fix

Add another allowed_error for this rarity.

Fixes: #8773

.gitattributes

@@ -0,0 +1,2 @@
+# allows for nicer hunk headers with git show
+*.rs diff=rust

.github/actionlint.yml

@@ -1,13 +1,15 @@
 self-hosted-runner:
   labels:
     - arm64
-    - gen3
     - large
     - large-arm64
     - small
     - small-arm64
     - us-east-2
 config-variables:
+  - BENCHMARK_PROJECT_ID_PUB
+  - BENCHMARK_PROJECT_ID_SUB
   - REMOTE_STORAGE_AZURE_CONTAINER
   - REMOTE_STORAGE_AZURE_REGION
   - SLACK_UPCOMING_RELEASE_CHANNEL_ID
+  - DEV_AWS_OIDC_ROLE_ARN

.github/actions/allure-report-generate/action.yml

@@ -183,7 +183,7 @@ runs:
       uses: actions/cache@v4
       with:
         path: ~/.cache/pypoetry/virtualenvs
-        key: v2-${{ runner.os }}-python-deps-${{ hashFiles('poetry.lock') }}
+        key: v2-${{ runner.os }}-${{ runner.arch }}-python-deps-${{ hashFiles('poetry.lock') }}
 
     - name: Store Allure test stat in the DB (new)
       if: ${{ !cancelled() && inputs.store-test-results-into-db == 'true' }}

.github/actions/download/action.yml

@@ -26,7 +26,7 @@ runs:
         TARGET: ${{ inputs.path }}
         ARCHIVE: /tmp/downloads/${{ inputs.name }}.tar.zst
         SKIP_IF_DOES_NOT_EXIST: ${{ inputs.skip-if-does-not-exist }}
-        PREFIX: artifacts/${{ inputs.prefix || format('{0}/{1}', github.run_id, github.run_attempt) }}
+        PREFIX: artifacts/${{ inputs.prefix || format('{0}/{1}/{2}', github.event.pull_request.head.sha || github.sha, github.run_id, github.run_attempt) }}
       run: |
         BUCKET=neon-github-public-dev
         FILENAME=$(basename $ARCHIVE)

.github/actions/neon-project-create/action.yml

@@ -9,16 +9,13 @@ inputs:
     description: 'Region ID, if not set the project will be created in the default region'
     default: aws-us-east-2
   postgres_version:
-    description: 'Postgres version; default is 15'
-    default: '15'
+    description: 'Postgres version; default is 16'
+    default: '16'
   api_host:
     description: 'Neon API host'
     default: console-stage.neon.build
-  provisioner:
-    description: 'k8s-pod or k8s-neonvm'
-    default: 'k8s-pod'
   compute_units:
-    description: '[Min, Max] compute units; Min and Max are used for k8s-neonvm with autoscaling, for k8s-pod values Min and Max should be equal'
+    description: '[Min, Max] compute units'
     default: '[1, 1]'
 
 outputs:
@@ -37,10 +34,6 @@ runs:
       # A shell without `set -x` to not to expose password/dsn in logs
       shell: bash -euo pipefail {0}
       run: |
-        if [ "${PROVISIONER}" == "k8s-pod" ] && [ "${MIN_CU}" != "${MAX_CU}" ]; then
-          echo >&2 "For k8s-pod provisioner MIN_CU should be equal to MAX_CU"
-        fi
-
         project=$(curl \
           "https://${API_HOST}/api/v2/projects" \
           --fail \
@@ -52,7 +45,7 @@ runs:
               \"name\": \"Created by actions/neon-project-create; GITHUB_RUN_ID=${GITHUB_RUN_ID}\",
               \"pg_version\": ${POSTGRES_VERSION},
               \"region_id\": \"${REGION_ID}\",
-              \"provisioner\": \"${PROVISIONER}\",
+              \"provisioner\": \"k8s-neonvm\",
               \"autoscaling_limit_min_cu\": ${MIN_CU},
               \"autoscaling_limit_max_cu\": ${MAX_CU},
               \"settings\": { }
@@ -75,6 +68,5 @@ runs:
         API_KEY: ${{ inputs.api_key }}
         REGION_ID: ${{ inputs.region_id }}
         POSTGRES_VERSION: ${{ inputs.postgres_version }}
-        PROVISIONER: ${{ inputs.provisioner }}
         MIN_CU: ${{ fromJSON(inputs.compute_units)[0] }}
         MAX_CU: ${{ fromJSON(inputs.compute_units)[1] }}

.github/actions/run-python-test-set/action.yml

@@ -43,7 +43,7 @@ inputs:
   pg_version:
     description: 'Postgres version to use for tests'
     required: false
-    default: 'v14'
+    default: 'v16'
   benchmark_durations:
     description: 'benchmark durations JSON'
     required: false
@@ -56,14 +56,14 @@ runs:
       if: inputs.build_type != 'remote'
       uses: ./.github/actions/download
       with:
-        name: neon-${{ runner.os }}-${{ inputs.build_type }}-artifact
+        name: neon-${{ runner.os }}-${{ runner.arch }}-${{ inputs.build_type }}-artifact
         path: /tmp/neon
 
