Merge branch 'ClickHouse:master' into opt_insertmanyfrom

.clang-tidy

@@ -10,7 +10,7 @@
 
 # TODO Let clang-tidy check headers in further directories
 #      --> HeaderFilterRegex: '^.*/(src|base|programs|utils)/.*(h|hpp)$'
-HeaderFilterRegex: '^.*/(base)/.*(h|hpp)$'
+HeaderFilterRegex: '^.*/(base|programs|utils)/.*(h|hpp)$'
 
 Checks: '*,
     -abseil-*,

.github/workflows/master.yml

@@ -305,7 +305,7 @@ jobs:
       runner_type: style-checker-aarch64
       data: ${{ needs.RunConfig.outputs.data }}
   MarkReleaseReady:
-    if: ${{ ! (contains(needs.*.result, 'skipped') || contains(needs.*.result, 'failure')) }}
+    if: ${{ !failure() && !cancelled() }}
     needs:
       - BuilderBinDarwin
       - BuilderBinDarwinAarch64
@@ -313,9 +313,25 @@ jobs:
       - BuilderDebAarch64
     runs-on: [self-hosted, style-checker]
     steps:
+      - name: Debug
+        run: |
+          echo need with different filters
+          cat << 'EOF'
+          ${{ toJSON(needs) }}
+          ${{ toJSON(needs.*.result) }}
+          no failures ${{ !contains(needs.*.result, 'failure') }}
+          no skips ${{ !contains(needs.*.result, 'skipped') }}
+          no both ${{ !(contains(needs.*.result, 'skipped') || contains(needs.*.result, 'failure')) }}
+          EOF
+      - name: Not ready
+        # fail the job to be able restart it
+        if: ${{ contains(needs.*.result, 'skipped') || contains(needs.*.result, 'failure') }}
+        run: exit 1
       - name: Check out repository code
+        if: ${{ ! (contains(needs.*.result, 'skipped') || contains(needs.*.result, 'failure')) }}
         uses: ClickHouse/checkout@v1
       - name: Mark Commit Release Ready
+        if: ${{ ! (contains(needs.*.result, 'skipped') || contains(needs.*.result, 'failure')) }}
         run: |
           cd "$GITHUB_WORKSPACE/tests/ci"
           python3 mark_release_ready.py

.github/workflows/pull_request.yml

@@ -13,9 +13,7 @@ on:  # yamllint disable-line rule:truthy
       - opened
     branches:
       - master
-##########################################################################################
-##################################### SMALL CHECKS #######################################
-##########################################################################################
+
 jobs:
   RunConfig:
     runs-on: [self-hosted, style-checker-aarch64]
@@ -70,13 +68,13 @@ jobs:
           python3 "$GITHUB_WORKSPACE/tests/ci/ci.py" --infile ${{ runner.temp }}/ci_run_data.json --post --job-name 'Style check'
   BuildDockers:
     needs: [RunConfig]
-    if: ${{ !failure() && !cancelled() }}
+    if: ${{ !failure() && !cancelled() && toJson(fromJson(needs.RunConfig.outputs.data).docker_data.missing_multi) != '[]' }}
     uses: ./.github/workflows/reusable_docker.yml
     with:
       data: ${{ needs.RunConfig.outputs.data }}
   StyleCheck:
     needs: [RunConfig, BuildDockers]
-    if: ${{ !failure() && !cancelled() }}
+    if: ${{ !failure() && !cancelled() && contains(fromJson(needs.RunConfig.outputs.data).jobs_data.jobs_to_do, 'Style check')}}
     uses: ./.github/workflows/reusable_test.yml
     with:
       test_name: Style check
@@ -89,19 +87,9 @@ jobs:
         ROBOT_CLICKHOUSE_SSH_KEY<<RCSK
         ${{secrets.ROBOT_CLICKHOUSE_SSH_KEY}}
         RCSK
-  DocsCheck:
-    needs: [RunConfig, StyleCheck]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Docs check
-      runner_type: func-tester-aarch64
-      data: ${{ needs.RunConfig.outputs.data }}
-      run_command: |
-        python3 docs_check.py
   FastTest:
     needs: [RunConfig, StyleCheck]
-    if: ${{ !failure() && !cancelled() }}
+    if: ${{ !failure() && !cancelled() && contains(fromJson(needs.RunConfig.outputs.data).jobs_data.jobs_to_do, 'Fast test') }}
     uses: ./.github/workflows/reusable_test.yml
     with:
       test_name: Fast test
@@ -109,818 +97,83 @@ jobs:
       data: ${{ needs.RunConfig.outputs.data }}
       run_command: |
           python3 fast_test_check.py
-  CompatibilityCheckX86:
-    needs: [RunConfig, BuilderDebRelease]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Compatibility check (amd64)
-      runner_type: style-checker
-      data: ${{ needs.RunConfig.outputs.data }}
-  CompatibilityCheckAarch64:
-    needs: [RunConfig, BuilderDebAarch64]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Compatibility check (aarch64)
-      runner_type: style-checker
-      data: ${{ needs.RunConfig.outputs.data }}
-#########################################################################################
-#################################### ORDINARY BUILDS ####################################
-#########################################################################################
-  BuilderDebDebug:
+
+  ################################# Main statges #################################
+  # for main CI chain
+  #
+  Builds_1:
     needs: [RunConfig, FastTest]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_build.yml
+    if: ${{ !failure() && !cancelled() && contains(fromJson(needs.RunConfig.outputs.data).stages_data.stages_to_do, 'Builds_1') }}
+    # using callable wf (reusable_stage.yml) allows to group all nested jobs under a tab
+    uses: ./.github/workflows/reusable_build_stage.yml
     with:
-      build_name: package_debug
+      stage: Builds_1
       data: ${{ needs.RunConfig.outputs.data }}
-  BuilderDebRelease:
-    needs: [RunConfig, FastTest]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_build.yml
+  Tests_1:
+    needs: [RunConfig, Builds_1]
+    if: ${{ !failure() && !cancelled() && contains(fromJson(needs.RunConfig.outputs.data).stages_data.stages_to_do, 'Tests_1') }}
+    # using callable wf (reusable_stage.yml) allows to group all nested jobs under a tab
+    uses: ./.github/workflows/reusable_test_stage.yml
     with:
-      build_name: package_release
-      checkout_depth: 0
+      stage: Tests_1
       data: ${{ needs.RunConfig.outputs.data }}
-  BuilderDebReleaseCoverage:
-    needs: [RunConfig, FastTest]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_build.yml
+  Builds_2:
+    needs: [RunConfig, Builds_1]
+    if: ${{ !failure() && !cancelled() && contains(fromJson(needs.RunConfig.outputs.data).stages_data.stages_to_do, 'Builds_2') }}
+    # using callable wf (reusable_stage.yml) allows to group all nested jobs under a tab
+    uses: ./.github/workflows/reusable_build_stage.yml
     with:
-      build_name: package_release_coverage
-      checkout_depth: 0
+      stage: Builds_2
       data: ${{ needs.RunConfig.outputs.data }}
-  BuilderDebAarch64:
-    needs: [RunConfig, FastTest]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_build.yml
+  Tests_2:
+    needs: [RunConfig, Builds_2]
+    if: ${{ !failure() && !cancelled() && contains(fromJson(needs.RunConfig.outputs.data).stages_data.stages_to_do, 'Tests_2') }}
+    # using callable wf (reusable_stage.yml) allows to group all nested jobs under a tab
+    uses: ./.github/workflows/reusable_test_stage.yml
     with:
-      build_name: package_aarch64
-      checkout_depth: 0
+      stage: Tests_2
       data: ${{ needs.RunConfig.outputs.data }}
-  BuilderBinRelease:
-    needs: [RunConfig, FastTest]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_build.yml
-    with:
-      build_name: binary_release
-      data: ${{ needs.RunConfig.outputs.data }}
-  BuilderDebAsan:
-    needs: [RunConfig, FastTest]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_build.yml
-    with:
-      build_name: package_asan
-      data: ${{ needs.RunConfig.outputs.data }}
-  BuilderDebUBsan:
-    needs: [RunConfig, FastTest]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_build.yml
-    with:
-      build_name: package_ubsan
-      data: ${{ needs.RunConfig.outputs.data }}
-  BuilderDebTsan:
-    needs: [RunConfig, FastTest]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_build.yml
-    with:
-      build_name: package_tsan
-      data: ${{ needs.RunConfig.outputs.data }}
-  BuilderDebMsan:
-    needs: [RunConfig, FastTest]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_build.yml
-    with:
-      build_name: package_msan
-      data: ${{ needs.RunConfig.outputs.data }}
-##########################################################################################
-##################################### SPECIAL BUILDS #####################################
-##########################################################################################
-  BuilderBinClangTidy:
-    needs: [RunConfig, FastTest]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_build.yml
-    with:
-      build_name: binary_tidy
-      data: ${{ needs.RunConfig.outputs.data }}
-  BuilderBinDarwin:
-    needs: [RunConfig, FastTest]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_build.yml
-    with:
-      build_name: binary_darwin
-      data: ${{ needs.RunConfig.outputs.data }}
-  BuilderBinAarch64:
-    needs: [RunConfig, FastTest]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_build.yml
-    with:
-      build_name: binary_aarch64
-      data: ${{ needs.RunConfig.outputs.data }}
-  BuilderBinFreeBSD:
-    needs: [RunConfig, FastTest]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_build.yml
-    with:
-      build_name: binary_freebsd
-      data: ${{ needs.RunConfig.outputs.data }}
-  BuilderBinDarwinAarch64:
-    needs: [RunConfig, FastTest]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_build.yml
-    with:
-      build_name: binary_darwin_aarch64
-      data: ${{ needs.RunConfig.outputs.data }}
-  BuilderBinPPC64:
-    needs: [RunConfig, FastTest]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_build.yml
-    with:
-      build_name: binary_ppc64le
-      data: ${{ needs.RunConfig.outputs.data }}
-  BuilderBinAmd64Compat:
-    needs: [RunConfig, FastTest]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_build.yml
-    with:
-      build_name: binary_amd64_compat
-      data: ${{ needs.RunConfig.outputs.data }}
-  BuilderBinAmd64Musl:
-    needs: [RunConfig, FastTest]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_build.yml
-    with:
-      build_name: binary_amd64_musl
-      data: ${{ needs.RunConfig.outputs.data }}
-  BuilderBinAarch64V80Compat:
-    needs: [RunConfig, FastTest]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_build.yml
-    with:
-      build_name: binary_aarch64_v80compat
-      data: ${{ needs.RunConfig.outputs.data }}
-  BuilderBinRISCV64:
-    needs: [RunConfig, FastTest]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_build.yml
-    with:
-      build_name: binary_riscv64
-      data: ${{ needs.RunConfig.outputs.data }}
-  BuilderBinS390X:
-    needs: [RunConfig, FastTest]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_build.yml
-    with:
-      build_name: binary_s390x
-      data: ${{ needs.RunConfig.outputs.data }}
-############################################################################################
-##################################### Docker images  #######################################
-############################################################################################
-  DockerServerImage:
-    needs: [RunConfig, BuilderDebRelease, BuilderDebAarch64]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Docker server image
-      runner_type: style-checker
-      data: ${{ needs.RunConfig.outputs.data }}
-  DockerKeeperImage:
-    needs: [RunConfig, BuilderDebRelease, BuilderDebAarch64]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Docker keeper image
-      runner_type: style-checker
-      data: ${{ needs.RunConfig.outputs.data }}
-############################################################################################
-##################################### BUILD REPORTER #######################################
-############################################################################################
-  BuilderReport:
+
+  ################################# Reports #################################
+  # Reports should by run even if Builds_1/2 fail, so put them separatly in wf (not in Tests_1/2)
+  Builds_1_Report:
     # run report check for failed builds to indicate the CI error
-    if: ${{ !cancelled() }}
+    if: ${{ !cancelled() && contains(fromJson(needs.RunConfig.outputs.data).jobs_data.jobs_to_do, 'ClickHouse build check') }}
     needs:
       - RunConfig
-      - BuilderDebAarch64
-      - BuilderDebAsan
-      - BuilderDebDebug
-      - BuilderDebMsan
-      - BuilderDebRelease
-      - BuilderDebTsan
-      - BuilderDebUBsan
+      - Builds_1
     uses: ./.github/workflows/reusable_test.yml
     with:
       test_name: ClickHouse build check
-      runner_type: style-checker
+      runner_type: style-checker-aarch64
       data: ${{ needs.RunConfig.outputs.data }}
-  BuilderSpecialReport:
+  Builds_2_Report:
     # run report check for failed builds to indicate the CI error
-    if: ${{ !cancelled() }}
+    if: ${{ !cancelled() && contains(fromJson(needs.RunConfig.outputs.data).jobs_data.jobs_to_do, 'ClickHouse special build check') }}
     needs:
       - RunConfig
-      - BuilderBinAarch64
-      - BuilderBinDarwin
-      - BuilderBinDarwinAarch64
-      - BuilderBinFreeBSD
-      - BuilderBinPPC64
-      - BuilderBinRISCV64
-      - BuilderBinS390X
-      - BuilderBinAmd64Compat
-      - BuilderBinAarch64V80Compat
-      - BuilderBinClangTidy
-      - BuilderDebReleaseCoverage
-      - BuilderBinRelease
+      - Builds_2
     uses: ./.github/workflows/reusable_test.yml
     with:
       test_name: ClickHouse special build check
-      runner_type: style-checker
-      data: ${{ needs.RunConfig.outputs.data }}
-############################################################################################
-#################################### INSTALL PACKAGES ######################################
-############################################################################################
-  InstallPackagesTestRelease:
-    needs: [RunConfig, BuilderDebRelease]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Install packages (amd64)
-      runner_type: style-checker
-      data: ${{ needs.RunConfig.outputs.data }}
-      run_command: |
-        python3 install_check.py "$CHECK_NAME"
-  InstallPackagesTestAarch64:
-    needs: [RunConfig, BuilderDebAarch64]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Install packages (arm64)
       runner_type: style-checker-aarch64
       data: ${{ needs.RunConfig.outputs.data }}
-      run_command: |
-        python3 install_check.py "$CHECK_NAME"
-##############################################################################################
-########################### FUNCTIONAl STATELESS TESTS #######################################
-##############################################################################################
-  FunctionalStatelessTestRelease:
-    needs: [RunConfig, BuilderDebRelease]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Stateless tests (release)
-      runner_type: func-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-  FunctionalStatelessTestReleaseAnalyzerS3Replicated:
-    needs: [RunConfig, BuilderDebRelease]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Stateless tests (release, analyzer, s3, DatabaseReplicated)
-      runner_type: func-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-  FunctionalStatelessTestS3Debug:
-    needs: [RunConfig, BuilderDebDebug]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Stateless tests (debug, s3 storage)
-      runner_type: func-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-  FunctionalStatelessTestS3Tsan:
-    needs: [RunConfig, BuilderDebTsan]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Stateless tests (tsan, s3 storage)
-      runner_type: func-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-  FunctionalStatelessTestAarch64:
-    needs: [RunConfig, BuilderDebAarch64]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Stateless tests (aarch64)
-      runner_type: func-tester-aarch64
-      data: ${{ needs.RunConfig.outputs.data }}
-  FunctionalStatelessTestAsan:
-    needs: [RunConfig, BuilderDebAsan]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Stateless tests (asan)
-      runner_type: func-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-  FunctionalStatelessTestTsan:
-    needs: [RunConfig, BuilderDebTsan]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Stateless tests (tsan)
-      runner_type: func-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-  FunctionalStatelessTestMsan:
-    needs: [RunConfig, BuilderDebMsan]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Stateless tests (msan)
-      runner_type: func-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-  FunctionalStatelessTestUBsan:
-    needs: [RunConfig, BuilderDebUBsan]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Stateless tests (ubsan)
-      runner_type: func-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-  FunctionalStatelessTestDebug:
-    needs: [RunConfig, BuilderDebDebug]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Stateless tests (debug)
-      runner_type: func-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-  FunctionalStatelessTestFlakyCheck:
-    needs: [RunConfig, BuilderDebAsan]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Stateless tests flaky check (asan)
-      data: ${{ needs.RunConfig.outputs.data }}
-      runner_type: func-tester
-  TestsBugfixCheck:
-    needs: [RunConfig, StyleCheck]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Bugfix validation
-      runner_type: func-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-##############################################################################################
-############################ FUNCTIONAl STATEFUL TESTS #######################################
-##############################################################################################
-  FunctionalStatefulTestRelease:
-    needs: [RunConfig, BuilderDebRelease]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Stateful tests (release)
-      runner_type: func-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-  FunctionalStatefulTestAarch64:
-    needs: [RunConfig, BuilderDebAarch64]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Stateful tests (aarch64)
-      runner_type: func-tester-aarch64
-      data: ${{ needs.RunConfig.outputs.data }}
-  FunctionalStatefulTestAsan:
-    needs: [RunConfig, BuilderDebAsan]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Stateful tests (asan)
-      runner_type: func-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-  FunctionalStatefulTestTsan:
-    needs: [RunConfig, BuilderDebTsan]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Stateful tests (tsan)
-      runner_type: func-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-  FunctionalStatefulTestMsan:
-    needs: [RunConfig, BuilderDebMsan]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Stateful tests (msan)
-      runner_type: func-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-  FunctionalStatefulTestUBsan:
-    needs: [RunConfig, BuilderDebUBsan]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Stateful tests (ubsan)
-      runner_type: func-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-  FunctionalStatefulTestDebug:
-    needs: [RunConfig, BuilderDebDebug]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Stateful tests (debug)
-      runner_type: func-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-  # Parallel replicas
-  FunctionalStatefulTestDebugParallelReplicas:
-    needs: [RunConfig, BuilderDebDebug]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Stateful tests (debug, ParallelReplicas)
-      runner_type: func-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-  FunctionalStatefulTestUBsanParallelReplicas:
-    needs: [RunConfig, BuilderDebUBsan]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Stateful tests (ubsan, ParallelReplicas)
-      runner_type: func-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-  FunctionalStatefulTestMsanParallelReplicas:
-    needs: [RunConfig, BuilderDebMsan]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Stateful tests (msan, ParallelReplicas)
-      runner_type: func-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-  FunctionalStatefulTestTsanParallelReplicas:
-    needs: [RunConfig, BuilderDebTsan]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Stateful tests (tsan, ParallelReplicas)
-      runner_type: func-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-  FunctionalStatefulTestAsanParallelReplicas:
-    needs: [RunConfig, BuilderDebAsan]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Stateful tests (asan, ParallelReplicas)
-      runner_type: func-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-  FunctionalStatefulTestReleaseParallelReplicas:
-    needs: [RunConfig, BuilderDebRelease]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Stateful tests (release, ParallelReplicas)
-      runner_type: func-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-##############################################################################################
-########################### ClickBench #######################################################
-##############################################################################################
-  ClickBenchAMD64:
-    needs: [RunConfig, BuilderDebRelease]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: ClickBench (amd64)
-      runner_type: func-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-      run_command: |
-        python3 clickbench.py "$CHECK_NAME"
-  ClickBenchAarch64:
-    needs: [RunConfig, BuilderDebAarch64]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: ClickBench (aarch64)
-      runner_type: func-tester-aarch64
-      data: ${{ needs.RunConfig.outputs.data }}
-      run_command: |
-        python3 clickbench.py "$CHECK_NAME"
-##############################################################################################
-######################################### STRESS TESTS #######################################
-##############################################################################################
-  StressTestAsan:
-    needs: [RunConfig, BuilderDebAsan]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Stress test (asan)
-      runner_type: stress-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-  StressTestTsan:
-    needs: [RunConfig, BuilderDebTsan]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Stress test (tsan)
-      runner_type: stress-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-  StressTestMsan:
-    needs: [RunConfig, BuilderDebMsan]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Stress test (msan)
-      runner_type: stress-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-  StressTestUBsan:
-    needs: [RunConfig, BuilderDebUBsan]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Stress test (ubsan)
-      runner_type: stress-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-  StressTestDebug:
-    needs: [RunConfig, BuilderDebDebug]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Stress test (debug)
-      runner_type: stress-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-##############################################################################################
-######################################### UPGRADE CHECK ######################################
-##############################################################################################
-  UpgradeCheckAsan:
-    needs: [RunConfig, BuilderDebAsan]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Upgrade check (asan)
-      runner_type: stress-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-  UpgradeCheckTsan:
-    needs: [RunConfig, BuilderDebTsan]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Upgrade check (tsan)
-      runner_type: stress-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-  UpgradeCheckMsan:
-    needs: [RunConfig, BuilderDebMsan]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Upgrade check (msan)
-      runner_type: stress-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-  UpgradeCheckDebug:
-    needs: [RunConfig, BuilderDebDebug]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Upgrade check (debug)
-      runner_type: stress-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-##############################################################################################
-##################################### AST FUZZERS ############################################
-##############################################################################################
-  ASTFuzzerTestAsan:
-    needs: [RunConfig, BuilderDebAsan]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: AST fuzzer (asan)
-      runner_type: fuzzer-unit-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-  ASTFuzzerTestTsan:
-    needs: [RunConfig, BuilderDebTsan]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: AST fuzzer (tsan)
-      runner_type: fuzzer-unit-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-  ASTFuzzerTestUBSan:
-    needs: [RunConfig, BuilderDebUBsan]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: AST fuzzer (ubsan)
-      runner_type: fuzzer-unit-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-  ASTFuzzerTestMSan:
-    needs: [RunConfig, BuilderDebMsan]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: AST fuzzer (msan)
-      runner_type: fuzzer-unit-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-  ASTFuzzerTestDebug:
-    needs: [RunConfig, BuilderDebDebug]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: AST fuzzer (debug)
-      runner_type: fuzzer-unit-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-#############################################################################################
-############################# INTEGRATION TESTS #############################################
-#############################################################################################
-  IntegrationTestsAnalyzerAsan:
-    needs: [RunConfig, BuilderDebAsan]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Integration tests (asan, analyzer)
-      runner_type: stress-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-  IntegrationTestsTsan:
-    needs: [RunConfig, BuilderDebTsan]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Integration tests (tsan)
-      runner_type: stress-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-  IntegrationTestsAarch64:
-    needs: [RunConfig, BuilderDebAarch64]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Integration tests (aarch64)
-      # FIXME: there is no stress-tester for aarch64. func-tester-aarch64 is ok?
-      runner_type: func-tester-aarch64
-      data: ${{ needs.RunConfig.outputs.data }}
-  IntegrationTestsFlakyCheck:
-    needs: [RunConfig, BuilderDebAsan]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Integration tests flaky check (asan)
-      runner_type: stress-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-#############################################################################################
-#################################### UNIT TESTS #############################################
-#############################################################################################
-  UnitTestsAsan:
-    needs: [RunConfig, BuilderDebAsan]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Unit tests (asan)
-      runner_type: fuzzer-unit-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-  UnitTestsRelease:
-    needs: [RunConfig, BuilderBinRelease]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Unit tests (release)
-      runner_type: fuzzer-unit-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-  UnitTestsTsan:
-    needs: [RunConfig, BuilderDebTsan]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Unit tests (tsan)
-      runner_type: fuzzer-unit-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-  UnitTestsMsan:
-    needs: [RunConfig, BuilderDebMsan]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Unit tests (msan)
-      runner_type: fuzzer-unit-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-  UnitTestsUBsan:
-    needs: [RunConfig, BuilderDebUBsan]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Unit tests (ubsan)
-      runner_type: fuzzer-unit-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-#############################################################################################
-#################################### PERFORMANCE TESTS ######################################
-#############################################################################################
-  PerformanceComparisonX86:
-    needs: [RunConfig, BuilderDebRelease]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Performance Comparison
-      runner_type: stress-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-  PerformanceComparisonAarch:
-    needs: [RunConfig, BuilderDebAarch64]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Performance Comparison Aarch64
-      runner_type: func-tester-aarch64
-      data: ${{ needs.RunConfig.outputs.data }}
-##############################################################################################
-###################################### SQLANCER FUZZERS ######################################
-##############################################################################################
-  SQLancerTestRelease:
-    needs: [RunConfig, BuilderDebRelease]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: SQLancer (release)
-      runner_type: fuzzer-unit-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-  SQLancerTestDebug:
-    needs: [RunConfig, BuilderDebDebug]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: SQLancer (debug)
-      runner_type: fuzzer-unit-tester
-      data: ${{ needs.RunConfig.outputs.data }}
+
+  ################################# Stage Final #################################
+  #
   FinishCheck:
     if: ${{ !failure() && !cancelled() }}
-    needs:
-      - BuilderReport
-      - BuilderSpecialReport
-      - DocsCheck
-      - FastTest
-      - TestsBugfixCheck
-      - FunctionalStatelessTestDebug
-      - FunctionalStatelessTestRelease
-      - FunctionalStatelessTestAarch64
-      - FunctionalStatelessTestAsan
-      - FunctionalStatelessTestTsan
-      - FunctionalStatelessTestMsan
-      - FunctionalStatelessTestUBsan
-      - FunctionalStatefulTestDebug
-      - FunctionalStatefulTestRelease
-      - FunctionalStatefulTestAarch64
-      - FunctionalStatefulTestAsan
-      - FunctionalStatefulTestTsan
-      - FunctionalStatefulTestMsan
-      - FunctionalStatefulTestUBsan
-      - FunctionalStatelessTestS3Debug
-      - FunctionalStatelessTestS3Tsan
-      - FunctionalStatelessTestReleaseAnalyzerS3Replicated
-      - FunctionalStatefulTestReleaseParallelReplicas
-      - FunctionalStatefulTestAsanParallelReplicas
-      - FunctionalStatefulTestTsanParallelReplicas
-      - FunctionalStatefulTestMsanParallelReplicas
-      - FunctionalStatefulTestUBsanParallelReplicas
-      - FunctionalStatefulTestDebugParallelReplicas
-      - StressTestDebug
-      - StressTestAsan
-      - StressTestTsan
-      - StressTestMsan
-      - StressTestUBsan
-      - UpgradeCheckAsan
-      - UpgradeCheckTsan
-      - UpgradeCheckMsan
-      - UpgradeCheckDebug
-      - ASTFuzzerTestDebug
-      - ASTFuzzerTestAsan
-      - ASTFuzzerTestTsan
-      - ASTFuzzerTestMSan
-      - ASTFuzzerTestUBSan
-      - IntegrationTestsAnalyzerAsan
-      - IntegrationTestsTsan
-      - IntegrationTestsAarch64
-      - IntegrationTestsFlakyCheck
-      - PerformanceComparisonX86
-      - PerformanceComparisonAarch
-      - UnitTestsAsan
-      - UnitTestsTsan
-      - UnitTestsMsan
-      - UnitTestsUBsan
-      - UnitTestsRelease
-      - CompatibilityCheckX86
-      - CompatibilityCheckAarch64
-      - SQLancerTestRelease
-      - SQLancerTestDebug
+    needs: [Tests_1, Tests_2]
     runs-on: [self-hosted, style-checker]
     steps:
       - name: Check out repository code
         uses: ClickHouse/checkout@v1
-        with:
-          clear-repository: true
       - name: Finish label
         run: |
           cd "$GITHUB_WORKSPACE/tests/ci"
           python3 finish_check.py
-          python3 merge_pr.py --check-approved
-##############################################################################################
-############################ SQLLOGIC TEST ###################################################
-##############################################################################################
-  SQLLogicTestRelease:
-    needs: [RunConfig, BuilderDebRelease]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: Sqllogic test (release)
-      runner_type: func-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-##############################################################################################
-##################################### SQL TEST ###############################################
-##############################################################################################
-  SQLTest:
-    needs: [RunConfig, BuilderDebRelease]
-    if: ${{ !failure() && !cancelled() }}
-    uses: ./.github/workflows/reusable_test.yml
-    with:
-      test_name: SQLTest
-      runner_type: fuzzer-unit-tester
-      data: ${{ needs.RunConfig.outputs.data }}
-#############################################################################################
-###################################### NOT IN FINISH ########################################
+
+
 #############################################################################################
 ###################################### JEPSEN TESTS #########################################
 #############################################################################################
@@ -931,19 +184,11 @@ jobs:
     # we need concurrency as the job uses dedicated instances in the cloud
     concurrency:
       group: jepsen
-    if: ${{ !failure() && !cancelled() }}
-    needs: [RunConfig, BuilderBinRelease]
+    if: ${{ !failure() && !cancelled() && contains(fromJson(needs.RunConfig.outputs.data).jobs_data.jobs_to_do, 'ClickHouse Keeper Jepsen') }}
+    # jepsen needs binary_release build which is in Builds_2
+    needs: [RunConfig, Builds_2]
     uses: ./.github/workflows/reusable_test.yml
     with:
       test_name: ClickHouse Keeper Jepsen
       runner_type: style-checker
       data: ${{ needs.RunConfig.outputs.data }}
-#############################################################################################
-####################################### libFuzzer ###########################################
-#############################################################################################
-  libFuzzer:
-    if: ${{ !failure() && !cancelled() }}
-    needs: [RunConfig, StyleCheck]
-    uses: ./.github/workflows/libfuzzer.yml
-    with:
-      data: ${{ needs.RunConfig.outputs.data }}

.github/workflows/release_branches.yml

@@ -206,7 +206,7 @@ jobs:
       runner_type: style-checker-aarch64
       data: ${{ needs.RunConfig.outputs.data }}
   MarkReleaseReady:
-    if: ${{ ! (contains(needs.*.result, 'skipped') || contains(needs.*.result, 'failure')) }}
+    if: ${{ !failure() && !cancelled() }}
     needs:
       - BuilderBinDarwin
       - BuilderBinDarwinAarch64
@@ -214,9 +214,25 @@ jobs:
       - BuilderDebAarch64
     runs-on: [self-hosted, style-checker-aarch64]
     steps:
+      - name: Debug
+        run: |
+          echo need with different filters
+          cat << 'EOF'
+          ${{ toJSON(needs) }}
+          ${{ toJSON(needs.*.result) }}
+          no failures ${{ !contains(needs.*.result, 'failure') }}
+          no skips ${{ !contains(needs.*.result, 'skipped') }}
+          no both ${{ !(contains(needs.*.result, 'skipped') || contains(needs.*.result, 'failure')) }}
+          EOF
+      - name: Not ready
+        # fail the job to be able restart it
+        if: ${{ contains(needs.*.result, 'skipped') || contains(needs.*.result, 'failure') }}
+        run: exit 1
       - name: Check out repository code
+        if: ${{ ! (contains(needs.*.result, 'skipped') || contains(needs.*.result, 'failure')) }}
         uses: ClickHouse/checkout@v1
       - name: Mark Commit Release Ready
+        if: ${{ ! (contains(needs.*.result, 'skipped') || contains(needs.*.result, 'failure')) }}
         run: |
           cd "$GITHUB_WORKSPACE/tests/ci"
           python3 mark_release_ready.py

.github/workflows/reusable_build.yml

@@ -43,7 +43,8 @@ jobs:
     runs-on: [self-hosted, '${{inputs.runner_type}}']
     steps:
       - name: Check out repository code
-        uses: ClickHouse/checkout@v1
+        # WIP: temporary try commit with limited perallelization of checkout
+        uses: ClickHouse/checkout@0be3f7b3098bae494d3ef5d29d2e0676fb606232
         with:
           clear-repository: true
           ref: ${{ fromJson(inputs.data).git_ref }}

.github/workflows/reusable_build_stage.yml

@@ -0,0 +1,32 @@
+### FIXME: merge reusable_test.yml and reusable_build.yml as they are almost identical
+#          and then merge reusable_build_stage.yml and reusable_test_stage.yml
+
+name: BuildStageWF
+'on':
+  workflow_call:
+    inputs:
+      stage:
+        description: stage name
+        type: string
+        required: true
+      data:
+        description: ci data
+        type: string
+        required: true
+
+jobs:
+  s:
+    if: ${{ !failure() && !cancelled() }}
+    strategy:
+      fail-fast: false
+      matrix:
+        job_name_and_runner_type: ${{ fromJson(inputs.data).stages_data[inputs.stage] }}
+    uses: ./.github/workflows/reusable_build.yml
+    with:
+      build_name: ${{ matrix.job_name_and_runner_type.job_name }}
+      runner_type: ${{ matrix.job_name_and_runner_type.runner_type }}
+      # don't forget to pass force flag (no ci cache/no reuse) - once it's needed
+      force: false
+      # for now let's do I deep checkout for builds
+      checkout_depth: 0
+      data: ${{ inputs.data }}

.github/workflows/reusable_test_stage.yml

@@ -0,0 +1,25 @@
+name: StageWF
+'on':
+  workflow_call:
+    inputs:
+      stage:
+        description: stage name
+        type: string
+        required: true
+      data:
+        description: ci data
+        type: string
+        required: true
+
+jobs:
+  s:
+    if: ${{ !failure() && !cancelled() }}
+    strategy:
+      fail-fast: false
+      matrix:
+        job_name_and_runner_type: ${{ fromJson(inputs.data).stages_data[inputs.stage] }}
+    uses: ./.github/workflows/reusable_test.yml
+    with:
+      test_name: ${{ matrix.job_name_and_runner_type.job_name }}
+      runner_type: ${{ matrix.job_name_and_runner_type.runner_type }}
+      data: ${{ inputs.data }}

.gitignore

@@ -165,7 +165,7 @@ tests/queries/0_stateless/*.expect.history
 tests/integration/**/_gen
 
 # rust
-/rust/**/target
+/rust/**/target*
 # It is autogenerated from *.in
 /rust/**/.cargo/config.toml
 /rust/**/vendor

CMakeLists.txt

@@ -319,7 +319,8 @@ if (COMPILER_CLANG)
     endif()
 endif ()
 
-set (COMPILER_FLAGS "${COMPILER_FLAGS}")
+# Disable floating-point expression contraction in order to get consistent floating point calculation results across platforms
+set (COMPILER_FLAGS "${COMPILER_FLAGS} -ffp-contract=off")
 
 # Our built-in unwinder only supports DWARF version up to 4.
 set (DEBUG_INFO_FLAGS "-g")

README.md

@@ -31,15 +31,30 @@ curl https://clickhouse.com/ | sh
 * [Static Analysis (SonarCloud)](https://sonarcloud.io/project/issues?resolved=false&id=ClickHouse_ClickHouse) proposes C++ quality improvements.
 * [Contacts](https://clickhouse.com/company/contact) can help to get your questions answered if there are any.
 
