chore: Commits for 2.5.0 (#2352)

Testing in: https://github.com/XRPLF/clio/pull/2345

.github/actions/build_docker_image/action.yml

@@ -17,6 +17,9 @@ inputs:
   platforms:
     description: Platforms to build the image for (e.g. linux/amd64,linux/arm64)
     required: true
+  build_args:
+    description: List of build-time variables
+    required: false
 
   dockerhub_repo:
     description: DockerHub repository name
@@ -46,7 +49,7 @@ runs:
     - uses: docker/setup-qemu-action@29109295f81e9208d7d86ff1c6c12d2833863392 # v3.6.0
       with:
         cache-image: false
-    - uses: docker/setup-buildx-action@b5ca514318bd6ebac0fb2aedd5d36ec1b5c232a2 # v3.10.0
+    - uses: docker/setup-buildx-action@e468171a9de216ec08956ac3ada2f0791b6bd435 # v3.11.1
 
     - uses: docker/metadata-action@902fa8ec7d6ecbf8d84d538b9b233a880e428804 # v5.7.0
       id: meta
@@ -61,13 +64,4 @@ runs:
         platforms: ${{ inputs.platforms }}
         push: ${{ inputs.push_image == 'true' }}
         tags: ${{ steps.meta.outputs.tags }}
-
-    - name: Update DockerHub description
-      if: ${{ inputs.push_image == 'true' && inputs.dockerhub_repo != '' }}
-      uses: peter-evans/dockerhub-description@432a30c9e07499fd01da9f8a49f0faf9e0ca5b77 # v4.0.2
-      with:
-        username: ${{ env.DOCKERHUB_USER }}
-        password: ${{ env.DOCKERHUB_PW }}
-        repository: ${{ inputs.dockerhub_repo }}
-        short-description: ${{ inputs.dockerhub_description }}
-        readme-filepath: ${{ inputs.directory }}/README.md
+        build-args: ${{ inputs.build_args }}

.github/actions/generate/action.yml

@@ -5,8 +5,8 @@ inputs:
   conan_profile:
     description: Conan profile name
     required: true
-  conan_cache_hit:
-    description: Whether conan cache has been downloaded
+  force_conan_source_build:
+    description: Whether conan should build all dependencies from source
     required: true
     default: "false"
   build_type:
@@ -25,15 +25,6 @@ inputs:
     description: Whether Clio is to be statically linked
     required: true
     default: "false"
-  sanitizer:
-    description: Sanitizer to use
-    required: true
-    default: "false"
-    choices:
-      - "false"
-      - "tsan"
-      - "asan"
-      - "ubsan"
   time_trace:
     description: Whether to enable compiler trace reports
     required: true
@@ -49,7 +40,7 @@ runs:
     - name: Run conan
       shell: bash
       env:
-        BUILD_OPTION: "${{ inputs.conan_cache_hit == 'true' && 'missing' || '*' }}"
+        CONAN_BUILD_OPTION: "${{ inputs.force_conan_source_build == 'true' && '*' || 'missing' }}"
         CODE_COVERAGE: "${{ inputs.code_coverage == 'true' && 'True' || 'False' }}"
         STATIC_OPTION: "${{ inputs.static == 'true' && 'True' || 'False' }}"
         INTEGRATION_TESTS_OPTION: "${{ inputs.build_integration_tests == 'true' && 'True' || 'False' }}"
@@ -59,7 +50,7 @@ runs:
         conan \
           install .. \
           -of . \
-          -b "$BUILD_OPTION" \
+          -b "$CONAN_BUILD_OPTION" \
           -s "build_type=${{ inputs.build_type }}" \
           -o "&:static=${STATIC_OPTION}" \
           -o "&:tests=True" \
@@ -74,9 +65,9 @@ runs:
       env:
         BUILD_TYPE: "${{ inputs.build_type }}"
         SANITIZER_OPTION: |-
-          ${{ inputs.sanitizer == 'tsan' && '-Dsan=thread' ||
-              inputs.sanitizer == 'ubsan' && '-Dsan=undefined' ||
-              inputs.sanitizer == 'asan' && '-Dsan=address' ||
+          ${{ endsWith(inputs.conan_profile, '.asan') && '-Dsan=address' ||
+              endsWith(inputs.conan_profile, '.tsan') && '-Dsan=thread' ||
+              endsWith(inputs.conan_profile, '.ubsan') && '-Dsan=undefined' ||
               '' }}
       run: |
         cd build

.github/actions/prepare_runner/action.yml

@@ -13,7 +13,7 @@ runs:
       if: ${{ runner.os == 'macOS' }}
       shell: bash
       run: |
-        brew install \
+        brew install --quiet \
           bison \
           ca-certificates \
           ccache \
@@ -31,7 +31,7 @@ runs:
       shell: bash
       run: |
         # Uninstall any existing cmake
-        brew uninstall cmake --ignore-dependencies || true
+        brew uninstall --formula cmake --ignore-dependencies || true
 
         # Download specific cmake formula
         FORMULA_URL="https://raw.githubusercontent.com/Homebrew/homebrew-core/b4e46db74e74a8c1650b38b1da222284ce1ec5ce/Formula/c/cmake.rb"
@@ -43,7 +43,7 @@ runs:
         echo "$FORMULA_EXPECTED_SHA256  /tmp/homebrew-formula/cmake.rb" | shasum -a 256 -c
 
         # Install cmake from the specific formula with force flag
-        brew install --formula --force /tmp/homebrew-formula/cmake.rb
+        brew install --formula --quiet --force /tmp/homebrew-formula/cmake.rb
 
     - name: Fix git permissions on Linux
       if: ${{ runner.os == 'Linux' }}

.github/actions/restore_cache/action.yml

@@ -1,10 +1,7 @@
 name: Restore cache
-description: Find and restores conan and ccache cache
+description: Find and restores ccache cache
 
 inputs:
-  conan_dir:
-    description: Path to Conan directory
-    required: true
   conan_profile:
     description: Conan profile name
     required: true
@@ -19,13 +16,8 @@ inputs:
     description: Whether code coverage is on
     required: true
     default: "false"
+
 outputs:
-  conan_hash:
-    description: Hash to use as a part of conan cache key
-    value: ${{ steps.conan_hash.outputs.hash }}
-  conan_cache_hit:
-    description: True if conan cache has been downloaded
-    value: ${{ steps.conan_cache.outputs.cache-hit }}
   ccache_cache_hit:
     description: True if ccache cache has been downloaded
     value: ${{ steps.ccache_cache.outputs.cache-hit }}
@@ -37,24 +29,6 @@ runs:
       id: git_common_ancestor
       uses: ./.github/actions/git_common_ancestor
 
-    - name: Calculate conan hash
-      id: conan_hash
-      shell: bash
-      run: |
-        conan graph info . --format json --out-file info.json -o '&:tests=True' --profile:all ${{ inputs.conan_profile }}
-        packages_info="$(cat info.json | jq -r '.graph.nodes[]?.ref' | grep -v 'clio')"
-        echo "$packages_info"
-        hash="$(echo "$packages_info" | shasum -a 256 | cut -d ' ' -f 1)"
-        rm info.json
-        echo "hash=$hash" >> $GITHUB_OUTPUT
-
-    - name: Restore conan cache
-      uses: actions/cache/restore@v4
-      id: conan_cache
-      with:
-        path: ${{ inputs.conan_dir }}/p
-        key: clio-conan_data-${{ runner.os }}-${{ inputs.build_type }}-${{ inputs.conan_profile }}-develop-${{ steps.conan_hash.outputs.hash }}
-
     - name: Restore ccache cache
       uses: actions/cache/restore@v4
       id: ccache_cache

.github/actions/save_cache/action.yml

@@ -1,27 +1,13 @@
 name: Save cache
-description: Save conan and ccache cache for develop branch
+description: Save ccache cache for develop branch
 
 inputs:
-  conan_dir:
-    description: Path to .conan directory
-    required: true
   conan_profile:
     description: Conan profile name
     required: true
-  conan_hash:
-    description: Hash to use as a part of conan cache key
-    required: true
-  conan_cache_hit:
-    description: Whether conan cache has been downloaded
-    required: true
   ccache_dir:
     description: Path to .ccache directory
     required: true
-  ccache_cache_hit:
-    description: Whether conan cache has been downloaded
-    required: true
-  ccache_cache_miss_rate:
-    description: How many cache misses happened
   build_type:
     description: Current build type (e.g. Release, Debug)
     required: true
@@ -31,6 +17,12 @@ inputs:
     required: true
     default: "false"
 
+  ccache_cache_hit:
+    description: Whether ccache cache has been downloaded
+    required: true
+  ccache_cache_miss_rate:
+    description: How many ccache cache misses happened
+
 runs:
   using: composite
   steps:
@@ -38,19 +30,6 @@ runs:
       id: git_common_ancestor
       uses: ./.github/actions/git_common_ancestor
 
-    - name: Cleanup conan directory from extra data
-      if: ${{ inputs.conan_cache_hit != 'true' }}
-      shell: bash
-      run: |
-        conan cache clean --source --build --temp
-
-    - name: Save conan cache
-      if: ${{ inputs.conan_cache_hit != 'true' }}
-      uses: actions/cache/save@v4
-      with:
-        path: ${{ inputs.conan_dir }}/p
-        key: clio-conan_data-${{ runner.os }}-${{ inputs.build_type }}-${{ inputs.conan_profile }}-develop-${{ inputs.conan_hash }}
-
     - name: Save ccache cache
       if: ${{ inputs.ccache_cache_hit != 'true' || inputs.ccache_cache_miss_rate == '100.0' }}
       uses: actions/cache/save@v4

.github/actions/setup_conan/action.yml

@@ -1,22 +0,0 @@
-name: Setup conan
-description: Setup conan profile and artifactory
-
-inputs:
-  conan_profile:
-    description: Conan profile name
-    required: true
-
-runs:
-  using: composite
-  steps:
-    - name: Create conan profile on macOS
-      if: ${{ runner.os == 'macOS' }}
-      shell: bash
-      run: |
-        conan profile detect --name "${{ inputs.conan_profile }}" --force
-        sed -i '' 's/compiler.cppstd=[^ ]*/compiler.cppstd=20/' "${{ env.CONAN_HOME }}/profiles/${{ inputs.conan_profile }}"
-
-    - name: Add artifactory remote
-      shell: bash
-      run: |
-        conan remote add --index 0 --force ripple http://18.143.149.228:8081/artifactory/api/conan/dev

.github/dependabot.yml

@@ -142,16 +142,3 @@ updates:
     commit-message:
       prefix: "ci: [DEPENDABOT] "
     target-branch: develop
-
-  - package-ecosystem: github-actions
-    directory: .github/actions/setup_conan/
-    schedule:
-      interval: weekly
-      day: monday
-      time: "04:00"
-      timezone: Etc/GMT
-    reviewers:
-      - XRPLF/clio-dev-team
-    commit-message:
-      prefix: "ci: [DEPENDABOT] "
-    target-branch: develop

.github/scripts/conan/apple-clang-ci.profile

@@ -0,0 +1,8 @@
+[settings]
+arch={{detect_api.detect_arch()}}
+build_type=Release
+compiler=apple-clang
+compiler.cppstd=20
+compiler.libcxx=libc++
+compiler.version=16
+os=Macos

.github/scripts/conan/apple-clang-local.profile

@@ -0,0 +1,11 @@
+[settings]
+arch={{detect_api.detect_arch()}}
+build_type=Release
+compiler=apple-clang
+compiler.cppstd=20
+compiler.libcxx=libc++
+compiler.version=17
+os=Macos
+
+[conf]
+grpc/1.50.1:tools.build:cxxflags+=["-Wno-missing-template-arg-list-after-template-kw"]

.github/scripts/conan/generate_matrix.py

@@ -0,0 +1,39 @@
+#!/usr/bin/env python3
+import itertools
+import json
+
+LINUX_OS = ["heavy", "heavy-arm64"]
+LINUX_CONTAINERS = ['{ "image": "ghcr.io/xrplf/clio-ci:latest" }']
+LINUX_COMPILERS = ["gcc", "clang"]
+
+MACOS_OS = ["macos15"]
+MACOS_CONTAINERS = [""]
+MACOS_COMPILERS = ["apple-clang"]
+
+BUILD_TYPES = ["Release", "Debug"]
+SANITIZER_EXT = [".asan", ".tsan", ".ubsan", ""]
+
+
+def generate_matrix():
+    configurations = []
+
+    for os, container, compiler in itertools.chain(
+        itertools.product(LINUX_OS, LINUX_CONTAINERS, LINUX_COMPILERS),
+        itertools.product(MACOS_OS, MACOS_CONTAINERS, MACOS_COMPILERS),
+    ):
+        for sanitizer_ext, build_type in itertools.product(SANITIZER_EXT, BUILD_TYPES):
+            configurations.append(
+                {
+                    "os": os,
+                    "container": container,
+                    "compiler": compiler,
+                    "sanitizer_ext": sanitizer_ext,
+                    "build_type": build_type,
+                }
+            )
+
+    return {"include": configurations}
+
+
+if __name__ == "__main__":
+    print(f"matrix={json.dumps(generate_matrix())}")

.github/scripts/conan/init.sh

@@ -0,0 +1,47 @@
+#!/bin/bash
+
+set -ex
+
+CURRENT_DIR="$(cd "$(dirname "$0")" && pwd)"
+REPO_DIR="$(cd "$CURRENT_DIR/../../../" && pwd)"
+
+CONAN_DIR="${CONAN_HOME:-$HOME/.conan2}"
+PROFILES_DIR="$CONAN_DIR/profiles"
+
+if [[ -z "$CI" ]]; then
+    APPLE_CLANG_PROFILE="$CURRENT_DIR/apple-clang-local.profile"
+else
+    APPLE_CLANG_PROFILE="$CURRENT_DIR/apple-clang-ci.profile"
+fi
+
+GCC_PROFILE="$REPO_DIR/docker/ci/conan/gcc.profile"
+CLANG_PROFILE="$REPO_DIR/docker/ci/conan/clang.profile"
+
+SANITIZER_TEMPLATE_FILE="$REPO_DIR/docker/ci/conan/sanitizer_template.profile"
+
+rm -rf "$CONAN_DIR"
+
+conan remote add --index 0 ripple http://18.143.149.228:8081/artifactory/api/conan/dev
+
+cp "$REPO_DIR/docker/ci/conan/global.conf" "$CONAN_DIR/global.conf"
+
+create_profile_with_sanitizers() {
+    profile_name="$1"
+    profile_source="$2"
+
+    cp "$profile_source" "$PROFILES_DIR/$profile_name"
+    cp "$SANITIZER_TEMPLATE_FILE" "$PROFILES_DIR/$profile_name.asan"
+    cp "$SANITIZER_TEMPLATE_FILE" "$PROFILES_DIR/$profile_name.tsan"
+    cp "$SANITIZER_TEMPLATE_FILE" "$PROFILES_DIR/$profile_name.ubsan"
+}
+
+mkdir -p "$PROFILES_DIR"
+
+if [[ "$(uname)" == "Darwin" ]]; then
+    create_profile_with_sanitizers "apple-clang" "$APPLE_CLANG_PROFILE"
+    echo "include(apple-clang)" > "$PROFILES_DIR/default"
+else
+    create_profile_with_sanitizers "clang" "$CLANG_PROFILE"
+    create_profile_with_sanitizers "gcc" "$GCC_PROFILE"
+    echo "include(gcc)" > "$PROFILES_DIR/default"
+fi

.github/scripts/update-libxrpl-version

@@ -1,28 +0,0 @@
-#!/bin/bash
-
-# Note: This script is intended to be run from the root of the repository.
-#
-# This script modifies conanfile.py such that the specified version of libXRPL is used.
-
-if [[ -z "$1" ]]; then
-    cat <<EOF
-
-                                    ERROR
------------------------------------------------------------------------------
-            Version should be passed as first argument to the script.
------------------------------------------------------------------------------
-
-EOF
-    exit 1
-fi
-
-VERSION=$1
-GNU_SED=$(sed --version 2>&1 | grep -q 'GNU' && echo true || echo false)
-
-echo "+ Updating required libXRPL version to $VERSION"
-
-if [[ "$GNU_SED" == "false" ]]; then
-    sed -i '' -E "s|'xrpl/[a-zA-Z0-9\\.\\-]+'|'xrpl/$VERSION'|g" conanfile.py
-else
-    sed -i -E "s|'xrpl/[a-zA-Z0-9\\.\\-]+'|'xrpl/$VERSION'|g" conanfile.py
-fi

.github/workflows/build.yml

@@ -2,9 +2,9 @@ name: Build
 
 on:
   push:
-    branches: [master, release/*, develop]
+    branches: [release/*, develop]
   pull_request:
-    branches: [master, release/*, develop]
+    branches: [release/*, develop]
     paths:
       - .github/workflows/build.yml
 
@@ -19,6 +19,7 @@ on:
 
       - CMakeLists.txt
       - conanfile.py
+      - conan.lock
       - "cmake/**"
       - "src/**"
       - "tests/**"
@@ -46,7 +47,7 @@ jobs:
 
         include:
           - os: macos15
-            conan_profile: default_apple_clang
+            conan_profile: apple-clang
             build_type: Release
             container: ""
             static: false
@@ -76,7 +77,6 @@ jobs:
       static: true
       upload_clio_server: false
       targets: all
-      sanitizer: "false"
       analyze_build_time: false
     secrets:
       CODECOV_TOKEN: ${{ secrets.CODECOV_TOKEN }}
@@ -99,23 +99,9 @@ jobs:
         shell: bash
         run: |
           repoConfigFile=docs/config-description.md
-          if ! [ -f "${repoConfigFile}" ]; then
-            echo "Config Description markdown file is missing in docs folder"
-            exit 1
-          fi
+          configDescriptionFile=config_description_new.md
 
           chmod +x ./clio_server
-          configDescriptionFile=config_description_new.md
           ./clio_server -d "${configDescriptionFile}"
 
-          configDescriptionHash=$(sha256sum "${configDescriptionFile}" | cut -d' ' -f1)
-          repoConfigHash=$(sha256sum "${repoConfigFile}" | cut -d' ' -f1)
-
-          if [ "${configDescriptionHash}" != "${repoConfigHash}" ]; then
-            echo "Markdown file is not up to date"
-            diff -u "${repoConfigFile}" "${configDescriptionFile}"
-            rm -f "${configDescriptionFile}"
-            exit 1
-          fi
-          rm -f "${configDescriptionFile}"
-          exit 0
+          diff -u "${repoConfigFile}" "${configDescriptionFile}"

.github/workflows/build_and_test.yml

@@ -24,7 +24,7 @@ on:
         type: string
 
       disable_cache:
-        description: Whether ccache and conan cache should be disabled
+        description: Whether ccache should be disabled
         required: false
         type: boolean
         default: false
@@ -57,12 +57,6 @@ on:
         type: string
         default: all
 
-      sanitizer:
-        description: Sanitizer to use
-        required: false
-        type: string
-        default: "false"
-
 jobs:
   build:
     uses: ./.github/workflows/build_impl.yml
@@ -76,7 +70,6 @@ jobs:
       static: ${{ inputs.static }}
       upload_clio_server: ${{ inputs.upload_clio_server }}
       targets: ${{ inputs.targets }}
-      sanitizer: ${{ inputs.sanitizer }}
       analyze_build_time: false
 
   test:
@@ -89,4 +82,3 @@ jobs:
       build_type: ${{ inputs.build_type }}
       run_unit_tests: ${{ inputs.run_unit_tests }}
       run_integration_tests: ${{ inputs.run_integration_tests }}
-      sanitizer: ${{ inputs.sanitizer }}

.github/workflows/build_impl.yml

@@ -24,7 +24,7 @@ on:
         type: string
 
       disable_cache:
-        description: Whether ccache and conan cache should be disabled
+        description: Whether ccache should be disabled
         required: false
         type: boolean
 
@@ -48,11 +48,6 @@ on:
         required: true
         type: string
 
-      sanitizer:
-        description: Sanitizer to use
-        required: true
-        type: string
-
       analyze_build_time:
         description: Whether to enable build time analysis
         required: true
@@ -64,7 +59,7 @@ on:
 
 jobs:
   build:
-    name: Build ${{ inputs.container != '' && 'in container' || 'natively' }}
+    name: Build
     runs-on: ${{ inputs.runs_on }}
     container: ${{ inputs.container != '' && fromJson(inputs.container) || null }}
 
@@ -82,17 +77,16 @@ jobs:
         with:
           disable_ccache: ${{ inputs.disable_cache }}
 
-      - name: Setup conan
-        uses: ./.github/actions/setup_conan
-        with:
-          conan_profile: ${{ inputs.conan_profile }}
+      - name: Setup conan on macOS
+        if: runner.os == 'macOS'
+        shell: bash
+        run: ./.github/scripts/conan/init.sh
 
       - name: Restore cache
         if: ${{ !inputs.disable_cache }}
         uses: ./.github/actions/restore_cache
         id: restore_cache
         with:
-          conan_dir: ${{ env.CONAN_HOME }}
           conan_profile: ${{ inputs.conan_profile }}
           ccache_dir: ${{ env.CCACHE_DIR }}
           build_type: ${{ inputs.build_type }}
@@ -102,11 +96,9 @@ jobs:
         uses: ./.github/actions/generate
         with:
           conan_profile: ${{ inputs.conan_profile }}
-          conan_cache_hit: ${{ !inputs.disable_cache && steps.restore_cache.outputs.conan_cache_hit }}
           build_type: ${{ inputs.build_type }}
           code_coverage: ${{ inputs.code_coverage }}
           static: ${{ inputs.static }}
-          sanitizer: ${{ inputs.sanitizer }}
           time_trace: ${{ inputs.analyze_build_time }}
 
       - name: Build Clio
@@ -140,11 +132,11 @@ jobs:
           cat /tmp/ccache.stats
 
       - name: Strip unit_tests
-        if: inputs.sanitizer == 'false' && !inputs.code_coverage && !inputs.analyze_build_time
+        if: ${{ !endsWith(inputs.conan_profile, 'san') && !inputs.code_coverage && !inputs.analyze_build_time }}
         run: strip build/clio_tests
 
       - name: Strip integration_tests
-        if: inputs.sanitizer == 'false' && !inputs.code_coverage && !inputs.analyze_build_time
+        if: ${{ !endsWith(inputs.conan_profile, 'san') && !inputs.code_coverage && !inputs.analyze_build_time }}
         run: strip build/clio_integration_tests
 
       - name: Upload clio_server
@@ -172,15 +164,13 @@ jobs:
         if: ${{ !inputs.disable_cache && github.ref == 'refs/heads/develop' }}
         uses: ./.github/actions/save_cache
         with:
-          conan_dir: ${{ env.CONAN_HOME }}
-          conan_hash: ${{ steps.restore_cache.outputs.conan_hash }}
-          conan_cache_hit: ${{ steps.restore_cache.outputs.conan_cache_hit }}
+          conan_profile: ${{ inputs.conan_profile }}
           ccache_dir: ${{ env.CCACHE_DIR }}
-          ccache_cache_hit: ${{ steps.restore_cache.outputs.ccache_cache_hit }}
-          ccache_cache_miss_rate: ${{ steps.ccache_stats.outputs.miss_rate }}
           build_type: ${{ inputs.build_type }}
           code_coverage: ${{ inputs.code_coverage }}
-          conan_profile: ${{ inputs.conan_profile }}
+
+          ccache_cache_hit: ${{ steps.restore_cache.outputs.ccache_cache_hit }}
+          ccache_cache_miss_rate: ${{ steps.ccache_stats.outputs.miss_rate }}
 
       # This is run as part of the build job, because it requires the following:
       # - source code

.github/workflows/check_libxrpl.yml

@@ -15,7 +15,7 @@ env:
 jobs:
   build:
     name: Build Clio / `libXRPL ${{ github.event.client_payload.version }}`
-    runs-on: [self-hosted, heavy]
+    runs-on: heavy
     container:
       image: ghcr.io/xrplf/clio-ci:latest
 
@@ -27,24 +27,23 @@ jobs:
       - name: Update libXRPL version requirement
         shell: bash
         run: |
-          ./.github/scripts/update-libxrpl-version ${{ github.event.client_payload.version }}
+          sed -i.bak -E "s|'xrpl/[a-zA-Z0-9\\.\\-]+'|'xrpl/${{ github.event.client_payload.version }}'|g" conanfile.py
+          rm -f conanfile.py.bak
+
+      - name: Update conan lockfile
+        shell: bash
+        run: |
+          conan lock create . -o '&:tests=True' -o '&:benchmark=True'
 
       - name: Prepare runner
         uses: ./.github/actions/prepare_runner
         with:
           disable_ccache: true
 
-      - name: Setup conan
-        uses: ./.github/actions/setup_conan
-        with:
-          conan_profile: ${{ env.CONAN_PROFILE }}
-
       - name: Run conan and cmake
         uses: ./.github/actions/generate
         with:
           conan_profile: ${{ env.CONAN_PROFILE }}
-          conan_cache_hit: ${{ steps.restore_cache.outputs.conan_cache_hit }}
-          build_type: Release
 
       - name: Build Clio
         uses: ./.github/actions/build_clio
@@ -61,7 +60,7 @@ jobs:
   run_tests:
     name: Run tests
     needs: build
-    runs-on: [self-hosted, heavy]
+    runs-on: heavy
     container:
       image: ghcr.io/xrplf/clio-ci:latest
 

.github/workflows/clang-tidy.yml

@@ -18,6 +18,7 @@ concurrency:
 
 env:
   CONAN_PROFILE: clang
+  LLVM_TOOLS_VERSION: 20
 
 jobs:
   clang_tidy:
@@ -40,25 +41,17 @@ jobs:
         with:
           disable_ccache: true
 
-      - name: Setup conan
-        uses: ./.github/actions/setup_conan
-        with:
-          conan_profile: ${{ env.CONAN_PROFILE }}
-
       - name: Restore cache
         uses: ./.github/actions/restore_cache
         id: restore_cache
         with:
-          conan_dir: ${{ env.CONAN_HOME }}
-          ccache_dir: ${{ env.CCACHE_DIR }}
           conan_profile: ${{ env.CONAN_PROFILE }}
+          ccache_dir: ${{ env.CCACHE_DIR }}
 
       - name: Run conan and cmake
         uses: ./.github/actions/generate
         with:
           conan_profile: ${{ env.CONAN_PROFILE }}
-          conan_cache_hit: ${{ steps.restore_cache.outputs.conan_cache_hit }}
-          build_type: Release
 
       - name: Get number of threads
         uses: ./.github/actions/get_number_of_threads
@@ -69,7 +62,7 @@ jobs:
         shell: bash
         id: run_clang_tidy
         run: |
-          run-clang-tidy-19 -p build -j "${{ steps.number_of_threads.outputs.threads_number }}" -fix -quiet 1>output.txt
+          run-clang-tidy-${{ env.LLVM_TOOLS_VERSION }} -p build -j "${{ steps.number_of_threads.outputs.threads_number }}" -fix -quiet 1>output.txt
 