     - name: Download Neon binaries for the previous release
       if: inputs.build_type != 'remote'
       uses: ./.github/actions/download
       with:
-        name: neon-${{ runner.os }}-${{ inputs.build_type }}-artifact
+        name: neon-${{ runner.os }}-${{ runner.arch }}-${{ inputs.build_type }}-artifact
         path: /tmp/neon-previous
         prefix: latest
 
@@ -83,13 +83,12 @@ runs:
       uses: actions/checkout@v4
       with:
         submodules: true
-        fetch-depth: 1
 
     - name: Cache poetry deps
       uses: actions/cache@v4
       with:
         path: ~/.cache/pypoetry/virtualenvs
-        key: v2-${{ runner.os }}-python-deps-${{ hashFiles('poetry.lock') }}
+        key: v2-${{ runner.os }}-${{ runner.arch }}-python-deps-${{ hashFiles('poetry.lock') }}
 
     - name: Install Python deps
       shell: bash -euxo pipefail {0}
@@ -114,6 +113,8 @@ runs:
         export PLATFORM=${PLATFORM:-github-actions-selfhosted}
         export POSTGRES_DISTRIB_DIR=${POSTGRES_DISTRIB_DIR:-/tmp/neon/pg_install}
         export DEFAULT_PG_VERSION=${PG_VERSION#v}
+        export LD_LIBRARY_PATH=${POSTGRES_DISTRIB_DIR}/v${DEFAULT_PG_VERSION}/lib
+        export BENCHMARK_CONNSTR=${BENCHMARK_CONNSTR:-}
 
         if [ "${BUILD_TYPE}" = "remote" ]; then
           export REMOTE_ENV=1
@@ -129,8 +130,8 @@ runs:
           exit 1
         fi
         if [[ "${{ inputs.run_in_parallel }}" == "true" ]]; then
-          # -n16 uses sixteen processes to run tests via pytest-xdist
-          EXTRA_PARAMS="-n16 $EXTRA_PARAMS"
+          # -n sets the number of parallel processes that pytest-xdist will run
+          EXTRA_PARAMS="-n12 $EXTRA_PARAMS"
 
           # --dist=loadgroup points tests marked with @pytest.mark.xdist_group
           # to the same worker to make @pytest.mark.order work with xdist
@@ -178,13 +179,20 @@ runs:
 
         # Wake up the cluster if we use remote neon instance
         if [ "${{ inputs.build_type }}" = "remote" ] && [ -n "${BENCHMARK_CONNSTR}" ]; then
-          ${POSTGRES_DISTRIB_DIR}/v${DEFAULT_PG_VERSION}/bin/psql ${BENCHMARK_CONNSTR} -c "SELECT version();"
+          QUERIES=("SELECT version()")
+          if [[ "${PLATFORM}" = "neon"* ]]; then
+            QUERIES+=("SHOW neon.tenant_id")
+            QUERIES+=("SHOW neon.timeline_id")
+          fi
+
+          for q in "${QUERIES[@]}"; do
+            ${POSTGRES_DISTRIB_DIR}/v${DEFAULT_PG_VERSION}/bin/psql ${BENCHMARK_CONNSTR} -c "${q}"
+          done
         fi
 
         # Run the tests.
         #
-        # The junit.xml file allows CI tools to display more fine-grained test information
-        # in its "Tests" tab in the results page.
+        # --alluredir saves test results in Allure format (in a specified directory)
         # --verbose prints name of each test (helpful when there are
         # multiple tests in one file)
         # -rA prints summary in the end
@@ -193,7 +201,6 @@ runs:
         #
         mkdir -p $TEST_OUTPUT/allure/results
         "${cov_prefix[@]}" ./scripts/pytest \
-          --junitxml=$TEST_OUTPUT/junit.xml \
           --alluredir=$TEST_OUTPUT/allure/results \
           --tb=short \
           --verbose \

