
@@ -123,6 +125,7 @@ void registerFormats()
     registerFileSegmentationEngineJSONEachRow(factory);
     registerFileSegmentationEngineRegexp(factory);
     registerFileSegmentationEngineJSONAsString(factory);
+    registerFileSegmentationEngineJSONAsObject(factory);
     registerFileSegmentationEngineJSONCompactEachRow(factory);
 
     registerInputFormatNative(factory);
@@ -207,6 +210,7 @@ void registerFormats()
     registerProtobufListSchemaReader(factory);
     registerLineAsStringSchemaReader(factory);
     registerJSONAsStringSchemaReader(factory);
+    registerJSONAsObjectSchemaReader(factory);
     registerRawBLOBSchemaReader(factory);
     registerMsgPackSchemaReader(factory);
     registerCapnProtoSchemaReader(factory);

src/Functions/caseWithExpression.cpp

@@ -43,6 +43,9 @@ public:
         for (size_t i = 2; i < args.size() - 1; i += 2)
             dst_array_types.push_back(args[i]);
 
+        // Type of the ELSE branch
+        dst_array_types.push_back(args.back());
+
         return getLeastSupertype(dst_array_types);
     }
 

src/Functions/kostikConsistentHash.cpp

@@ -7,9 +7,9 @@ namespace DB
 {
 
 /// An O(1) time and space consistent hash algorithm by Konstantin Oblakov
-struct YandexConsistentHashImpl
+struct KostikConsistentHashImpl
 {
-    static constexpr auto name = "yandexConsistentHash";
+    static constexpr auto name = "kostikConsistentHash";
 
     using HashType = UInt64;
     /// Actually it supports UInt64, but it is efficient only if n <= 32768
@@ -23,12 +23,12 @@ struct YandexConsistentHashImpl
     }
 };
 
-using FunctionYandexConsistentHash = FunctionConsistentHashImpl<YandexConsistentHashImpl>;
+using FunctionKostikConsistentHash = FunctionConsistentHashImpl<KostikConsistentHashImpl>;
 
-void registerFunctionYandexConsistentHash(FunctionFactory & factory)
+void registerFunctionKostikConsistentHash(FunctionFactory & factory)
 {
-    factory.registerFunction<FunctionYandexConsistentHash>();
+    factory.registerFunction<FunctionKostikConsistentHash>();
+    factory.registerAlias("yandexConsistentHash", "kostikConsistentHash");
 }
 