+## Monthly Release & Community Call
+
+Every month we get together with the community (users, contributors, customers, those interested in learning more about ClickHouse) to discuss what is coming in the latest release. If you are interested in sharing what you've built on ClickHouse, let us know.
+
+* [v24.3 Community Call](https://clickhouse.com/company/events/v24-3-community-release-call) - Mar 26
+* [v24.4 Community Call](https://clickhouse.com/company/events/v24-4-community-release-call) - Apr 30
+
 ## Upcoming Events
 
-Keep an eye out for upcoming meetups around the world. Somewhere else you want us to be? Please feel free to reach out to tyler `<at>` clickhouse `<dot>` com.
+Keep an eye out for upcoming meetups and eventsaround the world. Somewhere else you want us to be? Please feel free to reach out to tyler `<at>` clickhouse `<dot>` com. You can also peruse [ClickHouse Events](https://clickhouse.com/company/news-events) for a list of all upcoming trainings, meetups, speaking engagements, etc.
+
+* [ClickHouse Meetup in Bellevue](https://www.meetup.com/clickhouse-seattle-user-group/events/298650371/) - Mar 11
+* [ClickHouse Meetup at Ramp's Offices in NYC](https://www.meetup.com/clickhouse-new-york-user-group/events/298640542/) - Mar 19
+* [ClickHouse Melbourne Meetup](https://www.meetup.com/clickhouse-australia-user-group/events/299479750/) - Mar 20
+* [ClickHouse Meetup in Paris](https://www.meetup.com/clickhouse-france-user-group/events/298997115/) - Mar 21
+* [ClickHouse Meetup in Bengaluru](https://www.meetup.com/clickhouse-bangalore-user-group/events/299479850/) - Mar 23
+* [ClickHouse Meetup in Zurich](https://www.meetup.com/clickhouse-switzerland-meetup-group/events/299628922/) - Apr 16
+* [ClickHouse Meetup in Copenhagen](https://www.meetup.com/clickhouse-denmark-meetup-group/events/299629133/) - Apr 23
+* [ClickHouse Meetup in Dubai](https://www.meetup.com/clickhouse-dubai-meetup-group/events/299629189/) - May 28
+
 
 ## Recent Recordings
 * **Recent Meetup Videos**: [Meetup Playlist](https://www.youtube.com/playlist?list=PL0Z2YDlm0b3iNDUzpY1S3L_iV4nARda_U) Whenever possible recordings of the ClickHouse Community Meetups are edited and presented as individual talks. Current featuring "Modern SQL in 2023", "Fast, Concurrent, and Consistent Asynchronous INSERTS in ClickHouse", and "Full-Text Indices: Design and Experiments"
-* **Recording available**: [**v24.1 Release Webinar**](https://www.youtube.com/watch?v=pBF9g0wGAGs) All the features of 24.1, one convenient video! Watch it now!
-* **All release webinar recordings**: [YouTube playlist](https://www.youtube.com/playlist?list=PL0Z2YDlm0b3jAlSy1JxyP8zluvXaN3nxU)
-
+* **Recording available**: [**v24.2 Release Call**](https://www.youtube.com/watch?v=iN2y-TK8f3A) All the features of 24.2, one convenient video! Watch it now!
   
  ## Interested in joining ClickHouse and making it your full-time job? 
   

base/base/CMakeLists.txt

@@ -13,6 +13,7 @@ set (SRCS
     cgroupsv2.cpp
     coverage.cpp
     demangle.cpp
+    Decimal.cpp
     getAvailableMemoryAmount.cpp
     getFQDNOrHostName.cpp
     getMemoryAmount.cpp

base/base/Decimal.cpp

@@ -0,0 +1,87 @@
+#include <base/Decimal.h>
+#include <base/extended_types.h>
+
+namespace DB
+{
+
+/// Explicit template instantiations.
+
+#define FOR_EACH_UNDERLYING_DECIMAL_TYPE(M) \
+    M(Int32)  \
+    M(Int64)  \
+    M(Int128) \
+    M(Int256)
+
+#define FOR_EACH_UNDERLYING_DECIMAL_TYPE_PASS(M, X) \
+    M(Int32, X) \
+    M(Int64, X) \
+    M(Int128, X) \
+    M(Int256, X)
+
+template <typename T> const Decimal<T> & Decimal<T>::operator += (const T & x) { value += x; return *this; }
+template <typename T> const Decimal<T> & Decimal<T>::operator -= (const T & x) { value -= x; return *this; }
+template <typename T> const Decimal<T> & Decimal<T>::operator *= (const T & x) { value *= x; return *this; }
+template <typename T> const Decimal<T> & Decimal<T>::operator /= (const T & x) { value /= x; return *this; }
+template <typename T> const Decimal<T> & Decimal<T>::operator %= (const T & x) { value %= x; return *this; }
+
+template <typename T> void NO_SANITIZE_UNDEFINED Decimal<T>::addOverflow(const T & x) { value += x; }
+
+/// Maybe this explicit instantiation affects performance since operators cannot be inlined.
+
+template <typename T> template <typename U> const Decimal<T> & Decimal<T>::operator += (const Decimal<U> & x) { value += static_cast<T>(x.value); return *this; }
+template <typename T> template <typename U> const Decimal<T> & Decimal<T>::operator -= (const Decimal<U> & x) { value -= static_cast<T>(x.value); return *this; }
+template <typename T> template <typename U> const Decimal<T> & Decimal<T>::operator *= (const Decimal<U> & x) { value *= static_cast<T>(x.value); return *this; }
+template <typename T> template <typename U> const Decimal<T> & Decimal<T>::operator /= (const Decimal<U> & x) { value /= static_cast<T>(x.value); return *this; }
+template <typename T> template <typename U> const Decimal<T> & Decimal<T>::operator %= (const Decimal<U> & x) { value %= static_cast<T>(x.value); return *this; }
+
+#define DISPATCH(TYPE_T, TYPE_U) \
+    template const Decimal<TYPE_T> & Decimal<TYPE_T>::operator += (const Decimal<TYPE_U> & x); \
+    template const Decimal<TYPE_T> & Decimal<TYPE_T>::operator -= (const Decimal<TYPE_U> & x); \
+    template const Decimal<TYPE_T> & Decimal<TYPE_T>::operator *= (const Decimal<TYPE_U> & x); \
+    template const Decimal<TYPE_T> & Decimal<TYPE_T>::operator /= (const Decimal<TYPE_U> & x); \
+    template const Decimal<TYPE_T> & Decimal<TYPE_T>::operator %= (const Decimal<TYPE_U> & x);
+#define INVOKE(X) FOR_EACH_UNDERLYING_DECIMAL_TYPE_PASS(DISPATCH, X)
+FOR_EACH_UNDERLYING_DECIMAL_TYPE(INVOKE);
+#undef INVOKE
+#undef DISPATCH
+
+#define DISPATCH(TYPE) template struct Decimal<TYPE>;
+FOR_EACH_UNDERLYING_DECIMAL_TYPE(DISPATCH)
+#undef DISPATCH
+
+template <typename T> bool operator< (const Decimal<T> & x, const Decimal<T> & y) { return x.value < y.value; }
+template <typename T> bool operator> (const Decimal<T> & x, const Decimal<T> & y) { return x.value > y.value; }
+template <typename T> bool operator<= (const Decimal<T> & x, const Decimal<T> & y) { return x.value <= y.value; }
+template <typename T> bool operator>= (const Decimal<T> & x, const Decimal<T> & y) { return x.value >= y.value; }
+template <typename T> bool operator== (const Decimal<T> & x, const Decimal<T> & y) { return x.value == y.value; }
+template <typename T> bool operator!= (const Decimal<T> & x, const Decimal<T> & y) { return x.value != y.value; }
+
+#define DISPATCH(TYPE) \
+template bool operator< (const Decimal<TYPE> & x, const Decimal<TYPE> & y); \
+template bool operator> (const Decimal<TYPE> & x, const Decimal<TYPE> & y); \
+template bool operator<= (const Decimal<TYPE> & x, const Decimal<TYPE> & y); \
+template bool operator>= (const Decimal<TYPE> & x, const Decimal<TYPE> & y); \
+template bool operator== (const Decimal<TYPE> & x, const Decimal<TYPE> & y); \
+template bool operator!= (const Decimal<TYPE> & x, const Decimal<TYPE> & y);
+FOR_EACH_UNDERLYING_DECIMAL_TYPE(DISPATCH)
+#undef DISPATCH
+
+
+template <typename T> Decimal<T> operator+ (const Decimal<T> & x, const Decimal<T> & y) { return x.value + y.value; }
+template <typename T> Decimal<T> operator- (const Decimal<T> & x, const Decimal<T> & y) { return x.value - y.value; }
+template <typename T> Decimal<T> operator* (const Decimal<T> & x, const Decimal<T> & y) { return x.value * y.value; }
+template <typename T> Decimal<T> operator/ (const Decimal<T> & x, const Decimal<T> & y) { return x.value / y.value; }
+template <typename T> Decimal<T> operator- (const Decimal<T> & x) { return -x.value; }
+
+#define DISPATCH(TYPE) \
+template Decimal<TYPE> operator+ (const Decimal<TYPE> & x, const Decimal<TYPE> & y); \
+template Decimal<TYPE> operator- (const Decimal<TYPE> & x, const Decimal<TYPE> & y); \
+template Decimal<TYPE> operator* (const Decimal<TYPE> & x, const Decimal<TYPE> & y); \
+template Decimal<TYPE> operator/ (const Decimal<TYPE> & x, const Decimal<TYPE> & y); \
+template Decimal<TYPE> operator- (const Decimal<TYPE> & x);
+FOR_EACH_UNDERLYING_DECIMAL_TYPE(DISPATCH)
+#undef DISPATCH
+
+#undef FOR_EACH_UNDERLYING_DECIMAL_TYPE_PASS
+#undef FOR_EACH_UNDERLYING_DECIMAL_TYPE
+}

base/base/Decimal.h

@@ -2,6 +2,7 @@
 
 #include <base/extended_types.h>
 #include <base/Decimal_fwd.h>
+#include <base/types.h>
 #include <base/defines.h>
 
 
@@ -10,6 +11,18 @@ namespace DB
 template <class> struct Decimal;
 class DateTime64;
 
+#define FOR_EACH_UNDERLYING_DECIMAL_TYPE(M) \
+    M(Int32) \
+    M(Int64) \
+    M(Int128) \
+    M(Int256)
+
+#define FOR_EACH_UNDERLYING_DECIMAL_TYPE_PASS(M, X) \
+    M(Int32,  X) \
+    M(Int64,  X) \
+    M(Int128, X) \
+    M(Int256, X)
+
 using Decimal32 = Decimal<Int32>;
 using Decimal64 = Decimal<Int64>;
 using Decimal128 = Decimal<Int128>;
@@ -50,36 +63,73 @@ struct Decimal
             return static_cast<U>(value);
     }
 
-    const Decimal<T> & operator += (const T & x) { value += x; return *this; }
-    const Decimal<T> & operator -= (const T & x) { value -= x; return *this; }
-    const Decimal<T> & operator *= (const T & x) { value *= x; return *this; }
-    const Decimal<T> & operator /= (const T & x) { value /= x; return *this; }
-    const Decimal<T> & operator %= (const T & x) { value %= x; return *this; }
+    const Decimal<T> & operator += (const T & x);
+    const Decimal<T> & operator -= (const T & x);
+    const Decimal<T> & operator *= (const T & x);
+    const Decimal<T> & operator /= (const T & x);
+    const Decimal<T> & operator %= (const T & x);
 
-    template <typename U> const Decimal<T> & operator += (const Decimal<U> & x) { value += x.value; return *this; }
-    template <typename U> const Decimal<T> & operator -= (const Decimal<U> & x) { value -= x.value; return *this; }
-    template <typename U> const Decimal<T> & operator *= (const Decimal<U> & x) { value *= x.value; return *this; }
-    template <typename U> const Decimal<T> & operator /= (const Decimal<U> & x) { value /= x.value; return *this; }
-    template <typename U> const Decimal<T> & operator %= (const Decimal<U> & x) { value %= x.value; return *this; }
+    template <typename U> const Decimal<T> & operator += (const Decimal<U> & x);
+    template <typename U> const Decimal<T> & operator -= (const Decimal<U> & x);
+    template <typename U> const Decimal<T> & operator *= (const Decimal<U> & x);
+    template <typename U> const Decimal<T> & operator /= (const Decimal<U> & x);
+    template <typename U> const Decimal<T> & operator %= (const Decimal<U> & x);
 
     /// This is to avoid UB for sumWithOverflow()
-    void NO_SANITIZE_UNDEFINED addOverflow(const T & x) { value += x; }
+    void NO_SANITIZE_UNDEFINED addOverflow(const T & x);
 
     T value;
 };
 
-template <typename T> inline bool operator< (const Decimal<T> & x, const Decimal<T> & y) { return x.value < y.value; }
-template <typename T> inline bool operator> (const Decimal<T> & x, const Decimal<T> & y) { return x.value > y.value; }
-template <typename T> inline bool operator<= (const Decimal<T> & x, const Decimal<T> & y) { return x.value <= y.value; }
-template <typename T> inline bool operator>= (const Decimal<T> & x, const Decimal<T> & y) { return x.value >= y.value; }
-template <typename T> inline bool operator== (const Decimal<T> & x, const Decimal<T> & y) { return x.value == y.value; }
-template <typename T> inline bool operator!= (const Decimal<T> & x, const Decimal<T> & y) { return x.value != y.value; }
+#define DISPATCH(TYPE) extern template struct Decimal<TYPE>;
+FOR_EACH_UNDERLYING_DECIMAL_TYPE(DISPATCH)
+#undef DISPATCH
 
-template <typename T> inline Decimal<T> operator+ (const Decimal<T> & x, const Decimal<T> & y) { return x.value + y.value; }
-template <typename T> inline Decimal<T> operator- (const Decimal<T> & x, const Decimal<T> & y) { return x.value - y.value; }
-template <typename T> inline Decimal<T> operator* (const Decimal<T> & x, const Decimal<T> & y) { return x.value * y.value; }
-template <typename T> inline Decimal<T> operator/ (const Decimal<T> & x, const Decimal<T> & y) { return x.value / y.value; }
-template <typename T> inline Decimal<T> operator- (const Decimal<T> & x) { return -x.value; }
+#define DISPATCH(TYPE_T, TYPE_U) \
+    extern template const Decimal<TYPE_T> & Decimal<TYPE_T>::operator += (const Decimal<TYPE_U> & x); \
+    extern template const Decimal<TYPE_T> & Decimal<TYPE_T>::operator -= (const Decimal<TYPE_U> & x); \
+    extern template const Decimal<TYPE_T> & Decimal<TYPE_T>::operator *= (const Decimal<TYPE_U> & x); \
+    extern template const Decimal<TYPE_T> & Decimal<TYPE_T>::operator /= (const Decimal<TYPE_U> & x); \
+    extern template const Decimal<TYPE_T> & Decimal<TYPE_T>::operator %= (const Decimal<TYPE_U> & x);
+#define INVOKE(X) FOR_EACH_UNDERLYING_DECIMAL_TYPE_PASS(DISPATCH, X)
+FOR_EACH_UNDERLYING_DECIMAL_TYPE(INVOKE);
+#undef INVOKE
+#undef DISPATCH
+
+template <typename T> bool operator< (const Decimal<T> & x, const Decimal<T> & y);
+template <typename T> bool operator> (const Decimal<T> & x, const Decimal<T> & y);
+template <typename T> bool operator<= (const Decimal<T> & x, const Decimal<T> & y);
+template <typename T> bool operator>= (const Decimal<T> & x, const Decimal<T> & y);
+template <typename T> bool operator== (const Decimal<T> & x, const Decimal<T> & y);
+template <typename T> bool operator!= (const Decimal<T> & x, const Decimal<T> & y);
+
+#define DISPATCH(TYPE) \
+extern template bool operator< (const Decimal<TYPE> & x, const Decimal<TYPE> & y); \
+extern template bool operator> (const Decimal<TYPE> & x, const Decimal<TYPE> & y); \
+extern template bool operator<= (const Decimal<TYPE> & x, const Decimal<TYPE> & y); \
+extern template bool operator>= (const Decimal<TYPE> & x, const Decimal<TYPE> & y); \
+extern template bool operator== (const Decimal<TYPE> & x, const Decimal<TYPE> & y); \
+extern template bool operator!= (const Decimal<TYPE> & x, const Decimal<TYPE> & y);
+FOR_EACH_UNDERLYING_DECIMAL_TYPE(DISPATCH)
+#undef DISPATCH
+
+template <typename T> Decimal<T> operator+ (const Decimal<T> & x, const Decimal<T> & y);
+template <typename T> Decimal<T> operator- (const Decimal<T> & x, const Decimal<T> & y);
+template <typename T> Decimal<T> operator* (const Decimal<T> & x, const Decimal<T> & y);
+template <typename T> Decimal<T> operator/ (const Decimal<T> & x, const Decimal<T> & y);
+template <typename T> Decimal<T> operator- (const Decimal<T> & x);
+
+#define DISPATCH(TYPE) \
+extern template Decimal<TYPE> operator+ (const Decimal<TYPE> & x, const Decimal<TYPE> & y); \
+extern template Decimal<TYPE> operator- (const Decimal<TYPE> & x, const Decimal<TYPE> & y); \
+extern template Decimal<TYPE> operator* (const Decimal<TYPE> & x, const Decimal<TYPE> & y); \
+extern template Decimal<TYPE> operator/ (const Decimal<TYPE> & x, const Decimal<TYPE> & y); \
+extern template Decimal<TYPE> operator- (const Decimal<TYPE> & x);
+FOR_EACH_UNDERLYING_DECIMAL_TYPE(DISPATCH)
+#undef DISPATCH
+
+#undef FOR_EACH_UNDERLYING_DECIMAL_TYPE_PASS
+#undef FOR_EACH_UNDERLYING_DECIMAL_TYPE
 
 /// Distinguishable type to allow function resolution/deduction based on value type,
 /// but also relatively easy to convert to/from Decimal64.

base/base/extended_types.h

@@ -64,6 +64,44 @@ template <> struct is_arithmetic<UInt256> { static constexpr bool value = true;
 template <typename T>
 inline constexpr bool is_arithmetic_v = is_arithmetic<T>::value;
 
+#define FOR_EACH_ARITHMETIC_TYPE(M) \
+    M(DataTypeDate) \
+    M(DataTypeDate32) \
+    M(DataTypeDateTime) \
+    M(DataTypeInt8) \
+    M(DataTypeUInt8) \
+    M(DataTypeInt16) \
+    M(DataTypeUInt16) \
+    M(DataTypeInt32) \
+    M(DataTypeUInt32) \
+    M(DataTypeInt64) \
+    M(DataTypeUInt64) \
+    M(DataTypeInt128) \
+    M(DataTypeUInt128) \
+    M(DataTypeInt256) \
+    M(DataTypeUInt256) \
+    M(DataTypeFloat32) \
+    M(DataTypeFloat64)
+
+#define FOR_EACH_ARITHMETIC_TYPE_PASS(M, X) \
+    M(DataTypeDate, X) \
+    M(DataTypeDate32, X) \
+    M(DataTypeDateTime, X) \
+    M(DataTypeInt8, X) \
+    M(DataTypeUInt8, X) \
+    M(DataTypeInt16, X) \
+    M(DataTypeUInt16, X) \
+    M(DataTypeInt32, X) \
+    M(DataTypeUInt32, X) \
+    M(DataTypeInt64, X) \
+    M(DataTypeUInt64, X) \
+    M(DataTypeInt128, X) \
+    M(DataTypeUInt128, X) \
+    M(DataTypeInt256, X) \
+    M(DataTypeUInt256, X) \
+    M(DataTypeFloat32, X) \
+    M(DataTypeFloat64, X)
+
 template <typename T>
 struct make_unsigned // NOLINT(readability-identifier-naming)
 {

base/poco/Net/include/Poco/Net/HTTPChunkedStream.h

@@ -45,6 +45,8 @@ namespace Net
         ~HTTPChunkedStreamBuf();
         void close();
 
+        bool isComplete() const { return _chunk == std::char_traits<char>::eof(); }
+
     protected:
         int readFromDevice(char * buffer, std::streamsize length);
         int writeToDevice(const char * buffer, std::streamsize length);
@@ -68,6 +70,8 @@ namespace Net
         ~HTTPChunkedIOS();
         HTTPChunkedStreamBuf * rdbuf();
 
+        bool isComplete() const { return _buf.isComplete(); }
+
     protected:
         HTTPChunkedStreamBuf _buf;
     };

base/poco/Net/include/Poco/Net/HTTPClientSession.h

@@ -210,7 +210,7 @@ namespace Net
         void setKeepAliveTimeout(const Poco::Timespan & timeout);
         /// Sets the connection timeout for HTTP connections.
 
-        const Poco::Timespan & getKeepAliveTimeout() const;
+        Poco::Timespan getKeepAliveTimeout() const;
         /// Returns the connection timeout for HTTP connections.
 
         virtual std::ostream & sendRequest(HTTPRequest & request);
@@ -275,7 +275,7 @@ namespace Net
         /// This method should only be called if the request contains
         /// a "Expect: 100-continue" header.
 
-        void flushRequest();
+        virtual void flushRequest();
         /// Flushes the request stream.
         ///
         /// Normally this method does not need to be called.
@@ -283,7 +283,7 @@ namespace Net
         /// fully sent if receiveResponse() is not called, e.g.,
         /// because the underlying socket will be detached.
 
-        void reset();
+        virtual void reset();
         /// Resets the session and closes the socket.
         ///
         /// The next request will initiate a new connection,
@@ -303,6 +303,9 @@ namespace Net
         /// Returns true if the proxy should be bypassed
         /// for the current host.
 
+        const Poco::Timestamp & getLastRequest() const;
+        /// Returns time when connection has been used last time
+
     protected:
         enum
         {
@@ -338,6 +341,10 @@ namespace Net
         /// Calls proxyConnect() and attaches the resulting StreamSocket
         /// to the HTTPClientSession.
 
+        void setLastRequest(Poco::Timestamp time);
+
+        void assign(HTTPClientSession & session);
+
         HTTPSessionFactory _proxySessionFactory;
         /// Factory to create HTTPClientSession to proxy.
     private:
@@ -433,11 +440,20 @@ namespace Net
     }
 
 
-    inline const Poco::Timespan & HTTPClientSession::getKeepAliveTimeout() const
+    inline Poco::Timespan HTTPClientSession::getKeepAliveTimeout() const
     {
         return _keepAliveTimeout;
     }
 
+    inline const Poco::Timestamp & HTTPClientSession::getLastRequest() const
+    {
+        return _lastRequest;
+    }
+
+    inline void HTTPClientSession::setLastRequest(Poco::Timestamp time)
+    {
+        _lastRequest = time;
+    }
 
 }
 } // namespace Poco::Net

base/poco/Net/include/Poco/Net/HTTPFixedLengthStream.h

@@ -48,6 +48,8 @@ namespace Net
         HTTPFixedLengthStreamBuf(HTTPSession & session, ContentLength length, openmode mode);
         ~HTTPFixedLengthStreamBuf();
 
+        bool isComplete() const;
+
     protected:
         int readFromDevice(char * buffer, std::streamsize length);
         int writeToDevice(const char * buffer, std::streamsize length);
@@ -67,6 +69,8 @@ namespace Net
         ~HTTPFixedLengthIOS();
         HTTPFixedLengthStreamBuf * rdbuf();
 
+        bool isComplete() const { return _buf.isComplete(); }
+
     protected:
         HTTPFixedLengthStreamBuf _buf;
     };

base/poco/Net/include/Poco/Net/HTTPSession.h

@@ -64,6 +64,15 @@ namespace Net
         Poco::Timespan getTimeout() const;
         /// Returns the timeout for the HTTP session.
 
+        Poco::Timespan getConnectionTimeout() const;
+        /// Returns connection timeout for the HTTP session.
+
+        Poco::Timespan getSendTimeout() const;
+        /// Returns send timeout for the HTTP session.
+
+        Poco::Timespan getReceiveTimeout() const;
+        /// Returns receive timeout for the HTTP session.
+
         bool connected() const;
         /// Returns true if the underlying socket is connected.
 