.github/actions/set-docker-config-dir/action.yml

@@ -0,0 +1,36 @@
+name: "Set custom docker config directory"
+description: "Create a directory for docker config and set DOCKER_CONFIG"
+
+# Use custom DOCKER_CONFIG directory to avoid conflicts with default settings
+runs:
+  using: "composite"
+  steps:
+  - name: Show warning on GitHub-hosted runners
+    if: runner.environment == 'github-hosted'
+    shell: bash -euo pipefail {0}
+    run: |
+      # Using the following environment variables to find a path to the workflow file
+      # ${GITHUB_WORKFLOW_REF} - octocat/hello-world/.github/workflows/my-workflow.yml@refs/heads/my_branch
+      # ${GITHUB_REPOSITORY}   - octocat/hello-world
+      # ${GITHUB_REF}          - refs/heads/my_branch
+      # From https://docs.github.com/en/actions/writing-workflows/choosing-what-your-workflow-does/variables
+
+      filename_with_ref=${GITHUB_WORKFLOW_REF#"$GITHUB_REPOSITORY/"}
+      filename=${filename_with_ref%"@$GITHUB_REF"}
+
+      # https://docs.github.com/en/actions/writing-workflows/choosing-what-your-workflow-does/workflow-commands-for-github-actions#setting-a-warning-message
+      title='Unnecessary usage of `.github/actions/set-docker-config-dir`'
+      message='No need to use `.github/actions/set-docker-config-dir` action on GitHub-hosted runners'
+      echo "::warning file=${filename},title=${title}::${message}"
+
+  - uses: pyTooling/Actions/with-post-step@74afc5a42a17a046c90c68cb5cfa627e5c6c5b6b # v1.0.7
+    env:
+      DOCKER_CONFIG: .docker-custom-${{ github.run_id }}-${{ github.run_attempt }}
+    with:
+      main: |
+        mkdir -p "${DOCKER_CONFIG}"
+        echo DOCKER_CONFIG=${DOCKER_CONFIG} | tee -a $GITHUB_ENV
+      post: |
+        if [ -d "${DOCKER_CONFIG}" ]; then
+          rm -r "${DOCKER_CONFIG}"
+        fi

.github/actions/upload/action.yml

@@ -8,7 +8,7 @@ inputs:
     description: "A directory or file to upload"
     required: true
   prefix:
-    description: "S3 prefix. Default is '${GITHUB_RUN_ID}/${GITHUB_RUN_ATTEMPT}'"
+    description: "S3 prefix. Default is '${GITHUB_SHA}/${GITHUB_RUN_ID}/${GITHUB_RUN_ATTEMPT}'"
     required: false
 
 runs:
@@ -45,7 +45,7 @@ runs:
       env:
         SOURCE: ${{ inputs.path }}
         ARCHIVE: /tmp/uploads/${{ inputs.name }}.tar.zst
-        PREFIX: artifacts/${{ inputs.prefix || format('{0}/{1}', github.run_id, github.run_attempt) }}
+        PREFIX: artifacts/${{ inputs.prefix || format('{0}/{1}/{2}', github.event.pull_request.head.sha || github.sha, github.run_id , github.run_attempt) }}
       run: |
         BUCKET=neon-github-public-dev
         FILENAME=$(basename $ARCHIVE)