       - name: Fix local includes and clang-format style
         if: ${{ steps.run_clang_tidy.outcome != 'success' }}

.github/workflows/docs.yml

@@ -22,6 +22,11 @@ jobs:
         with:
           lfs: true
 
+      - name: Prepare runner
+        uses: ./.github/actions/prepare_runner
+        with:
+          disable_ccache: true
+
       - name: Create build directory
         run: mkdir build_docs
 

.github/workflows/nightly.yml

@@ -32,28 +32,24 @@ jobs:
       matrix:
         include:
           - os: macos15
-            conan_profile: default_apple_clang
+            conan_profile: apple-clang
             build_type: Release
             static: false
-            sanitizer: "false"
           - os: heavy
             conan_profile: gcc
             build_type: Release
             static: true
             container: '{ "image": "ghcr.io/xrplf/clio-ci:latest" }'
-            sanitizer: "false"
           - os: heavy
             conan_profile: gcc
             build_type: Debug
             static: true
             container: '{ "image": "ghcr.io/xrplf/clio-ci:latest" }'
-            sanitizer: "false"
           - os: heavy
             conan_profile: gcc.ubsan
             build_type: Release
             static: false
             container: '{ "image": "ghcr.io/xrplf/clio-ci:latest" }'
-            sanitizer: "ubsan"
 
     uses: ./.github/workflows/build_and_test.yml
     with:
@@ -66,7 +62,6 @@ jobs:
       run_integration_tests: true
       upload_clio_server: true
       disable_cache: true
-      sanitizer: ${{ matrix.sanitizer }}
 
   analyze_build_time:
     name: Analyze Build Time
@@ -80,7 +75,7 @@ jobs:
             container: '{ "image": "ghcr.io/xrplf/clio-ci:latest" }'
             static: true
           - os: macos15
-            conan_profile: default_apple_clang
+            conan_profile: apple-clang
             container: ""
             static: false
     uses: ./.github/workflows/build_impl.yml
@@ -94,7 +89,6 @@ jobs:
       static: ${{ matrix.static }}
       upload_clio_server: false
       targets: all
-      sanitizer: "false"
       analyze_build_time: true
 
   nightly_release:
@@ -102,6 +96,7 @@ jobs:
     uses: ./.github/workflows/release_impl.yml
     with:
       overwrite_release: true
+      prerelease: true
       title: "Clio development (nightly) build"
       version: nightly
       header: >

.github/workflows/pre-commit-autoupdate.yml

@@ -26,14 +26,25 @@ jobs:
       - run: pre-commit autoupdate --freeze
       - run: pre-commit run --all-files || true
 
+      - uses: crazy-max/ghaction-import-gpg@e89d40939c28e39f97cf32126055eeae86ba74ec # v6.3.0
+        if: github.event_name != 'pull_request'
+        with:
+          gpg_private_key: ${{ secrets.ACTIONS_GPG_PRIVATE_KEY }}
+          passphrase: ${{ secrets.ACTIONS_GPG_PASSPHRASE }}
+          git_user_signingkey: true
+          git_commit_gpgsign: true
+
       - uses: peter-evans/create-pull-request@271a8d0340265f705b14b6d32b9829c1cb33d45e # v7.0.8
         if: always()
         env:
           GH_REPO: ${{ github.repository }}
           GH_TOKEN: ${{ github.token }}
         with:
-          branch: update/pre-commit-hooks
-          title: "style: Update pre-commit hooks"
           commit-message: "style: Update pre-commit hooks"
+          committer: Clio CI <skuznetsov@ripple.com>
+          branch: update/pre-commit-hooks
+          branch-suffix: timestamp
+          delete-branch: true
+          title: "style: Update pre-commit hooks"
           body: Update versions of pre-commit hooks to latest version.
           reviewers: "godexsoft,kuznetsss,PeterChen13579,mathbunnyru"

.github/workflows/pre-commit.yml

@@ -3,8 +3,7 @@ name: Run pre-commit hooks
 on:
   pull_request:
   push:
-    branches:
-      - develop
+    branches: [develop]
   workflow_dispatch:
 
 jobs:

.github/workflows/release.yml

@@ -22,7 +22,7 @@ jobs:
       matrix:
         include:
           - os: macos15
-            conan_profile: default_apple_clang
+            conan_profile: apple-clang
             build_type: Release
             static: false
           - os: heavy
@@ -48,6 +48,7 @@ jobs:
     uses: ./.github/workflows/release_impl.yml
     with:
       overwrite_release: false
+      prerelease: ${{ contains(github.ref_name, '-') }}
       title: "${{ github.ref_name}}"
       version: "${{ github.ref_name }}"
       header: >

.github/workflows/release_impl.yml

@@ -8,6 +8,11 @@ on:
         required: true
         type: boolean
 
+      prerelease:
+        description: "Create a prerelease"
+        required: true
+        type: boolean
+
       title:
         description: "Release title"
         required: true
@@ -25,12 +30,12 @@ on:
 
       generate_changelog:
         description: "Generate changelog"
-        required: false
+        required: true
         type: boolean
 
       draft:
         description: "Create a draft release"
-        required: false
+        required: true
         type: boolean
 
 jobs:
@@ -69,9 +74,9 @@ jobs:
         shell: bash
         if: ${{ inputs.generate_changelog }}
         run: |
-          LAST_TAG=$(gh release view --json tagName -q .tagName)
-          LAST_TAG_COMMIT=$(git rev-parse $LAST_TAG)
-          BASE_COMMIT=$(git merge-base HEAD $LAST_TAG_COMMIT)
+          LAST_TAG="$(gh release view --json tagName -q .tagName)"
+          LAST_TAG_COMMIT="$(git rev-parse $LAST_TAG)"
+          BASE_COMMIT="$(git merge-base HEAD $LAST_TAG_COMMIT)"
           git-cliff "${BASE_COMMIT}..HEAD" --ignore-tags "nightly|-b"
           cat CHANGELOG.md >> "${RUNNER_TEMP}/release_notes.md"
 
@@ -108,10 +113,10 @@ jobs:
         if: ${{ github.event_name != 'pull_request' }}
         shell: bash
         run: |
-          gh release create ${{ inputs.version }} \
-            ${{ inputs.overwrite_release && '--prerelease' || '' }} \
+          gh release create "${{ inputs.version }}" \
+            ${{ inputs.prerelease && '--prerelease' || '' }} \
             --title "${{ inputs.title }}" \
-            --target $GITHUB_SHA \
+            --target "${GITHUB_SHA}" \
             ${{ inputs.draft && '--draft' || '' }} \
             --notes-file "${RUNNER_TEMP}/release_notes.md" \
             ./release_artifacts/clio_server*

.github/workflows/sanitizers.yml

@@ -19,6 +19,7 @@ on:
 
       - CMakeLists.txt
       - conanfile.py
+      - conan.lock
       - "cmake/**"
       # We don't run sanitizer on code change, because it takes too long
       # - "src/**"
@@ -36,24 +37,20 @@ jobs:
     strategy:
       fail-fast: false
       matrix:
-        include:
-          - sanitizer: tsan
-            compiler: gcc
-          - sanitizer: asan
-            compiler: gcc
-          - sanitizer: ubsan
-            compiler: gcc
+        compiler: [gcc, clang]
+        sanitizer_ext: [.asan, .tsan, .ubsan]
+        build_type: [Release, Debug]
 
     uses: ./.github/workflows/build_and_test.yml
     with:
       runs_on: heavy
       container: '{ "image": "ghcr.io/xrplf/clio-ci:latest" }'
       disable_cache: true
-      conan_profile: ${{ matrix.compiler }}.${{ matrix.sanitizer }}
-      build_type: Release
+      conan_profile: ${{ matrix.compiler }}${{ matrix.sanitizer_ext }}
+      build_type: ${{ matrix.build_type }}
       static: false
-      run_unit_tests: true
+      # Currently, both gcc.tsan and clang.tsan unit tests hang
+      run_unit_tests: ${{ matrix.sanitizer_ext != '.tsan' }}
       run_integration_tests: false
       upload_clio_server: false
       targets: clio_tests clio_integration_tests
-      sanitizer: ${{ matrix.sanitizer }}

.github/workflows/test_impl.yml

@@ -33,22 +33,17 @@ on:
         required: true
         type: boolean
 
-      sanitizer:
-        description: Sanitizer to use
-        required: true
-        type: string
-
 jobs:
   unit_tests:
-    name: Unit testing ${{ inputs.container != '' && 'in container' || 'natively' }}
+    name: Unit testing
     runs-on: ${{ inputs.runs_on }}
     container: ${{ inputs.container != '' && fromJson(inputs.container) || null }}
 
     if: inputs.run_unit_tests
 
     env:
-      # TODO: remove when we have fixed all currently existing issues from sanitizers
-      SANITIZER_IGNORE_ERRORS: ${{ inputs.sanitizer != 'false' && inputs.sanitizer != 'ubsan' }}
+      # TODO: remove completely when we have fixed all currently existing issues with sanitizers
+      SANITIZER_IGNORE_ERRORS: ${{ endsWith(inputs.conan_profile, '.asan') || endsWith(inputs.conan_profile, '.tsan') }}
 
     steps:
       - name: Clean workdir
@@ -90,7 +85,7 @@ jobs:
         if: env.SANITIZER_IGNORE_ERRORS == 'true' && steps.check_report.outputs.found_report == 'true'
         uses: actions/upload-artifact@v4
         with:
-          name: ${{ inputs.conan_profile }}_report
+          name: sanitizer_report_${{ runner.os }}_${{ inputs.build_type }}_${{ inputs.conan_profile }}
           path: .sanitizer-report/*
           include-hidden-files: true
 
@@ -109,7 +104,7 @@ jobs:
             Reports are available as artifacts.
 
   integration_tests:
-    name: Integration testing ${{ inputs.container != '' && 'in container' || 'natively' }}
+    name: Integration testing
     runs-on: ${{ inputs.runs_on }}
     container: ${{ inputs.container != '' && fromJson(inputs.container) || null }}
 

.github/workflows/update_docker_ci.yml

@@ -23,14 +23,33 @@ on:
   workflow_dispatch:
 
 concurrency:
-  # Only cancel in-progress jobs or runs for the current workflow - matches against branch & tags
+  # Only matches runs for the current workflow - matches against branch & tags
   group: ${{ github.workflow }}-${{ github.ref }}
-  cancel-in-progress: true
+  # We want to execute all builds sequentially in develop
+  cancel-in-progress: false
+
+env:
+  CLANG_MAJOR_VERSION: 19
+  GCC_MAJOR_VERSION: 14
+  GCC_VERSION: 14.3.0
 
 jobs:
-  gcc:
-    name: Build and push GCC docker image
-    runs-on: [self-hosted, heavy]
+  repo:
+    name: Calculate repo name
+    runs-on: ubuntu-latest
+    outputs:
+      GHCR_REPO: ${{ steps.set-ghcr-repo.outputs.GHCR_REPO }}
+
+    steps:
+      - name: Set GHCR_REPO
+        id: set-ghcr-repo
+        run: |
+          echo "GHCR_REPO=$(echo ghcr.io/${{ github.repository_owner }} | tr '[:upper:]' '[:lower:]')" >> ${GITHUB_OUTPUT}
+
+  gcc-amd64:
+    name: Build and push GCC docker image (amd64)
+    runs-on: heavy
+    needs: repo
 
     steps:
       - uses: actions/checkout@v4
@@ -42,30 +61,116 @@ jobs:
           files: "docker/compilers/gcc/**"
 
       - uses: ./.github/actions/build_docker_image
-        # Skipping this build for now, because CI environment is not stable
-        if: false && steps.changed-files.outputs.any_changed == 'true'
+        if: steps.changed-files.outputs.any_changed == 'true'
         env:
           GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
           DOCKERHUB_USER: ${{ secrets.DOCKERHUB_USER }}
           DOCKERHUB_PW: ${{ secrets.DOCKERHUB_PW }}
         with:
           images: |
-            ghcr.io/xrplf/clio-gcc
+            ${{ needs.repo.outputs.GHCR_REPO }}/clio-gcc
             rippleci/clio_gcc
           push_image: ${{ github.event_name != 'pull_request' }}
           directory: docker/compilers/gcc
           tags: |
-            type=raw,value=latest
-            type=raw,value=12
-            type=raw,value=12.3.0
-            type=raw,value=${{ github.sha }}
-          platforms: linux/amd64,linux/arm64
+            type=raw,value=amd64-latest
+            type=raw,value=amd64-${{ env.GCC_MAJOR_VERSION }}
+            type=raw,value=amd64-${{ env.GCC_VERSION }}
+            type=raw,value=amd64-${{ github.sha }}
+          platforms: linux/amd64
+          build_args: |
+            GCC_MAJOR_VERSION=${{ env.GCC_MAJOR_VERSION }}
+            GCC_VERSION=${{ env.GCC_VERSION }}
           dockerhub_repo: rippleci/clio_gcc
           dockerhub_description: GCC compiler for XRPLF/clio.
 
+  gcc-arm64:
+    name: Build and push GCC docker image (arm64)
+    runs-on: heavy-arm64
+    needs: repo
+
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Get changed files
+        id: changed-files
+        uses: tj-actions/changed-files@ed68ef82c095e0d48ec87eccea555d944a631a4c # v46.0.5
+        with:
+          files: "docker/compilers/gcc/**"
+
+      - uses: ./.github/actions/build_docker_image
+        if: steps.changed-files.outputs.any_changed == 'true'
+        env:
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+          DOCKERHUB_USER: ${{ secrets.DOCKERHUB_USER }}
+          DOCKERHUB_PW: ${{ secrets.DOCKERHUB_PW }}
+        with:
+          images: |
+            ${{ needs.repo.outputs.GHCR_REPO }}/clio-gcc
+            rippleci/clio_gcc
+          push_image: ${{ github.event_name != 'pull_request' }}
+          directory: docker/compilers/gcc
+          tags: |
+            type=raw,value=arm64-latest
+            type=raw,value=arm64-${{ env.GCC_MAJOR_VERSION }}
+            type=raw,value=arm64-${{ env.GCC_VERSION }}
+            type=raw,value=arm64-${{ github.sha }}
+          platforms: linux/arm64
+          build_args: |
+            GCC_MAJOR_VERSION=${{ env.GCC_MAJOR_VERSION }}
+            GCC_VERSION=${{ env.GCC_VERSION }}
+          dockerhub_repo: rippleci/clio_gcc
+          dockerhub_description: GCC compiler for XRPLF/clio.
+
+  gcc-merge:
+    name: Merge and push multi-arch GCC docker image
+    runs-on: heavy
+    needs: [repo, gcc-amd64, gcc-arm64]
+
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Get changed files
+        id: changed-files
+        uses: tj-actions/changed-files@ed68ef82c095e0d48ec87eccea555d944a631a4c # v46.0.5
+        with:
+          files: "docker/compilers/gcc/**"
+
+      - name: Set up Docker Buildx
+        uses: docker/setup-buildx-action@v3
+
+      - name: Login to GitHub Container Registry
+        if: github.event_name != 'pull_request'
+        uses: docker/login-action@74a5d142397b4f367a81961eba4e8cd7edddf772 # v3.4.0
+        with:
+          registry: ghcr.io
+          username: ${{ github.repository_owner }}
+          password: ${{ secrets.GITHUB_TOKEN }}
+
+      - name: Login to DockerHub
+        if: github.event_name != 'pull_request'
+        uses: docker/login-action@v3
+        with:
+          username: ${{ secrets.DOCKERHUB_USER }}
+          password: ${{ secrets.DOCKERHUB_PW }}
+
+      - name: Create and push multi-arch manifest
+        if: github.event_name != 'pull_request' && steps.changed-files.outputs.any_changed == 'true'
+        run: |
+          for image in ${{ needs.repo.outputs.GHCR_REPO }}/clio-gcc rippleci/clio_gcc; do
+            docker buildx imagetools create \
+              -t $image:latest \
+              -t $image:${{ env.GCC_MAJOR_VERSION }} \
+              -t $image:${{ env.GCC_VERSION }} \
+              -t $image:${{ github.sha }} \
+              $image:arm64-latest \
+              $image:amd64-latest
+          done
+
   clang:
     name: Build and push Clang docker image
-    runs-on: [self-hosted, heavy]
+    runs-on: heavy
+    needs: repo
 
     steps:
       - uses: actions/checkout@v4
@@ -84,21 +189,24 @@ jobs:
           DOCKERHUB_PW: ${{ secrets.DOCKERHUB_PW }}
         with:
           images: |
-            ghcr.io/xrplf/clio-clang
+            ${{ needs.repo.outputs.GHCR_REPO }}/clio-clang
             rippleci/clio_clang
           push_image: ${{ github.event_name != 'pull_request' }}
           directory: docker/compilers/clang
           tags: |
             type=raw,value=latest
-            type=raw,value=16
+            type=raw,value=${{ env.CLANG_MAJOR_VERSION }}
             type=raw,value=${{ github.sha }}
           platforms: linux/amd64,linux/arm64
+          build_args: |
+            CLANG_MAJOR_VERSION=${{ env.CLANG_MAJOR_VERSION }}
           dockerhub_repo: rippleci/clio_clang
           dockerhub_description: Clang compiler for XRPLF/clio.
 
-  tools:
-    name: Build and push tools docker image
-    runs-on: [self-hosted, heavy]
+  tools-amd64:
+    name: Build and push tools docker image (amd64)
+    runs-on: heavy
+    needs: [repo, gcc-merge]
 
     steps:
       - uses: actions/checkout@v4
@@ -115,18 +223,87 @@ jobs:
           GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
         with:
           images: |
-            ghcr.io/xrplf/clio-tools
+            ${{ needs.repo.outputs.GHCR_REPO }}/clio-tools
           push_image: ${{ github.event_name != 'pull_request' }}
           directory: docker/tools
           tags: |
-            type=raw,value=latest
-            type=raw,value=${{ github.sha }}
-          platforms: linux/amd64,linux/arm64
+            type=raw,value=amd64-latest
+            type=raw,value=amd64-${{ github.sha }}
+          platforms: linux/amd64
+          build_args: |
+            GHCR_REPO=${{ needs.repo.outputs.GHCR_REPO }}
+            GCC_VERSION=${{ env.GCC_VERSION }}
+
+  tools-arm64:
+    name: Build and push tools docker image (arm64)
+    runs-on: heavy-arm64
+    needs: [repo, gcc-merge]
+
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Get changed files
+        id: changed-files
+        uses: tj-actions/changed-files@ed68ef82c095e0d48ec87eccea555d944a631a4c # v46.0.5
+        with:
+          files: "docker/tools/**"
+
+      - uses: ./.github/actions/build_docker_image
+        if: steps.changed-files.outputs.any_changed == 'true'
+        env:
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+        with:
+          images: |
+            ${{ needs.repo.outputs.GHCR_REPO }}/clio-tools
+          push_image: ${{ github.event_name != 'pull_request' }}
+          directory: docker/tools
+          tags: |
+            type=raw,value=arm64-latest
+            type=raw,value=arm64-${{ github.sha }}
+          platforms: linux/arm64
+          build_args: |
+            GHCR_REPO=${{ needs.repo.outputs.GHCR_REPO }}
+            GCC_VERSION=${{ env.GCC_VERSION }}
+
+  tools-merge:
+    name: Merge and push multi-arch tools docker image
+    runs-on: heavy
+    needs: [repo, tools-amd64, tools-arm64]
+
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Get changed files
+        id: changed-files
+        uses: tj-actions/changed-files@ed68ef82c095e0d48ec87eccea555d944a631a4c # v46.0.5
+        with:
+          files: "docker/tools/**"
+
+      - name: Set up Docker Buildx
+        uses: docker/setup-buildx-action@v3
+
+      - name: Login to GitHub Container Registry
+        if: github.event_name != 'pull_request'
+        uses: docker/login-action@74a5d142397b4f367a81961eba4e8cd7edddf772 # v3.4.0
+        with:
+          registry: ghcr.io
+          username: ${{ github.repository_owner }}
+          password: ${{ secrets.GITHUB_TOKEN }}
+
+      - name: Create and push multi-arch manifest
+        if: github.event_name != 'pull_request' && steps.changed-files.outputs.any_changed == 'true'
+        run: |
+          image=${{ needs.repo.outputs.GHCR_REPO }}/clio-tools
+          docker buildx imagetools create \
+            -t $image:latest \
+            -t $image:${{ github.sha }} \
+            $image:arm64-latest \
+            $image:amd64-latest
 
   ci:
     name: Build and push CI docker image
-    runs-on: [self-hosted, heavy]
-    needs: [gcc, clang, tools]
+    runs-on: heavy
+    needs: [repo, gcc-merge, clang, tools-merge]
 
     steps:
       - uses: actions/checkout@v4
@@ -137,14 +314,19 @@ jobs:
           DOCKERHUB_PW: ${{ secrets.DOCKERHUB_PW }}
         with:
           images: |
+            ${{ needs.repo.outputs.GHCR_REPO }}/clio-ci
             rippleci/clio_ci
-            ghcr.io/xrplf/clio-ci
           push_image: ${{ github.event_name != 'pull_request' }}
           directory: docker/ci
           tags: |
             type=raw,value=latest
-            type=raw,value=gcc_12_clang_16
+            type=raw,value=gcc_${{ env.GCC_MAJOR_VERSION }}_clang_${{ env.CLANG_MAJOR_VERSION }}
             type=raw,value=${{ github.sha }}
           platforms: linux/amd64,linux/arm64
+          build_args: |
+            GHCR_REPO=${{ needs.repo.outputs.GHCR_REPO }}
+            CLANG_MAJOR_VERSION=${{ env.CLANG_MAJOR_VERSION }}
+            GCC_MAJOR_VERSION=${{ env.GCC_MAJOR_VERSION }}
+            GCC_VERSION=${{ env.GCC_VERSION }}
           dockerhub_repo: rippleci/clio_ci
           dockerhub_description: CI image for XRPLF/clio.

.github/workflows/upload_conan_deps.yml

@@ -0,0 +1,100 @@
+name: Upload Conan Dependencies
+
+on:
+  schedule:
+    - cron: "0 9 * * 1-5"
+  workflow_dispatch:
+    inputs:
+      force_source_build:
+        description: "Force source build of all dependencies"
+        required: false
+        default: false
+        type: boolean
+  pull_request:
+    branches: [develop]
+    paths:
+      - .github/workflows/upload_conan_deps.yml
+
+      - .github/actions/generate/action.yml
+      - .github/actions/prepare_runner/action.yml
+      - ".github/scripts/conan/**"
+      - "!.github/scripts/conan/apple-clang-local.profile"
+
+      - conanfile.py
+      - conan.lock
+  push:
+    branches: [develop]
+    paths:
+      - .github/workflows/upload_conan_deps.yml
+
+      - .github/actions/generate/action.yml
+      - .github/actions/prepare_runner/action.yml
+      - ".github/scripts/conan/**"
+      - "!.github/scripts/conan/apple-clang-local.profile"
+
+      - conanfile.py
+      - conan.lock
+
+concurrency:
+  group: ${{ github.workflow }}-${{ github.ref }}
+  cancel-in-progress: true
+
+jobs:
+  generate-matrix:
+    runs-on: ubuntu-latest
+    outputs:
+      matrix: ${{ steps.set-matrix.outputs.matrix }}
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Calculate conan matrix
+        id: set-matrix
+        run: .github/scripts/conan/generate_matrix.py >> "${GITHUB_OUTPUT}"
+
+  upload-conan-deps:
+    name: Build ${{ matrix.compiler }}${{ matrix.sanitizer_ext }} ${{ matrix.build_type }}
+
+    needs: generate-matrix
+
+    strategy:
+      fail-fast: false
+      matrix: ${{ fromJson(needs.generate-matrix.outputs.matrix) }}
+
+    runs-on: ${{ matrix.os }}
+    container: ${{ matrix.container != '' && fromJson(matrix.container) || null }}
+
+    env:
+      CONAN_PROFILE: ${{ matrix.compiler }}${{ matrix.sanitizer_ext }}
+
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Prepare runner
+        uses: ./.github/actions/prepare_runner
+        with:
+          disable_ccache: true
+
+      - name: Setup conan on macOS
+        if: runner.os == 'macOS'
+        shell: bash
+        run: ./.github/scripts/conan/init.sh
+
+      - name: Show conan profile
+        run: conan profile show --profile:all ${{ env.CONAN_PROFILE }}
+
+      - name: Run conan and cmake
+        uses: ./.github/actions/generate
+        with:
+          conan_profile: ${{ env.CONAN_PROFILE }}
+          # We check that everything builds fine from source on scheduled runs
+          # But we do build and upload packages with build=missing by default
+          force_conan_source_build: ${{ github.event_name == 'schedule' || github.event.inputs.force_source_build == 'true' }}
+          build_type: ${{ matrix.build_type }}
+
+      - name: Login to Conan
+        if: github.event_name != 'pull_request'
+        run: conan remote login -p ${{ secrets.CONAN_PASSWORD }} ripple ${{ secrets.CONAN_USERNAME }}
+
+      - name: Upload Conan packages
+        if: github.event_name != 'pull_request' && github.event_name != 'schedule'
+        run: conan upload "*" -r=ripple --confirm

.pre-commit-config.yaml

@@ -11,7 +11,7 @@
 #
 # See https://pre-commit.com for more information
 # See https://pre-commit.com/hooks.html for more hooks
-exclude: ^docs/doxygen-awesome-theme/
+exclude: ^(docs/doxygen-awesome-theme/|conan\.lock$)
 
 repos:
   # `pre-commit sample-config` default hooks
@@ -26,12 +26,12 @@ repos:
 
   # Autoformat: YAML, JSON, Markdown, etc.
   - repo: https://github.com/rbubley/mirrors-prettier
-    rev: 787fb9f542b140ba0b2aced38e6a3e68021647a3 # frozen: v3.5.3
+    rev: 5ba47274f9b181bce26a5150a725577f3c336011 # frozen: v3.6.2
     hooks:
       - id: prettier
 
   - repo: https://github.com/igorshubovych/markdownlint-cli
-    rev: 586c3ea3f51230da42bab657c6a32e9e66c364f0 # frozen: v0.44.0
+    rev: 192ad822316c3a22fb3d3cc8aa6eafa0b8488360 # frozen: v0.45.0
     hooks:
       - id: markdownlint-fix
         exclude: LICENSE.md
@@ -55,17 +55,38 @@ repos:
             --ignore-words=pre-commit-hooks/codespell_ignore.txt,
           ]
 
