style: Fix hadolint issues (#2777)

Fixes #2413.

.clang-format

@@ -49,6 +49,7 @@ IndentFunctionDeclarationAfterType: false
 IndentWidth: 4
 IndentWrappedFunctionNames: false
 IndentRequiresClause: true
+InsertNewlineAtEOF: true
 RequiresClausePosition: OwnLine
 KeepEmptyLinesAtTheStartOfBlocks: false
 MaxEmptyLinesToKeep: 1

.cmake-format.yaml

@@ -54,7 +54,7 @@ format:
   _help_max_pargs_hwrap:
     - If a positional argument group contains more than this many
     - arguments, then force it to a vertical layout.
-  max_pargs_hwrap: 6
+  max_pargs_hwrap: 5
   _help_max_rows_cmdline:
     - If a cmdline positional group consumes more than this many
     - lines without nesting, then invalidate the layout (and nest)

.github/actions/build-clio/action.yml

@@ -5,25 +5,27 @@ inputs:
   targets:
     description: Space-separated build target names
     default: all
-  subtract_threads:
-    description: An option for the action get-threads-number.
+  nproc_subtract:
+    description: The number of processors to subtract when calculating parallelism.
     required: true
     default: "0"
 
 runs:
   using: composite
   steps:
-    - name: Get number of threads
-      uses: ./.github/actions/get-threads-number
-      id: number_of_threads
+    - name: Get number of processors
+      uses: XRPLF/actions/.github/actions/get-nproc@046b1620f6bfd6cd0985dc82c3df02786801fe0a
+      id: nproc
       with:
-        subtract_threads: ${{ inputs.subtract_threads }}
+        subtract: ${{ inputs.nproc_subtract }}
 
     - name: Build targets
       shell: bash
+      env:
+        CMAKE_TARGETS: ${{ inputs.targets }}
       run: |
         cd build
         cmake \
           --build . \
-          --parallel "${{ steps.number_of_threads.outputs.threads_number }}" \
-          --target ${{ inputs.targets }}
+          --parallel "${{ steps.nproc.outputs.nproc }}" \
+          --target ${CMAKE_TARGETS}

.github/actions/build-docker-image/action.yml

@@ -47,12 +47,12 @@ runs:
         username: ${{ github.repository_owner }}
         password: ${{ env.GITHUB_TOKEN }}
 
-    - uses: docker/setup-qemu-action@29109295f81e9208d7d86ff1c6c12d2833863392 # v3.6.0
+    - uses: docker/setup-qemu-action@c7c53464625b32c7a7e944ae62b3e17d2b600130 # v3.7.0
       with:
         cache-image: false
     - uses: docker/setup-buildx-action@e468171a9de216ec08956ac3ada2f0791b6bd435 # v3.11.1
 
-    - uses: docker/metadata-action@c1e51972afc2121e065aed6d45c65596fe445f3f # v5.8.0
+    - uses: docker/metadata-action@318604b99e75e41977312d83839a89be02ca4893 # v5.9.0
       id: meta
       with:
         images: ${{ inputs.images }}

.github/actions/code-coverage/action.yml

@@ -24,7 +24,7 @@ runs:
           -j8 --exclude-throw-branches
 
     - name: Archive coverage report
-      uses: actions/upload-artifact@v4
+      uses: actions/upload-artifact@330a01c490aca151604b8cf639adc76d48f6c5d4 # v5.0.0
       with:
         name: coverage-report.xml
         path: build/coverage_report.xml

.github/actions/create-issue/action.yml

@@ -28,12 +28,17 @@ runs:
     - name: Create an issue
       id: create_issue
       shell: bash
+      env:
+        ISSUE_BODY: ${{ inputs.body }}
+        ISSUE_ASSIGNEES: ${{ inputs.assignees }}
+        ISSUE_LABELS: ${{ inputs.labels }}
+        ISSUE_TITLE: ${{ inputs.title }}
       run: |
-        echo -e '${{ inputs.body }}' > issue.md
+        echo -e "${ISSUE_BODY}" > issue.md
         gh issue create \
-          --assignee '${{ inputs.assignees }}' \
-          --label '${{ inputs.labels }}' \
-          --title '${{ inputs.title }}' \
+          --assignee "${ISSUE_ASSIGNEES}" \
+          --label "${ISSUE_LABELS}" \
+          --title "${ISSUE_TITLE}" \
           --body-file ./issue.md \
           > create_issue.log
         created_issue="$(sed 's|.*/||' create_issue.log)"

.github/actions/get-threads-number/action.yml

@@ -1,36 +0,0 @@
-name: Get number of threads
-description: Determines number of threads to use on macOS and Linux
-
-inputs:
-  subtract_threads:
-    description: How many threads to subtract from the calculated number
-    required: true
-    default: "0"
-outputs:
-  threads_number:
-    description: Number of threads to use
-    value: ${{ steps.number_of_threads_export.outputs.num }}
-
-runs:
-  using: composite
-  steps:
-    - name: Get number of threads on mac
-      id: mac_threads
-      if: ${{ runner.os == 'macOS' }}
-      shell: bash
-      run: echo "num=$(($(sysctl -n hw.logicalcpu) - 2))" >> $GITHUB_OUTPUT
-
-    - name: Get number of threads on Linux
-      id: linux_threads
-      if: ${{ runner.os == 'Linux' }}
-      shell: bash
-      run: echo "num=$(($(nproc) - 2))" >> $GITHUB_OUTPUT
-
-    - name: Shift and export number of threads
-      id: number_of_threads_export
-      shell: bash
-      run: |
-        num_of_threads="${{ steps.mac_threads.outputs.num || steps.linux_threads.outputs.num }}"
-        shift_by="${{ inputs.subtract_threads }}"
-        shifted="$((num_of_threads - shift_by))"
-        echo "num=$(( shifted > 1 ? shifted : 1 ))" >> $GITHUB_OUTPUT

.github/actions/restore-cache/action.yml

@@ -30,7 +30,7 @@ runs:
       uses: ./.github/actions/git-common-ancestor
 
     - name: Restore ccache cache
-      uses: actions/cache/restore@v4
+      uses: actions/cache/restore@0057852bfaa89a56745cba8c7296529d2fc39830 # v4.3.0
       id: ccache_cache
       if: ${{ env.CCACHE_DISABLE != '1' }}
       with:

.github/actions/save-cache/action.yml

@@ -32,7 +32,7 @@ runs:
 
     - name: Save ccache cache
       if: ${{ inputs.ccache_cache_hit != 'true' || inputs.ccache_cache_miss_rate == '100.0' }}
-      uses: actions/cache/save@v4
+      uses: actions/cache/save@0057852bfaa89a56745cba8c7296529d2fc39830 # v4.3.0
       with:
         path: ${{ inputs.ccache_dir }}
         key: clio-ccache-${{ runner.os }}-${{ inputs.build_type }}${{ inputs.code_coverage == 'true' && '-code_coverage' || '' }}-${{ inputs.conan_profile }}-develop-${{ steps.git_common_ancestor.outputs.commit }}

.github/dependabot.yml

@@ -91,19 +91,6 @@ updates:
       prefix: "ci: [DEPENDABOT] "
     target-branch: develop
 
-  - package-ecosystem: github-actions
-    directory: .github/actions/get-threads-number/
-    schedule:
-      interval: weekly
-      day: monday
-      time: "04:00"
-      timezone: Etc/GMT
-    reviewers:
-      - XRPLF/clio-dev-team
-    commit-message:
-      prefix: "ci: [DEPENDABOT] "
-    target-branch: develop
-
   - package-ecosystem: github-actions
     directory: .github/actions/git-common-ancestor/
     schedule:

.github/scripts/conan/apple-clang-17.profile

@@ -4,7 +4,7 @@ build_type=Release
 compiler=apple-clang
 compiler.cppstd=20
 compiler.libcxx=libc++
-compiler.version=17
+compiler.version=17.0
 os=Macos
 
 [conf]

.github/scripts/conan/generate_matrix.py

@@ -4,7 +4,7 @@ import json
 
 LINUX_OS = ["heavy", "heavy-arm64"]
 LINUX_CONTAINERS = [
-    '{ "image": "ghcr.io/xrplf/clio-ci:213752862ca95ecadeb59a6176c3db91a7864b3e" }'
+    '{ "image": "ghcr.io/xrplf/clio-ci:62369411404eb32b0140603a785ff05e1dc36ce8" }'
 ]
 LINUX_COMPILERS = ["gcc", "clang"]
 

.github/scripts/execute-tests-under-sanitizer.sh

@@ -31,15 +31,16 @@ TESTS=$($TEST_BINARY --gtest_list_tests | awk '/^  / {print suite $1} !/^  / {su
 OUTPUT_DIR="./.sanitizer-report"
 mkdir -p "$OUTPUT_DIR"
 
-for TEST in $TESTS; do
-  OUTPUT_FILE="$OUTPUT_DIR/${TEST//\//_}"
-  export TSAN_OPTIONS="log_path=\"$OUTPUT_FILE\" die_after_fork=0"
-  export ASAN_OPTIONS="log_path=\"$OUTPUT_FILE\""
-  export UBSAN_OPTIONS="log_path=\"$OUTPUT_FILE\""
+export TSAN_OPTIONS="die_after_fork=0"
 export MallocNanoZone='0' # for MacOSX
-  $TEST_BINARY --gtest_filter="$TEST" > /dev/null 2>&1
+
+for TEST in $TESTS; do
+  OUTPUT_FILE="$OUTPUT_DIR/${TEST//\//_}.log"
+  $TEST_BINARY --gtest_filter="$TEST" > "$OUTPUT_FILE" 2>&1
 
   if [ $? -ne 0 ]; then
     echo "'$TEST' failed a sanitizer check."
+  else
+    rm "$OUTPUT_FILE"
   fi
 done

.github/workflows/build-clio-docker-image.yml

@@ -38,32 +38,37 @@ on:
         description: Whether to strip clio binary
         default: true
 
+defaults:
+  run:
+    shell: bash
+
 jobs:
   build_and_publish_image:
     name: Build and publish image
     runs-on: ubuntu-latest
 
     steps:
-      - uses: actions/checkout@v4
+      - uses: actions/checkout@08c6903cd8c0fde910a37f88322edcfb5dd907a8 # v5.0.0
 
       - name: Download Clio binary from artifact
         if: ${{ inputs.artifact_name != null }}
-        uses: actions/download-artifact@v5
+        uses: actions/download-artifact@018cc2cf5baa6db3ef3c5f8a56943fffe632ef53 # v6.0.0
         with:
           name: ${{ inputs.artifact_name }}
           path: ./docker/clio/artifact/
 
       - name: Download Clio binary from url
         if: ${{ inputs.clio_server_binary_url != null }}
-        shell: bash
+        env:
+          BINARY_URL: ${{ inputs.clio_server_binary_url }}
+          BINARY_SHA256: ${{ inputs.binary_sha256 }}
         run: |
-          wget "${{inputs.clio_server_binary_url}}" -P ./docker/clio/artifact/
-          if [ "$(sha256sum ./docker/clio/clio_server | awk '{print $1}')" != "${{inputs.binary_sha256}}" ]; then
+          wget "${BINARY_URL}" -P ./docker/clio/artifact/
+          if [ "$(sha256sum ./docker/clio/clio_server | awk '{print $1}')" != "${BINARY_SHA256}" ]; then
             echo "Binary sha256 sum doesn't match"
             exit 1
           fi
       - name: Unpack binary
-        shell: bash
         run: |
           sudo apt update && sudo apt install -y tar unzip
           cd docker/clio/artifact
@@ -80,7 +85,6 @@ jobs:
 
       - name: Strip binary
         if: ${{ inputs.strip_binary }}
-        shell: bash
         run: strip ./docker/clio/clio_server
 
       - name: Set GHCR_REPO

.github/workflows/build.yml

@@ -33,6 +33,10 @@ concurrency:
   group: ${{ github.workflow }}-${{ github.ref }}-${{ github.ref == 'refs/heads/develop' && github.run_number || 'branch' }}
   cancel-in-progress: true
 
+defaults:
+  run:
+    shell: bash
+
 jobs:
   build-and-test:
     name: Build and Test
@@ -45,7 +49,7 @@ jobs:
         build_type: [Release, Debug]
         container:
           [
-            '{ "image": "ghcr.io/xrplf/clio-ci:213752862ca95ecadeb59a6176c3db91a7864b3e" }',
+            '{ "image": "ghcr.io/xrplf/clio-ci:62369411404eb32b0140603a785ff05e1dc36ce8" }',
           ]
         static: [true]
 
@@ -75,7 +79,7 @@ jobs:
     uses: ./.github/workflows/reusable-build.yml
     with:
       runs_on: heavy
-      container: '{ "image": "ghcr.io/xrplf/clio-ci:213752862ca95ecadeb59a6176c3db91a7864b3e" }'
+      container: '{ "image": "ghcr.io/xrplf/clio-ci:62369411404eb32b0140603a785ff05e1dc36ce8" }'
       conan_profile: gcc
       build_type: Debug
       download_ccache: true
@@ -94,7 +98,7 @@ jobs:
     uses: ./.github/workflows/reusable-build.yml
     with:
       runs_on: heavy
-      container: '{ "image": "ghcr.io/xrplf/clio-ci:213752862ca95ecadeb59a6176c3db91a7864b3e" }'
+      container: '{ "image": "ghcr.io/xrplf/clio-ci:62369411404eb32b0140603a785ff05e1dc36ce8" }'
       conan_profile: gcc
       build_type: Release
       download_ccache: true
@@ -111,17 +115,16 @@ jobs:
     needs: build-and-test
     runs-on: heavy
     container:
-      image: ghcr.io/xrplf/clio-ci:213752862ca95ecadeb59a6176c3db91a7864b3e
+      image: ghcr.io/xrplf/clio-ci:62369411404eb32b0140603a785ff05e1dc36ce8
 
     steps:
-      - uses: actions/checkout@v4
+      - uses: actions/checkout@08c6903cd8c0fde910a37f88322edcfb5dd907a8 # v5.0.0
 
-      - uses: actions/download-artifact@v5
+      - uses: actions/download-artifact@018cc2cf5baa6db3ef3c5f8a56943fffe632ef53 # v6.0.0
         with:
           name: clio_server_Linux_Release_gcc
 
       - name: Compare Config Description
-        shell: bash
         run: |
           repoConfigFile=docs/config-description.md
           configDescriptionFile=config_description_new.md

.github/workflows/check-libxrpl.yml

@@ -12,31 +12,33 @@ concurrency:
 env:
   CONAN_PROFILE: gcc
 
+defaults:
+  run:
+    shell: bash
+
 jobs:
   build:
     name: Build Clio / `libXRPL ${{ github.event.client_payload.version }}`
     runs-on: heavy
     container:
-      image: ghcr.io/xrplf/clio-ci:213752862ca95ecadeb59a6176c3db91a7864b3e
+      image: ghcr.io/xrplf/clio-ci:62369411404eb32b0140603a785ff05e1dc36ce8
 
     steps:
-      - uses: actions/checkout@v4
+      - uses: actions/checkout@08c6903cd8c0fde910a37f88322edcfb5dd907a8 # v5.0.0
         with:
           fetch-depth: 0
 
       - name: Prepare runner
-        uses: XRPLF/actions/.github/actions/prepare-runner@7951b682e5a2973b28b0719a72f01fc4b0d0c34f
+        uses: XRPLF/actions/.github/actions/prepare-runner@8abb0722cbff83a9a2dc7d06c473f7a4964b7382
         with:
           disable_ccache: true
 
       - name: Update libXRPL version requirement
-        shell: bash
         run: |
           sed -i.bak -E "s|'xrpl/[a-zA-Z0-9\\.\\-]+'|'xrpl/${{ github.event.client_payload.conan_ref }}'|g" conanfile.py
           rm -f conanfile.py.bak
 
       - name: Update conan lockfile
-        shell: bash
         run: |
           conan lock create . --profile:all ${{ env.CONAN_PROFILE }}
 
@@ -57,7 +59,7 @@ jobs:
         run: strip build/clio_tests
 
       - name: Upload clio_tests
-        uses: actions/upload-artifact@v4
+        uses: actions/upload-artifact@330a01c490aca151604b8cf639adc76d48f6c5d4 # v5.0.0
         with:
           name: clio_tests_check_libxrpl
           path: build/clio_tests
@@ -67,10 +69,10 @@ jobs:
     needs: build
     runs-on: heavy
     container:
-      image: ghcr.io/xrplf/clio-ci:213752862ca95ecadeb59a6176c3db91a7864b3e
+      image: ghcr.io/xrplf/clio-ci:62369411404eb32b0140603a785ff05e1dc36ce8
 
     steps:
-      - uses: actions/download-artifact@v5
+      - uses: actions/download-artifact@018cc2cf5baa6db3ef3c5f8a56943fffe632ef53 # v6.0.0
         with:
           name: clio_tests_check_libxrpl
 
@@ -90,7 +92,7 @@ jobs:
       issues: write
 
     steps:
-      - uses: actions/checkout@v4
+      - uses: actions/checkout@08c6903cd8c0fde910a37f88322edcfb5dd907a8 # v5.0.0
 
       - name: Create an issue
         uses: ./.github/actions/create-issue

.github/workflows/check-pr-title.yml

@@ -5,20 +5,26 @@ on:
     types: [opened, edited, reopened, synchronize]
     branches: [develop]
 
+defaults:
+  run:
+    shell: bash
+
 jobs:
   check_title:
     runs-on: ubuntu-latest
 
     steps:
-      - uses: ytanikin/pr-conventional-commits@b72758283dcbee706975950e96bc4bf323a8d8c0 # v1.4.2
+      - uses: ytanikin/pr-conventional-commits@b72758283dcbee706975950e96bc4bf323a8d8c0 # 1.4.2
         with:
           task_types: '["build","feat","fix","docs","test","ci","style","refactor","perf","chore"]'
           add_label: false
           custom_labels: '{"build":"build", "feat":"enhancement", "fix":"bug", "docs":"documentation", "test":"testability", "ci":"ci", "style":"refactoring", "refactor":"refactoring", "perf":"performance", "chore":"tooling"}'
 
       - name: Check if message starts with upper-case letter
+        env:
+          PR_TITLE: ${{ github.event.pull_request.title }}
         run: |
-          if [[ ! "${{ github.event.pull_request.title }}" =~ ^[a-z]+:\ [\[A-Z] ]]; then
+          if [[ ! "${PR_TITLE}" =~ ^[a-z]+:\ [\[A-Z] ]]; then
             echo "Error: PR title must start with an upper-case letter."
             exit 1
           fi

.github/workflows/clang-tidy.yml

@@ -22,12 +22,16 @@ env:
   CONAN_PROFILE: clang
   LLVM_TOOLS_VERSION: 20
 
+defaults:
+  run:
+    shell: bash
+
 jobs:
   clang_tidy:
     if: github.event_name != 'push' || contains(github.event.head_commit.message, 'clang-tidy auto fixes')
     runs-on: heavy
     container:
-      image: ghcr.io/xrplf/clio-ci:213752862ca95ecadeb59a6176c3db91a7864b3e
+      image: ghcr.io/xrplf/clio-ci:62369411404eb32b0140603a785ff05e1dc36ce8
 
     permissions:
       contents: write
@@ -35,12 +39,12 @@ jobs:
       pull-requests: write
 
     steps:
-      - uses: actions/checkout@v4
+      - uses: actions/checkout@08c6903cd8c0fde910a37f88322edcfb5dd907a8 # v5.0.0
         with:
           fetch-depth: 0
 
       - name: Prepare runner
-        uses: XRPLF/actions/.github/actions/prepare-runner@7951b682e5a2973b28b0719a72f01fc4b0d0c34f
+        uses: XRPLF/actions/.github/actions/prepare-runner@8abb0722cbff83a9a2dc7d06c473f7a4964b7382
         with:
           disable_ccache: true
 
@@ -61,27 +65,24 @@ jobs:
         with:
           conan_profile: ${{ env.CONAN_PROFILE }}
 
-      - name: Get number of threads
-        uses: ./.github/actions/get-threads-number
-        id: number_of_threads
+      - name: Get number of processors
+        uses: XRPLF/actions/.github/actions/get-nproc@046b1620f6bfd6cd0985dc82c3df02786801fe0a
+        id: nproc
 
       - name: Run clang-tidy
         continue-on-error: true
-        shell: bash
         id: run_clang_tidy
         run: |
-          run-clang-tidy-${{ env.LLVM_TOOLS_VERSION }} -p build -j "${{ steps.number_of_threads.outputs.threads_number }}" -fix -quiet 1>output.txt
+          run-clang-tidy-${{ env.LLVM_TOOLS_VERSION }} -p build -j "${{ steps.nproc.outputs.nproc }}" -fix -quiet 1>output.txt
 
       - name: Fix local includes and clang-format style
         if: ${{ steps.run_clang_tidy.outcome != 'success' }}
-        shell: bash
         run: |
           pre-commit run --all-files fix-local-includes || true
           pre-commit run --all-files clang-format || true
 
       - name: Print issues found
         if: ${{ steps.run_clang_tidy.outcome != 'success' }}
-        shell: bash
         run: |
           sed -i '/error\||/!d' ./output.txt
           cat output.txt
@@ -126,5 +127,4 @@ jobs:
 
       - name: Fail the job
         if: ${{ steps.run_clang_tidy.outcome != 'success' }}
-        shell: bash
         run: exit 1

.github/workflows/docs.yml

@@ -10,20 +10,24 @@ concurrency:
   group: ${{ github.workflow }}-${{ github.ref }}
   cancel-in-progress: true
 
+defaults:
+  run:
+    shell: bash
+
 jobs:
   build:
     runs-on: ubuntu-latest
     container:
-      image: ghcr.io/xrplf/clio-ci:213752862ca95ecadeb59a6176c3db91a7864b3e
+      image: ghcr.io/xrplf/clio-ci:62369411404eb32b0140603a785ff05e1dc36ce8
 
     steps:
       - name: Checkout
-        uses: actions/checkout@v4
+        uses: actions/checkout@08c6903cd8c0fde910a37f88322edcfb5dd907a8 # v5.0.0
         with:
           lfs: true
 