@@ -217,12 +226,25 @@ namespace Net
         return _keepAlive;
     }
 
-
     inline Poco::Timespan HTTPSession::getTimeout() const
     {
         return _receiveTimeout;
     }
 
+    inline Poco::Timespan HTTPSession::getConnectionTimeout() const
+    {
+        return _connectionTimeout;
+    }
+
+    inline Poco::Timespan HTTPSession::getSendTimeout() const
+    {
+        return _sendTimeout;
+    }
+
+    inline Poco::Timespan HTTPSession::getReceiveTimeout() const
+    {
+        return _receiveTimeout;
+    }
 
     inline StreamSocket & HTTPSession::socket()
     {

base/poco/Net/include/Poco/Net/HTTPStream.h

@@ -63,6 +63,8 @@ namespace Net
         ~HTTPIOS();
         HTTPStreamBuf * rdbuf();
 
+        bool isComplete() const { return false; }
+
     protected:
         HTTPStreamBuf _buf;
     };

base/poco/Net/src/HTTPChunkedStream.cpp

@@ -49,10 +49,12 @@ HTTPChunkedStreamBuf::~HTTPChunkedStreamBuf()
 
 void HTTPChunkedStreamBuf::close()
 {
-	if (_mode & std::ios::out)
+	if (_mode & std::ios::out && _chunk != std::char_traits<char>::eof())
 	{
 		sync();
 		_session.write("0\r\n\r\n", 5);
+
+        _chunk = std::char_traits<char>::eof();
 	}
 }
 

base/poco/Net/src/HTTPClientSession.cpp

@@ -227,7 +227,7 @@ void HTTPClientSession::setKeepAliveTimeout(const Poco::Timespan& timeout)
 std::ostream& HTTPClientSession::sendRequest(HTTPRequest& request)
 {
 	_pRequestStream = 0;
-	_pResponseStream = 0;
+    _pResponseStream = 0;
 	clearException();
 	_responseReceived = false;
 
@@ -501,5 +501,26 @@ bool HTTPClientSession::bypassProxy() const
 	else return false;
 }
 
+void HTTPClientSession::assign(Poco::Net::HTTPClientSession & session)
+{
+    poco_assert (this != &session);
+
+    if (session.buffered())
+        throw Poco::LogicException("assign a session with not empty buffered data");
+
+    if (buffered())
+        throw Poco::LogicException("assign to a session with not empty buffered data");
+
+    attachSocket(session.detachSocket());
+    setLastRequest(session.getLastRequest());
+    setResolvedHost(session.getResolvedHost());
+    setKeepAlive(session.getKeepAlive());
+
+    setTimeout(session.getConnectionTimeout(), session.getSendTimeout(), session.getReceiveTimeout());
+    setKeepAliveTimeout(session.getKeepAliveTimeout());
+    setProxyConfig(session.getProxyConfig());
+
+    session.reset();
+}
 
 } } // namespace Poco::Net

base/poco/Net/src/HTTPFixedLengthStream.cpp

@@ -43,6 +43,12 @@ HTTPFixedLengthStreamBuf::~HTTPFixedLengthStreamBuf()
 }
 
 
+bool HTTPFixedLengthStreamBuf::isComplete() const
+{
+    return _count == _length;
+}
+
+
 int HTTPFixedLengthStreamBuf::readFromDevice(char* buffer, std::streamsize length)
 {
 	int n = 0;

contrib/aws

@@ -1 +1 @@
-Subproject commit 5f0542b3ad7eef25b0540d37d778207e0345ea8f
+Subproject commit 32870e234cac03e0ac46370c26858b0ffdf14200

docker/images.json

@@ -1,8 +1,12 @@
 {
-    "docker/packager/binary": {
+    "docker/packager/binary-builder": {
         "name": "clickhouse/binary-builder",
         "dependent": []
     },
+    "docker/packager/cctools": {
+        "name": "clickhouse/cctools",
+        "dependent": []
+    },
     "docker/test/compatibility/centos": {
         "name": "clickhouse/test-old-centos",
         "dependent": []
@@ -30,7 +34,6 @@
     "docker/test/util": {
         "name": "clickhouse/test-util",
         "dependent": [
-            "docker/packager/binary",
             "docker/test/base",
             "docker/test/fasttest"
         ]
@@ -67,7 +70,9 @@
     },
     "docker/test/fasttest": {
         "name": "clickhouse/fasttest",
-        "dependent": []
+        "dependent": [
+            "docker/packager/binary-builder"
+        ]
     },
     "docker/test/style": {
         "name": "clickhouse/style-test",

docker/packager/binary-builder/Dockerfile

@@ -1,43 +1,6 @@
 # docker build -t clickhouse/binary-builder .
 ARG FROM_TAG=latest
-FROM clickhouse/test-util:latest AS cctools
-# The cctools are built always from the clickhouse/test-util:latest and cached inline
-# Theoretically, it should improve rebuild speed significantly
-ENV CC=clang-${LLVM_VERSION}
-ENV CXX=clang++-${LLVM_VERSION}
-# !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
-# DO NOT PUT ANYTHING BEFORE THE NEXT TWO `RUN` DIRECTIVES
-# THE MOST HEAVY OPERATION MUST BE THE FIRST IN THE CACHE
-# !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
-# libtapi is required to support .tbh format from recent MacOS SDKs
-RUN git clone https://github.com/tpoechtrager/apple-libtapi.git \
-    && cd apple-libtapi \
-    && git checkout 15dfc2a8c9a2a89d06ff227560a69f5265b692f9 \
-    && INSTALLPREFIX=/cctools ./build.sh \
-    && ./install.sh \
-    && cd .. \
-    && rm -rf apple-libtapi
-
-# Build and install tools for cross-linking to Darwin (x86-64)
-# Build and install tools for cross-linking to Darwin (aarch64)
-RUN git clone https://github.com/tpoechtrager/cctools-port.git \
-    && cd cctools-port/cctools \
-    && git checkout 2a3e1c2a6ff54a30f898b70cfb9ba1692a55fad7 \
-    && ./configure --prefix=/cctools --with-libtapi=/cctools \
-        --target=x86_64-apple-darwin \
-    && make install -j$(nproc) \
-    && make clean \
-    && ./configure --prefix=/cctools --with-libtapi=/cctools \
-        --target=aarch64-apple-darwin \
-    && make install -j$(nproc) \
-    && cd ../.. \
-    && rm -rf cctools-port
-
-# !!!!!!!!!!!
-# END COMPILE
-# !!!!!!!!!!!
-
-FROM clickhouse/test-util:$FROM_TAG
+FROM clickhouse/fasttest:$FROM_TAG
 ENV CC=clang-${LLVM_VERSION}
 ENV CXX=clang++-${LLVM_VERSION}
 
@@ -110,7 +73,8 @@ RUN curl -Lo /usr/bin/clang-tidy-cache \
         "https://raw.githubusercontent.com/matus-chochlik/ctcache/$CLANG_TIDY_SHA1/clang-tidy-cache" \
     && chmod +x /usr/bin/clang-tidy-cache
 
-COPY --from=cctools /cctools /cctools
+# If the cctools is updated, then first build it in the CI, then update here in a different commit
+COPY --from=clickhouse/cctools:5a908f73878a /cctools /cctools
 
 RUN mkdir /workdir && chmod 777 /workdir
 WORKDIR /workdir

docker/packager/cctools/Dockerfile

@@ -0,0 +1,31 @@
+# This is a hack to significantly reduce the build time of the clickhouse/binary-builder
+# It's based on the assumption that we don't care of the cctools version so much
+# It event does not depend on the clickhouse/fasttest in the `docker/images.json`
+ARG FROM_TAG=latest
+FROM clickhouse/fasttest:$FROM_TAG
+
+ENV CC=clang-${LLVM_VERSION}
+ENV CXX=clang++-${LLVM_VERSION}
+
+RUN git clone https://github.com/tpoechtrager/apple-libtapi.git \
+    && cd apple-libtapi \
+    && git checkout 15dfc2a8c9a2a89d06ff227560a69f5265b692f9 \
+    && INSTALLPREFIX=/cctools ./build.sh \
+    && ./install.sh \
+    && cd .. \
+    && rm -rf apple-libtapi
+
+# Build and install tools for cross-linking to Darwin (x86-64)
+# Build and install tools for cross-linking to Darwin (aarch64)
+RUN git clone https://github.com/tpoechtrager/cctools-port.git \
+    && cd cctools-port/cctools \
+    && git checkout 2a3e1c2a6ff54a30f898b70cfb9ba1692a55fad7 \
+    && ./configure --prefix=/cctools --with-libtapi=/cctools \
+        --target=x86_64-apple-darwin \
+    && make install -j$(nproc) \
+    && make clean \
+    && ./configure --prefix=/cctools --with-libtapi=/cctools \
+        --target=aarch64-apple-darwin \
+    && make install -j$(nproc) \
+    && cd ../.. \
+    && rm -rf cctools-port

docker/packager/packager

@@ -1,16 +1,16 @@
 #!/usr/bin/env python3
 # -*- coding: utf-8 -*-
-import subprocess
-import os
 import argparse
 import logging
+import os
+import subprocess
 import sys
 from pathlib import Path
 from typing import List, Optional
 
 SCRIPT_PATH = Path(__file__).absolute()
-IMAGE_TYPE = "binary"
-IMAGE_NAME = f"clickhouse/{IMAGE_TYPE}-builder"
+IMAGE_TYPE = "binary-builder"
+IMAGE_NAME = f"clickhouse/{IMAGE_TYPE}"
 
 
 class BuildException(Exception):

docker/test/fasttest/Dockerfile

@@ -6,9 +6,18 @@ FROM clickhouse/test-util:$FROM_TAG
 RUN apt-get update \
     && apt-get install \
         brotli \
+        clang-${LLVM_VERSION} \
+        clang-tidy-${LLVM_VERSION} \
+        cmake \
         expect \
         file \
+        libclang-${LLVM_VERSION}-dev \
+        libclang-rt-${LLVM_VERSION}-dev \
+        lld-${LLVM_VERSION} \
+        llvm-${LLVM_VERSION} \
+        llvm-${LLVM_VERSION}-dev \
         lsof \
+        ninja-build \
         odbcinst \
         psmisc \
         python3 \
@@ -26,14 +35,50 @@ RUN apt-get update \
 
 RUN pip3 install numpy==1.26.3 scipy==1.12.0 pandas==1.5.3 Jinja2==3.1.3
 
-ARG odbc_driver_url="https://github.com/ClickHouse/clickhouse-odbc/releases/download/v1.1.4.20200302/clickhouse-odbc-1.1.4-Linux.tar.gz"
+# This symlink is required by gcc to find the lld linker
+RUN ln -s /usr/bin/lld-${LLVM_VERSION} /usr/bin/ld.lld
+# for external_symbolizer_path
+RUN ln -s /usr/bin/llvm-symbolizer-${LLVM_VERSION} /usr/bin/llvm-symbolizer
+# FIXME: workaround for "The imported target "merge-fdata" references the file" error
+# https://salsa.debian.org/pkg-llvm-team/llvm-toolchain/-/commit/992e52c0b156a5ba9c6a8a54f8c4857ddd3d371d
+RUN sed -i '/_IMPORT_CHECK_FILES_FOR_\(mlir-\|llvm-bolt\|merge-fdata\|MLIR\)/ {s|^|#|}' /usr/lib/llvm-${LLVM_VERSION}/lib/cmake/llvm/LLVMExports-*.cmake
 
-RUN mkdir -p /tmp/clickhouse-odbc-tmp \
-  && wget -nv -O - ${odbc_driver_url} | tar --strip-components=1 -xz -C /tmp/clickhouse-odbc-tmp \
-  && cp /tmp/clickhouse-odbc-tmp/lib64/*.so /usr/local/lib/ \
-  && odbcinst -i -d -f /tmp/clickhouse-odbc-tmp/share/doc/clickhouse-odbc/config/odbcinst.ini.sample \
-  && odbcinst -i -s -l -f /tmp/clickhouse-odbc-tmp/share/doc/clickhouse-odbc/config/odbc.ini.sample \
-  && rm -rf /tmp/clickhouse-odbc-tmp
+ARG CCACHE_VERSION=4.6.1
+RUN mkdir /tmp/ccache \
+    && cd /tmp/ccache \
+    && curl -L \
+        -O https://github.com/ccache/ccache/releases/download/v$CCACHE_VERSION/ccache-$CCACHE_VERSION.tar.xz \
+        -O https://github.com/ccache/ccache/releases/download/v$CCACHE_VERSION/ccache-$CCACHE_VERSION.tar.xz.asc \
+    && gpg --recv-keys --keyserver hkps://keyserver.ubuntu.com 5A939A71A46792CF57866A51996DDA075594ADB8 \
+    && gpg --verify ccache-4.6.1.tar.xz.asc \
+    && tar xf ccache-$CCACHE_VERSION.tar.xz \
+    && cd /tmp/ccache/ccache-$CCACHE_VERSION \
+    && cmake -DCMAKE_INSTALL_PREFIX=/usr \
+        -DCMAKE_BUILD_TYPE=None \
+        -DZSTD_FROM_INTERNET=ON \
+        -DREDIS_STORAGE_BACKEND=OFF \
+        -Wno-dev \
+        -B build \
+        -S . \
+    && make VERBOSE=1 -C build \
+    && make install -C build \
+    && cd / \
+    && rm -rf /tmp/ccache
+
+ARG TARGETARCH
+ARG SCCACHE_VERSION=v0.7.7
+ENV SCCACHE_IGNORE_SERVER_IO_ERROR=1
+# sccache requires a value for the region. So by default we use The Default Region
+ENV SCCACHE_REGION=us-east-1
+RUN arch=${TARGETARCH:-amd64} \
+  && case $arch in \
+    amd64) rarch=x86_64 ;; \
+    arm64) rarch=aarch64 ;; \
+  esac \
+  && curl -Ls "https://github.com/mozilla/sccache/releases/download/$SCCACHE_VERSION/sccache-$SCCACHE_VERSION-$rarch-unknown-linux-musl.tar.gz" | \
+    tar xz -C /tmp \
+  && mv "/tmp/sccache-$SCCACHE_VERSION-$rarch-unknown-linux-musl/sccache" /usr/bin \
+  && rm "/tmp/sccache-$SCCACHE_VERSION-$rarch-unknown-linux-musl" -r
 
 # Give suid to gdb to grant it attach permissions
 # chmod 777 to make the container user independent

docker/test/fuzzer/run-fuzzer.sh

@@ -247,6 +247,12 @@ quit
     fuzzer_pid=$!
     echo "Fuzzer pid is $fuzzer_pid"
 
+    # The fuzzer_pid belongs to the timeout process.
+    actual_fuzzer_pid=$(ps -o pid= --ppid "$fuzzer_pid")
+
+    echo "Attaching gdb to the fuzzer itself"
+    gdb -batch -command script.gdb -p $actual_fuzzer_pid &
+
     # Wait for the fuzzer to complete.
     # Note that the 'wait || ...' thing is required so that the script doesn't
     # exit because of 'set -e' when 'wait' returns nonzero code.
@@ -337,10 +343,9 @@ quit
         # which is confusing.
         task_exit_code=$fuzzer_exit_code
         echo "failure" > status.txt
-        { rg -ao "Found error:.*" fuzzer.log \
-            || rg -ao "Exception:.*" fuzzer.log \
-            || echo "Fuzzer failed ($fuzzer_exit_code). See the logs." ; } \
-            | tail -1 > description.txt
+        echo "Achtung!" > description.txt
+        echo "Fuzzer went wrong with error code: ($fuzzer_exit_code). Its process died somehow when the server stayed alive. The server log probably won't tell you much so try to find information in other files." >>description.txt
+        { rg -ao "Found error:.*" fuzzer.log || rg -ao "Exception:.*" fuzzer.log; } | tail -1 >>description.txt
     fi
 
     if test -f core.*; then
@@ -387,7 +392,7 @@ if [ -f core.zst ]; then
 fi
 
 # Keep all the lines in the paragraphs containing <Fatal> that either contain <Fatal> or don't start with 20... (year)
-sed -n '/<Fatal>/,/^$/p' s.log | awk '/<Fatal>/ || !/^20/' server.log > fatal.log ||:
+sed -n '/<Fatal>/,/^$/p' server.log | awk '/<Fatal>/ || !/^20/' > fatal.log ||:
 FATAL_LINK=''
 if [ -s fatal.log ]; then
     FATAL_LINK='<a href="fatal.log">fatal.log</a>'

docker/test/sqllogic/Dockerfile

@@ -24,17 +24,18 @@ RUN pip3 install \
     deepdiff \
     sqlglot
 
-ARG odbc_repo="https://github.com/ClickHouse/clickhouse-odbc.git"
+ARG odbc_driver_url="https://github.com/ClickHouse/clickhouse-odbc/releases/download/v1.1.6.20200320/clickhouse-odbc-1.1.6-Linux.tar.gz"
+
+RUN mkdir -p /tmp/clickhouse-odbc-tmp \
+  && cd /tmp/clickhouse-odbc-tmp \
+  && curl -L ${odbc_driver_url} | tar --strip-components=1 -xz clickhouse-odbc-1.1.6-Linux \
+  && mkdir /usr/local/lib64 -p \
+  && cp /tmp/clickhouse-odbc-tmp/lib64/*.so /usr/local/lib64/ \
+  && odbcinst -i -d -f /tmp/clickhouse-odbc-tmp/share/doc/clickhouse-odbc/config/odbcinst.ini.sample \
+  && odbcinst -i -s -l -f /tmp/clickhouse-odbc-tmp/share/doc/clickhouse-odbc/config/odbc.ini.sample \
+  && sed -i 's"=libclickhouseodbc"=/usr/local/lib64/libclickhouseodbc"' /etc/odbcinst.ini \
+  && rm -rf /tmp/clickhouse-odbc-tmp
 
-RUN git clone --recursive ${odbc_repo} \
-    && mkdir -p /clickhouse-odbc/build \
-    && cmake -S /clickhouse-odbc -B /clickhouse-odbc/build \
-    && ls /clickhouse-odbc/build/driver \
-    && make -j 10 -C /clickhouse-odbc/build \
-    && ls /clickhouse-odbc/build/driver \
-    && mkdir -p /usr/local/lib64/ && cp /clickhouse-odbc/build/driver/lib*.so /usr/local/lib64/ \
-    && odbcinst -i -d -f /clickhouse-odbc/packaging/odbcinst.ini.sample \
-    && odbcinst -i -s -l -f /clickhouse-odbc/packaging/odbc.ini.sample
 
 ENV TZ=Europe/Amsterdam
 ENV MAX_RUN_TIME=9000

docker/test/stateless/.gitignore

@@ -0,0 +1 @@
+/minio_data

docker/test/stateless/Dockerfile

@@ -3,7 +3,7 @@
 ARG FROM_TAG=latest
 FROM clickhouse/test-base:$FROM_TAG
 
-ARG odbc_driver_url="https://github.com/ClickHouse/clickhouse-odbc/releases/download/v1.1.4.20200302/clickhouse-odbc-1.1.4-Linux.tar.gz"
+ARG odbc_driver_url="https://github.com/ClickHouse/clickhouse-odbc/releases/download/v1.1.6.20200320/clickhouse-odbc-1.1.6-Linux.tar.gz"
 
 # golang version 1.13 on Ubuntu 20 is enough for tests
 RUN apt-get update -y \
@@ -35,7 +35,6 @@ RUN apt-get update -y \
             sudo \
             tree \
             unixodbc \
-            wget \
             rustc \
             cargo \
             zstd \
@@ -50,11 +49,14 @@ RUN apt-get update -y \
 RUN pip3 install numpy==1.26.3 scipy==1.12.0 pandas==1.5.3 Jinja2==3.1.3 pyarrow==15.0.0
 
 RUN mkdir -p /tmp/clickhouse-odbc-tmp \
-   && wget -nv -O - ${odbc_driver_url} | tar --strip-components=1 -xz -C /tmp/clickhouse-odbc-tmp \
-   && cp /tmp/clickhouse-odbc-tmp/lib64/*.so /usr/local/lib/ \
-   && odbcinst -i -d -f /tmp/clickhouse-odbc-tmp/share/doc/clickhouse-odbc/config/odbcinst.ini.sample \
-   && odbcinst -i -s -l -f /tmp/clickhouse-odbc-tmp/share/doc/clickhouse-odbc/config/odbc.ini.sample \
-   && rm -rf /tmp/clickhouse-odbc-tmp
+  && cd /tmp/clickhouse-odbc-tmp \
+  && curl -L ${odbc_driver_url} | tar --strip-components=1 -xz clickhouse-odbc-1.1.6-Linux \
+  && mkdir /usr/local/lib64 -p \
+  && cp /tmp/clickhouse-odbc-tmp/lib64/*.so /usr/local/lib64/ \
+  && odbcinst -i -d -f /tmp/clickhouse-odbc-tmp/share/doc/clickhouse-odbc/config/odbcinst.ini.sample \
+  && odbcinst -i -s -l -f /tmp/clickhouse-odbc-tmp/share/doc/clickhouse-odbc/config/odbc.ini.sample \
+  && sed -i 's"=libclickhouseodbc"=/usr/local/lib64/libclickhouseodbc"' /etc/odbcinst.ini \
+  && rm -rf /tmp/clickhouse-odbc-tmp
 
 ENV TZ=Europe/Amsterdam
 RUN ln -snf /usr/share/zoneinfo/$TZ /etc/localtime && echo $TZ > /etc/timezone
@@ -70,11 +72,11 @@ ARG TARGETARCH
 
 # Download Minio-related binaries
 RUN arch=${TARGETARCH:-amd64} \
-    && wget "https://dl.min.io/server/minio/release/linux-${arch}/archive/minio.RELEASE.${MINIO_SERVER_VERSION}" -O ./minio \
-    && wget "https://dl.min.io/client/mc/release/linux-${arch}/archive/mc.RELEASE.${MINIO_CLIENT_VERSION}" -O ./mc \
+    && curl -L "https://dl.min.io/server/minio/release/linux-${arch}/archive/minio.RELEASE.${MINIO_SERVER_VERSION}" -o ./minio \
+    && curl -L "https://dl.min.io/client/mc/release/linux-${arch}/archive/mc.RELEASE.${MINIO_CLIENT_VERSION}" -o ./mc \
     && chmod +x ./mc ./minio
 
-RUN wget --no-verbose 'https://archive.apache.org/dist/hadoop/common/hadoop-3.3.1/hadoop-3.3.1.tar.gz' \
+RUN curl -L --no-verbose -O 'https://archive.apache.org/dist/hadoop/common/hadoop-3.3.1/hadoop-3.3.1.tar.gz' \
     && tar -xvf hadoop-3.3.1.tar.gz \
     && rm -rf hadoop-3.3.1.tar.gz
 

docker/test/style/Dockerfile

@@ -18,7 +18,8 @@ RUN apt-get update && env DEBIAN_FRONTEND=noninteractive apt-get install --yes \
     python3-pip \
     yamllint \
     locales \
-    && pip3 install black==23.1.0 boto3 codespell==2.2.1 mypy==1.3.0 PyGithub unidiff pylint==2.6.2 \
+    && pip3 install black==23.12.0 boto3 codespell==2.2.1 mypy==1.8.0 PyGithub unidiff pylint==3.1.0 \
+        requests types-requests \
     && apt-get clean \
     && rm -rf /var/lib/apt/lists/* /var/cache/debconf /tmp/* \
     && rm -rf /root/.cache/pip
@@ -59,5 +60,4 @@ RUN arch=${TARGETARCH:-amd64} \
 
 
 COPY run.sh /
-COPY process_style_check_result.py /
 CMD ["/bin/bash", "/run.sh"]

docker/test/upgrade/run.sh

@@ -79,6 +79,18 @@ remove_keeper_config "async_replication" "1"
 # create_if_not_exists feature flag doesn't exist on some older versions
 remove_keeper_config "create_if_not_exists" "[01]"
 
+#todo: remove these after 24.3 released.
+sudo cat /etc/clickhouse-server/config.d/azure_storage_conf.xml \
+  | sed "s|<object_storage_type>azure|<object_storage_type>azure_blob_storage|" \
+  > /etc/clickhouse-server/config.d/azure_storage_conf.xml.tmp
+sudo mv /etc/clickhouse-server/config.d/azure_storage_conf.xml.tmp /etc/clickhouse-server/config.d/azure_storage_conf.xml
+
+#todo: remove these after 24.3 released.
+sudo cat /etc/clickhouse-server/config.d/storage_conf.xml \
+  | sed "s|<object_storage_type>local|<object_storage_type>local_blob_storage|" \
+  > /etc/clickhouse-server/config.d/storage_conf.xml.tmp
+sudo mv /etc/clickhouse-server/config.d/storage_conf.xml.tmp /etc/clickhouse-server/config.d/storage_conf.xml
+
 # latest_logs_cache_size_threshold setting doesn't exist on some older versions
 remove_keeper_config "latest_logs_cache_size_threshold" "[[:digit:]]\+"
 
@@ -113,6 +125,18 @@ sudo cat /etc/clickhouse-server/config.d/keeper_port.xml \
   > /etc/clickhouse-server/config.d/keeper_port.xml.tmp
 sudo mv /etc/clickhouse-server/config.d/keeper_port.xml.tmp /etc/clickhouse-server/config.d/keeper_port.xml
 
+#todo: remove these after 24.3 released.
+sudo cat /etc/clickhouse-server/config.d/azure_storage_conf.xml \
+  | sed "s|<object_storage_type>azure|<object_storage_type>azure_blob_storage|" \
+  > /etc/clickhouse-server/config.d/azure_storage_conf.xml.tmp
+sudo mv /etc/clickhouse-server/config.d/azure_storage_conf.xml.tmp /etc/clickhouse-server/config.d/azure_storage_conf.xml
+
+#todo: remove these after 24.3 released.
+sudo cat /etc/clickhouse-server/config.d/storage_conf.xml \
+  | sed "s|<object_storage_type>local|<object_storage_type>local_blob_storage|" \
+  > /etc/clickhouse-server/config.d/storage_conf.xml.tmp
+sudo mv /etc/clickhouse-server/config.d/storage_conf.xml.tmp /etc/clickhouse-server/config.d/storage_conf.xml
+
 # async_replication setting doesn't exist on some older versions
 remove_keeper_config "async_replication" "1"
 

docker/test/util/Dockerfile

@@ -41,20 +41,11 @@ RUN apt-get update \
         bash \
         bsdmainutils \
         build-essential \
-        clang-${LLVM_VERSION} \
-        clang-tidy-${LLVM_VERSION} \
-        cmake \
         gdb \
         git \
         gperf \
-        libclang-rt-${LLVM_VERSION}-dev \
-        lld-${LLVM_VERSION} \
-        llvm-${LLVM_VERSION} \
-        llvm-${LLVM_VERSION}-dev \
-        libclang-${LLVM_VERSION}-dev \
         moreutils \
         nasm \
-        ninja-build \
         pigz \
         rename \
         software-properties-common \
@@ -63,49 +54,4 @@ RUN apt-get update \
     && apt-get clean \
     && rm -rf /var/lib/apt/lists/* /var/cache/debconf /tmp/*
 
-# This symlink is required by gcc to find the lld linker
-RUN ln -s /usr/bin/lld-${LLVM_VERSION} /usr/bin/ld.lld
-# for external_symbolizer_path
-RUN ln -s /usr/bin/llvm-symbolizer-${LLVM_VERSION} /usr/bin/llvm-symbolizer
-# FIXME: workaround for "The imported target "merge-fdata" references the file" error
-# https://salsa.debian.org/pkg-llvm-team/llvm-toolchain/-/commit/992e52c0b156a5ba9c6a8a54f8c4857ddd3d371d
-RUN sed -i '/_IMPORT_CHECK_FILES_FOR_\(mlir-\|llvm-bolt\|merge-fdata\|MLIR\)/ {s|^|#|}' /usr/lib/llvm-${LLVM_VERSION}/lib/cmake/llvm/LLVMExports-*.cmake
-
-ARG CCACHE_VERSION=4.6.1
-RUN mkdir /tmp/ccache \
-    && cd /tmp/ccache \
-    && curl -L \
-        -O https://github.com/ccache/ccache/releases/download/v$CCACHE_VERSION/ccache-$CCACHE_VERSION.tar.xz \
-        -O https://github.com/ccache/ccache/releases/download/v$CCACHE_VERSION/ccache-$CCACHE_VERSION.tar.xz.asc \
-    && gpg --recv-keys --keyserver hkps://keyserver.ubuntu.com 5A939A71A46792CF57866A51996DDA075594ADB8 \
-    && gpg --verify ccache-4.6.1.tar.xz.asc \
-    && tar xf ccache-$CCACHE_VERSION.tar.xz \
-    && cd /tmp/ccache/ccache-$CCACHE_VERSION \
-    && cmake -DCMAKE_INSTALL_PREFIX=/usr \
-        -DCMAKE_BUILD_TYPE=None \
-        -DZSTD_FROM_INTERNET=ON \
-        -DREDIS_STORAGE_BACKEND=OFF \
-        -Wno-dev \
-        -B build \
-        -S . \
-    && make VERBOSE=1 -C build \
-    && make install -C build \
-    && cd / \
-    && rm -rf /tmp/ccache
-
-ARG TARGETARCH
-ARG SCCACHE_VERSION=v0.5.4
-ENV SCCACHE_IGNORE_SERVER_IO_ERROR=1
-# sccache requires a value for the region. So by default we use The Default Region
-ENV SCCACHE_REGION=us-east-1
-RUN arch=${TARGETARCH:-amd64} \
-  && case $arch in \
-    amd64) rarch=x86_64 ;; \
-    arm64) rarch=aarch64 ;; \
-  esac \
-  && curl -Ls "https://github.com/mozilla/sccache/releases/download/$SCCACHE_VERSION/sccache-$SCCACHE_VERSION-$rarch-unknown-linux-musl.tar.gz" | \
-    tar xz -C /tmp \
-  && mv "/tmp/sccache-$SCCACHE_VERSION-$rarch-unknown-linux-musl/sccache" /usr/bin \
-  && rm "/tmp/sccache-$SCCACHE_VERSION-$rarch-unknown-linux-musl" -r
-
 COPY process_functional_tests_result.py /