.github/workflows/_benchmarking_preparation.yml

@@ -0,0 +1,154 @@
+name: Prepare benchmarking databases by restoring dumps
+
+on:
+  workflow_call:
+    # no inputs needed
+    
+defaults:
+  run:
+    shell: bash -euxo pipefail {0}
+
+jobs:
+  setup-databases:
+    strategy:
+      fail-fast: false
+      matrix:
+        platform: [ aws-rds-postgres, aws-aurora-serverless-v2-postgres, neon ] 
+        database: [ clickbench, tpch, userexample ]
+  
+    env:
+      LD_LIBRARY_PATH: /tmp/neon/pg_install/v16/lib
+      PLATFORM: ${{ matrix.platform }}
+      PG_BINARIES: /tmp/neon/pg_install/v16/bin
+
+    runs-on: [ self-hosted, us-east-2, x64 ]
+    container:
+      image: 369495373322.dkr.ecr.eu-central-1.amazonaws.com/build-tools:pinned
+      options: --init
+
+    steps:
+    - name: Set up Connection String
+      id: set-up-prep-connstr
+      run: |
+        case "${PLATFORM}" in
+          neon)
+            CONNSTR=${{ secrets.BENCHMARK_CAPTEST_CONNSTR }} 
+            ;;
+          aws-rds-postgres)
+            CONNSTR=${{ secrets.BENCHMARK_RDS_POSTGRES_CONNSTR }} 
+            ;;
+          aws-aurora-serverless-v2-postgres)
+            CONNSTR=${{ secrets.BENCHMARK_RDS_AURORA_CONNSTR }} 
+            ;;
+          *)
+            echo >&2 "Unknown PLATFORM=${PLATFORM}"
+            exit 1
+            ;;
+        esac
+
+        echo "connstr=${CONNSTR}" >> $GITHUB_OUTPUT  
+
+    - uses: actions/checkout@v4
+
+    - name: Download Neon artifact
+      uses: ./.github/actions/download
+      with:
+        name: neon-${{ runner.os }}-${{ runner.arch }}-release-artifact
+        path: /tmp/neon/
+        prefix: latest
+
+    # we create a table that has one row for each database that we want to restore with the status whether the restore is done    
+    - name: Create benchmark_restore_status table if it does not exist
+      env:
+        BENCHMARK_CONNSTR: ${{ steps.set-up-prep-connstr.outputs.connstr }}
+        DATABASE_NAME: ${{ matrix.database }}
+      # to avoid a race condition of multiple jobs trying to create the table at the same time, 
+      # we use an advisory lock
+      run: |
+        ${PG_BINARIES}/psql "${{ env.BENCHMARK_CONNSTR }}" -c "
+        SELECT pg_advisory_lock(4711);  
+        CREATE TABLE IF NOT EXISTS benchmark_restore_status (
+        databasename text primary key,
+        restore_done boolean
+        );
+        SELECT pg_advisory_unlock(4711);
+        "
+    
+    - name: Check if restore is already done
+      id: check-restore-done
+      env:
+        BENCHMARK_CONNSTR: ${{ steps.set-up-prep-connstr.outputs.connstr }}
+        DATABASE_NAME: ${{ matrix.database }}
+      run: |
+        skip=false
+        if ${PG_BINARIES}/psql "${{ env.BENCHMARK_CONNSTR }}" -tAc "SELECT 1 FROM benchmark_restore_status WHERE databasename='${{ env.DATABASE_NAME }}' AND restore_done=true;" | grep -q 1; then
+          echo "Restore already done for database ${{ env.DATABASE_NAME }} on platform ${{ env.PLATFORM }}. Skipping this database."
+          skip=true
+        fi
+        echo "skip=${skip}" | tee -a $GITHUB_OUTPUT
+
+    - name: Check and create database if it does not exist
+      if: steps.check-restore-done.outputs.skip != 'true'
+      env:
+        BENCHMARK_CONNSTR: ${{ steps.set-up-prep-connstr.outputs.connstr }}
+        DATABASE_NAME: ${{ matrix.database }}
+      run: |
+        DB_EXISTS=$(${PG_BINARIES}/psql "${{ env.BENCHMARK_CONNSTR }}" -tAc "SELECT 1 FROM pg_database WHERE datname='${{ env.DATABASE_NAME }}'")
+        if [ "$DB_EXISTS" != "1" ]; then
+          echo "Database ${{ env.DATABASE_NAME }} does not exist. Creating it..."
+          ${PG_BINARIES}/psql "${{ env.BENCHMARK_CONNSTR }}" -c "CREATE DATABASE \"${{ env.DATABASE_NAME }}\";"
+        else
+          echo "Database ${{ env.DATABASE_NAME }} already exists."
+        fi
+
+    - name: Download dump from S3 to /tmp/dumps
+      if: steps.check-restore-done.outputs.skip != 'true'
+      env:
+        DATABASE_NAME: ${{ matrix.database }}
+      run: |
+        mkdir -p /tmp/dumps
+        aws s3 cp s3://neon-github-dev/performance/pgdumps/$DATABASE_NAME/$DATABASE_NAME.pg_dump /tmp/dumps/ 
+
+    - name: Replace database name in connection string
+      if: steps.check-restore-done.outputs.skip != 'true'
+      id: replace-dbname
+      env:
+        DATABASE_NAME: ${{ matrix.database }}
+        BENCHMARK_CONNSTR: ${{ steps.set-up-prep-connstr.outputs.connstr }}
+      run: |
+        # Extract the part before the database name
+        base_connstr="${BENCHMARK_CONNSTR%/*}"
+        # Extract the query parameters (if any) after the database name
+        query_params="${BENCHMARK_CONNSTR#*\?}"
+        # Reconstruct the new connection string
+        if [ "$query_params" != "$BENCHMARK_CONNSTR" ]; then
+          new_connstr="${base_connstr}/${DATABASE_NAME}?${query_params}"
+        else
+          new_connstr="${base_connstr}/${DATABASE_NAME}"
+        fi
+        echo "database_connstr=${new_connstr}" >> $GITHUB_OUTPUT  
+
+    - name: Restore dump
+      if: steps.check-restore-done.outputs.skip != 'true'
+      env:
+        DATABASE_NAME: ${{ matrix.database }}
+        DATABASE_CONNSTR: ${{ steps.replace-dbname.outputs.database_connstr }}
+        # the following works only with larger computes: 
+        # PGOPTIONS: "-c maintenance_work_mem=8388608 -c max_parallel_maintenance_workers=7"
+        # we add the || true because:
+        # the dumps were created with Neon and contain neon extensions that are not 
+        # available in RDS, so we will always report an error, but we can ignore it
+      run: |
+        ${PG_BINARIES}/pg_restore --clean --if-exists --no-owner --jobs=4 \
+        -d "${DATABASE_CONNSTR}" /tmp/dumps/${DATABASE_NAME}.pg_dump || true
+
+    - name: Update benchmark_restore_status table
+      if: steps.check-restore-done.outputs.skip != 'true'
+      env:
+        BENCHMARK_CONNSTR: ${{ steps.set-up-prep-connstr.outputs.connstr }}
+        DATABASE_NAME: ${{ matrix.database }}
+      run: |
+        ${PG_BINARIES}/psql "${{ env.BENCHMARK_CONNSTR }}" -c "
+        INSERT INTO benchmark_restore_status (databasename, restore_done) VALUES ('${{ env.DATABASE_NAME }}', true)
+        ON CONFLICT (databasename) DO UPDATE SET restore_done = true;
+        "