       - name: Prepare runner
-        uses: XRPLF/actions/.github/actions/prepare-runner@7951b682e5a2973b28b0719a72f01fc4b0d0c34f
+        uses: XRPLF/actions/.github/actions/prepare-runner@8abb0722cbff83a9a2dc7d06c473f7a4964b7382
         with:
           disable_ccache: true
 
@@ -39,10 +43,10 @@ jobs:
         run: cmake --build . --target docs
 
       - name: Setup Pages
-        uses: actions/configure-pages@v5
+        uses: actions/configure-pages@983d7736d9b0ae728b81ab479565c72886d7745b # v5.0.0
 
       - name: Upload artifact
-        uses: actions/upload-pages-artifact@v4
+        uses: actions/upload-pages-artifact@7b1f4a764d45c48632c6b24a0339c27f5614fb0b # v4.0.0
         with:
           path: build_docs/html
           name: docs-develop
@@ -62,6 +66,6 @@ jobs:
     steps:
       - name: Deploy to GitHub Pages
         id: deployment
-        uses: actions/deploy-pages@v4
+        uses: actions/deploy-pages@d6db90164ac5ed86f2b6aed7e0febac5b3c0c03e # v4.0.5
         with:
           artifact_name: docs-develop

.github/workflows/nightly.yml

@@ -23,6 +23,10 @@ concurrency:
   group: ${{ github.workflow }}-${{ github.ref }}
   cancel-in-progress: true
 
+defaults:
+  run:
+    shell: bash
+
 jobs:
   build-and-test:
     name: Build and Test
@@ -39,17 +43,17 @@ jobs:
             conan_profile: gcc
             build_type: Release
             static: true
-            container: '{ "image": "ghcr.io/xrplf/clio-ci:213752862ca95ecadeb59a6176c3db91a7864b3e" }'
+            container: '{ "image": "ghcr.io/xrplf/clio-ci:62369411404eb32b0140603a785ff05e1dc36ce8" }'
           - os: heavy
             conan_profile: gcc
             build_type: Debug
             static: true
-            container: '{ "image": "ghcr.io/xrplf/clio-ci:213752862ca95ecadeb59a6176c3db91a7864b3e" }'
+            container: '{ "image": "ghcr.io/xrplf/clio-ci:62369411404eb32b0140603a785ff05e1dc36ce8" }'
           - os: heavy
             conan_profile: gcc.ubsan
             build_type: Release
             static: false
-            container: '{ "image": "ghcr.io/xrplf/clio-ci:213752862ca95ecadeb59a6176c3db91a7864b3e" }'
+            container: '{ "image": "ghcr.io/xrplf/clio-ci:62369411404eb32b0140603a785ff05e1dc36ce8" }'
 
     uses: ./.github/workflows/reusable-build-test.yml
     with:
@@ -73,7 +77,7 @@ jobs:
         include:
           - os: heavy
             conan_profile: clang
-            container: '{ "image": "ghcr.io/xrplf/clio-ci:213752862ca95ecadeb59a6176c3db91a7864b3e" }'
+            container: '{ "image": "ghcr.io/xrplf/clio-ci:62369411404eb32b0140603a785ff05e1dc36ce8" }'
             static: true
           - os: macos15
             conan_profile: apple-clang
@@ -93,18 +97,29 @@ jobs:
       targets: all
       analyze_build_time: true
 
+  get_date:
+    name: Get Date
+    runs-on: ubuntu-latest
+    outputs:
+      date: ${{ steps.get_date.outputs.date }}
+    steps:
+      - name: Get current date
+        id: get_date
+        run: |
+          echo "date=$(date +'%Y%m%d')" >> $GITHUB_OUTPUT
+
   nightly_release:
-    needs: build-and-test
+    needs: [build-and-test, get_date]
     uses: ./.github/workflows/reusable-release.yml
     with:
-      overwrite_release: true
+      delete_pattern: "nightly-*"
       prerelease: true
-      title: "Clio development (nightly) build"
-      version: nightly
+      title: "Clio development build (nightly-${{ needs.get_date.outputs.date }})"
+      version: nightly-${{ needs.get_date.outputs.date }}
       header: >
         > **Note:** Please remember that this is a development release and it is not recommended for production use.
 
-        Changelog (including previous releases): <https://github.com/XRPLF/clio/commits/nightly>
+        Changelog (including previous releases): <https://github.com/XRPLF/clio/commits/nightly-${{ needs.get_date.outputs.date }}>
       generate_changelog: false
       draft: false
 
@@ -130,7 +145,7 @@ jobs:
       issues: write
 
     steps:
-      - uses: actions/checkout@v4
+      - uses: actions/checkout@08c6903cd8c0fde910a37f88322edcfb5dd907a8 # v5.0.0
 
       - name: Create an issue
         uses: ./.github/actions/create-issue

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

@@ -1,8 +1,8 @@
 name: Pre-commit auto-update
 
 on:
-  # every first day of the month
   schedule:
+    # every first day of the month
     - cron: "0 0 1 * *"
   pull_request:
     branches: [release/*, develop]

.github/workflows/pre-commit.yml

@@ -8,7 +8,7 @@ on:
 
 jobs:
   run-hooks:
-    uses: XRPLF/actions/.github/workflows/pre-commit.yml@afbcbdafbe0ce5439492fb87eda6441371086386
+    uses: XRPLF/actions/.github/workflows/pre-commit.yml@34790936fae4c6c751f62ec8c06696f9c1a5753a
     with:
       runs_on: heavy
-      container: '{ "image": "ghcr.io/xrplf/clio-pre-commit:213752862ca95ecadeb59a6176c3db91a7864b3e" }'
+      container: '{ "image": "ghcr.io/xrplf/clio-pre-commit:c117f470f2ef954520ab5d1c8a5ed2b9e68d6f8a" }'

.github/workflows/release.yml

@@ -29,7 +29,7 @@ jobs:
             conan_profile: gcc
             build_type: Release
             static: true
-            container: '{ "image": "ghcr.io/xrplf/clio-ci:213752862ca95ecadeb59a6176c3db91a7864b3e" }'
+            container: '{ "image": "ghcr.io/xrplf/clio-ci:62369411404eb32b0140603a785ff05e1dc36ce8" }'
 
     uses: ./.github/workflows/reusable-build-test.yml
     with:
@@ -49,7 +49,7 @@ jobs:
     needs: build-and-test
     uses: ./.github/workflows/reusable-release.yml
     with:
-      overwrite_release: false
+      delete_pattern: ""
       prerelease: ${{ contains(github.ref_name, '-') }}
       title: "${{ github.ref_name }}"
       version: "${{ github.ref_name }}"

.github/workflows/reusable-build.yml

@@ -75,6 +75,10 @@ on:
       CODECOV_TOKEN:
         required: false
 
+defaults:
+  run:
+    shell: bash
+
 jobs:
   build:
     name: Build
@@ -86,7 +90,7 @@ jobs:
         if: ${{ runner.os == 'macOS' }}
         uses: XRPLF/actions/.github/actions/cleanup-workspace@ea9970b7c211b18f4c8bcdb28c29f5711752029f
 
-      - uses: actions/checkout@v4
+      - uses: actions/checkout@08c6903cd8c0fde910a37f88322edcfb5dd907a8 # v5.0.0
         with:
           fetch-depth: 0
           # We need to fetch tags to have correct version in the release
@@ -95,13 +99,12 @@ jobs:
           ref: ${{ github.ref }}
 
       - name: Prepare runner
-        uses: XRPLF/actions/.github/actions/prepare-runner@7951b682e5a2973b28b0719a72f01fc4b0d0c34f
+        uses: XRPLF/actions/.github/actions/prepare-runner@8abb0722cbff83a9a2dc7d06c473f7a4964b7382
         with:
           disable_ccache: ${{ !inputs.download_ccache }}
 
       - name: Setup conan on macOS
         if: ${{ runner.os == 'macOS' }}
-        shell: bash
         run: ./.github/scripts/conan/init.sh
 
       - name: Restore cache
@@ -141,18 +144,16 @@ jobs:
           ClangBuildAnalyzer --all build/ build_time_report.bin
           ClangBuildAnalyzer --analyze build_time_report.bin > build_time_report.txt
           cat build_time_report.txt
-        shell: bash
 
       - name: Upload build time analyze report
         if: ${{ inputs.analyze_build_time }}
-        uses: actions/upload-artifact@v4
+        uses: actions/upload-artifact@330a01c490aca151604b8cf639adc76d48f6c5d4 # v5.0.0
         with:
           name: build_time_report_${{ runner.os }}_${{ inputs.build_type }}_${{ inputs.conan_profile }}
           path: build_time_report.txt
 
       - name: Show ccache's statistics
         if: ${{ inputs.download_ccache }}
-        shell: bash
         id: ccache_stats
         run: |
           ccache -s > /tmp/ccache.stats
@@ -170,28 +171,28 @@ jobs:
 
       - name: Upload clio_server
         if: ${{ inputs.upload_clio_server && !inputs.code_coverage && !inputs.analyze_build_time }}
-        uses: actions/upload-artifact@v4
+        uses: actions/upload-artifact@330a01c490aca151604b8cf639adc76d48f6c5d4 # v5.0.0
         with:
           name: clio_server_${{ runner.os }}_${{ inputs.build_type }}_${{ inputs.conan_profile }}
           path: build/clio_server
 
       - name: Upload clio_tests
         if: ${{ !inputs.code_coverage && !inputs.analyze_build_time && !inputs.package }}
-        uses: actions/upload-artifact@v4
+        uses: actions/upload-artifact@330a01c490aca151604b8cf639adc76d48f6c5d4 # v5.0.0
         with:
           name: clio_tests_${{ runner.os }}_${{ inputs.build_type }}_${{ inputs.conan_profile }}
           path: build/clio_tests
 
       - name: Upload clio_integration_tests
         if: ${{ !inputs.code_coverage && !inputs.analyze_build_time && !inputs.package }}
-        uses: actions/upload-artifact@v4
+        uses: actions/upload-artifact@330a01c490aca151604b8cf639adc76d48f6c5d4 # v5.0.0
         with:
           name: clio_integration_tests_${{ runner.os }}_${{ inputs.build_type }}_${{ inputs.conan_profile }}
           path: build/clio_integration_tests
 
       - name: Upload Clio Linux package
         if: ${{ inputs.package }}
-        uses: actions/upload-artifact@v4
+        uses: actions/upload-artifact@330a01c490aca151604b8cf639adc76d48f6c5d4 # v5.0.0
         with:
           name: clio_deb_package_${{ runner.os }}_${{ inputs.build_type }}_${{ inputs.conan_profile }}
           path: build/*.deb
@@ -220,13 +221,14 @@ jobs:
 
       - name: Verify expected version
         if: ${{ inputs.expected_version != '' }}
-        shell: bash
+        env:
+          INPUT_EXPECTED_VERSION: ${{ inputs.expected_version }}
         run: |
           set -e
-          EXPECTED_VERSION="clio-${{ inputs.expected_version }}"
+          EXPECTED_VERSION="clio-${INPUT_EXPECTED_VERSION}"
           actual_version=$(./build/clio_server --version)
           if [[ "$actual_version" != "$EXPECTED_VERSION" ]]; then
-            echo "Expected version '$EXPECTED_VERSION', but got '$actual_version'"
+            echo "Expected version '${EXPECTED_VERSION}', but got '${actual_version}'"
             exit 1
           fi
 

.github/workflows/reusable-release.yml

@@ -3,10 +3,10 @@ name: Make release
 on:
   workflow_call:
     inputs:
-      overwrite_release:
-        description: "Overwrite the current release and tag"
+      delete_pattern:
+        description: "Pattern to delete previous releases"
         required: true
-        type: boolean
+        type: string
 
       prerelease:
         description: "Create a prerelease"
@@ -38,11 +38,15 @@ on:
         required: true
         type: boolean
 
+defaults:
+  run:
+    shell: bash
+
 jobs:
   release:
     runs-on: heavy
     container:
-      image: ghcr.io/xrplf/clio-ci:213752862ca95ecadeb59a6176c3db91a7864b3e
+      image: ghcr.io/xrplf/clio-ci:62369411404eb32b0140603a785ff05e1dc36ce8
     env:
       GH_REPO: ${{ github.repository }}
       GH_TOKEN: ${{ github.token }}
@@ -51,29 +55,29 @@ jobs:
       contents: write
 
     steps:
-      - uses: actions/checkout@v4
+      - uses: actions/checkout@08c6903cd8c0fde910a37f88322edcfb5dd907a8 # v5.0.0
         with:
           fetch-depth: 0
 
       - name: Prepare runner
-        uses: XRPLF/actions/.github/actions/prepare-runner@7951b682e5a2973b28b0719a72f01fc4b0d0c34f
+        uses: XRPLF/actions/.github/actions/prepare-runner@8abb0722cbff83a9a2dc7d06c473f7a4964b7382
         with:
           disable_ccache: true
 
-      - uses: actions/download-artifact@v5
+      - uses: actions/download-artifact@018cc2cf5baa6db3ef3c5f8a56943fffe632ef53 # v6.0.0
         with:
           path: release_artifacts
           pattern: clio_server_*
 
       - name: Create release notes
-        shell: bash
+        env:
+          RELEASE_HEADER: ${{ inputs.header }}
         run: |
           echo "# Release notes" > "${RUNNER_TEMP}/release_notes.md"
           echo "" >> "${RUNNER_TEMP}/release_notes.md"
-          printf '%s\n' "${{ inputs.header }}" >> "${RUNNER_TEMP}/release_notes.md"
+          printf '%s\n' "${RELEASE_HEADER}" >> "${RUNNER_TEMP}/release_notes.md"
 
       - name: Generate changelog
-        shell: bash
         if: ${{ inputs.generate_changelog }}
         run: |
           LAST_TAG="$(gh release view --json tagName -q .tagName --repo XRPLF/clio)"
@@ -83,30 +87,39 @@ jobs:
           cat CHANGELOG.md >> "${RUNNER_TEMP}/release_notes.md"
 
       - name: Prepare release artifacts
-        shell: bash
         run: .github/scripts/prepare-release-artifacts.sh release_artifacts
 
       - name: Upload release notes
-        uses: actions/upload-artifact@v4
+        uses: actions/upload-artifact@330a01c490aca151604b8cf639adc76d48f6c5d4 # v5.0.0
         with:
           name: release_notes_${{ inputs.version }}
           path: "${RUNNER_TEMP}/release_notes.md"
 
-      - name: Remove current release and tag
-        if: ${{ github.event_name != 'pull_request' && inputs.overwrite_release }}
-        shell: bash
+      - name: Remove previous release with a pattern
+        if: ${{ github.event_name != 'pull_request' && inputs.delete_pattern != '' }}
+        env:
+          DELETE_PATTERN: ${{ inputs.delete_pattern }}
         run: |
-          gh release delete ${{ inputs.version }} --yes || true
-          git push origin :${{ inputs.version }} || true
+          RELEASES_TO_DELETE=$(gh release list --limit 50 --repo "${GH_REPO}" | grep -E "${DELETE_PATTERN}" | awk -F'\t' '{print $3}' || true)
+          if [ -n "$RELEASES_TO_DELETE" ]; then
+            for RELEASE in $RELEASES_TO_DELETE; do
+              echo "Deleting release: $RELEASE"
+              gh release delete "$RELEASE" --repo "${GH_REPO}" --yes --cleanup-tag
+            done
+          fi
 
       - name: Publish release
         if: ${{ github.event_name != 'pull_request' }}
-        shell: bash
+        env:
+          RELEASE_VERSION: ${{ inputs.version }}
+          PRERELEASE_OPTION: ${{ inputs.prerelease && '--prerelease' || '' }}
+          RELEASE_TITLE: ${{ inputs.title }}
+          DRAFT_OPTION: ${{ inputs.draft && '--draft' || '' }}
         run: |
-          gh release create "${{ inputs.version }}" \
-            ${{ inputs.prerelease && '--prerelease' || '' }} \
-            --title "${{ inputs.title }}" \
+          gh release create "${RELEASE_VERSION}" \
+            ${PRERELEASE_OPTION} \
+            --title "${RELEASE_TITLE}" \
             --target "${GITHUB_SHA}" \
-            ${{ inputs.draft && '--draft' || '' }} \
+            ${DRAFT_OPTION} \
             --notes-file "${RUNNER_TEMP}/release_notes.md" \
             ./release_artifacts/clio_server*

.github/workflows/reusable-test.yml

@@ -33,6 +33,10 @@ on:
         required: true
         type: boolean
 
+defaults:
+  run:
+    shell: bash
+
 jobs:
   unit_tests:
     name: Unit testing
@@ -43,23 +47,22 @@ jobs:
 
     env:
       # TODO: remove completely when we have fixed all currently existing issues with sanitizers
-      SANITIZER_IGNORE_ERRORS: ${{ endsWith(inputs.conan_profile, '.tsan') || inputs.conan_profile == 'clang.asan' || (inputs.conan_profile == 'gcc.asan' && inputs.build_type == 'Release') }}
+      SANITIZER_IGNORE_ERRORS: ${{ endsWith(inputs.conan_profile, '.tsan') }}
 
     steps:
       - name: Cleanup workspace
         if: ${{ runner.os == 'macOS' }}
         uses: XRPLF/actions/.github/actions/cleanup-workspace@ea9970b7c211b18f4c8bcdb28c29f5711752029f
 
-      - uses: actions/checkout@v4
+      - uses: actions/checkout@08c6903cd8c0fde910a37f88322edcfb5dd907a8 # v5.0.0
         with:
           fetch-depth: 0
 
-      - uses: actions/download-artifact@v5
+      - uses: actions/download-artifact@018cc2cf5baa6db3ef3c5f8a56943fffe632ef53 # v6.0.0
         with:
           name: clio_tests_${{ runner.os }}_${{ inputs.build_type }}_${{ inputs.conan_profile }}
 
       - name: Make clio_tests executable
-        shell: bash
         run: chmod +x ./clio_tests
 
       - name: Run clio_tests (regular)
@@ -68,11 +71,10 @@ jobs:
 
       - name: Run clio_tests (sanitizer errors ignored)
         if: ${{ env.SANITIZER_IGNORE_ERRORS == 'true' }}
-        run: ./.github/scripts/execute-tests-under-sanitizer ./clio_tests
+        run: ./.github/scripts/execute-tests-under-sanitizer.sh ./clio_tests
 
       - name: Check for sanitizer report
         if: ${{ env.SANITIZER_IGNORE_ERRORS == 'true' }}
-        shell: bash
         id: check_report
         run: |
           if ls .sanitizer-report/* 1> /dev/null 2>&1; then
@@ -83,7 +85,7 @@ jobs:
 
       - name: Upload sanitizer report
         if: ${{ env.SANITIZER_IGNORE_ERRORS == 'true' && steps.check_report.outputs.found_report == 'true' }}
-        uses: actions/upload-artifact@v4
+        uses: actions/upload-artifact@330a01c490aca151604b8cf639adc76d48f6c5d4 # v5.0.0
         with:
           name: sanitizer_report_${{ runner.os }}_${{ inputs.build_type }}_${{ inputs.conan_profile }}
           path: .sanitizer-report/*
@@ -144,7 +146,7 @@ jobs:
             sleep 5
           done
 
-      - uses: actions/download-artifact@v5
+      - uses: actions/download-artifact@018cc2cf5baa6db3ef3c5f8a56943fffe632ef53 # v6.0.0
         with:
           name: clio_integration_tests_${{ runner.os }}_${{ inputs.build_type }}_${{ inputs.conan_profile }}
 

.github/workflows/reusable-upload-coverage-report.yml

@@ -6,18 +6,22 @@ on:
       CODECOV_TOKEN:
         required: true
 
+defaults:
+  run:
+    shell: bash
+
 jobs:
   upload_report:
     name: Upload report
     runs-on: ubuntu-latest
 
     steps:
-      - uses: actions/checkout@v4
+      - uses: actions/checkout@08c6903cd8c0fde910a37f88322edcfb5dd907a8 # v5.0.0
         with:
           fetch-depth: 0
 
       - name: Download report artifact
-        uses: actions/download-artifact@v5
+        uses: actions/download-artifact@018cc2cf5baa6db3ef3c5f8a56943fffe632ef53 # v6.0.0
         with:
           name: coverage-report.xml
           path: build

.github/workflows/sanitizers.yml

@@ -15,7 +15,7 @@ on:
       - ".github/actions/**"
       - "!.github/actions/build-docker-image/**"
       - "!.github/actions/create-issue/**"
-      - .github/scripts/execute-tests-under-sanitizer
+      - .github/scripts/execute-tests-under-sanitizer.sh
 
       - CMakeLists.txt
       - conanfile.py
@@ -44,7 +44,7 @@ jobs:
     uses: ./.github/workflows/reusable-build-test.yml
     with:
       runs_on: heavy
-      container: '{ "image": "ghcr.io/xrplf/clio-ci:213752862ca95ecadeb59a6176c3db91a7864b3e" }'
+      container: '{ "image": "ghcr.io/xrplf/clio-ci:62369411404eb32b0140603a785ff05e1dc36ce8" }'
       download_ccache: false
       upload_ccache: false
       conan_profile: ${{ matrix.compiler }}${{ matrix.sanitizer_ext }}

.github/workflows/update-docker-ci.yml

@@ -33,6 +33,10 @@ env:
   GCC_MAJOR_VERSION: 15
   GCC_VERSION: 15.2.0
 
+defaults:
+  run:
+    shell: bash
+
 jobs:
   repo:
     name: Calculate repo name
@@ -52,7 +56,7 @@ jobs:
     needs: repo
 
     steps:
-      - uses: actions/checkout@v4
+      - uses: actions/checkout@08c6903cd8c0fde910a37f88322edcfb5dd907a8 # v5.0.0
 
       - name: Get changed files
         id: changed-files
@@ -90,7 +94,7 @@ jobs:
     needs: repo
 
     steps:
-      - uses: actions/checkout@v4
+      - uses: actions/checkout@08c6903cd8c0fde910a37f88322edcfb5dd907a8 # v5.0.0
 
       - name: Get changed files
         id: changed-files
@@ -128,7 +132,7 @@ jobs:
     needs: [repo, gcc-amd64, gcc-arm64]
 
     steps:
-      - uses: actions/checkout@v4
+      - uses: actions/checkout@08c6903cd8c0fde910a37f88322edcfb5dd907a8 # v5.0.0
 