 }
-

src/Functions/makeDate.cpp

@@ -0,0 +1,155 @@
+#include <Functions/IFunction.h>
+#include <Functions/FunctionFactory.h>
+#include <DataTypes/DataTypeDate.h>
+#include <DataTypes/DataTypeDate32.h>
+#include <DataTypes/DataTypesNumber.h>
+#include <Columns/ColumnsNumber.h>
+#include <Interpreters/castColumn.h>
+
+#include <Common/DateLUT.h>
+#include <Common/typeid_cast.h>
+
+#include <array>
+
+namespace DB
+{
+namespace ErrorCodes
+{
+    extern const int NUMBER_OF_ARGUMENTS_DOESNT_MATCH;
+    extern const int ILLEGAL_TYPE_OF_ARGUMENT;
+}
+
+namespace
+{
+
+// A helper function to simplify comparisons of valid YYYY-MM-DD values for <,>,=
+inline constexpr Int64 YearMonthDayToSingleInt(Int64 year, Int64 month, Int64 day)
+{
+    return year * 512 + month * 32 + day;
+}
+
+// Common implementation for makeDate, makeDate32
+template <typename Traits>
+class FunctionMakeDate : public IFunction
+{
+private:
+    static constexpr std::array<const char*, 3> argument_names = {"year", "month", "day"};
+
+public:
+    static constexpr auto name = Traits::name;
+
+    static FunctionPtr create(ContextPtr) { return std::make_shared<FunctionMakeDate>(); }
+
+    String getName() const override { return name; }
+
+    bool isVariadic() const override { return false; }
+
+    size_t getNumberOfArguments() const override { return argument_names.size(); }
+
+    bool isInjective(const ColumnsWithTypeAndName &) const override
+    {
+        return false; // {year,month,day} that are out of supported range are converted into a default value
+    }
+
+    bool isSuitableForShortCircuitArgumentsExecution(const DataTypesWithConstInfo & /*arguments*/) const override { return true; }
+
+    bool useDefaultImplementationForNulls() const override { return true; }
+
+    bool useDefaultImplementationForConstants() const override { return true; }
+
+    DataTypePtr getReturnTypeImpl(const DataTypes & arguments) const override
+    {
+        if (arguments.size() != argument_names.size())
+            throw Exception(ErrorCodes::NUMBER_OF_ARGUMENTS_DOESNT_MATCH,
+                "Function {} requires 3 arguments, but {} given", getName(), arguments.size());
+
+        for (size_t i = 0; i < argument_names.size(); ++i)
+        {
+            DataTypePtr argument_type = arguments[i];
+            if (!isNumber(argument_type))
+                throw Exception(ErrorCodes::ILLEGAL_TYPE_OF_ARGUMENT,
+                    "Argument '{}' for function {} must be number", std::string(argument_names[i]), getName());
+        }
+
+        return std::make_shared<typename Traits::ReturnDataType>();
+    }
+
+    ColumnPtr executeImpl(const ColumnsWithTypeAndName & arguments, const DataTypePtr &, size_t input_rows_count) const override
+    {
+        const DataTypePtr converted_argument_type = std::make_shared<DataTypeFloat32>();
+        Columns converted_arguments;
+        converted_arguments.reserve(arguments.size());
+        for (const auto & argument : arguments)
+        {
+            ColumnPtr argument_column = castColumn(argument, converted_argument_type);
+            argument_column = argument_column->convertToFullColumnIfConst();
+            converted_arguments.push_back(argument_column);
+        }
+
+        auto res_column = Traits::ReturnColumnType::create(input_rows_count);
+        auto & result_data = res_column->getData();
+
+        const auto & year_data = typeid_cast<const ColumnFloat32 &>(*converted_arguments[0]).getData();
+        const auto & month_data = typeid_cast<const ColumnFloat32 &>(*converted_arguments[1]).getData();
+        const auto & day_data = typeid_cast<const ColumnFloat32 &>(*converted_arguments[2]).getData();
+
+        const auto & date_lut = DateLUT::instance();
+
+        for (size_t i = 0; i < input_rows_count; ++i)
+        {
+            const auto year = year_data[i];
+            const auto month = month_data[i];
+            const auto day = day_data[i];
+
+            Int32 day_num = 0;
+
+            if (year >= Traits::MIN_YEAR &&
+                year <= Traits::MAX_YEAR &&
+                month >= 1 && month <= 12 &&
+                day >= 1 && day <= 31 &&
+                YearMonthDayToSingleInt(year, month, day) <= Traits::MAX_DATE)
+            {
+                day_num = date_lut.makeDayNum(year, month, day);
+            }
+
+            result_data[i] = day_num;
+        }
+
+        return res_column;
+    }
+};
+
+// makeDate(year, month, day)
+struct MakeDateTraits
+{
+    static constexpr auto name = "makeDate";
+    using ReturnDataType = DataTypeDate;
+    using ReturnColumnType = ColumnUInt16;
+
+    static constexpr auto MIN_YEAR = 1970;
+    static constexpr auto MAX_YEAR = 2149;
+    // This date has the maximum day number that fits in 16-bit uint
+    static constexpr auto MAX_DATE = YearMonthDayToSingleInt(MAX_YEAR, 6, 6);
+};
+
+// makeDate32(year, month, day)
+struct MakeDate32Traits
+{
+    static constexpr auto name = "makeDate32";
+    using ReturnDataType = DataTypeDate32;
+    using ReturnColumnType = ColumnInt32;
+
+    static constexpr auto MIN_YEAR = 1925;
+    static constexpr auto MAX_YEAR = 2283;
+    static constexpr auto MAX_DATE = YearMonthDayToSingleInt(MAX_YEAR, 11, 11);
+};
+
+}
+
+void registerFunctionsMakeDate(FunctionFactory & factory)
+{
+    factory.registerFunction<FunctionMakeDate<MakeDateTraits>>();
+    factory.registerFunction<FunctionMakeDate<MakeDate32Traits>>();
+}
+
+}