.github/workflows/_build-and-test-locally.yml

@@ -0,0 +1,290 @@
+name: Build and Test Locally
+
+on:
+  workflow_call:
+    inputs:
+      arch:
+        description: 'x64 or arm64'
+        required: true
+        type: string
+      build-tag:
+        description: 'build tag'
+        required: true
+        type: string
+      build-tools-image:
+        description: 'build-tools image'
+        required: true
+        type: string
+      build-type:
+        description: 'debug or release'
+        required: true
+        type: string
+      pg-versions:
+        description: 'a json array of postgres versions to run regression tests on'
+        required: true
+        type: string
+
+defaults:
+  run:
+    shell: bash -euxo pipefail {0}
+
+env:
+  RUST_BACKTRACE: 1
+  COPT: '-Werror'
+  AWS_ACCESS_KEY_ID: ${{ secrets.AWS_ACCESS_KEY_DEV }}
+  AWS_SECRET_ACCESS_KEY: ${{ secrets.AWS_SECRET_KEY_DEV }}
+
+jobs:
+  build-neon:
+    runs-on: ${{ fromJson(format('["self-hosted", "{0}"]', inputs.arch == 'arm64' && 'large-arm64' || 'large')) }}
+    container:
+      image: ${{ inputs.build-tools-image }}
+      credentials:
+        username: ${{ secrets.NEON_DOCKERHUB_USERNAME }}
+        password: ${{ secrets.NEON_DOCKERHUB_PASSWORD }}
+      # Raise locked memory limit for tokio-epoll-uring.
+      # On 5.10 LTS kernels < 5.10.162 (and generally mainline kernels < 5.12),
+      # io_uring will account the memory of the CQ and SQ as locked.
+      # More details: https://github.com/neondatabase/neon/issues/6373#issuecomment-1905814391
+      options: --init --shm-size=512mb --ulimit memlock=67108864:67108864
+    env:
+      BUILD_TYPE: ${{ inputs.build-type }}
+      GIT_VERSION: ${{ github.event.pull_request.head.sha || github.sha }}
+      BUILD_TAG: ${{ inputs.build-tag }}
+
+    steps:
+      - name: Fix git ownership
+        run: |
+          # Workaround for `fatal: detected dubious ownership in repository at ...`
+          #
+          # Use both ${{ github.workspace }} and ${GITHUB_WORKSPACE} because they're different on host and in containers
+          #   Ref https://github.com/actions/checkout/issues/785
+          #
+          git config --global --add safe.directory ${{ github.workspace }}
+          git config --global --add safe.directory ${GITHUB_WORKSPACE}
+          for r in 14 15 16; do
+            git config --global --add safe.directory "${{ github.workspace }}/vendor/postgres-v$r"
+            git config --global --add safe.directory "${GITHUB_WORKSPACE}/vendor/postgres-v$r"
+          done
+
+      - uses: actions/checkout@v4
+        with:
+          submodules: true
+
+      - name: Set pg 14 revision for caching
+        id: pg_v14_rev
+        run: echo pg_rev=$(git rev-parse HEAD:vendor/postgres-v14) >> $GITHUB_OUTPUT
+
+      - name: Set pg 15 revision for caching
+        id: pg_v15_rev
+        run: echo pg_rev=$(git rev-parse HEAD:vendor/postgres-v15) >> $GITHUB_OUTPUT
+
+      - name: Set pg 16 revision for caching
+        id: pg_v16_rev
+        run: echo pg_rev=$(git rev-parse HEAD:vendor/postgres-v16) >> $GITHUB_OUTPUT
+
+      # Set some environment variables used by all the steps.
+      #
+      # CARGO_FLAGS is extra options to pass to "cargo build", "cargo test" etc.
+      #   It also includes --features, if any
+      #
+      # CARGO_FEATURES is passed to "cargo metadata". It is separate from CARGO_FLAGS,
+      #   because "cargo metadata" doesn't accept --release or --debug options
+      #
+      # We run tests with addtional features, that are turned off by default (e.g. in release builds), see
+      # corresponding Cargo.toml files for their descriptions.
+      - name: Set env variables
+        run: |
+          CARGO_FEATURES="--features testing"
+          if [[ $BUILD_TYPE == "debug" ]]; then
+            cov_prefix="scripts/coverage --profraw-prefix=$GITHUB_JOB --dir=/tmp/coverage run"