       - name: Get changed files
         id: changed-files
@@ -137,7 +141,7 @@ jobs:
           files: "docker/compilers/gcc/**"
 
       - name: Set up Docker Buildx
-        uses: docker/setup-buildx-action@v3
+        uses: docker/setup-buildx-action@e468171a9de216ec08956ac3ada2f0791b6bd435 # v3.11.1
 
       - name: Login to GitHub Container Registry
         if: ${{ github.event_name != 'pull_request' }}
@@ -179,7 +183,7 @@ jobs:
     needs: repo
 
     steps:
-      - uses: actions/checkout@v4
+      - uses: actions/checkout@08c6903cd8c0fde910a37f88322edcfb5dd907a8 # v5.0.0
 
       - name: Get changed files
         id: changed-files
@@ -215,7 +219,7 @@ jobs:
     needs: [repo, gcc-merge]
 
     steps:
-      - uses: actions/checkout@v4
+      - uses: actions/checkout@08c6903cd8c0fde910a37f88322edcfb5dd907a8 # v5.0.0
 
       - name: Get changed files
         id: changed-files
@@ -246,7 +250,7 @@ jobs:
     needs: [repo, gcc-merge]
 
     steps:
-      - uses: actions/checkout@v4
+      - uses: actions/checkout@08c6903cd8c0fde910a37f88322edcfb5dd907a8 # v5.0.0
 
       - name: Get changed files
         id: changed-files
@@ -277,7 +281,7 @@ jobs:
     needs: [repo, tools-amd64, tools-arm64]
 
     steps:
-      - uses: actions/checkout@v4
+      - uses: actions/checkout@08c6903cd8c0fde910a37f88322edcfb5dd907a8 # v5.0.0
 
       - name: Get changed files
         id: changed-files
@@ -286,7 +290,7 @@ jobs:
           files: "docker/tools/**"
 
       - name: Set up Docker Buildx
-        uses: docker/setup-buildx-action@v3
+        uses: docker/setup-buildx-action@e468171a9de216ec08956ac3ada2f0791b6bd435 # v3.11.1
 
       - name: Login to GitHub Container Registry
         if: ${{ github.event_name != 'pull_request' }}
@@ -312,7 +316,7 @@ jobs:
     needs: [repo, tools-merge]
 
     steps:
-      - uses: actions/checkout@v4
+      - uses: actions/checkout@08c6903cd8c0fde910a37f88322edcfb5dd907a8 # v5.0.0
       - uses: ./.github/actions/build-docker-image
         env:
           GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
@@ -334,7 +338,7 @@ jobs:
     needs: [repo, gcc-merge, clang, tools-merge]
 
     steps:
-      - uses: actions/checkout@v4
+      - uses: actions/checkout@08c6903cd8c0fde910a37f88322edcfb5dd907a8 # v5.0.0
       - uses: ./.github/actions/build-docker-image
         env:
           GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}

.github/workflows/upload-conan-deps.yml

@@ -40,13 +40,17 @@ concurrency:
   group: ${{ github.workflow }}-${{ github.ref }}
   cancel-in-progress: true
 
+defaults:
+  run:
+    shell: bash
+
 jobs:
   generate-matrix:
     runs-on: ubuntu-latest
     outputs:
       matrix: ${{ steps.set-matrix.outputs.matrix }}
     steps:
-      - uses: actions/checkout@v4
+      - uses: actions/checkout@08c6903cd8c0fde910a37f88322edcfb5dd907a8 # v5.0.0
 
       - name: Calculate conan matrix
         id: set-matrix
@@ -69,16 +73,15 @@ jobs:
       CONAN_PROFILE: ${{ matrix.compiler }}${{ matrix.sanitizer_ext }}
 
     steps:
-      - uses: actions/checkout@v4
+      - uses: actions/checkout@08c6903cd8c0fde910a37f88322edcfb5dd907a8 # v5.0.0
 
       - name: Prepare runner
-        uses: XRPLF/actions/.github/actions/prepare-runner@7951b682e5a2973b28b0719a72f01fc4b0d0c34f
+        uses: XRPLF/actions/.github/actions/prepare-runner@8abb0722cbff83a9a2dc7d06c473f7a4964b7382
         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
@@ -99,4 +102,6 @@ jobs:
 
       - name: Upload Conan packages
         if: ${{ github.repository_owner == 'XRPLF' && github.event_name != 'pull_request' && github.event_name != 'schedule' }}
-        run: conan upload "*" -r=xrplf --confirm ${{ github.event.inputs.force_upload == 'true' && '--force' || '' }}
+        env:
+          FORCE_OPTION: ${{ github.event.inputs.force_upload == 'true' && '--force' || '' }}
+        run: conan upload "*" -r=xrplf --confirm ${FORCE_OPTION}

.pre-commit-config.yaml

@@ -11,7 +11,10 @@
 #
 # See https://pre-commit.com for more information
 # See https://pre-commit.com/hooks.html for more hooks
-exclude: ^(docs/doxygen-awesome-theme/|conan\.lock$)
+exclude: |
+  (?x)^(
+    docs/doxygen-awesome-theme/.*
+  )$
 
 repos:
   # `pre-commit sample-config` default hooks

CMakeLists.txt

@@ -75,10 +75,11 @@ if (san)
   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>
-  )
+  if (san STREQUAL "address")
+    # ASAN needs these definitions as well as correct b2 flags in conan profile for sanitizers
+    target_compile_definitions(clio_options INTERFACE BOOST_USE_ASAN=1 BOOST_USE_UCONTEXT=1)
+  endif ()
+
   target_link_libraries(clio_options INTERFACE ${SAN_FLAG} ${SAN_LIB})
 endif ()
 

README.md

@@ -34,7 +34,6 @@ Below are some useful docs to learn more about Clio.
 
 - [How to configure Clio and rippled](./docs/configure-clio.md)
 - [How to run Clio](./docs/run-clio.md)
-- [Logging](./docs/logging.md)
 - [Troubleshooting guide](./docs/trouble_shooting.md)
 
 **General reference material:**

conan.lock

@@ -3,6 +3,7 @@
     "requires": [
         "zlib/1.3.1#b8bc2603263cf7eccbd6e17e66b0ed76%1756234269.497",
         "xxhash/0.8.3#681d36a0a6111fc56e5e45ea182c19cc%1756234289.683",
+        "xrpl/3.0.0-rc1#f5c8ecd42bdf511ad36f57bc702dacd2%1762975621.294",
         "xrpl/2.6.1#973af2bf9631f239941dd9f5a100bb84%1759275059.342",
         "sqlite3/3.49.1#8631739a4c9b93bd3d6b753bac548a63%1756234266.869",
         "spdlog/1.15.3#3ca0e9e6b83af4d0151e26541d140c86%1754401846.61",

conanfile.py

@@ -18,7 +18,7 @@ class ClioConan(ConanFile):
         'protobuf/3.21.12',
         'grpc/1.50.1',
         'openssl/1.1.1w',
-        'xrpl/2.6.1',
+        'xrpl/3.0.0-rc1',
         'zlib/1.3.1',
         'libbacktrace/cci.20210118',
         'spdlog/1.15.3',

docker/ci/Dockerfile

@@ -55,8 +55,11 @@ RUN pip install -q --no-cache-dir \
         # lxml 6.0.0 is not compatible with our image
         'lxml<6.0.0' \
         cmake \
-        conan==2.20.1 \
-        gcovr
+        conan==2.22.1 \
+        gcovr \
+        # We're adding pre-commit to this image as well,
+        # because clang-tidy workflow requires it
+        pre-commit
 
 # Install LLVM tools
 ARG LLVM_TOOLS_VERSION=20

docker/ci/README.md

@@ -5,17 +5,17 @@ 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
+- ccache 4.12.1
 - Clang 19
 - ClangBuildAnalyzer 1.6.0
-- Conan 2.20.1
-- Doxygen 1.14
+- Conan 2.22.1
+- Doxygen 1.15.0
 - GCC 15.2.0
 - GDB 16.3
-- gh 2.74
-- git-cliff 2.9.1
-- mold 2.40.1
-- Python 3.13
+- gh 2.82.1
+- git-cliff 2.10.1
+- mold 2.40.4
+- Python 3.8
 - and some other useful tools
 
 Conan is set up to build Clio without any additional steps.

docker/ci/conan/sanitizer_template.profile

@@ -3,6 +3,9 @@
 {% set sanitizer_opt_map = {"asan": "address", "tsan": "thread", "ubsan": "undefined"} %}
 {% set sanitizer = sanitizer_opt_map[sani] %}
 
+{% set sanitizer_b2_flags_map = {"address": "define=BOOST_USE_ASAN=1 context-impl=ucontext address-sanitizer=on", "thread": "thread-sanitizer=on", "undefined": "undefined-sanitizer=on"} %}
+{% set sanitizer_b2_flags_str = sanitizer_b2_flags_map[sanitizer] %}
+
 {% 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 %}
@@ -11,7 +14,8 @@
 include({{ compiler }})
 
 [options]
-boost/*:extra_b2_flags="cxxflags=\"{{ sanitizer_build_flags_str }}\" linkflags=\"{{ sanitizer_link_flags_str }}\""
+boost/*:extra_b2_flags="{{ sanitizer_b2_flags_str }}"
+boost/*:without_context=False
 boost/*:without_stacktrace=True
 
 [conf]

docker/compilers/gcc/Dockerfile

@@ -8,7 +8,7 @@ ARG UBUNTU_VERSION
 
 ARG GCC_MAJOR_VERSION
 
-ARG BUILD_VERSION=1
+ARG BUILD_VERSION=0
 
 ARG DEBIAN_FRONTEND=noninteractive
 ARG TARGETARCH
@@ -34,6 +34,7 @@ RUN wget --progress=dot:giga https://gcc.gnu.org/pub/gcc/releases/gcc-$GCC_VERSI
 WORKDIR /gcc-$GCC_VERSION
 RUN ./contrib/download_prerequisites
 
+# hadolint ignore=DL3059
 RUN mkdir /gcc-build
 WORKDIR /gcc-build
 RUN /gcc-$GCC_VERSION/configure \

docker/develop/compose.yaml

@@ -1,6 +1,6 @@
 services:
   clio_develop:
-    image: ghcr.io/xrplf/clio-ci:213752862ca95ecadeb59a6176c3db91a7864b3e
+    image: ghcr.io/xrplf/clio-ci:62369411404eb32b0140603a785ff05e1dc36ce8
     volumes:
       - clio_develop_conan_data:/root/.conan2/p
       - clio_develop_ccache:/root/.ccache

docker/pre-commit/Dockerfile

@@ -17,6 +17,7 @@ RUN apt-get update \
     && apt-get install -y --no-install-recommends --no-install-suggests \
         curl \
         git \
+        libatomic1 \
         software-properties-common \
     && apt-get clean \
     && rm -rf /var/lib/apt/lists/*

docker/tools/Dockerfile

@@ -8,7 +8,7 @@ ARG TARGETARCH
 
 SHELL ["/bin/bash", "-o", "pipefail", "-c"]
 
-ARG BUILD_VERSION=2
+ARG BUILD_VERSION=0
 
 RUN apt-get update \
     && apt-get install -y --no-install-recommends --no-install-suggests \
@@ -24,7 +24,7 @@ RUN apt-get update \
 
 WORKDIR /tmp
 
-ARG MOLD_VERSION=2.40.1
+ARG MOLD_VERSION=2.40.4
 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}" \
@@ -34,7 +34,7 @@ RUN wget --progress=dot:giga "https://github.com/rui314/mold/archive/refs/tags/v
     && ninja install \
     && rm -rf /tmp/* /var/tmp/*
 
-ARG CCACHE_VERSION=4.11.3
+ARG CCACHE_VERSION=4.12.1
 RUN wget --progress=dot:giga "https://github.com/ccache/ccache/releases/download/v${CCACHE_VERSION}/ccache-${CCACHE_VERSION}.tar.gz" \
     && tar xf "ccache-${CCACHE_VERSION}.tar.gz" \
     && cd "ccache-${CCACHE_VERSION}" \
@@ -51,7 +51,7 @@ RUN apt-get update \
     && apt-get clean \
     && rm -rf /var/lib/apt/lists/*
 
-ARG DOXYGEN_VERSION=1.14.0
+ARG DOXYGEN_VERSION=1.15.0
 RUN wget --progress=dot:giga "https://github.com/doxygen/doxygen/releases/download/Release_${DOXYGEN_VERSION//./_}/doxygen-${DOXYGEN_VERSION}.src.tar.gz" \
     && tar xf "doxygen-${DOXYGEN_VERSION}.src.tar.gz" \
     && cd "doxygen-${DOXYGEN_VERSION}" \
@@ -71,13 +71,13 @@ RUN wget --progress=dot:giga "https://github.com/aras-p/ClangBuildAnalyzer/archi
     && ninja install \
     && rm -rf /tmp/* /var/tmp/*
 
-ARG GIT_CLIFF_VERSION=2.9.1
+ARG GIT_CLIFF_VERSION=2.10.1
 RUN wget --progress=dot:giga "https://github.com/orhun/git-cliff/releases/download/v${GIT_CLIFF_VERSION}/git-cliff-${GIT_CLIFF_VERSION}-x86_64-unknown-linux-musl.tar.gz" \
     && tar xf git-cliff-${GIT_CLIFF_VERSION}-x86_64-unknown-linux-musl.tar.gz \
     && mv git-cliff-${GIT_CLIFF_VERSION}/git-cliff /usr/local/bin/git-cliff \
     && rm -rf /tmp/* /var/tmp/*
 
-ARG GH_VERSION=2.74.0
+ARG GH_VERSION=2.82.1
 RUN wget --progress=dot:giga "https://github.com/cli/cli/releases/download/v${GH_VERSION}/gh_${GH_VERSION}_linux_${TARGETARCH}.tar.gz" \
     && tar xf gh_${GH_VERSION}_linux_${TARGETARCH}.tar.gz \
     && mv gh_${GH_VERSION}_linux_${TARGETARCH}/bin/gh /usr/local/bin/gh \

docs/build-clio.md

@@ -191,7 +191,7 @@ Open the `index.html` file in your browser to see the documentation pages.
 It is also possible to build Clio using [Docker](https://www.docker.com/) if you don't want to install all the dependencies on your machine.
 
 ```sh
-docker run -it ghcr.io/xrplf/clio-ci:213752862ca95ecadeb59a6176c3db91a7864b3e
+docker run -it ghcr.io/xrplf/clio-ci:62369411404eb32b0140603a785ff05e1dc36ce8
 git clone https://github.com/XRPLF/clio
 cd clio
 ```

docs/config-description.md

@@ -293,7 +293,7 @@ This document provides a list of all available Clio configuration properties in
 
 - **Required**: True
 - **Type**: int
-- **Default value**: `1`
+- **Default value**: `1000`
 - **Constraints**: The minimum value is `1`. The maximum value is `4294967295`.
 - **Description**: The maximum size of the server's request queue. If set to `0`, this means there is no queue size limit.
 
@@ -441,6 +441,22 @@ This document provides a list of all available Clio configuration properties in
 - **Constraints**: The value must be one of the following: `sync`, `async`, `none`.
 - **Description**: The strategy used for Cache loading.
 
+### cache.file.path
+
+- **Required**: False
+- **Type**: string
+- **Default value**: None
+- **Constraints**: None
+- **Description**: The path to a file where cache will be saved to on shutdown and loaded from on startup. If the file couldn't be read Clio will load cache as usual (from DB or from rippled).
+
+### cache.file.max_sequence_age
+
+- **Required**: True
+- **Type**: int
+- **Default value**: `5000`
+- **Constraints**: None
+- **Description**: Max allowed difference between the latest sequence in DB and in cache file. If the cache file is too old (contains too low latest sequence) Clio will reject using it.
+
 ### log.channels.[].channel
 
 - **Required**: False

docs/doxygen-awesome-theme/doxygen-awesome.css

@@ -951,7 +951,7 @@ span.arrowhead {
     border-color: var(--primary-color);
 }
 
-#nav-tree ul li:first-child > div > a {
+#nav-tree-contents > ul > li:first-child > div > a {
     opacity: 0;
     pointer-events: none;
 }

docs/examples/config/example-config.json

@@ -61,7 +61,7 @@
     "ip": "0.0.0.0",
     "port": 51233,
     // Max number of requests to queue up before rejecting further requests.
-    // Defaults to 0, which disables the limit.
+    // Defaults to 1000 (use 0 to make the queue unbound).
     "max_queue_size": 500,
     // If request contains header with authorization, Clio will check if it matches the prefix 'Password ' + this value's sha256 hash
     // If matches, the request will be considered as admin request
@@ -137,7 +137,11 @@
     // "num_cursors_from_account": 3200, // Read the cursors from the account table until we have enough cursors to partition the ledger to load concurrently.
     "num_markers": 48, // The number of markers is the number of coroutines to load the cache concurrently.
     "page_fetch_size": 512, // The number of rows to load for each page.
-    "load": "async" // "sync" to load cache synchronously  or "async" to load cache asynchronously or "none"/"no" to turn off the cache.
+    "load": "async", // "sync" to load cache synchronously  or "async" to load cache asynchronously or "none"/"no" to turn off the cache.
+    "file": {
+      "path": "./cache.bin",
+      "max_sequence_age": 5000
+    }
   },
   "prometheus": {
     "enabled": true,

pre-commit-hooks/json_in_cpp.py

@@ -44,8 +44,13 @@ def fix_colon_spacing(cpp_content: str) -> str:
 
 
 def fix_indentation(cpp_content: str) -> str:
+    if "JSON(" not in cpp_content:
+        return cpp_content
+
     lines = cpp_content.splitlines()
 
+    ends_with_newline = cpp_content.endswith('\n')
+
     def find_indentation(line: str) -> int:
         return len(line) - len(line.lstrip())
 
@@ -66,7 +71,11 @@ def fix_indentation(cpp_content: str) -> str:
                         break
                     lines[i] = lines[i][by_how_much:] if by_how_much > 0 else " " * (-by_how_much) + lines[i]
 
-    return "\n".join(lines) + "\n"
+    result = "\n".join(lines)
+
+    if ends_with_newline:
+        result += "\n"
+    return result
 
 
 def process_file(file_path: Path, dry_run: bool) -> bool:

src/CMakeLists.txt

@@ -2,7 +2,6 @@ add_subdirectory(util)
 add_subdirectory(data)
 add_subdirectory(cluster)
 add_subdirectory(etl)
-add_subdirectory(etlng)
 add_subdirectory(feed)
 add_subdirectory(rpc)
 add_subdirectory(web)

src/app/CMakeLists.txt

@@ -5,10 +5,9 @@ target_link_libraries(
   clio_app
   PUBLIC clio_cluster
          clio_etl
-         clio_etlng
          clio_feed
-         clio_web
-         clio_rpc
          clio_migration
+         clio_rpc
+         clio_web
   PRIVATE Boost::program_options
 )

src/app/ClioApplication.cpp

@@ -28,8 +28,6 @@
 #include "etl/ETLService.hpp"
 #include "etl/LoadBalancer.hpp"
 #include "etl/NetworkValidatedLedgers.hpp"
-#include "etlng/LoadBalancer.hpp"
-#include "etlng/LoadBalancerInterface.hpp"
 #include "feed/SubscriptionManager.hpp"
 #include "migration/MigrationInspectorFactory.hpp"
 #include "rpc/Counters.hpp"
@@ -57,6 +55,7 @@
 #include <cstdlib>
 #include <memory>
 #include <optional>
+#include <string>
 #include <thread>
 #include <utility>
 #include <vector>
@@ -103,7 +102,7 @@ ClioApplication::run(bool const useNgWebServer)
     // This is not the only io context in the application.
     boost::asio::io_context ioc{threads};
 
-    // Similarly we need a context to run ETLng on
+    // Similarly we need a context to run ETL on
     // In the future we can remove the raw ioc and use ctx instead
     util::async::CoroExecutionContext ctx{threads};
 
@@ -112,7 +111,23 @@ ClioApplication::run(bool const useNgWebServer)
     auto const dosguardWeights = web::dosguard::Weights::make(config_);
     auto dosGuard = web::dosguard::DOSGuard{config_, whitelistHandler, dosguardWeights};
     auto sweepHandler = web::dosguard::IntervalSweepHandler{config_, ioc, dosGuard};
+
     auto cache = data::LedgerCache{};
+    appStopper_.setOnStop([&cache, this](auto&&) {
+        // TODO(kuznetsss): move this into Stopper::makeOnStopCallback()
+        auto const cacheFilePath = config_.maybeValue<std::string>("cache.file.path");
+        if (not cacheFilePath.has_value()) {
+            return;
+        }
+
+        LOG(util::LogService::info()) << "Saving ledger cache to " << *cacheFilePath;
+        if (auto const [success, duration_ms] = util::timed([&]() { return cache.saveToFile(*cacheFilePath); });
+            success.has_value()) {
+            LOG(util::LogService::info()) << "Successfully saved ledger cache in " << duration_ms << " ms";
+        } else {
+            LOG(util::LogService::error()) << "Error saving LedgerCache to file";
+        }
+    });
 
     // Interface to the database
     auto backend = data::makeBackend(config_, cache);
@@ -142,20 +157,12 @@ ClioApplication::run(bool const useNgWebServer)
     // ETL uses the balancer to extract data.
     // The server uses the balancer to forward RPCs to a rippled node.
     // The balancer itself publishes to streams (transactions_proposed and accounts_proposed)
-    auto balancer = [&] -> std::shared_ptr<etlng::LoadBalancerInterface> {
-        if (config_.get<bool>("__ng_etl")) {
-            return etlng::LoadBalancer::makeLoadBalancer(
+    auto balancer = etl::LoadBalancer::makeLoadBalancer(
         config_, ioc, backend, subscriptions, std::make_unique<util::MTRandomGenerator>(), ledgers
     );
-        }
-
-        return etl::LoadBalancer::makeLoadBalancer(
-            config_, ioc, backend, subscriptions, std::make_unique<util::MTRandomGenerator>(), ledgers
-        );
-    }();
 