-  # Running fix-local-includes before clang-format
-  # to ensure that the include order is correct.
+  - repo: https://github.com/trufflesecurity/trufflehog
+    rev: 6641d4ba5b684fffe195b9820345de1bf19f3181 # frozen: v3.89.2
+    hooks:
+      - id: trufflehog
+        entry: trufflehog git file://. --since-commit HEAD --no-verification --fail
+
+  # Running some C++ hooks before clang-format
+  # to ensure that the style is consistent.
   - repo: local
     hooks:
+      - id: json-in-cpp
+        name: Fix JSON style in C++
+        entry: pre-commit-hooks/json_in_cpp.py
+        types: [c++]
+        language: python
+        exclude: |
+          (?x)^(
+            tests/unit/etl/SubscriptionSourceTests.cpp|
+            tests/unit/web/ServerTests.cpp|
+            tests/unit/web/impl/ErrorHandlingTests.cpp|
+            tests/unit/web/ng/ServerTests.cpp|
+            tests/unit/web/ng/impl/ErrorHandlingTests.cpp
+          )$
+
       - id: fix-local-includes
         name: Fix Local Includes
         entry: pre-commit-hooks/fix-local-includes.sh
         types: [c++]
         language: script
+
   - repo: https://github.com/pre-commit/mirrors-clang-format
-    rev: f9a52e87b6cdcb01b0a62b8611d9ba9f2dad0067 # frozen: v19.1.7
+    rev: 6b9072cd80691b1b48d80046d884409fb1d962d1 # frozen: v20.1.7
     hooks:
       - id: clang-format
         args: [--style=file]

CMakeLists.txt

@@ -69,15 +69,17 @@ endif ()
 # Enable selected sanitizer if enabled via `san`
 if (san)
   set(SUPPORTED_SANITIZERS "address" "thread" "memory" "undefined")
-  list(FIND SUPPORTED_SANITIZERS "${san}" INDEX)
-  if (INDEX EQUAL -1)
+  if (NOT san IN_LIST SUPPORTED_SANITIZERS)
     message(FATAL_ERROR "Error: Unsupported sanitizer '${san}'. Supported values are: ${SUPPORTED_SANITIZERS}.")
   endif ()
 
-  target_compile_options(
-    clio_options INTERFACE # Sanitizers recommend minimum of -O1 for reasonable performance
-                           $<$<CONFIG:Debug>:-O1> ${SAN_FLAG} -fno-omit-frame-pointer
-  )
+  # Sanitizers recommend minimum of -O1 for reasonable performance so we enable it for debug builds
+  set(SAN_OPTIMIZATION_FLAG "")
+  if (CMAKE_BUILD_TYPE STREQUAL "Debug")
+    set(SAN_OPTIMIZATION_FLAG -O1)
+  endif ()
+  target_compile_options(clio_options INTERFACE ${SAN_OPTIMIZATION_FLAG} ${SAN_FLAG} -fno-omit-frame-pointer)
+
   target_compile_definitions(
     clio_options INTERFACE $<$<STREQUAL:${san},address>:SANITIZER=ASAN> $<$<STREQUAL:${san},thread>:SANITIZER=TSAN>
                            $<$<STREQUAL:${san},memory>:SANITIZER=MSAN> $<$<STREQUAL:${san},undefined>:SANITIZER=UBSAN>

cmake/ClangTidy.cmake

@@ -8,7 +8,7 @@ if (lint)
     endif ()
     message(STATUS "Using clang-tidy from CLIO_CLANG_TIDY_BIN")
   else ()
-    find_program(_CLANG_TIDY_BIN NAMES "clang-tidy-19" "clang-tidy" REQUIRED)
+    find_program(_CLANG_TIDY_BIN NAMES "clang-tidy-20" "clang-tidy" REQUIRED)
   endif ()
 
   if (NOT _CLANG_TIDY_BIN)

cmake/ClioVersion.cmake

@@ -4,39 +4,42 @@
 
 find_package(Git REQUIRED)
 
-set(GIT_COMMAND rev-parse --short HEAD)
+set(GIT_COMMAND describe --tags --exact-match)
 execute_process(
-  COMMAND ${GIT_EXECUTABLE} ${GIT_COMMAND} WORKING_DIRECTORY ${CMAKE_SOURCE_DIR} OUTPUT_VARIABLE REV
+  COMMAND ${GIT_EXECUTABLE} ${GIT_COMMAND}
+  WORKING_DIRECTORY ${CMAKE_SOURCE_DIR}
+  OUTPUT_VARIABLE TAG
+  RESULT_VARIABLE RC
   OUTPUT_STRIP_TRAILING_WHITESPACE
 )
 
-set(GIT_COMMAND branch --show-current)
-execute_process(
-  COMMAND ${GIT_EXECUTABLE} ${GIT_COMMAND} WORKING_DIRECTORY ${CMAKE_SOURCE_DIR} OUTPUT_VARIABLE BRANCH
-  OUTPUT_STRIP_TRAILING_WHITESPACE
-)
+if (RC EQUAL 0)
+  # if we are on a tag, use the tag name
+  set(CLIO_VERSION "${TAG}")
+  set(DOC_CLIO_VERSION "${TAG}")
+else ()
+  # if not, use YYYYMMDDHMS-<branch>-<git-rev>
 
-if (BRANCH STREQUAL "")
-  set(BRANCH "dev")
-endif ()
-
-if (NOT (BRANCH MATCHES master OR BRANCH MATCHES release/*)) # for develop and any other branch name
-                                                             # YYYYMMDDHMS-<branch>-<git-rev>
   set(GIT_COMMAND show -s --date=format:%Y%m%d%H%M%S --format=%cd)
   execute_process(
-    COMMAND ${GIT_EXECUTABLE} ${GIT_COMMAND} WORKING_DIRECTORY ${CMAKE_CURRENT_SOURCE_DIR} OUTPUT_VARIABLE DATE
-    OUTPUT_STRIP_TRAILING_WHITESPACE
+    COMMAND ${GIT_EXECUTABLE} ${GIT_COMMAND} WORKING_DIRECTORY ${CMAKE_SOURCE_DIR} OUTPUT_VARIABLE DATE
+    OUTPUT_STRIP_TRAILING_WHITESPACE COMMAND_ERROR_IS_FATAL ANY
   )
+
+  set(GIT_COMMAND branch --show-current)
+  execute_process(
+    COMMAND ${GIT_EXECUTABLE} ${GIT_COMMAND} WORKING_DIRECTORY ${CMAKE_SOURCE_DIR} OUTPUT_VARIABLE BRANCH
+    OUTPUT_STRIP_TRAILING_WHITESPACE COMMAND_ERROR_IS_FATAL ANY
+  )
+
+  set(GIT_COMMAND rev-parse --short HEAD)
+  execute_process(
+    COMMAND ${GIT_EXECUTABLE} ${GIT_COMMAND} WORKING_DIRECTORY ${CMAKE_SOURCE_DIR} OUTPUT_VARIABLE REV
+    OUTPUT_STRIP_TRAILING_WHITESPACE COMMAND_ERROR_IS_FATAL ANY
+  )
+
   set(CLIO_VERSION "${DATE}-${BRANCH}-${REV}")
   set(DOC_CLIO_VERSION "develop")
-else ()
-  set(GIT_COMMAND describe --tags)
-  execute_process(
-    COMMAND ${GIT_EXECUTABLE} ${GIT_COMMAND} WORKING_DIRECTORY ${CMAKE_SOURCE_DIR} OUTPUT_VARIABLE CLIO_TAG_VERSION
-    OUTPUT_STRIP_TRAILING_WHITESPACE
-  )
-  set(CLIO_VERSION "${CLIO_TAG_VERSION}")
-  set(DOC_CLIO_VERSION "${CLIO_TAG_VERSION}")
 endif ()
 
 if (CMAKE_BUILD_TYPE MATCHES Debug)

cmake/Settings.cmake

@@ -1,21 +1,22 @@
 set(COMPILER_FLAGS
+    -pedantic
     -Wall
     -Wcast-align
     -Wdouble-promotion
-    -Wextra
     -Werror
+    -Wextra
     -Wformat=2
     -Wimplicit-fallthrough
     -Wmisleading-indentation
-    -Wno-narrowing
-    -Wno-deprecated-declarations
     -Wno-dangling-else
+    -Wno-deprecated-declarations
+    -Wno-narrowing
     -Wno-unused-but-set-variable
     -Wnon-virtual-dtor
     -Wnull-dereference
     -Wold-style-cast
-    -pedantic
     -Wpedantic
+    -Wunreachable-code
     -Wunused
     # FIXME: The following bunch are needed for gcc12 atm.
     -Wno-missing-requires

conan.lock

@@ -0,0 +1,57 @@
+{
+    "version": "0.5",
+    "requires": [
+        "zlib/1.3.1#b8bc2603263cf7eccbd6e17e66b0ed76%1752006674.465",
+        "xxhash/0.8.2#7856c968c985b2981b707ee8f2413b2b%1752006674.334",
+        "xrpl/2.5.0#7880d1696f11fceb1d498570f1a184c8%1752006708.218",
+        "sqlite3/3.47.0#7a0904fd061f5f8a2366c294f9387830%1752006674.338",
+        "soci/4.0.3#a9f8d773cd33e356b5879a4b0564f287%1752006674.465",
+        "re2/20230301#dfd6e2bf050eb90ddd8729cfb4c844a4%1752006674.077",
+        "rapidjson/cci.20220822#1b9d8c2256876a154172dc5cfbe447c6%1752006673.227",
+        "protobuf/3.21.12#d927114e28de9f4691a6bbcdd9a529d1%1752006673.172",
+        "openssl/1.1.1v#216374e4fb5b2e0f5ab1fb6f27b5b434%1752006673.069",
+        "nudb/2.0.8#63990d3e517038e04bf529eb8167f69f%1752006673.862",
+        "minizip/1.2.13#9e87d57804bd372d6d1e32b1871517a3%1752006672.983",
+        "lz4/1.10.0#59fc63cac7f10fbe8e05c7e62c2f3504%1752006672.825",
+        "libuv/1.46.0#78565d142ac7102776256328a26cdf60%1752006672.827",
+        "libiconv/1.17#1e65319e945f2d31941a9d28cc13c058%1752006672.826",
+        "libbacktrace/cci.20210118#a7691bfccd8caaf66309df196790a5a1%1752006672.822",
+        "libarchive/3.7.6#e0453864b2a4d225f06b3304903cb2b7%1752006672.917",
+        "http_parser/2.9.4#98d91690d6fd021e9e624218a85d9d97%1752006672.658",
+        "gtest/1.14.0#f8f0757a574a8dd747d16af62d6eb1b7%1752006671.555",
+        "grpc/1.50.1#02291451d1e17200293a409410d1c4e1%1752006671.777",
+        "fmt/11.2.0#579bb2cdf4a7607621beea4eb4651e0f%1752006671.557",
+        "date/3.0.3#cf28fe9c0aab99fe12da08aa42df65e1%1752006671.553",
+        "cassandra-cpp-driver/2.17.0#e50919efac8418c26be6671fd702540a%1752006671.654",
+        "c-ares/1.34.5#b78b91e7cfb1f11ce777a285bbf169c6%1752006671.554",
+        "bzip2/1.0.8#00b4a4658791c1f06914e087f0e792f5%1752006671.549",
+        "boost/1.83.0#5bcb2a14a35875e328bf312e080d3562%1752006671.557",
+        "benchmark/1.8.3#1a2ce62c99e2b3feaa57b1f0c15a8c46%1752006671.408",
+        "abseil/20230802.1#f0f91485b111dc9837a68972cb19ca7b%1752006671.555"
+    ],
+    "build_requires": [
+        "zlib/1.3.1#b8bc2603263cf7eccbd6e17e66b0ed76%1752006674.465",
+        "protobuf/3.21.12#d927114e28de9f4691a6bbcdd9a529d1%1752006673.172",
+        "protobuf/3.21.9#64ce20e1d9ea24f3d6c504015d5f6fa8%1752006673.173",
+        "cmake/3.31.7#57c3e118bcf267552c0ea3f8bee1e7d5%1752006671.64",
+        "b2/5.3.2#7b5fabfe7088ae933fb3e78302343ea0%1752006671.407"
+    ],
+    "python_requires": [],
+    "overrides": {
+        "boost/1.83.0": [
+            null,
+            "boost/1.83.0#5bcb2a14a35875e328bf312e080d3562"
+        ],
+        "protobuf/3.21.9": [
+            null,
+            "protobuf/3.21.12"
+        ],
+        "lz4/1.9.4": [
+            "lz4/1.10.0"
+        ],
+        "sqlite3/3.44.2": [
+            "sqlite3/3.47.0"
+        ]
+    },
+    "config_requires": []
+}

conanfile.py

@@ -11,7 +11,6 @@ class ClioConan(ConanFile):
     settings = 'os', 'compiler', 'build_type', 'arch'
     options = {
         'static': [True, False],              # static linkage
-        'fPIC': [True, False],                # unused?
         'verbose': [True, False],
         'tests': [True, False],               # build unit tests; create `clio_tests` binary
         'integration_tests': [True, False],   # build integration tests; create `clio_integration_tests` binary
@@ -27,18 +26,17 @@ class ClioConan(ConanFile):
     requires = [
         'boost/1.83.0',
         'cassandra-cpp-driver/2.17.0',
-        'fmt/10.1.1',
+        'fmt/11.2.0',
         'protobuf/3.21.12',
         'grpc/1.50.1',
         'openssl/1.1.1v',
-        'xrpl/2.5.0-rc1',
+        'xrpl/2.5.0',
         'zlib/1.3.1',
         'libbacktrace/cci.20210118'
     ]
 
     default_options = {
         'static': False,
-        'fPIC': True,
         'verbose': False,
         'tests': False,
         'integration_tests': False,
@@ -89,21 +87,8 @@ class ClioConan(ConanFile):
 
     def generate(self):
         tc = CMakeToolchain(self)
-        tc.variables['verbose'] = self.options.verbose
-        tc.variables['static'] = self.options.static
-        tc.variables['tests'] = self.options.tests
-        tc.variables['integration_tests'] = self.options.integration_tests
-        tc.variables['coverage'] = self.options.coverage
-        tc.variables['lint'] = self.options.lint
-        tc.variables['docs'] = self.options.docs
-        tc.variables['packaging'] = self.options.packaging
-        tc.variables['benchmark'] = self.options.benchmark
-        tc.variables['snapshot'] = self.options.snapshot
-        tc.variables['time_trace'] = self.options.time_trace
-
-        if self.settings.compiler == 'clang' and self.settings.compiler.version == 16:
-            tc.extra_cxxflags = ["-DBOOST_ASIO_DISABLE_CONCEPTS"]
-
+        for option_name, option_value in self.options.items():
+            tc.variables[option_name] = option_value
         tc.generate()
 
     def build(self):

docker/ci/Dockerfile

@@ -1,8 +1,11 @@
-# TODO: change this when we are able to push gcc image to ghcr.io
-FROM rippleci/clio_gcc:12.3.0 AS clio-gcc
-FROM ghcr.io/xrplf/clio-tools:latest AS clio-tools
+ARG GHCR_REPO=invalid
+ARG CLANG_MAJOR_VERSION=invalid
+ARG GCC_VERSION=invalid
 
-FROM ghcr.io/xrplf/clio-clang:16
+FROM ${GHCR_REPO}/clio-gcc:${GCC_VERSION} AS clio-gcc
+FROM ${GHCR_REPO}/clio-tools:latest AS clio-tools
+
+FROM ${GHCR_REPO}/clio-clang:${CLANG_MAJOR_VERSION}
 
 ARG DEBIAN_FRONTEND=noninteractive
 
@@ -17,7 +20,7 @@ SHELL ["/bin/bash", "-o", "pipefail", "-c"]
 USER root
 WORKDIR /root
 
-ARG LLVM_TOOLS_VERSION=19
+ARG LLVM_TOOLS_VERSION=20
 
 # Add repositories
 RUN apt-get update \
@@ -33,10 +36,8 @@ RUN apt-get update \
 # Install packages
 RUN apt-get update \
     && apt-get install -y --no-install-recommends --no-install-suggests \
-        bison \
         clang-tidy-${LLVM_TOOLS_VERSION} \
         clang-tools-${LLVM_TOOLS_VERSION} \
-        flex \
         git \
         git-lfs \
         graphviz \
@@ -48,6 +49,10 @@ RUN apt-get update \
         zip \
     && pip3 install -q --upgrade --no-cache-dir pip \
     && pip3 install -q --no-cache-dir \
+        # TODO: Remove this once we switch to newer Ubuntu base image
+        # lxml 6.0.0 is not compatible with our image
+        'lxml<6.0.0' \
+        \
         cmake==3.31.6 \
         conan==2.17.0 \
         gcovr \
@@ -55,29 +60,33 @@ RUN apt-get update \
     && apt-get clean \
     && rm -rf /var/lib/apt/lists/*
 
-# Install gcc-12 and make ldconfig aware of the new libstdc++ location (for gcc)
+ARG GCC_MAJOR_VERSION=invalid
+
+# Install custom-built gcc and make ldconfig aware of the new libstdc++ location (for gcc)
 # Note: Clang is using libc++ instead
-COPY --from=clio-gcc /gcc12.deb /
+COPY --from=clio-gcc /gcc${GCC_MAJOR_VERSION}.deb /
 RUN apt-get update \
     && apt-get install -y --no-install-recommends --no-install-suggests \
         binutils \
         libc6-dev \
     && apt-get clean \
     && rm -rf /var/lib/apt/lists/* \
-    && dpkg -i /gcc12.deb \
-    && rm -rf /gcc12.deb \
+    && dpkg -i /gcc${GCC_MAJOR_VERSION}.deb \
+    && rm -rf /gcc${GCC_MAJOR_VERSION}.deb \
     && ldconfig
 
-# Rewire to use gcc-12 as default compiler
-RUN update-alternatives --install /usr/bin/g++ g++ /usr/bin/g++-12 100 \
-    && update-alternatives --install /usr/bin/c++ c++ /usr/bin/g++-12 100 \
-    && update-alternatives --install /usr/bin/gcc gcc /usr/bin/gcc-12 100 \
-    && update-alternatives --install /usr/bin/cc cc /usr/bin/gcc-12 100 \
-    && update-alternatives --install /usr/bin/gcov gcov /usr/bin/gcov-12 100 \
-    && update-alternatives --install /usr/bin/gcov-dump gcov-dump /usr/bin/gcov-dump-12 100 \
-    && update-alternatives --install /usr/bin/gcov-tool gcov-tool /usr/bin/gcov-tool-12 100
+# Rewire to use our custom-built gcc as default compiler
+RUN update-alternatives --install /usr/bin/g++ g++ /usr/bin/g++-${GCC_MAJOR_VERSION} 100 \
+    && update-alternatives --install /usr/bin/c++ c++ /usr/bin/g++-${GCC_MAJOR_VERSION} 100 \
+    && update-alternatives --install /usr/bin/gcc gcc /usr/bin/gcc-${GCC_MAJOR_VERSION} 100 \
+    && update-alternatives --install /usr/bin/cc cc /usr/bin/gcc-${GCC_MAJOR_VERSION} 100 \
+    && update-alternatives --install /usr/bin/gcov gcov /usr/bin/gcov-${GCC_MAJOR_VERSION} 100 \
+    && update-alternatives --install /usr/bin/gcov-dump gcov-dump /usr/bin/gcov-dump-${GCC_MAJOR_VERSION} 100 \
+    && update-alternatives --install /usr/bin/gcov-tool gcov-tool /usr/bin/gcov-tool-${GCC_MAJOR_VERSION} 100
 
 COPY --from=clio-tools \
+    /usr/local/bin/mold \
+    /usr/local/bin/ld.mold \
     /usr/local/bin/ccache \
     /usr/local/bin/doxygen \
     /usr/local/bin/ClangBuildAnalyzer \
@@ -90,6 +99,9 @@ WORKDIR /root
 # Setup conan
 RUN conan remote add --index 0 ripple http://18.143.149.228:8081/artifactory/api/conan/dev
 
+WORKDIR /root/.conan2
+COPY conan/global.conf ./global.conf
+
 WORKDIR /root/.conan2/profiles
 
 COPY conan/clang.profile ./clang

docker/ci/README.md

@@ -6,13 +6,14 @@ It is used in [Clio Github Actions](https://github.com/XRPLF/clio/actions) but c
 The image is based on Ubuntu 20.04 and contains:
 
 - ccache 4.11.3
-- clang 16.0.6
+- Clang 19
 - ClangBuildAnalyzer 1.6.0
-- conan 2.17.0
-- doxygen 1.12
-- gcc 12.3.0
+- Conan 2.17.0
+- Doxygen 1.12
+- GCC 14.3.0
 - gh 2.74
 - git-cliff 2.9.1
+- mold 2.40.1
 - and some other useful tools
 
 Conan is set up to build Clio without any additional steps.

docker/ci/conan/clang.profile

@@ -1,11 +1,12 @@
 [settings]
-arch=x86_64
+arch={{detect_api.detect_arch()}}
 build_type=Release
 compiler=clang
 compiler.cppstd=20
 compiler.libcxx=libc++
-compiler.version=16
+compiler.version=19
 os=Linux
 
 [conf]
-tools.build:compiler_executables={'c': '/usr/bin/clang-16', 'cpp': '/usr/bin/clang++-16'}
+tools.build:compiler_executables={"c": "/usr/bin/clang-19", "cpp": "/usr/bin/clang++-19"}
+grpc/1.50.1:tools.build:cxxflags+=["-Wno-missing-template-arg-list-after-template-kw"]

docker/ci/conan/gcc.profile

@@ -1,11 +1,11 @@
 [settings]
-arch=x86_64
+arch={{detect_api.detect_arch()}}
 build_type=Release
 compiler=gcc
 compiler.cppstd=20
 compiler.libcxx=libstdc++11
-compiler.version=12
+compiler.version=14
 os=Linux
 
 [conf]
-tools.build:compiler_executables={'c': '/usr/bin/gcc-12', 'cpp': '/usr/bin/g++-12'}
+tools.build:compiler_executables={"c": "/usr/bin/gcc-14", "cpp": "/usr/bin/g++-14"}

docker/ci/conan/global.conf

@@ -0,0 +1,2 @@
+core.download:parallel={{os.cpu_count()}}
+core.upload:parallel={{os.cpu_count()}}

docker/ci/conan/sanitizer_template.profile

@@ -1,15 +1,23 @@
 {% set compiler, sani = profile_name.split('.') %}
 
-{% set sanitizer_opt_map = {'asan': 'address', 'tsan': 'thread', 'ubsan': 'undefined'} %}
+{% set sanitizer_opt_map = {"asan": "address", "tsan": "thread", "ubsan": "undefined"} %}
 {% set sanitizer = sanitizer_opt_map[sani] %}
 
+{% set sanitizer_build_flags_str = "-fsanitize=" ~ sanitizer ~ " -g -O1 -fno-omit-frame-pointer" %}
+{% set sanitizer_build_flags = sanitizer_build_flags_str.split(' ') %}
+{% set sanitizer_link_flags_str = "-fsanitize=" ~ sanitizer %}
+{% set sanitizer_link_flags = sanitizer_link_flags_str.split(' ') %}
+
 include({{ compiler }})
 
 [options]
-boost/*:extra_b2_flags="cxxflags=\"-fsanitize={{ sanitizer }}\" linkflags=\"-fsanitize={{ sanitizer }}\""
+boost/*:extra_b2_flags="cxxflags=\"{{ sanitizer_build_flags_str }}\" linkflags=\"{{ sanitizer_link_flags_str }}\""
 boost/*:without_stacktrace=True
 
 [conf]
-tools.build:cflags+=["-fsanitize={{ sanitizer }}"]
-tools.build:cxxflags+=["-fsanitize={{ sanitizer }}"]
-tools.build:exelinkflags+=["-fsanitize={{ sanitizer }}"]
+tools.build:cflags+={{ sanitizer_build_flags }}
+tools.build:cxxflags+={{ sanitizer_build_flags }}
+tools.build:exelinkflags+={{ sanitizer_link_flags }}
+tools.build:sharedlinkflags+={{ sanitizer_link_flags }}
+
+tools.info.package_id:confs+=["tools.build:cflags", "tools.build:cxxflags", "tools.build:exelinkflags", "tools.build:sharedlinkflags"]

docker/compilers/clang/Dockerfile

@@ -8,8 +8,6 @@ SHELL ["/bin/bash", "-c"]
 USER root
 WORKDIR /root
 
-ARG CLANG_VERSION=16
-
 RUN apt-get update \
     && apt-get install -y --no-install-recommends --no-install-suggests \
         wget \
@@ -18,13 +16,17 @@ RUN apt-get update \
     && apt-get clean \
     && rm -rf /var/lib/apt/lists/*
 
+ARG CLANG_MAJOR_VERSION=invalid
+# Bump this version to force rebuild of the image
+ARG BUILD_VERSION=0
+
 RUN wget --progress=dot:giga https://apt.llvm.org/llvm.sh \
     && chmod +x llvm.sh \
-    && ./llvm.sh ${CLANG_VERSION} \
+    && ./llvm.sh ${CLANG_MAJOR_VERSION} \
     && rm -rf llvm.sh \
     && apt-get update \
     && apt-get install -y --no-install-recommends --no-install-suggests \
-        libc++-${CLANG_VERSION}-dev \
-        libc++abi-${CLANG_VERSION}-dev \
+        libc++-${CLANG_MAJOR_VERSION}-dev \
+        libc++abi-${CLANG_MAJOR_VERSION}-dev \
     && apt-get clean \
     && rm -rf /var/lib/apt/lists/*

docker/compilers/gcc/Dockerfile

@@ -1,10 +1,17 @@
-FROM ubuntu:20.04 AS build
+ARG UBUNTU_VERSION=20.04
+
+ARG GCC_MAJOR_VERSION=invalid
+
+FROM ubuntu:$UBUNTU_VERSION AS build
+
+ARG UBUNTU_VERSION
+
+ARG GCC_MAJOR_VERSION
+
+ARG BUILD_VERSION=0
 
 ARG DEBIAN_FRONTEND=noninteractive
 ARG TARGETARCH
-ARG UBUNTU_VERSION=20.04
-ARG GCC_VERSION=12.3.0
-ARG BUILD_VERSION=2
 
 RUN apt-get update \
     && apt-get install -y --no-install-recommends --no-install-suggests \
@@ -18,18 +25,23 @@ RUN apt-get update \
     && apt-get clean \
     && rm -rf /var/lib/apt/lists/*
 
-RUN wget --progress=dot:giga https://gcc.gnu.org/pub/gcc/releases/gcc-$GCC_VERSION/gcc-$GCC_VERSION.tar.gz \
-    && tar xf gcc-$GCC_VERSION.tar.gz \
-    && cd /gcc-$GCC_VERSION && ./contrib/download_prerequisites
+ARG GCC_VERSION
 