docs/changelogs/v23.3.20.27-lts.md

@@ -0,0 +1,29 @@
+---
+sidebar_position: 1
+sidebar_label: 2024
+---
+
+# 2024 Changelog
+
+### ClickHouse release v23.3.20.27-lts (cc974ba4f81) FIXME as compared to v23.3.19.32-lts (c4d4ca8ec02)
+
+#### Improvement
+* Backported in [#58818](https://github.com/ClickHouse/ClickHouse/issues/58818): Add `SYSTEM JEMALLOC PURGE` for purging unused jemalloc pages, `SYSTEM JEMALLOC [ ENABLE | DISABLE | FLUSH ] PROFILE` for controlling jemalloc profile if the profiler is enabled. Add jemalloc-related 4LW command in Keeper: `jmst` for dumping jemalloc stats, `jmfp`, `jmep`, `jmdp` for controlling jemalloc profile if the profiler is enabled. [#58665](https://github.com/ClickHouse/ClickHouse/pull/58665) ([Antonio Andelic](https://github.com/antonio2368)).
+
+#### Build/Testing/Packaging Improvement
+* Backported in [#59877](https://github.com/ClickHouse/ClickHouse/issues/59877): If you want to run initdb scripts every time when ClickHouse container is starting you shoud initialize environment varible CLICKHOUSE_ALWAYS_RUN_INITDB_SCRIPTS. [#59808](https://github.com/ClickHouse/ClickHouse/pull/59808) ([Alexander Nikolaev](https://github.com/AlexNik)).
+
+#### Bug Fix (user-visible misbehavior in an official stable release)
+
+* Fix working with read buffers in StreamingFormatExecutor [#57438](https://github.com/ClickHouse/ClickHouse/pull/57438) ([Kruglov Pavel](https://github.com/Avogar)).
+* Fix double destroy call on exception throw in addBatchLookupTable8 [#58745](https://github.com/ClickHouse/ClickHouse/pull/58745) ([Raúl Marín](https://github.com/Algunenano)).
+* Fix: LIMIT BY and LIMIT in distributed query [#59153](https://github.com/ClickHouse/ClickHouse/pull/59153) ([Igor Nikonov](https://github.com/devcrafter)).
+* Fix translate() with FixedString input [#59356](https://github.com/ClickHouse/ClickHouse/pull/59356) ([Raúl Marín](https://github.com/Algunenano)).
+* Fix leftPad / rightPad function with FixedString input [#59739](https://github.com/ClickHouse/ClickHouse/pull/59739) ([Raúl Marín](https://github.com/Algunenano)).
+* Fix cosineDistance crash with Nullable [#60150](https://github.com/ClickHouse/ClickHouse/pull/60150) ([Raúl Marín](https://github.com/Algunenano)).
+
+#### NOT FOR CHANGELOG / INSIGNIFICANT
+
+* Fix possible race in ManyAggregatedData dtor. [#58624](https://github.com/ClickHouse/ClickHouse/pull/58624) ([Nikolai Kochetov](https://github.com/KochetovNicolai)).
+* Make ZooKeeper actually sequentialy consistent [#59735](https://github.com/ClickHouse/ClickHouse/pull/59735) ([Alexander Tokmakov](https://github.com/tavplubix)).
+

docs/changelogs/v23.8.10.43-lts.md

@@ -0,0 +1,39 @@
+---
+sidebar_position: 1
+sidebar_label: 2024
+---
+
+# 2024 Changelog
+
+### ClickHouse release v23.8.10.43-lts (a278225bba9) FIXME as compared to v23.8.9.54-lts (192a1d231fa)
+
+#### Improvement
+* Backported in [#58819](https://github.com/ClickHouse/ClickHouse/issues/58819): Add `SYSTEM JEMALLOC PURGE` for purging unused jemalloc pages, `SYSTEM JEMALLOC [ ENABLE | DISABLE | FLUSH ] PROFILE` for controlling jemalloc profile if the profiler is enabled. Add jemalloc-related 4LW command in Keeper: `jmst` for dumping jemalloc stats, `jmfp`, `jmep`, `jmdp` for controlling jemalloc profile if the profiler is enabled. [#58665](https://github.com/ClickHouse/ClickHouse/pull/58665) ([Antonio Andelic](https://github.com/antonio2368)).
+* Backported in [#60286](https://github.com/ClickHouse/ClickHouse/issues/60286): Copy S3 file GCP fallback to buffer copy in case GCP returned `Internal Error` with `GATEWAY_TIMEOUT` HTTP error code. [#60164](https://github.com/ClickHouse/ClickHouse/pull/60164) ([Maksim Kita](https://github.com/kitaisreal)).
+
+#### Build/Testing/Packaging Improvement
+* Backported in [#59879](https://github.com/ClickHouse/ClickHouse/issues/59879): If you want to run initdb scripts every time when ClickHouse container is starting you shoud initialize environment varible CLICKHOUSE_ALWAYS_RUN_INITDB_SCRIPTS. [#59808](https://github.com/ClickHouse/ClickHouse/pull/59808) ([Alexander Nikolaev](https://github.com/AlexNik)).
+
+#### Bug Fix (user-visible misbehavior in an official stable release)
+
+* Background merges correctly use temporary data storage in the cache [#57275](https://github.com/ClickHouse/ClickHouse/pull/57275) ([vdimir](https://github.com/vdimir)).
+* MergeTree mutations reuse source part index granularity [#57352](https://github.com/ClickHouse/ClickHouse/pull/57352) ([Maksim Kita](https://github.com/kitaisreal)).
+* Fix double destroy call on exception throw in addBatchLookupTable8 [#58745](https://github.com/ClickHouse/ClickHouse/pull/58745) ([Raúl Marín](https://github.com/Algunenano)).
+* Fix JSONExtract function for LowCardinality(Nullable) columns [#58808](https://github.com/ClickHouse/ClickHouse/pull/58808) ([vdimir](https://github.com/vdimir)).
+* Fix: LIMIT BY and LIMIT in distributed query [#59153](https://github.com/ClickHouse/ClickHouse/pull/59153) ([Igor Nikonov](https://github.com/devcrafter)).
+* Fix translate() with FixedString input [#59356](https://github.com/ClickHouse/ClickHouse/pull/59356) ([Raúl Marín](https://github.com/Algunenano)).
+* Fix error "Read beyond last offset" for AsynchronousBoundedReadBuffer [#59630](https://github.com/ClickHouse/ClickHouse/pull/59630) ([Vitaly Baranov](https://github.com/vitlibar)).
+* Fix query start time on non initial queries [#59662](https://github.com/ClickHouse/ClickHouse/pull/59662) ([Raúl Marín](https://github.com/Algunenano)).
+* Fix leftPad / rightPad function with FixedString input [#59739](https://github.com/ClickHouse/ClickHouse/pull/59739) ([Raúl Marín](https://github.com/Algunenano)).
+* rabbitmq: fix having neither acked nor nacked messages [#59775](https://github.com/ClickHouse/ClickHouse/pull/59775) ([Kseniia Sumarokova](https://github.com/kssenii)).
+* Fix cosineDistance crash with Nullable [#60150](https://github.com/ClickHouse/ClickHouse/pull/60150) ([Raúl Marín](https://github.com/Algunenano)).
+
+#### NOT FOR CHANGELOG / INSIGNIFICANT
+
+* Fix rare race in external sort/aggregation with temporary data in cache [#58013](https://github.com/ClickHouse/ClickHouse/pull/58013) ([Anton Popov](https://github.com/CurtizJ)).
+* Fix possible race in ManyAggregatedData dtor. [#58624](https://github.com/ClickHouse/ClickHouse/pull/58624) ([Nikolai Kochetov](https://github.com/KochetovNicolai)).
+* Fix 02720_row_policy_column_with_dots [#59453](https://github.com/ClickHouse/ClickHouse/pull/59453) ([Duc Canh Le](https://github.com/canhld94)).
+* Pin python dependencies in stateless tests [#59663](https://github.com/ClickHouse/ClickHouse/pull/59663) ([Raúl Marín](https://github.com/Algunenano)).
+* Make ZooKeeper actually sequentialy consistent [#59735](https://github.com/ClickHouse/ClickHouse/pull/59735) ([Alexander Tokmakov](https://github.com/tavplubix)).
+* Remove broken test while we fix it [#60547](https://github.com/ClickHouse/ClickHouse/pull/60547) ([Raúl Marín](https://github.com/Algunenano)).
+

docs/en/interfaces/schema-inference.md

@@ -549,6 +549,48 @@ Result:
 └───────┴─────────────────────────────────────────────────────────────────────────────────────────────────┴──────────────┴────────────────────┴─────────┴──────────────────┴────────────────┘
 ```
 
+##### input_format_json_use_string_type_for_ambiguous_paths_in_named_tuples_inference_from_objects
+
+Enabling this setting allows to use String type for ambiguous paths during named tuples inference from JSON objects (when `input_format_json_try_infer_named_tuples_from_objects` is enabled) instead of an exception.
+It allows to read JSON objects as named Tuples even if there are ambiguous paths.
+
+Disabled by default.
+
+**Examples**
+
+With disabled setting:
+```sql
+SET input_format_json_try_infer_named_tuples_from_objects = 1;
+SET input_format_json_use_string_type_for_ambiguous_paths_in_named_tuples_inference_from_objects = 0;
+DESC format(JSONEachRow, '{"obj" : {"a" : 42}}, {"obj" : {"a" : {"b" : "Hello"}}}');
+```
+Result:
+
+```text
+Code: 636. DB::Exception: The table structure cannot be extracted from a JSONEachRow format file. Error:
+Code: 117. DB::Exception: JSON objects have ambiguous data: in some objects path 'a' has type 'Int64' and in some - 'Tuple(b String)'. You can enable setting input_format_json_use_string_type_for_ambiguous_paths_in_named_tuples_inference_from_objects to use String type for path 'a'. (INCORRECT_DATA) (version 24.3.1.1).
+You can specify the structure manually. (CANNOT_EXTRACT_TABLE_STRUCTURE)
+```
+
+With enabled setting:
+```sql
+SET input_format_json_try_infer_named_tuples_from_objects = 1;
+SET input_format_json_use_string_type_for_ambiguous_paths_in_named_tuples_inference_from_objects = 1;
+DESC format(JSONEachRow, '{"obj" : "a" : 42}, {"obj" : {"a" : {"b" : "Hello"}}}');
+SELECT * FROM format(JSONEachRow, '{"obj" : {"a" : 42}}, {"obj" : {"a" : {"b" : "Hello"}}}');
+```
+
+Result:
+```text
+┌─name─┬─type──────────────────────────┬─default_type─┬─default_expression─┬─comment─┬─codec_expression─┬─ttl_expression─┐
+│ obj  │ Tuple(a Nullable(String))     │              │                    │         │                  │                │
+└──────┴───────────────────────────────┴──────────────┴────────────────────┴─────────┴──────────────────┴────────────────┘
+┌─obj─────────────────┐
+│ ('42')              │
+│ ('{"b" : "Hello"}') │
+└─────────────────────┘
+```
+
 ##### input_format_json_read_objects_as_strings
 
 Enabling this setting allows reading nested JSON objects as strings.
@@ -1554,6 +1596,28 @@ DESC format(JSONEachRow, $$
 └──────┴──────────────────┴──────────────┴────────────────────┴─────────┴──────────────────┴────────────────┘
 ```
 
+#### input_format_try_infer_exponent_floats
+
+If enabled, ClickHouse will try to infer floats in exponential form for text formats (except JSON where numbers in exponential form are always inferred).
+
+Disabled by default.
+
+**Example**
+
+```sql
+SET input_format_try_infer_exponent_floats = 1;
+DESC format(CSV,
+$$1.1E10
+2.3e-12
+42E00
+$$)
+```
+```response
+┌─name─┬─type──────────────┬─default_type─┬─default_expression─┬─comment─┬─codec_expression─┬─ttl_expression─┐
+│ c1   │ Nullable(Float64) │              │                    │         │                  │                │
+└──────┴───────────────────┴──────────────┴────────────────────┴─────────┴──────────────────┴────────────────┘
+```
+
 ## Self describing formats {#self-describing-formats}
 
 Self-describing formats contain information about the structure of the data in the data itself,

docs/en/operations/settings/settings-formats.md

@@ -467,7 +467,7 @@ Enabled by default.
 
 Allow to use String type for JSON keys that contain only `Null`/`{}`/`[]` in data sample during schema inference.
 In JSON formats any value can be read as String, and we can avoid errors like `Cannot determine type for column 'column_name' by first 25000 rows of data, most likely this column contains only Nulls or empty Arrays/Maps` during schema inference
-by using String type for keys with unknown types. 
+by using String type for keys with unknown types.
 
 Example:
 
@@ -891,7 +891,7 @@ Default value: `,`.
 
 If it is set to true, allow strings in single quotes.
 
-Enabled by default.
+Disabled by default.
 
 ### format_csv_allow_double_quotes {#format_csv_allow_double_quotes}
 
@@ -1605,7 +1605,7 @@ possible values:
 -   `1` — Enabled. Pretty formats will use ANSI escape sequences except for `NoEscapes` formats.
 -   `auto` - Enabled if `stdout` is a terminal except for `NoEscapes` formats.
 
-Default value is `auto`. 
+Default value is `auto`.
 
 ### output_format_pretty_grid_charset {#output_format_pretty_grid_charset}
 

docs/en/operations/settings/settings.md

@@ -3954,6 +3954,7 @@ Possible values:
 - `none` — Is similar to throw, but distributed DDL query returns no result set.
 - `null_status_on_timeout` — Returns `NULL` as execution status in some rows of result set instead of throwing `TIMEOUT_EXCEEDED` if query is not finished on the corresponding hosts.
 - `never_throw` — Do not throw `TIMEOUT_EXCEEDED` and do not rethrow exceptions if query has failed on some hosts.
+- `none_only_active` - similar to `none`, but doesn't wait for inactive replicas of the `Replicated` database. Note: with this mode it's impossible to figure out that the query was not executed on some replica and will be executed in background.
 - `null_status_on_timeout_only_active` — similar to `null_status_on_timeout`, but doesn't wait for inactive replicas of the `Replicated` database
 - `throw_only_active` — similar to `throw`, but doesn't wait for inactive replicas of the `Replicated` database
 

docs/en/operations/utilities/clickhouse-local.md

@@ -201,12 +201,12 @@ Arguments:
 
 - `-S`, `--structure` — table structure for input data.
 - `--input-format` — input format, `TSV` by default.
-- `-f`, `--file` — path to data, `stdin` by default.
+- `-F`, `--file` — path to data, `stdin` by default.
 - `-q`, `--query` — queries to execute with `;` as delimiter. `--query` can be specified multiple times, e.g. `--query "SELECT 1" --query "SELECT 2"`. Cannot be used simultaneously with `--queries-file`.
 - `--queries-file` - file path with queries to execute. `--queries-file` can be specified multiple times, e.g. `--query queries1.sql --query queries2.sql`. Cannot be used simultaneously with `--query`.
 - `--multiquery, -n` – If specified, multiple queries separated by semicolons can be listed after the `--query` option. For convenience, it is also possible to omit `--query` and pass the queries directly after `--multiquery`.
 - `-N`, `--table` — table name where to put output data, `table` by default.
-- `--format`, `--output-format` — output format, `TSV` by default.
+- `-f`, `--format`, `--output-format` — output format, `TSV` by default.
 - `-d`, `--database` — default database, `_local` by default.
 - `--stacktrace` — whether to dump debug output in case of exception.
 - `--echo` — print query before execution.

docs/en/sql-reference/aggregate-functions/reference/grouparraysorted.md

@@ -14,8 +14,6 @@
 
  -   `N` – The number of elements to return.
 
- If the parameter is omitted, default value is the size of input.
-
  -   `column` – The value (Integer, String, Float and other Generic types).
 
  **Example**
@@ -36,13 +34,12 @@
  Gets all the String implementations of all numbers in column:
 
  ``` sql
-SELECT groupArraySorted(str) FROM (SELECT toString(number) as str FROM numbers(5));
+SELECT groupArraySorted(5)(str) FROM (SELECT toString(number) as str FROM numbers(5));
 
  ```
 
  ``` text
- ┌─groupArraySorted(str)────────┐
- │ ['0','1','2','3','4']        │
- └──────────────────────────────┘
- ```
- 
+┌─groupArraySorted(5)(str)─┐
+│ ['0','1','2','3','4']    │
+└──────────────────────────┘
+ ```

docs/en/sql-reference/functions/date-time-functions.md

@@ -394,8 +394,7 @@ Result:
 
 ## toYear
 
-Converts a date or date with time to the year number (AD) as `UInt16` value.
-
+Returns the year component (AD) of a date or date with time.
 
 **Syntax**
 
@@ -431,7 +430,7 @@ Result:
 
 ## toQuarter
 
-Converts a date or date with time to the quarter number (1-4) as `UInt8` value.
+Returns the quarter (1-4) of a date or date with time.
 
 **Syntax**
 
@@ -465,10 +464,9 @@ Result:
 └──────────────────────────────────────────────┘
 ```
 
-
 ## toMonth
 
-Converts a date or date with time to the month number (1-12) as `UInt8` value.
+Returns the month component (1-12) of a date or date with time.
 
 **Syntax**
 
@@ -504,7 +502,7 @@ Result:
 
 ## toDayOfYear
 
-Converts a date or date with time to the number of the day of the year (1-366) as `UInt16` value.
+Returns the number of the day within the year (1-366) of a date or date with time.
 
 **Syntax**
 
@@ -540,7 +538,7 @@ Result:
 
 ## toDayOfMonth
 
-Converts a date or date with time to the number of the day in the month (1-31) as `UInt8` value.
+Returns the number of the day within the month (1-31) of a date or date with time.
 
 **Syntax**
 
@@ -576,7 +574,7 @@ Result:
 
 ## toDayOfWeek
 
-Converts a date or date with time to the number of the day in the week as `UInt8` value.
+Returns the number of the day within the week of a date or date with time.
 
 The two-argument form of `toDayOfWeek()` enables you to specify whether the week starts on Monday or Sunday, and whether the return value should be in the range from 0 to 6 or 1 to 7. If the mode argument is omitted, the default mode is 0. The time zone of the date can be specified as the third argument.
 
@@ -627,7 +625,7 @@ Result:
 
 ## toHour
 
-Converts a date with time to the number of the hour in 24-hour time (0-23) as `UInt8` value.
+Returns the hour component (0-24) of a date with time.
 
 Assumes that if clocks are moved ahead, it is by one hour and occurs at 2 a.m., and if clocks are moved back, it is by one hour and occurs at 3 a.m. (which is not always exactly when it occurs - it depends on the timezone).
 
@@ -641,7 +639,7 @@ Alias: `HOUR`
 
 **Arguments**
 
-- `value` - a [Date](../data-types/date.md), [Date32](../data-types/date32.md), [DateTime](../data-types/datetime.md) or [DateTime64](../data-types/datetime64.md)
+- `value` - a [DateTime](../data-types/datetime.md) or [DateTime64](../data-types/datetime64.md)
 
 **Returned value**
 
@@ -665,7 +663,7 @@ Result:
 
 ## toMinute
 
-Converts a date with time to the number of the minute of the hour (0-59) as `UInt8` value.
+Returns the minute component (0-59) a date with time.
 
 **Syntax**
 
@@ -677,7 +675,7 @@ Alias: `MINUTE`
 
 **Arguments**
 
-- `value` - a [Date](../data-types/date.md), [Date32](../data-types/date32.md), [DateTime](../data-types/datetime.md) or [DateTime64](../data-types/datetime64.md)
+- `value` - a [DateTime](../data-types/datetime.md) or [DateTime64](../data-types/datetime64.md)
 
 **Returned value**
 
@@ -701,7 +699,7 @@ Result:
 
 ## toSecond
 
-Converts a date with time to the second in the minute (0-59) as `UInt8` value. Leap seconds are not considered.
+Returns the second component (0-59) of a date with time. Leap seconds are not considered.
 
 **Syntax**
 
@@ -713,7 +711,7 @@ Alias: `SECOND`
 
 **Arguments**
 
-- `value` - a [Date](../data-types/date.md), [Date32](../data-types/date32.md), [DateTime](../data-types/datetime.md) or [DateTime64](../data-types/datetime64.md)
+- `value` - a [DateTime](../data-types/datetime.md) or [DateTime64](../data-types/datetime64.md)
 
 **Returned value**
 
@@ -735,6 +733,40 @@ Result:
 └─────────────────────────────────────────────┘
 ```
 
+## toMillisecond
+
+Returns the millisecond component (0-999) of a date with time.
+
+**Syntax**
+
+```sql
+toMillisecond(value)
+```
+
+*Arguments**
+
+- `value` - [DateTime](../data-types/datetime.md) or [DateTime64](../data-types/datetime64.md)
+
+Alias: `MILLISECOND`
+
+```sql
+SELECT toMillisecond(toDateTime64('2023-04-21 10:20:30.456', 3))
+```
+
+Result:
+
+```response
+┌──toMillisecond(toDateTime64('2023-04-21 10:20:30.456', 3))─┐
+│                                                        456 │
+└────────────────────────────────────────────────────────────┘
+```
+
+**Returned value**
+
+- The millisecond in the minute (0 - 59) of the given date/time
+
+Type: `UInt16`
+
 ## toUnixTimestamp
 
 Converts a string, a date or a date with time to the [Unix Timestamp](https://en.wikipedia.org/wiki/Unix_time) in `UInt32` representation.

docs/en/sql-reference/statements/alter/column.md

@@ -272,10 +272,16 @@ ALTER TABLE table_name MODIFY COLUMN column_name RESET SETTING max_compress_bloc
 
 ## MATERIALIZE COLUMN
 
-Materializes or updates a column with an expression for a default value (`DEFAULT` or `MATERIALIZED`).
-It is used if it is necessary to add or update a column with a complicated expression, because evaluating such an expression directly on `SELECT` executing turns out to be expensive.
+Materializes a column with a `DEFAULT` or `MATERIALIZED` value expression.
+This statement can be used to rewrite existing column data after a `DEFAULT` or `MATERIALIZED` expression has been added or updated (which only updates the metadata but does not change existing data).
 Implemented as a [mutation](/docs/en/sql-reference/statements/alter/index.md#mutations).
 
+For columns with a new or updated `MATERIALIZED` value expression, all existing rows are rewritten.
+
+For columns with a new or updated `DEFAULT` value expression, the behavior depends on the ClickHouse version:
+- In ClickHouse < v24.2, all existing rows are rewritten.
+- ClickHouse >= v24.2 distinguishes if a row value in a column with `DEFAULT` value expression was explicitly specified when it was inserted, or not, i.e. calculated from the `DEFAULT` value expression. If the value was explicitly specified, ClickHouse keeps it as is. If the value was was calculated, ClickHouse changes it to the new or updated `MATERIALIZED` value expression.
+
 Syntax:
 
 ```sql

docs/en/sql-reference/statements/grant.md

@@ -202,6 +202,13 @@ Hierarchy of privileges:
     - `S3`
 - [dictGet](#grant-dictget)
 - [displaySecretsInShowAndSelect](#grant-display-secrets)
+- [NAMED COLLECTION ADMIN](#grant-named-collection-admin)
+    - `CREATE NAMED COLLECTION`
+    - `DROP NAMED COLLECTION`
+    - `ALTER NAMED COLLECTION`
+    - `SHOW NAMED COLLECTIONS`
+    - `SHOW NAMED COLLECTIONS SECRETS`
+    - `NAMED COLLECTION`
 
 Examples of how this hierarchy is treated:
 
@@ -498,6 +505,25 @@ and
 [`format_display_secrets_in_show_and_select` format setting](../../operations/settings/formats#format_display_secrets_in_show_and_select)
 are turned on.
 
+### NAMED COLLECTION ADMIN
+
+Allows a certain operation on a specified named collection. Before version 23.7 it was called NAMED COLLECTION CONTROL, and after 23.7 NAMED COLLECTION ADMIN was added and NAMED COLLECTION CONTROL is preserved as an alias.
+
+- `NAMED COLLECTION ADMIN`. Level: `NAMED_COLLECTION`. Aliases: `NAMED COLLECTION CONTROL`
+    - `CREATE NAMED COLLECTION`. Level: `NAMED_COLLECTION`
+    - `DROP NAMED COLLECTION`. Level: `NAMED_COLLECTION`
+    - `ALTER NAMED COLLECTION`. Level: `NAMED_COLLECTION`
+    - `SHOW NAMED COLLECTIONS`. Level: `NAMED_COLLECTION`. Aliases: `SHOW NAMED COLLECTIONS`
+    - `SHOW NAMED COLLECTIONS SECRETS`. Level: `NAMED_COLLECTION`. Aliases: `SHOW NAMED COLLECTIONS SECRETS`
+    - `NAMED COLLECTION`. Level: `NAMED_COLLECTION`. Aliases: `NAMED COLLECTION USAGE, USE NAMED COLLECTION`
+
+Unlike all other grants (CREATE, DROP, ALTER, SHOW) grant NAMED COLLECTION was added only in 23.7, while all others were added earlier - in 22.12.
+
+**Examples**
+
+Assuming a named collection is called abc, we grant privilege CREATE NAMED COLLECTION to user john.
+- `GRANT CREATE NAMED COLLECTION ON abc TO john`
+
 ### ALL
 
 Grants all the privileges on regulated entity to a user account or a role.

docs/en/sql-reference/window-functions/index.md

@@ -5,7 +5,12 @@ sidebar_label: Window Functions
 title: Window Functions
 ---
 
-ClickHouse supports the standard grammar for defining windows and window functions. The following features are currently supported:
+Windows functions let you perform calculations across a set of rows that are related to the current row.
+Some of the calculations that you can do are similar to those that can be done with an aggregate function, but a window function doesn't cause rows to be grouped into a single output - the individual rows are still returned.
+
+## Standard Window Functions
+
+ClickHouse supports the standard grammar for defining windows and window functions. The table below indicates whether a feature is currently supported.
 
 | Feature                                                                            | Support or workaround                                                                                                                                                                       |
 |------------------------------------------------------------------------------------|---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
@@ -25,6 +30,8 @@ ClickHouse supports the standard grammar for defining windows and window functio
 
 ## ClickHouse-specific Window Functions
 
+There are also the following window function that's specific to ClickHouse:
+
 ### nonNegativeDerivative(metric_column, timestamp_column[, INTERVAL X UNITS])
 
 Finds non-negative derivative for given `metric_column` by `timestamp_column`. 
@@ -33,40 +40,6 @@ The computed value is the following for each row:
 - `0` for 1st row,
 - ${metric_i - metric_{i-1} \over timestamp_i - timestamp_{i-1}}  * interval$ for $i_th$ row.
 
-## References
-
-### GitHub Issues
-
-The roadmap for the initial support of window functions is [in this issue](https://github.com/ClickHouse/ClickHouse/issues/18097).
-
-All GitHub issues related to window functions have the [comp-window-functions](https://github.com/ClickHouse/ClickHouse/labels/comp-window-functions) tag.
-
-### Tests
-
-These tests contain the examples of the currently supported grammar:
-
-https://github.com/ClickHouse/ClickHouse/blob/master/tests/performance/window_functions.xml
-
-https://github.com/ClickHouse/ClickHouse/blob/master/tests/queries/0_stateless/01591_window_functions.sql
-
-### Postgres Docs
-
-https://www.postgresql.org/docs/current/sql-select.html#SQL-WINDOW
-
-https://www.postgresql.org/docs/devel/sql-expressions.html#SYNTAX-WINDOW-FUNCTIONS
-
-https://www.postgresql.org/docs/devel/functions-window.html
-
-https://www.postgresql.org/docs/devel/tutorial-window.html
-
-### MySQL Docs
-
-https://dev.mysql.com/doc/refman/8.0/en/window-function-descriptions.html
-
-https://dev.mysql.com/doc/refman/8.0/en/window-functions-usage.html
-
-https://dev.mysql.com/doc/refman/8.0/en/window-functions-frames.html
-
 ## Syntax
 
 ```text
@@ -80,20 +53,7 @@ WINDOW window_name as ([[PARTITION BY grouping_column] [ORDER BY sorting_column]
 - `PARTITION BY` - defines how to break a resultset into groups.
 - `ORDER BY` - defines how to order rows inside the group during calculation aggregate_function.
 - `ROWS or RANGE` - defines bounds of a frame, aggregate_function is calculated within a frame.
-- `WINDOW` - allows to reuse a window definition with multiple expressions.
-
-### Functions
-
-These functions can be used only as a window function.
-
-- `row_number()` - Number the current row within its partition starting from 1.
-- `first_value(x)` - Return the first non-NULL value evaluated within its ordered frame.
-- `last_value(x)` -	Return the last non-NULL value evaluated within its ordered frame.
-- `nth_value(x, offset)` - Return the first non-NULL value evaluated against the nth row (offset) in its ordered frame.
-- `rank()` - Rank the current row within its partition with gaps.
-- `dense_rank()` - Rank the current row within its partition without gaps.
-- `lagInFrame(x)` - Return a value evaluated at the row that is at a specified physical offset row before the current row within the ordered frame.
-- `leadInFrame(x)` - Return a value evaluated at the row that is offset rows after the current row within the ordered frame.
+- `WINDOW` - allows multiple expressions to use the same window definition.
 