     // ETL is responsible for writing and publishing to streams. In read-only mode, ETL only publishes
-    auto etl = etl::ETLService::makeETLService(config_, ioc, ctx, backend, subscriptions, balancer, ledgers);
+    auto etl = etl::ETLService::makeETLService(config_, ctx, backend, subscriptions, balancer, ledgers);
 
     auto workQueue = rpc::WorkQueue::makeWorkQueue(config_);
     auto counters = rpc::Counters::makeCounters(workQueue);

src/app/Stopper.hpp

@@ -20,8 +20,8 @@
 #pragma once
 
 #include "data/BackendInterface.hpp"
-#include "etlng/ETLServiceInterface.hpp"
-#include "etlng/LoadBalancerInterface.hpp"
+#include "etl/ETLServiceInterface.hpp"
+#include "etl/LoadBalancerInterface.hpp"
 #include "feed/SubscriptionManagerInterface.hpp"
 #include "util/CoroutineGroup.hpp"
 #include "util/log/Logger.hpp"
@@ -78,8 +78,8 @@ public:
     static std::function<void(boost::asio::yield_context)>
     makeOnStopCallback(
         ServerType& server,
-        etlng::LoadBalancerInterface& balancer,
-        etlng::ETLServiceInterface& etl,
+        etl::LoadBalancerInterface& balancer,
+        etl::ETLServiceInterface& etl,
         feed::SubscriptionManagerInterface& subscriptions,
         data::BackendInterface& backend,
         boost::asio::io_context& ioc

src/data/AmendmentCenter.hpp

@@ -147,6 +147,11 @@ struct Amendments {
     REGISTER(fixAMMClawbackRounding);
     REGISTER(fixMPTDeliveredAmount);
     REGISTER(fixPriceOracleOrder);
+    REGISTER(DynamicMPT);
+    REGISTER(fixDelegateV1_1);
+    REGISTER(fixDirectoryLimit);
+    REGISTER(fixIncludeKeyletFields);
+    REGISTER(fixTokenEscrowV1);
 
     // Obsolete but supported by libxrpl
     REGISTER(CryptoConditionsSuite);

src/data/BackendInterface.cpp

@@ -270,7 +270,7 @@ BackendInterface::updateRange(uint32_t newMax)
 {
     std::scoped_lock const lck(rngMtx_);
 
-    if (range_.has_value() && newMax < range_->maxSequence) {
+    if (range_.has_value() and newMax < range_->maxSequence) {
         ASSERT(
             false,
             "Range shouldn't exist yet or newMax should be at least range->maxSequence. newMax = {}, "
@@ -280,11 +280,14 @@ BackendInterface::updateRange(uint32_t newMax)
         );
     }
 
-    if (!range_.has_value()) {
-        range_ = {.minSequence = newMax, .maxSequence = newMax};
-    } else {
-        range_->maxSequence = newMax;
+    updateRangeImpl(newMax);
 }
+
+void
+BackendInterface::forceUpdateRange(uint32_t newMax)
+{
+    std::scoped_lock const lck(rngMtx_);
+    updateRangeImpl(newMax);
 }
 
 void
@@ -410,4 +413,14 @@ BackendInterface::fetchFees(std::uint32_t const seq, boost::asio::yield_context
     return fees;
 }
 
+void
+BackendInterface::updateRangeImpl(uint32_t newMax)
+{
+    if (!range_.has_value()) {
+        range_ = {.minSequence = newMax, .maxSequence = newMax};
+    } else {
+        range_->maxSequence = newMax;
+    }
+}
+
 }  // namespace data

src/data/BackendInterface.hpp

@@ -249,6 +249,15 @@ public:
     void
     updateRange(uint32_t newMax);
 
+    /**
+     * @brief Updates the range of sequences that are stored in the DB without any checks
+     * @note In the most cases you should use updateRange() instead
+     *
+     * @param newMax The new maximum sequence available
+     */
+    void
+    forceUpdateRange(uint32_t newMax);
+
     /**
      * @brief Sets the range of sequences that are stored in the DB.
      *
@@ -776,6 +785,9 @@ private:
      */
     virtual bool
     doFinishWrites() = 0;
+
+    void
+    updateRangeImpl(uint32_t newMax);
 };
 
 }  // namespace data

src/data/CMakeLists.txt

@@ -14,6 +14,9 @@ target_sources(
           cassandra/impl/SslContext.cpp
           cassandra/Handle.cpp
           cassandra/SettingsProvider.cpp
+          impl/InputFile.cpp
+          impl/LedgerCacheFile.cpp
+          impl/OutputFile.cpp
 )
 
 target_link_libraries(clio_data PUBLIC cassandra-cpp-driver::cassandra-cpp-driver clio_util)

src/data/LedgerCache.cpp

@@ -20,16 +20,22 @@
 #include "data/LedgerCache.hpp"
 
 #include "data/Types.hpp"
-#include "etlng/Models.hpp"
+#include "data/impl/LedgerCacheFile.hpp"
+#include "etl/Models.hpp"
 #include "util/Assert.hpp"
 
 #include <xrpl/basics/base_uint.h>
 
 #include <cstddef>
 #include <cstdint>
+#include <cstdlib>
+#include <cstring>
+#include <map>
 #include <mutex>
 #include <optional>
 #include <shared_mutex>
+#include <string>
+#include <utility>
 #include <vector>
 
 namespace data {
@@ -89,7 +95,7 @@ LedgerCache::update(std::vector<LedgerObject> const& objs, uint32_t seq, bool is
 }
 
 void
-LedgerCache::update(std::vector<etlng::model::Object> const& objs, uint32_t seq)
+LedgerCache::update(std::vector<etl::model::Object> const& objs, uint32_t seq)
 {
     if (disabled_)
         return;
@@ -251,4 +257,34 @@ LedgerCache::getSuccessorHitRate() const
     return static_cast<float>(successorHitCounter_.get().value()) / successorReqCounter_.get().value();
 }
 
+std::expected<void, std::string>
+LedgerCache::saveToFile(std::string const& path) const
+{
+    if (not isFull()) {
+        return std::unexpected{"Ledger cache is not full"};
+    }
+
+    impl::LedgerCacheFile file{path};
+    std::unique_lock lock{mtx_};
+    impl::LedgerCacheFile::DataView data{.latestSeq = latestSeq_, .map = map_, .deleted = deleted_};
+    return file.write(data);
+}
+
+std::expected<void, std::string>
+LedgerCache::loadFromFile(std::string const& path, uint32_t minLatestSequence)
+{
+    impl::LedgerCacheFile file{path};
+    auto data = file.read(minLatestSequence);
+    if (not data.has_value()) {
+        return std::unexpected(std::move(data).error());
+    }
+    auto [latestSeq, map, deleted] = std::move(data).value();
+    std::unique_lock lock{mtx_};
+    latestSeq_ = latestSeq;
+    map_ = std::move(map);
+    deleted_ = std::move(deleted);
+    full_ = true;
+    return {};
+}
+
 }  // namespace data

src/data/LedgerCache.hpp

@@ -21,7 +21,7 @@
 
 #include "data/LedgerCacheInterface.hpp"
 #include "data/Types.hpp"
-#include "etlng/Models.hpp"
+#include "etl/Models.hpp"
 #include "util/prometheus/Bool.hpp"
 #include "util/prometheus/Counter.hpp"
 #include "util/prometheus/Label.hpp"
@@ -37,6 +37,7 @@
 #include <map>
 #include <optional>
 #include <shared_mutex>
+#include <string>
 #include <unordered_set>
 #include <vector>
 
@@ -46,11 +47,16 @@ namespace data {
  * @brief Cache for an entire ledger.
  */
 class LedgerCache : public LedgerCacheInterface {
+public:
+    /** @brief An entry of the cache */
     struct CacheEntry {
         uint32_t seq = 0;
         Blob blob;
     };
 
+    using CacheMap = std::map<ripple::uint256, CacheEntry>;
+
+private:
     // counters for fetchLedgerObject(s) hit rate
     std::reference_wrapper<util::prometheus::CounterInt> objectReqCounter_{PrometheusService::counterInt(
         "ledger_cache_counter_total_number",
@@ -73,8 +79,8 @@ class LedgerCache : public LedgerCacheInterface {
         util::prometheus::Labels({{"type", "cache_hit"}, {"fetch", "successor_key"}})
     )};
 
-    std::map<ripple::uint256, CacheEntry> map_;
-    std::map<ripple::uint256, CacheEntry> deleted_;
+    CacheMap map_;
+    CacheMap deleted_;
 
     mutable std::shared_mutex mtx_;
     std::condition_variable_any cv_;
@@ -98,7 +104,7 @@ public:
     update(std::vector<LedgerObject> const& objs, uint32_t seq, bool isBackground) override;
 
     void
-    update(std::vector<etlng::model::Object> const& objs, uint32_t seq) override;
+    update(std::vector<etl::model::Object> const& objs, uint32_t seq) override;
 
     std::optional<Blob>
     get(ripple::uint256 const& key, uint32_t seq) const override;
@@ -138,6 +144,19 @@ public:
 
     void
     waitUntilCacheContainsSeq(uint32_t seq) override;
+
+    /**
+     * @brief Save the cache to file
+     * @note This operation takes about 7 seconds and it keeps mtx_ exclusively locked
+     *
+     * @param path The file path to save the cache to
+     * @return An error as a string if any
+     */
+    std::expected<void, std::string>
+    saveToFile(std::string const& path) const;
+
+    std::expected<void, std::string>
+    loadFromFile(std::string const& path, uint32_t minLatestSequence) override;
 };
 
 }  // namespace data

src/data/LedgerCacheInterface.hpp

@@ -20,14 +20,16 @@
 #pragma once
 
 #include "data/Types.hpp"
-#include "etlng/Models.hpp"
+#include "etl/Models.hpp"
 
 #include <xrpl/basics/base_uint.h>
 #include <xrpl/basics/hardened_hash.h>
 
 #include <cstddef>
 #include <cstdint>
+#include <expected>
 #include <optional>
+#include <string>
 #include <vector>
 
 namespace data {
@@ -63,7 +65,7 @@ public:
      * @param seq The sequence to update cache for
      */
     virtual void
-    update(std::vector<etlng::model::Object> const& objs, uint32_t seq) = 0;
+    update(std::vector<etl::model::Object> const& objs, uint32_t seq) = 0;
 
     /**
      * @brief Fetch a cached object by its key and sequence number.
@@ -168,6 +170,17 @@ public:
      */
     virtual void
     waitUntilCacheContainsSeq(uint32_t seq) = 0;
+
+    /**
+     * @brief Load the cache from file
+     * @note This operation takes about 7 seconds and it keeps mtx_ exclusively locked
+     *
+     * @param path The file path to load data from
+     * @param minLatestSequence The minimum allowed value of the latestLedgerSequence in cache file
+     * @return An error as a string if any
+     */
+    [[nodiscard]] virtual std::expected<void, std::string>
+    loadFromFile(std::string const& path, uint32_t minLatestSequence) = 0;
 };
 
 }  // namespace data

src/data/Types.hpp

@@ -247,6 +247,9 @@ struct MPTHoldersAndCursor {
 struct LedgerRange {
     std::uint32_t minSequence = 0;
     std::uint32_t maxSequence = 0;
+
+    bool
+    operator==(LedgerRange const&) const = default;
 };
 
 /**

src/data/impl/InputFile.cpp

@@ -0,0 +1,58 @@
+//------------------------------------------------------------------------------
+/*
+    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.
+*/
+//==============================================================================
+
+#include "data/impl/InputFile.hpp"
+
+#include <xrpl/basics/base_uint.h>
+
+#include <cstddef>
+#include <cstring>
+#include <ios>
+#include <iosfwd>
+#include <string>
+#include <utility>
+
+namespace data::impl {
+
+InputFile::InputFile(std::string const& path) : file_(path, std::ios::binary | std::ios::in)
+{
+}
+
+bool
+InputFile::isOpen() const
+{
+    return file_.is_open();
+}
+
+bool
+InputFile::readRaw(char* data, size_t size)
+{
+    file_.read(data, size);
+    shasum_.update(data, size);
+    return not file_.fail();
+}
+
+ripple::uint256
+InputFile::hash() const
+{
+    auto sum = shasum_;
+    return std::move(sum).finalize();
+}
+
+}  // namespace data::impl

src/data/impl/InputFile.hpp

@@ -1,7 +1,7 @@
 //------------------------------------------------------------------------------
 /*
     This file is part of clio: https://github.com/XRPLF/clio
-    Copyright (c) 2023, the clio developers.
+    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
@@ -19,24 +19,39 @@
 
 #pragma once
 
-#include "etl/impl/LedgerLoader.hpp"
-#include "util/FakeFetchResponse.hpp"
+#include "util/Shasum.hpp"
 
-#include <gmock/gmock.h>
-#include <xrpl/protocol/LedgerHeader.h>
+#include <xrpl/basics/base_uint.h>
 
-#include <cstdint>
-#include <optional>
+#include <cstddef>
+#include <cstring>
+#include <fstream>
+#include <iosfwd>
+#include <string>
 
-struct MockLedgerLoader {
-    using GetLedgerResponseType = FakeFetchResponse;
-    using RawLedgerObjectType = FakeLedgerObject;
+namespace data::impl {
 
-    MOCK_METHOD(
-        FormattedTransactionsData,
-        insertTransactions,
-        (ripple::LedgerHeader const&, GetLedgerResponseType& data),
-        ()
-    );
-    MOCK_METHOD(std::optional<ripple::LedgerHeader>, loadInitialLedger, (uint32_t sequence), ());
+class InputFile {
+    std::ifstream file_;
+    util::Sha256sum shasum_;
+
+public:
+    InputFile(std::string const& path);
+
+    bool
+    isOpen() const;
+
+    template <typename T>
+    bool
+    read(T& t)
+    {
+        return readRaw(reinterpret_cast<char*>(&t), sizeof(T));
+    }
+
+    bool
+    readRaw(char* data, size_t size);
+
+    ripple::uint256
+    hash() const;
 };
+}  // namespace data::impl

src/data/impl/LedgerCacheFile.cpp

@@ -0,0 +1,210 @@
+//------------------------------------------------------------------------------
+/*
+    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.
+*/
+//==============================================================================
+
+#include "data/impl/LedgerCacheFile.hpp"
+
+#include "data/LedgerCache.hpp"
+#include "data/Types.hpp"
+
+#include <fmt/format.h>
+#include <xrpl/basics/base_uint.h>
+
+#include <algorithm>
+#include <array>
+#include <cstddef>
+#include <cstdint>
+#include <exception>
+#include <filesystem>
+#include <string>
+#include <utility>
+
+namespace data::impl {
+
+using Hash = ripple::uint256;
+using Separator = std::array<char, 16>;
+static constexpr Separator kSEPARATOR = {0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0};
+
+namespace {
+
+std::expected<std::pair<ripple::uint256, LedgerCache::CacheEntry>, std::string>
+readCacheEntry(InputFile& file, size_t i)
+{
+    ripple::uint256 key;
+    if (not file.readRaw(reinterpret_cast<char*>(key.data()), ripple::base_uint<256>::bytes)) {
+        return std::unexpected(fmt::format("Failed to read key at index {}", i));
+    }
+
+    uint32_t seq{};
+    if (not file.read(seq)) {
+        return std::unexpected(fmt::format("Failed to read sequence at index {}", i));
+    }
+
+    size_t blobSize{};
+    if (not file.read(blobSize)) {
+        return std::unexpected(fmt::format("Failed to read blob size at index {}", i));
+    }
+
+    Blob blob(blobSize);
+    if (not file.readRaw(reinterpret_cast<char*>(blob.data()), blobSize)) {
+        return std::unexpected(fmt::format("Failed to read blob data at index {}", i));
+    }
+    return std::make_pair(key, LedgerCache::CacheEntry{.seq = seq, .blob = std::move(blob)});
+}
+
+std::expected<void, std::string>
+verifySeparator(Separator const& s)
+{
+    if (not std::ranges::all_of(s, [](char c) { return c == 0; })) {
+        return std::unexpected{"Separator verification failed - data corruption detected"};
+    }
+    return {};
+}
+
+}  // anonymous namespace
+
+LedgerCacheFile::LedgerCacheFile(std::string path) : path_(std::move(path))
+{
+}
+
+std::expected<void, std::string>
+LedgerCacheFile::write(DataView dataView)
+{
+    auto const newFilePath = fmt::format("{}.new", path_);
+    auto file = OutputFile{newFilePath};
+    if (not file.isOpen()) {
+        return std::unexpected{fmt::format("Couldn't open file: {}", newFilePath)};
+    }
+
+    Header const header{
+        .latestSeq = dataView.latestSeq, .mapSize = dataView.map.size(), .deletedSize = dataView.deleted.size()
+    };
+    file.write(header);
+    file.write(kSEPARATOR);
+
+    for (auto const& [k, v] : dataView.map) {
+        file.write(k.data(), decltype(k)::bytes);
+        file.write(v.seq);
+        file.write(v.blob.size());
+        file.writeRaw(reinterpret_cast<char const*>(v.blob.data()), v.blob.size());
+    }
+    file.write(kSEPARATOR);
+
+    for (auto const& [k, v] : dataView.deleted) {
+        file.write(k.data(), decltype(k)::bytes);
+        file.write(v.seq);
+        file.write(v.blob.size());
+        file.writeRaw(reinterpret_cast<char const*>(v.blob.data()), v.blob.size());
+    }
+    file.write(kSEPARATOR);
+    auto const hash = file.hash();
+    file.write(hash.data(), decltype(hash)::bytes);
+
+    try {
+        std::filesystem::rename(newFilePath, path_);
+    } catch (std::exception const& e) {
+        return std::unexpected{fmt::format("Error moving cache file from {} to {}: {}", newFilePath, path_, e.what())};
+    }
+
+    return {};
+}
+
+std::expected<LedgerCacheFile::Data, std::string>
+LedgerCacheFile::read(uint32_t minLatestSequence)
+{
+    try {
+        auto file = InputFile{path_};
+        if (not file.isOpen()) {
+            return std::unexpected{fmt::format("Couldn't open file: {}", path_)};
+        }
+
+        Data result;
+
+        Header header{};
+        if (not file.read(header)) {
+            return std::unexpected{"Error reading cache header"};
+        }
+        if (header.version != kVERSION) {
+            return std::unexpected{
+                fmt::format("Cache has wrong version: expected {} found {}", kVERSION, header.version)
+            };
+        }
+        if (header.latestSeq < minLatestSequence) {
+            return std::unexpected{fmt::format("Latest sequence ({}) in the cache file is too low.", header.latestSeq)};
+        }
+        result.latestSeq = header.latestSeq;
+
+        Separator separator{};
+        if (not file.readRaw(separator.data(), separator.size())) {
+            return std::unexpected{"Error reading cache header"};
+        }
+        if (auto verificationResult = verifySeparator(separator); not verificationResult.has_value()) {
+            return std::unexpected{std::move(verificationResult).error()};
+        }
+
+        for (size_t i = 0; i < header.mapSize; ++i) {
+            auto cacheEntryExpected = readCacheEntry(file, i);
+            if (not cacheEntryExpected.has_value()) {
+                return std::unexpected{std::move(cacheEntryExpected).error()};
+            }
+            // Using insert with hint here to decrease insert operation complexity to the amortized constant instead of
+            // logN
+            result.map.insert(result.map.end(), std::move(cacheEntryExpected).value());
+        }
+
+        if (not file.readRaw(separator.data(), separator.size())) {
+            return std::unexpected{"Error reading separator"};
+        }
+        if (auto verificationResult = verifySeparator(separator); not verificationResult.has_value()) {
+            return std::unexpected{std::move(verificationResult).error()};
+        }
+
+        for (size_t i = 0; i < header.deletedSize; ++i) {
+            auto cacheEntryExpected = readCacheEntry(file, i);
+            if (not cacheEntryExpected.has_value()) {
+                return std::unexpected{std::move(cacheEntryExpected).error()};
+            }
+            result.deleted.insert(result.deleted.end(), std::move(cacheEntryExpected).value());
+        }
+
+        if (not file.readRaw(separator.data(), separator.size())) {
+            return std::unexpected{"Error reading separator"};
+        }
+        if (auto verificationResult = verifySeparator(separator); not verificationResult.has_value()) {
+            return std::unexpected{std::move(verificationResult).error()};
+        }
+
+        auto const dataHash = file.hash();
+        ripple::uint256 hashFromFile{};
+        if (not file.readRaw(reinterpret_cast<char*>(hashFromFile.data()), decltype(hashFromFile)::bytes)) {
+            return std::unexpected{"Error reading hash"};
+        }
+
+        if (dataHash != hashFromFile) {
+            return std::unexpected{"Hash file corruption detected"};
+        }
+
+        return result;
+    } catch (std::exception const& e) {
+        return std::unexpected{fmt::format(" Error reading cache file: {}", e.what())};
+    } catch (...) {
+        return std::unexpected{fmt::format(" Error reading cache file")};
+    }
+}
+
+}  // namespace data::impl

src/data/impl/LedgerCacheFile.hpp

@@ -0,0 +1,70 @@
+//------------------------------------------------------------------------------
+/*
+    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/LedgerCache.hpp"
+#include "data/impl/InputFile.hpp"
+#include "data/impl/OutputFile.hpp"
+
+#include <fmt/format.h>
+#include <xrpl/basics/base_uint.h>
+
+#include <cstddef>
+#include <cstdint>
+#include <cstring>
+#include <string>
+
+namespace data::impl {
+
+class LedgerCacheFile {
+public:
+    struct Header {
+        uint32_t version = kVERSION;
+        uint32_t latestSeq{};
+        uint64_t mapSize{};
+        uint64_t deletedSize{};
+    };
+
+private:
+    static constexpr uint32_t kVERSION = 1;
+
+    std::string path_;
+
+public:
+    template <typename T>
+    struct DataBase {
+        uint32_t latestSeq{0};
+        T map;
+        T deleted;
+    };
+
+    using DataView = DataBase<LedgerCache::CacheMap const&>;
+    using Data = DataBase<LedgerCache::CacheMap>;
+
+    LedgerCacheFile(std::string path);
+
+    std::expected<void, std::string>
+    write(DataView dataView);
+
+    std::expected<Data, std::string>
+    read(uint32_t minLatestSequence);
+};
+
+}  // namespace data::impl

src/data/impl/OutputFile.cpp

@@ -0,0 +1,62 @@
+//------------------------------------------------------------------------------
+/*
+    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.
+*/
+//==============================================================================
+
+#include "data/impl/OutputFile.hpp"
+
+#include <xrpl/basics/base_uint.h>
+
+#include <cstddef>
+#include <cstring>
+#include <ios>
+#include <string>
+#include <utility>
+
+namespace data::impl {
+
+OutputFile::OutputFile(std::string const& path) : file_(path, std::ios::binary | std::ios::out)
+{
+}
+
+bool
+OutputFile::isOpen() const
+{
+    return file_.is_open();
+}
+
+void
+OutputFile::writeRaw(char const* data, size_t size)
+{
+    writeToFile(data, size);
+}
+
+void
+OutputFile::writeToFile(char const* data, size_t size)
+{
+    file_.write(data, size);
+    shasum_.update(data, size);
+}
+
+ripple::uint256
+OutputFile::hash() const
+{
+    auto sum = shasum_;
+    return std::move(sum).finalize();
+}
+
+}  // namespace data::impl

src/data/impl/OutputFile.hpp

@@ -0,0 +1,68 @@
+//------------------------------------------------------------------------------
+/*
+    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 "util/Shasum.hpp"
+
+#include <xrpl/basics/base_uint.h>
+
+#include <cstddef>
+#include <cstring>
+#include <fstream>
+#include <string>
+
+namespace data::impl {
+
+class OutputFile {
+    std::ofstream file_;
+    util::Sha256sum shasum_;
+
+public:
+    OutputFile(std::string const& path);
+
+    bool
+    isOpen() const;
+
+    template <typename T>
+    void
+    write(T&& data)
+    {
+        writeRaw(reinterpret_cast<char const*>(&data), sizeof(T));
+    }
+
+    template <typename T>
+    void
+    write(T const* data, size_t const size)
+    {
+        writeRaw(reinterpret_cast<char const*>(data), size);
+    }
+
+    void
+    writeRaw(char const* data, size_t size);
+
+    ripple::uint256
+    hash() const;
+
+private:
+    void
+    writeToFile(char const* data, size_t size);
+};
+
+}  // namespace data::impl

src/etl/AmendmentBlockHandlerInterface.hpp

@@ -19,7 +19,7 @@
 
 #pragma once
 
-namespace etlng {
+namespace etl {
 
 /**
  * @brief The interface of a handler for amendment blocking
@@ -40,4 +40,4 @@ struct AmendmentBlockHandlerInterface {
     stop() = 0;
 };
 