-RUN mkdir /${TARGETARCH}-gcc-12
-WORKDIR /${TARGETARCH}-gcc-12
+WORKDIR /
+RUN wget --progress=dot:giga https://gcc.gnu.org/pub/gcc/releases/gcc-$GCC_VERSION/gcc-$GCC_VERSION.tar.gz \
+    && tar xf gcc-$GCC_VERSION.tar.gz
+
+WORKDIR /gcc-$GCC_VERSION
+RUN ./contrib/download_prerequisites
+
+RUN mkdir /gcc-build
+WORKDIR /gcc-build
 RUN /gcc-$GCC_VERSION/configure \
         --with-pkgversion="clio-build-$BUILD_VERSION https://github.com/XRPLF/clio" \
         --enable-languages=c,c++ \
         --prefix=/usr \
         --with-gcc-major-version-only \
-        --program-suffix=-12 \
+        --program-suffix=-${GCC_MAJOR_VERSION} \
         --enable-shared \
         --enable-linker-build-id \
         --libexecdir=/usr/lib \
@@ -53,38 +65,55 @@ RUN /gcc-$GCC_VERSION/configure \
         --enable-cet \
         --disable-multilib \
         --without-cuda-driver \
-        --enable-checking=release \
-    && make -j "$(nproc)" \
-    && make install-strip DESTDIR=/gcc-$GCC_VERSION-$BUILD_VERSION-ubuntu-$UBUNTU_VERSION \
-    && mkdir -p /gcc-$GCC_VERSION-$BUILD_VERSION-ubuntu-$UBUNTU_VERSION/usr/share/gdb/auto-load/usr/lib64 \
-    && mv /gcc-$GCC_VERSION-$BUILD_VERSION-ubuntu-$UBUNTU_VERSION/usr/lib64/libstdc++.so.6.0.30-gdb.py /gcc-$GCC_VERSION-$BUILD_VERSION-ubuntu-$UBUNTU_VERSION/usr/share/gdb/auto-load/usr/lib64/libstdc++.so.6.0.30-gdb.py
+        --enable-checking=release
+
+RUN make -j "$(nproc)"
+RUN make install-strip DESTDIR=/gcc-$GCC_VERSION-$BUILD_VERSION-ubuntu-$UBUNTU_VERSION
+
+RUN export GDB_AUTOLOAD_DIR="/gcc-$GCC_VERSION-$BUILD_VERSION-ubuntu-$UBUNTU_VERSION/usr/share/gdb/auto-load/usr/lib64" \
+    && mkdir -p "$GDB_AUTOLOAD_DIR" \
+    && mv \
+        /gcc-$GCC_VERSION-$BUILD_VERSION-ubuntu-$UBUNTU_VERSION/usr/lib64/libstdc++.so.*-gdb.py \
+        $GDB_AUTOLOAD_DIR/
 
 # Generate deb
 WORKDIR /
 COPY control.m4 /
-COPY ld.so.conf /gcc-$GCC_VERSION-$BUILD_VERSION-ubuntu-$UBUNTU_VERSION/etc/ld.so.conf.d/1-gcc-12.conf
+COPY ld.so.conf /gcc-$GCC_VERSION-$BUILD_VERSION-ubuntu-$UBUNTU_VERSION/etc/ld.so.conf.d/1-gcc-${GCC_MAJOR_VERSION}.conf
 
 RUN mkdir /gcc-$GCC_VERSION-$BUILD_VERSION-ubuntu-$UBUNTU_VERSION/DEBIAN \
-    && m4 -P -DUBUNTU_VERSION=$UBUNTU_VERSION -DVERSION=$GCC_VERSION-$BUILD_VERSION -DTARGETARCH=$TARGETARCH control.m4 > /gcc-$GCC_VERSION-$BUILD_VERSION-ubuntu-$UBUNTU_VERSION/DEBIAN/control \
-    && dpkg-deb --build --root-owner-group /gcc-$GCC_VERSION-$BUILD_VERSION-ubuntu-$UBUNTU_VERSION /gcc12.deb
+    && m4 \
+        -P \
+        -DUBUNTU_VERSION=$UBUNTU_VERSION \
+        -DVERSION=$GCC_VERSION-$BUILD_VERSION \
+        -DTARGETARCH=$TARGETARCH \
+        control.m4 > /gcc-$GCC_VERSION-$BUILD_VERSION-ubuntu-$UBUNTU_VERSION/DEBIAN/control \
+    && dpkg-deb \
+        --build \
+        --root-owner-group \
+        /gcc-$GCC_VERSION-$BUILD_VERSION-ubuntu-$UBUNTU_VERSION \
+        /gcc${GCC_MAJOR_VERSION}.deb
 
 # Create final image
-FROM ubuntu:20.04
-COPY --from=build /gcc12.deb /
+FROM ubuntu:$UBUNTU_VERSION
 
-# Make gcc-12 available but also leave gcc12.deb for others to copy if needed
+ARG GCC_MAJOR_VERSION
+
+COPY --from=build /gcc${GCC_MAJOR_VERSION}.deb /
+
+# Install gcc-${GCC_MAJOR_VERSION}, but also leave gcc${GCC_MAJOR_VERSION}.deb for others to copy if needed
 RUN apt-get update \
     && apt-get install -y --no-install-recommends --no-install-suggests \
         binutils \
         libc6-dev \
     && apt-get clean \
     && rm -rf /var/lib/apt/lists/* \
-    && dpkg -i /gcc12.deb
+    && dpkg -i /gcc${GCC_MAJOR_VERSION}.deb
 
-RUN update-alternatives --install /usr/bin/g++ g++ /usr/bin/g++-12 100 \
-    && update-alternatives --install /usr/bin/c++ c++ /usr/bin/g++-12 100 \
-    && update-alternatives --install /usr/bin/gcc gcc /usr/bin/gcc-12 100 \
-    && update-alternatives --install /usr/bin/cc cc /usr/bin/gcc-12 100 \
-    && update-alternatives --install /usr/bin/gcov gcov /usr/bin/gcov-12 100 \
-    && update-alternatives --install /usr/bin/gcov-dump gcov-dump /usr/bin/gcov-dump-12 100 \
-    && update-alternatives --install /usr/bin/gcov-tool gcov-tool /usr/bin/gcov-tool-12 100
+RUN update-alternatives --install /usr/bin/g++ g++ /usr/bin/g++-${GCC_MAJOR_VERSION} 100 \
+    && update-alternatives --install /usr/bin/c++ c++ /usr/bin/g++-${GCC_MAJOR_VERSION} 100 \
+    && update-alternatives --install /usr/bin/gcc gcc /usr/bin/gcc-${GCC_MAJOR_VERSION} 100 \
+    && update-alternatives --install /usr/bin/cc cc /usr/bin/gcc-${GCC_MAJOR_VERSION} 100 \
+    && update-alternatives --install /usr/bin/gcov gcov /usr/bin/gcov-${GCC_MAJOR_VERSION} 100 \
+    && update-alternatives --install /usr/bin/gcov-dump gcov-dump /usr/bin/gcov-dump-${GCC_MAJOR_VERSION} 100 \
+    && update-alternatives --install /usr/bin/gcov-tool gcov-tool /usr/bin/gcov-tool-${GCC_MAJOR_VERSION} 100

docker/compilers/gcc/README.md

@@ -1,3 +1,3 @@
-# gcc compiler
+# GCC compiler
 
-This image contains gcc compiler to build <https://github.com/XRPLF/clio>.
+This image contains GCC compiler to build <https://github.com/XRPLF/clio>.

docker/compilers/gcc/control.m4

@@ -1,6 +1,7 @@
-Package: gcc-12-ubuntu-UBUNTUVERSION
+Package: gcc-14-ubuntu-UBUNTUVERSION
 Version: VERSION
 Architecture: TARGETARCH
 Maintainer: Alex Kremer <akremer@ripple.com>
-Description: Gcc VERSION build for ubuntu UBUNTUVERSION
+Uploaders: Ayaz Salikhov <asalikhov@ripple.com>
+Description: GCC VERSION build for ubuntu UBUNTUVERSION
 Depends: binutils, libc6-dev

docker/develop/compose.yaml

@@ -2,7 +2,7 @@ services:
   clio_develop:
     image: ghcr.io/xrplf/clio-ci:latest
     volumes:
-      - clio_develop_conan_data:/root/.conan/data
+      - clio_develop_conan_data:/root/.conan2/p
       - clio_develop_ccache:/root/.ccache
       - ../../:/root/clio
       - clio_develop_build:/root/clio/build_docker

docker/tools/Dockerfile

@@ -1,24 +1,41 @@
-FROM ubuntu:20.04
+ARG GHCR_REPO=invalid
+ARG GCC_VERSION=invalid
+
+FROM ${GHCR_REPO}/clio-gcc:${GCC_VERSION}
 
 ARG DEBIAN_FRONTEND=noninteractive
 ARG TARGETARCH
 
 SHELL ["/bin/bash", "-o", "pipefail", "-c"]
 
+ARG BUILD_VERSION=1
+
 RUN apt-get update \
     && apt-get install -y --no-install-recommends --no-install-suggests \
         bison \
-        build-essential \
-        cmake \
         flex \
         ninja-build \
+        python3 \
+        python3-pip \
         software-properties-common \
         wget \
+    && pip3 install -q --no-cache-dir \
+        cmake==3.31.6 \
     && apt-get clean \
     && rm -rf /var/lib/apt/lists/*
 
 WORKDIR /tmp
 
+ARG MOLD_VERSION=2.40.1
+RUN wget --progress=dot:giga "https://github.com/rui314/mold/archive/refs/tags/v${MOLD_VERSION}.tar.gz" \
+    && tar xf "v${MOLD_VERSION}.tar.gz" \
+    && cd "mold-${MOLD_VERSION}" \
+    && mkdir build \
+    && cd build \
+    && cmake -GNinja -DCMAKE_BUILD_TYPE=Release .. \
+    && ninja install \
+    && rm -rf /tmp/* /var/tmp/*
+
 ARG CCACHE_VERSION=4.11.3
 RUN wget --progress=dot:giga "https://github.com/ccache/ccache/releases/download/v${CCACHE_VERSION}/ccache-${CCACHE_VERSION}.tar.gz" \
     && tar xf "ccache-${CCACHE_VERSION}.tar.gz" \
@@ -26,7 +43,7 @@ RUN wget --progress=dot:giga "https://github.com/ccache/ccache/releases/download
     && mkdir build \
     && cd build \
     && cmake -GNinja -DCMAKE_BUILD_TYPE=Release -DENABLE_TESTING=False .. \
-    && cmake --build . --target install \
+    && ninja install \
     && rm -rf /tmp/* /var/tmp/*
 
 ARG DOXYGEN_VERSION=1.12.0
@@ -36,7 +53,7 @@ RUN wget --progress=dot:giga "https://github.com/doxygen/doxygen/releases/downlo
     && mkdir build \
     && cd build \
     && cmake -GNinja -DCMAKE_BUILD_TYPE=Release .. \
-    && cmake --build . --target install \
+    && ninja install \
     && rm -rf /tmp/* /var/tmp/*
 
 ARG CLANG_BUILD_ANALYZER_VERSION=1.6.0
@@ -46,7 +63,7 @@ RUN wget --progress=dot:giga "https://github.com/aras-p/ClangBuildAnalyzer/archi
     && mkdir build \
     && cd build \
     && cmake -GNinja -DCMAKE_BUILD_TYPE=Release .. \
-    && cmake --build . --target install \
+    && ninja install \
     && rm -rf /tmp/* /var/tmp/*
 
 ARG GIT_CLIFF_VERSION=2.9.1

docs/build-clio.md

@@ -6,7 +6,7 @@
 ## Minimum Requirements
 
 - [Python 3.7](https://www.python.org/downloads/)
-- [Conan 1.55, <2.0](https://conan.io/downloads.html)
+- [Conan 2.17.0](https://conan.io/downloads.html)
 - [CMake 3.20, <4.0](https://cmake.org/download/)
 - [**Optional**] [GCovr](https://gcc.gnu.org/onlinedocs/gcc/Gcov.html): needed for code coverage generation
 - [**Optional**] [CCache](https://ccache.dev/): speeds up compilation if you are going to compile Clio often
@@ -19,32 +19,43 @@
 
 ### Conan Configuration
 
-Clio requires `compiler.cppstd=20` in your Conan profile (`~/.conan/profiles/default`).
+By default, Conan uses `~/.conan2` as it's home folder.
+You can change it by using `$CONAN_HOME` env variable.
+[More info about Conan home](https://docs.conan.io/2/reference/environment.html#conan-home).
 
-> [!NOTE]
-> Although Clio is built using C++23, it's required to set `compiler.cppstd=20` for the time being as some of Clio's dependencies are not yet capable of building under C++23.
+> [!TIP]
+> To setup Conan automatically, you can run `.github/scripts/conan/init.sh`.
+> This will delete Conan home directory (if it exists), set up profiles and add Artifactory remote.
 
-**Mac apple-clang 16 example**:
+The instruction below assumes that `$CONAN_HOME` is not set.
+
+#### Profiles
+
+The default profile is the file in `~/.conan2/profiles/default`.
+
+Here are some examples of possible profiles:
+
+**Mac apple-clang 17 example**:
 
 ```text
 [settings]
-arch=armv8
+arch={{detect_api.detect_arch()}}
 build_type=Release
 compiler=apple-clang
 compiler.cppstd=20
 compiler.libcxx=libc++
-compiler.version=16
+compiler.version=17
 os=Macos
 
 [conf]
-tools.build:cxxflags+=["-Wno-missing-template-arg-list-after-template-kw"]
+grpc/1.50.1:tools.build:cxxflags+=["-Wno-missing-template-arg-list-after-template-kw"]
 ```
 
 **Linux gcc-12 example**:
 
 ```text
 [settings]
-arch=x86_64
+arch={{detect_api.detect_arch()}}
 build_type=Release
 compiler=gcc
 compiler.cppstd=20
@@ -53,7 +64,19 @@ compiler.version=12
 os=Linux
 
 [conf]
-tools.build:compiler_executables={'c': '/usr/bin/gcc-12', 'cpp': '/usr/bin/g++-12'}
+tools.build:compiler_executables={"c": "/usr/bin/gcc-12", "cpp": "/usr/bin/g++-12"}
+```
+
+> [!NOTE]
+> Although Clio is built using C++23, it's required to set `compiler.cppstd=20` in your profile for the time being as some of Clio's dependencies are not yet capable of building under C++23.
+
+#### global.conf file
+
+To increase the speed of downloading and uploading packages, add the following to the `~/.conan2/global.conf` file:
+
+```text
+core.download:parallel={{os.cpu_count()}}
+core.upload:parallel={{os.cpu_count()}}
 ```
 
 #### Artifactory
@@ -64,15 +87,21 @@ Make sure artifactory is setup with Conan.
 conan remote add --index 0 ripple http://18.143.149.228:8081/artifactory/api/conan/dev
 ```
 
-Now you should be able to download the prebuilt `xrpl` package on some platforms.
+Now you should be able to download the prebuilt dependencies (including `xrpl` package) on supported platforms.
 
-> [!NOTE]
-> You may need to edit the `~/.conan/remotes.json` file to ensure that this newly added artifactory is listed last. Otherwise, you could see compilation errors when building the project with gcc version 13 (or newer).
+#### Conan lockfile
 
-Remove old packages you may have cached.
+To achieve reproducible dependencies, we use [Conan lockfile](https://docs.conan.io/2/tutorial/versioning/lockfiles.html).
 
-```sh
-conan remove -f xrpl
+The `conan.lock` file in the repository contains a "snapshot" of the current dependencies.
+It is implicitly used when running `conan` commands, you don't need to specify it.
+
+You have to update this file every time you add a new dependency or change a revision or version of an existing dependency.
+
+To do that, run the following command in the repository root:
+
+```bash
+conan lock create . -o '&:tests=True' -o '&:benchmark=True'
 ```
 
 ## Building Clio
@@ -81,6 +110,7 @@ Navigate to Clio's root directory and run:
 
 ```sh
 mkdir build && cd build
+# You can also specify profile explicitly by adding `--profile:all <PROFILE_NAME>`
 conan install .. --output-folder . --build missing --settings build_type=Release -o '&:tests=True'
 # You can also add -GNinja to use Ninja build system instead of Make
 cmake -DCMAKE_TOOLCHAIN_FILE:FILEPATH=build/generators/conan_toolchain.cmake -DCMAKE_BUILD_TYPE=Release ..
@@ -152,24 +182,24 @@ If you wish to develop against a `rippled` instance running in standalone mode t
 
 Sometimes, during development, you need to build against a custom version of `libxrpl`. (For example, you may be developing compatibility for a proposed amendment that is not yet merged to the main `rippled` codebase.) To build Clio with compatibility for a custom fork or branch of `rippled`, follow these steps:
 
-1. First, pull/clone the appropriate `rippled` fork and switch to the branch you want to build.
-   The following example uses an in-development build with [XLS-33d Multi-Purpose Tokens](https://github.com/XRPLF/XRPL-Standards/tree/master/XLS-0033d-multi-purpose-tokens):
+1. First, pull/clone the appropriate `rippled` version and switch to the branch you want to build.
+   The following example uses a `2.5.0-rc1` tag of rippled in the main branch:
 
    ```sh
-   git clone https://github.com/shawnxie999/rippled/
+   git clone https://github.com/XRPLF/rippled/
    cd rippled
-   git switch mpt-1.1
+   git checkout 2.5.0-rc1
    ```
 
 2. Export a custom package to your local Conan store using a user/channel:
 
    ```sh
-   conan export . my/feature
+   conan export . --user=my --channel=feature
    ```
 
 3. Patch your local Clio build to use the right package.
 
-   Edit `conanfile.py` (from the Clio repository root). Replace the `xrpl` requirement with the custom package version from the previous step. This must also include the current version number from your `rippled` branch. For example:
+   Edit `conanfile.py` in the Clio repository root. Replace the `xrpl` requirement with the custom package version from the previous step. This must also include the current version number from your `rippled` branch. For example:
 
    ```py
    # ... (excerpt from conanfile.py)
@@ -180,7 +210,7 @@ Sometimes, during development, you need to build against a custom version of `li
        'protobuf/3.21.9',
        'grpc/1.50.1',
        'openssl/1.1.1v',
-       'xrpl/2.3.0-b1@my/feature', # Update this line
+       'xrpl/2.5.0-rc1@my/feature', # Use your exported version here
        'zlib/1.3.1',
        'libbacktrace/cci.20210118'
    ]
@@ -192,18 +222,16 @@ Sometimes, during development, you need to build against a custom version of `li
 
 ## Using `clang-tidy` for static analysis
 
-The minimum [clang-tidy](https://clang.llvm.org/extra/clang-tidy/) version required is 19.0.
-
 Clang-tidy can be run by CMake when building the project.
 To achieve this, you just need to provide the option `-o '&:lint=True'` for the `conan install` command:
 
 ```sh
-conan install .. --output-folder . --build missing --settings build_type=Release -o '&:tests=True' -o '&:lint=True'
+conan install .. --output-folder . --build missing --settings build_type=Release -o '&:tests=True' -o '&:lint=True' --profile:all clang
 ```
 
 By default CMake will try to find `clang-tidy` automatically in your system.
 To force CMake to use your desired binary, set the `CLIO_CLANG_TIDY_BIN` environment variable to the path of the `clang-tidy` binary. For example:
 