 ```text
       PARTITION
@@ -112,8 +72,23 @@ These functions can be used only as a window function.
 └─────────────────┘  <--- UNBOUNDED FOLLOWING (END of the PARTITION)
 ```
 
+### Functions
+
+These functions can be used only as a window function.
+
+- `row_number()` - Number the current row within its partition starting from 1.
+- `first_value(x)` - Return the first non-NULL value evaluated within its ordered frame.
+- `last_value(x)` -	Return the last non-NULL value evaluated within its ordered frame.
+- `nth_value(x, offset)` - Return the first non-NULL value evaluated against the nth row (offset) in its ordered frame.
+- `rank()` - Rank the current row within its partition with gaps.
+- `dense_rank()` - Rank the current row within its partition without gaps.
+- `lagInFrame(x)` - Return a value evaluated at the row that is at a specified physical offset row before the current row within the ordered frame.
+- `leadInFrame(x)` - Return a value evaluated at the row that is offset rows after the current row within the ordered frame.
+
 ## Examples
 
+Let's have a look at some examples of how window functions can be used.
+
 ```sql
 CREATE TABLE wf_partition
 (
@@ -589,6 +564,41 @@ ORDER BY
 └──────────────┴─────────────────────┴───────┴─────────────────────────┘
 ```
 
+## References
+
+### GitHub Issues
+
+The roadmap for the initial support of window functions is [in this issue](https://github.com/ClickHouse/ClickHouse/issues/18097).
+
+All GitHub issues related to window functions have the [comp-window-functions](https://github.com/ClickHouse/ClickHouse/labels/comp-window-functions) tag.
+
+### Tests
+
+These tests contain the examples of the currently supported grammar:
+
+https://github.com/ClickHouse/ClickHouse/blob/master/tests/performance/window_functions.xml
+
+https://github.com/ClickHouse/ClickHouse/blob/master/tests/queries/0_stateless/01591_window_functions.sql
+
+### Postgres Docs
+
+https://www.postgresql.org/docs/current/sql-select.html#SQL-WINDOW
+
+https://www.postgresql.org/docs/devel/sql-expressions.html#SYNTAX-WINDOW-FUNCTIONS
+
+https://www.postgresql.org/docs/devel/functions-window.html
+
+https://www.postgresql.org/docs/devel/tutorial-window.html
+
+### MySQL Docs
+
+https://dev.mysql.com/doc/refman/8.0/en/window-function-descriptions.html
+
+https://dev.mysql.com/doc/refman/8.0/en/window-functions-usage.html
+
+https://dev.mysql.com/doc/refman/8.0/en/window-functions-frames.html
+
+
 ## Related Content
 
 - Blog: [Working with time series data in ClickHouse](https://clickhouse.com/blog/working-with-time-series-data-and-functions-ClickHouse)

docs/ru/operations/clickhouse-keeper.md

@@ -38,6 +38,7 @@ ClickHouse Keeper может использоваться как равноце
 -    `dead_session_check_period_ms` — частота, с которой ClickHouse Keeper проверяет мертвые сессии и удаляет их, в миллисекундах (по умолчанию: 500).
 -    `election_timeout_lower_bound_ms` — время, после которого последователь может инициировать перевыбор лидера, если не получил от него контрольный сигнал (по умолчанию: 1000).
 -    `election_timeout_upper_bound_ms` — время, после которого последователь должен инициировать перевыбор лидера, если не получил от него контрольный сигнал (по умолчанию: 2000).
+-    `leadership_expiry_ms` — Если лидер не получает ответа от достаточного количества последователей в течение этого промежутка времени, он добровольно отказывается от своего руководства. При настройке 0 автоматически устанавливается 20 - кратное значение `heart_beat_interval_ms`, а при настройке меньше 0 лидер не отказывается от лидерства (по умолчанию 0).
 -    `force_sync` — вызывать `fsync` при каждой записи в журнал координации (по умолчанию: true).
 -    `four_letter_word_white_list` — список разрешенных 4-х буквенных команд (по умолчанию: "conf,cons,crst,envi,ruok,srst,srvr,stat,wchc,wchs,dirs,mntr,isro").
 -    `fresh_log_gap` — минимальное отставание от лидера в количестве записей журнала после которого последователь считает себя актуальным (по умолчанию: 200).
@@ -209,6 +210,7 @@ dead_session_check_period_ms=500
 heart_beat_interval_ms=500
 election_timeout_lower_bound_ms=1000
 election_timeout_upper_bound_ms=2000
+leadership_expiry_ms=0
 reserved_log_items=1000000000000000
 snapshot_distance=10000
 auto_forwarding=true

docs/zh/operations/clickhouse-keeper.md

@@ -45,6 +45,7 @@ ClickHouse Keeper 完全可以作为ZooKeeper的独立替代品或者作为Click
 -    `heart_beat_interval_ms` — ClickHouse Keeper的leader发送心跳频率(毫秒)(默认为500)。
 -    `election_timeout_lower_bound_ms` — 如果follower在此间隔内没有收到leader的心跳，那么它可以启动leader选举(默认为1000).
 -    `election_timeout_upper_bound_ms` — 如果follower在此间隔内没有收到leader的心跳，那么它必须启动leader选举(默认为2000)。
+-    `leadership_expiry_ms` — 如果leader在此间隔内没有收到足够的follower回复，那么他会主动放弃领导权。当被设置为0时会自动设置为`heart_beat_interval_ms`的20倍，当被设置小于0时leader不会主动放弃领导权（默认为0）。
 -    `rotate_log_storage_interval` — 单个文件中存储的日志记录数量(默认100000条)。
 -    `reserved_log_items` — 在压缩之前需要存储多少协调日志记录(默认100000)。
 -    `snapshot_distance` — ClickHouse Keeper创建新快照的频率(以日志记录的数量为单位)(默认100000)。
@@ -214,6 +215,7 @@ dead_session_check_period_ms=500
 heart_beat_interval_ms=500
 election_timeout_lower_bound_ms=1000
 election_timeout_upper_bound_ms=2000
+leadership_expiry_ms=0
 reserved_log_items=1000000000000000
 snapshot_distance=10000
 auto_forwarding=true

docs/zh/operations/settings/settings.md

@@ -649,11 +649,22 @@ log_query_threads=1
 
 ## max_query_size {#settings-max_query_size}
 
-查询的最大部分，可以被带到RAM用于使用SQL解析器进行解析。
-插入查询还包含由单独的流解析器（消耗O(1)RAM）处理的插入数据，这些数据不包含在此限制中。
+SQL 解析器解析的查询字符串的最大字节数。 INSERT 查询的 VALUES 子句中的数据由单独的流解析器（消耗 O(1) RAM）处理，并且不受此限制的影响。
 
 默认值：256KiB。
 
+
+## max_parser_depth {#max_parser_depth}
+
+限制递归下降解析器中的最大递归深度。允许控制堆栈大小。
+
+可能的值：
+
+- 正整数。
+- 0 — 递归深度不受限制。
+
+默认值：1000。
+
 ## interactive_delay {#interactive-delay}
 
 以微秒为单位的间隔，用于检查请求执行是否已被取消并发送进度。
@@ -1064,6 +1075,28 @@ ClickHouse生成异常
 
 默认值：0。
 
+## optimize_functions_to_subcolumns {#optimize_functions_to_subcolumns}
+
+启用或禁用通过将某些函数转换为读取子列的优化。这减少了要读取的数据量。
+
+这些函数可以转化为：
+
+- [length](../../sql-reference/functions/array-functions.md/#array_functions-length) 读取 [size0](../../sql-reference/data-types/array.md/#array-size）子列。
+- [empty](../../sql-reference/functions/array-functions.md/#empty函数) 读取 [size0](../../sql-reference/data-types/array.md/#array-size）子列。
+- [notEmpty](../../sql-reference/functions/array-functions.md/#notempty函数) 读取 [size0](../../sql-reference/data-types/array.md/#array-size）子列。
+- [isNull](../../sql-reference/operators/index.md#operator-is-null) 读取 [null](../../sql-reference/data-types/nullable. md/#finding-null) 子列。
+- [isNotNull](../../sql-reference/operators/index.md#is-not-null) 读取 [null](../../sql-reference/data-types/nullable. md/#finding-null) 子列。
+- [count](../../sql-reference/aggregate-functions/reference/count.md) 读取 [null](../../sql-reference/data-types/nullable.md/#finding-null) 子列。
+- [mapKeys](../../sql-reference/functions/tuple-map-functions.mdx/#mapkeys) 读取 [keys](../../sql-reference/data-types/map.md/#map-subcolumns) 子列。
+- [mapValues](../../sql-reference/functions/tuple-map-functions.mdx/#mapvalues) 读取 [values](../../sql-reference/data-types/map.md/#map-subcolumns) 子列。
+
+可能的值：
+
+- 0 — 禁用优化。
+- 1 — 优化已启用。
+
+默认值：`0`。
+
 ## distributed_replica_error_half_life {#settings-distributed_replica_error_half_life}
 
 -   类型：秒

docs/zh/sql-reference/data-types/array.md

@@ -1,7 +1,7 @@
 ---
 slug: /zh/sql-reference/data-types/array
 ---
-# 阵列(T) {#data-type-array}
+# 数组(T) {#data-type-array}
 
 由 `T` 类型元素组成的数组。
 
@@ -66,3 +66,27 @@ SELECT array(1, 'a')
 Received exception from server (version 1.1.54388):
 Code: 386. DB::Exception: Received from localhost:9000, 127.0.0.1. DB::Exception: There is no supertype for types UInt8, String because some of them are String/FixedString and some of them are not.
 ```
+
+## 数组大小 {#array-size}
+
+可以使用 `size0` 子列找到数组的大小，而无需读取整个列。对于多维数组，您可以使用 `sizeN-1`，其中 `N` 是所需的维度。
+
+**例子**
+
+SQL查询：
+
+```sql
+CREATE TABLE t_arr (`arr` Array(Array(Array(UInt32)))) ENGINE = MergeTree ORDER BY tuple();
+
+INSERT INTO t_arr VALUES ([[[12, 13, 0, 1],[12]]]);
+
+SELECT arr.size0, arr.size1, arr.size2 FROM t_arr;
+```
+
+结果：
+
+``` text
+┌─arr.size0─┬─arr.size1─┬─arr.size2─┐
+│         1 │ [2]       │ [[4,1]]   │
+└───────────┴───────────┴───────────┘
+```