src/Functions/registerFunctions.cpp

@@ -8,6 +8,7 @@ namespace DB
 void registerFunctionsArithmetic(FunctionFactory &);
 void registerFunctionsArray(FunctionFactory &);
 void registerFunctionsTuple(FunctionFactory &);
+void registerFunctionsMakeDate(FunctionFactory &);
 void registerFunctionsMap(FunctionFactory &);
 void registerFunctionsBitmap(FunctionFactory &);
 void registerFunctionsBinaryRepr(FunctionFactory &);
@@ -73,6 +74,7 @@ void registerFunctions()
     registerFunctionsArithmetic(factory);
     registerFunctionsArray(factory);
     registerFunctionsTuple(factory);
+    registerFunctionsMakeDate(factory);
     registerFunctionsMap(factory);
     registerFunctionsBitmap(factory);
     registerFunctionsBinaryRepr(factory);

src/Functions/registerFunctionsConsistentHashing.cpp

@@ -2,12 +2,12 @@ namespace DB
 {
 class FunctionFactory;
 
-void registerFunctionYandexConsistentHash(FunctionFactory & factory);
+void registerFunctionKostikConsistentHash(FunctionFactory & factory);
 void registerFunctionJumpConsistentHash(FunctionFactory & factory);
 
 void registerFunctionsConsistentHashing(FunctionFactory & factory)
 {
-    registerFunctionYandexConsistentHash(factory);
+    registerFunctionKostikConsistentHash(factory);
     registerFunctionJumpConsistentHash(factory);
 }
 

src/IO/ReadBufferFromFileDescriptor.h

@@ -39,6 +39,10 @@ public:
     {
     }
 
+    virtual ~ReadBufferFromFileDescriptor() override
+    {
+    }
+
     int getFD() const
     {
         return fd;
@@ -80,6 +84,9 @@ public:
     {
         use_pread = true;
     }
+    virtual ~ReadBufferFromFileDescriptorPRead() override
+    {
+    }
 };
 
 }

src/IO/ReadBufferFromString.h

@@ -15,4 +15,11 @@ public:
 
     explicit ReadBufferFromString(std::string_view s) : ReadBufferFromMemory(s.data(), s.size()) {}
 };
+
+class ReadBufferFromOwnString : public String, public ReadBufferFromString
+{
+public:
+    explicit ReadBufferFromOwnString(const String & s_): String(s_), ReadBufferFromString(*this) {}
+};
+
 }

src/IO/ReadWriteBufferFromHTTP.h

@@ -663,7 +663,7 @@ public:
 
         Range range{from, to};
         from = to;
-        return std::move(range);
+        return range;
     }
 
 private:

src/IO/examples/read_buffer.cpp

@@ -2,18 +2,15 @@
 
 #include <iostream>
 
-#include <base/types.h>
+#include <IO/ReadBuffer.h>
-#include <IO/ReadHelpers.h>
 #include <IO/ReadBufferFromString.h>
+#include <IO/ReadHelpers.h>
+#include <base/types.h>
 