 ```sh
-export CLIO_CLANG_TIDY_BIN=/opt/homebrew/opt/llvm@19/bin/clang-tidy
+export CLIO_CLANG_TIDY_BIN=/opt/homebrew/opt/llvm/bin/clang-tidy
 ```

docs/configure-clio.md

@@ -5,7 +5,6 @@
 Clio needs access to a `rippled` server in order to work. The following configurations are required for Clio and `rippled` to communicate:
 
 1. In the Clio config file, provide the following:
-
    - The IP of the `rippled` server
 
    - The port on which `rippled` is accepting unencrypted WebSocket connections
@@ -13,7 +12,6 @@ Clio needs access to a `rippled` server in order to work. The following configur
    - The port on which `rippled` is handling gRPC requests
 
 2. In the `rippled` config file, you need to open:
-
    - A port to accept unencrypted WebSocket connections
 
    - A port to handle gRPC requests, with the IP(s) of Clio specified in the `secure_gateway` entry

pre-commit-hooks/json_in_cpp.py

@@ -0,0 +1,76 @@
+#!/usr/bin/env python3
+
+import argparse
+import re
+from pathlib import Path
+
+
+def fix_json_style(cpp_content: str) -> str:
+    cpp_content = cpp_content.replace('R"json(', 'R"JSON(').replace(')json"', ')JSON"')
+
+    pattern = r'R"JSON\((.*?)\)JSON"'
+
+    def replace_json(match):
+        raw_json = match.group(1)
+
+        raw_json = (
+            raw_json.replace(" :", ":")
+            .replace(" ,", ",")
+            .replace(" null", "null")
+            .replace(':"', ': "')
+            .replace(',"', ', "')
+            .replace('":{', '": {')
+            .replace('":[', '": [')
+            .replace('":true', '": true')
+            .replace('":false', '": false')
+            .replace('":null', '": null')
+        )
+        for digit in range(10):
+            raw_json = raw_json.replace(f'":{digit}', f'": {digit}')
+        return f'R"JSON({raw_json})JSON"'
+
+    return re.sub(pattern, replace_json, cpp_content, flags=re.DOTALL)
+
+
+def process_file(file_path: Path, dry_run: bool) -> bool:
+    content = file_path.read_text(encoding="utf-8")
+    new_content = fix_json_style(content)
+
+    if new_content != content:
+        print(f"Processing file: {file_path}")
+        if dry_run:
+            print("Dry run: changes won't be written to the file.")
+        else:
+            print("Writing changes to file.")
+            file_path.write_text(new_content, encoding="utf-8")
+    return new_content == content
+
+
+def main():
+    parser = argparse.ArgumentParser(
+        description="Fix JSON style in C++ files",
+    )
+    parser.add_argument(
+        "--dry-run",
+        default=False,
+        action="store_true",
+        help="Don't modify files, just print what would be changed",
+    )
+    parser.add_argument(
+        "files",
+        nargs="*",
+        help="Specific files to process",
+    )
+
+    args = parser.parse_args()
+
+    success = True
+    for file in args.files:
+        success = success and process_file(Path(file), dry_run=args.dry_run)
+    if not success:
+        print("Errors occurred while processing files.")
+        exit(1)
+
+
+if __name__ == "__main__":
+    main()

src/app/CMakeLists.txt

@@ -10,4 +10,5 @@ target_link_libraries(
          clio_web
          clio_rpc
          clio_migration
+  PRIVATE Boost::program_options
 )

src/app/CliArgs.cpp

@@ -60,17 +60,17 @@ CliArgs::parse(int argc, char const* argv[])
     po::store(po::command_line_parser(argc, argv).options(description).positional(positional).run(), parsed);
     po::notify(parsed);
 
-    if (parsed.count("help") != 0u) {
+    if (parsed.contains("help")) {
         std::cout << "Clio server " << util::build::getClioFullVersionString() << "\n\n" << description;
         return Action{Action::Exit{EXIT_SUCCESS}};
     }
 
-    if (parsed.count("version") != 0u) {
+    if (parsed.contains("version")) {
         std::cout << util::build::getClioFullVersionString() << '\n';
         return Action{Action::Exit{EXIT_SUCCESS}};
     }
 
-    if (parsed.count("config-description") != 0u) {
+    if (parsed.contains("config-description")) {
         std::filesystem::path const filePath = parsed["config-description"].as<std::string>();
 
         auto const res = util::config::ClioConfigDescription::generateConfigDescriptionToFile(filePath);
@@ -83,18 +83,17 @@ CliArgs::parse(int argc, char const* argv[])
 
     auto configPath = parsed["conf"].as<std::string>();
 
-    if (parsed.count("migrate") != 0u) {
+    if (parsed.contains("migrate")) {
         auto const opt = parsed["migrate"].as<std::string>();
         if (opt == "status")
             return Action{Action::Migrate{.configPath = std::move(configPath), .subCmd = MigrateSubCmd::status()}};
         return Action{Action::Migrate{.configPath = std::move(configPath), .subCmd = MigrateSubCmd::migration(opt)}};
     }
 
-    if (parsed.count("verify") != 0u)
+    if (parsed.contains("verify"))
         return Action{Action::VerifyConfig{.configPath = std::move(configPath)}};
 
-    return Action{Action::Run{.configPath = std::move(configPath), .useNgWebServer = parsed.count("ng-web-server") != 0}
-    };
+    return Action{Action::Run{.configPath = std::move(configPath), .useNgWebServer = parsed.contains("ng-web-server")}};
 }
 
 }  // namespace app

src/app/ClioApplication.cpp

@@ -41,7 +41,6 @@
 #include "util/build/Build.hpp"
 #include "util/config/ConfigDefinition.hpp"
 #include "util/log/Logger.hpp"
-#include "util/prometheus/Prometheus.hpp"
 #include "web/AdminVerificationStrategy.hpp"
 #include "web/RPCServerHandler.hpp"
 #include "web/Server.hpp"
@@ -91,7 +90,6 @@ ClioApplication::ClioApplication(util::config::ClioConfigDefinition const& confi
     : config_(config), signalsHandler_{config_}
 {
     LOG(util::LogService::info()) << "Clio version: " << util::build::getClioFullVersionString();
-    PrometheusService::init(config);
     signalsHandler_.subscribeToStop([this]() { appStopper_.stop(); });
 }
 

src/app/WebHandlers.cpp

@@ -97,7 +97,7 @@ HealthCheckHandler::operator()(
     boost::asio::yield_context
 )
 {
-    static auto constexpr kHEALTH_CHECK_HTML = R"html(
+    static constexpr auto kHEALTH_CHECK_HTML = R"html(
     <!DOCTYPE html>
     <html>
         <head><title>Test page for Clio</title></head>

src/data/AmendmentCenter.cpp

@@ -78,17 +78,20 @@ WritingAmendmentKey::WritingAmendmentKey(std::string amendmentName) : AmendmentK
 
 }  // namespace impl
 
-AmendmentKey::operator std::string const&() const
+AmendmentKey::
+operator std::string const&() const
 {
     return name;
 }
 
-AmendmentKey::operator std::string_view() const
+AmendmentKey::
+operator std::string_view() const
 {
     return name;
 }
 
-AmendmentKey::operator ripple::uint256() const
+AmendmentKey::
+operator ripple::uint256() const
 {
     return Amendment::getAmendmentId(name);
 }

src/data/BackendCounters.cpp

@@ -49,35 +49,45 @@ durationInMillisecondsSince(std::chrono::steady_clock::time_point const startTim
 using namespace util::prometheus;
 
 BackendCounters::BackendCounters()
-    : tooBusyCounter_(PrometheusService::counterInt(
-          "backend_too_busy_total_number",
-          Labels(),
-          "The total number of times the backend was too busy to process a request"
-      ))
-    , writeSyncCounter_(PrometheusService::counterInt(
-          "backend_operations_total_number",
-          Labels({Label{"operation", "write_sync"}}),
-          "The total number of times the backend had to write synchronously"
-      ))
-    , writeSyncRetryCounter_(PrometheusService::counterInt(
-          "backend_operations_total_number",
-          Labels({Label{"operation", "write_sync_retry"}}),
-          "The total number of times the backend had to retry a synchronous write"
-      ))
+    : tooBusyCounter_(
+          PrometheusService::counterInt(
+              "backend_too_busy_total_number",
+              Labels(),
+              "The total number of times the backend was too busy to process a request"
+          )
+      )
+    , writeSyncCounter_(
+          PrometheusService::counterInt(
+              "backend_operations_total_number",
+              Labels({Label{"operation", "write_sync"}}),
+              "The total number of times the backend had to write synchronously"
+          )
+      )
+    , writeSyncRetryCounter_(
+          PrometheusService::counterInt(
+              "backend_operations_total_number",
+              Labels({Label{"operation", "write_sync_retry"}}),
+              "The total number of times the backend had to retry a synchronous write"
+          )
+      )
     , asyncWriteCounters_{"write_async"}
     , asyncReadCounters_{"read_async"}
-    , readDurationHistogram_(PrometheusService::histogramInt(
-          "backend_duration_milliseconds_histogram",
-          Labels({Label{"operation", "read"}}),
-          kHISTOGRAM_BUCKETS,
-          "The duration of backend read operations including retries"
-      ))
-    , writeDurationHistogram_(PrometheusService::histogramInt(
-          "backend_duration_milliseconds_histogram",
-          Labels({Label{"operation", "write"}}),
-          kHISTOGRAM_BUCKETS,
-          "The duration of backend write operations including retries"
-      ))
+    , readDurationHistogram_(
+          PrometheusService::histogramInt(
+              "backend_duration_milliseconds_histogram",
+              Labels({Label{"operation", "read"}}),
+              kHISTOGRAM_BUCKETS,
+              "The duration of backend read operations including retries"
+          )
+      )
+    , writeDurationHistogram_(
+          PrometheusService::histogramInt(
+              "backend_duration_milliseconds_histogram",
+              Labels({Label{"operation", "write"}}),
+              kHISTOGRAM_BUCKETS,
+              "The duration of backend write operations including retries"
+          )
+      )
 {
 }
 
@@ -170,26 +180,34 @@ BackendCounters::report() const
 
 BackendCounters::AsyncOperationCounters::AsyncOperationCounters(std::string name)
     : name_(std::move(name))
-    , pendingCounter_(PrometheusService::gaugeInt(
-          "backend_operations_current_number",
-          Labels({{"operation", name_}, {"status", "pending"}}),
-          "The current number of pending " + name_ + " operations"
-      ))
-    , completedCounter_(PrometheusService::counterInt(
-          "backend_operations_total_number",
-          Labels({{"operation", name_}, {"status", "completed"}}),
-          "The total number of completed " + name_ + " operations"
-      ))
-    , retryCounter_(PrometheusService::counterInt(
-          "backend_operations_total_number",
-          Labels({{"operation", name_}, {"status", "retry"}}),
-          "The total number of retried " + name_ + " operations"
-      ))
-    , errorCounter_(PrometheusService::counterInt(
-          "backend_operations_total_number",
-          Labels({{"operation", name_}, {"status", "error"}}),
-          "The total number of errored " + name_ + " operations"
-      ))
+    , pendingCounter_(
+          PrometheusService::gaugeInt(
+              "backend_operations_current_number",
+              Labels({{"operation", name_}, {"status", "pending"}}),
+              "The current number of pending " + name_ + " operations"
+          )
+      )
+    , completedCounter_(
+          PrometheusService::counterInt(
+              "backend_operations_total_number",
+              Labels({{"operation", name_}, {"status", "completed"}}),
+              "The total number of completed " + name_ + " operations"
+          )
+      )
+    , retryCounter_(
+          PrometheusService::counterInt(
+              "backend_operations_total_number",
+              Labels({{"operation", name_}, {"status", "retry"}}),
+              "The total number of retried " + name_ + " operations"
+          )
+      )
+    , errorCounter_(
+          PrometheusService::counterInt(
+              "backend_operations_total_number",
+              Labels({{"operation", name_}, {"status", "error"}}),
+              "The total number of errored " + name_ + " operations"
+          )
+      )
 {
 }
 

src/data/BackendInterface.hpp

@@ -234,8 +234,12 @@ public:
      * @return A vector of ripple::uint256 representing the account roots
      */
     virtual std::vector<ripple::uint256>
-    fetchAccountRoots(std::uint32_t number, std::uint32_t pageSize, std::uint32_t seq, boost::asio::yield_context yield)
-        const = 0;
+    fetchAccountRoots(
+        std::uint32_t number,
+        std::uint32_t pageSize,
+        std::uint32_t seq,
+        boost::asio::yield_context yield
+    ) const = 0;
 
     /**
      * @brief Updates the range of sequences that are stored in the DB.
@@ -459,8 +463,11 @@ public:
      * @return The sequence in unit32_t on success; nullopt otherwise
      */
     virtual std::optional<std::uint32_t>
-    doFetchLedgerObjectSeq(ripple::uint256 const& key, std::uint32_t sequence, boost::asio::yield_context yield)
-        const = 0;
+    doFetchLedgerObjectSeq(
+        ripple::uint256 const& key,
+        std::uint32_t sequence,
+        boost::asio::yield_context yield
+    ) const = 0;
 
     /**
      * @brief The database-specific implementation for fetching ledger objects.

src/data/CassandraBackend.hpp

@@ -40,7 +40,7 @@
 #include <boost/uuid/string_generator.hpp>
 #include <boost/uuid/uuid.hpp>
 #include <cassandra.h>
-#include <fmt/core.h>
+#include <fmt/format.h>
 #include <xrpl/basics/Blob.h>
 #include <xrpl/basics/base_uint.h>
 #include <xrpl/basics/strHex.h>
@@ -361,8 +361,10 @@ public:
     }
 
     std::vector<ripple::uint256>
-    fetchAllTransactionHashesInLedger(std::uint32_t const ledgerSequence, boost::asio::yield_context yield)
-        const override
+    fetchAllTransactionHashesInLedger(
+        std::uint32_t const ledgerSequence,
+        boost::asio::yield_context yield
+    ) const override
     {
         auto start = std::chrono::system_clock::now();
         auto const res = executor_.read(yield, schema_->selectAllTransactionHashesInLedger, ledgerSequence);
@@ -392,8 +394,11 @@ public:
     }
 
     std::optional<NFT>
-    fetchNFT(ripple::uint256 const& tokenID, std::uint32_t const ledgerSequence, boost::asio::yield_context yield)
-        const override
+    fetchNFT(
+        ripple::uint256 const& tokenID,
+        std::uint32_t const ledgerSequence,
+        boost::asio::yield_context yield
+    ) const override
     {
         auto const res = executor_.read(yield, schema_->selectNFT, tokenID, ledgerSequence);
         if (not res)
@@ -554,10 +559,9 @@ public:
         selectNFTStatements.reserve(nftIDs.size());
 
         std::transform(
-            std::cbegin(nftIDs),
-            std::cend(nftIDs),
-            std::back_inserter(selectNFTStatements),
-            [&](auto const& nftID) { return schema_->selectNFT.bind(nftID, ledgerSequence); }
+            std::cbegin(nftIDs), std::cend(nftIDs), std::back_inserter(selectNFTStatements), [&](auto const& nftID) {
+                return schema_->selectNFT.bind(nftID, ledgerSequence);
+            }
         );
 
         auto const nftInfos = executor_.readEach(yield, selectNFTStatements);
@@ -566,10 +570,9 @@ public:
         selectNFTURIStatements.reserve(nftIDs.size());
 
         std::transform(
-            std::cbegin(nftIDs),
-            std::cend(nftIDs),
-            std::back_inserter(selectNFTURIStatements),
-            [&](auto const& nftID) { return schema_->selectNFTURI.bind(nftID, ledgerSequence); }
+            std::cbegin(nftIDs), std::cend(nftIDs), std::back_inserter(selectNFTURIStatements), [&](auto const& nftID) {
+                return schema_->selectNFTURI.bind(nftID, ledgerSequence);
+            }
         );
 
         auto const nftUris = executor_.readEach(yield, selectNFTURIStatements);
@@ -626,8 +629,11 @@ public:
     }
 
     std::optional<Blob>
-    doFetchLedgerObject(ripple::uint256 const& key, std::uint32_t const sequence, boost::asio::yield_context yield)
-        const override
+    doFetchLedgerObject(
+        ripple::uint256 const& key,
+        std::uint32_t const sequence,
+        boost::asio::yield_context yield
+    ) const override
     {
         LOG(log_.debug()) << "Fetching ledger object for seq " << sequence << ", key = " << ripple::to_string(key);
         if (auto const res = executor_.read(yield, schema_->selectObject, key, sequence); res) {
@@ -645,8 +651,11 @@ public:
     }
 
     std::optional<std::uint32_t>
-    doFetchLedgerObjectSeq(ripple::uint256 const& key, std::uint32_t const sequence, boost::asio::yield_context yield)
-        const override
+    doFetchLedgerObjectSeq(
+        ripple::uint256 const& key,
+        std::uint32_t const sequence,
+        boost::asio::yield_context yield
+    ) const override
     {
         LOG(log_.debug()) << "Fetching ledger object for seq " << sequence << ", key = " << ripple::to_string(key);
         if (auto const res = executor_.read(yield, schema_->selectObject, key, sequence); res) {
@@ -680,8 +689,11 @@ public:
     }
 
     std::optional<ripple::uint256>
-    doFetchSuccessorKey(ripple::uint256 key, std::uint32_t const ledgerSequence, boost::asio::yield_context yield)
-        const override
+    doFetchSuccessorKey(
+        ripple::uint256 key,
+        std::uint32_t const ledgerSequence,
+        boost::asio::yield_context yield
+    ) const override
     {
         if (auto const res = executor_.read(yield, schema_->selectSuccessor, key, ledgerSequence); res) {
             if (auto const result = res->template get<ripple::uint256>(); result) {
@@ -714,10 +726,9 @@ public:
         auto const timeDiff = util::timed([this, yield, &results, &hashes, &statements]() {
             // TODO: seems like a job for "hash IN (list of hashes)" instead?
             std::transform(
-                std::cbegin(hashes),
-                std::cend(hashes),
-                std::back_inserter(statements),
-                [this](auto const& hash) { return schema_->selectTransaction.bind(hash); }
+                std::cbegin(hashes), std::cend(hashes), std::back_inserter(statements), [this](auto const& hash) {
+                    return schema_->selectTransaction.bind(hash);
+                }
             );
 
             auto const entries = executor_.readEach(yield, statements);
@@ -761,18 +772,14 @@ public:
 
         // TODO: seems like a job for "key IN (list of keys)" instead?
         std::transform(
-            std::cbegin(keys),
-            std::cend(keys),
-            std::back_inserter(statements),
-            [this, &sequence](auto const& key) { return schema_->selectObject.bind(key, sequence); }
+            std::cbegin(keys), std::cend(keys), std::back_inserter(statements), [this, &sequence](auto const& key) {
+                return schema_->selectObject.bind(key, sequence);
+            }
         );
 
         auto const entries = executor_.readEach(yield, statements);
         std::transform(
-            std::cbegin(entries),
-            std::cend(entries),
-            std::back_inserter(results),
-            [](auto const& res) -> Blob {
+            std::cbegin(entries), std::cend(entries), std::back_inserter(results), [](auto const& res) -> Blob {
                 if (auto const maybeValue = res.template get<Blob>(); maybeValue)
                     return *maybeValue;
 
@@ -785,8 +792,12 @@ public:
     }
 
     std::vector<ripple::uint256>
-    fetchAccountRoots(std::uint32_t number, std::uint32_t pageSize, std::uint32_t seq, boost::asio::yield_context yield)
-        const override
+    fetchAccountRoots(
+        std::uint32_t number,
+        std::uint32_t pageSize,
+        std::uint32_t seq,
+        boost::asio::yield_context yield
+    ) const override
     {
         std::vector<ripple::uint256> liveAccounts;
         std::optional<ripple::AccountID> lastItem;

src/data/DBHelpers.hpp

@@ -198,39 +198,6 @@ struct MPTHolderData {
     ripple::AccountID holder;
 };
 
-/**
- * @brief Check whether the supplied object is an offer.
- *
- * @param object The object to check
- * @return true if the object is an offer; false otherwise
- */
-template <typename T>
-inline bool
-isOffer(T const& object)
-{
-    static constexpr short kOFFER_OFFSET = 0x006f;
-    static constexpr short kSHIFT = 8;
-
-    short offerBytes = (object[1] << kSHIFT) | object[2];
-    return offerBytes == kOFFER_OFFSET;
-}
-
-/**
- * @brief Check whether the supplied hex represents an offer object.
- *
- * @param object The object to check
- * @return true if the object is an offer; false otherwise
- */
-template <typename T>
-inline bool
-isOfferHex(T const& object)
-{
-    auto blob = ripple::strUnHex(4, object.begin(), object.begin() + 4);
-    if (blob)
-        return isOffer(*blob);
-    return false;
-}
-
 /**
  * @brief Check whether the supplied object is a dir node.
  *
@@ -241,6 +208,10 @@ template <typename T>
 inline bool
 isDirNode(T const& object)
 {
+    static constexpr auto kMIN_SIZE_REQUIRED = 3;
+    if (std::size(object) < kMIN_SIZE_REQUIRED)
+        return false;
+
     static constexpr short kDIR_NODE_SPACE_KEY = 0x0064;
     short const spaceKey = (object.data()[1] << 8) | object.data()[2];
     return spaceKey == kDIR_NODE_SPACE_KEY;
@@ -264,23 +235,6 @@ isBookDir(T const& key, R const& object)
     return !sle[~ripple::sfOwner].has_value();
 }
 
-/**
- * @brief Get the book out of an offer object.
- *
- * @param offer The offer to get the book for
- * @return Book as ripple::uint256
- */
-template <typename T>
-inline ripple::uint256
-getBook(T const& offer)
-{
-    ripple::SerialIter it{offer.data(), offer.size()};
-    ripple::SLE const sle{it, {}};
-    ripple::uint256 book = sle.getFieldH256(ripple::sfBookDirectory);
-
-    return book;
-}
-
 /**
  * @brief Get the book base.
  *

src/data/README.md

@@ -6,7 +6,7 @@ To support additional database types, you can create new classes that implement
 
 ## Data Model
 
-The data model used by Clio to read and write ledger data is different from what `rippled` uses. `rippled` uses a novel data structure named [_SHAMap_](https://github.com/ripple/rippled/blob/master/src/ripple/shamap/README.md), which is a combination of a Merkle Tree and a Radix Trie. In a SHAMap, ledger objects are stored in the root vertices of the tree. Thus, looking up a record located at the leaf node of the SHAMap executes a tree search, where the path from the root node to the leaf node is the key of the record.
+The data model used by Clio to read and write ledger data is different from what `rippled` uses. `rippled` uses a novel data structure named [_SHAMap_](https://github.com/XRPLF/rippled/blob/develop/src/xrpld/shamap/README.md), which is a combination of a Merkle Tree and a Radix Trie. In a SHAMap, ledger objects are stored in the root vertices of the tree. Thus, looking up a record located at the leaf node of the SHAMap executes a tree search, where the path from the root node to the leaf node is the key of the record.
 
 `rippled` nodes can also generate a proof-tree by forming a subtree with all the path nodes and their neighbors, which can then be used to prove the existence of the leaf node data to other `rippled` nodes. In short, the main purpose of the SHAMap data structure is to facilitate the fast validation of data integrity between different decentralized `rippled` nodes.
 

src/data/cassandra/Handle.hpp

@@ -99,7 +99,7 @@ public:
     connect() const;
 
     /**
-     * @brief Connect to the the specified keyspace asynchronously.
+     * @brief Connect to the specified keyspace asynchronously.
      *
      * @param keyspace The keyspace to use
      * @return A future
@@ -137,7 +137,7 @@ public:
     disconnect() const;
 
     /**
-     * @brief Reconnect to the the specified keyspace asynchronously.
+     * @brief Reconnect to the specified keyspace asynchronously.
      *
      * @param keyspace The keyspace to use
      * @return A future

src/data/cassandra/impl/Cluster.cpp

@@ -25,7 +25,7 @@
 #include "util/log/Logger.hpp"
 
 #include <cassandra.h>
-#include <fmt/core.h>
+#include <fmt/format.h>
 
 #include <stdexcept>
 #include <string>
@@ -45,7 +45,8 @@ Cluster::Cluster(Settings const& settings) : ManagedObject{cass_cluster_new(), k
 
     cass_cluster_set_token_aware_routing(*this, cass_true);
     if (auto const rc = cass_cluster_set_protocol_version(*this, CASS_PROTOCOL_VERSION_V4); rc != CASS_OK) {
-        throw std::runtime_error(fmt::format("Error setting cassandra protocol version to v4: {}", cass_error_desc(rc))
+        throw std::runtime_error(
+            fmt::format("Error setting cassandra protocol version to v4: {}", cass_error_desc(rc))
         );
     }
 

src/data/cassandra/impl/RetryPolicy.hpp

@@ -45,11 +45,13 @@ public:
      * @brief Create a new retry policy instance with the io_context provided
      */
     ExponentialBackoffRetryPolicy(boost::asio::io_context& ioc)
-        : retry_(util::makeRetryExponentialBackoff(
-              std::chrono::milliseconds(1),
-              std::chrono::seconds(1),
-              boost::asio::make_strand(ioc)
-          ))
+        : retry_(
+              util::makeRetryExponentialBackoff(
+                  std::chrono::milliseconds(1),
+                  std::chrono::seconds(1),
+                  boost::asio::make_strand(ioc)
+              )
+          )
     {
     }
 

src/data/cassandra/impl/Statement.hpp

@@ -28,7 +28,7 @@
 #include <boost/uuid/uuid.hpp>
 #include <boost/uuid/uuid_io.hpp>
 #include <cassandra.h>
-#include <fmt/core.h>
+#include <fmt/format.h>
 #include <xrpl/basics/base_uint.h>
 #include <xrpl/protocol/AccountID.h>
 #include <xrpl/protocol/STAccount.h>

src/etl/ETLService.cpp

@@ -22,7 +22,6 @@
 #include "data/BackendInterface.hpp"
 #include "etl/CacheLoader.hpp"
 #include "etl/CorruptionDetector.hpp"
-#include "etl/ETLState.hpp"
 #include "etl/LoadBalancer.hpp"
 #include "etl/NetworkValidatedLedgersInterface.hpp"
 #include "etl/SystemState.hpp"
@@ -38,6 +37,7 @@
 #include "etlng/LoadBalancer.hpp"
 #include "etlng/LoadBalancerInterface.hpp"
 #include "etlng/impl/LedgerPublisher.hpp"
+#include "etlng/impl/MonitorProvider.hpp"
 #include "etlng/impl/TaskManagerProvider.hpp"
 #include "etlng/impl/ext/Cache.hpp"
 #include "etlng/impl/ext/Core.hpp"
@@ -86,6 +86,7 @@ ETLService::makeETLService(
         );
 
         auto state = std::make_shared<etl::SystemState>();
+        state->isStrictReadonly = config.get<bool>("read_only");
 
         auto fetcher = std::make_shared<etl::impl::LedgerFetcher>(backend, balancer);
         auto extractor = std::make_shared<etlng::impl::Extractor>(fetcher);
@@ -93,6 +94,9 @@ ETLService::makeETLService(
         auto cacheLoader = std::make_shared<etl::CacheLoader<>>(config, backend, backend->cache());
         auto cacheUpdater = std::make_shared<etlng::impl::CacheUpdater>(backend->cache());
         auto amendmentBlockHandler = std::make_shared<etlng::impl::AmendmentBlockHandler>(ctx, *state);
+        auto monitorProvider = std::make_shared<etlng::impl::MonitorProvider>();
+
+        backend->setCorruptionDetector(CorruptionDetector{*state, backend->cache()});
 
         auto loader = std::make_shared<etlng::impl::Loader>(
             backend,
@@ -104,7 +108,8 @@ ETLService::makeETLService(
                 etlng::impl::NFTExt{backend},
                 etlng::impl::MPTExt{backend}
             ),
-            amendmentBlockHandler
+            amendmentBlockHandler,
+            state
         );
 
         auto taskManagerProvider = std::make_shared<etlng::impl::TaskManagerProvider>(*ledgers, extractor, loader);
@@ -122,6 +127,7 @@ ETLService::makeETLService(
             loader,  // loader itself