-}  // namespace etlng
+}  // namespace etl

src/etl/CMakeLists.txt

@@ -7,14 +7,24 @@ target_sources(
           ETLService.cpp
           ETLState.cpp
           LoadBalancer.cpp
+          MPTHelpers.cpp
           NetworkValidatedLedgers.cpp
           NFTHelpers.cpp
           Source.cpp
-          MPTHelpers.cpp
           impl/AmendmentBlockHandler.cpp
+          impl/AsyncGrpcCall.cpp
+          impl/Extraction.cpp
           impl/ForwardingSource.cpp
           impl/GrpcSource.cpp
+          impl/Loading.cpp
+          impl/Monitor.cpp
           impl/SubscriptionSource.cpp
+          impl/TaskManager.cpp
+          impl/ext/Cache.cpp
+          impl/ext/Core.cpp
+          impl/ext/MPT.cpp
+          impl/ext/NFT.cpp
+          impl/ext/Successor.cpp
 )
 
 target_link_libraries(clio_etl PUBLIC clio_data)

src/etl/CacheLoader.hpp

@@ -21,16 +21,20 @@
 
 #include "data/BackendInterface.hpp"
 #include "data/LedgerCacheInterface.hpp"
+#include "data/Types.hpp"
+#include "etl/CacheLoaderInterface.hpp"
 #include "etl/CacheLoaderSettings.hpp"
 #include "etl/impl/CacheLoader.hpp"
 #include "etl/impl/CursorFromAccountProvider.hpp"
 #include "etl/impl/CursorFromDiffProvider.hpp"
 #include "etl/impl/CursorFromFixDiffNumProvider.hpp"
-#include "etlng/CacheLoaderInterface.hpp"
 #include "util/Assert.hpp"
+#include "util/Profiler.hpp"
 #include "util/async/context/BasicExecutionContext.hpp"
+#include "util/config/ConfigDefinition.hpp"
 #include "util/log/Logger.hpp"
 
+#include <algorithm>
 #include <cstdint>
 #include <functional>
 #include <memory>
@@ -48,7 +52,7 @@ namespace etl {
  * @tparam ExecutionContextType The type of the execution context to use
  */
 template <typename ExecutionContextType = util::async::CoroExecutionContext>
-class CacheLoader : public etlng::CacheLoaderInterface {
+class CacheLoader : public CacheLoaderInterface {
     using CacheLoaderType = impl::CacheLoaderImpl<data::LedgerCacheInterface>;
 
     util::Logger log_{"ETL"};
@@ -98,6 +102,10 @@ public:
             return;
         }
 
+        if (loadCacheFromFile()) {
+            return;
+        }
+
         std::shared_ptr<impl::BaseCursorProvider> provider;
         if (settings_.numCacheCursorsFromDiff != 0) {
             LOG(log_.info()) << "Loading cache with cursor from num_cursors_from_diff="
@@ -149,6 +157,36 @@ public:
         if (loader_ != nullptr)
             loader_->wait();
     }
+
+private:
+    bool
+    loadCacheFromFile()
+    {
+        if (not settings_.cacheFileSettings.has_value()) {
+            return false;
+        }
+        LOG(log_.info()) << "Loading ledger cache from " << settings_.cacheFileSettings->path;
+        auto const minLatestSequence =
+            backend_->fetchLedgerRange()
+                .transform([this](data::LedgerRange const& range) {
+                    return std::max(range.maxSequence - settings_.cacheFileSettings->maxAge, range.minSequence);
+                })
+                .value_or(0);
+
+        auto const [success, duration_ms] = util::timed([&]() {
+            return cache_.get().loadFromFile(settings_.cacheFileSettings->path, minLatestSequence);
+        });
+
+        if (not success.has_value()) {
+            LOG(log_.warn()) << "Error loading cache from file: " << success.error();
+            return false;
+        }
+
+        LOG(log_.info()) << "Loaded cache from file in " << duration_ms
+                         << " ms. Latest sequence: " << cache_.get().latestLedgerSequence();
+        backend_->forceUpdateRange(cache_.get().latestLedgerSequence());
+        return true;
+    }
 };
 
 }  // namespace etl

src/etl/CacheLoaderInterface.hpp

@@ -21,7 +21,7 @@
 
 #include <cstdint>
 
-namespace etlng {
+namespace etl {
 
 /**
  * @brief An interface for the Cache Loader
@@ -50,4 +50,4 @@ struct CacheLoaderInterface {
     wait() noexcept = 0;
 };
 
-}  // namespace etlng
+}  // namespace etl

src/etl/CacheLoaderSettings.cpp

@@ -26,6 +26,7 @@
 #include <cstddef>
 #include <cstdint>
 #include <string>
+#include <utility>
 
 namespace etl {
 
@@ -63,6 +64,12 @@ makeCacheLoaderSettings(util::config::ClioConfigDefinition const& config)
     settings.numCacheMarkers = cache.get<std::size_t>("num_markers");
     settings.cachePageFetchSize = cache.get<std::size_t>("page_fetch_size");
 
+    if (auto filePath = cache.maybeValue<std::string>("file.path"); filePath.has_value()) {
+        settings.cacheFileSettings = CacheLoaderSettings::CacheFileSettings{
+            .path = std::move(filePath).value(), .maxAge = cache.get<uint32_t>("file.max_sequence_age")
+        };
+    }
+
     auto const entry = cache.get<std::string>("load");
     if (boost::iequals(entry, "sync"))
         settings.loadStyle = CacheLoaderSettings::LoadStyle::SYNC;

src/etl/CacheLoaderSettings.hpp

@@ -22,6 +22,9 @@
 #include "util/config/ConfigDefinition.hpp"
 
 #include <cstddef>
+#include <cstdint>
+#include <optional>
+#include <string>
 
 namespace etl {
 
@@ -32,6 +35,15 @@ struct CacheLoaderSettings {
     /** @brief Ways to load the cache */
     enum class LoadStyle { ASYNC, SYNC, NONE };
 
+    /** @brief Settings for cache file operations */
+    struct CacheFileSettings {
+        std::string path;       /**< path to the file to load cache from on start and save cache to on shutdown */
+        uint32_t maxAge = 5000; /**< max difference between latest sequence in cache file and DB */
+
+        auto
+        operator<=>(CacheFileSettings const&) const = default;
+    };
+
     size_t numCacheDiffs = 32;             /**< number of diffs to use to generate cursors */
     size_t numCacheMarkers = 48;           /**< number of markers to use at one time to traverse the ledger */
     size_t cachePageFetchSize = 512;       /**< number of ledger objects to fetch concurrently per marker */
@@ -40,6 +52,7 @@ struct CacheLoaderSettings {
     size_t numCacheCursorsFromAccount = 0; /**< number of cursors to fetch from account_tx */
 
     LoadStyle loadStyle = LoadStyle::ASYNC;             /**< how to load the cache */
+    std::optional<CacheFileSettings> cacheFileSettings; /**< optional settings for cache file operations */
 
     auto
     operator<=>(CacheLoaderSettings const&) const = default;

src/etl/CacheUpdaterInterface.hpp

@@ -20,12 +20,12 @@
 #pragma once
 
 #include "data/Types.hpp"
-#include "etlng/Models.hpp"
+#include "etl/Models.hpp"
 
 #include <cstdint>
 #include <vector>
 
-namespace etlng {
+namespace etl {
 
 /**
  * @brief An interface for the Cache Updater
@@ -63,4 +63,4 @@ struct CacheUpdaterInterface {
     setFull() = 0;
 };
 
-}  // namespace etlng
+}  // namespace etl

src/etl/ETLService.cpp

@@ -1,7 +1,7 @@
 //------------------------------------------------------------------------------
 /*
     This file is part of clio: https://github.com/XRPLF/clio
-    Copyright (c) 2023, the clio developers.
+    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
@@ -20,101 +20,103 @@
 #include "etl/ETLService.hpp"
 
 #include "data/BackendInterface.hpp"
+#include "data/LedgerCacheInterface.hpp"
+#include "data/Types.hpp"
 #include "etl/CacheLoader.hpp"
+#include "etl/CacheLoaderInterface.hpp"
+#include "etl/CacheUpdaterInterface.hpp"
 #include "etl/CorruptionDetector.hpp"
-#include "etl/LoadBalancer.hpp"
+#include "etl/ETLServiceInterface.hpp"
+#include "etl/ETLState.hpp"
+#include "etl/ExtractorInterface.hpp"
+#include "etl/InitialLoadObserverInterface.hpp"
+#include "etl/LedgerPublisherInterface.hpp"
+#include "etl/LoadBalancerInterface.hpp"
+#include "etl/LoaderInterface.hpp"
+#include "etl/MonitorInterface.hpp"
+#include "etl/MonitorProviderInterface.hpp"
 #include "etl/NetworkValidatedLedgersInterface.hpp"
 #include "etl/SystemState.hpp"
+#include "etl/TaskManagerProviderInterface.hpp"
 #include "etl/impl/AmendmentBlockHandler.hpp"
-#include "etl/impl/ExtractionDataPipe.hpp"
-#include "etl/impl/Extractor.hpp"
+#include "etl/impl/CacheUpdater.hpp"
+#include "etl/impl/Extraction.hpp"
 #include "etl/impl/LedgerFetcher.hpp"
-#include "etl/impl/LedgerLoader.hpp"
 #include "etl/impl/LedgerPublisher.hpp"
-#include "etl/impl/Transformer.hpp"
-#include "etlng/ETLService.hpp"
-#include "etlng/ETLServiceInterface.hpp"
-#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"
-#include "etlng/impl/ext/MPT.hpp"
-#include "etlng/impl/ext/NFT.hpp"
-#include "etlng/impl/ext/Successor.hpp"
+#include "etl/impl/Loading.hpp"
+#include "etl/impl/MonitorProvider.hpp"
+#include "etl/impl/Registry.hpp"
+#include "etl/impl/Scheduling.hpp"
+#include "etl/impl/TaskManager.hpp"
+#include "etl/impl/TaskManagerProvider.hpp"
+#include "etl/impl/ext/Cache.hpp"
+#include "etl/impl/ext/Core.hpp"
+#include "etl/impl/ext/MPT.hpp"
+#include "etl/impl/ext/NFT.hpp"
+#include "etl/impl/ext/Successor.hpp"
 #include "feed/SubscriptionManagerInterface.hpp"
 #include "util/Assert.hpp"
-#include "util/Constants.hpp"
+#include "util/Profiler.hpp"
 #include "util/async/AnyExecutionContext.hpp"
-#include "util/config/ConfigDefinition.hpp"
 #include "util/log/Logger.hpp"
 
-#include <boost/asio/io_context.hpp>
-#include <xrpl/beast/core/CurrentThreadName.h>
+#include <boost/json/object.hpp>
+#include <boost/signals2/connection.hpp>
 #include <xrpl/protocol/LedgerHeader.h>
 
 #include <chrono>
 #include <cstddef>
 #include <cstdint>
+#include <functional>
 #include <memory>
 #include <optional>
-#include <stdexcept>
-#include <thread>
+#include <string>
 #include <utility>
-#include <vector>
 
 namespace etl {
 
-std::shared_ptr<etlng::ETLServiceInterface>
+std::shared_ptr<ETLServiceInterface>
 ETLService::makeETLService(
     util::config::ClioConfigDefinition const& config,
-    boost::asio::io_context& ioc,
     util::async::AnyExecutionContext ctx,
     std::shared_ptr<BackendInterface> backend,
     std::shared_ptr<feed::SubscriptionManagerInterface> subscriptions,
-    std::shared_ptr<etlng::LoadBalancerInterface> balancer,
+    std::shared_ptr<LoadBalancerInterface> balancer,
     std::shared_ptr<NetworkValidatedLedgersInterface> ledgers
 )
 {
-    std::shared_ptr<etlng::ETLServiceInterface> ret;
+    std::shared_ptr<ETLServiceInterface> ret;
 
-    if (config.get<bool>("__ng_etl")) {
-        ASSERT(
-            std::dynamic_pointer_cast<etlng::LoadBalancer>(balancer), "LoadBalancer type must be etlng::LoadBalancer"
-        );
-
-        auto state = std::make_shared<etl::SystemState>();
+    auto state = std::make_shared<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);
-        auto publisher = std::make_shared<etlng::impl::LedgerPublisher>(ioc, backend, subscriptions, *state);
-        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>();
+    auto fetcher = std::make_shared<impl::LedgerFetcher>(backend, balancer);
+    auto extractor = std::make_shared<impl::Extractor>(fetcher);
+    auto publisher = std::make_shared<impl::LedgerPublisher>(ctx, backend, subscriptions, *state);
+    auto cacheLoader = std::make_shared<CacheLoader<>>(config, backend, backend->cache());
+    auto cacheUpdater = std::make_shared<impl::CacheUpdater>(backend->cache());
+    auto amendmentBlockHandler = std::make_shared<impl::AmendmentBlockHandler>(ctx, *state);
+    auto monitorProvider = std::make_shared<impl::MonitorProvider>();
 
     backend->setCorruptionDetector(CorruptionDetector{*state, backend->cache()});
 
-        auto loader = std::make_shared<etlng::impl::Loader>(
+    auto loader = std::make_shared<impl::Loader>(
         backend,
-            etlng::impl::makeRegistry(
+        impl::makeRegistry(
             *state,
-                etlng::impl::CacheExt{cacheUpdater},
-                etlng::impl::CoreExt{backend},
-                etlng::impl::SuccessorExt{backend, backend->cache()},
-                etlng::impl::NFTExt{backend},
-                etlng::impl::MPTExt{backend}
+            impl::CacheExt{cacheUpdater},
+            impl::CoreExt{backend},
+            impl::SuccessorExt{backend, backend->cache()},
+            impl::NFTExt{backend},
+            impl::MPTExt{backend}
         ),
         amendmentBlockHandler,
         state
     );
 
-        auto taskManagerProvider = std::make_shared<etlng::impl::TaskManagerProvider>(*ledgers, extractor, loader);
+    auto taskManagerProvider = std::make_shared<impl::TaskManagerProvider>(*ledgers, extractor, loader);
 
-        ret = std::make_shared<etlng::ETLService>(
+    ret = std::make_shared<ETLService>(
         ctx,
         config,
         backend,
@@ -130,261 +132,280 @@ ETLService::makeETLService(
         monitorProvider,
         state
     );
-    } else {
-        ASSERT(std::dynamic_pointer_cast<etl::LoadBalancer>(balancer), "LoadBalancer type must be etl::LoadBalancer");
-        ret = std::make_shared<etl::ETLService>(config, ioc, backend, subscriptions, balancer, ledgers);
-    }
 
     // inject networkID into subscriptions, as transaction feed require it to inject CTID in response
-    if (auto const state = ret->getETLState(); state)
-        subscriptions->setNetworkID(state->networkID);
+    if (auto const etlState = ret->getETLState(); etlState)
+        subscriptions->setNetworkID(etlState->networkID);
 
     ret->run();
     return ret;
 }
 
-// Database must be populated when this starts
-std::optional<uint32_t>
-ETLService::runETLPipeline(uint32_t startSequence, uint32_t numExtractors)
-{
-    if (finishSequence_ && startSequence > *finishSequence_)
-        return {};
-
-    LOG(log_.debug()) << "Wait for cache containing seq " << startSequence - 1
-                      << " current cache last seq =" << backend_->cache().latestLedgerSequence();
-    backend_->cache().waitUntilCacheContainsSeq(startSequence - 1);
-
-    LOG(log_.debug()) << "Starting etl pipeline";
-    state_.isWriting = true;
-
-    auto const rng = backend_->hardFetchLedgerRangeNoThrow();
-    ASSERT(rng.has_value(), "Parent ledger range can't be null");
-    ASSERT(
-        rng->maxSequence >= startSequence - 1,
-        "Got not parent ledger. rnd->maxSequence = {}, startSequence = {}",
-        rng->maxSequence,
-        startSequence
-    );
-
-    auto const begin = std::chrono::system_clock::now();
-    auto extractors = std::vector<std::unique_ptr<ExtractorType>>{};
-    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_
+ETLService::ETLService(
+    util::async::AnyExecutionContext ctx,
+    std::reference_wrapper<util::config::ClioConfigDefinition const> config,
+    std::shared_ptr<data::BackendInterface> backend,
+    std::shared_ptr<LoadBalancerInterface> balancer,
+    std::shared_ptr<NetworkValidatedLedgersInterface> ledgers,
+    std::shared_ptr<LedgerPublisherInterface> publisher,
+    std::shared_ptr<CacheLoaderInterface> cacheLoader,
+    std::shared_ptr<CacheUpdaterInterface> cacheUpdater,
+    std::shared_ptr<ExtractorInterface> extractor,
+    std::shared_ptr<LoaderInterface> loader,
+    std::shared_ptr<InitialLoadObserverInterface> initialLoadObserver,
+    std::shared_ptr<TaskManagerProviderInterface> taskManagerProvider,
+    std::shared_ptr<MonitorProviderInterface> monitorProvider,
+    std::shared_ptr<SystemState> state
 )
-        );
-    }
-
-    auto transformer =
-        TransformerType{pipe, backend_, ledgerLoader_, ledgerPublisher_, amendmentBlockHandler_, startSequence, state_};
-    transformer.waitTillFinished();  // suspend current thread until exit condition is met
-    pipe.cleanup();                  // TODO: this should probably happen automatically using destructor
-
-    // wait for all of the extractors to stop
-    for (auto& t : extractors)
-        t->waitTillFinished();
-
-    auto const end = std::chrono::system_clock::now();
-    auto const lastPublishedSeq = ledgerPublisher_.getLastPublishedSequence();
-    static constexpr auto kNANOSECONDS_PER_SECOND = 1'000'000'000.0;
-    LOG(log_.debug()) << "Extracted and wrote " << lastPublishedSeq.value_or(startSequence) - startSequence << " in "
-                      << ((end - begin).count()) / kNANOSECONDS_PER_SECOND;
-
-    state_.isWriting = false;
-
-    LOG(log_.debug()) << "Stopping etl pipeline";
-    return lastPublishedSeq;
-}
-
-// Main loop of ETL.
-// The software begins monitoring the ledgers that are validated by the network.
-// The member networkValidatedLedgers_ keeps track of the sequences of ledgers validated by the network.
-// Whenever a ledger is validated by the network, the software looks for that ledger in the database. Once the ledger is
-// found in the database, the software publishes that ledger to the ledgers stream. If a network validated ledger is not
-// found in the database after a certain amount of time, then the software attempts to take over responsibility of the
-// ETL process, where it writes new ledgers to the database. The software will relinquish control of the ETL process if
-// it detects that another process has taken over ETL.
-void
-ETLService::monitor()
+    : ctx_(std::move(ctx))
+    , config_(config)
+    , backend_(std::move(backend))
+    , balancer_(std::move(balancer))
+    , ledgers_(std::move(ledgers))
+    , publisher_(std::move(publisher))
+    , cacheLoader_(std::move(cacheLoader))
+    , cacheUpdater_(std::move(cacheUpdater))
+    , extractor_(std::move(extractor))
+    , 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"))
 {
-    auto rng = backend_->hardFetchLedgerRangeNoThrow();
-    if (!rng) {
-        LOG(log_.info()) << "Database is empty. Will download a ledger from the network.";
-        std::optional<ripple::LedgerHeader> ledger;
+    ASSERT(not state_->isWriting, "ETL should never start in writer mode");
 
-        try {
-            if (startSequence_) {
-                LOG(log_.info()) << "ledger sequence specified in config. "
-                                 << "Will begin ETL process starting with ledger " << *startSequence_;
-                ledger = ledgerLoader_.loadInitialLedger(*startSequence_);
-            } else {
-                LOG(log_.info()) << "Waiting for next ledger to be validated by network...";
-                std::optional<uint32_t> mostRecentValidated = networkValidatedLedgers_->getMostRecent();
+    if (startSequence_.has_value())
+        LOG(log_.info()) << "Start sequence: " << *startSequence_;
 