docs/zh/sql-reference/data-types/nullable.md

@@ -20,6 +20,33 @@ slug: /zh/sql-reference/data-types/nullable
 
 掩码文件中的条目允许ClickHouse区分每个表行的对应数据类型的«NULL»和默认值由于有额外的文件，«Nullable»列比普通列消耗更多的存储空间
 
+## null子列 {#finding-null}
+
+通过使用 `null` 子列可以在列中查找 `NULL` 值，而无需读取整个列。如果对应的值为 `NULL`，则返回 `1`，否则返回 `0`。
+
+**示例**
+
+SQL查询:
+
+``` sql
+CREATE TABLE nullable (`n` Nullable(UInt32)) ENGINE = MergeTree ORDER BY tuple();
+
+INSERT INTO nullable VALUES (1) (NULL) (2) (NULL);
+
+SELECT n.null FROM nullable;
+```
+
+结果:
+
+``` text
+┌─n.null─┐
+│      0 │
+│      1 │
+│      0 │
+│      1 │
+└────────┘
+```
+
 ## 用法示例 {#yong-fa-shi-li}
 
 ``` sql

programs/copier/Internals.cpp

@@ -259,7 +259,7 @@ ShardPriority getReplicasPriority(const Cluster::Addresses & replicas, const std
     res.is_remote = 1;
     for (const auto & replica : replicas)
     {
-        if (isLocalAddress(DNSResolver::instance().resolveHost(replica.host_name)))
+        if (isLocalAddress(DNSResolver::instance().resolveHostAllInOriginOrder(replica.host_name).front()))
         {
             res.is_remote = 0;
             break;

programs/copier/Internals.h

@@ -102,7 +102,7 @@ struct TaskStateWithOwner
         return TaskStateWithOwner(state, owner).toString();
     }
 
-    String toString()
+    String toString() const
     {
         WriteBufferFromOwnString wb;
         wb << static_cast<UInt32>(state) << "\n" << escape << owner;

programs/copier/ZooKeeperStaff.h

@@ -180,7 +180,7 @@ public:
                     auto logger = getLogger("ClusterCopier");
                     if (rsp.error == Coordination::Error::ZOK)
                     {
-                        switch (rsp.type)
+                        switch (rsp.type) /// NOLINT(bugprone-switch-missing-default-case)
                         {
                             case Coordination::CREATED:
                                 LOG_DEBUG(logger, "CleanStateClock change: CREATED, at {}", rsp.path);

programs/library-bridge/LibraryBridgeHandlers.cpp

@@ -1,6 +1,5 @@
 #include "LibraryBridgeHandlers.h"
 
-#include "CatBoostLibraryHandler.h"
 #include "CatBoostLibraryHandlerFactory.h"
 #include "Common/ProfileEvents.h"
 #include "ExternalDictionaryLibraryHandler.h"
@@ -11,10 +10,8 @@
 #include <IO/ReadHelpers.h>
 #include <Common/BridgeProtocolVersion.h>
 #include <IO/WriteHelpers.h>
-#include <Poco/Net/HTMLForm.h>
 #include <Poco/Net/HTTPServerRequest.h>
 #include <Poco/Net/HTTPServerResponse.h>
-#include <Poco/ThreadPool.h>
 #include <Processors/Executors/CompletedPipelineExecutor.h>
 #include <Processors/Executors/PullingPipelineExecutor.h>
 #include <Processors/Formats/IInputFormat.h>

programs/local/LocalServer.cpp

@@ -841,7 +841,7 @@ void LocalServer::addOptions(OptionsDescription & options_description)
 
         /// If structure argument is omitted then initial query is not generated
         ("structure,S", po::value<std::string>(), "structure of the initial table (list of column and type names)")
-        ("file,f", po::value<std::string>(), "path to file with data of the initial table (stdin if not specified)")
+        ("file,F", po::value<std::string>(), "path to file with data of the initial table (stdin if not specified)")
 
         ("input-format", po::value<std::string>(), "input format of the initial table data")
         ("output-format", po::value<std::string>(), "default output format")

programs/odbc-bridge/ColumnInfoHandler.cpp

@@ -8,13 +8,11 @@
 #include <IO/ReadHelpers.h>
 #include <IO/WriteHelpers.h>
 #include <Parsers/ParserQueryWithOutput.h>
-#include <Parsers/parseQuery.h>
 #include <Server/HTTP/HTMLForm.h>
 #include <Poco/Net/HTTPServerRequest.h>
 #include <Poco/Net/HTTPServerResponse.h>
 #include <Poco/NumberParser.h>
 #include <Common/logger_useful.h>
-#include <base/scope_guard.h>
 #include <Common/BridgeProtocolVersion.h>
 #include <Common/quoteString.h>
 #include "getIdentifierQuote.h"

programs/odbc-bridge/IdentifierQuoteHandler.cpp

@@ -7,13 +7,10 @@
 #include <Server/HTTP/WriteBufferFromHTTPServerResponse.h>
 #include <IO/ReadHelpers.h>
 #include <IO/WriteHelpers.h>
-#include <Parsers/ParserQueryWithOutput.h>
-#include <Parsers/parseQuery.h>
 #include <Poco/Net/HTTPServerRequest.h>
 #include <Poco/Net/HTTPServerResponse.h>
 #include <Common/BridgeProtocolVersion.h>
 #include <Common/logger_useful.h>
-#include <base/scope_guard.h>
 #include "getIdentifierQuote.h"
 #include "validateODBCConnectionString.h"
 #include "ODBCPooledConnectionFactory.h"

programs/odbc-bridge/ODBCBlockInputStream.cpp

@@ -3,12 +3,8 @@
 #include <IO/ReadBufferFromString.h>
 #include <DataTypes/DataTypeNullable.h>
 #include <DataTypes/DataTypeDateTime64.h>
-#include <Columns/ColumnNullable.h>
-#include <Columns/ColumnString.h>
-#include <Columns/ColumnsNumber.h>
 #include <Common/assert_cast.h>
 #include <IO/ReadHelpers.h>
-#include <IO/WriteHelpers.h>
 #include <Common/logger_useful.h>
 
 
@@ -54,21 +50,7 @@ Chunk ODBCSource::generate()
             const auto & sample = description.sample_block.getByPosition(idx);
 
             if (!result.is_null(idx))
-            {
-                bool is_nullable = description.types[idx].second;
-
-                if (is_nullable)
-                {
-                    ColumnNullable & column_nullable = assert_cast<ColumnNullable &>(*columns[idx]);
-                    const auto & data_type = assert_cast<const DataTypeNullable &>(*sample.type);
-                    insertValue(column_nullable.getNestedColumn(), data_type.getNestedType(), description.types[idx].first, result, idx);
-                    column_nullable.getNullMapData().emplace_back(0);
-                }
-                else
-                {
-                    insertValue(*columns[idx], sample.type, description.types[idx].first, result, idx);
-                }
-            }
+                insertValue(*columns[idx], removeNullable(sample.type), description.types[idx].first, result, idx);
             else
                 insertDefaultValue(*columns[idx], *sample.column);
         }
@@ -87,59 +69,60 @@ void ODBCSource::insertValue(
     switch (type)
     {
         case ValueType::vtUInt8:
-            assert_cast<ColumnUInt8 &>(column).insertValue(row.get<uint16_t>(idx));
+            column.insert(row.get<uint16_t>(idx));
             break;
         case ValueType::vtUInt16:
-            assert_cast<ColumnUInt16 &>(column).insertValue(row.get<uint16_t>(idx));
+            column.insert(row.get<uint16_t>(idx));
             break;
         case ValueType::vtUInt32:
-            assert_cast<ColumnUInt32 &>(column).insertValue(row.get<uint32_t>(idx));
+            column.insert(row.get<uint32_t>(idx));
             break;
         case ValueType::vtUInt64:
-            assert_cast<ColumnUInt64 &>(column).insertValue(row.get<uint64_t>(idx));
+            column.insert(row.get<uint64_t>(idx));
             break;
         case ValueType::vtInt8:
-            assert_cast<ColumnInt8 &>(column).insertValue(row.get<int16_t>(idx));
+            column.insert(row.get<int16_t>(idx));
             break;
         case ValueType::vtInt16:
-            assert_cast<ColumnInt16 &>(column).insertValue(row.get<int16_t>(idx));
+            column.insert(row.get<int16_t>(idx));
             break;
         case ValueType::vtInt32:
-            assert_cast<ColumnInt32 &>(column).insertValue(row.get<int32_t>(idx));
+            column.insert(row.get<int32_t>(idx));
             break;
         case ValueType::vtInt64:
-            assert_cast<ColumnInt64 &>(column).insertValue(row.get<int64_t>(idx));
+            column.insert(row.get<int64_t>(idx));
             break;
         case ValueType::vtFloat32:
-            assert_cast<ColumnFloat32 &>(column).insertValue(row.get<float>(idx));
+            column.insert(row.get<float>(idx));
             break;
         case ValueType::vtFloat64:
-            assert_cast<ColumnFloat64 &>(column).insertValue(row.get<double>(idx));
+            column.insert(row.get<double>(idx));
             break;
-        case ValueType::vtFixedString:[[fallthrough]];
+        case ValueType::vtFixedString:
         case ValueType::vtEnum8:
         case ValueType::vtEnum16:
         case ValueType::vtString:
-            assert_cast<ColumnString &>(column).insert(row.get<std::string>(idx));
+            column.insert(row.get<std::string>(idx));
             break;
         case ValueType::vtUUID:
         {
             auto value = row.get<std::string>(idx);
-            assert_cast<ColumnUInt128 &>(column).insert(parse<UUID>(value.data(), value.size()));
+            column.insert(parse<UUID>(value.data(), value.size()));
             break;
         }
         case ValueType::vtDate:
-            assert_cast<ColumnUInt16 &>(column).insertValue(UInt16{LocalDate{row.get<std::string>(idx)}.getDayNum()});
+            column.insert(UInt16{LocalDate{row.get<std::string>(idx)}.getDayNum()});
             break;
         case ValueType::vtDateTime:
         {
             auto value = row.get<std::string>(idx);
             ReadBufferFromString in(value);
             time_t time = 0;
-            readDateTimeText(time, in, assert_cast<const DataTypeDateTime *>(data_type.get())->getTimeZone());
+            const DataTypeDateTime & datetime_type = assert_cast<const DataTypeDateTime &>(*data_type);
+            readDateTimeText(time, in, datetime_type.getTimeZone());
             if (time < 0)
                 time = 0;
-            assert_cast<ColumnUInt32 &>(column).insertValue(static_cast<UInt32>(time));
+            column.insert(static_cast<UInt32>(time));
             break;
         }
         case ValueType::vtDateTime64:
@@ -147,14 +130,14 @@ void ODBCSource::insertValue(
             auto value = row.get<std::string>(idx);
             ReadBufferFromString in(value);
             DateTime64 time = 0;
-            const auto * datetime_type = assert_cast<const DataTypeDateTime64 *>(data_type.get());
-            readDateTime64Text(time, datetime_type->getScale(), in, datetime_type->getTimeZone());
-            assert_cast<DataTypeDateTime64::ColumnType &>(column).insertValue(time);
+            const DataTypeDateTime64 & datetime_type = assert_cast<const DataTypeDateTime64 &>(*data_type);
+            readDateTime64Text(time, datetime_type.getScale(), in, datetime_type.getTimeZone());
+            column.insert(time);
             break;
         }
-        case ValueType::vtDecimal32: [[fallthrough]];
-        case ValueType::vtDecimal64: [[fallthrough]];
-        case ValueType::vtDecimal128: [[fallthrough]];
+        case ValueType::vtDecimal32:
+        case ValueType::vtDecimal64:
+        case ValueType::vtDecimal128:
         case ValueType::vtDecimal256:
         {
             auto value = row.get<std::string>(idx);

programs/odbc-bridge/ODBCBlockOutputStream.cpp

@@ -1,6 +1,5 @@
 #include "ODBCBlockOutputStream.h"
 
-#include <Common/logger_useful.h>
 #include <IO/WriteBufferFromString.h>
 #include <Interpreters/Context.h>
 #include <Processors/Formats/IOutputFormat.h>

programs/odbc-bridge/ODBCPooledConnectionFactory.h

@@ -40,7 +40,6 @@ public:
 
     explicit ConnectionHolder(const String & connection_string_)
         : pool(nullptr)
-        , connection()
         , connection_string(connection_string_)
     {
         updateConnection();
@@ -143,7 +142,7 @@ public:
     {
         std::lock_guard lock(mutex);
 
-        if (!factory.count(connection_string))
+        if (!factory.contains(connection_string))
             factory.emplace(std::make_pair(connection_string, std::make_shared<nanodbc::Pool>(pool_size)));
 
         auto & pool = factory[connection_string];

programs/server/.gitignore

@@ -2,6 +2,7 @@
 /metadata_dropped
 /data
 /store
+/disks
 /access
 /flags
 /dictionaries_lib

programs/server/Server.cpp

@@ -46,6 +46,7 @@
 #include <Common/makeSocketAddress.h>
 #include <Common/FailPoint.h>
 #include <Common/CPUID.h>
+#include <Common/HTTPConnectionPool.h>
 #include <Server/waitServersToFinish.h>
 #include <Interpreters/Cache/FileCacheFactory.h>
 #include <Core/ServerUUID.h>
@@ -184,7 +185,7 @@ static bool jemallocOptionEnabled(const char *name)
     return value;
 }
 #else
-static bool jemallocOptionEnabled(const char *) { return 0; }
+static bool jemallocOptionEnabled(const char *) { return false; }
 #endif
 
 int mainEntryClickHouseServer(int argc, char ** argv)
@@ -1547,6 +1548,23 @@ try
 
             FileCacheFactory::instance().updateSettingsFromConfig(*config);
 
+            HTTPConnectionPools::instance().setLimits(
+                HTTPConnectionPools::Limits{
+                    new_server_settings.disk_connections_soft_limit,
+                    new_server_settings.disk_connections_warn_limit,
+                    new_server_settings.disk_connections_store_limit,
+                },
+                HTTPConnectionPools::Limits{
+                    new_server_settings.storage_connections_soft_limit,
+                    new_server_settings.storage_connections_warn_limit,
+                    new_server_settings.storage_connections_store_limit,
+                },
+                HTTPConnectionPools::Limits{
+                    new_server_settings.http_connections_soft_limit,
+                    new_server_settings.http_connections_warn_limit,
+                    new_server_settings.http_connections_store_limit,
+                });
+
             ProfileEvents::increment(ProfileEvents::MainConfigLoads);
 
             /// Must be the last.

programs/static-files-disk-uploader/static-files-disk-uploader.cpp

@@ -65,7 +65,7 @@ void processFile(const fs::path & file_path, const fs::path & dst_path, bool tes
 
         /// test mode for integration tests.
         if (test_mode)
-            dst_buf = std::make_shared<WriteBufferFromHTTP>(Poco::URI(dst_file_path), Poco::Net::HTTPRequest::HTTP_PUT);
+            dst_buf = std::make_shared<WriteBufferFromHTTP>(HTTPConnectionGroupType::HTTP, Poco::URI(dst_file_path), Poco::Net::HTTPRequest::HTTP_PUT);
         else
             dst_buf = std::make_shared<WriteBufferFromFile>(dst_file_path);
 
@@ -88,7 +88,7 @@ void processTableFiles(const fs::path & data_path, fs::path dst_path, bool test_
     {
         dst_path /= "store";
         auto files_root = dst_path / prefix;
-        root_meta = std::make_shared<WriteBufferFromHTTP>(Poco::URI(files_root / ".index"), Poco::Net::HTTPRequest::HTTP_PUT);
+        root_meta = std::make_shared<WriteBufferFromHTTP>(HTTPConnectionGroupType::HTTP, Poco::URI(files_root / ".index"), Poco::Net::HTTPRequest::HTTP_PUT);
     }
     else
     {
@@ -112,7 +112,7 @@ void processTableFiles(const fs::path & data_path, fs::path dst_path, bool test_
             if (test_mode)
             {
                 auto files_root = dst_path / prefix;
-                directory_meta = std::make_shared<WriteBufferFromHTTP>(Poco::URI(dst_path / directory_prefix / ".index"), Poco::Net::HTTPRequest::HTTP_PUT);
+                directory_meta = std::make_shared<WriteBufferFromHTTP>(HTTPConnectionGroupType::HTTP, Poco::URI(dst_path / directory_prefix / ".index"), Poco::Net::HTTPRequest::HTTP_PUT);
             }
             else
             {

pyproject.toml

@@ -1,6 +1,4 @@
-# vim: ft=config
-
-[BASIC]
+[tool.pylint.BASIC]
 max-module-lines=2000
 # due to SQL
 max-line-length=200
@@ -9,11 +7,13 @@ max-branches=50
 max-nested-blocks=10
 max-statements=200
 
-[FORMAT]
-ignore-long-lines = (# )?<?https?://\S+>?$
+[tool.pylint.FORMAT]
+#ignore-long-lines = (# )?<?https?://\S+>?$
 
-[MESSAGES CONTROL]
-disable = missing-docstring,
+[tool.pylint.'MESSAGES CONTROL']
+          # pytest.mark.parametrize is not callable (not-callable)
+disable = '''
+          missing-docstring,
           too-few-public-methods,
           invalid-name,
           too-many-arguments,
@@ -26,18 +26,15 @@ disable = missing-docstring,
           wildcard-import,
           unused-wildcard-import,
           singleton-comparison,
-          # pytest.mark.parametrize is not callable (not-callable)
           not-callable,
-          # https://github.com/PyCQA/pylint/issues/3882
-          # [Python 3.9] Value 'Optional' is unsubscriptable (unsubscriptable-object) (also Union)
-          unsubscriptable-object,
-          # Drop them one day:
           redefined-outer-name,
           broad-except,
           bare-except,
           no-else-return,
           global-statement
+          '''
 
-[SIMILARITIES]
+[tool.pylint.SIMILARITIES]
 # due to SQL
 min-similarity-lines=1000
+

rust/Cargo.lock

@@ -6,6 +6,7 @@ version = 3
 name = "_ch_rust_prql"
 version = "0.1.0"
 dependencies = [
+ "anstream",
  "prqlc",
  "serde_json",
 ]
@@ -698,9 +699,9 @@ checksum = "68354c5c6bd36d73ff3feceb05efa59b6acb7626617f4962be322a825e61f79a"
 
 [[package]]
 name = "miniz_oxide"
-version = "0.7.1"
+version = "0.7.2"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "e7810e0be55b428ada41041c41f32c9f1a42817901b4ccf45fa3d4b6561e74c7"
+checksum = "9d811f3e15f28568be3407c8e7fdb6514c1cda3cb30683f15b6a1a1dc4ea14a7"
 dependencies = [
  "adler",
 ]
@@ -751,9 +752,9 @@ dependencies = [
 
 [[package]]
 name = "object"
-version = "0.32.1"
+version = "0.32.2"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "9cf5f9dd3933bd50a9e1f149ec995f39ae2c496d31fd772c1fd45ebc27e902b0"
+checksum = "a6a622008b6e321afc04970976f62ee297fdbaa6f95318ca343e3eebb9648441"
 dependencies = [
  "memchr",
 ]

rust/prql/Cargo.toml

@@ -4,6 +4,7 @@ name = "_ch_rust_prql"
 version = "0.1.0"
 
 [dependencies]
+anstream = {version = "0.6.12"}
 prqlc = {version = "0.11.3", default-features = false}
 serde_json = "1.0"
 

rust/prql/src/lib.rs

@@ -39,6 +39,11 @@ pub unsafe extern "C" fn prql_to_sql_impl(
     };
 
     if let Ok(sql_str) = prqlc::compile(&query_str, &opts) {
+        // NOTE: Over at PRQL we're considering to un-deprecate & re-enable the
+        // `color: false` option. If that happens, we can remove the `strip_str`
+        // here, which strips color codes from the output.
+        use anstream::adapter::strip_str;
+        let sql_str = strip_str(&sql_str).to_string();
         set_output(sql_str, out, out_size);
         0
     } else {
@@ -54,17 +59,50 @@ pub unsafe extern "C" fn prql_to_sql(
     out: *mut *mut u8,
     out_size: *mut u64,
 ) -> i64 {
-    let ret = panic::catch_unwind(|| {
-        return prql_to_sql_impl(query, size, out, out_size);
-    });
-    return match ret {
-        // NOTE: using cxxbridge we can return proper Result<> type.
-        Err(_err) => 1,
-        Ok(res) => res,
-    }
+    // NOTE: using cxxbridge we can return proper Result<> type.
+    panic::catch_unwind(|| prql_to_sql_impl(query, size, out, out_size)).unwrap_or_else(|_| {
+        set_output("prqlc panicked".to_string(), out, out_size);
+        1
+    })
 }
 
 #[no_mangle]
 pub unsafe extern "C" fn prql_free_pointer(ptr_to_free: *mut u8) {
     std::mem::drop(CString::from_raw(ptr_to_free as *mut c_char));
 }
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+    use std::ffi::{CStr, CString};
+
+    /// A test helper to offer a rust interface to the C bindings
+    fn run_compile(query: &str) -> (String, i64) {
+        let query_cstr = CString::new(query).unwrap();
+        let query_ptr = query_cstr.as_ptr() as *const u8;
+        let query_size = query_cstr.to_bytes_with_nul().len() as u64 - 1; // Excluding the null terminator
+
+        let mut out: *mut u8 = std::ptr::null_mut();
+        let mut out_size = 0_u64;
+
+        unsafe {
+            let success = prql_to_sql(query_ptr, query_size, &mut out, &mut out_size);
+            let output = CStr::from_ptr(out as *const i8)
+                .to_str()
+                .unwrap()
+                .to_string();
+            prql_free_pointer(out);
+            (output, success)
+        }
+    }
+
+    #[test]
+    fn test_prql_to_sql() {
+        assert!(run_compile("from x").0.contains("SELECT"));
+        assert!(run_compile("asdf").1 == 1);
+        // In prqlc 0.11.3, this is a panic, so that allows us to test that the
+        // panic is caught. When we upgrade prqlc, it won't be a panic any
+        // longer.
+        assert!(run_compile("x -> y").1 == 1);
+    }
+}

src/Access/Common/AccessType.h

@@ -154,7 +154,8 @@ enum class AccessType
     M(SET_DEFINER, "", USER_NAME, ALL) \
     \
     M(SYSTEM_SHUTDOWN, "SYSTEM KILL, SHUTDOWN", GLOBAL, SYSTEM) \
-    M(SYSTEM_DROP_DNS_CACHE, "SYSTEM DROP DNS, DROP DNS CACHE, DROP DNS", GLOBAL, SYSTEM_DROP_CACHE) \
+    M(SYSTEM_DROP_DNS_CACHE, "SYSTEM DROP DNS, DROP DNS CACHE, DROP DNS", GLOBAL, SYSTEM_DROP_CACHE)  \
+    M(SYSTEM_DROP_CONNECTIONS_CACHE, "SYSTEM DROP CONNECTIONS CACHE, DROP CONNECTIONS CACHE", GLOBAL, SYSTEM_DROP_CACHE)  \
     M(SYSTEM_DROP_MARK_CACHE, "SYSTEM DROP MARK, DROP MARK CACHE, DROP MARKS", GLOBAL, SYSTEM_DROP_CACHE) \
     M(SYSTEM_DROP_UNCOMPRESSED_CACHE, "SYSTEM DROP UNCOMPRESSED, DROP UNCOMPRESSED CACHE, DROP UNCOMPRESSED", GLOBAL, SYSTEM_DROP_CACHE) \
     M(SYSTEM_DROP_MMAP_CACHE, "SYSTEM DROP MMAP, DROP MMAP CACHE, DROP MMAP", GLOBAL, SYSTEM_DROP_CACHE) \

src/Access/Common/AllowedClientHosts.cpp

@@ -55,7 +55,7 @@ namespace
     {
         IPAddress addr_v6 = toIPv6(address);
 
-        auto host_addresses = DNSResolver::instance().resolveHostAll(host);
+        auto host_addresses = DNSResolver::instance().resolveHostAllInOriginOrder(host);
 
         for (const auto & addr : host_addresses)
         {

src/Access/Common/QuotaDefs.cpp

@@ -49,71 +49,135 @@ String QuotaTypeInfo::valueToStringWithName(QuotaValue value) const
 
 const QuotaTypeInfo & QuotaTypeInfo::get(QuotaType type)
 {
-    static constexpr auto make_info = [](const char * raw_name_, UInt64 output_denominator_)
+    static constexpr auto make_info = [](const char * raw_name_, String current_usage_description_, String max_allowed_usage_description_, UInt64 output_denominator_)
     {
         String init_name = raw_name_;
         boost::to_lower(init_name);
         String init_keyword = raw_name_;
         boost::replace_all(init_keyword, "_", " ");
         bool init_output_as_float = (output_denominator_ != 1);
-        return QuotaTypeInfo{raw_name_, std::move(init_name), std::move(init_keyword), init_output_as_float, output_denominator_};
+        return QuotaTypeInfo
+        {
+            .raw_name = raw_name_,
+            .name = std::move(init_name),
+            .keyword = std::move(init_keyword),
+            .current_usage_description = std::move(current_usage_description_),
+            .max_allowed_usage_description = std::move(max_allowed_usage_description_),
+            .output_as_float = init_output_as_float,
+            .output_denominator = output_denominator_
+        };
     };
 
     switch (type)
     {
         case QuotaType::QUERIES:
         {
-            static const auto info = make_info("QUERIES", 1);
+            static const auto info = make_info(
+                "QUERIES",
+                "The current number of executed queries.",
+                "The maximum allowed number of queries of all types allowed to be executed.",
+                1
+            );
             return info;
         }
         case QuotaType::QUERY_SELECTS:
         {
-            static const auto info = make_info("QUERY_SELECTS", 1);
+            static const auto info = make_info(
+                "QUERY_SELECTS",
+                "The current number of executed SELECT queries.",
+                "The maximum allowed number of SELECT queries allowed to be executed.",
+                1
+            );
             return info;
         }
         case QuotaType::QUERY_INSERTS:
         {
-            static const auto info = make_info("QUERY_INSERTS", 1);
+            static const auto info = make_info(
+                "QUERY_INSERTS",
+                "The current number of executed INSERT queries.",
+                "The maximum allowed number of INSERT queries allowed to be executed.",
+                1
+            );
             return info;
         }
         case QuotaType::ERRORS:
         {
-            static const auto info = make_info("ERRORS", 1);
+            static const auto info = make_info(
+                "ERRORS",
+                "The current number of queries resulted in an error.",
+                "The maximum number of queries resulted in an error allowed within the specified period of time.",
+                1
+            );
             return info;
         }
         case QuotaType::RESULT_ROWS:
         {
-            static const auto info = make_info("RESULT_ROWS", 1);
+            static const auto info = make_info(
+                "RESULT_ROWS",
+                "The current total number of rows in the result set of all queries within the current period of time.",
+                "The maximum total number of rows in the result set of all queries allowed within the specified period of time.",
+                1
+            );
             return info;
         }
         case QuotaType::RESULT_BYTES:
         {
-            static const auto info = make_info("RESULT_BYTES", 1);
+            static const auto info = make_info(
+                "RESULT_BYTES",
+                "The current total number of bytes in the result set of all queries within the current period of time.",
+                "The maximum total number of bytes in the result set of all queries allowed within the specified period of time.",
+                1
+            );
             return info;
         }
         case QuotaType::READ_ROWS:
         {
-            static const auto info = make_info("READ_ROWS", 1);
+            static const auto info = make_info(
+                "READ_ROWS",
+                "The current total number of rows read during execution of all queries within the current period of time.",
+                "The maximum number of rows to read during execution of all queries allowed within the specified period of time.",
+                1
+            );
             return info;
         }
         case QuotaType::READ_BYTES:
         {
-            static const auto info = make_info("READ_BYTES", 1);
+            static const auto info = make_info(
+                "READ_BYTES",
+                "The current total number of bytes read during execution of all queries within the current period of time.",
+                "The maximum number of bytes to read during execution of all queries allowed within the specified period of time.",
+                1
+            );
             return info;
         }
         case QuotaType::EXECUTION_TIME:
         {
-            static const auto info = make_info("EXECUTION_TIME", 1000000000 /* execution_time is stored in nanoseconds */);
+            static const auto info = make_info(
+                "EXECUTION_TIME",
+                "The current total amount of time (in nanoseconds) spent to execute queries within the current period of time",
+                "The maximum amount of time (in nanoseconds) allowed for all queries to execute within the specified period of time",
+                1000000000 /* execution_time is stored in nanoseconds */
+            );
             return info;
         }
         case QuotaType::WRITTEN_BYTES:
         {
-            static const auto info = make_info("WRITTEN_BYTES", 1);
+            static const auto info = make_info(
+                "WRITTEN_BYTES",
+                "The current total number of bytes written during execution of all queries within the current period of time.",
+                "The maximum number of bytes to written during execution of all queries allowed within the specified period of time.",
+                1
+            );
             return info;
         }
         case QuotaType::FAILED_SEQUENTIAL_AUTHENTICATIONS:
         {
-            static const auto info = make_info("FAILED_SEQUENTIAL_AUTHENTICATIONS", 1);
+            static const auto info = make_info(
+                "FAILED_SEQUENTIAL_AUtheNTICATIONS",
+                "The current number of consecutive authentication failures within the current period of time.",
+                "The maximum number of consecutive authentication failures allowed within the specified period of time.",
+                1
+            );
             return info;
         }
         case QuotaType::MAX: break;

src/Access/Common/QuotaDefs.h

@@ -33,6 +33,8 @@ struct QuotaTypeInfo
     const char * const raw_name = "";
     const String name;    /// Lowercased with underscores, e.g. "result_rows".
     const String keyword; /// Uppercased with spaces, e.g. "RESULT ROWS".
+    const String current_usage_description;
+    const String max_allowed_usage_description;
     const bool output_as_float = false;
     const UInt64 output_denominator = 1;
     String valueToString(QuotaValue value) const;

src/Access/Common/RowPolicyDefs.cpp

@@ -33,7 +33,7 @@ String toString(RowPolicyFilterType type)
 
 const RowPolicyFilterTypeInfo & RowPolicyFilterTypeInfo::get(RowPolicyFilterType type_)
 {
-    static constexpr auto make_info = [](const char * raw_name_)
+    static constexpr auto make_info = [](const char * raw_name_, const String & comment_)
     {
         String init_name = raw_name_;
         boost::to_lower(init_name);
@@ -41,14 +41,17 @@ const RowPolicyFilterTypeInfo & RowPolicyFilterTypeInfo::get(RowPolicyFilterType
         String init_command = init_name.substr(0, underscore_pos);
         boost::to_upper(init_command);
         bool init_is_check = (std::string_view{init_name}.substr(underscore_pos + 1) == "check");
-        return RowPolicyFilterTypeInfo{raw_name_, std::move(init_name), std::move(init_command), init_is_check};
+        return RowPolicyFilterTypeInfo{raw_name_, std::move(init_name), std::move(init_command), comment_, init_is_check};
     };
 
     switch (type_)
     {
         case RowPolicyFilterType::SELECT_FILTER:
         {
-            static const auto info = make_info("SELECT_FILTER");
+            static const auto info = make_info(
+                "SELECT_FILTER",
+                "Expression which is used for filtering in SELECT queries."
+            );
             return info;
         }
 #if 0 /// Row-level security for INSERT, UPDATE, DELETE is not implemented yet.

src/Access/Common/RowPolicyDefs.h

@@ -52,6 +52,7 @@ struct RowPolicyFilterTypeInfo
     const char * const raw_name;
     const String name;    /// Lowercased with underscores, e.g. "select_filter".
     const String command; /// Uppercased without last word, e.g. "SELECT".
+    const String description;
     const bool is_check;  /// E.g. false for SELECT_FILTER.
     static const RowPolicyFilterTypeInfo & get(RowPolicyFilterType type);
 };

src/Access/HTTPAuthClient.h

@@ -36,7 +36,7 @@ public:
 
     Result authenticateRequest(Poco::Net::HTTPRequest & request) const
     {
-        auto session = makeHTTPSession(uri, timeouts);
+        auto session = makeHTTPSession(HTTPConnectionGroupType::HTTP, uri, timeouts);
         Poco::Net::HTTPResponse response;
 
         auto milliseconds_to_wait = retry_initial_backoff_ms;

src/Analyzer/ConstantNode.cpp

@@ -1,5 +1,7 @@
 #include <Analyzer/ConstantNode.h>
 
+#include <Analyzer/FunctionNode.h>
+
 #include <Common/assert_cast.h>
 #include <Common/FieldVisitorToString.h>
 #include <Common/SipHash.h>
@@ -38,52 +40,9 @@ ConstantNode::ConstantNode(Field value_)
     : ConstantNode(value_, applyVisitor(FieldToDataType(), value_))
 {}
 
-void ConstantNode::dumpTreeImpl(WriteBuffer & buffer, FormatState & format_state, size_t indent) const
-{
-    buffer << std::string(indent, ' ') << "CONSTANT id: " << format_state.getNodeId(this);
-
-    if (hasAlias())
-        buffer << ", alias: " << getAlias();
-
-    buffer << ", constant_value: " << constant_value->getValue().dump();
-    buffer << ", constant_value_type: " << constant_value->getType()->getName();
-
-    if (getSourceExpression())
-    {
-        buffer << '\n' << std::string(indent + 2, ' ') << "EXPRESSION" << '\n';
-        getSourceExpression()->dumpTreeImpl(buffer, format_state, indent + 4);
-    }
-}
-
-bool ConstantNode::isEqualImpl(const IQueryTreeNode & rhs) const
-{
-    const auto & rhs_typed = assert_cast<const ConstantNode &>(rhs);
-    return *constant_value == *rhs_typed.constant_value && value_string == rhs_typed.value_string;
-}
-
-void ConstantNode::updateTreeHashImpl(HashState & hash_state) const
-{
-    auto type_name = constant_value->getType()->getName();
-    hash_state.update(type_name.size());
-    hash_state.update(type_name);
-
-    hash_state.update(value_string.size());
-    hash_state.update(value_string);
-}
-
-QueryTreeNodePtr ConstantNode::cloneImpl() const
-{
-    return std::make_shared<ConstantNode>(constant_value, source_expression);
-}
-
-ASTPtr ConstantNode::toASTImpl(const ConvertToASTOptions & options) const
+bool ConstantNode::requiresCastCall() const
 {
     const auto & constant_value_literal = constant_value->getValue();
-    auto constant_value_ast = std::make_shared<ASTLiteral>(constant_value_literal);
-
-    if (!options.add_cast_for_constants)
-        return constant_value_ast;
-
     bool need_to_add_cast_function = false;
     auto constant_value_literal_type = constant_value_literal.getType();
     WhichDataType constant_value_type(constant_value->getType());
@@ -131,7 +90,72 @@ ASTPtr ConstantNode::toASTImpl(const ConvertToASTOptions & options) const
     // Add cast if constant was created as a result of constant folding.
     // Constant folding may lead to type transformation and literal on shard
     // may have a different type.
-    if (need_to_add_cast_function || source_expression != nullptr)
+    return need_to_add_cast_function || source_expression != nullptr;
+}
+
+bool ConstantNode::receivedFromInitiatorServer() const
+{
+    if (!hasSourceExpression())
+        return false;
+
+    auto * cast_function = getSourceExpression()->as<FunctionNode>();
+    if (!cast_function || cast_function->getFunctionName() != "_CAST")
+        return false;
+    return true;
+}
+
+void ConstantNode::dumpTreeImpl(WriteBuffer & buffer, FormatState & format_state, size_t indent) const
+{
+    buffer << std::string(indent, ' ') << "CONSTANT id: " << format_state.getNodeId(this);
+
+    if (hasAlias())
+        buffer << ", alias: " << getAlias();
+
+    buffer << ", constant_value: ";
+    if (mask_id)
+        buffer << "[HIDDEN id: " << mask_id << "]";
+    else
+        buffer << constant_value->getValue().dump();
+
+    buffer << ", constant_value_type: " << constant_value->getType()->getName();
+
+    if (!mask_id && getSourceExpression())
+    {
+        buffer << '\n' << std::string(indent + 2, ' ') << "EXPRESSION" << '\n';
+        getSourceExpression()->dumpTreeImpl(buffer, format_state, indent + 4);
+    }
+}
+
+bool ConstantNode::isEqualImpl(const IQueryTreeNode & rhs) const
+{
+    const auto & rhs_typed = assert_cast<const ConstantNode &>(rhs);
+    return *constant_value == *rhs_typed.constant_value && value_string == rhs_typed.value_string;
+}
+
+void ConstantNode::updateTreeHashImpl(HashState & hash_state) const
+{
+    auto type_name = constant_value->getType()->getName();
+    hash_state.update(type_name.size());
+    hash_state.update(type_name);
+
+    hash_state.update(value_string.size());
+    hash_state.update(value_string);
+}
+
+QueryTreeNodePtr ConstantNode::cloneImpl() const
+{
+    return std::make_shared<ConstantNode>(constant_value, source_expression);
+}
+
+ASTPtr ConstantNode::toASTImpl(const ConvertToASTOptions & options) const
+{
+    const auto & constant_value_literal = constant_value->getValue();
+    auto constant_value_ast = std::make_shared<ASTLiteral>(constant_value_literal);
+
+    if (!options.add_cast_for_constants)
+        return constant_value_ast;
+
+    if (requiresCastCall())
     {
         auto constant_type_name_ast = std::make_shared<ASTLiteral>(constant_value->getType()->getName());
         return makeASTFunction("_CAST", std::move(constant_value_ast), std::move(constant_type_name_ast));

src/Analyzer/ConstantNode.h

@@ -75,6 +75,17 @@ public:
         return constant_value->getType();
     }
 
+    /// Check if conversion to AST requires wrapping with _CAST function.
+    bool requiresCastCall() const;
+
+    /// Check if constant is a result of _CAST function constant folding.
+    bool receivedFromInitiatorServer() const;
+
+    void setMaskId(size_t id)
+    {
+        mask_id = id;
+    }
+
     void dumpTreeImpl(WriteBuffer & buffer, FormatState & format_state, size_t indent) const override;
 
 protected:
@@ -90,6 +101,7 @@ private:
     ConstantValuePtr constant_value;
     String value_string;
     QueryTreeNodePtr source_expression;
+    size_t mask_id = 0;
 
     static constexpr size_t children_size = 0;
 };

src/Analyzer/FunctionSecretArgumentsFinderTreeNode.h