             loader,  // initial load observer
             taskManagerProvider,
+            monitorProvider,
             state
         );
     } else {
@@ -165,9 +171,11 @@ ETLService::runETLPipeline(uint32_t startSequence, uint32_t numExtractors)
     auto pipe = DataPipeType{numExtractors, startSequence};
 
     for (auto i = 0u; i < numExtractors; ++i) {
-        extractors.push_back(std::make_unique<ExtractorType>(
-            pipe, networkValidatedLedgers_, ledgerFetcher_, startSequence + i, finishSequence_, state_
-        ));
+        extractors.push_back(
+            std::make_unique<ExtractorType>(
+                pipe, networkValidatedLedgers_, ledgerFetcher_, startSequence + i, finishSequence_, state_
+            )
+        );
     }
 
     auto transformer =
@@ -346,7 +354,7 @@ ETLService::doWork()
     worker_ = std::thread([this]() {
         beast::setCurrentThreadName("ETLService worker");
 
-        if (state_.isReadOnly) {
+        if (state_.isStrictReadonly) {
             monitorReadOnly();
         } else {
             monitor();
@@ -373,7 +381,7 @@ ETLService::ETLService(
 {
     startSequence_ = config.maybeValue<uint32_t>("start_sequence");
     finishSequence_ = config.maybeValue<uint32_t>("finish_sequence");
-    state_.isReadOnly = config.get<bool>("read_only");
+    state_.isStrictReadonly = config.get<bool>("read_only");
     extractorThreads_ = config.get<uint32_t>("extractor_threads");
 
     // This should probably be done in the backend factory but we don't have state available until here

src/etl/ETLService.hpp

@@ -239,7 +239,7 @@ public:
 
         result["etl_sources"] = loadBalancer_->toJson();
         result["is_writer"] = static_cast<int>(state_.isWriting);
-        result["read_only"] = static_cast<int>(state_.isReadOnly);
+        result["read_only"] = static_cast<int>(state_.isStrictReadonly);
         auto last = ledgerPublisher_.getLastPublish();
         if (last.time_since_epoch().count() != 0)
             result["last_publish_age_seconds"] = std::to_string(ledgerPublisher_.lastPublishAgeSeconds());

src/etl/LoadBalancer.cpp

@@ -44,7 +44,7 @@
 #include <boost/json/object.hpp>
 #include <boost/json/value.hpp>
 #include <boost/json/value_to.hpp>
-#include <fmt/core.h>
+#include <fmt/format.h>
 
 #include <algorithm>
 #include <chrono>
@@ -57,7 +57,6 @@
 #include <string>
 #include <thread>
 #include <utility>
-#include <variant>
 #include <vector>
 
 using namespace util::config;
@@ -184,11 +183,14 @@ LoadBalancer::LoadBalancer(
             LOG(log_.warn()) << "Failed to fetch ETL state from source = " << source->toString()
                              << " Please check the configuration and network";
         } else if (etlState_ && etlState_->networkID != stateOpt->networkID) {
-            checkOnETLFailure(fmt::format(
-                "ETL sources must be on the same network. Source network id = {} does not match others network id = {}",
-                stateOpt->networkID,
-                etlState_->networkID
-            ));
+            checkOnETLFailure(
+                fmt::format(
+                    "ETL sources must be on the same network. Source network id = {} does not match others network id "
+                    "= {}",
+                    stateOpt->networkID,
+                    etlState_->networkID
+                )
+            );
         } else {
             etlState_ = stateOpt;
         }
@@ -275,42 +277,46 @@ LoadBalancer::forwardToRippled(
         return std::unexpected{rpc::ClioError::RpcCommandIsMissing};
 
     auto const cmd = boost::json::value_to<std::string>(request.at("command"));
-
-    if (forwardingCache_ and forwardingCache_->shouldCache(cmd)) {
-        bool servedFromCache = true;
-        auto updater =
-            [this, &request, &clientIp, &servedFromCache, isAdmin](boost::asio::yield_context yield
-            ) -> std::expected<util::ResponseExpirationCache::EntryData, util::ResponseExpirationCache::Error> {
-            servedFromCache = false;
-            auto result = forwardToRippledImpl(request, clientIp, isAdmin, yield);
-            if (result.has_value()) {
-                return util::ResponseExpirationCache::EntryData{
-                    .lastUpdated = std::chrono::steady_clock::now(), .response = std::move(result).value()
-                };
-            }
-            return std::unexpected{
-                util::ResponseExpirationCache::Error{.status = rpc::Status{result.error()}, .warnings = {}}
-            };
-        };
-
-        auto result = forwardingCache_->getOrUpdate(
-            yield,
-            cmd,
-            std::move(updater),
-            [](util::ResponseExpirationCache::EntryData const& entry) { return not entry.response.contains("error"); }
-        );
-        if (servedFromCache) {
-            ++forwardingCounters_.cacheHit.get();
+    if (forwardingCache_) {
+        if (auto cachedResponse = forwardingCache_->get(cmd); cachedResponse) {
+            forwardingCounters_.cacheHit.get() += 1;
+            return std::move(cachedResponse).value();
         }
-        if (result.has_value()) {
-            return std::move(result).value();
+    }
+    forwardingCounters_.cacheMiss.get() += 1;
+
+    ASSERT(not sources_.empty(), "ETL sources must be configured to forward requests.");
+    std::size_t sourceIdx = randomGenerator_->uniform(0ul, sources_.size() - 1);
+
+    auto numAttempts = 0u;
+
+    auto xUserValue = isAdmin ? kADMIN_FORWARDING_X_USER_VALUE : kUSER_FORWARDING_X_USER_VALUE;
+
+    std::optional<boost::json::object> response;
+    rpc::ClioError error = rpc::ClioError::EtlConnectionError;
+    while (numAttempts < sources_.size()) {
+        auto [res, duration] =
+            util::timed([&]() { return sources_[sourceIdx]->forwardToRippled(request, clientIp, xUserValue, yield); });
+        if (res) {
+            forwardingCounters_.successDuration.get() += duration;
+            response = std::move(res).value();
+            break;
         }
-        auto const combinedError = result.error().status.code;
-        ASSERT(std::holds_alternative<rpc::ClioError>(combinedError), "There could be only ClioError here");
-        return std::unexpected{std::get<rpc::ClioError>(combinedError)};
+        forwardingCounters_.failDuration.get() += duration;
+        ++forwardingCounters_.retries.get();
+        error = std::max(error, res.error());  // Choose the best result between all sources
+
+        sourceIdx = (sourceIdx + 1) % sources_.size();
+        ++numAttempts;
     }
 
-    return forwardToRippledImpl(request, clientIp, isAdmin, yield);
+    if (response) {
+        if (forwardingCache_ and not response->contains("error"))
+            forwardingCache_->put(cmd, *response);
+        return std::move(response).value();
+    }
+
+    return std::unexpected{error};
 }
 
 boost::json::value
@@ -401,47 +407,4 @@ LoadBalancer::chooseForwardingSource()
     }
 }
 
-std::expected<boost::json::object, rpc::CombinedError>
-LoadBalancer::forwardToRippledImpl(
-    boost::json::object const& request,
-    std::optional<std::string> const& clientIp,
-    bool const isAdmin,
-    boost::asio::yield_context yield
-)
-{
-    ++forwardingCounters_.cacheMiss.get();
-
-    ASSERT(not sources_.empty(), "ETL sources must be configured to forward requests.");
-    std::size_t sourceIdx = randomGenerator_->uniform(0ul, sources_.size() - 1);
-
-    auto numAttempts = 0u;
-
-    auto xUserValue = isAdmin ? kADMIN_FORWARDING_X_USER_VALUE : kUSER_FORWARDING_X_USER_VALUE;
-
-    std::optional<boost::json::object> response;
-    rpc::ClioError error = rpc::ClioError::EtlConnectionError;
-    while (numAttempts < sources_.size()) {
-        auto [res, duration] =
-            util::timed([&]() { return sources_[sourceIdx]->forwardToRippled(request, clientIp, xUserValue, yield); });
-
-        if (res) {
-            forwardingCounters_.successDuration.get() += duration;
-            response = std::move(res).value();
-            break;
-        }
-        forwardingCounters_.failDuration.get() += duration;
-        ++forwardingCounters_.retries.get();
-        error = std::max(error, res.error());  // Choose the best result between all sources
-
-        sourceIdx = (sourceIdx + 1) % sources_.size();
-        ++numAttempts;
-    }
-
-    if (response.has_value()) {
-        return std::move(response).value();
-    }
-
-    return std::unexpected{error};
-}
-
 }  // namespace etl

src/etl/LoadBalancer.hpp

@@ -49,6 +49,7 @@
 #include <concepts>
 #include <cstdint>
 #include <expected>
+#include <functional>
 #include <memory>
 #include <optional>
 #include <string>
@@ -172,19 +173,21 @@ public:
      * @return A std::vector<std::string> The ledger data
      */
     std::vector<std::string>
-    loadInitialLedger(uint32_t sequence, std::chrono::steady_clock::duration retryAfter = std::chrono::seconds{2})
-        override;
+    loadInitialLedger(
+        uint32_t sequence,
+        std::chrono::steady_clock::duration retryAfter = std::chrono::seconds{2}
+    ) override;
 
     /**
      * @brief Load the initial ledger, writing data to the queue.
-     * @note This function will retry indefinitely until the ledger is downloaded.
+     * @note This function will retry indefinitely until the ledger is downloaded or the download is cancelled.
      *
      * @param sequence Sequence of ledger to download
      * @param observer The observer to notify of progress
      * @param retryAfter Time to wait between retries (2 seconds by default)
-     * @return A std::vector<std::string> The ledger data
+     * @return A std::expected with ledger edge keys on success, or InitialLedgerLoadError on failure
      */
-    std::vector<std::string>
+    etlng::InitialLedgerLoadResult
     loadInitialLedger(
         [[maybe_unused]] uint32_t sequence,
         [[maybe_unused]] etlng::InitialLoadObserverInterface& observer,
@@ -279,14 +282,6 @@ private:
      */
     void
     chooseForwardingSource();
-
-    std::expected<boost::json::object, rpc::CombinedError>
-    forwardToRippledImpl(
-        boost::json::object const& request,
-        std::optional<std::string> const& clientIp,
-        bool isAdmin,
-        boost::asio::yield_context yield
-    );
 };
 
 }  // namespace etl

src/etl/NFTHelpers.cpp

@@ -18,8 +18,9 @@
 //==============================================================================
 
 #include "data/DBHelpers.hpp"
+#include "util/Assert.hpp"
 
-#include <fmt/core.h>
+#include <fmt/format.h>
 #include <xrpl/basics/base_uint.h>
 #include <xrpl/basics/strHex.h>
 #include <xrpl/protocol/AccountID.h>
@@ -138,7 +139,8 @@ getNFTokenMintData(ripple::TxMeta const& txMeta, ripple::STTx const& sttx)
     // There should always be a difference so the returned finalIDs
     // iterator should never be end().  But better safe than sorry.
     if (finalIDs.size() != prevIDs.size() + 1 || diff.first == finalIDs.end() || !owner) {
-        throw std::runtime_error(fmt::format(" - unexpected NFTokenMint data in tx {}", strHex(sttx.getTransactionID()))
+        throw std::runtime_error(
+            fmt::format(" - unexpected NFTokenMint data in tx {}", strHex(sttx.getTransactionID()))
         );
     }
 
@@ -358,14 +360,18 @@ getNFTDataFromTx(ripple::TxMeta const& txMeta, ripple::STTx const& sttx)
 std::vector<NFTsData>
 getNFTDataFromObj(std::uint32_t const seq, std::string const& key, std::string const& blob)
 {
-    std::vector<NFTsData> nfts;
-    ripple::STLedgerEntry const sle =
+    // https://github.com/XRPLF/XRPL-Standards/tree/master/XLS-0020-non-fungible-tokens#tokenpage-id-format
+    ASSERT(key.size() == ripple::uint256::size(), "The size of the key (token) is expected to fit uint256 exactly");
+
+    auto const sle =
         ripple::STLedgerEntry(ripple::SerialIter{blob.data(), blob.size()}, ripple::uint256::fromVoid(key.data()));
 
     if (sle.getFieldU16(ripple::sfLedgerEntryType) != ripple::ltNFTOKEN_PAGE)
-        return nfts;
+        return {};
 
     auto const owner = ripple::AccountID::fromVoid(key.data());
+    std::vector<NFTsData> nfts;
+
     for (ripple::STObject const& node : sle.getFieldArray(ripple::sfNFTokens))
         nfts.emplace_back(node.getFieldH256(ripple::sfNFTokenID), seq, owner, node.getFieldVL(ripple::sfURI));
 

src/etl/SystemState.hpp

@@ -37,7 +37,7 @@ struct SystemState {
      * In strict read-only mode, the process will never attempt to become the ETL writer, and will only publish ledgers
      * as they are written to the database.
      */
-    util::prometheus::Bool isReadOnly = PrometheusService::boolMetric(
+    util::prometheus::Bool isStrictReadonly = PrometheusService::boolMetric(
         "read_only",
         util::prometheus::Labels{},
         "Whether the process is in strict read-only mode"

src/etl/impl/Extractor.hpp

@@ -94,8 +94,8 @@ private:
         double totalTime = 0.0;
         auto currentSequence = startSequence_;
 
-        while (!shouldFinish(currentSequence) && networkValidatedLedgers_->waitUntilValidatedByNetwork(currentSequence)
-        ) {
+        while (!shouldFinish(currentSequence) &&
+               networkValidatedLedgers_->waitUntilValidatedByNetwork(currentSequence)) {
             auto [fetchResponse, time] = ::util::timed<std::chrono::duration<double>>([this, currentSequence]() {
                 return ledgerFetcher_.get().fetchDataAndDiff(currentSequence);
             });

src/etl/impl/ForwardingSource.cpp

@@ -28,7 +28,7 @@
 #include <boost/json/object.hpp>
 #include <boost/json/parse.hpp>
 #include <boost/json/serialize.hpp>
-#include <fmt/core.h>
+#include <fmt/format.h>
 
 #include <chrono>
 #include <exception>

src/etl/impl/GrpcSource.cpp

@@ -26,7 +26,7 @@
 
 #include <boost/asio/io_context.hpp>
 #include <boost/asio/ip/tcp.hpp>
-#include <fmt/core.h>
+#include <fmt/format.h>
 #include <grpcpp/client_context.h>
 #include <grpcpp/security/credentials.h>
 #include <grpcpp/support/channel_arguments.h>

src/etl/impl/LedgerLoader.hpp

@@ -209,47 +209,49 @@ public:
             size_t numWrites = 0;
             backend_->cache().setFull();
 
-            auto seconds = ::util::timed<std::chrono::seconds>([this, keys = std::move(edgeKeys), sequence, &numWrites](
-                                                               ) mutable {
-                for (auto& key : keys) {
-                    LOG(log_.debug()) << "Writing edge key = " << ripple::strHex(key);
-                    auto succ = backend_->cache().getSuccessor(*ripple::uint256::fromVoidChecked(key), sequence);
-                    if (succ)
-                        backend_->writeSuccessor(std::move(key), sequence, uint256ToString(succ->key));
-                }
-
-                ripple::uint256 prev = data::kFIRST_KEY;
-                while (auto cur = backend_->cache().getSuccessor(prev, sequence)) {
-                    ASSERT(cur.has_value(), "Successor for key {} must exist", ripple::strHex(prev));
-                    if (prev == data::kFIRST_KEY)
-                        backend_->writeSuccessor(uint256ToString(prev), sequence, uint256ToString(cur->key));
-
-                    if (isBookDir(cur->key, cur->blob)) {
-                        auto base = getBookBase(cur->key);
-                        // make sure the base is not an actual object
-                        if (!backend_->cache().get(base, sequence)) {
-                            auto succ = backend_->cache().getSuccessor(base, sequence);
-                            ASSERT(succ.has_value(), "Book base {} must have a successor", ripple::strHex(base));
-                            if (succ->key == cur->key) {
-                                LOG(log_.debug()) << "Writing book successor = " << ripple::strHex(base) << " - "
-                                                  << ripple::strHex(cur->key);
-
-                                backend_->writeSuccessor(uint256ToString(base), sequence, uint256ToString(cur->key));
-                            }
-                        }
-
-                        ++numWrites;
+            auto seconds =
+                ::util::timed<std::chrono::seconds>([this, keys = std::move(edgeKeys), sequence, &numWrites]() mutable {
+                    for (auto& key : keys) {
+                        LOG(log_.debug()) << "Writing edge key = " << ripple::strHex(key);
+                        auto succ = backend_->cache().getSuccessor(*ripple::uint256::fromVoidChecked(key), sequence);
+                        if (succ)
+                            backend_->writeSuccessor(std::move(key), sequence, uint256ToString(succ->key));
                     }
 
-                    prev = cur->key;
-                    static constexpr std::size_t kLOG_INTERVAL = 100000;
-                    if (numWrites % kLOG_INTERVAL == 0 && numWrites != 0)
-                        LOG(log_.info()) << "Wrote " << numWrites << " book successors";
-                }
+                    ripple::uint256 prev = data::kFIRST_KEY;
+                    while (auto cur = backend_->cache().getSuccessor(prev, sequence)) {
+                        ASSERT(cur.has_value(), "Successor for key {} must exist", ripple::strHex(prev));
+                        if (prev == data::kFIRST_KEY)
+                            backend_->writeSuccessor(uint256ToString(prev), sequence, uint256ToString(cur->key));
 
-                backend_->writeSuccessor(uint256ToString(prev), sequence, uint256ToString(data::kLAST_KEY));
-                ++numWrites;
-            });
+                        if (isBookDir(cur->key, cur->blob)) {
+                            auto base = getBookBase(cur->key);
+                            // make sure the base is not an actual object
+                            if (!backend_->cache().get(base, sequence)) {
+                                auto succ = backend_->cache().getSuccessor(base, sequence);
+                                ASSERT(succ.has_value(), "Book base {} must have a successor", ripple::strHex(base));
+                                if (succ->key == cur->key) {
+                                    LOG(log_.debug()) << "Writing book successor = " << ripple::strHex(base) << " - "
+                                                      << ripple::strHex(cur->key);
+
+                                    backend_->writeSuccessor(
+                                        uint256ToString(base), sequence, uint256ToString(cur->key)
+                                    );
+                                }
+                            }
+
+                            ++numWrites;
+                        }
+
+                        prev = cur->key;
+                        static constexpr std::size_t kLOG_STRIDE = 100000;
+                        if (numWrites % kLOG_STRIDE == 0 && numWrites != 0)
+                            LOG(log_.info()) << "Wrote " << numWrites << " book successors";
+                    }
+
+                    backend_->writeSuccessor(uint256ToString(prev), sequence, uint256ToString(data::kLAST_KEY));
+                    ++numWrites;
+                });
 
             LOG(log_.info()) << "Looping through cache and submitting all writes took " << seconds
                              << " seconds. numWrites = " << std::to_string(numWrites);

src/etl/impl/LedgerPublisher.hpp

@@ -249,8 +249,9 @@ public:
     std::chrono::time_point<std::chrono::system_clock>
     getLastPublish() const override
     {
-        return std::chrono::time_point<std::chrono::system_clock>{std::chrono::seconds{lastPublishSeconds_.get().value()
-        }};
+        return std::chrono::time_point<std::chrono::system_clock>{
+            std::chrono::seconds{lastPublishSeconds_.get().value()}
+        };
     }
 
     /**

src/etl/impl/SubscriptionSource.cpp

@@ -40,7 +40,7 @@
 #include <boost/json/parse.hpp>
 #include <boost/json/serialize.hpp>
 #include <boost/json/value_to.hpp>
-#include <fmt/core.h>
+#include <fmt/format.h>
 #include <xrpl/protocol/jss.h>
 
 #include <algorithm>
@@ -79,11 +79,13 @@ SubscriptionSource::SubscriptionSource(
     , onConnect_(std::move(onConnect))
     , onDisconnect_(std::move(onDisconnect))
     , onLedgerClosed_(std::move(onLedgerClosed))
-    , lastMessageTimeSecondsSinceEpoch_(PrometheusService::gaugeInt(
-          "subscription_source_last_message_time",
-          util::prometheus::Labels({{"source", fmt::format("{}:{}", ip, wsPort)}}),
-          "Seconds since epoch of the last message received from rippled subscription streams"
-      ))
+    , lastMessageTimeSecondsSinceEpoch_(
+          PrometheusService::gaugeInt(
+              "subscription_source_last_message_time",
+              util::prometheus::Labels({{"source", fmt::format("{}:{}", ip, wsPort)}}),
+              "Seconds since epoch of the last message received from rippled subscription streams"
+          )
+      )
 {
     wsConnectionBuilder_.addHeader({boost::beast::http::field::user_agent, "clio-client"})
         .addHeader({"X-User", "clio-client"})
@@ -329,9 +331,13 @@ SubscriptionSource::setValidatedRange(std::string range)
             pairs.emplace_back(sequence, sequence);
         } else {
             if (minAndMax.size() != 2) {
-                throw std::runtime_error(fmt::format(
-                    "Error parsing range: {}.Min and max should be of size 2. Got size = {}", range, minAndMax.size()
-                ));
+                throw std::runtime_error(
+                    fmt::format(
+                        "Error parsing range: {}.Min and max should be of size 2. Got size = {}",
+                        range,
+                        minAndMax.size()
+                    )
+                );
             }
             uint32_t const min = std::stoll(minAndMax[0]);
             uint32_t const max = std::stoll(minAndMax[1]);

src/etl/impl/SubscriptionSource.hpp

@@ -34,7 +34,7 @@
 #include <boost/asio/spawn.hpp>
 #include <boost/asio/strand.hpp>
 #include <boost/beast/http/field.hpp>
-#include <fmt/core.h>
+#include <fmt/format.h>
 
 #include <atomic>
 #include <chrono>

src/etlng/ETLService.cpp

@@ -35,13 +35,13 @@
 #include "etlng/LoadBalancerInterface.hpp"
 #include "etlng/LoaderInterface.hpp"
 #include "etlng/MonitorInterface.hpp"
+#include "etlng/MonitorProviderInterface.hpp"
 #include "etlng/TaskManagerProviderInterface.hpp"
 #include "etlng/impl/AmendmentBlockHandler.hpp"
 #include "etlng/impl/CacheUpdater.hpp"
 #include "etlng/impl/Extraction.hpp"
 #include "etlng/impl/LedgerPublisher.hpp"
 #include "etlng/impl/Loading.hpp"
-#include "etlng/impl/Monitor.hpp"
 #include "etlng/impl/Registry.hpp"
 #include "etlng/impl/Scheduling.hpp"
 #include "etlng/impl/TaskManager.hpp"
@@ -56,7 +56,7 @@
 
 #include <boost/json/object.hpp>
 #include <boost/signals2/connection.hpp>
-#include <fmt/core.h>
+#include <xrpl/protocol/LedgerHeader.h>
 
 #include <chrono>
 #include <cstddef>
@@ -82,6 +82,7 @@ ETLService::ETLService(
     std::shared_ptr<LoaderInterface> loader,
     std::shared_ptr<InitialLoadObserverInterface> initialLoadObserver,
     std::shared_ptr<etlng::TaskManagerProviderInterface> taskManagerProvider,
+    std::shared_ptr<etlng::MonitorProviderInterface> monitorProvider,
     std::shared_ptr<etl::SystemState> state
 )
     : ctx_(std::move(ctx))
@@ -96,9 +97,20 @@ ETLService::ETLService(
     , loader_(std::move(loader))
     , initialLoadObserver_(std::move(initialLoadObserver))
     , taskManagerProvider_(std::move(taskManagerProvider))
+    , monitorProvider_(std::move(monitorProvider))
     , state_(std::move(state))
+    , startSequence_(config.get().maybeValue<uint32_t>("start_sequence"))
+    , finishSequence_(config.get().maybeValue<uint32_t>("finish_sequence"))
 {
-    LOG(log_.info()) << "Creating ETLng...";
+    ASSERT(not state_->isWriting, "ETL should never start in writer mode");
+
+    if (startSequence_.has_value())
+        LOG(log_.info()) << "Start sequence: " << *startSequence_;
+
+    if (finishSequence_.has_value())
+        LOG(log_.info()) << "Finish sequence: " << *finishSequence_;
+
+    LOG(log_.info()) << "Starting in " << (state_->isStrictReadonly ? "STRICT READONLY MODE" : "WRITE MODE");
 }
 
 ETLService::~ETLService()
@@ -112,12 +124,7 @@ ETLService::run()
 {
     LOG(log_.info()) << "Running ETLng...";
 
-    // TODO: write-enabled node should start in readonly and do the 10 second dance to become a writer
     mainLoop_.emplace(ctx_.execute([this] {
-        state_->isWriting =
-            not state_->isReadOnly;  // TODO: this is now needed because we don't have a mechanism for readonly or
-                                     // ETL writer node. remove later in favor of real mechanism
-
         auto const rng = loadInitialLedgerIfNeeded();
 
         LOG(log_.info()) << "Waiting for next ledger to be validated by network...";
@@ -129,15 +136,18 @@ ETLService::run()
             return;
         }
 
-        ASSERT(rng.has_value(), "Ledger range can't be null");
+        if (not rng.has_value()) {
+            LOG(log_.warn()) << "Initial ledger download got cancelled - stopping ETL service";
+            return;
+        }
+
         auto const nextSequence = rng->maxSequence + 1;
 
         LOG(log_.debug()) << "Database is populated. Starting monitor loop. sequence = " << nextSequence;
         startMonitor(nextSequence);
 
-        // TODO: we only want to run the full ETL task man if we are POSSIBLY a write node
-        // but definitely not in strict readonly
-        if (not state_->isReadOnly)
+        // If we are a writer as the result of loading the initial ledger - start loading
+        if (state_->isWriting)
             startLoading(nextSequence);
     }));
 }
@@ -147,6 +157,8 @@ ETLService::stop()
 {
     LOG(log_.info()) << "Stop called";
 
+    if (mainLoop_)
+        mainLoop_->wait();
     if (taskMan_)
         taskMan_->stop();
     if (monitor_)
@@ -160,7 +172,7 @@ ETLService::getInfo() const
 
     result["etl_sources"] = balancer_->toJson();
     result["is_writer"] = static_cast<int>(state_->isWriting);
-    result["read_only"] = static_cast<int>(state_->isReadOnly);
+    result["read_only"] = static_cast<int>(state_->isStrictReadonly);
     auto last = publisher_->getLastPublish();
     if (last.time_since_epoch().count() != 0)
         result["last_publish_age_seconds"] = std::to_string(publisher_->lastPublishAgeSeconds());
@@ -196,21 +208,40 @@ ETLService::loadInitialLedgerIfNeeded()
 {
     auto rng = backend_->hardFetchLedgerRangeNoThrow();
     if (not rng.has_value()) {
-        LOG(log_.info()) << "Database is empty. Will download a ledger from the network.";
+        ASSERT(
+            not state_->isStrictReadonly,
+            "Database is empty but this node is in strict readonly mode. Can't write initial ledger."
+        );
 
-        LOG(log_.info()) << "Waiting for next ledger to be validated by network...";
-        if (auto const mostRecentValidated = ledgers_->getMostRecent(); mostRecentValidated.has_value()) {
-            auto const seq = *mostRecentValidated;
-            LOG(log_.info()) << "Ledger " << seq << " has been validated. Downloading... ";
+        LOG(log_.info()) << "Database is empty. Will download a ledger from the network.";
+        state_->isWriting = true;  // immediately become writer as the db is empty
+
+        auto const getMostRecent = [this]() {
+            LOG(log_.info()) << "Waiting for next ledger to be validated by network...";
+            return ledgers_->getMostRecent();
+        };
+
+        if (auto const maybeSeq = startSequence_.or_else(getMostRecent); maybeSeq.has_value()) {
+            auto const seq = *maybeSeq;