-                if (mostRecentValidated) {
-                    LOG(log_.info()) << "Ledger " << *mostRecentValidated << " has been validated. Downloading...";
-                    ledger = ledgerLoader_.loadInitialLedger(*mostRecentValidated);
-                } else {
-                    LOG(log_.info()) << "The wait for the next validated ledger has been aborted. "
-                                        "Exiting monitor loop";
-                    return;
-                }
-            }
-        } catch (std::runtime_error const& e) {
-            LOG(log_.fatal()) << "Failed to load initial ledger: " << e.what();
-            amendmentBlockHandler_.notifyAmendmentBlocked();
-            return;
+    if (finishSequence_.has_value())
+        LOG(log_.info()) << "Finish sequence: " << *finishSequence_;
+
+    LOG(log_.info()) << "Starting in " << (state_->isStrictReadonly ? "STRICT READONLY MODE" : "WRITE MODE");
 }
 
-        if (ledger) {
-            rng = backend_->hardFetchLedgerRangeNoThrow();
-        } else {
-            LOG(log_.error()) << "Failed to load initial ledger. Exiting monitor loop";
-            return;
-        }
-    } else {
-        if (startSequence_)
-            LOG(log_.warn()) << "start sequence specified but db is already populated";
-
-        LOG(log_.info()) << "Database already populated. Picking up from the tip of history";
-        cacheLoader_.load(rng->maxSequence);
-    }
-
-    ASSERT(rng.has_value(), "Ledger range can't be null");
-    uint32_t nextSequence = rng->maxSequence + 1;
-
-    LOG(log_.debug()) << "Database is populated. Starting monitor loop. sequence = " << nextSequence;
-
-    while (not isStopping()) {
-        nextSequence = publishNextSequence(nextSequence);
-    }
-}
-
-uint32_t
-ETLService::publishNextSequence(uint32_t nextSequence)
+ETLService::~ETLService()
 {
-    if (auto rng = backend_->hardFetchLedgerRangeNoThrow(); rng && rng->maxSequence >= nextSequence) {
-        ledgerPublisher_.publish(nextSequence, {});
-        ++nextSequence;
-    } else if (networkValidatedLedgers_->waitUntilValidatedByNetwork(nextSequence, util::kMILLISECONDS_PER_SECOND)) {
-        LOG(log_.info()) << "Ledger with sequence = " << nextSequence << " has been validated by the network. "
-                         << "Attempting to find in database and publish";
-
-        // Attempt to take over responsibility of ETL writer after 10 failed
-        // attempts to publish the ledger. publishLedger() fails if the
-        // ledger that has been validated by the network is not found in the
-        // database after the specified number of attempts. publishLedger()
-        // waits one second between each attempt to read the ledger from the
-        // database
-        constexpr size_t kTIMEOUT_SECONDS = 10;
-        bool const success = ledgerPublisher_.publish(nextSequence, kTIMEOUT_SECONDS);
-
-        if (!success) {
-            LOG(log_.warn()) << "Failed to publish ledger with sequence = " << nextSequence << " . Beginning ETL";
-
-            // returns the most recent sequence published. empty optional if no sequence was published
-            std::optional<uint32_t> lastPublished = runETLPipeline(nextSequence, extractorThreads_);
-            LOG(log_.info()) << "Aborting ETL. Falling back to publishing";
-
-            // if no ledger was published, don't increment nextSequence
-            if (lastPublished)
-                nextSequence = *lastPublished + 1;
-        } else {
-            ++nextSequence;
-        }
-    }
-    return nextSequence;
-}
-
-void
-ETLService::monitorReadOnly()
-{
-    LOG(log_.debug()) << "Starting reporting in strict read only mode";
-
-    auto const latestSequenceOpt = [this]() -> std::optional<uint32_t> {
-        auto rng = backend_->hardFetchLedgerRangeNoThrow();
-
-        if (!rng) {
-            if (auto net = networkValidatedLedgers_->getMostRecent()) {
-                return net;
-            }
-            return std::nullopt;
-        }
-
-        return rng->maxSequence;
-    }();
-
-    if (!latestSequenceOpt.has_value()) {
-        return;
-    }
-
-    uint32_t latestSequence = *latestSequenceOpt;
-
-    cacheLoader_.load(latestSequence);
-    latestSequence++;
-
-    while (not isStopping()) {
-        if (auto rng = backend_->hardFetchLedgerRangeNoThrow(); rng && rng->maxSequence >= latestSequence) {
-            ledgerPublisher_.publish(latestSequence, {});
-            latestSequence = latestSequence + 1;
-        } else {
-            // if we can't, wait until it's validated by the network, or 1 second passes, whichever occurs
-            // first. Even if we don't hear from rippled, if ledgers are being written to the db, we publish
-            // them.
-            networkValidatedLedgers_->waitUntilValidatedByNetwork(latestSequence, util::kMILLISECONDS_PER_SECOND);
-        }
-    }
+    stop();
+    LOG(log_.debug()) << "Destroying ETL";
 }
 
 void
 ETLService::run()
 {
-    LOG(log_.info()) << "Starting reporting etl";
-    state_.isStopping = false;
+    LOG(log_.info()) << "Running ETL...";
 
-    doWork();
+    mainLoop_.emplace(ctx_.execute([this] {
+        auto const rng = loadInitialLedgerIfNeeded();
+
+        LOG(log_.info()) << "Waiting for next ledger to be validated by network...";
+        std::optional<uint32_t> const mostRecentValidated = ledgers_->getMostRecent();
+
+        if (not mostRecentValidated) {
+            LOG(log_.info()) << "The wait for the next validated ledger has been aborted. "
+                                "Exiting monitor loop";
+            return;
+        }
+
+        if (not rng.has_value()) {
+            LOG(log_.warn()) << "Initial ledger download got cancelled - stopping ETL service";
+            return;
+        }
+
+        auto nextSequence = rng->maxSequence + 1;
+        if (backend_->cache().latestLedgerSequence() != 0) {
+            nextSequence = backend_->cache().latestLedgerSequence();
+        }
+
+        LOG(log_.debug()) << "Database is populated. Starting monitor loop. sequence = " << nextSequence;
+        startMonitor(nextSequence);
+
+        // If we are a writer as the result of loading the initial ledger - start loading
+        if (state_->isWriting)
+            startLoading(nextSequence);
+    }));
 }
 
 void
-ETLService::doWork()
+ETLService::stop()
 {
-    worker_ = std::thread([this]() {
-        beast::setCurrentThreadName("ETLService worker");
+    LOG(log_.info()) << "Stop called";
 
-        if (state_.isStrictReadonly) {
-            monitorReadOnly();
-        } else {
-            monitor();
+    if (mainLoop_)
+        mainLoop_->wait();
+    if (taskMan_)
+        taskMan_->stop();
+    if (monitor_)
+        monitor_->stop();
 }
+
+boost::json::object
+ETLService::getInfo() const
+{
+    boost::json::object result;
+
+    result["etl_sources"] = balancer_->toJson();
+    result["is_writer"] = static_cast<int>(state_->isWriting);
+    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());
+    return result;
+}
+
+bool
+ETLService::isAmendmentBlocked() const
+{
+    return state_->isAmendmentBlocked;
+}
+
+bool
+ETLService::isCorruptionDetected() const
+{
+    return state_->isCorruptionDetected;
+}
+
+std::optional<ETLState>
+ETLService::getETLState() const
+{
+    return balancer_->getETLState();
+}
+
+std::uint32_t
+ETLService::lastCloseAgeSeconds() const
+{
+    return publisher_->lastCloseAgeSeconds();
+}
+
+std::optional<data::LedgerRange>
+ETLService::loadInitialLedgerIfNeeded()
+{
+    auto rng = backend_->hardFetchLedgerRangeNoThrow();
+    if (not rng.has_value()) {
+        ASSERT(
+            not state_->isStrictReadonly,
+            "Database is empty but this node is in strict readonly mode. Can't write initial ledger."
+        );
+
+        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) -> 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;
+                        }
+
+                        ASSERT(res.has_value(), "Initial ledger retry logic failed");
+                        data.edgeKeys = std::move(res).value();
+
+                        return loader_->loadInitialLedger(data);
+                    }
+                );
             });
+
+            if (not ledger.has_value()) {
+                LOG(log_.error()) << "Failed to load initial ledger. Exiting monitor loop";
+                return std::nullopt;
             }
 
-ETLService::ETLService(
-    util::config::ClioConfigDefinition const& config,
-    boost::asio::io_context& ioc,
-    std::shared_ptr<BackendInterface> backend,
-    std::shared_ptr<feed::SubscriptionManagerInterface> subscriptions,
-    std::shared_ptr<etlng::LoadBalancerInterface> balancer,
-    std::shared_ptr<NetworkValidatedLedgersInterface> ledgers
-)
-    : backend_(backend)
-    , loadBalancer_(balancer)
-    , networkValidatedLedgers_(std::move(ledgers))
-    , cacheLoader_(config, backend, backend->cache())
-    , ledgerFetcher_(backend, balancer)
-    , ledgerLoader_(backend, balancer, ledgerFetcher_, state_)
-    , ledgerPublisher_(ioc, backend, backend->cache(), subscriptions, state_)
-    , amendmentBlockHandler_(ioc, state_)
+            LOG(log_.debug()) << "Time to download and store ledger = " << timeDiff;
+            LOG(log_.info()) << "Finished loadInitialLedger. cache size = " << backend_->cache().size();
+
+            return backend_->hardFetchLedgerRangeNoThrow();
+        }
+
+        LOG(log_.info()) << "The wait for the next validated ledger has been aborted. "
+                            "Exiting monitor loop";
+        return std::nullopt;
+    }
+
+    LOG(log_.info()) << "Database already populated. Picking up from the tip of history";
+    if (not backend_->cache().isFull()) {
+        cacheLoader_->load(rng->maxSequence);
+    }
+
+    return rng;
+}
+
+void
+ETLService::startMonitor(uint32_t seq)
 {
-    startSequence_ = config.maybeValue<uint32_t>("start_sequence");
-    finishSequence_ = config.maybeValue<uint32_t>("finish_sequence");
-    state_.isStrictReadonly = config.get<bool>("read_only");
-    extractorThreads_ = config.get<uint32_t>("extractor_threads");
+    monitor_ = monitorProvider_->make(ctx_, backend_, ledgers_, seq);
 
-    // This should probably be done in the backend factory but we don't have state available until here
-    backend_->setCorruptionDetector(CorruptionDetector{state_, backend->cache()});
+    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();
         }
+
+        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)
+{
+    ASSERT(not state_->isStrictReadonly, "This should only happen on writer nodes");
+    taskMan_ = taskManagerProvider_->make(ctx_, *monitor_, seq, finishSequence_);
+
+    // FIXME: this legacy name "extractor_threads" is no longer accurate (we have coroutines now)
+    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 etl

src/etl/ETLService.hpp

@@ -1,7 +1,7 @@
 //------------------------------------------------------------------------------
 /*
     This file is part of clio: https://github.com/XRPLF/clio
-    Copyright (c) 2023, the clio developers.
+    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
@@ -20,57 +20,64 @@
 #pragma once
 
 #include "data/BackendInterface.hpp"
-#include "etl/CacheLoader.hpp"
+#include "data/Types.hpp"
+#include "etl/CacheLoaderInterface.hpp"
+#include "etl/CacheUpdaterInterface.hpp"
+#include "etl/ETLServiceInterface.hpp"
 #include "etl/ETLState.hpp"
+#include "etl/ExtractorInterface.hpp"
+#include "etl/InitialLoadObserverInterface.hpp"
+#include "etl/LedgerPublisherInterface.hpp"
+#include "etl/LoadBalancerInterface.hpp"
+#include "etl/LoaderInterface.hpp"
+#include "etl/MonitorInterface.hpp"
+#include "etl/MonitorProviderInterface.hpp"
 #include "etl/NetworkValidatedLedgersInterface.hpp"
 #include "etl/SystemState.hpp"
+#include "etl/TaskManagerInterface.hpp"
+#include "etl/TaskManagerProviderInterface.hpp"
 #include "etl/impl/AmendmentBlockHandler.hpp"
-#include "etl/impl/ExtractionDataPipe.hpp"
-#include "etl/impl/Extractor.hpp"
+#include "etl/impl/CacheUpdater.hpp"
+#include "etl/impl/Extraction.hpp"
 #include "etl/impl/LedgerFetcher.hpp"
-#include "etl/impl/LedgerLoader.hpp"
 #include "etl/impl/LedgerPublisher.hpp"
-#include "etl/impl/Transformer.hpp"
-#include "etlng/ETLServiceInterface.hpp"
-#include "etlng/LoadBalancerInterface.hpp"
-#include "etlng/impl/LedgerPublisher.hpp"
-#include "etlng/impl/TaskManagerProvider.hpp"
+#include "etl/impl/Loading.hpp"
+#include "etl/impl/Registry.hpp"
+#include "etl/impl/Scheduling.hpp"
+#include "etl/impl/TaskManager.hpp"
+#include "etl/impl/ext/Cache.hpp"
+#include "etl/impl/ext/Core.hpp"
+#include "etl/impl/ext/NFT.hpp"
+#include "etl/impl/ext/Successor.hpp"
 #include "feed/SubscriptionManagerInterface.hpp"
 #include "util/async/AnyExecutionContext.hpp"
+#include "util/async/AnyOperation.hpp"
+#include "util/config/ConfigDefinition.hpp"
 #include "util/log/Logger.hpp"
 
 #include <boost/asio/io_context.hpp>
 #include <boost/json/object.hpp>
-#include <grpcpp/grpcpp.h>
-#include <org/xrpl/rpc/v1/get_ledger.pb.h>
-#include <xrpl/proto/org/xrpl/rpc/v1/xrp_ledger.grpc.pb.h>
+#include <boost/signals2/connection.hpp>
+#include <fmt/format.h>
+#include <xrpl/basics/Blob.h>
+#include <xrpl/basics/base_uint.h>
+#include <xrpl/basics/strHex.h>
+#include <xrpl/proto/org/xrpl/rpc/v1/get_ledger.pb.h>
+#include <xrpl/proto/org/xrpl/rpc/v1/ledger.pb.h>
+#include <xrpl/protocol/LedgerHeader.h>
+#include <xrpl/protocol/STTx.h>
+#include <xrpl/protocol/TxFormats.h>
+#include <xrpl/protocol/TxMeta.h>
 
 #include <cstddef>
 #include <cstdint>
+#include <functional>
 #include <memory>
 #include <optional>
 #include <string>
-#include <thread>
 
-struct AccountTransactionsData;
-struct NFTTransactionsData;
-struct NFTsData;
-
-/**
- * @brief This namespace contains everything to do with the ETL and ETL sources.
- */
 namespace etl {
 
-/**
- * @brief A tag class to help identify ETLService in templated code.
- */
-struct ETLServiceTag {
-    virtual ~ETLServiceTag() = default;
-};
-
-template <typename T>
-concept SomeETLService = std::derived_from<T, ETLServiceTag>;
-
 /**
  * @brief This class is responsible for continuously extracting data from a p2p node, and writing that data to the
  * databases.
@@ -84,71 +91,42 @@ concept SomeETLService = std::derived_from<T, ETLServiceTag>;
  * the others will fall back to monitoring/publishing. In this sense, this class dynamically transitions from monitoring
  * to writing and from writing to monitoring, based on the activity of other processes running on different machines.
  */
-class ETLService : public etlng::ETLServiceInterface, ETLServiceTag {
-    // TODO: make these template parameters in ETLService
-    using DataPipeType = etl::impl::ExtractionDataPipe<org::xrpl::rpc::v1::GetLedgerResponse>;
-    using CacheLoaderType = etl::CacheLoader<>;
-    using LedgerFetcherType = etl::impl::LedgerFetcher;
-    using ExtractorType = etl::impl::Extractor<DataPipeType, LedgerFetcherType>;
-    using LedgerLoaderType = etl::impl::LedgerLoader<LedgerFetcherType>;
-    using LedgerPublisherType = etl::impl::LedgerPublisher;
-    using AmendmentBlockHandlerType = etl::impl::AmendmentBlockHandler;
-    using TransformerType =
-        etl::impl::Transformer<DataPipeType, LedgerLoaderType, LedgerPublisherType, AmendmentBlockHandlerType>;
-
+class ETLService : public ETLServiceInterface {
     util::Logger log_{"ETL"};
 
+    util::async::AnyExecutionContext ctx_;
+    std::reference_wrapper<util::config::ClioConfigDefinition const> config_;
     std::shared_ptr<BackendInterface> backend_;
-    std::shared_ptr<etlng::LoadBalancerInterface> loadBalancer_;
-    std::shared_ptr<NetworkValidatedLedgersInterface> networkValidatedLedgers_;
+    std::shared_ptr<LoadBalancerInterface> balancer_;
+    std::shared_ptr<NetworkValidatedLedgersInterface> ledgers_;
+    std::shared_ptr<LedgerPublisherInterface> publisher_;
+    std::shared_ptr<CacheLoaderInterface> cacheLoader_;
+    std::shared_ptr<CacheUpdaterInterface> cacheUpdater_;
+    std::shared_ptr<ExtractorInterface> extractor_;
+    std::shared_ptr<LoaderInterface> loader_;
+    std::shared_ptr<InitialLoadObserverInterface> initialLoadObserver_;
+    std::shared_ptr<TaskManagerProviderInterface> taskManagerProvider_;
+    std::shared_ptr<MonitorProviderInterface> monitorProvider_;
+    std::shared_ptr<SystemState> state_;
 
-    std::uint32_t extractorThreads_ = 1;
-    std::thread worker_;
-
-    CacheLoaderType cacheLoader_;
-    LedgerFetcherType ledgerFetcher_;
-    LedgerLoaderType ledgerLoader_;
-    LedgerPublisherType ledgerPublisher_;
-    AmendmentBlockHandlerType amendmentBlockHandler_;
-
-    SystemState state_;
-
-    size_t numMarkers_ = 2;
     std::optional<uint32_t> startSequence_;
     std::optional<uint32_t> finishSequence_;
 
+    std::unique_ptr<MonitorInterface> monitor_;
+    std::unique_ptr<TaskManagerInterface> taskMan_;
+
+    boost::signals2::scoped_connection monitorNewSeqSubscription_;
+    boost::signals2::scoped_connection monitorDbStalledSubscription_;
+
+    std::optional<util::async::AnyOperation<void>> mainLoop_;
+
 public:
-    /**
-     * @brief Create an instance of ETLService.
-     *
-     * @param config The configuration to use
-     * @param ioc io context to run on
-     * @param backend BackendInterface implementation
-     * @param subscriptions Subscription manager
-     * @param balancer Load balancer to use
-     * @param ledgers The network validated ledgers datastructure
-     */
-    ETLService(
-        util::config::ClioConfigDefinition const& config,
-        boost::asio::io_context& ioc,
-        std::shared_ptr<BackendInterface> backend,
-        std::shared_ptr<feed::SubscriptionManagerInterface> subscriptions,
-        std::shared_ptr<etlng::LoadBalancerInterface> balancer,
-        std::shared_ptr<NetworkValidatedLedgersInterface> ledgers
-    );
-
-    /**
-     * @brief Move constructor is deleted because ETL service shares its fields by reference
-     */
-    ETLService(ETLService&&) = delete;
-
     /**
      * @brief A factory function to spawn new ETLService instances.
      *
      * Creates and runs the ETL service.
      *
      * @param config The configuration to use
-     * @param ioc io context to run on
      * @param ctx Execution context for asynchronous operations
      * @param backend BackendInterface implementation
      * @param subscriptions Subscription manager
@@ -156,182 +134,89 @@ public:
      * @param ledgers The network validated ledgers datastructure
      * @return A shared pointer to a new instance of ETLService
      */
-    static std::shared_ptr<etlng::ETLServiceInterface>
+    static std::shared_ptr<ETLServiceInterface>
     makeETLService(
         util::config::ClioConfigDefinition const& config,
-        boost::asio::io_context& ioc,
         util::async::AnyExecutionContext ctx,
         std::shared_ptr<BackendInterface> backend,
         std::shared_ptr<feed::SubscriptionManagerInterface> subscriptions,
-        std::shared_ptr<etlng::LoadBalancerInterface> balancer,
+        std::shared_ptr<LoadBalancerInterface> balancer,
         std::shared_ptr<NetworkValidatedLedgersInterface> ledgers
     );
 
     /**
-     * @brief Stops components and joins worker thread.
-     */
-    ~ETLService() override
-    {
-        if (not state_.isStopping)
-            stop();
-    }
-
-    /**
-     * @brief Stop the ETL service.
-     * @note This method blocks until the ETL service has stopped.
-     */
-    void
-    stop() override
-    {
-        LOG(log_.info()) << "Stop called";
-
-        state_.isStopping = true;
-        cacheLoader_.stop();
-
-        if (worker_.joinable())
-            worker_.join();
-
-        LOG(log_.debug()) << "Joined ETLService worker thread";
-    }
-
-    /**
-     * @brief Get time passed since last ledger close, in seconds.
+     * @brief Create an instance of ETLService.
      *
-     * @return Time passed since last ledger close
+     * @param ctx The execution context for asynchronous operations
+     * @param config The Clio configuration definition
+     * @param backend Interface to the backend database
+     * @param balancer Load balancer for distributing work
+     * @param ledgers Interface for accessing network validated ledgers
+     * @param publisher Interface for publishing ledger data
+     * @param cacheLoader Interface for loading cache data
+     * @param cacheUpdater Interface for updating cache data
+     * @param extractor The extractor to use
+     * @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
      */
-    std::uint32_t
-    lastCloseAgeSeconds() const override
-    {
-        return ledgerPublisher_.lastCloseAgeSeconds();
-    }
+    ETLService(
+        util::async::AnyExecutionContext ctx,
+        std::reference_wrapper<util::config::ClioConfigDefinition const> config,
+        std::shared_ptr<data::BackendInterface> backend,
+        std::shared_ptr<LoadBalancerInterface> balancer,
+        std::shared_ptr<NetworkValidatedLedgersInterface> ledgers,
+        std::shared_ptr<LedgerPublisherInterface> publisher,
+        std::shared_ptr<CacheLoaderInterface> cacheLoader,
+        std::shared_ptr<CacheUpdaterInterface> cacheUpdater,
+        std::shared_ptr<ExtractorInterface> extractor,
+        std::shared_ptr<LoaderInterface> loader,
+        std::shared_ptr<InitialLoadObserverInterface> initialLoadObserver,
+        std::shared_ptr<TaskManagerProviderInterface> taskManagerProvider,
+        std::shared_ptr<MonitorProviderInterface> monitorProvider,
+        std::shared_ptr<SystemState> state
+    );
 