@@ -0,0 +1,372 @@
+#pragma once
+
+#include <Parsers/FunctionSecretArgumentsFinder.h>
+#include <Analyzer/ConstantNode.h>
+#include <Analyzer/FunctionNode.h>
+#include <Analyzer/IQueryTreeNode.h>
+#include <Analyzer/IdentifierNode.h>
+#include <Analyzer/ListNode.h>
+#include <Common/KnownObjectNames.h>
+#include <Core/QualifiedTableName.h>
+
+#include <boost/algorithm/string/predicate.hpp>
+
+
+namespace DB
+{
+
+
+/// Finds arguments of a specified function which should not be displayed for most users for security reasons.
+/// That involves passwords and secret keys.
+class FunctionSecretArgumentsFinderTreeNode
+{
+public:
+    explicit FunctionSecretArgumentsFinderTreeNode(const FunctionNode & function_) : function(function_), arguments(function.getArguments())
+    {
+        if (arguments.getNodes().empty())
+            return;
+
+        findFunctionSecretArguments();
+    }
+
+    struct Result
+    {
+        /// Result constructed by default means no arguments will be hidden.
+        size_t start = static_cast<size_t>(-1);
+        size_t count = 0; /// Mostly it's either 0 or 1. There are only a few cases where `count` can be greater than 1 (e.g. see `encrypt`).
+                            /// In all known cases secret arguments are consecutive
+        bool are_named = false; /// Arguments like `password = 'password'` are considered as named arguments.
+        /// E.g. "headers" in `url('..', headers('foo' = '[HIDDEN]'))`
+        std::vector<std::string> nested_maps;
+
+        bool hasSecrets() const
+        {
+            return count != 0 || !nested_maps.empty();
+        }
+    };
+
+    FunctionSecretArgumentsFinder::Result getResult() const { return result; }
+
+private:
+    const FunctionNode & function;
+    const ListNode & arguments;
+    FunctionSecretArgumentsFinder::Result result;
+
+    void markSecretArgument(size_t index, bool argument_is_named = false)
+    {
+        if (index >= arguments.getNodes().size())
+            return;
+        if (!result.count)
+        {
+            result.start = index;
+            result.are_named = argument_is_named;
+        }
+        chassert(index >= result.start); /// We always check arguments consecutively
+        result.count = index + 1 - result.start;
+        if (!argument_is_named)
+            result.are_named = false;
+    }
+
+    void findFunctionSecretArguments()
+    {
+        const auto & name = function.getFunctionName();
+
+        if ((name == "mysql") || (name == "postgresql") || (name == "mongodb"))
+        {
+            /// mysql('host:port', 'database', 'table', 'user', 'password', ...)
+            /// postgresql('host:port', 'database', 'table', 'user', 'password', ...)
+            /// mongodb('host:port', 'database', 'collection', 'user', 'password', ...)
+            findMySQLFunctionSecretArguments();
+        }
+        else if ((name == "s3") || (name == "cosn") || (name == "oss") ||
+                    (name == "deltaLake") || (name == "hudi") || (name == "iceberg"))
+        {
+            /// s3('url', 'aws_access_key_id', 'aws_secret_access_key', ...)
+            findS3FunctionSecretArguments(/* is_cluster_function= */ false);
+        }
+        else if (name == "s3Cluster")
+        {
+            /// s3Cluster('cluster_name', 'url', 'aws_access_key_id', 'aws_secret_access_key', ...)
+            findS3FunctionSecretArguments(/* is_cluster_function= */ true);
+        }
+        else if ((name == "remote") || (name == "remoteSecure"))
+        {
+            /// remote('addresses_expr', 'db', 'table', 'user', 'password', ...)
+            findRemoteFunctionSecretArguments();
+        }
+        else if ((name == "encrypt") || (name == "decrypt") ||
+                    (name == "aes_encrypt_mysql") || (name == "aes_decrypt_mysql") ||
+                    (name == "tryDecrypt"))
+        {
+            /// encrypt('mode', 'plaintext', 'key' [, iv, aad])
+            findEncryptionFunctionSecretArguments();
+        }
+        else if (name == "url")
+        {
+            findURLSecretArguments();
+        }
+    }
+
+    void findMySQLFunctionSecretArguments()
+    {
+        if (isNamedCollectionName(0))
+        {
+            /// mysql(named_collection, ..., password = 'password', ...)
+            findSecretNamedArgument("password", 1);
+        }
+        else
+        {
+            /// mysql('host:port', 'database', 'table', 'user', 'password', ...)
+            markSecretArgument(4);
+        }
+    }
+
+    /// Returns the number of arguments excluding "headers" and "extra_credentials" (which should
+    /// always be at the end). Marks "headers" as secret, if found.
+    size_t excludeS3OrURLNestedMaps()
+    {
+        const auto & nodes = arguments.getNodes();
+        size_t count = nodes.size();
+        while (count > 0)
+        {
+            const FunctionNode * f = nodes.at(count - 1)->as<FunctionNode>();
+            if (!f)
+                break;
+            if (f->getFunctionName() == "headers")
+                result.nested_maps.push_back(f->getFunctionName());
+            else if (f->getFunctionName() != "extra_credentials")
+                break;
+            count -= 1;
+        }
+        return count;
+    }
+
+    void findS3FunctionSecretArguments(bool is_cluster_function)
+    {
+        /// s3Cluster('cluster_name', 'url', ...) has 'url' as its second argument.
+        size_t url_arg_idx = is_cluster_function ? 1 : 0;
+
+        if (!is_cluster_function && isNamedCollectionName(0))
+        {
+            /// s3(named_collection, ..., secret_access_key = 'secret_access_key', ...)
+            findSecretNamedArgument("secret_access_key", 1);
+            return;
+        }
+
+        /// We should check other arguments first because we don't need to do any replacement in case of
+        /// s3('url', NOSIGN, 'format' [, 'compression'] [, extra_credentials(..)] [, headers(..)])
+        /// s3('url', 'format', 'structure' [, 'compression'] [, extra_credentials(..)] [, headers(..)])
+        size_t count = excludeS3OrURLNestedMaps();
+        if ((url_arg_idx + 3 <= count) && (count <= url_arg_idx + 4))
+        {
+            String second_arg;
+            if (tryGetStringFromArgument(url_arg_idx + 1, &second_arg))
+            {
+                if (boost::iequals(second_arg, "NOSIGN"))
+                    return; /// The argument after 'url' is "NOSIGN".
+
+                if (second_arg == "auto" || KnownFormatNames::instance().exists(second_arg))
+                    return; /// The argument after 'url' is a format: s3('url', 'format', ...)
+            }
+        }
+
+        /// We're going to replace 'aws_secret_access_key' with '[HIDDEN]' for the following signatures:
+        /// s3('url', 'aws_access_key_id', 'aws_secret_access_key', ...)
+        /// s3Cluster('cluster_name', 'url', 'aws_access_key_id', 'aws_secret_access_key', 'format', 'compression')
+        if (url_arg_idx + 2 < count)
+            markSecretArgument(url_arg_idx + 2);
+    }
+
+    void findURLSecretArguments()
+    {
+        if (!isNamedCollectionName(0))
+            excludeS3OrURLNestedMaps();
+    }
+
+    bool tryGetStringFromArgument(size_t arg_idx, String * res, bool allow_identifier = true) const
+    {
+        if (arg_idx >= arguments.getNodes().size())
+            return false;
+
+        return tryGetStringFromArgument(arguments.getNodes()[arg_idx], res, allow_identifier);
+    }
+
+    static bool tryGetStringFromArgument(const QueryTreeNodePtr argument, String * res, bool allow_identifier = true)
+    {
+        if (const auto * literal = argument->as<ConstantNode>())
+        {
+            if (literal->getValue().getType() != Field::Types::String)
+                return false;
+            if (res)
+                *res = literal->getValue().safeGet<String>();
+            return true;
+        }
+
+        if (allow_identifier)
+        {
+            if (const auto * id = argument->as<IdentifierNode>())
+            {
+                if (res)
+                    *res = id->getIdentifier().getFullName();
+                return true;
+            }
+        }
+
+        return false;
+    }
+
+    void findRemoteFunctionSecretArguments()
+    {
+        if (isNamedCollectionName(0))
+        {
+            /// remote(named_collection, ..., password = 'password', ...)
+            findSecretNamedArgument("password", 1);
+            return;
+        }
+
+        /// We're going to replace 'password' with '[HIDDEN'] for the following signatures:
+        /// remote('addresses_expr', db.table, 'user' [, 'password'] [, sharding_key])
+        /// remote('addresses_expr', 'db', 'table', 'user' [, 'password'] [, sharding_key])
+        /// remote('addresses_expr', table_function(), 'user' [, 'password'] [, sharding_key])
+
+        /// But we should check the number of arguments first because we don't need to do any replacements in case of
+        /// remote('addresses_expr', db.table)
+        if (arguments.getNodes().size() < 3)
+            return;
+
+        size_t arg_num = 1;
+
+        /// Skip 1 or 2 arguments with table_function() or db.table or 'db', 'table'.
+        const auto * table_function = arguments.getNodes()[arg_num]->as<FunctionNode>();
+        if (table_function && KnownTableFunctionNames::instance().exists(table_function->getFunctionName()))
+        {
+            ++arg_num;
+        }
+        else
+        {
+            std::optional<String> database;
+            std::optional<QualifiedTableName> qualified_table_name;
+            if (!tryGetDatabaseNameOrQualifiedTableName(arg_num, database, qualified_table_name))
+            {
+                /// We couldn't evaluate the argument so we don't know whether it is 'db.table' or just 'db'.
+                /// Hence we can't figure out whether we should skip one argument 'user' or two arguments 'table', 'user'
+                /// before the argument 'password'. So it's safer to wipe two arguments just in case.
+                /// The last argument can be also a `sharding_key`, so we need to check that argument is a literal string
+                /// before wiping it (because the `password` argument is always a literal string).
+                if (tryGetStringFromArgument(arg_num + 2, nullptr, /* allow_identifier= */ false))
+                {
+                    /// Wipe either `password` or `user`.
+                    markSecretArgument(arg_num + 2);
+                }
+                if (tryGetStringFromArgument(arg_num + 3, nullptr, /* allow_identifier= */ false))
+                {
+                    /// Wipe either `password` or `sharding_key`.
+                    markSecretArgument(arg_num + 3);
+                }
+                return;
+            }
+
+            /// Skip the current argument (which is either a database name or a qualified table name).
+            ++arg_num;
+            if (database)
+            {
+                /// Skip the 'table' argument if the previous argument was a database name.
+                ++arg_num;
+            }
+        }
+
+        /// Skip username.
+        ++arg_num;
+
+        /// Do our replacement:
+        /// remote('addresses_expr', db.table, 'user', 'password', ...) -> remote('addresses_expr', db.table, 'user', '[HIDDEN]', ...)
+        /// The last argument can be also a `sharding_key`, so we need to check that argument is a literal string
+        /// before wiping it (because the `password` argument is always a literal string).
+        bool can_be_password = tryGetStringFromArgument(arg_num, nullptr, /* allow_identifier= */ false);
+        if (can_be_password)
+            markSecretArgument(arg_num);
+    }
+
+    /// Tries to get either a database name or a qualified table name from an argument.
+    /// Empty string is also allowed (it means the default database).
+    /// The function is used by findRemoteFunctionSecretArguments() to determine how many arguments to skip before a password.
+    bool tryGetDatabaseNameOrQualifiedTableName(
+        size_t arg_idx,
+        std::optional<String> & res_database,
+        std::optional<QualifiedTableName> & res_qualified_table_name) const
+    {
+        res_database.reset();
+        res_qualified_table_name.reset();
+
+        String str;
+        if (!tryGetStringFromArgument(arg_idx, &str, /* allow_identifier= */ true))
+            return false;
+
+        if (str.empty())
+        {
+            res_database = "";
+            return true;
+        }
+
+        auto qualified_table_name = QualifiedTableName::tryParseFromString(str);
+        if (!qualified_table_name)
+            return false;
+
+        if (qualified_table_name->database.empty())
+            res_database = std::move(qualified_table_name->table);
+        else
+            res_qualified_table_name = std::move(qualified_table_name);
+        return true;
+    }
+
+    void findEncryptionFunctionSecretArguments()
+    {
+        if (arguments.getNodes().empty())
+            return;
+
+        /// We replace all arguments after 'mode' with '[HIDDEN]':
+        /// encrypt('mode', 'plaintext', 'key' [, iv, aad]) -> encrypt('mode', '[HIDDEN]')
+        result.start = 1;
+        result.count = arguments.getNodes().size() - 1;
+    }
+
+
+    /// Whether a specified argument can be the name of a named collection?
+    bool isNamedCollectionName(size_t arg_idx) const
+    {
+        if (arguments.getNodes().size() <= arg_idx)
+            return false;
+
+        const auto * identifier = arguments.getNodes()[arg_idx]->as<IdentifierNode>();
+        return identifier != nullptr;
+    }
+
+    /// Looks for a secret argument with a specified name. This function looks for arguments in format `key=value` where the key is specified.
+    void findSecretNamedArgument(const std::string_view & key, size_t start = 0)
+    {
+        for (size_t i = start; i < arguments.getNodes().size(); ++i)
+        {
+            const auto & argument = arguments.getNodes()[i];
+            const auto * equals_func = argument->as<FunctionNode>();
+            if (!equals_func || (equals_func->getFunctionName() != "equals"))
+                continue;
+
+            const auto * expr_list = equals_func->getArguments().as<ListNode>();
+            if (!expr_list)
+                continue;
+
+            const auto & equal_args = expr_list->getNodes();
+            if (equal_args.size() != 2)
+                continue;
+
+            String found_key;
+            if (!tryGetStringFromArgument(equal_args[0], &found_key))
+                continue;
+
+            if (found_key == key)
+                markSecretArgument(i, /* argument_is_named= */ true);
+        }
+    }
+};
+
+}

src/Analyzer/Passes/OptimizeDateOrDateTimeConverterWithPreimagePass.cpp

@@ -94,7 +94,8 @@ public:
         if (!func_node || func_node->getArguments().getNodes().size() != 1)
             return;
 
-        const auto * column_id = func_node->getArguments().getNodes()[0]->as<ColumnNode>();
+        const auto & argument_node = func_node->getArguments().getNodes()[0];
+        const auto * column_id = argument_node->as<ColumnNode>();
         if (!column_id)
             return;
 
@@ -119,7 +120,7 @@ public:
         if (!preimage_range)
             return;
 
-        const auto new_node = generateOptimizedDateFilter(comparator, *column_id, *preimage_range);
+        const auto new_node = generateOptimizedDateFilter(comparator, argument_node, *preimage_range);
 
         if (!new_node)
             return;
@@ -128,20 +129,22 @@ public:
     }
 
 private:
-    QueryTreeNodePtr
-    generateOptimizedDateFilter(const String & comparator, const ColumnNode & column_node, const std::pair<Field, Field> & range) const
+    QueryTreeNodePtr generateOptimizedDateFilter(
+        const String & comparator, const QueryTreeNodePtr & column_node, const std::pair<Field, Field> & range) const
     {
         const DateLUTImpl & date_lut = DateLUT::instance("UTC");
 
         String start_date_or_date_time;
         String end_date_or_date_time;
 
-        if (isDateOrDate32(column_node.getColumnType().get()))
+        const auto & column_node_typed = column_node->as<ColumnNode &>();
+        const auto & column_type = column_node_typed.getColumnType().get();
+        if (isDateOrDate32(column_type))
         {
             start_date_or_date_time = date_lut.dateToString(range.first.get<DateLUTImpl::Time>());
             end_date_or_date_time = date_lut.dateToString(range.second.get<DateLUTImpl::Time>());
         }
-        else if (isDateTime(column_node.getColumnType().get()) || isDateTime64(column_node.getColumnType().get()))
+        else if (isDateTime(column_type) || isDateTime64(column_type))
         {
             start_date_or_date_time = date_lut.timeToString(range.first.get<DateLUTImpl::Time>());
             end_date_or_date_time = date_lut.timeToString(range.second.get<DateLUTImpl::Time>());
@@ -151,69 +154,29 @@ private:
 
         if (comparator == "equals")
         {
-            const auto lhs = std::make_shared<FunctionNode>("greaterOrEquals");
-            lhs->getArguments().getNodes().push_back(std::make_shared<ColumnNode>(column_node.getColumn(), column_node.getColumnSource()));
-            lhs->getArguments().getNodes().push_back(std::make_shared<ConstantNode>(start_date_or_date_time));
-            resolveOrdinaryFunctionNode(*lhs, lhs->getFunctionName());
-
-            const auto rhs = std::make_shared<FunctionNode>("less");
-            rhs->getArguments().getNodes().push_back(std::make_shared<ColumnNode>(column_node.getColumn(), column_node.getColumnSource()));
-            rhs->getArguments().getNodes().push_back(std::make_shared<ConstantNode>(end_date_or_date_time));
-            resolveOrdinaryFunctionNode(*rhs, rhs->getFunctionName());
-
-            const auto new_date_filter = std::make_shared<FunctionNode>("and");
-            new_date_filter->getArguments().getNodes() = {lhs, rhs};
-            resolveOrdinaryFunctionNode(*new_date_filter, new_date_filter->getFunctionName());
-
-            return new_date_filter;
+            return createFunctionNode(
+                "and",
+                createFunctionNode("greaterOrEquals", column_node, std::make_shared<ConstantNode>(start_date_or_date_time)),
+                createFunctionNode("less", column_node, std::make_shared<ConstantNode>(end_date_or_date_time)));
         }
         else if (comparator == "notEquals")
         {
-            const auto lhs = std::make_shared<FunctionNode>("less");
-            lhs->getArguments().getNodes().push_back(std::make_shared<ColumnNode>(column_node.getColumn(), column_node.getColumnSource()));
-            lhs->getArguments().getNodes().push_back(std::make_shared<ConstantNode>(start_date_or_date_time));
-            resolveOrdinaryFunctionNode(*lhs, lhs->getFunctionName());
-
-            const auto rhs = std::make_shared<FunctionNode>("greaterOrEquals");
-            rhs->getArguments().getNodes().push_back(std::make_shared<ColumnNode>(column_node.getColumn(), column_node.getColumnSource()));
-            rhs->getArguments().getNodes().push_back(std::make_shared<ConstantNode>(end_date_or_date_time));
-            resolveOrdinaryFunctionNode(*rhs, rhs->getFunctionName());
-
-            const auto new_date_filter = std::make_shared<FunctionNode>("or");
-            new_date_filter->getArguments().getNodes() = {lhs, rhs};
-            resolveOrdinaryFunctionNode(*new_date_filter, new_date_filter->getFunctionName());
-
-            return new_date_filter;
+            return createFunctionNode(
+                "or",
+                createFunctionNode("less", column_node, std::make_shared<ConstantNode>(start_date_or_date_time)),
+                createFunctionNode("greaterOrEquals", column_node, std::make_shared<ConstantNode>(end_date_or_date_time)));
         }
         else if (comparator == "greater")
         {
-            const auto new_date_filter = std::make_shared<FunctionNode>("greaterOrEquals");
-            new_date_filter->getArguments().getNodes().push_back(
-                std::make_shared<ColumnNode>(column_node.getColumn(), column_node.getColumnSource()));
-            new_date_filter->getArguments().getNodes().push_back(std::make_shared<ConstantNode>(end_date_or_date_time));
-            resolveOrdinaryFunctionNode(*new_date_filter, new_date_filter->getFunctionName());
-
-            return new_date_filter;
+            return createFunctionNode("greaterOrEquals", column_node, std::make_shared<ConstantNode>(end_date_or_date_time));
         }
         else if (comparator == "lessOrEquals")
         {
-            const auto new_date_filter = std::make_shared<FunctionNode>("less");
-            new_date_filter->getArguments().getNodes().push_back(
-                std::make_shared<ColumnNode>(column_node.getColumn(), column_node.getColumnSource()));
-            new_date_filter->getArguments().getNodes().push_back(std::make_shared<ConstantNode>(end_date_or_date_time));
-            resolveOrdinaryFunctionNode(*new_date_filter, new_date_filter->getFunctionName());
-
-            return new_date_filter;
+            return createFunctionNode("less", column_node, std::make_shared<ConstantNode>(end_date_or_date_time));
         }
         else if (comparator == "less" || comparator == "greaterOrEquals")
         {
-            const auto new_date_filter = std::make_shared<FunctionNode>(comparator);
-            new_date_filter->getArguments().getNodes().push_back(
-                std::make_shared<ColumnNode>(column_node.getColumn(), column_node.getColumnSource()));
-            new_date_filter->getArguments().getNodes().push_back(std::make_shared<ConstantNode>(start_date_or_date_time));
-            resolveOrdinaryFunctionNode(*new_date_filter, new_date_filter->getFunctionName());
-
-            return new_date_filter;
+            return createFunctionNode(comparator, column_node, std::make_shared<ConstantNode>(start_date_or_date_time));
         }
         else [[unlikely]]
         {
@@ -224,10 +187,17 @@ private:
         }
     }
 
-    void resolveOrdinaryFunctionNode(FunctionNode & function_node, const String & function_name) const
+    template <typename... Args>
+    QueryTreeNodePtr createFunctionNode(const String & function_name, Args &&... args) const
     {
         auto function = FunctionFactory::instance().get(function_name, getContext());
-        function_node.resolveAsFunction(function->build(function_node.getArgumentColumns()));
+        const auto function_node = std::make_shared<FunctionNode>(function_name);
+        auto & new_arguments = function_node->getArguments().getNodes();
+        new_arguments.reserve(sizeof...(args));
+        (new_arguments.push_back(std::forward<Args>(args)), ...);
+        function_node->resolveAsFunction(function->build(function_node->getArgumentColumns()));
+
+        return function_node;
     }
 };
 

src/Analyzer/Passes/QueryAnalysisPass.cpp

@@ -3,6 +3,7 @@
 #include <Common/checkStackSize.h>
 #include <Common/NamePrompter.h>
 #include <Common/ProfileEvents.h>
+#include <Analyzer/FunctionSecretArgumentsFinderTreeNode.h>
 
 #include <IO/WriteBuffer.h>
 #include <IO/WriteHelpers.h>
@@ -706,7 +707,10 @@ struct IdentifierResolveScope
         {
             subquery_depth = parent_scope->subquery_depth;
             context = parent_scope->context;
+            projection_mask_map = parent_scope->projection_mask_map;
         }
+        else
+            projection_mask_map = std::make_shared<std::map<IQueryTreeNode::Hash, size_t>>();
 
         if (auto * union_node = scope_node->as<UnionNode>())
         {
@@ -718,6 +722,11 @@ struct IdentifierResolveScope
             group_by_use_nulls = context->getSettingsRef().group_by_use_nulls &&
                 (query_node->isGroupByWithGroupingSets() || query_node->isGroupByWithRollup() || query_node->isGroupByWithCube());
         }
+
+        if (context)
+            join_use_nulls = context->getSettingsRef().join_use_nulls;
+        else if (parent_scope)
+            join_use_nulls = parent_scope->join_use_nulls;
     }
 
     QueryTreeNodePtr scope_node;
@@ -772,6 +781,8 @@ struct IdentifierResolveScope
 
     /// Apply nullability to aggregation keys
     bool group_by_use_nulls = false;
+    /// Join retutns NULLs instead of default values
+    bool join_use_nulls = false;
 
     /// JOINs count
     size_t joins_count = 0;
@@ -784,6 +795,9 @@ struct IdentifierResolveScope
       */
     QueryTreeNodePtr expression_join_tree_node;
 
+    /// Node hash to mask id map
+    std::shared_ptr<std::map<IQueryTreeNode::Hash, size_t>> projection_mask_map;
+
     [[maybe_unused]] const IdentifierResolveScope * getNearestQueryScope() const
     {
         const IdentifierResolveScope * scope_to_check = this;
@@ -1068,6 +1082,8 @@ private:
 class QueryAnalyzer
 {
 public:
+    explicit QueryAnalyzer(bool only_analyze_) : only_analyze(only_analyze_) {}
+
     void resolve(QueryTreeNodePtr & node, const QueryTreeNodePtr & table_expression, ContextPtr context)
     {
         IdentifierResolveScope scope(node, nullptr /*parent_scope*/);
@@ -1430,6 +1446,7 @@ private:
     /// Global scalar subquery to scalar value map
     std::unordered_map<QueryTreeNodePtrWithHash, Block> scalar_subquery_to_scalar_value;
 
+    const bool only_analyze;
 };
 
 /// Utility functions implementation
@@ -1977,80 +1994,96 @@ void QueryAnalyzer::evaluateScalarSubqueryIfNeeded(QueryTreeNodePtr & node, Iden
         auto interpreter = std::make_unique<InterpreterSelectQueryAnalyzer>(node->toAST(), subquery_context, subquery_context->getViewSource(), options);
 
         auto io = interpreter->execute();
-
         PullingAsyncPipelineExecutor executor(io.pipeline);
         io.pipeline.setProgressCallback(context->getProgressCallback());
         io.pipeline.setProcessListElement(context->getProcessListElement());
 
-        Block block;
-
-        while (block.rows() == 0 && executor.pull(block))
+        if (only_analyze)
         {
-        }
-
-        if (block.rows() == 0)
-        {
-            auto types = interpreter->getSampleBlock().getDataTypes();
-            if (types.size() != 1)
-                types = {std::make_shared<DataTypeTuple>(types)};
-
-            auto & type = types[0];
-            if (!type->isNullable())
+            /// If query is only analyzed, then constants are not correct.
+            scalar_block = interpreter->getSampleBlock();
+            for (auto & column : scalar_block)
             {
-                if (!type->canBeInsideNullable())
-                    throw Exception(ErrorCodes::INCORRECT_RESULT_OF_SCALAR_SUBQUERY,
-                        "Scalar subquery returned empty result of type {} which cannot be Nullable",
-                        type->getName());
-
-                type = makeNullable(type);
+                if (column.column->empty())
+                {
+                    auto mut_col = column.column->cloneEmpty();
+                    mut_col->insertDefault();
+                    column.column = std::move(mut_col);
+                }
             }
-
-            auto scalar_column = type->createColumn();
-            scalar_column->insert(Null());
-            scalar_block.insert({std::move(scalar_column), type, "null"});
         }
         else
         {
-            if (block.rows() != 1)
-                throw Exception(ErrorCodes::INCORRECT_RESULT_OF_SCALAR_SUBQUERY, "Scalar subquery returned more than one row");
+            Block block;
 
-            Block tmp_block;
-            while (tmp_block.rows() == 0 && executor.pull(tmp_block))
+            while (block.rows() == 0 && executor.pull(block))
             {
             }
 
-            if (tmp_block.rows() != 0)
-                throw Exception(ErrorCodes::INCORRECT_RESULT_OF_SCALAR_SUBQUERY, "Scalar subquery returned more than one row");
-
-            block = materializeBlock(block);
-            size_t columns = block.columns();
-
-            if (columns == 1)
+            if (block.rows() == 0)
             {
-                auto & column = block.getByPosition(0);
-                /// Here we wrap type to nullable if we can.
-                /// It is needed cause if subquery return no rows, it's result will be Null.
-                /// In case of many columns, do not check it cause tuple can't be nullable.
-                if (!column.type->isNullable() && column.type->canBeInsideNullable())
+                auto types = interpreter->getSampleBlock().getDataTypes();
+                if (types.size() != 1)
+                    types = {std::make_shared<DataTypeTuple>(types)};
+
+                auto & type = types[0];
+                if (!type->isNullable())
                 {
-                    column.type = makeNullable(column.type);
-                    column.column = makeNullable(column.column);
+                    if (!type->canBeInsideNullable())
+                        throw Exception(ErrorCodes::INCORRECT_RESULT_OF_SCALAR_SUBQUERY,
+                            "Scalar subquery returned empty result of type {} which cannot be Nullable",
+                            type->getName());
+
+                    type = makeNullable(type);
                 }
 
-                scalar_block = block;
+                auto scalar_column = type->createColumn();
+                scalar_column->insert(Null());
+                scalar_block.insert({std::move(scalar_column), type, "null"});
             }
             else
             {
-                /** Make unique column names for tuple.
-                  *
-                  * Example: SELECT (SELECT 2 AS x, x)
-                  */
-                makeUniqueColumnNamesInBlock(block);
+                if (block.rows() != 1)
+                    throw Exception(ErrorCodes::INCORRECT_RESULT_OF_SCALAR_SUBQUERY, "Scalar subquery returned more than one row");
 
-                scalar_block.insert({
-                    ColumnTuple::create(block.getColumns()),
-                    std::make_shared<DataTypeTuple>(block.getDataTypes(), block.getNames()),
-                    "tuple"});
+                Block tmp_block;
+                while (tmp_block.rows() == 0 && executor.pull(tmp_block))
+                {
+                }
+
+                if (tmp_block.rows() != 0)
+                    throw Exception(ErrorCodes::INCORRECT_RESULT_OF_SCALAR_SUBQUERY, "Scalar subquery returned more than one row");
+
+                block = materializeBlock(block);
+                size_t columns = block.columns();
+
+                if (columns == 1)
+                {
+                    auto & column = block.getByPosition(0);
+                    /// Here we wrap type to nullable if we can.
+                    /// It is needed cause if subquery return no rows, it's result will be Null.
+                    /// In case of many columns, do not check it cause tuple can't be nullable.
+                    if (!column.type->isNullable() && column.type->canBeInsideNullable())
+                    {
+                        column.type = makeNullable(column.type);
+                        column.column = makeNullable(column.column);
+                    }
+
+                    scalar_block = block;
+                }
+                else
+                {
+                    /** Make unique column names for tuple.
+                    *
+                    * Example: SELECT (SELECT 2 AS x, x)
+                    */
+                    makeUniqueColumnNamesInBlock(block);
+
+                    scalar_block.insert({
+                        ColumnTuple::create(block.getColumns()),
+                        std::make_shared<DataTypeTuple>(block.getDataTypes(), block.getNames()),
+                        "tuple"});
+                }
             }
         }
 
@@ -3286,7 +3319,6 @@ QueryTreeNodePtr QueryAnalyzer::tryResolveIdentifierFromJoin(const IdentifierLoo
     QueryTreeNodePtr resolved_identifier;
 
     JoinKind join_kind = from_join_node.getKind();
-    bool join_use_nulls = scope.context->getSettingsRef().join_use_nulls;
 
     /// If columns from left or right table were missed Object(Nullable('json')) subcolumns, they will be replaced
     /// to ConstantNode(NULL), which can't be cast to ColumnNode, so we resolve it here.
@@ -3451,7 +3483,7 @@ QueryTreeNodePtr QueryAnalyzer::tryResolveIdentifierFromJoin(const IdentifierLoo
     if (join_node_in_resolve_process || !resolved_identifier)
         return resolved_identifier;
 
-    if (join_use_nulls)
+    if (scope.join_use_nulls)
     {
         resolved_identifier = resolved_identifier->clone();
         convertJoinedColumnTypeToNullIfNeeded(resolved_identifier, join_kind, resolved_side);
@@ -4439,7 +4471,7 @@ ProjectionNames QueryAnalyzer::resolveMatcher(QueryTreeNodePtr & matcher_node, I
     else
         matched_expression_nodes_with_names = resolveUnqualifiedMatcher(matcher_node, scope);
 
-    if (scope.context->getSettingsRef().join_use_nulls)
+    if (scope.join_use_nulls)
     {
         /** If we are resolving matcher came from the result of JOIN and `join_use_nulls` is set,
           * we need to convert joined column type to Nullable.
@@ -5124,22 +5156,31 @@ ProjectionNames QueryAnalyzer::resolveFunction(QueryTreeNodePtr & node, Identifi
     }
 
     /// Resolve function arguments
-
     bool allow_table_expressions = is_special_function_in;
     auto arguments_projection_names = resolveExpressionNodeList(function_node_ptr->getArgumentsNode(),
         scope,
         true /*allow_lambda_expression*/,
         allow_table_expressions /*allow_table_expression*/);
 
-    if (function_node_ptr->toAST()->hasSecretParts())
+    /// Mask arguments if needed
+    if (!scope.context->getSettingsRef().format_display_secrets_in_show_and_select)
     {
-        for (auto & argument : arguments_projection_names)
+        if (FunctionSecretArgumentsFinder::Result secret_arguments = FunctionSecretArgumentsFinderTreeNode(*function_node_ptr).getResult(); secret_arguments.count)
         {
-            SipHash hash;
-            hash.update(argument);
-            argument = getHexUIntLowercase(hash.get128());
+            auto & argument_nodes = function_node_ptr->getArgumentsNode()->as<ListNode &>().getNodes();
+
+            for (size_t n = secret_arguments.start; n < secret_arguments.start + secret_arguments.count; ++n)
+            {
+                if (auto * constant = argument_nodes[n]->as<ConstantNode>())
+                {
+                    auto mask = scope.projection_mask_map->insert({constant->getTreeHash(), scope.projection_mask_map->size() + 1}).first->second;
+                    constant->setMaskId(mask);
+                    arguments_projection_names[n] = "[HIDDEN id: " + std::to_string(mask) + "]";
+                }
+            }
         }
     }
+
     auto & function_node = *function_node_ptr;
 
     /// Replace right IN function argument if it is table or table function with subquery that read ordinary columns
@@ -6651,7 +6692,6 @@ void QueryAnalyzer::initializeTableExpressionData(const QueryTreeNodePtr & table
             if (column_default && column_default->kind == ColumnDefaultKind::Alias)
             {
                 auto alias_expression = buildQueryTree(column_default->expression, scope.context);
-                alias_expression = buildCastFunction(alias_expression, column_name_and_type.type, scope.context, false /*resolve*/);
                 auto column_node = std::make_shared<ColumnNode>(column_name_and_type, std::move(alias_expression), table_expression_node);
                 column_name_to_column_node.emplace(column_name_and_type.name, column_node);
                 alias_columns_to_resolve.emplace_back(column_name_and_type.name, column_node);
@@ -6684,7 +6724,9 @@ void QueryAnalyzer::initializeTableExpressionData(const QueryTreeNodePtr & table
                 alias_column_resolve_scope,
                 false /*allow_lambda_expression*/,
                 false /*allow_table_expression*/);
-
+            auto & resolved_expression = alias_column_to_resolve->getExpression();
+            if (!resolved_expression->getResultType()->equals(*alias_column_to_resolve->getResultType()))
+                resolved_expression = buildCastFunction(resolved_expression, alias_column_to_resolve->getResultType(), scope.context, true);
             column_name_to_column_node = std::move(alias_column_resolve_scope.column_name_to_column_node);
             column_name_to_column_node[alias_column_to_resolve_name] = alias_column_to_resolve;
         }
@@ -7558,8 +7600,22 @@ void QueryAnalyzer::resolveQuery(const QueryTreeNodePtr & query_node, Identifier
     }
 
     if (query_node_typed.getPrewhere())
+    {
+        /** Expression in PREWHERE with JOIN should not be modified by join_use_nulls.