+            LOG(log_.info()) << "Starting from sequence " << seq
+                             << ". Initial ledger download and extraction can take a while...";
 
             auto [ledger, timeDiff] = ::util::timed<std::chrono::duration<double>>([this, seq]() {
-                return extractor_->extractLedgerOnly(seq).and_then([this, seq](auto&& data) {
-                    // TODO: loadInitialLedger in balancer should be called fetchEdgeKeys or similar
-                    data.edgeKeys = balancer_->loadInitialLedger(seq, *initialLoadObserver_);
+                return extractor_->extractLedgerOnly(seq).and_then(
+                    [this, seq](auto&& data) -> std::optional<ripple::LedgerHeader> {
+                        // TODO: loadInitialLedger in balancer should be called fetchEdgeKeys or similar
+                        auto res = balancer_->loadInitialLedger(seq, *initialLoadObserver_);
+                        if (not res.has_value() and res.error() == InitialLedgerLoadError::Cancelled) {
+                            LOG(log_.debug()) << "Initial ledger load got cancelled";
+                            return std::nullopt;
+                        }
 
-                    // TODO: this should be interruptible for graceful shutdown
-                    return loader_->loadInitialLedger(data);
-                });
+                        ASSERT(res.has_value(), "Initial ledger retry logic failed");
+                        data.edgeKeys = std::move(res).value();
+
+                        return loader_->loadInitialLedger(data);
+                    }
+                );
             });
 
             if (not ledger.has_value()) {
@@ -238,28 +269,64 @@ ETLService::loadInitialLedgerIfNeeded()
 void
 ETLService::startMonitor(uint32_t seq)
 {
-    monitor_ = std::make_unique<impl::Monitor>(ctx_, backend_, ledgers_, seq);
-    monitorSubscription_ = monitor_->subscribe([this](uint32_t seq) {
-        log_.info() << "MONITOR got new seq from db: " << seq;
+    monitor_ = monitorProvider_->make(ctx_, backend_, ledgers_, seq);
+
+    monitorNewSeqSubscription_ = monitor_->subscribeToNewSequence([this](uint32_t seq) {
+        LOG(log_.info()) << "ETLService (via Monitor) got new seq from db: " << seq;
+
+        if (state_->writeConflict) {
+            LOG(log_.info()) << "Got a write conflict; Giving up writer seat immediately";
+            giveUpWriter();
+        }
 
-        // FIXME: is this the best way?
         if (not state_->isWriting) {
             auto const diff = data::synchronousAndRetryOnTimeout([this, seq](auto yield) {
                 return backend_->fetchLedgerDiff(seq, yield);
             });
+
             cacheUpdater_->update(seq, diff);
+            backend_->updateRange(seq);
         }
 
         publisher_->publish(seq, {});
     });
+
+    monitorDbStalledSubscription_ = monitor_->subscribeToDbStalled([this]() {
+        LOG(log_.warn()) << "ETLService received DbStalled signal from Monitor";
+        if (not state_->isStrictReadonly and not state_->isWriting)
+            attemptTakeoverWriter();
+    });
+
     monitor_->run();
 }
 
 void
 ETLService::startLoading(uint32_t seq)
 {
-    taskMan_ = taskManagerProvider_->make(ctx_, *monitor_, seq);
+    ASSERT(not state_->isStrictReadonly, "This should only happen on writer nodes");
+    taskMan_ = taskManagerProvider_->make(ctx_, *monitor_, seq, finishSequence_);
     taskMan_->run(config_.get().get<std::size_t>("extractor_threads"));
 }
 
+void
+ETLService::attemptTakeoverWriter()
+{
+    ASSERT(not state_->isStrictReadonly, "This should only happen on writer nodes");
+    auto rng = backend_->hardFetchLedgerRangeNoThrow();
+    ASSERT(rng.has_value(), "Ledger range can't be null");
+
+    state_->isWriting = true;  // switch to writer
+    LOG(log_.info()) << "Taking over the ETL writer seat";
+    startLoading(rng->maxSequence + 1);
+}
+
+void
+ETLService::giveUpWriter()
+{
+    ASSERT(not state_->isStrictReadonly, "This should only happen on writer nodes");
+    state_->isWriting = false;
+    state_->writeConflict = false;
+    taskMan_ = nullptr;
+}
+
 }  // namespace etlng

src/etlng/ETLService.hpp

@@ -35,6 +35,7 @@
 #include "etlng/LoadBalancerInterface.hpp"
 #include "etlng/LoaderInterface.hpp"
 #include "etlng/MonitorInterface.hpp"
+#include "etlng/MonitorProviderInterface.hpp"
 #include "etlng/TaskManagerInterface.hpp"
 #include "etlng/TaskManagerProviderInterface.hpp"
 #include "etlng/impl/AmendmentBlockHandler.hpp"
@@ -42,7 +43,6 @@
 #include "etlng/impl/Extraction.hpp"
 #include "etlng/impl/LedgerPublisher.hpp"
 #include "etlng/impl/Loading.hpp"
-#include "etlng/impl/Monitor.hpp"
 #include "etlng/impl/Registry.hpp"
 #include "etlng/impl/Scheduling.hpp"
 #include "etlng/impl/TaskManager.hpp"
@@ -58,7 +58,7 @@
 #include <boost/asio/io_context.hpp>
 #include <boost/json/object.hpp>
 #include <boost/signals2/connection.hpp>
-#include <fmt/core.h>
+#include <fmt/format.h>
 #include <xrpl/basics/Blob.h>
 #include <xrpl/basics/base_uint.h>
 #include <xrpl/basics/strHex.h>
@@ -106,12 +106,17 @@ class ETLService : public ETLServiceInterface {
     std::shared_ptr<LoaderInterface> loader_;
     std::shared_ptr<InitialLoadObserverInterface> initialLoadObserver_;
     std::shared_ptr<etlng::TaskManagerProviderInterface> taskManagerProvider_;
+    std::shared_ptr<etlng::MonitorProviderInterface> monitorProvider_;
     std::shared_ptr<etl::SystemState> state_;
 
+    std::optional<uint32_t> startSequence_;
+    std::optional<uint32_t> finishSequence_;
+
     std::unique_ptr<MonitorInterface> monitor_;
     std::unique_ptr<TaskManagerInterface> taskMan_;
 
-    boost::signals2::scoped_connection monitorSubscription_;
+    boost::signals2::scoped_connection monitorNewSeqSubscription_;
+    boost::signals2::scoped_connection monitorDbStalledSubscription_;
 
     std::optional<util::async::AnyOperation<void>> mainLoop_;
 
@@ -131,6 +136,7 @@ public:
      * @param loader Interface for loading data
      * @param initialLoadObserver The observer for initial data loading
      * @param taskManagerProvider The provider of the task manager instance
+     * @param monitorProvider The provider of the monitor instance
      * @param state System state tracking object
      */
     ETLService(
@@ -146,6 +152,7 @@ public:
         std::shared_ptr<LoaderInterface> loader,
         std::shared_ptr<InitialLoadObserverInterface> initialLoadObserver,
         std::shared_ptr<etlng::TaskManagerProviderInterface> taskManagerProvider,
+        std::shared_ptr<etlng::MonitorProviderInterface> monitorProvider,
         std::shared_ptr<etl::SystemState> state
     );
 
@@ -173,7 +180,6 @@ public:
     lastCloseAgeSeconds() const override;
 
 private:
-    // TODO: this better be std::expected
     std::optional<data::LedgerRange>
     loadInitialLedgerIfNeeded();
 
@@ -182,6 +188,12 @@ private:
 
     void
     startLoading(uint32_t seq);
+
+    void
+    attemptTakeoverWriter();
+
+    void
+    giveUpWriter();
 };
 
 }  // namespace etlng

src/etlng/LoadBalancer.cpp

@@ -45,7 +45,7 @@
 #include <boost/json/object.hpp>
 #include <boost/json/value.hpp>
 #include <boost/json/value_to.hpp>
-#include <fmt/core.h>
+#include <fmt/format.h>
 
 #include <algorithm>
 #include <chrono>
@@ -58,7 +58,6 @@
 #include <string>
 #include <thread>
 #include <utility>
-#include <variant>
 #include <vector>
 
 using namespace util::config;
@@ -184,11 +183,14 @@ LoadBalancer::LoadBalancer(
             LOG(log_.warn()) << "Failed to fetch ETL state from source = " << source->toString()
                              << " Please check the configuration and network";
         } else if (etlState_ && etlState_->networkID != stateOpt->networkID) {
-            checkOnETLFailure(fmt::format(
-                "ETL sources must be on the same network. Source network id = {} does not match others network id = {}",
-                stateOpt->networkID,
-                etlState_->networkID
-            ));
+            checkOnETLFailure(
+                fmt::format(
+                    "ETL sources must be on the same network. Source network id = {} does not match others network id "
+                    "= {}",
+                    stateOpt->networkID,
+                    etlState_->networkID
+                )
+            );
         } else {
             etlState_ = stateOpt;
         }
@@ -210,30 +212,32 @@ LoadBalancer::LoadBalancer(
     }
 }
 
-std::vector<std::string>
+InitialLedgerLoadResult
 LoadBalancer::loadInitialLedger(
     uint32_t sequence,
     etlng::InitialLoadObserverInterface& loadObserver,
     std::chrono::steady_clock::duration retryAfter
 )
 {
-    std::vector<std::string> response;
+    InitialLedgerLoadResult response;
+
     execute(
         [this, &response, &sequence, &loadObserver](auto& source) {
-            auto [data, res] = source->loadInitialLedger(sequence, downloadRanges_, loadObserver);
+            auto res = source->loadInitialLedger(sequence, downloadRanges_, loadObserver);
 
-            if (!res) {
+            if (not res.has_value() and res.error() == InitialLedgerLoadError::Errored) {
                 LOG(log_.error()) << "Failed to download initial ledger."
                                   << " Sequence = " << sequence << " source = " << source->toString();
-            } else {
-                response = std::move(data);
+                return false;  // should retry on error
             }
 
-            return res;
+            response = std::move(res);  // cancelled or data received
+            return true;
         },
         sequence,
         retryAfter
     );
+
     return response;
 }
 
@@ -279,42 +283,46 @@ LoadBalancer::forwardToRippled(
         return std::unexpected{rpc::ClioError::RpcCommandIsMissing};
 
     auto const cmd = boost::json::value_to<std::string>(request.at("command"));
-
-    if (forwardingCache_ and forwardingCache_->shouldCache(cmd)) {
-        bool servedFromCache = true;
-        auto updater =
-            [this, &request, &clientIp, &servedFromCache, isAdmin](boost::asio::yield_context yield
-            ) -> std::expected<util::ResponseExpirationCache::EntryData, util::ResponseExpirationCache::Error> {
-            servedFromCache = false;
-            auto result = forwardToRippledImpl(request, clientIp, isAdmin, yield);
-            if (result.has_value()) {
-                return util::ResponseExpirationCache::EntryData{
-                    .lastUpdated = std::chrono::steady_clock::now(), .response = std::move(result).value()
-                };
-            }
-            return std::unexpected{
-                util::ResponseExpirationCache::Error{.status = rpc::Status{result.error()}, .warnings = {}}
-            };
-        };
-
-        auto result = forwardingCache_->getOrUpdate(
-            yield,
-            cmd,
-            std::move(updater),
-            [](util::ResponseExpirationCache::EntryData const& entry) { return not entry.response.contains("error"); }
-        );
-        if (servedFromCache) {
-            ++forwardingCounters_.cacheHit.get();
+    if (forwardingCache_) {
+        if (auto cachedResponse = forwardingCache_->get(cmd); cachedResponse) {
+            forwardingCounters_.cacheHit.get() += 1;
+            return std::move(cachedResponse).value();
         }
-        if (result.has_value()) {
-            return std::move(result).value();
+    }
+    forwardingCounters_.cacheMiss.get() += 1;
+
+    ASSERT(not sources_.empty(), "ETL sources must be configured to forward requests.");
+    std::size_t sourceIdx = randomGenerator_->uniform(0ul, sources_.size() - 1);
+
+    auto numAttempts = 0u;
+
+    auto xUserValue = isAdmin ? kADMIN_FORWARDING_X_USER_VALUE : kUSER_FORWARDING_X_USER_VALUE;
+
+    std::optional<boost::json::object> response;
+    rpc::ClioError error = rpc::ClioError::EtlConnectionError;
+    while (numAttempts < sources_.size()) {
+        auto [res, duration] =
+            util::timed([&]() { return sources_[sourceIdx]->forwardToRippled(request, clientIp, xUserValue, yield); });
+        if (res) {
+            forwardingCounters_.successDuration.get() += duration;
+            response = std::move(res).value();
+            break;
         }
-        auto const combinedError = result.error().status.code;
-        ASSERT(std::holds_alternative<rpc::ClioError>(combinedError), "There could be only ClioError here");
-        return std::unexpected{std::get<rpc::ClioError>(combinedError)};
+        forwardingCounters_.failDuration.get() += duration;
+        ++forwardingCounters_.retries.get();
+        error = std::max(error, res.error());  // Choose the best result between all sources
+
+        sourceIdx = (sourceIdx + 1) % sources_.size();
+        ++numAttempts;
     }
 
-    return forwardToRippledImpl(request, clientIp, isAdmin, yield);
+    if (response) {
+        if (forwardingCache_ and not response->contains("error"))
+            forwardingCache_->put(cmd, *response);
+        return std::move(response).value();
+    }
+
+    return std::unexpected{error};
 }
 
 boost::json::value
@@ -405,47 +413,4 @@ LoadBalancer::chooseForwardingSource()
     }
 }
 
-std::expected<boost::json::object, rpc::CombinedError>
-LoadBalancer::forwardToRippledImpl(
-    boost::json::object const& request,
-    std::optional<std::string> const& clientIp,
-    bool isAdmin,
-    boost::asio::yield_context yield
-)
-{
-    ++forwardingCounters_.cacheMiss.get();
-
-    ASSERT(not sources_.empty(), "ETL sources must be configured to forward requests.");
-    std::size_t sourceIdx = randomGenerator_->uniform(0ul, sources_.size() - 1);
-
-    auto numAttempts = 0u;
-
-    auto xUserValue = isAdmin ? kADMIN_FORWARDING_X_USER_VALUE : kUSER_FORWARDING_X_USER_VALUE;
-
-    std::optional<boost::json::object> response;
-    rpc::ClioError error = rpc::ClioError::EtlConnectionError;
-    while (numAttempts < sources_.size()) {
-        auto [res, duration] =
-            util::timed([&]() { return sources_[sourceIdx]->forwardToRippled(request, clientIp, xUserValue, yield); });
-
-        if (res) {
-            forwardingCounters_.successDuration.get() += duration;
-            response = std::move(res).value();
-            break;
-        }
-        forwardingCounters_.failDuration.get() += duration;
-        ++forwardingCounters_.retries.get();
-        error = std::max(error, res.error());  // Choose the best result between all sources
-
-        sourceIdx = (sourceIdx + 1) % sources_.size();
-        ++numAttempts;
-    }
-
-    if (response.has_value()) {
-        return std::move(response).value();
-    }
-
-    return std::unexpected{error};
-}
-
 }  // namespace etlng

src/etlng/LoadBalancer.hpp

@@ -49,6 +49,7 @@
 #include <concepts>
 #include <cstdint>
 #include <expected>
+#include <functional>
 #include <memory>
 #include <optional>
 #include <string>
@@ -183,14 +184,14 @@ public:
 
     /**
      * @brief Load the initial ledger, writing data to the queue.
-     * @note This function will retry indefinitely until the ledger is downloaded.
+     * @note This function will retry indefinitely until the ledger is downloaded or the download is cancelled.
      *
      * @param sequence Sequence of ledger to download
      * @param observer The observer to notify of progress
      * @param retryAfter Time to wait between retries (2 seconds by default)
-     * @return A std::vector<std::string> The ledger data
+     * @return A std::expected with ledger edge keys on success, or InitialLedgerLoadError on failure
      */
-    std::vector<std::string>
+    InitialLedgerLoadResult
     loadInitialLedger(
         uint32_t sequence,
         etlng::InitialLoadObserverInterface& observer,
@@ -281,14 +282,6 @@ private:
      */
     void
     chooseForwardingSource();
-
-    std::expected<boost::json::object, rpc::CombinedError>
-    forwardToRippledImpl(
-        boost::json::object const& request,
-        std::optional<std::string> const& clientIp,
-        bool isAdmin,
-        boost::asio::yield_context yield
-    );
 };
 
 }  // namespace etlng

src/etlng/LoadBalancerInterface.hpp

@@ -39,6 +39,20 @@
 
 namespace etlng {
 
+/**
+ * @brief Represents possible errors for initial ledger load
+ */
+enum class InitialLedgerLoadError {
+    Cancelled, /*< Indicating the initial load got cancelled by user */
+    Errored,   /*< Indicating some error happened during initial ledger load */
+};
+
+/**
+ * @brief The result type of the initial ledger load
+ * @note The successful value represents edge keys
+ */
+using InitialLedgerLoadResult = std::expected<std::vector<std::string>, InitialLedgerLoadError>;
+
 /**
  * @brief An interface for LoadBalancer
  */
@@ -52,14 +66,14 @@ public:
 
     /**
      * @brief Load the initial ledger, writing data to the queue.
-     * @note This function will retry indefinitely until the ledger is downloaded.
+     * @note This function will retry indefinitely until the ledger is downloaded or the download is cancelled.
      *
      * @param sequence Sequence of ledger to download
      * @param loader InitialLoadObserverInterface implementation
      * @param retryAfter Time to wait between retries (2 seconds by default)
-     * @return A std::vector<std::string> The ledger data
+     * @return A std::expected with ledger edge keys on success, or InitialLedgerLoadError on failure
      */
-    virtual std::vector<std::string>
+    [[nodiscard]] virtual InitialLedgerLoadResult
     loadInitialLedger(
         uint32_t sequence,
         etlng::InitialLoadObserverInterface& loader,
@@ -74,7 +88,7 @@ public:
      * @param retryAfter Time to wait between retries (2 seconds by default)
      * @return A std::vector<std::string> The ledger data
      */
-    virtual std::vector<std::string>
+    [[nodiscard]] virtual std::vector<std::string>
     loadInitialLedger(uint32_t sequence, std::chrono::steady_clock::duration retryAfter = std::chrono::seconds{2}) = 0;
 
     /**
@@ -90,7 +104,7 @@ public:
      * @return The extracted data, if extraction was successful. If the ledger was found
      * in the database or the server is shutting down, the optional will be empty
      */
-    virtual OptionalGetLedgerResponseType
+    [[nodiscard]] virtual OptionalGetLedgerResponseType
     fetchLedger(
         uint32_t ledgerSequence,
         bool getObjects,
@@ -103,7 +117,7 @@ public:
      *
      * @return JSON representation of the state of this load balancer.
      */
-    virtual boost::json::value
+    [[nodiscard]] virtual boost::json::value
     toJson() const = 0;
 
     /**
@@ -115,7 +129,7 @@ public:
      * @param yield The coroutine context
      * @return Response received from rippled node as JSON object on success or error on failure
      */
-    virtual std::expected<boost::json::object, rpc::CombinedError>
+    [[nodiscard]] virtual std::expected<boost::json::object, rpc::CombinedError>
     forwardToRippled(
         boost::json::object const& request,
         std::optional<std::string> const& clientIp,
@@ -127,7 +141,7 @@ public:
      * @brief Return state of ETL nodes.
      * @return ETL state, nullopt if etl nodes not available
      */
-    virtual std::optional<etl::ETLState>
+    [[nodiscard]] virtual std::optional<etl::ETLState>
     getETLState() noexcept = 0;
 
     /**

src/etlng/LoaderInterface.hpp

@@ -23,10 +23,19 @@
 
 #include <xrpl/protocol/LedgerHeader.h>
 
+#include <expected>
 #include <optional>
 
 namespace etlng {
 
+/**
+ * @brief Enumeration of possible errors that can occur during loading operations
+ */
+enum class LoaderError {
+    AmendmentBlocked, /*< Error indicating that an operation is blocked by an amendment */
+    WriteConflict,    /*< Error indicating that a write operation resulted in a conflict */
+};
+
 /**
  * @brief An interface for a ETL Loader
  */
@@ -36,8 +45,9 @@ struct LoaderInterface {
     /**
      * @brief Load ledger data
      * @param data The data to load
+     * @return Nothing or error as std::expected
      */
-    virtual void
+    [[nodiscard]] virtual std::expected<void, LoaderError>
     load(model::LedgerData const& data) = 0;
 
     /**

src/etlng/Models.hpp

@@ -22,7 +22,7 @@
 #include "util/Concepts.hpp"
 
 #include <boost/json/object.hpp>
-#include <fmt/core.h>
+#include <fmt/format.h>
 #include <xrpl/basics/Blob.h>
 #include <xrpl/basics/base_uint.h>
 #include <xrpl/proto/org/xrpl/rpc/v1/get_ledger.pb.h>

src/etlng/MonitorInterface.hpp

@@ -36,7 +36,8 @@ namespace etlng {
 class MonitorInterface {
 public:
     static constexpr auto kDEFAULT_REPEAT_INTERVAL = std::chrono::seconds{1};
-    using SignalType = boost::signals2::signal<void(uint32_t)>;
+    using NewSequenceSignalType = boost::signals2::signal<void(uint32_t)>;
+    using DbStalledSignalType = boost::signals2::signal<void()>;
 
     virtual ~MonitorInterface() = default;
 
@@ -45,7 +46,14 @@ public:
      * @param seq The ledger sequence loaded
      */
     virtual void
-    notifyLedgerLoaded(uint32_t seq) = 0;
+    notifySequenceLoaded(uint32_t seq) = 0;
+
+    /**
+     * @brief Notifies the monitor of a write conflict
+     * @param seq The sequence number of the ledger that encountered a write conflict
+     */
+    virtual void
+    notifyWriteConflict(uint32_t seq) = 0;
 
     /**
      * @brief Allows clients to get notified when a new ledger becomes available in Clio's database
@@ -54,7 +62,16 @@ public:
      * @return A connection object that automatically disconnects the subscription once destroyed
      */
     [[nodiscard]] virtual boost::signals2::scoped_connection
-    subscribe(SignalType::slot_type const& subscriber) = 0;
+    subscribeToNewSequence(NewSequenceSignalType::slot_type const& subscriber) = 0;
+
+    /**
+     * @brief Allows clients to get notified when no database update is detected for a configured period.
+     *
+     * @param subscriber The slot to connect
+     * @return A connection object that automatically disconnects the subscription once destroyed
+     */
+    [[nodiscard]] virtual boost::signals2::scoped_connection
+    subscribeToDbStalled(DbStalledSignalType::slot_type const& subscriber) = 0;
 
     /**
      * @brief Run the monitor service

src/etlng/MonitorProviderInterface.hpp

@@ -0,0 +1,64 @@
+//------------------------------------------------------------------------------
+/*
+    This file is part of clio: https://github.com/XRPLF/clio
+    Copyright (c) 2025, the clio developers.
+
+    Permission to use, copy, modify, and distribute this software for any
+    purpose with or without fee is hereby granted, provided that the above
+    copyright notice and this permission notice appear in all copies.
+
+    THE  SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
+    WITH  REGARD  TO  THIS  SOFTWARE  INCLUDING  ALL  IMPLIED  WARRANTIES  OF
+    MERCHANTABILITY  AND  FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
+    ANY  SPECIAL,  DIRECT,  INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
+    WHATSOEVER  RESULTING  FROM  LOSS  OF USE, DATA OR PROFITS, WHETHER IN AN
+    ACTION  OF  CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
+    OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
+*/
+//==============================================================================
+
+#pragma once
+
+#include "data/BackendInterface.hpp"
+#include "etl/NetworkValidatedLedgersInterface.hpp"
+#include "etlng/MonitorInterface.hpp"
+#include "util/async/AnyExecutionContext.hpp"
+
+#include <chrono>
+#include <cstdint>
+#include <memory>
+
+namespace etlng {
+
+/**
+ * @brief An interface for providing Monitor instances
+ */
+struct MonitorProviderInterface {
+    /**
+     * @brief The time Monitor should wait before reporting absence of updates to the database
+     */
+    static constexpr auto kDEFAULT_DB_STALLED_REPORT_DELAY = std::chrono::seconds{10};
+
+    virtual ~MonitorProviderInterface() = default;
+
+    /**
+     * @brief Create a new Monitor instance
+     *
+     * @param ctx The execution context for asynchronous operations
+     * @param backend Interface to the backend database
+     * @param validatedLedgers Interface for accessing network validated ledgers
+     * @param startSequence The sequence number to start monitoring from
+     * @param dbStalledReportDelay The timeout duration after which to signal no database updates
+     * @return A unique pointer to a Monitor implementation
+     */
+    [[nodiscard]] virtual std::unique_ptr<MonitorInterface>
+    make(
+        util::async::AnyExecutionContext ctx,
+        std::shared_ptr<BackendInterface> backend,
+        std::shared_ptr<etl::NetworkValidatedLedgersInterface> validatedLedgers,
+        uint32_t startSequence,
+        std::chrono::steady_clock::duration dbStalledReportDelay = kDEFAULT_DB_STALLED_REPORT_DELAY
+    ) = 0;
+};
+
+}  // namespace etlng

src/etlng/Source.cpp

@@ -20,8 +20,8 @@
 #include "etlng/Source.hpp"
 
 #include "etl/NetworkValidatedLedgersInterface.hpp"
-#include "etl/impl/ForwardingSource.hpp"
 #include "etl/impl/SubscriptionSource.hpp"
+#include "etlng/impl/ForwardingSource.hpp"
 #include "etlng/impl/GrpcSource.hpp"
 #include "etlng/impl/SourceImpl.hpp"
 #include "feed/SubscriptionManagerInterface.hpp"
@@ -52,7 +52,7 @@ makeSource(
     auto const wsPort = config.get<std::string>("ws_port");
     auto const grpcPort = config.get<std::string>("grpc_port");
 
-    etl::impl::ForwardingSource forwardingSource{ip, wsPort, forwardingTimeout};
+    etlng::impl::ForwardingSource forwardingSource{ip, wsPort, forwardingTimeout};
     impl::GrpcSource grpcSource{ip, grpcPort};
     auto subscriptionSource = std::make_unique<etl::impl::SubscriptionSource>(
         ioc,

src/etlng/Source.hpp

@@ -19,9 +19,9 @@
 
 #pragma once
 
-#include "data/BackendInterface.hpp"
 #include "etl/NetworkValidatedLedgersInterface.hpp"
 #include "etlng/InitialLoadObserverInterface.hpp"
+#include "etlng/LoadBalancerInterface.hpp"
 #include "feed/SubscriptionManagerInterface.hpp"
 #include "rpc/Errors.hpp"
 #include "util/config/ObjectView.hpp"
@@ -131,7 +131,7 @@ public:
      * @param loader InitialLoadObserverInterface implementation
      * @return A std::pair of the data and a bool indicating whether the download was successful
      */
-    virtual std::pair<std::vector<std::string>, bool>
+    virtual InitialLedgerLoadResult
     loadInitialLedger(uint32_t sequence, std::uint32_t numMarkers, etlng::InitialLoadObserverInterface& loader) = 0;
 
     /**

src/etlng/TaskManagerProviderInterface.hpp

@@ -27,6 +27,7 @@
 #include <cstdint>
 #include <functional>
 #include <memory>
+#include <optional>
 
 namespace etlng {
 
@@ -41,11 +42,17 @@ struct TaskManagerProviderInterface {
      *
      * @param ctx The async context to associate the task manager instance with
      * @param monitor The monitor to notify when ledger is loaded
-     * @param seq The sequence to start at
+     * @param startSeq The sequence to start at
+     * @param finishSeq The sequence to stop at if specified
      * @return A unique pointer to a TaskManager implementation
      */
-    virtual std::unique_ptr<TaskManagerInterface>
-    make(util::async::AnyExecutionContext ctx, std::reference_wrapper<MonitorInterface> monitor, uint32_t seq) = 0;
+    [[nodiscard]] virtual std::unique_ptr<TaskManagerInterface>
+    make(
+        util::async::AnyExecutionContext ctx,
+        std::reference_wrapper<MonitorInterface> monitor,
+        uint32_t startSeq,
+        std::optional<uint32_t> finishSeq = std::nullopt
+    ) = 0;
 };
 
 }  // namespace etlng

src/etlng/impl/ForwardingSource.cpp

@@ -28,7 +28,7 @@