-    /**
-     * @brief Check for the amendment blocked state.
-     *
-     * @return true if currently amendment blocked; false otherwise
-     */
-    bool
-    isAmendmentBlocked() const override
-    {
-        return state_.isAmendmentBlocked;
-    }
+    ~ETLService() override;
 
-    /**
-     * @brief Check whether Clio detected DB corruptions.
-     *
-     * @return true if corruption of DB was detected and cache was stopped.
-     */
-    bool
-    isCorruptionDetected() const override
-    {
-        return state_.isCorruptionDetected;
-    }
-
-    /**
-     * @brief Get state of ETL as a JSON object
-     *
-     * @return The state of ETL as a JSON object
-     */
-    boost::json::object
-    getInfo() const override
-    {
-        boost::json::object result;
-
-        result["etl_sources"] = loadBalancer_->toJson();
-        result["is_writer"] = static_cast<int>(state_.isWriting);
-        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());
-        return result;
-    }
-
-    /**
-     * @brief Get the etl nodes' state
-     * @return The etl nodes' state, nullopt if etl nodes are not connected
-     */
-    std::optional<etl::ETLState>
-    getETLState() const noexcept override
-    {
-        return loadBalancer_->getETLState();
-    }
-
-    /**
-     * @brief Start all components to run ETL service.
-     */
     void
     run() override;
 
-private:
-    /**
-     * @brief Run the ETL pipeline.
-     *
-     * Extracts ledgers and writes them to the database, until a write conflict occurs (or the server shuts down).
-     * @note database must already be populated when this function is called
-     *
-     * @param startSequence the first ledger to extract
-     * @param numExtractors number of extractors to use
-     * @return The last ledger written to the database, if any
-     */
-    std::optional<uint32_t>
-    runETLPipeline(uint32_t startSequence, uint32_t numExtractors);
-
-    /**
-     * @brief Monitor the network for newly validated ledgers.
-     *
-     * Also monitor the database to see if any process is writing those ledgers.
-     * This function is called when the application starts, and will only return when the application is shutting down.
-     * If the software detects the database is empty, this function will call loadInitialLedger(). If the software
-     * detects ledgers are not being written, this function calls runETLPipeline(). Otherwise, this function publishes
-     * ledgers as they are written to the database.
-     */
     void
-    monitor();
+    stop() override;
 
-    /**
-     * @brief Monitor the network for newly validated ledgers and publish them to the ledgers stream
-     *
-     * @param nextSequence the ledger sequence to publish
-     * @return The next ledger sequence to publish
-     */
-    uint32_t
-    publishNextSequence(uint32_t nextSequence);
+    boost::json::object
+    getInfo() const override;
 
-    /**
-     * @brief Monitor the database for newly written ledgers.
-     *
-     * Similar to the monitor(), except this function will never call runETLPipeline() or loadInitialLedger().
-     * This function only publishes ledgers as they are written to the database.
-     */
-    void
-    monitorReadOnly();
-
-    /**
-     * @return true if stopping; false otherwise
-     */
     bool
-    isStopping() const
-    {
-        return state_.isStopping;
-    }
+    isAmendmentBlocked() const override;
+
+    bool
+    isCorruptionDetected() const override;
+
+    std::optional<ETLState>
+    getETLState() const override;
 
-    /**
-     * @brief Get the number of markers to use during the initial ledger download.
-     *
-     * This is equivalent to the degree of parallelism during the initial ledger download.
-     *
-     * @return The number of markers
-     */
     std::uint32_t
-    getNumMarkers() const
-    {
-        return numMarkers_;
-    }
+    lastCloseAgeSeconds() const override;
+
+private:
+    std::optional<data::LedgerRange>
+    loadInitialLedgerIfNeeded();
 
-    /**
-     * @brief Spawn the worker thread and start monitoring.
-     */
     void
-    doWork();
+    startMonitor(uint32_t seq);
+
+    void
+    startLoading(uint32_t seq);
+
+    void
+    attemptTakeoverWriter();
+
+    void
+    giveUpWriter();
 };
+
 }  // namespace etl

src/etl/ETLServiceInterface.hpp

@@ -26,7 +26,7 @@
 #include <cstdint>
 #include <optional>
 
-namespace etlng {
+namespace etl {
 
 /**
  * @brief This is a base class for any ETL service implementations.
@@ -77,7 +77,7 @@ struct ETLServiceInterface {
      * @brief Get the etl nodes' state
      * @return The etl nodes' state, nullopt if etl nodes are not connected
      */
-    [[nodiscard]] virtual std::optional<etl::ETLState>
+    [[nodiscard]] virtual std::optional<ETLState>
     getETLState() const = 0;
 
     /**
@@ -89,4 +89,4 @@ struct ETLServiceInterface {
     lastCloseAgeSeconds() const = 0;
 };
 
-}  // namespace etlng
+}  // namespace etl

src/etl/ExtractorInterface.hpp

@@ -19,12 +19,12 @@
 
 #pragma once
 
-#include "etlng/Models.hpp"
+#include "etl/Models.hpp"
 
 #include <cstdint>
 #include <optional>
 
-namespace etlng {
+namespace etl {
 
 /**
  * @brief An interface for the Extractor
@@ -51,4 +51,4 @@ struct ExtractorInterface {
     extractLedgerOnly(uint32_t seq) = 0;
 };
 
-}  // namespace etlng
+}  // namespace etl

src/etl/InitialLoadObserverInterface.hpp

@@ -19,7 +19,7 @@
 
 #pragma once
 
-#include "etlng/Models.hpp"
+#include "etl/Models.hpp"
 
 #include <xrpl/protocol/LedgerHeader.h>
 
@@ -28,7 +28,7 @@
 #include <string>
 #include <vector>
 
-namespace etlng {
+namespace etl {
 
 /**
  * @brief The interface for observing the initial ledger load
@@ -51,4 +51,4 @@ struct InitialLoadObserverInterface {
     ) = 0;
 };
 
-}  // namespace etlng
+}  // namespace etl

src/etl/LedgerPublisherInterface.hpp

@@ -23,7 +23,7 @@
 #include <cstdint>
 #include <optional>
 
-namespace etlng {
+namespace etl {
 
 /**
  * @brief The interface of a scheduler for the extraction process
@@ -71,4 +71,4 @@ struct LedgerPublisherInterface {
     lastPublishAgeSeconds() const = 0;
 };
 
-}  // namespace etlng
+}  // namespace etl

src/etl/LoadBalancer.cpp

@@ -21,9 +21,10 @@
 
 #include "data/BackendInterface.hpp"
 #include "etl/ETLState.hpp"
+#include "etl/InitialLoadObserverInterface.hpp"
+#include "etl/LoadBalancerInterface.hpp"
 #include "etl/NetworkValidatedLedgersInterface.hpp"
 #include "etl/Source.hpp"
-#include "etlng/LoadBalancerInterface.hpp"
 #include "feed/SubscriptionManagerInterface.hpp"
 #include "rpc/Errors.hpp"
 #include "util/Assert.hpp"
@@ -64,7 +65,7 @@ using util::prometheus::Labels;
 
 namespace etl {
 
-std::shared_ptr<etlng::LoadBalancerInterface>
+std::shared_ptr<LoadBalancerInterface>
 LoadBalancer::makeLoadBalancer(
     ClioConfigDefinition const& config,
     boost::asio::io_context& ioc,
@@ -158,7 +159,6 @@ LoadBalancer::LoadBalancer(
         auto source = sourceFactory(
             *it,
             ioc,
-            backend,
             subscriptions,
             validatedLedgers,
             forwardingTimeout,
@@ -212,26 +212,32 @@ LoadBalancer::LoadBalancer(
     }
 }
 
-std::vector<std::string>
-LoadBalancer::loadInitialLedger(uint32_t sequence, std::chrono::steady_clock::duration retryAfter)
+InitialLedgerLoadResult
+LoadBalancer::loadInitialLedger(
+    uint32_t sequence,
+    InitialLoadObserverInterface& loadObserver,
+    std::chrono::steady_clock::duration retryAfter
+)
 {
-    std::vector<std::string> response;
-    execute(
-        [this, &response, &sequence](auto& source) {
-            auto [data, res] = source->loadInitialLedger(sequence, downloadRanges_);
+    InitialLedgerLoadResult response;
 
-            if (!res) {
+    execute(
+        [this, &response, &sequence, &loadObserver](auto& source) {
+            auto res = source->loadInitialLedger(sequence, downloadRanges_, loadObserver);
+
+            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;
 }
 

src/etl/LoadBalancer.hpp

@@ -21,13 +21,12 @@
 
 #include "data/BackendInterface.hpp"
 #include "etl/ETLState.hpp"
+#include "etl/InitialLoadObserverInterface.hpp"
+#include "etl/LoadBalancerInterface.hpp"
 #include "etl/NetworkValidatedLedgersInterface.hpp"
 #include "etl/Source.hpp"
-#include "etlng/InitialLoadObserverInterface.hpp"
-#include "etlng/LoadBalancerInterface.hpp"
 #include "feed/SubscriptionManagerInterface.hpp"
 #include "rpc/Errors.hpp"
-#include "util/Assert.hpp"
 #include "util/Mutex.hpp"
 #include "util/Random.hpp"
 #include "util/ResponseExpirationCache.hpp"
@@ -54,7 +53,6 @@
 #include <optional>
 #include <string>
 #include <string_view>
-#include <utility>
 #include <vector>
 
 namespace etl {
@@ -76,7 +74,7 @@ concept SomeLoadBalancer = std::derived_from<T, LoadBalancerTag>;
  * which ledgers have been validated by the network, and the range of ledgers each etl source has). This class also
  * allows requests for ledger data to be load balanced across all possible ETL sources.
  */
-class LoadBalancer : public etlng::LoadBalancerInterface, LoadBalancerTag {
+class LoadBalancer : public LoadBalancerInterface, LoadBalancerTag {
 public:
     using RawLedgerObjectType = org::xrpl::rpc::v1::RawLedgerObject;
     using GetLedgerResponseType = org::xrpl::rpc::v1::GetLedgerResponse;
@@ -164,20 +162,6 @@ public:
         SourceFactory sourceFactory = makeSource
     );
 
-    /**
-     * @brief Load the initial ledger, writing data to the queue.
-     * @note This function will retry indefinitely until the ledger is downloaded.
-     *
-     * @param sequence Sequence of ledger to download
-     * @param retryAfter Time to wait between retries (2 seconds by default)
-     * @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;
-
     /**
      * @brief Load the initial ledger, writing data to the queue.
      * @note This function will retry indefinitely until the ledger is downloaded or the download is cancelled.
@@ -187,16 +171,12 @@ public:
      * @param retryAfter Time to wait between retries (2 seconds by default)
      * @return A std::expected with ledger edge keys on success, or InitialLedgerLoadError on failure
      */
-    etlng::InitialLedgerLoadResult
+    InitialLedgerLoadResult
     loadInitialLedger(
-        [[maybe_unused]] uint32_t sequence,
-        [[maybe_unused]] etlng::InitialLoadObserverInterface& observer,
-        [[maybe_unused]] std::chrono::steady_clock::duration retryAfter
-    ) override
-    {
-        ASSERT(false, "Not available for old ETL");
-        std::unreachable();
-    }
+        uint32_t sequence,
+        InitialLoadObserverInterface& observer,
+        std::chrono::steady_clock::duration retryAfter
+    ) override;
 
     /**
      * @brief Fetch data for a specific ledger.

src/etl/LoadBalancerInterface.hpp

@@ -21,7 +21,7 @@
 #pragma once
 
 #include "etl/ETLState.hpp"
-#include "etlng/InitialLoadObserverInterface.hpp"
+#include "etl/InitialLoadObserverInterface.hpp"
 #include "rpc/Errors.hpp"
 
 #include <boost/asio/spawn.hpp>
@@ -37,7 +37,7 @@
 #include <string>
 #include <vector>
 
-namespace etlng {
+namespace etl {
 
 /**
  * @brief Represents possible errors for initial ledger load
@@ -76,21 +76,10 @@ public:
     [[nodiscard]] virtual InitialLedgerLoadResult
     loadInitialLedger(
         uint32_t sequence,
-        etlng::InitialLoadObserverInterface& loader,
+        InitialLoadObserverInterface& loader,
         std::chrono::steady_clock::duration retryAfter = std::chrono::seconds{2}
     ) = 0;
 
-    /**
-     * @brief Load the initial ledger, writing data to the queue.
-     * @note This function will retry indefinitely until the ledger is downloaded.
-     *
-     * @param sequence Sequence of ledger to download
-     * @param retryAfter Time to wait between retries (2 seconds by default)
-     * @return A std::vector<std::string> The ledger data
-     */
-    [[nodiscard]] virtual std::vector<std::string>
-    loadInitialLedger(uint32_t sequence, std::chrono::steady_clock::duration retryAfter = std::chrono::seconds{2}) = 0;
-
     /**
      * @brief Fetch data for a specific ledger.
      *
@@ -141,7 +130,7 @@ public:
      * @brief Return state of ETL nodes.
      * @return ETL state, nullopt if etl nodes not available
      */
-    [[nodiscard]] virtual std::optional<etl::ETLState>
+    [[nodiscard]] virtual std::optional<ETLState>
     getETLState() noexcept = 0;
 
     /**
@@ -154,4 +143,4 @@ public:
     stop(boost::asio::yield_context yield) = 0;
 };
 
-}  // namespace etlng
+}  // namespace etl

src/etl/LoaderInterface.hpp

@@ -19,14 +19,14 @@
 
 #pragma once
 
-#include "etlng/Models.hpp"
+#include "etl/Models.hpp"
 
 #include <xrpl/protocol/LedgerHeader.h>
 
 #include <expected>
 #include <optional>
 
-namespace etlng {
+namespace etl {
 
 /**
  * @brief Enumeration of possible errors that can occur during loading operations
@@ -59,4 +59,4 @@ struct LoaderInterface {
     loadInitialLedger(model::LedgerData const& data) = 0;
 };
 
-}  // namespace etlng
+}  // namespace etl

src/etl/Models.hpp

@@ -38,7 +38,7 @@
 #include <string>
 #include <vector>
 
-namespace etlng::model {
+namespace etl::model {
 
 /**
  * @brief A specification for the Registry.
@@ -190,4 +190,4 @@ struct Task {
     uint32_t seq;
 };
 
-}  // namespace etlng::model
+}  // namespace etl::model

src/etl/MonitorInterface.hpp

@@ -26,7 +26,7 @@
 #include <chrono>
 #include <cstdint>
 
-namespace etlng {
+namespace etl {
 
 /**
  * @brief An interface for the monitor service
@@ -88,4 +88,4 @@ public:
     stop() = 0;
 };
 
-}  // namespace etlng
+}  // namespace etl

src/etl/MonitorProviderInterface.hpp

@@ -20,15 +20,15 @@
 #pragma once
 
 #include "data/BackendInterface.hpp"
+#include "etl/MonitorInterface.hpp"
 #include "etl/NetworkValidatedLedgersInterface.hpp"
-#include "etlng/MonitorInterface.hpp"
 #include "util/async/AnyExecutionContext.hpp"
 
 #include <chrono>
 #include <cstdint>
 #include <memory>
 
-namespace etlng {
+namespace etl {
 
 /**
  * @brief An interface for providing Monitor instances
@@ -55,10 +55,10 @@ struct MonitorProviderInterface {
     make(
         util::async::AnyExecutionContext ctx,
         std::shared_ptr<BackendInterface> backend,
-        std::shared_ptr<etl::NetworkValidatedLedgersInterface> validatedLedgers,
+        std::shared_ptr<NetworkValidatedLedgersInterface> validatedLedgers,
         uint32_t startSequence,
         std::chrono::steady_clock::duration dbStalledReportDelay = kDEFAULT_DB_STALLED_REPORT_DELAY
     ) = 0;
 };
 
-}  // namespace etlng
+}  // namespace etl

src/etl/README.md

@@ -6,7 +6,7 @@ A single Clio node has one or more ETL sources specified in the config file. Cli
 
 Upon receiving a message on the stream, Clio fetches the data associated with the newly validated ledger from one of the ETL sources. The fetch is performed via a gRPC request called `GetLedger`. This request returns the ledger header, transactions and metadata blobs, and every ledger object added/modified/deleted as part of this ledger. The ETL subsystem then writes all of this data to the databases, and moves on to the next ledger.
 
-If the database is not empty, clio will first come up in a "soft" read-only mode. In read-only mode, the server does not perform ETL and simply publishes new ledgers as they are written to the database. If the database is not updated within a certain time period (currently hard coded at 20 seconds), clio will begin the ETL process and start writing to the database. The database will report an error when trying to write a record with a key that already exists. ETL uses this error to determine that another process is writing to the database, and subsequently falls back to a soft read-only mode. clio can also operate in strict read-only mode, in which case they will never write to the database.
+If the database is not empty, clio will first come up in a "soft" read-only mode. In read-only mode, the server does not perform ETL and simply publishes new ledgers as they are written to the database. If the database is not updated within a certain time period (currently hard coded at 10 seconds), clio will begin the ETL process and start writing to the database. The database will report an error when trying to write a record with a key that already exists. ETL uses this error to determine that another process is writing to the database, and subsequently falls back to a soft read-only mode. clio can also operate in strict read-only mode, in which case they will never write to the database.
 
 ## Ledger cache
 

src/etl/RegistryInterface.hpp

@@ -19,13 +19,13 @@
 
 #pragma once
 
-#include "etlng/Models.hpp"
+#include "etl/Models.hpp"
 
 #include <cstdint>
 #include <string>
 #include <vector>
 
-namespace etlng {
+namespace etl {
 
 /**
  * @brief The interface for a registry that can dispatch transactions and objects to extensions.
@@ -36,15 +36,15 @@ namespace etlng {
  * The registry itself consists of Extensions.
  * Each extension must define at least one valid hook:
  * - for ongoing ETL dispatch:
- *   - void onLedgerData(etlng::model::LedgerData const&)
- *   - void onTransaction(uint32_t, etlng::model::Transaction const&)
- *   - void onObject(uint32_t, etlng::model::Object const&)
+ *   - void onLedgerData(etl::model::LedgerData const&)
+ *   - void onTransaction(uint32_t, etl::model::Transaction const&)
+ *   - void onObject(uint32_t, etl::model::Object const&)
  * - for initial ledger load
- *   - void onInitialData(etlng::model::LedgerData const&)
- *   - void onInitialTransaction(uint32_t, etlng::model::Transaction const&)
+ *   - void onInitialData(etl::model::LedgerData const&)
+ *   - void onInitialTransaction(uint32_t, etl::model::Transaction const&)
  * - for initial objects (called for each downloaded batch)
- *   - void onInitialObjects(uint32_t, std::vector<etlng::model::Object> const&, std::string)
- *   - void onInitialObject(uint32_t, etlng::model::Object const&)
+ *   - void onInitialObjects(uint32_t, std::vector<etl::model::Object> const&, std::string)
+ *   - void onInitialObject(uint32_t, etl::model::Object const&)
  *
  * When the registry dispatches (initial)data or objects, each of the above hooks will be called in order on each
  * registered extension.
@@ -57,7 +57,7 @@ namespace etlng {
  * The specification is setup like so:
  * @code{.cpp}
  * struct Ext {
- *   using spec = etlng::model::Spec<
+ *   using spec = etl::model::Spec<
  *     ripple::TxType::ttNFTOKEN_BURN,
  *     ripple::TxType::ttNFTOKEN_ACCEPT_OFFER,
  *     ripple::TxType::ttNFTOKEN_CREATE_OFFER,
@@ -65,7 +65,7 @@ namespace etlng {
  *     ripple::TxType::ttNFTOKEN_MINT>;
  *
  *   static void
- *   onInitialTransaction(uint32_t, etlng::model::Transaction const&);
+ *   onInitialTransaction(uint32_t, etl::model::Transaction const&);
  * };
  * @endcode
  */
@@ -105,4 +105,4 @@ struct RegistryInterface {
     dispatch(model::LedgerData const& data) = 0;
 };
 
-}  // namespace etlng
+}  // namespace etl

src/etl/SchedulerInterface.hpp

@@ -19,11 +19,11 @@
 
 #pragma once
 
-#include "etlng/Models.hpp"
+#include "etl/Models.hpp"
 
 #include <optional>
 
-namespace etlng {
+namespace etl {
 
 /**
  * @brief The interface of a scheduler for the extraction process
@@ -39,4 +39,4 @@ struct SchedulerInterface {
     next() = 0;
 };
 
-}  // namespace etlng
+}  // namespace etl

src/etl/Source.cpp

@@ -19,7 +19,6 @@
 
 #include "etl/Source.hpp"
 
-#include "data/BackendInterface.hpp"
 #include "etl/NetworkValidatedLedgersInterface.hpp"
 #include "etl/impl/ForwardingSource.hpp"
 #include "etl/impl/GrpcSource.hpp"
@@ -41,7 +40,6 @@ SourcePtr
 makeSource(
     util::config::ObjectView const& config,
     boost::asio::io_context& ioc,
-    std::shared_ptr<BackendInterface> backend,
     std::shared_ptr<feed::SubscriptionManagerInterface> subscriptions,
     std::shared_ptr<NetworkValidatedLedgersInterface> validatedLedgers,
     std::chrono::steady_clock::duration forwardingTimeout,
@@ -55,7 +53,7 @@ makeSource(
     auto const grpcPort = config.get<std::string>("grpc_port");
 
     impl::ForwardingSource forwardingSource{ip, wsPort, forwardingTimeout};
-    impl::GrpcSource grpcSource{ip, grpcPort, std::move(backend)};
+    impl::GrpcSource grpcSource{ip, grpcPort};
     auto subscriptionSource = std::make_unique<impl::SubscriptionSource>(
         ioc,
         ip,

src/etl/Source.hpp

@@ -19,7 +19,8 @@
 
 #pragma once
 
-#include "data/BackendInterface.hpp"
+#include "etl/InitialLoadObserverInterface.hpp"
+#include "etl/LoadBalancerInterface.hpp"
 #include "etl/NetworkValidatedLedgersInterface.hpp"
 #include "feed/SubscriptionManagerInterface.hpp"
 #include "rpc/Errors.hpp"
@@ -47,7 +48,6 @@ namespace etl {
 