+          * Example: SELECT * FROM t1 JOIN t2 USING (id) PREWHERE a = 1
+          * Column `a` should be resolved from table and should not change its type to Nullable.
+          */
+        bool join_use_nulls = scope.join_use_nulls;
+        bool use_identifier_lookup_to_result_cache = scope.use_identifier_lookup_to_result_cache;
+        scope.join_use_nulls = false;
+        scope.use_identifier_lookup_to_result_cache = false;
+
         resolveExpressionNode(query_node_typed.getPrewhere(), scope, false /*allow_lambda_expression*/, false /*allow_table_expression*/);
 
+        scope.join_use_nulls = join_use_nulls;
+        scope.use_identifier_lookup_to_result_cache = use_identifier_lookup_to_result_cache;
+    }
+
     if (query_node_typed.getWhere())
         resolveExpressionNode(query_node_typed.getWhere(), scope, false /*allow_lambda_expression*/, false /*allow_table_expression*/);
 
@@ -7749,13 +7805,16 @@ void QueryAnalyzer::resolveUnion(const QueryTreeNodePtr & union_node, Identifier
 
 }
 
-QueryAnalysisPass::QueryAnalysisPass(QueryTreeNodePtr table_expression_)
+QueryAnalysisPass::QueryAnalysisPass(QueryTreeNodePtr table_expression_, bool only_analyze_)
     : table_expression(std::move(table_expression_))
+    , only_analyze(only_analyze_)
 {}
 
+QueryAnalysisPass::QueryAnalysisPass(bool only_analyze_) : only_analyze(only_analyze_) {}
+
 void QueryAnalysisPass::run(QueryTreeNodePtr & query_tree_node, ContextPtr context)
 {
-    QueryAnalyzer analyzer;
+    QueryAnalyzer analyzer(only_analyze);
     analyzer.resolve(query_tree_node, table_expression, context);
     createUniqueTableAliases(query_tree_node, table_expression, context);
 }

src/Analyzer/Passes/QueryAnalysisPass.h

@@ -71,13 +71,13 @@ public:
     /** Construct query analysis pass for query or union analysis.
       * Available columns are extracted from query node join tree.
       */
-    QueryAnalysisPass() = default;
+    explicit QueryAnalysisPass(bool only_analyze_ = false);
 
     /** Construct query analysis pass for expression or list of expressions analysis.
       * Available expression columns are extracted from table expression.
       * Table expression node must have query, union, table, table function type.
       */
-    explicit QueryAnalysisPass(QueryTreeNodePtr table_expression_);
+    QueryAnalysisPass(QueryTreeNodePtr table_expression_, bool only_analyze_ = false);
 
     String getName() override
     {
@@ -93,6 +93,7 @@ public:
 
 private:
     QueryTreeNodePtr table_expression;
+    const bool only_analyze;
 };
 
 }

src/Analyzer/QueryTreePassManager.cpp

@@ -246,9 +246,9 @@ void QueryTreePassManager::dump(WriteBuffer & buffer, size_t up_to_pass_index)
     }
 }
 
-void addQueryTreePasses(QueryTreePassManager & manager)
+void addQueryTreePasses(QueryTreePassManager & manager, bool only_analyze)
 {
-    manager.addPass(std::make_unique<QueryAnalysisPass>());
+    manager.addPass(std::make_unique<QueryAnalysisPass>(only_analyze));
     manager.addPass(std::make_unique<GroupingFunctionsResolvePass>());
 
     manager.addPass(std::make_unique<RemoveUnusedProjectionColumnsPass>());

src/Analyzer/QueryTreePassManager.h

@@ -47,6 +47,6 @@ private:
     std::vector<QueryTreePassPtr> passes;
 };
 
-void addQueryTreePasses(QueryTreePassManager & manager);
+void addQueryTreePasses(QueryTreePassManager & manager, bool only_analyze = false);
 
 }

src/Backups/BackupsWorker.cpp

@@ -18,6 +18,7 @@
 #include <Interpreters/executeDDLQueryOnCluster.h>
 #include <Parsers/ASTBackupQuery.h>
 #include <Parsers/ASTFunction.h>
+#include <Common/CurrentThread.h>
 #include <Common/Exception.h>
 #include <Common/Macros.h>
 #include <Common/logger_useful.h>
@@ -486,7 +487,7 @@ OperationID BackupsWorker::startMakingBackup(const ASTPtr & query, const Context
             /// process_list_element_holder is used to make an element in ProcessList live while BACKUP is working asynchronously.
             auto process_list_element = context_in_use->getProcessListElement();
 
-            scheduleFromThreadPool<void>(
+            thread_pool.scheduleOrThrowOnError(
                 [this,
                  backup_query,
                  backup_id,
@@ -502,6 +503,8 @@ OperationID BackupsWorker::startMakingBackup(const ASTPtr & query, const Context
                     BackupMutablePtr backup_async;
                     try
                     {
+                        setThreadName("BackupWorker");
+                        CurrentThread::QueryScope query_scope(context_in_use);
                         doBackup(
                             backup_async,
                             backup_query,
@@ -517,8 +520,7 @@ OperationID BackupsWorker::startMakingBackup(const ASTPtr & query, const Context
                     {
                         on_exception(backup_async, backup_id, backup_name_for_logging, backup_settings, backup_coordination);
                     }
-                },
-                thread_pool, "BackupWorker");
+                });
         }
         else
         {
@@ -864,7 +866,7 @@ OperationID BackupsWorker::startRestoring(const ASTPtr & query, ContextMutablePt
             /// process_list_element_holder is used to make an element in ProcessList live while RESTORE is working asynchronously.
             auto process_list_element = context_in_use->getProcessListElement();
 
-            scheduleFromThreadPool<void>(
+            thread_pool.scheduleOrThrowOnError(
                 [this,
                  restore_query,
                  restore_id,
@@ -878,6 +880,8 @@ OperationID BackupsWorker::startRestoring(const ASTPtr & query, ContextMutablePt
                 {
                     try
                     {
+                        setThreadName("RestorerWorker");
+                        CurrentThread::QueryScope query_scope(context_in_use);
                         doRestore(
                             restore_query,
                             restore_id,
@@ -891,9 +895,7 @@ OperationID BackupsWorker::startRestoring(const ASTPtr & query, ContextMutablePt
                     {
                         on_exception(restore_id, backup_name_for_logging, restore_settings, restore_coordination);
                     }
-                },
-                thread_pool,
-                "RestoreWorker");
+                });
         }
         else
         {

src/Backups/registerBackupEngineS3.cpp

@@ -15,8 +15,6 @@
 
 namespace DB
 {
-namespace fs = std::filesystem;
-
 namespace ErrorCodes
 {
     extern const int BAD_ARGUMENTS;
@@ -65,13 +63,13 @@ void registerBackupEngineS3(BackupFactory & factory)
             secret_access_key = config.getString(config_prefix + ".secret_access_key", "");
 
             if (config.has(config_prefix + ".filename"))
-                s3_uri = fs::path(s3_uri) / config.getString(config_prefix + ".filename");
+                s3_uri = std::filesystem::path(s3_uri) / config.getString(config_prefix + ".filename");
 
             if (args.size() > 1)
                 throw Exception(ErrorCodes::NUMBER_OF_ARGUMENTS_DOESNT_MATCH, "Backup S3 requires 1 or 2 arguments: named_collection, [filename]");
 
             if (args.size() == 1)
-                s3_uri = fs::path(s3_uri) / args[0].safeGet<String>();
+                s3_uri = std::filesystem::path(s3_uri) / args[0].safeGet<String>();
         }
         else
         {

src/BridgeHelper/CatBoostLibraryBridgeHelper.cpp

@@ -58,8 +58,12 @@ bool CatBoostLibraryBridgeHelper::bridgeHandShake()
     String result;
     try
     {
-        ReadWriteBufferFromHTTP buf(getPingURI(), Poco::Net::HTTPRequest::HTTP_GET, {}, http_timeouts, credentials);
-        readString(result, buf);
+        auto buf = BuilderRWBufferFromHTTP(getPingURI())
+                       .withConnectionGroup(HTTPConnectionGroupType::STORAGE)
+                       .withTimeouts(http_timeouts)
+                       .create(credentials);
+
+        readString(result, *buf);
     }
     catch (...)
     {
@@ -79,29 +83,29 @@ ExternalModelInfos CatBoostLibraryBridgeHelper::listModels()
 {
     startBridgeSync();
 
-    ReadWriteBufferFromHTTP buf(
-        createRequestURI(CATBOOST_LIST_METHOD),
-        Poco::Net::HTTPRequest::HTTP_POST,
-        [](std::ostream &) {},
-        http_timeouts, credentials);
+    auto buf = BuilderRWBufferFromHTTP(createRequestURI(CATBOOST_LIST_METHOD))
+                   .withConnectionGroup(HTTPConnectionGroupType::STORAGE)
+                   .withMethod(Poco::Net::HTTPRequest::HTTP_POST)
+                   .withTimeouts(http_timeouts)
+                   .create(credentials);
 
     ExternalModelInfos result;
 
     UInt64 num_rows;
-    readIntBinary(num_rows, buf);
+    readIntBinary(num_rows, *buf);
 
     for (UInt64 i = 0; i < num_rows; ++i)
     {
         ExternalModelInfo info;
 
-        readStringBinary(info.model_path, buf);
-        readStringBinary(info.model_type, buf);
+        readStringBinary(info.model_path, *buf);
+        readStringBinary(info.model_type, *buf);
 
         UInt64 t;
-        readIntBinary(t, buf);
+        readIntBinary(t, *buf);
         info.loading_start_time = std::chrono::system_clock::from_time_t(t);
 
-        readIntBinary(t, buf);
+        readIntBinary(t, *buf);
         info.loading_duration = std::chrono::milliseconds(t);
 
         result.push_back(info);
@@ -116,17 +120,19 @@ void CatBoostLibraryBridgeHelper::removeModel()
 
     assert(model_path);
 
-    ReadWriteBufferFromHTTP buf(
-        createRequestURI(CATBOOST_REMOVEMODEL_METHOD),
-        Poco::Net::HTTPRequest::HTTP_POST,
-        [this](std::ostream & os)
-        {
-            os << "model_path=" << escapeForFileName(*model_path);
-        },
-        http_timeouts, credentials);
+    auto buf = BuilderRWBufferFromHTTP(createRequestURI(CATBOOST_REMOVEMODEL_METHOD))
+                   .withConnectionGroup(HTTPConnectionGroupType::STORAGE)
+                   .withMethod(Poco::Net::HTTPRequest::HTTP_POST)
+                   .withTimeouts(http_timeouts)
+                   .withOutCallback(
+                       [this](std::ostream & os)
+                       {
+                           os << "model_path=" << escapeForFileName(*model_path);
+                       })
+                   .create(credentials);
 
     String result;
-    readStringBinary(result, buf);
+    readStringBinary(result, *buf);
     assert(result == "1");
 }
 
@@ -134,14 +140,14 @@ void CatBoostLibraryBridgeHelper::removeAllModels()
 {
     startBridgeSync();
 
-    ReadWriteBufferFromHTTP buf(
-        createRequestURI(CATBOOST_REMOVEALLMODELS_METHOD),
-        Poco::Net::HTTPRequest::HTTP_POST,
-        [](std::ostream &){},
-        http_timeouts, credentials);
+    auto buf = BuilderRWBufferFromHTTP(createRequestURI(CATBOOST_REMOVEALLMODELS_METHOD))
+                   .withConnectionGroup(HTTPConnectionGroupType::STORAGE)
+                   .withMethod(Poco::Net::HTTPRequest::HTTP_POST)
+                   .withTimeouts(http_timeouts)
+                   .create(credentials);
 
     String result;
-    readStringBinary(result, buf);
+    readStringBinary(result, *buf);
     assert(result == "1");
 }
 
@@ -151,18 +157,20 @@ size_t CatBoostLibraryBridgeHelper::getTreeCount()
 
     assert(model_path && library_path);
 
-    ReadWriteBufferFromHTTP buf(
-        createRequestURI(CATBOOST_GETTREECOUNT_METHOD),
-        Poco::Net::HTTPRequest::HTTP_POST,
-        [this](std::ostream & os)
-        {
-            os << "library_path=" << escapeForFileName(*library_path) << "&";
-            os << "model_path=" << escapeForFileName(*model_path);
-        },
-        http_timeouts, credentials);
+    auto buf = BuilderRWBufferFromHTTP(createRequestURI(CATBOOST_GETTREECOUNT_METHOD))
+                   .withConnectionGroup(HTTPConnectionGroupType::STORAGE)
+                   .withMethod(Poco::Net::HTTPRequest::HTTP_POST)
+                   .withTimeouts(http_timeouts)
+                   .withOutCallback(
+                        [this](std::ostream & os)
+                        {
+                           os << "library_path=" << escapeForFileName(*library_path) << "&";
+                           os << "model_path=" << escapeForFileName(*model_path);
+                        })
+                   .create(credentials);
 
     size_t result;
-    readIntBinary(result, buf);
+    readIntBinary(result, *buf);
     return result;
 }
 
@@ -177,17 +185,19 @@ ColumnPtr CatBoostLibraryBridgeHelper::evaluate(const ColumnsWithTypeAndName & c
 
     assert(model_path);
 
-    ReadWriteBufferFromHTTP buf(
-        createRequestURI(CATBOOST_LIB_EVALUATE_METHOD),
-        Poco::Net::HTTPRequest::HTTP_POST,
-        [this, serialized = string_write_buf.str()](std::ostream & os)
-        {
-            os << "model_path=" << escapeForFileName(*model_path) << "&";
-            os << "data=" << escapeForFileName(serialized);
-        },
-        http_timeouts, credentials);
+    auto buf = BuilderRWBufferFromHTTP(createRequestURI(CATBOOST_LIB_EVALUATE_METHOD))
+                   .withConnectionGroup(HTTPConnectionGroupType::STORAGE)
+                   .withMethod(Poco::Net::HTTPRequest::HTTP_POST)
+                   .withTimeouts(http_timeouts)
+                   .withOutCallback(
+                       [this, serialized = string_write_buf.str()](std::ostream & os)
+                       {
+                           os << "model_path=" << escapeForFileName(*model_path) << "&";
+                           os << "data=" << escapeForFileName(serialized);
+                       })
+                   .create(credentials);
 
-    NativeReader deserializer(buf, /*server_revision*/ 0);
+    NativeReader deserializer(*buf, /*server_revision*/ 0);
     Block block_read = deserializer.read();
 
     return block_read.getColumns()[0];

src/BridgeHelper/ExternalDictionaryLibraryBridgeHelper.cpp

@@ -71,8 +71,12 @@ bool ExternalDictionaryLibraryBridgeHelper::bridgeHandShake()
     String result;
     try
     {
-        ReadWriteBufferFromHTTP buf(getPingURI(), Poco::Net::HTTPRequest::HTTP_GET, {}, http_timeouts, credentials);
-        readString(result, buf);
+        auto buf = BuilderRWBufferFromHTTP(getPingURI())
+                       .withConnectionGroup(HTTPConnectionGroupType::STORAGE)
+                       .withTimeouts(http_timeouts)
+                       .create(credentials);
+
+        readString(result, *buf);
     }
     catch (...)
     {
@@ -247,30 +251,28 @@ QueryPipeline ExternalDictionaryLibraryBridgeHelper::loadKeys(const Block & requ
 
 bool ExternalDictionaryLibraryBridgeHelper::executeRequest(const Poco::URI & uri, ReadWriteBufferFromHTTP::OutStreamCallback out_stream_callback) const
 {
-    ReadWriteBufferFromHTTP buf(
-        uri,
-        Poco::Net::HTTPRequest::HTTP_POST,
-        std::move(out_stream_callback),
-        http_timeouts, credentials);
+    auto buf = BuilderRWBufferFromHTTP(uri)
+                   .withConnectionGroup(HTTPConnectionGroupType::STORAGE)
+                   .withMethod(Poco::Net::HTTPRequest::HTTP_POST)
+                   .withTimeouts(http_timeouts)
+                   .withOutCallback(std::move(out_stream_callback))
+                   .create(credentials);
 
     bool res;
-    readBoolText(res, buf);
+    readBoolText(res, *buf);
     return res;
 }
 
 
 QueryPipeline ExternalDictionaryLibraryBridgeHelper::loadBase(const Poco::URI & uri, ReadWriteBufferFromHTTP::OutStreamCallback out_stream_callback)
 {
-    auto read_buf_ptr = std::make_unique<ReadWriteBufferFromHTTP>(
-        uri,
-        Poco::Net::HTTPRequest::HTTP_POST,
-        std::move(out_stream_callback),
-        http_timeouts,
-        credentials,
-        0,
-        DBMS_DEFAULT_BUFFER_SIZE,
-        getContext()->getReadSettings(),
-        HTTPHeaderEntries{});
+    auto read_buf_ptr = BuilderRWBufferFromHTTP(uri)
+                            .withConnectionGroup(HTTPConnectionGroupType::STORAGE)
+                            .withMethod(Poco::Net::HTTPRequest::HTTP_POST)
+                            .withSettings(getContext()->getReadSettings())
+                            .withTimeouts(http_timeouts)
+                            .withOutCallback(std::move(out_stream_callback))
+                            .create(credentials);
 
     auto source = FormatFactory::instance().getInput(ExternalDictionaryLibraryBridgeHelper::DEFAULT_FORMAT, *read_buf_ptr, sample_block, getContext(), DEFAULT_BLOCK_SIZE);
     source->addBuffer(std::move(read_buf_ptr));

src/BridgeHelper/XDBCBridgeHelper.h

@@ -97,8 +97,12 @@ protected:
     {
         try
         {
-            ReadWriteBufferFromHTTP buf(getPingURI(), Poco::Net::HTTPRequest::HTTP_GET, {}, getHTTPTimeouts(), credentials);
-            return checkString(PING_OK_ANSWER, buf);
+            auto buf = BuilderRWBufferFromHTTP(getPingURI())
+                           .withConnectionGroup(HTTPConnectionGroupType::STORAGE)
+                           .withTimeouts(getHTTPTimeouts())
+                           .create(credentials);
+
+            return checkString(PING_OK_ANSWER, *buf);
         }
         catch (...)
         {
@@ -198,10 +202,14 @@ protected:
             uri.addQueryParameter("connection_string", getConnectionString());
             uri.addQueryParameter("use_connection_pooling", toString(use_connection_pooling));
 
-            ReadWriteBufferFromHTTP buf(uri, Poco::Net::HTTPRequest::HTTP_POST, {}, getHTTPTimeouts(), credentials);
+            auto buf = BuilderRWBufferFromHTTP(uri)
+                           .withConnectionGroup(HTTPConnectionGroupType::STORAGE)
+                           .withMethod(Poco::Net::HTTPRequest::HTTP_POST)
+                           .withTimeouts(getHTTPTimeouts())
+                           .create(credentials);
 
-            bool res;
-            readBoolText(res, buf);
+            bool res = false;
+            readBoolText(res, *buf);
             is_schema_allowed = res;
         }
 
@@ -220,10 +228,14 @@ protected:
             uri.addQueryParameter("connection_string", getConnectionString());
             uri.addQueryParameter("use_connection_pooling", toString(use_connection_pooling));
 
-            ReadWriteBufferFromHTTP buf(uri, Poco::Net::HTTPRequest::HTTP_POST, {}, getHTTPTimeouts(), credentials);
+            auto buf = BuilderRWBufferFromHTTP(uri)
+                           .withConnectionGroup(HTTPConnectionGroupType::STORAGE)
+                           .withMethod(Poco::Net::HTTPRequest::HTTP_POST)
+                           .withTimeouts(getHTTPTimeouts())
+                           .create(credentials);
 
             std::string character;
-            readStringBinary(character, buf);
+            readStringBinary(character, *buf);
             if (character.length() > 1)
                 throw Exception(ErrorCodes::ILLEGAL_TYPE_OF_ARGUMENT, "Failed to parse quoting style from '{}' for service {}",
                     character, BridgeHelperMixin::serviceAlias());

src/CMakeLists.txt

@@ -174,6 +174,8 @@ endif ()
 
 add_library(clickhouse_common_io ${clickhouse_common_io_headers} ${clickhouse_common_io_sources})
 
+set_source_files_properties(Common/ThreadFuzzer.cpp PROPERTIES COMPILE_FLAGS "-fomit-frame-pointer -momit-leaf-frame-pointer")
+
 add_library (clickhouse_malloc OBJECT Common/malloc.cpp)
 set_source_files_properties(Common/malloc.cpp PROPERTIES COMPILE_FLAGS "-fno-builtin")
 

src/Client/Connection.cpp

@@ -153,6 +153,12 @@ void Connection::connect(const ConnectionTimeouts & timeouts)
                 current_resolved_address = *it;
                 break;
             }
+            catch (DB::NetException &)
+            {
+                if (++it == addresses.end())
+                    throw;
+                continue;
+            }
             catch (Poco::Net::NetException &)
             {
                 if (++it == addresses.end())
@@ -199,6 +205,17 @@ void Connection::connect(const ConnectionTimeouts & timeouts)
         LOG_TRACE(log_wrapper.get(), "Connected to {} server version {}.{}.{}.",
             server_name, server_version_major, server_version_minor, server_version_patch);
     }
+    catch (DB::NetException & e)
+    {
+        disconnect();
+
+        /// Remove this possible stale entry from cache
+        DNSResolver::instance().removeHostFromCache(host);
+
+        /// Add server address to exception. Exception will preserve stack trace.
+        e.addMessage("({})", getDescription());
+        throw;
+    }
     catch (Poco::Net::NetException & e)
     {
         disconnect();
@@ -206,7 +223,7 @@ void Connection::connect(const ConnectionTimeouts & timeouts)
         /// Remove this possible stale entry from cache
         DNSResolver::instance().removeHostFromCache(host);
 
-        /// Add server address to exception. Also Exception will remember stack trace. It's a pity that more precise exception type is lost.
+        /// Add server address to exception. Also Exception will remember new stack trace. It's a pity that more precise exception type is lost.
         throw NetException(ErrorCodes::NETWORK_ERROR, "{} ({})", e.displayText(), getDescription());
     }
     catch (Poco::TimeoutException & e)
@@ -216,7 +233,7 @@ void Connection::connect(const ConnectionTimeouts & timeouts)
         /// Remove this possible stale entry from cache
         DNSResolver::instance().removeHostFromCache(host);
 
-        /// Add server address to exception. Also Exception will remember stack trace. It's a pity that more precise exception type is lost.
+        /// Add server address to exception. Also Exception will remember new stack trace. It's a pity that more precise exception type is lost.
         /// This exception can only be thrown from socket->connect(), so add information about connection timeout.
         const auto & connection_timeout = static_cast<bool>(secure) ? timeouts.secure_connection_timeout : timeouts.connection_timeout;
         throw NetException(

src/Client/ConnectionParameters.cpp

@@ -115,7 +115,7 @@ ConnectionParameters::ConnectionParameters(const Poco::Util::AbstractConfigurati
     /// At the same time, I want clickhouse-local to always work, regardless.
     /// TODO: get rid of glibc, or replace getaddrinfo to c-ares.
 
-    compression = config.getBool("compression", host != "localhost" && !isLocalAddress(DNSResolver::instance().resolveHost(host)))
+    compression = config.getBool("compression", host != "localhost" && !isLocalAddress(DNSResolver::instance().resolveHostAllInOriginOrder(host).front()))
                   ? Protocol::Compression::Enable : Protocol::Compression::Disable;
 
     timeouts = ConnectionTimeouts()

src/Client/QueryFuzzer.cpp

@@ -232,7 +232,7 @@ ASTPtr QueryFuzzer::getRandomColumnLike()
         return nullptr;
     }
 
-    ASTPtr new_ast = column_like[fuzz_rand() % column_like.size()]->clone();
+    ASTPtr new_ast = column_like[fuzz_rand() % column_like.size()].second->clone();
     new_ast->setAlias("");
 
     return new_ast;
@@ -272,7 +272,7 @@ void QueryFuzzer::replaceWithTableLike(ASTPtr & ast)
         return;
     }
 
-    ASTPtr new_ast = table_like[fuzz_rand() % table_like.size()]->clone();
+    ASTPtr new_ast = table_like[fuzz_rand() % table_like.size()].second->clone();
 
     std::string old_alias = ast->tryGetAlias();
     new_ast->setAlias(old_alias);
@@ -1214,57 +1214,46 @@ void QueryFuzzer::fuzz(ASTPtr & ast)
     }
 }
 
+#define AST_FUZZER_PART_TYPE_CAP 1000
+
 /*
  * This functions collects various parts of query that we can then substitute
  * to a query being fuzzed.
- *
- * TODO: we just stop remembering new parts after our corpus reaches certain size.
- * This is boring, should implement a random replacement of existing parst with
- * small probability. Do this after we add this fuzzer to CI and fix all the
- * problems it can routinely find even in this boring version.
  */
 void QueryFuzzer::collectFuzzInfoMain(ASTPtr ast)
 {
     collectFuzzInfoRecurse(ast);
-
-    aliases.clear();
-    for (const auto & alias : aliases_set)
-    {
-        aliases.push_back(alias);
-    }
-
-    column_like.clear();
-    for (const auto & [name, value] : column_like_map)
-    {
-        column_like.push_back(value);
-    }
-
-    table_like.clear();
-    for (const auto & [name, value] : table_like_map)
-    {
-        table_like.push_back(value);
-    }
 }
 
 void QueryFuzzer::addTableLike(ASTPtr ast)
 {
-    if (table_like_map.size() > 1000)
+    if (table_like_map.size() > AST_FUZZER_PART_TYPE_CAP)
     {
-        table_like_map.clear();
+        const auto iter = std::next(table_like.begin(), fuzz_rand() % table_like.size());
+        const auto ast_del = *iter;
+        table_like.erase(iter);
+        table_like_map.erase(ast_del.first);
     }
 
     const auto name = ast->formatForErrorMessage();
     if (name.size() < 200)
     {
-        table_like_map.insert({name, ast});
+        const auto res = table_like_map.insert({name, ast});
+        if (res.second)
+        {
+            table_like.push_back({name, ast});
+        }
     }
 }
 
 void QueryFuzzer::addColumnLike(ASTPtr ast)
 {
-    if (column_like_map.size() > 1000)
+    if (column_like_map.size() > AST_FUZZER_PART_TYPE_CAP)
     {
-        column_like_map.clear();
+        const auto iter = std::next(column_like.begin(), fuzz_rand() % column_like.size());
+        const auto ast_del = *iter;
+        column_like.erase(iter);
+        column_like_map.erase(ast_del.first);
     }
 
     const auto name = ast->formatForErrorMessage();
@@ -1279,22 +1268,16 @@ void QueryFuzzer::addColumnLike(ASTPtr ast)
     }
     if (name.size() < 200)
     {
-        column_like_map.insert({name, ast});
+        const auto res = column_like_map.insert({name, ast});
+        if (res.second)
+        {
+            column_like.push_back({name, ast});
+        }
     }
 }
 
 void QueryFuzzer::collectFuzzInfoRecurse(ASTPtr ast)
 {
-    if (auto * impl = dynamic_cast<ASTWithAlias *>(ast.get()))
-    {
-        if (aliases_set.size() > 1000)
-        {
-            aliases_set.clear();
-        }
-
-        aliases_set.insert(impl->alias);
-    }
-
     if (typeid_cast<ASTLiteral *>(ast.get()))
     {
         addColumnLike(ast);

src/Client/QueryFuzzer.h

@@ -50,14 +50,12 @@ struct QueryFuzzer
     // we are currently fuzzing. We add some part from each new query we are asked
     // to fuzz, and keep this state between queries, so the fuzzing output becomes
     // more interesting over time, as the queries mix.
-    std::unordered_set<std::string> aliases_set;
-    std::vector<std::string> aliases;
-
+    // The hash tables are used for collection, and the vectors are used for random access.
     std::unordered_map<std::string, ASTPtr> column_like_map;
-    std::vector<ASTPtr> column_like;
+    std::vector<std::pair<std::string, ASTPtr>> column_like;
 
     std::unordered_map<std::string, ASTPtr> table_like_map;
-    std::vector<ASTPtr> table_like;
+    std::vector<std::pair<std::string, ASTPtr>> table_like;
 
     // Some debug fields for detecting problematic ASTs with loops.
     // These are reset for each fuzzMain call.

src/Columns/ColumnUnique.cpp

@@ -0,0 +1,25 @@
+#include <Columns/ColumnUnique.h>
+
+namespace DB
+{
+
+/// Explicit template instantiations.
+template class ColumnUnique<ColumnInt8>;
+template class ColumnUnique<ColumnUInt8>;
+template class ColumnUnique<ColumnInt16>;
+template class ColumnUnique<ColumnUInt16>;
+template class ColumnUnique<ColumnInt32>;
+template class ColumnUnique<ColumnUInt32>;
+template class ColumnUnique<ColumnInt64>;
+template class ColumnUnique<ColumnUInt64>;
+template class ColumnUnique<ColumnInt128>;
+template class ColumnUnique<ColumnUInt128>;
+template class ColumnUnique<ColumnInt256>;
+template class ColumnUnique<ColumnUInt256>;
+template class ColumnUnique<ColumnFloat32>;
+template class ColumnUnique<ColumnFloat64>;
+template class ColumnUnique<ColumnString>;
+template class ColumnUnique<ColumnFixedString>;
+template class ColumnUnique<ColumnDateTime64>;
+
+}

src/Columns/ColumnUnique.h

@@ -15,6 +15,8 @@
 #include <Common/typeid_cast.h>
 #include <Common/assert_cast.h>
 #include <Common/FieldVisitors.h>
+#include "Columns/ColumnsDateTime.h"
+#include "Columns/ColumnsNumber.h"
 
 #include <base/range.h>
 #include <base/unaligned.h>
@@ -736,4 +738,23 @@ UInt128 ColumnUnique<ColumnType>::IncrementalHash::getHash(const ColumnType & co
     return cur_hash;
 }
 
+
+extern template class ColumnUnique<ColumnInt8>;
+extern template class ColumnUnique<ColumnUInt8>;
+extern template class ColumnUnique<ColumnInt16>;
+extern template class ColumnUnique<ColumnUInt16>;
+extern template class ColumnUnique<ColumnInt32>;
+extern template class ColumnUnique<ColumnUInt32>;
+extern template class ColumnUnique<ColumnInt64>;
+extern template class ColumnUnique<ColumnUInt64>;
+extern template class ColumnUnique<ColumnInt128>;
+extern template class ColumnUnique<ColumnUInt128>;
+extern template class ColumnUnique<ColumnInt256>;
+extern template class ColumnUnique<ColumnUInt256>;
+extern template class ColumnUnique<ColumnFloat32>;
+extern template class ColumnUnique<ColumnFloat64>;
+extern template class ColumnUnique<ColumnString>;
+extern template class ColumnUnique<ColumnFixedString>;
+extern template class ColumnUnique<ColumnDateTime64>;
+
 }

src/Common/CurrentMetrics.cpp

@@ -274,7 +274,19 @@
     M(DistrCacheUsedConnections, "Number of currently used connections to Distributed Cache") \
     M(DistrCacheReadRequests, "Number of executed Read requests to Distributed Cache") \
     M(DistrCacheWriteRequests, "Number of executed Write requests to Distributed Cache") \
-    M(DistrCacheServerConnections, "Number of open connections to ClickHouse server from Distributed Cache")
+    M(DistrCacheServerConnections, "Number of open connections to ClickHouse server from Distributed Cache") \
+    \
+    M(StorageConnectionsStored, "Total count of sessions stored in the session pool for storages") \
+    M(StorageConnectionsTotal, "Total count of all sessions: stored in the pool and actively used right now for storages") \
+    \
+    M(DiskConnectionsStored, "Total count of sessions stored in the session pool for disks") \
+    M(DiskConnectionsTotal, "Total count of all sessions: stored in the pool and actively used right now for disks") \
+    \
+    M(HTTPConnectionsStored, "Total count of sessions stored in the session pool for http hosts") \
+    M(HTTPConnectionsTotal, "Total count of all sessions: stored in the pool and actively used right now for http hosts") \
+    \
+    M(AddressesActive, "Total count of addresses which are used for creation connections with connection pools") \
+
 
 #ifdef APPLY_FOR_EXTERNAL_METRICS
     #define APPLY_FOR_METRICS(M) APPLY_FOR_BUILTIN_METRICS(M) APPLY_FOR_EXTERNAL_METRICS(M)

src/Common/DNSResolver.cpp

@@ -1,6 +1,7 @@
 #include "DNSResolver.h"
 #include <Common/CacheBase.h>
 #include <Common/Exception.h>
+#include <Common/NetException.h>
 #include <Common/ProfileEvents.h>
 #include <Common/thread_local_rng.h>
 #include <Common/logger_useful.h>
@@ -108,7 +109,7 @@ DNSResolver::IPAddresses hostByName(const std::string & host)
     if (addresses.empty())
     {
         ProfileEvents::increment(ProfileEvents::DNSError);
-        throw Exception(ErrorCodes::DNS_ERROR, "Not found address of host: {}", host);
+        throw DB::NetException(ErrorCodes::DNS_ERROR, "Not found address of host: {}", host);
     }
 
     return addresses;
@@ -202,10 +203,10 @@ DNSResolver::DNSResolver() : impl(std::make_unique<DNSResolver::Impl>()), log(ge
 
 Poco::Net::IPAddress DNSResolver::resolveHost(const std::string & host)
 {
-    return pickAddress(resolveHostAll(host));
+    return pickAddress(resolveHostAll(host)); // random order -> random pick
 }
 
-DNSResolver::IPAddresses DNSResolver::resolveHostAll(const std::string & host)
+DNSResolver::IPAddresses DNSResolver::resolveHostAllInOriginOrder(const std::string & host)
 {
     if (impl->disable_cache)
         return resolveIPAddressImpl(host);
@@ -214,6 +215,13 @@ DNSResolver::IPAddresses DNSResolver::resolveHostAll(const std::string & host)
     return resolveIPAddressWithCache(impl->cache_host, host);
 }
 
+DNSResolver::IPAddresses DNSResolver::resolveHostAll(const std::string & host)
+{
+    auto addresses = resolveHostAllInOriginOrder(host);
+    std::shuffle(addresses.begin(), addresses.end(), thread_local_rng);
+    return addresses;
+}
+
 Poco::Net::SocketAddress DNSResolver::resolveAddress(const std::string & host_and_port)
 {
     if (impl->disable_cache)

src/Common/DNSResolver.h

@@ -34,6 +34,9 @@ public:
     Poco::Net::IPAddress resolveHost(const std::string & host);
 
     /// Accepts host names like 'example.com' or '127.0.0.1' or '::1' and resolves all its IPs
+    /// resolveHostAllInOriginOrder returns addresses with the same order as system call returns it
+    IPAddresses resolveHostAllInOriginOrder(const std::string & host);
+    /// resolveHostAll returns addresses in random order
     IPAddresses resolveHostAll(const std::string & host);
 
     /// Accepts host names like 'example.com:port' or '127.0.0.1:port' or '[::1]:port' and resolves its IP and port

src/Common/DateLUTImpl.h

@@ -3,13 +3,13 @@
 #include <base/DayNum.h>
 #include <base/defines.h>
 #include <base/types.h>
+#include <Core/DecimalFunctions.h>
 
 #include <ctime>
 #include <cassert>
 #include <string>
 #include <type_traits>
 
-
 #define DATE_SECONDS_PER_DAY 86400 /// Number of seconds in a day, 60 * 60 * 24
 
 #define DATE_LUT_MIN_YEAR 1900 /// 1900 since majority of financial organizations consider 1900 as an initial year.
@@ -280,9 +280,9 @@ private:
         static_assert(std::is_integral_v<DateOrTime> && std::is_integral_v<Divisor>);
         assert(divisor > 0);
 
-        if (likely(offset_is_whole_number_of_hours_during_epoch))
+        if (offset_is_whole_number_of_hours_during_epoch) [[likely]]
         {
-            if (likely(x >= 0))
+            if (x >= 0) [[likely]]
                 return static_cast<DateOrTime>(x / divisor * divisor);
 
             /// Integer division for negative numbers rounds them towards zero (up).
@@ -576,10 +576,10 @@ public:
 
     unsigned toSecond(Time t) const
     {
-        if (likely(offset_is_whole_number_of_minutes_during_epoch))
+        if (offset_is_whole_number_of_minutes_during_epoch) [[likely]]
         {
             Time res = t % 60;
-            if (likely(res >= 0))
+            if (res >= 0) [[likely]]
                 return static_cast<unsigned>(res);
             return static_cast<unsigned>(res) + 60;
         }
@@ -593,6 +593,30 @@ public:
         return time % 60;
     }
 
+    template <typename DateOrTime>
+    unsigned toMillisecond(const DateOrTime & datetime, Int64 scale_multiplier) const
+    {
+        constexpr Int64 millisecond_multiplier = 1'000;
+        constexpr Int64 microsecond_multiplier = 1'000 * millisecond_multiplier;
+        constexpr Int64 divider = microsecond_multiplier / millisecond_multiplier;
+
+        auto components = DB::DecimalUtils::splitWithScaleMultiplier(datetime, scale_multiplier);
+
+        if (datetime.value < 0 && components.fractional)
+        {
+            components.fractional = scale_multiplier + (components.whole ? Int64(-1) : Int64(1)) * components.fractional;
+            --components.whole;
+        }
+        Int64 fractional = components.fractional;
+        if (scale_multiplier > microsecond_multiplier)
+            fractional = fractional / (scale_multiplier / microsecond_multiplier);
+        else if (scale_multiplier < microsecond_multiplier)
+            fractional = fractional * (microsecond_multiplier / scale_multiplier);
+
+        UInt16 millisecond = static_cast<UInt16>(fractional / divider);
+        return millisecond;
+    }
+
     unsigned toMinute(Time t) const
     {
         if (t >= 0 && offset_is_whole_number_of_hours_during_epoch)
@@ -1122,9 +1146,9 @@ public:
     DateOrTime toStartOfMinuteInterval(DateOrTime t, UInt64 minutes) const
     {
         Int64 divisor = 60 * minutes;
-        if (likely(offset_is_whole_number_of_minutes_during_epoch))
+        if (offset_is_whole_number_of_minutes_during_epoch) [[likely]]
         {
-            if (likely(t >= 0))
+            if (t >= 0) [[likely]]
                 return static_cast<DateOrTime>(t / divisor * divisor);
             return static_cast<DateOrTime>((t + 1 - divisor) / divisor * divisor);
         }
@@ -1339,7 +1363,7 @@ public:
 
     UInt8 saturateDayOfMonth(Int16 year, UInt8 month, UInt8 day_of_month) const
     {
-        if (likely(day_of_month <= 28))
+        if (day_of_month <= 28) [[likely]]
             return day_of_month;
 
         UInt8 days_in_month = daysInMonth(year, month);