 #include <boost/json/object.hpp>
 #include <boost/json/parse.hpp>
 #include <boost/json/serialize.hpp>
-#include <fmt/core.h>
+#include <fmt/format.h>
 
 #include <chrono>
 #include <exception>

src/etlng/impl/GrpcSource.cpp

@@ -20,12 +20,14 @@
 #include "etlng/impl/GrpcSource.hpp"
 
 #include "etlng/InitialLoadObserverInterface.hpp"
+#include "etlng/LoadBalancerInterface.hpp"
 #include "etlng/impl/AsyncGrpcCall.hpp"
 #include "util/Assert.hpp"
 #include "util/log/Logger.hpp"
 #include "web/Resolver.hpp"
 
-#include <fmt/core.h>
+#include <boost/asio/spawn.hpp>
+#include <fmt/format.h>
 #include <grpcpp/client_context.h>
 #include <grpcpp/security/credentials.h>
 #include <grpcpp/support/channel_arguments.h>
@@ -33,9 +35,12 @@
 #include <org/xrpl/rpc/v1/get_ledger.pb.h>
 #include <org/xrpl/rpc/v1/xrp_ledger.grpc.pb.h>
 
+#include <atomic>
 #include <cstddef>
 #include <cstdint>
 #include <exception>
+#include <expected>
+#include <memory>
 #include <stdexcept>
 #include <string>
 #include <utility>
@@ -60,6 +65,7 @@ namespace etlng::impl {
 
 GrpcSource::GrpcSource(std::string const& ip, std::string const& grpcPort)
     : log_(fmt::format("ETL_Grpc[{}:{}]", ip, grpcPort))
+    , initialLoadShouldStop_(std::make_unique<std::atomic_bool>(false))
 {
     try {
         grpc::ChannelArguments chArgs;
@@ -103,15 +109,18 @@ GrpcSource::fetchLedger(uint32_t sequence, bool getObjects, bool getObjectNeighb
     return {status, std::move(response)};
 }
 
-std::pair<std::vector<std::string>, bool>
+InitialLedgerLoadResult
 GrpcSource::loadInitialLedger(
     uint32_t const sequence,
     uint32_t const numMarkers,
     etlng::InitialLoadObserverInterface& observer
 )
 {
+    if (*initialLoadShouldStop_)
+        return std::unexpected{InitialLedgerLoadError::Cancelled};
+
     if (!stub_)
-        return {{}, false};
+        return std::unexpected{InitialLedgerLoadError::Errored};
 
     std::vector<AsyncGrpcCall> calls = AsyncGrpcCall::makeAsyncCalls(sequence, numMarkers);
 
@@ -131,9 +140,9 @@ GrpcSource::loadInitialLedger(
         ASSERT(tag != nullptr, "Tag can't be null.");
         auto ptr = static_cast<AsyncGrpcCall*>(tag);
 
-        if (!ok) {
-            LOG(log_.error()) << "loadInitialLedger - ok is false";
-            return {{}, false};  // cancelled
+        if (not ok or *initialLoadShouldStop_) {
+            LOG(log_.error()) << "loadInitialLedger cancelled";
+            return std::unexpected{InitialLedgerLoadError::Cancelled};
         }
 
         LOG(log_.trace()) << "Marker prefix = " << ptr->getMarkerPrefix();
@@ -151,7 +160,16 @@ GrpcSource::loadInitialLedger(
             abort = true;
     }
 
-    return {std::move(edgeKeys), !abort};
+    if (abort)
+        return std::unexpected{InitialLedgerLoadError::Errored};
+
+    return edgeKeys;
+}
+
+void
+GrpcSource::stop(boost::asio::yield_context)
+{
+    initialLoadShouldStop_->store(true);
 }
 
 }  // namespace etlng::impl

src/etlng/impl/GrpcSource.hpp

@@ -20,23 +20,26 @@
 #pragma once
 
 #include "etlng/InitialLoadObserverInterface.hpp"
+#include "etlng/LoadBalancerInterface.hpp"
 #include "util/log/Logger.hpp"
 
+#include <boost/asio/spawn.hpp>
 #include <grpcpp/support/status.h>
 #include <org/xrpl/rpc/v1/get_ledger.pb.h>
 #include <xrpl/proto/org/xrpl/rpc/v1/xrp_ledger.grpc.pb.h>
 
+#include <atomic>
 #include <cstdint>
 #include <memory>
 #include <string>
 #include <utility>
-#include <vector>
 
 namespace etlng::impl {
 
 class GrpcSource {
     util::Logger log_;
     std::unique_ptr<org::xrpl::rpc::v1::XRPLedgerAPIService::Stub> stub_;
+    std::unique_ptr<std::atomic_bool> initialLoadShouldStop_;
 
 public:
     GrpcSource(std::string const& ip, std::string const& grpcPort);
@@ -61,10 +64,18 @@ public:
      * @param sequence Sequence of the ledger to download
      * @param numMarkers Number of markers to generate for async calls
      * @param observer InitialLoadObserverInterface implementation
-     * @return A std::pair of the data and a bool indicating whether the download was successful
+     * @return Downloaded data or an indication of error or cancellation
      */
-    std::pair<std::vector<std::string>, bool>
+    InitialLedgerLoadResult
     loadInitialLedger(uint32_t sequence, uint32_t numMarkers, etlng::InitialLoadObserverInterface& observer);
+
+    /**
+     * @brief Stop any ongoing operations
+     * @note This is used to cancel any ongoing initial ledger downloads
+     * @param yield The coroutine context
+     */
+    void
+    stop(boost::asio::yield_context yield);
 };
 
 }  // namespace etlng::impl

src/etlng/impl/LedgerPublisher.hpp

@@ -21,7 +21,6 @@
 
 #include "data/BackendInterface.hpp"
 #include "data/DBHelpers.hpp"
-#include "data/Types.hpp"
 #include "etl/SystemState.hpp"
 #include "etlng/LedgerPublisherInterface.hpp"
 #include "etlng/impl/Loading.hpp"
@@ -35,6 +34,7 @@
 #include <boost/asio/io_context.hpp>
 #include <boost/asio/post.hpp>
 #include <boost/asio/strand.hpp>
+#include <fmt/format.h>
 #include <xrpl/basics/chrono.h>
 #include <xrpl/protocol/Fees.h>
 #include <xrpl/protocol/LedgerHeader.h>
@@ -164,10 +164,6 @@ public:
         boost::asio::post(publishStrand_, [this, lgrInfo = lgrInfo]() {
             LOG(log_.info()) << "Publishing ledger " << std::to_string(lgrInfo.seq);
 
-            // TODO: This should probably not be part of publisher in the future
-            if (not state_.get().isWriting)
-                backend_->updateRange(lgrInfo.seq);  // This can't be unit tested atm.
-
             setLastClose(lgrInfo.closeTime);
             auto age = lastCloseAgeSeconds();
 
@@ -231,8 +227,9 @@ public:
     std::chrono::time_point<std::chrono::system_clock>
     getLastPublish() const override
     {
-        return std::chrono::time_point<std::chrono::system_clock>{std::chrono::seconds{lastPublishSeconds_.get().value()
-        }};
+        return std::chrono::time_point<std::chrono::system_clock>{
+            std::chrono::seconds{lastPublishSeconds_.get().value()}
+        };
     }
 
     /**

src/etlng/impl/Loading.cpp

@@ -20,11 +20,14 @@
 #include "etlng/impl/Loading.hpp"
 
 #include "data/BackendInterface.hpp"
+#include "etl/SystemState.hpp"
 #include "etl/impl/LedgerLoader.hpp"
 #include "etlng/AmendmentBlockHandlerInterface.hpp"
+#include "etlng/LoaderInterface.hpp"
 #include "etlng/Models.hpp"
 #include "etlng/RegistryInterface.hpp"
 #include "util/Assert.hpp"
+#include "util/Constants.hpp"
 #include "util/LedgerUtils.hpp"
 #include "util/Profiler.hpp"
 #include "util/log/Logger.hpp"
@@ -46,29 +49,45 @@ namespace etlng::impl {
 Loader::Loader(
     std::shared_ptr<BackendInterface> backend,
     std::shared_ptr<RegistryInterface> registry,
-    std::shared_ptr<AmendmentBlockHandlerInterface> amendmentBlockHandler
+    std::shared_ptr<AmendmentBlockHandlerInterface> amendmentBlockHandler,
+    std::shared_ptr<etl::SystemState> state
 )
     : backend_(std::move(backend))
     , registry_(std::move(registry))
     , amendmentBlockHandler_(std::move(amendmentBlockHandler))
+    , state_(std::move(state))
 {
 }
 
-void
+std::expected<void, LoaderError>
 Loader::load(model::LedgerData const& data)
 {
     try {
-        // perform cache updates and all writes from extensions
+        // Perform cache updates and all writes from extensions
+        // TODO: maybe this readonly logic should be removed?
         registry_->dispatch(data);
 
-        auto [success, duration] =
-            ::util::timed<std::chrono::duration<double>>([&]() { return backend_->finishWrites(data.seq); });
-        LOG(log_.info()) << "Finished writes to DB for " << data.seq << ": " << (success ? "YES" : "NO") << "; took "
-                         << duration;
+        // Only a writer should attempt to commit to DB
+        // This is also where conflicts with other writer nodes will be detected
+        if (state_->isWriting) {
+            auto [success, duration] =
+                ::util::timed<std::chrono::milliseconds>([&]() { return backend_->finishWrites(data.seq); });
+            LOG(log_.info()) << "Finished writes to DB for " << data.seq << ": " << (success ? "YES" : "NO")
+                             << "; took " << duration << "ms";
+
+            if (not success) {
+                state_->writeConflict = true;
+                LOG(log_.warn()) << "Another node wrote a ledger into the DB - we have a write conflict";
+                return std::unexpected(LoaderError::WriteConflict);
+            }
+        }
     } catch (std::runtime_error const& e) {
         LOG(log_.fatal()) << "Failed to load " << data.seq << ": " << e.what();
         amendmentBlockHandler_->notifyAmendmentBlocked();
+        return std::unexpected(LoaderError::AmendmentBlocked);
     }
+
+    return {};
 };
 
 void
@@ -78,13 +97,32 @@ Loader::onInitialLoadGotMoreObjects(
     std::optional<std::string> lastKey
 )
 {
+    static constexpr std::size_t kLOG_STRIDE = 1000u;
+    static auto kINITIAL_LOAD_START_TIME = std::chrono::steady_clock::now();
+
     try {
-        LOG(log_.debug()) << "On initial load: got more objects for seq " << seq << ". size = " << data.size();
+        LOG(log_.trace()) << "On initial load: got more objects for seq " << seq << ". size = " << data.size();
         registry_->dispatchInitialObjects(
             seq,
             data,
             std::move(lastKey).value_or(std::string{})  // TODO: perhaps use optional all the way to extensions?
         );
+
+        initialLoadWrittenObjects_ += data.size();
+        ++initialLoadWrites_;
+        if (initialLoadWrites_ % kLOG_STRIDE == 0u && initialLoadWrites_ != 0u) {
+            auto elapsedSinceStart = std::chrono::duration_cast<std::chrono::milliseconds>(
+                std::chrono::steady_clock::now() - kINITIAL_LOAD_START_TIME
+            );
+            auto elapsedSeconds = elapsedSinceStart.count() / static_cast<double>(util::kMILLISECONDS_PER_SECOND);
+            auto objectsPerSecond =
+                elapsedSeconds > 0.0 ? static_cast<double>(initialLoadWrittenObjects_) / elapsedSeconds : 0.0;
+
+            LOG(log_.info()) << "Wrote " << initialLoadWrittenObjects_
+                             << " initial ledger objects so far with average rate of " << objectsPerSecond
+                             << " objects per second";
+        }
+
     } catch (std::runtime_error const& e) {
         LOG(log_.fatal()) << "Failed to load initial objects for " << seq << ": " << e.what();
         amendmentBlockHandler_->notifyAmendmentBlocked();
@@ -95,9 +133,7 @@ std::optional<ripple::LedgerHeader>
 Loader::loadInitialLedger(model::LedgerData const& data)
 {
     try {
-        // check that database is actually empty
-        auto rng = backend_->hardFetchLedgerRangeNoThrow();
-        if (rng) {
+        if (auto const rng = backend_->hardFetchLedgerRangeNoThrow(); rng.has_value()) {
             ASSERT(false, "Database is not empty");
             return std::nullopt;
         }

src/etlng/impl/Loading.hpp

@@ -20,7 +20,7 @@
 #pragma once
 
 #include "data/BackendInterface.hpp"
-#include "etl/LedgerFetcherInterface.hpp"
+#include "etl/SystemState.hpp"
 #include "etl/impl/LedgerLoader.hpp"
 #include "etlng/AmendmentBlockHandlerInterface.hpp"
 #include "etlng/InitialLoadObserverInterface.hpp"
@@ -39,6 +39,7 @@
 #include <xrpl/protocol/Serializer.h>
 #include <xrpl/protocol/TxMeta.h>
 
+#include <cstddef>
 #include <cstdint>
 #include <memory>
 #include <optional>
@@ -51,7 +52,10 @@ class Loader : public LoaderInterface, public InitialLoadObserverInterface {
     std::shared_ptr<BackendInterface> backend_;
     std::shared_ptr<RegistryInterface> registry_;
     std::shared_ptr<AmendmentBlockHandlerInterface> amendmentBlockHandler_;
+    std::shared_ptr<etl::SystemState> state_;
 
+    std::size_t initialLoadWrittenObjects_{0u};
+    std::size_t initialLoadWrites_{0u};
     util::Logger log_{"ETL"};
 
 public:
@@ -62,7 +66,8 @@ public:
     Loader(
         std::shared_ptr<BackendInterface> backend,
         std::shared_ptr<RegistryInterface> registry,
-        std::shared_ptr<AmendmentBlockHandlerInterface> amendmentBlockHandler
+        std::shared_ptr<AmendmentBlockHandlerInterface> amendmentBlockHandler,
+        std::shared_ptr<etl::SystemState> state
     );
 
     Loader(Loader const&) = delete;
@@ -72,7 +77,7 @@ public:
     Loader&
     operator=(Loader&&) = delete;
 
-    void
+    std::expected<void, LoaderError>
     load(model::LedgerData const& data) override;
 
     void

src/etlng/impl/Monitor.cpp

@@ -23,11 +23,11 @@
 #include "etl/NetworkValidatedLedgersInterface.hpp"
 #include "util/Assert.hpp"
 #include "util/async/AnyExecutionContext.hpp"
-#include "util/async/AnyOperation.hpp"
 #include "util/log/Logger.hpp"
 
 #include <boost/signals2/connection.hpp>
 
+#include <algorithm>
 #include <chrono>
 #include <cstddef>
 #include <cstdint>
@@ -41,12 +41,18 @@ Monitor::Monitor(
     util::async::AnyExecutionContext ctx,
     std::shared_ptr<BackendInterface> backend,
     std::shared_ptr<etl::NetworkValidatedLedgersInterface> validatedLedgers,
-    uint32_t startSequence
+    uint32_t startSequence,
+    std::chrono::steady_clock::duration dbStalledReportDelay
 )
     : strand_(ctx.makeStrand())
     , backend_(std::move(backend))
     , validatedLedgers_(std::move(validatedLedgers))
     , nextSequence_(startSequence)
+    , updateData_({
+          .dbStalledReportDelay = dbStalledReportDelay,
+          .lastDbCheckTime = std::chrono::steady_clock::now(),
+          .lastSeenMaxSeqInDb = startSequence > 0 ? startSequence - 1 : 0,
+      })
 {
 }
 
@@ -55,20 +61,37 @@ Monitor::~Monitor()
     stop();
 }
 
-// TODO: think about using signals perhaps? maybe combining with onNextSequence?
-// also, how do we not double invoke or does it not matter
 void
-Monitor::notifyLedgerLoaded(uint32_t seq)
+Monitor::notifySequenceLoaded(uint32_t seq)
 {
-    LOG(log_.debug()) << "Loader notified about newly committed ledger " << seq;
-    repeatedTask_->invoke();  // force-invoke immediately
+    LOG(log_.debug()) << "Loader notified Monitor about newly committed ledger " << seq;
+    {
+        auto lck = updateData_.lock();
+        lck->lastSeenMaxSeqInDb = std::max(seq, lck->lastSeenMaxSeqInDb);
+        lck->lastDbCheckTime = std::chrono::steady_clock::now();
+    }
+    repeatedTask_->invoke();  // force-invoke doWork immediately
 };
 
+void
+Monitor::notifyWriteConflict(uint32_t seq)
+{
+    LOG(log_.warn()) << "Loader notified Monitor about write conflict at " << seq;
+    nextSequence_ = seq + 1;  //  we already loaded the cache for seq just before we detected conflict
+    LOG(log_.warn()) << "Resume monitoring from " << nextSequence_;
+}
+
 void
 Monitor::run(std::chrono::steady_clock::duration repeatInterval)
 {
     ASSERT(not repeatedTask_.has_value(), "Monitor attempted to run more than once");
-    LOG(log_.debug()) << "Starting monitor";
+    {
+        auto lck = updateData_.lock();
+        LOG(log_.debug()) << "Starting monitor with repeat interval: "
+                          << std::chrono::duration_cast<std::chrono::seconds>(repeatInterval).count()
+                          << "s and dbStalledReportDelay: "
+                          << std::chrono::duration_cast<std::chrono::seconds>(lck->dbStalledReportDelay).count() << "s";
+    }
 
     repeatedTask_ = strand_.executeRepeatedly(repeatInterval, std::bind_front(&Monitor::doWork, this));
     subscription_ = validatedLedgers_->subscribe(std::bind_front(&Monitor::onNextSequence, this));
@@ -80,28 +103,65 @@ Monitor::stop()
     if (repeatedTask_.has_value())
         repeatedTask_->abort();
 
+    subscription_ = std::nullopt;
     repeatedTask_ = std::nullopt;
 }
 
 boost::signals2::scoped_connection
-Monitor::subscribe(SignalType::slot_type const& subscriber)
+Monitor::subscribeToNewSequence(NewSequenceSignalType::slot_type const& subscriber)
 {
     return notificationChannel_.connect(subscriber);
 }
 
+boost::signals2::scoped_connection
+Monitor::subscribeToDbStalled(DbStalledSignalType::slot_type const& subscriber)
+{
+    return dbStalledChannel_.connect(subscriber);
+}
+
 void
 Monitor::onNextSequence(uint32_t seq)
 {
-    LOG(log_.debug()) << "rippled published sequence " << seq;
+    ASSERT(repeatedTask_.has_value(), "Ledger subscription without repeated task is a logic error");
+    LOG(log_.debug()) << "Notified about new sequence on the network: " << seq;
     repeatedTask_->invoke();  // force-invoke immediately
 }
 
 void
 Monitor::doWork()
 {
-    if (auto rng = backend_->hardFetchLedgerRangeNoThrow(); rng) {
-        while (rng->maxSequence >= nextSequence_)
+    auto rng = backend_->hardFetchLedgerRangeNoThrow();
+    bool dbProgressedThisCycle = false;
+    auto lck = updateData_.lock();
+
+    if (rng.has_value()) {
+        if (rng->maxSequence > lck->lastSeenMaxSeqInDb) {
+            LOG(log_.trace()) << "DB progressed. Old max seq = " << lck->lastSeenMaxSeqInDb
+                              << ", new max seq = " << rng->maxSequence;
+            lck->lastSeenMaxSeqInDb = rng->maxSequence;
+            dbProgressedThisCycle = true;
+        }
+
+        while (lck->lastSeenMaxSeqInDb >= nextSequence_) {
+            LOG(log_.trace()) << "Publishing from Monitor::doWork. nextSequence_ = " << nextSequence_
+                              << ", lastSeenMaxSeqInDb_ = " << lck->lastSeenMaxSeqInDb;
             notificationChannel_(nextSequence_++);
+            dbProgressedThisCycle = true;
+        }
+    } else {
+        LOG(log_.trace()) << "DB range is not available or empty. lastSeenMaxSeqInDb_ = " << lck->lastSeenMaxSeqInDb
+                          << ", nextSequence_ = " << nextSequence_;
+    }
+
+    if (dbProgressedThisCycle) {
+        lck->lastDbCheckTime = std::chrono::steady_clock::now();
+    } else if (std::chrono::steady_clock::now() - lck->lastDbCheckTime > lck->dbStalledReportDelay) {
+        LOG(log_.info()) << "No DB update detected for "
+                         << std::chrono::duration_cast<std::chrono::seconds>(lck->dbStalledReportDelay).count()
+                         << " seconds. Firing dbStalledChannel. Last seen max seq in DB: " << lck->lastSeenMaxSeqInDb
+                         << ". Expecting next: " << nextSequence_;
+        dbStalledChannel_();
+        lck->lastDbCheckTime = std::chrono::steady_clock::now();
     }
 }
 

src/etlng/impl/Monitor.hpp

@@ -22,6 +22,7 @@
 #include "data/BackendInterface.hpp"
 #include "etl/NetworkValidatedLedgersInterface.hpp"
 #include "etlng/MonitorInterface.hpp"
+#include "util/Mutex.hpp"
 #include "util/async/AnyExecutionContext.hpp"
 #include "util/async/AnyOperation.hpp"
 #include "util/async/AnyStrand.hpp"
@@ -30,6 +31,7 @@
 #include <boost/signals2/connection.hpp>
 #include <xrpl/protocol/TxFormats.h>
 
+#include <atomic>
 #include <chrono>
 #include <cstddef>
 #include <cstdint>
@@ -43,11 +45,20 @@ class Monitor : public MonitorInterface {
     std::shared_ptr<BackendInterface> backend_;
     std::shared_ptr<etl::NetworkValidatedLedgersInterface> validatedLedgers_;
 
-    uint32_t nextSequence_;
+    std::atomic_uint32_t nextSequence_;
     std::optional<util::async::AnyOperation<void>> repeatedTask_;
     std::optional<boost::signals2::scoped_connection> subscription_;  // network validated ledgers subscription
 
-    SignalType notificationChannel_;
+    NewSequenceSignalType notificationChannel_;
+    DbStalledSignalType dbStalledChannel_;
+
+    struct UpdateData {
+        std::chrono::steady_clock::duration dbStalledReportDelay;
+        std::chrono::steady_clock::time_point lastDbCheckTime;
+        uint32_t lastSeenMaxSeqInDb = 0u;
+    };
+
+    util::Mutex<UpdateData> updateData_;
 
     util::Logger log_{"ETL"};
 
@@ -56,12 +67,16 @@ public:
         util::async::AnyExecutionContext ctx,
         std::shared_ptr<BackendInterface> backend,
         std::shared_ptr<etl::NetworkValidatedLedgersInterface> validatedLedgers,
-        uint32_t startSequence
+        uint32_t startSequence,
+        std::chrono::steady_clock::duration dbStalledReportDelay
     );
     ~Monitor() override;
 
     void
-    notifyLedgerLoaded(uint32_t seq) override;
+    notifySequenceLoaded(uint32_t seq) override;
+
+    void
+    notifyWriteConflict(uint32_t seq) override;
 
     void
     run(std::chrono::steady_clock::duration repeatInterval) override;
@@ -70,7 +85,10 @@ public:
     stop() override;
 
     boost::signals2::scoped_connection
-    subscribe(SignalType::slot_type const& subscriber) override;
+    subscribeToNewSequence(NewSequenceSignalType::slot_type const& subscriber) override;
+
+    boost::signals2::scoped_connection
+    subscribeToDbStalled(DbStalledSignalType::slot_type const& subscriber) override;
 
 private:
     void

src/etlng/impl/MonitorProvider.hpp

@@ -0,0 +1,53 @@
+//------------------------------------------------------------------------------
+/*
+    This file is part of clio: https://github.com/XRPLF/clio
+    Copyright (c) 2025, the clio developers.
+
+    Permission to use, copy, modify, and distribute this software for any
+    purpose with or without fee is hereby granted, provided that the above
+    copyright notice and this permission notice appear in all copies.
+
+    THE  SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
+    WITH  REGARD  TO  THIS  SOFTWARE  INCLUDING  ALL  IMPLIED  WARRANTIES  OF
+    MERCHANTABILITY  AND  FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
+    ANY  SPECIAL,  DIRECT,  INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
+    WHATSOEVER  RESULTING  FROM  LOSS  OF USE, DATA OR PROFITS, WHETHER IN AN
+    ACTION  OF  CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
+    OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
+*/
+//==============================================================================
+
+#pragma once
+
+#include "data/BackendInterface.hpp"
+#include "etl/NetworkValidatedLedgersInterface.hpp"
+#include "etlng/MonitorInterface.hpp"
+#include "etlng/MonitorProviderInterface.hpp"
+#include "etlng/impl/Monitor.hpp"
+#include "util/async/AnyExecutionContext.hpp"
+
+#include <chrono>
+#include <cstdint>
+#include <memory>
+#include <utility>
+
+namespace etlng::impl {
+
+class MonitorProvider : public MonitorProviderInterface {
+public:
+    std::unique_ptr<MonitorInterface>
+    make(
+        util::async::AnyExecutionContext ctx,
+        std::shared_ptr<BackendInterface> backend,
+        std::shared_ptr<etl::NetworkValidatedLedgersInterface> validatedLedgers,
+        uint32_t startSequence,
+        std::chrono::steady_clock::duration dbStalledReportDelay
+    ) override
+    {
+        return std::make_unique<Monitor>(
+            std::move(ctx), std::move(backend), std::move(validatedLedgers), startSequence, dbStalledReportDelay
+        );
+    }
+};
+
+}  // namespace etlng::impl

src/etlng/impl/SourceImpl.hpp

@@ -19,10 +19,11 @@
 
 #pragma once
 
-#include "etl/impl/ForwardingSource.hpp"
 #include "etl/impl/SubscriptionSource.hpp"
 #include "etlng/InitialLoadObserverInterface.hpp"
+#include "etlng/LoadBalancerInterface.hpp"
 #include "etlng/Source.hpp"
+#include "etlng/impl/ForwardingSource.hpp"
 #include "etlng/impl/GrpcSource.hpp"
 #include "rpc/Errors.hpp"
 
@@ -53,7 +54,7 @@ namespace etlng::impl {
 template <
     typename GrpcSourceType = GrpcSource,
     typename SubscriptionSourceTypePtr = std::unique_ptr<etl::impl::SubscriptionSource>,
-    typename ForwardingSourceType = etl::impl::ForwardingSource>
+    typename ForwardingSourceType = etlng::impl::ForwardingSource>
 class SourceImpl : public SourceBase {
     std::string ip_;
     std::string wsPort_;
@@ -107,6 +108,7 @@ public:
     stop(boost::asio::yield_context yield) final
     {
         subscriptionSource_->stop(yield);
+        grpcSource_.stop(yield);
     }
 
     /**
@@ -202,7 +204,7 @@ public:
      * @param loader InitialLoadObserverInterface implementation
      * @return A std::pair of the data and a bool indicating whether the download was successful
      */
-    std::pair<std::vector<std::string>, bool>
+    InitialLedgerLoadResult
     loadInitialLedger(uint32_t sequence, std::uint32_t numMarkers, etlng::InitialLoadObserverInterface& loader) final
     {
         return grpcSource_.loadInitialLedger(sequence, numMarkers, loader);

src/etlng/impl/TaskManager.cpp

@@ -26,6 +26,7 @@
 #include "etlng/SchedulerInterface.hpp"
 #include "etlng/impl/Monitor.hpp"
 #include "etlng/impl/TaskQueue.hpp"
+#include "util/Constants.hpp"
 #include "util/LedgerUtils.hpp"
 #include "util/Profiler.hpp"
 #include "util/async/AnyExecutionContext.hpp"
@@ -102,29 +103,49 @@ TaskManager::spawnExtractor(TaskQueue& queue)
                         if (stopRequested)
                             break;
                     }
-                } else {
-                    // TODO: how do we signal to the loaders that it's time to shutdown? some special task?
-                    break;  // TODO: handle server shutdown or other node took over ETL
                 }
             } else {
                 // TODO (https://github.com/XRPLF/clio/issues/1852)
                 std::this_thread::sleep_for(kDELAY_BETWEEN_ATTEMPTS);
             }
         }
+
+        LOG(log_.info()) << "Extractor (one of) coroutine stopped";
     });
 }
 
 util::async::AnyOperation<void>
 TaskManager::spawnLoader(TaskQueue& queue)
 {
-    static constexpr auto kNANO_TO_SECOND = 1.0e9;
-
     return ctx_.execute([this, &queue](auto stopRequested) {
         while (not stopRequested) {
             // TODO (https://github.com/XRPLF/clio/issues/66): does not tell the loader whether it's out of order or not
             if (auto data = queue.dequeue(); data.has_value()) {
-                auto nanos = util::timed<std::chrono::nanoseconds>([this, data = *data] { loader_.get().load(data); });
-                auto const seconds = nanos / kNANO_TO_SECOND;
+                auto [expectedSuccess, nanos] =
+                    util::timed<std::chrono::nanoseconds>([&] { return loader_.get().load(*data); });
+
+                auto const shouldExitOnError = [&] {
+                    if (expectedSuccess.has_value())
+                        return false;
+
+                    switch (expectedSuccess.error()) {
+                        case LoaderError::WriteConflict:
+                            LOG(log_.warn()) << "Immediately stopping loader on write conflict"
+                                             << "; latest ledger cache loaded for " << data->seq;
+                            monitor_.get().notifyWriteConflict(data->seq);
+                            return true;
+                        case LoaderError::AmendmentBlocked:
+                            LOG(log_.warn()) << "Immediately stopping loader on amendment block";
+                            return true;
+                    }
+
+                    std::unreachable();
+                }();
+
+                if (shouldExitOnError)
+                    break;
+
+                auto const seconds = nanos / util::kNANO_PER_SECOND;
                 auto const txnCount = data->transactions.size();
                 auto const objCount = data->objects.size();
 
@@ -133,9 +154,11 @@ TaskManager::spawnLoader(TaskQueue& queue)
                                  << " seconds;"
                                  << " tps[" << txnCount / seconds << "], ops[" << objCount / seconds << "]";
 
-                monitor_.get().notifyLedgerLoaded(data->seq);
+                monitor_.get().notifySequenceLoaded(data->seq);
             }
         }
+
+        LOG(log_.info()) << "Loader coroutine stopped";
     });
 }