 /**
  * @brief Provides an implementation of a ETL source
- *
  */
 class SourceBase {
 public:
@@ -77,7 +77,7 @@ public:
      *
      * @return true if source is connected; false otherwise
      */
-    virtual bool
+    [[nodiscard]] virtual bool
     isConnected() const = 0;
 
     /**
@@ -93,11 +93,11 @@ public:
      *
      * @return JSON representation of the source
      */
-    virtual boost::json::object
+    [[nodiscard]] virtual boost::json::object
     toJson() const = 0;
 
     /** @return String representation of the source (for debug) */
-    virtual std::string
+    [[nodiscard]] virtual std::string
     toString() const = 0;
 
     /**
@@ -106,7 +106,7 @@ public:
      * @param sequence The ledger sequence to check
      * @return true if ledger is in the range of this source; false otherwise
      */
-    virtual bool
+    [[nodiscard]] virtual bool
     hasLedger(uint32_t sequence) const = 0;
 
     /**
@@ -120,7 +120,7 @@ public:
      * @param getObjectNeighbors Whether to request object neighbors; defaults to false
      * @return A std::pair of the response status and the response itself
      */
-    virtual std::pair<grpc::Status, org::xrpl::rpc::v1::GetLedgerResponse>
+    [[nodiscard]] virtual std::pair<grpc::Status, org::xrpl::rpc::v1::GetLedgerResponse>
     fetchLedger(uint32_t sequence, bool getObjects = true, bool getObjectNeighbors = false) = 0;
 
     /**
@@ -128,10 +128,11 @@ public:
      *
      * @param sequence Sequence of the ledger to download
      * @param numMarkers Number of markers to generate for async calls
+     * @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>
-    loadInitialLedger(uint32_t sequence, std::uint32_t numMarkers) = 0;
+    virtual InitialLedgerLoadResult
+    loadInitialLedger(uint32_t sequence, std::uint32_t numMarkers, InitialLoadObserverInterface& loader) = 0;
 
     /**
      * @brief Forward a request to rippled.
@@ -142,7 +143,7 @@ public:
      * @param yield The coroutine context
      * @return Response on success or error on failure
      */
-    virtual std::expected<boost::json::object, rpc::ClioError>
+    [[nodiscard]] virtual std::expected<boost::json::object, rpc::ClioError>
     forwardToRippled(
         boost::json::object const& request,
         std::optional<std::string> const& forwardToRippledClientIp,
@@ -156,7 +157,6 @@ using SourcePtr = std::unique_ptr<SourceBase>;
 using SourceFactory = std::function<SourcePtr(
     util::config::ObjectView const& config,
     boost::asio::io_context& ioc,
-    std::shared_ptr<BackendInterface> backend,
     std::shared_ptr<feed::SubscriptionManagerInterface> subscriptions,
     std::shared_ptr<NetworkValidatedLedgersInterface> validatedLedgers,
     std::chrono::steady_clock::duration forwardingTimeout,
@@ -170,7 +170,6 @@ using SourceFactory = std::function<SourcePtr(
  *
  * @param config The configuration to use
  * @param ioc The io_context to run on
- * @param backend BackendInterface implementation
  * @param subscriptions Subscription manager
  * @param validatedLedgers The network validated ledgers data structure
  * @param forwardingTimeout The timeout for forwarding to rippled
@@ -180,11 +179,10 @@ using SourceFactory = std::function<SourcePtr(
  * as forwarding.
  * @return The created source
  */
-SourcePtr
+[[nodiscard]] SourcePtr
 makeSource(
     util::config::ObjectView const& config,
     boost::asio::io_context& ioc,
-    std::shared_ptr<BackendInterface> backend,
     std::shared_ptr<feed::SubscriptionManagerInterface> subscriptions,
     std::shared_ptr<NetworkValidatedLedgersInterface> validatedLedgers,
     std::chrono::steady_clock::duration forwardingTimeout,

src/etl/TaskManagerInterface.hpp

@@ -21,7 +21,7 @@
 
 #include <cstddef>
 
-namespace etlng {
+namespace etl {
 
 /**
  * @brief An interface for the Task Manager
@@ -43,4 +43,4 @@ struct TaskManagerInterface {
     stop() = 0;
 };
 
-}  // namespace etlng
+}  // namespace etl

src/etl/TaskManagerProviderInterface.hpp

@@ -19,8 +19,8 @@
 
 #pragma once
 
-#include "etlng/MonitorInterface.hpp"
-#include "etlng/TaskManagerInterface.hpp"
+#include "etl/MonitorInterface.hpp"
+#include "etl/TaskManagerInterface.hpp"
 #include "util/async/AnyExecutionContext.hpp"
 
 #include <cstddef>
@@ -29,7 +29,7 @@
 #include <memory>
 #include <optional>
 
-namespace etlng {
+namespace etl {
 
 /**
  * @brief An interface for providing the Task Manager
@@ -55,4 +55,4 @@ struct TaskManagerProviderInterface {
     ) = 0;
 };
 
-}  // namespace etlng
+}  // namespace etl

src/etl/impl/AmendmentBlockHandler.cpp

@@ -20,12 +20,12 @@
 #include "etl/impl/AmendmentBlockHandler.hpp"
 
 #include "etl/SystemState.hpp"
+#include "util/async/AnyExecutionContext.hpp"
 #include "util/log/Logger.hpp"
 
-#include <boost/asio/io_context.hpp>
-
 #include <chrono>
 #include <functional>
+#include <optional>
 #include <utility>
 
 namespace etl::impl {
@@ -37,20 +37,35 @@ AmendmentBlockHandler::ActionType const AmendmentBlockHandler::kDEFAULT_AMENDMEN
 };
 
 AmendmentBlockHandler::AmendmentBlockHandler(
-    boost::asio::io_context& ioc,
+    util::async::AnyExecutionContext ctx,
     SystemState& state,
     std::chrono::steady_clock::duration interval,
     ActionType action
 )
-    : state_{std::ref(state)}, repeat_{ioc}, interval_{interval}, action_{std::move(action)}
+    : state_{std::ref(state)}, interval_{interval}, ctx_{std::move(ctx)}, action_{std::move(action)}
 {
 }
 
+AmendmentBlockHandler::~AmendmentBlockHandler()
+{
+    stop();
+}
+
 void
 AmendmentBlockHandler::notifyAmendmentBlocked()
 {
     state_.get().isAmendmentBlocked = true;
-    repeat_.start(interval_, action_);
+    if (not operation_.has_value())
+        operation_.emplace(ctx_.executeRepeatedly(interval_, action_));
+}
+
+void
+AmendmentBlockHandler::stop()
+{
+    if (operation_.has_value()) {
+        operation_->abort();
+        operation_.reset();
+    }
 }
 
 }  // namespace etl::impl

src/etl/impl/AmendmentBlockHandler.hpp

@@ -1,7 +1,7 @@
 //------------------------------------------------------------------------------
 /*
     This file is part of clio: https://github.com/XRPLF/clio
-    Copyright (c) 2023, the clio developers.
+    Copyright (c) 2024, 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
@@ -19,8 +19,10 @@
 
 #pragma once
 
+#include "etl/AmendmentBlockHandlerInterface.hpp"
 #include "etl/SystemState.hpp"
-#include "util/Repeat.hpp"
+#include "util/async/AnyExecutionContext.hpp"
+#include "util/async/AnyOperation.hpp"
 
 #include <boost/asio/io_context.hpp>
 #include <boost/asio/post.hpp>
@@ -28,17 +30,19 @@
 
 #include <chrono>
 #include <functional>
+#include <optional>
 
 namespace etl::impl {
 
-class AmendmentBlockHandler {
+class AmendmentBlockHandler : public AmendmentBlockHandlerInterface {
 public:
     using ActionType = std::function<void()>;
 
 private:
     std::reference_wrapper<SystemState> state_;
-    util::Repeat repeat_;
     std::chrono::steady_clock::duration interval_;
+    util::async::AnyExecutionContext ctx_;
+    std::optional<util::async::AnyOperation<void>> operation_;
 
     ActionType action_;
 
@@ -46,14 +50,19 @@ public:
     static ActionType const kDEFAULT_AMENDMENT_BLOCK_ACTION;
 
     AmendmentBlockHandler(
-        boost::asio::io_context& ioc,
+        util::async::AnyExecutionContext ctx,
         SystemState& state,
         std::chrono::steady_clock::duration interval = std::chrono::seconds{1},
         ActionType action = kDEFAULT_AMENDMENT_BLOCK_ACTION
     );
 
+    ~AmendmentBlockHandler() override;
+
     void
-    notifyAmendmentBlocked();
+    stop() override;
+
+    void
+    notifyAmendmentBlocked() override;
 };
 
 }  // namespace etl::impl

src/etl/impl/AsyncData.hpp

@@ -1,221 +0,0 @@
-//------------------------------------------------------------------------------
-/*
-    This file is part of clio: https://github.com/XRPLF/clio
-    Copyright (c) 2023, 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 "data/Types.hpp"
-#include "etl/ETLHelpers.hpp"
-#include "etl/MPTHelpers.hpp"
-#include "etl/NFTHelpers.hpp"
-#include "util/Assert.hpp"
-#include "util/log/Logger.hpp"
-
-#include <grpcpp/client_context.h>
-#include <grpcpp/grpcpp.h>
-#include <grpcpp/support/status.h>
-#include <org/xrpl/rpc/v1/get_ledger_data.pb.h>
-#include <xrpl/basics/base_uint.h>
-#include <xrpl/basics/strHex.h>
-#include <xrpl/proto/org/xrpl/rpc/v1/xrp_ledger.grpc.pb.h>
-
-#include <cstddef>
-#include <cstdint>
-#include <memory>
-#include <optional>
-#include <string>
-#include <utility>
-#include <vector>
-
-namespace etl::impl {
-
-class AsyncCallData {
-    util::Logger log_{"ETL"};
-
-    std::unique_ptr<org::xrpl::rpc::v1::GetLedgerDataResponse> cur_;
-    std::unique_ptr<org::xrpl::rpc::v1::GetLedgerDataResponse> next_;
-
-    org::xrpl::rpc::v1::GetLedgerDataRequest request_;
-    std::unique_ptr<grpc::ClientContext> context_;
-
-    grpc::Status status_;
-    unsigned char nextPrefix_;
-
-    std::string lastKey_;
-
-public:
-    AsyncCallData(uint32_t seq, ripple::uint256 const& marker, std::optional<ripple::uint256> const& nextMarker)
-    {
-        request_.mutable_ledger()->set_sequence(seq);
-        if (marker.isNonZero()) {
-            request_.set_marker(marker.data(), ripple::uint256::size());
-        }
-        request_.set_user("ETL");
-        nextPrefix_ = 0x00;
-        if (nextMarker)
-            nextPrefix_ = nextMarker->data()[0];
-
-        unsigned char const prefix = marker.data()[0];
-
-        LOG(log_.debug()) << "Setting up AsyncCallData. marker = " << ripple::strHex(marker)
-                          << " . prefix = " << ripple::strHex(std::string(1, prefix))
-                          << " . nextPrefix_ = " << ripple::strHex(std::string(1, nextPrefix_));
-
-        ASSERT(
-            nextPrefix_ > prefix || nextPrefix_ == 0x00,
-            "Next prefix must be greater than current prefix. Got: nextPrefix_ = {}, prefix = {}",
-            nextPrefix_,
-            prefix
-        );
-
-        cur_ = std::make_unique<org::xrpl::rpc::v1::GetLedgerDataResponse>();
-        next_ = std::make_unique<org::xrpl::rpc::v1::GetLedgerDataResponse>();
-        context_ = std::make_unique<grpc::ClientContext>();
-    }
-
-    enum class CallStatus { MORE, DONE, ERRORED };
-
-    CallStatus
-    process(
-        std::unique_ptr<org::xrpl::rpc::v1::XRPLedgerAPIService::Stub>& stub,
-        grpc::CompletionQueue& cq,
-        BackendInterface& backend,
-        bool abort,
-        bool cacheOnly = false
-    )
-    {
-        LOG(log_.trace()) << "Processing response. "
-                          << "Marker prefix = " << getMarkerPrefix();
-        if (abort) {
-            LOG(log_.error()) << "AsyncCallData aborted";
-            return CallStatus::ERRORED;
-        }
-        if (!status_.ok()) {
-            LOG(log_.error()) << "AsyncCallData status_ not ok: code = " << status_.error_code()
-                              << " message = " << status_.error_message();
-            return CallStatus::ERRORED;
-        }
-        if (!next_->is_unlimited()) {
-            LOG(log_.warn()) << "AsyncCallData is_unlimited is false. "
-                             << "Make sure secure_gateway is set correctly at the ETL source";
-        }
-
-        std::swap(cur_, next_);
-
-        bool more = true;
-
-        // if no marker returned, we are done
-        if (cur_->marker().empty())
-            more = false;
-
-        // if returned marker is greater than our end, we are done
-        unsigned char const prefix = cur_->marker()[0];
-        if (nextPrefix_ != 0x00 && prefix >= nextPrefix_)
-            more = false;
-
-        // if we are not done, make the next async call
-        if (more) {
-            request_.set_marker(cur_->marker());
-            call(stub, cq);
-        }
-
-        auto const numObjects = cur_->ledger_objects().objects_size();
-        LOG(log_.debug()) << "Writing " << numObjects << " objects";
-
-        std::vector<data::LedgerObject> cacheUpdates;
-        cacheUpdates.reserve(numObjects);
-
-        for (int i = 0; i < numObjects; ++i) {
-            auto& obj = *(cur_->mutable_ledger_objects()->mutable_objects(i));
-            if (!more && nextPrefix_ != 0x00) {
-                if (static_cast<unsigned char>(obj.key()[0]) >= nextPrefix_)
-                    continue;
-            }
-            cacheUpdates.push_back(
-                {*ripple::uint256::fromVoidChecked(obj.key()), {obj.data().begin(), obj.data().end()}}
-            );
-            if (!cacheOnly) {
-                if (!lastKey_.empty())
-                    backend.writeSuccessor(std::move(lastKey_), request_.ledger().sequence(), std::string{obj.key()});
-                lastKey_ = obj.key();
-                backend.writeNFTs(getNFTDataFromObj(request_.ledger().sequence(), obj.key(), obj.data()));
-
-                auto const maybeMPTHolder = getMPTHolderFromObj(obj.key(), obj.data());
-                if (maybeMPTHolder)
-                    backend.writeMPTHolders({*maybeMPTHolder});
-
-                backend.writeLedgerObject(
-                    std::move(*obj.mutable_key()), request_.ledger().sequence(), std::move(*obj.mutable_data())
-                );
-            }
-        }
-        backend.cache().update(cacheUpdates, request_.ledger().sequence(), cacheOnly);
-        LOG(log_.debug()) << "Wrote " << numObjects << " objects. Got more: " << (more ? "YES" : "NO");
-
-        return more ? CallStatus::MORE : CallStatus::DONE;
-    }
-
-    void
-    call(std::unique_ptr<org::xrpl::rpc::v1::XRPLedgerAPIService::Stub>& stub, grpc::CompletionQueue& cq)
-    {
-        context_ = std::make_unique<grpc::ClientContext>();
-
-        std::unique_ptr<grpc::ClientAsyncResponseReader<org::xrpl::rpc::v1::GetLedgerDataResponse>> rpc(
-            stub->PrepareAsyncGetLedgerData(context_.get(), request_, &cq)
-        );
-
-        rpc->StartCall();
-
-        rpc->Finish(next_.get(), &status_, this);
-    }
-
-    std::string
-    getMarkerPrefix()
-    {
-        if (next_->marker().empty()) {
-            return "";
-        }
-        return ripple::strHex(std::string{next_->marker().data()[0]});
-    }
-
-    std::string
-    getLastKey()
-    {
-        return lastKey_;
-    }
-};
-
-inline std::vector<AsyncCallData>
-makeAsyncCallData(uint32_t const sequence, uint32_t const numMarkers)
-{
-    auto const markers = getMarkers(numMarkers);
-
-    std::vector<AsyncCallData> result;
-    result.reserve(markers.size());
-
-    for (size_t i = 0; i + 1 < markers.size(); ++i) {
-        result.emplace_back(sequence, markers[i], markers[i + 1]);
-    }
-    if (not markers.empty()) {
-        result.emplace_back(sequence, markers.back(), std::nullopt);
-    }
-    return result;
-}
-
-}  // namespace etl::impl

src/etl/impl/AsyncGrpcCall.cpp

@@ -17,12 +17,12 @@
 */
 //==============================================================================
 
-#include "etlng/impl/AsyncGrpcCall.hpp"
+#include "etl/impl/AsyncGrpcCall.hpp"
 
 #include "etl/ETLHelpers.hpp"
-#include "etlng/InitialLoadObserverInterface.hpp"
-#include "etlng/Models.hpp"
-#include "etlng/impl/Extraction.hpp"
+#include "etl/InitialLoadObserverInterface.hpp"
+#include "etl/Models.hpp"
+#include "etl/impl/Extraction.hpp"
 #include "util/Assert.hpp"
 #include "util/log/Logger.hpp"
 
@@ -41,7 +41,7 @@
 #include <utility>
 #include <vector>
 
-namespace etlng::impl {
+namespace etl::impl {
 
 AsyncGrpcCall::AsyncGrpcCall(
     uint32_t seq,
@@ -78,7 +78,7 @@ AsyncGrpcCall::CallStatus
 AsyncGrpcCall::process(
     std::unique_ptr<AsyncGrpcCall::StubType>& stub,
     grpc::CompletionQueue& cq,
-    etlng::InitialLoadObserverInterface& loader,
+    InitialLoadObserverInterface& loader,
     bool abort
 )
 {
@@ -121,7 +121,7 @@ AsyncGrpcCall::process(
     }
 
     auto const numObjects = cur_->ledger_objects().objects_size();
-    std::vector<etlng::model::Object> data;
+    std::vector<model::Object> data;
     data.reserve(numObjects);
 
     for (int i = 0; i < numObjects; ++i) {
@@ -132,7 +132,7 @@ AsyncGrpcCall::process(
         }
 
         lastKey_ = obj.key();  // this will end up the last key we actually touched eventually
-        data.push_back(etlng::impl::extractObj(std::move(obj)));
+        data.push_back(impl::extractObj(std::move(obj)));
     }
 
     if (not data.empty())
@@ -171,7 +171,7 @@ AsyncGrpcCall::getLastKey()
 std::vector<AsyncGrpcCall>
 AsyncGrpcCall::makeAsyncCalls(uint32_t const sequence, uint32_t const numMarkers)
 {
-    auto const markers = etl::getMarkers(numMarkers);
+    auto const markers = getMarkers(numMarkers);
 
     std::vector<AsyncGrpcCall> result;
     result.reserve(markers.size());
@@ -185,4 +185,4 @@ AsyncGrpcCall::makeAsyncCalls(uint32_t const sequence, uint32_t const numMarkers
     return result;
 }
 
-}  // namespace etlng::impl
+}  // namespace etl::impl

src/etl/impl/AsyncGrpcCall.hpp

@@ -19,7 +19,7 @@
 
 #pragma once
 
-#include "etlng/InitialLoadObserverInterface.hpp"
+#include "etl/InitialLoadObserverInterface.hpp"
 #include "util/log/Logger.hpp"
 
 #include <grpcpp/client_context.h>
@@ -34,7 +34,7 @@
 #include <string>
 #include <vector>
 
-namespace etlng::impl {
+namespace etl::impl {
 
 class AsyncGrpcCall {
 public:
@@ -68,7 +68,7 @@ public:
     process(
         std::unique_ptr<StubType>& stub,
         grpc::CompletionQueue& cq,
-        etlng::InitialLoadObserverInterface& loader,
+        InitialLoadObserverInterface& loader,
         bool abort
     );
 
@@ -82,4 +82,4 @@ public:
     getLastKey();
 };
 
-}  // namespace etlng::impl
+}  // namespace etl::impl

src/etl/impl/CacheLoader.hpp

@@ -53,7 +53,7 @@ class CacheLoaderImpl {
     std::shared_ptr<BackendInterface> backend_;
     std::reference_wrapper<CacheType> cache_;
 
-    etl::ThreadSafeQueue<CursorPair> queue_;
+    ThreadSafeQueue<CursorPair> queue_;
     std::atomic_int16_t remaining_;
 
     std::chrono::steady_clock::time_point startTime_ = std::chrono::steady_clock::now();

src/etl/impl/CacheUpdater.hpp

@@ -21,15 +21,15 @@
 
 #include "data/LedgerCacheInterface.hpp"
 #include "data/Types.hpp"
-#include "etlng/CacheUpdaterInterface.hpp"
-#include "etlng/Models.hpp"
+#include "etl/CacheUpdaterInterface.hpp"
+#include "etl/Models.hpp"
 #include "util/log/Logger.hpp"
 
 #include <cstdint>
 #include <functional>
 #include <vector>
 
-namespace etlng::impl {
+namespace etl::impl {
 
 class CacheUpdater : public CacheUpdaterInterface {
     std::reference_wrapper<data::LedgerCacheInterface> cache_;
@@ -66,4 +66,4 @@ public:
     }
 };
 
-}  // namespace etlng::impl
+}  // namespace etl::impl

src/etl/impl/Extraction.cpp

@@ -17,13 +17,13 @@
 */
 //==============================================================================
 
-#include "etlng/impl/Extraction.hpp"
+#include "etl/impl/Extraction.hpp"
 
 #include "data/DBHelpers.hpp"
 #include "data/Types.hpp"
 #include "etl/LedgerFetcherInterface.hpp"
+#include "etl/Models.hpp"
 #include "etl/impl/LedgerFetcher.hpp"
-#include "etlng/Models.hpp"
 #include "util/Assert.hpp"
 #include "util/LedgerUtils.hpp"
 #include "util/Profiler.hpp"
@@ -47,7 +47,7 @@
 #include <utility>
 #include <vector>
 
-namespace etlng::impl {
+namespace etl::impl {
 
 model::Object::ModType
 extractModType(PBModType type)
@@ -216,4 +216,4 @@ Extractor::extractLedgerOnly(uint32_t seq)
     return batch;
 }
 
-}  // namespace etlng::impl
+}  // namespace etl::impl