+            CARGO_FLAGS="--locked"
+          elif [[ $BUILD_TYPE == "release" ]]; then
+            cov_prefix=""
+            CARGO_FLAGS="--locked --release"
+          fi
+          {
+            echo "cov_prefix=${cov_prefix}"
+            echo "CARGO_FEATURES=${CARGO_FEATURES}"
+            echo "CARGO_FLAGS=${CARGO_FLAGS}"
+            echo "CARGO_HOME=${GITHUB_WORKSPACE}/.cargo"
+          } >> $GITHUB_ENV
+
+      - name: Cache postgres v14 build
+        id: cache_pg_14
+        uses: actions/cache@v4
+        with:
+          path: pg_install/v14
+          key: v1-${{ runner.os }}-${{ runner.arch }}-${{ inputs.build-type }}-pg-${{ steps.pg_v14_rev.outputs.pg_rev }}-${{ hashFiles('Makefile', 'Dockerfile.build-tools') }}
+
+      - name: Cache postgres v15 build
+        id: cache_pg_15
+        uses: actions/cache@v4
+        with:
+          path: pg_install/v15
+          key: v1-${{ runner.os }}-${{ runner.arch }}-${{ inputs.build-type }}-pg-${{ steps.pg_v15_rev.outputs.pg_rev }}-${{ hashFiles('Makefile', 'Dockerfile.build-tools') }}
+
+      - name: Cache postgres v16 build
+        id: cache_pg_16
+        uses: actions/cache@v4
+        with:
+          path: pg_install/v16
+          key: v1-${{ runner.os }}-${{ runner.arch }}-${{ inputs.build-type }}-pg-${{ steps.pg_v16_rev.outputs.pg_rev }}-${{ hashFiles('Makefile', 'Dockerfile.build-tools') }}
+
+      - name: Build postgres v14
+        if: steps.cache_pg_14.outputs.cache-hit != 'true'
+        run: mold -run make postgres-v14 -j$(nproc)
+
+      - name: Build postgres v15
+        if: steps.cache_pg_15.outputs.cache-hit != 'true'
+        run: mold -run make postgres-v15 -j$(nproc)
+
+      - name: Build postgres v16
+        if: steps.cache_pg_16.outputs.cache-hit != 'true'
+        run: mold -run make postgres-v16 -j$(nproc)
+
+      - name: Build neon extensions
+        run: mold -run make neon-pg-ext -j$(nproc)
+
+      - name: Build walproposer-lib
+        run: mold -run make walproposer-lib -j$(nproc)
+
+      - name: Run cargo build
+        run: |
+          PQ_LIB_DIR=$(pwd)/pg_install/v16/lib
+          export PQ_LIB_DIR
+          ${cov_prefix} mold -run cargo build $CARGO_FLAGS $CARGO_FEATURES --bins --tests
+
+      # Do install *before* running rust tests because they might recompile the
+      # binaries with different features/flags.
+      - name: Install rust binaries
+        run: |
+          # Install target binaries
+          mkdir -p /tmp/neon/bin/
+          binaries=$(
+            ${cov_prefix} cargo metadata $CARGO_FEATURES --format-version=1 --no-deps |
+            jq -r '.packages[].targets[] | select(.kind | index("bin")) | .name'
+          )
+          for bin in $binaries; do
+            SRC=target/$BUILD_TYPE/$bin
+            DST=/tmp/neon/bin/$bin
+            cp "$SRC" "$DST"
+          done
+
+          # Install test executables and write list of all binaries (for code coverage)
+          if [[ $BUILD_TYPE == "debug" ]]; then
+            # Keep bloated coverage data files away from the rest of the artifact
+            mkdir -p /tmp/coverage/
+
+            mkdir -p /tmp/neon/test_bin/
+
+            test_exe_paths=$(
+              ${cov_prefix} cargo test $CARGO_FLAGS $CARGO_FEATURES --message-format=json --no-run |
+              jq -r '.executable | select(. != null)'
+            )
+            for bin in $test_exe_paths; do
+              SRC=$bin
+              DST=/tmp/neon/test_bin/$(basename $bin)
+
+              # We don't need debug symbols for code coverage, so strip them out to make
+              # the artifact smaller.
+              strip "$SRC" -o "$DST"
+              echo "$DST" >> /tmp/coverage/binaries.list
+            done
+
+            for bin in $binaries; do
+              echo "/tmp/neon/bin/$bin" >> /tmp/coverage/binaries.list
+            done
+          fi
+
+      - name: Run rust tests
+        env:
+          NEXTEST_RETRIES: 3
+        run: |
+          PQ_LIB_DIR=$(pwd)/pg_install/v16/lib
+          export PQ_LIB_DIR
+          LD_LIBRARY_PATH=$(pwd)/pg_install/v16/lib
+          export LD_LIBRARY_PATH
+
+          #nextest does not yet support running doctests
+          ${cov_prefix} cargo test --doc $CARGO_FLAGS $CARGO_FEATURES
+
+          for io_engine in std-fs tokio-epoll-uring ; do
+            NEON_PAGESERVER_UNIT_TEST_VIRTUAL_FILE_IOENGINE=$io_engine ${cov_prefix} cargo nextest run $CARGO_FLAGS $CARGO_FEATURES
+          done
+
+          # Run separate tests for real S3
+          export ENABLE_REAL_S3_REMOTE_STORAGE=nonempty
+          export REMOTE_STORAGE_S3_BUCKET=neon-github-ci-tests
+          export REMOTE_STORAGE_S3_REGION=eu-central-1
+          ${cov_prefix} cargo nextest run $CARGO_FLAGS $CARGO_FEATURES -E 'package(remote_storage)' -E 'test(test_real_s3)'
+
+          # Run separate tests for real Azure Blob Storage
+          # XXX: replace region with `eu-central-1`-like region
+          export ENABLE_REAL_AZURE_REMOTE_STORAGE=y
+          export AZURE_STORAGE_ACCOUNT="${{ secrets.AZURE_STORAGE_ACCOUNT_DEV }}"
+          export AZURE_STORAGE_ACCESS_KEY="${{ secrets.AZURE_STORAGE_ACCESS_KEY_DEV }}"
+          export REMOTE_STORAGE_AZURE_CONTAINER="${{ vars.REMOTE_STORAGE_AZURE_CONTAINER }}"
+          export REMOTE_STORAGE_AZURE_REGION="${{ vars.REMOTE_STORAGE_AZURE_REGION }}"
+          ${cov_prefix} cargo nextest run $CARGO_FLAGS $CARGO_FEATURES -E 'package(remote_storage)' -E 'test(test_real_azure)'
+
+      - name: Install postgres binaries
+        run: cp -a pg_install /tmp/neon/pg_install
+
+      - name: Upload Neon artifact
+        uses: ./.github/actions/upload
+        with:
+          name: neon-${{ runner.os }}-${{ runner.arch }}-${{ inputs.build-type }}-artifact
+          path: /tmp/neon
+
+      # XXX: keep this after the binaries.list is formed, so the coverage can properly work later
+      - name: Merge and upload coverage data
+        if: inputs.build-type == 'debug'
+        uses: ./.github/actions/save-coverage-data
+
+  regress-tests:
+    # Run test on x64 only
+    if: inputs.arch == 'x64'
+    needs: [ build-neon ]
+    runs-on: ${{ fromJson(format('["self-hosted", "{0}"]', inputs.arch == 'arm64' && 'large-arm64' || 'large')) }}
+    container:
+      image: ${{ inputs.build-tools-image }}
+      credentials:
+        username: ${{ secrets.NEON_DOCKERHUB_USERNAME }}
+        password: ${{ secrets.NEON_DOCKERHUB_PASSWORD }}
+      # for changed limits, see comments on `options:` earlier in this file
+      options: --init --shm-size=512mb --ulimit memlock=67108864:67108864
+    strategy:
+      fail-fast: false
+      matrix:
+        pg_version: ${{ fromJson(inputs.pg-versions) }}
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          submodules: true
+
+      - name: Pytest regression tests
+        uses: ./.github/actions/run-python-test-set
+        timeout-minutes: 60
+        with:
+          build_type: ${{ inputs.build-type }}
+          test_selection: regress
+          needs_postgres_source: true
+          run_with_real_s3: true
+          real_s3_bucket: neon-github-ci-tests
+          real_s3_region: eu-central-1
+          rerun_flaky: true
+          pg_version: ${{ matrix.pg_version }}
+        env:
+          TEST_RESULT_CONNSTR: ${{ secrets.REGRESS_TEST_RESULT_CONNSTR_NEW }}
+          CHECK_ONDISK_DATA_COMPATIBILITY: nonempty
+          BUILD_TAG: ${{ inputs.build-tag }}
+          PAGESERVER_VIRTUAL_FILE_IO_ENGINE: tokio-epoll-uring
+
+      # Temporary disable this step until we figure out why it's so flaky
+      # Ref https://github.com/neondatabase/neon/issues/4540
+      - name: Merge and upload coverage data
+        if: |
+          false &&
+          inputs.build-type == 'debug' && matrix.pg_version == 'v16'
+        uses: ./.github/actions/save-coverage-data