-
+int readAndPrint(DB::ReadBuffer & in)
-int main(int, char **)
 {
     try
     {
-        std::string s = "-123456 123.456 вася пе\\tтя\t'\\'xyz\\\\'";
-        DB::ReadBufferFromString in(s);
-
         DB::Int64 a;
         DB::Float64 b;
         DB::String c, d;
@@ -31,12 +28,32 @@ int main(int, char **)
 
         std::cout << a << ' ' << b << ' ' << c << '\t' << '\'' << d << '\'' << std::endl;
         std::cout << in.count() << std::endl;
+        return 0;
     }
     catch (const DB::Exception & e)
     {
         std::cerr << e.what() << ", " << e.displayText() << std::endl;
         return 1;
     }
+}
+
+int main(int, char **)
+{
+    {
+        std::string s = "-123456 123.456 вася пе\\tтя\t'\\'xyz\\\\'";
+        DB::ReadBufferFromString in(s);
+        if (readAndPrint(in))
+            std::cout << "readAndPrint from ReadBufferFromString failed" << std::endl;
+    }
+
+
+    std::shared_ptr<DB::ReadBufferFromOwnString> in;
+    {
+        std::string s = "-123456 123.456 вася пе\\tтя\t'\\'xyz\\\\'";
+        in = std::make_shared<DB::ReadBufferFromOwnString>(s);
+    }
+    if (readAndPrint(*in))
+        std::cout << "readAndPrint from ReadBufferFromOwnString failed" << std::endl;
 
     return 0;
 }

src/Interpreters/AsynchronousInsertQueue.cpp

@@ -32,6 +32,7 @@ namespace CurrentMetrics
 namespace ProfileEvents
 {
     extern const Event AsyncInsertQuery;
+    extern const Event AsyncInsertBytes;
 }
 
 namespace DB
@@ -222,7 +223,9 @@ void AsynchronousInsertQueue::pushImpl(InsertData::EntryPtr entry, QueueIterator
     if (!data)
         data = std::make_unique<InsertData>();
 
-    data->size += entry->bytes.size();
+    size_t entry_data_size = entry->bytes.size();
+
+    data->size += entry_data_size;
     data->last_update = std::chrono::steady_clock::now();
     data->entries.emplace_back(entry);
 
@@ -239,6 +242,7 @@ void AsynchronousInsertQueue::pushImpl(InsertData::EntryPtr entry, QueueIterator
 
     CurrentMetrics::add(CurrentMetrics::PendingAsyncInsert);
     ProfileEvents::increment(ProfileEvents::AsyncInsertQuery);
+    ProfileEvents::increment(ProfileEvents::AsyncInsertBytes, entry_data_size);
 }
 
 void AsynchronousInsertQueue::waitForProcessingQuery(const String & query_id, const Milliseconds & timeout)

src/Interpreters/AsynchronousMetrics.cpp

@@ -12,6 +12,7 @@
 #include <Storages/MarkCache.h>
 #include <Storages/StorageMergeTree.h>
 #include <Storages/StorageReplicatedMergeTree.h>
+#include <Storages/MergeTree/MergeTreeMetadataCache.h>
 #include <IO/UncompressedCache.h>
 #include <IO/MMappedFileCache.h>
 #include <IO/ReadHelpers.h>
@@ -607,6 +608,15 @@ void AsynchronousMetrics::update(std::chrono::system_clock::time_point update_ti
         }
     }
 
+#if USE_ROCKSDB
+    {
+        if (auto metadata_cache = getContext()->tryGetMergeTreeMetadataCache())
+        {
+            new_values["MergeTreeMetadataCacheSize"] = metadata_cache->getEstimateNumKeys();
+        }
+    }
+#endif
+
 #if USE_EMBEDDED_COMPILER
     {
         if (auto * compiled_expression_cache = CompiledExpressionCacheFactory::instance().tryGetCache())
@@ -617,6 +627,7 @@ void AsynchronousMetrics::update(std::chrono::system_clock::time_point update_ti
     }
 #endif
 
+
     new_values["Uptime"] = getContext()->getUptimeSeconds();
 
     /// Process process memory usage according to OS