.github/workflows/actionlint.yml

@@ -36,15 +36,16 @@ jobs:
           fail_on_error: true
           filter_mode: nofilter
           level: error
-      - run: |
+
+      - name: Disallow 'ubuntu-latest' runners
+        run: |
           PAT='^\s*runs-on:.*-latest'
-          if grep -ERq $PAT .github/workflows
-          then
+          if grep -ERq $PAT .github/workflows; then
             grep -ERl $PAT .github/workflows |\
             while read -r f
             do
-              l=$(grep -nE $PAT .github/workflows/release.yml | awk -F: '{print $1}' | head -1)
-              echo "::error file=$f,line=$l::Please, do not use ubuntu-latest images to run on, use LTS instead."
+              l=$(grep -nE $PAT $f | awk -F: '{print $1}' | head -1)
+              echo "::error file=$f,line=$l::Please use 'ubuntu-22.04' instead of 'ubuntu-latest'"
             done
             exit 1
           fi

.github/workflows/benchmarking.yml

@@ -56,28 +56,53 @@ concurrency:
 jobs:
   bench:
     if: ${{ github.event.inputs.run_only_pgvector_tests == 'false' || github.event.inputs.run_only_pgvector_tests == null }}
+    permissions:
+      contents: write
+      statuses: write
+      id-token: write # Required for OIDC authentication in azure runners
+    strategy:
+      fail-fast: false
+      matrix:
+        include:
+          - DEFAULT_PG_VERSION: 16
+            PLATFORM: "neon-staging"
+            region_id: ${{ github.event.inputs.region_id || 'aws-us-east-2' }}
+            RUNNER: [ self-hosted, us-east-2, x64 ]
+            IMAGE: 369495373322.dkr.ecr.eu-central-1.amazonaws.com/build-tools:pinned
+          - DEFAULT_PG_VERSION: 16
+            PLATFORM: "azure-staging"
+            region_id: 'azure-eastus2'
+            RUNNER: [ self-hosted, eastus2, x64 ]
+            IMAGE: neondatabase/build-tools:pinned
     env:
       TEST_PG_BENCH_DURATIONS_MATRIX: "300"
       TEST_PG_BENCH_SCALES_MATRIX: "10,100"
       POSTGRES_DISTRIB_DIR: /tmp/neon/pg_install
-      DEFAULT_PG_VERSION: 14
+      DEFAULT_PG_VERSION: ${{ matrix.DEFAULT_PG_VERSION }}
       TEST_OUTPUT: /tmp/test_output
       BUILD_TYPE: remote
       SAVE_PERF_REPORT: ${{ github.event.inputs.save_perf_report || ( github.ref_name == 'main' ) }}
-      PLATFORM: "neon-staging"
+      PLATFORM: ${{ matrix.PLATFORM }}
 
-    runs-on: [ self-hosted, us-east-2, x64 ]
+    runs-on: ${{ matrix.RUNNER }}
     container:
-      image: 369495373322.dkr.ecr.eu-central-1.amazonaws.com/build-tools:pinned
+      image: ${{ matrix.IMAGE }}
       options: --init
 
     steps:
     - uses: actions/checkout@v4
 
+    - name: Configure AWS credentials # necessary on Azure runners
+      uses: aws-actions/configure-aws-credentials@v4
+      with:
+        aws-region: eu-central-1
+        role-to-assume: ${{ vars.DEV_AWS_OIDC_ROLE_ARN }}
+        role-duration-seconds: 18000 # 5 hours
+
     - name: Download Neon artifact
       uses: ./.github/actions/download
       with:
-        name: neon-${{ runner.os }}-release-artifact
+        name: neon-${{ runner.os }}-${{ runner.arch }}-release-artifact
         path: /tmp/neon/
         prefix: latest
 
@@ -85,7 +110,7 @@ jobs:
       id: create-neon-project
       uses: ./.github/actions/neon-project-create
       with:
-        region_id: ${{ github.event.inputs.region_id || 'aws-us-east-2' }}
+        region_id: ${{ matrix.region_id }}
         postgres_version: ${{ env.DEFAULT_PG_VERSION }}
         api_key: ${{ secrets.NEON_STAGING_API_KEY }}
 
@@ -96,10 +121,18 @@ jobs:
         test_selection: performance
         run_in_parallel: false
         save_perf_report: ${{ env.SAVE_PERF_REPORT }}
+        pg_version: ${{ env.DEFAULT_PG_VERSION }}
         # Set --sparse-ordering option of pytest-order plugin
         # to ensure tests are running in order of appears in the file.
         # It's important for test_perf_pgbench.py::test_pgbench_remote_* tests
-        extra_params: -m remote_cluster --sparse-ordering --timeout 5400 --ignore test_runner/performance/test_perf_olap.py --ignore test_runner/performance/test_perf_pgvector_queries.py
+        extra_params:
+          -m remote_cluster
+          --sparse-ordering
+          --timeout 14400
+          --ignore test_runner/performance/test_perf_olap.py
+          --ignore test_runner/performance/test_perf_pgvector_queries.py
+          --ignore test_runner/performance/test_logical_replication.py
+          --ignore test_runner/performance/test_physical_replication.py
       env:
         BENCHMARK_CONNSTR: ${{ steps.create-neon-project.outputs.dsn }}
         VIP_VAP_ACCESS_TOKEN: "${{ secrets.VIP_VAP_ACCESS_TOKEN }}"
@@ -113,6 +146,7 @@ jobs:
         api_key: ${{ secrets.NEON_STAGING_API_KEY }}
 
     - name: Create Allure report
+      id: create-allure-report
       if: ${{ !cancelled() }}
       uses: ./.github/actions/allure-report-generate
 
@@ -121,7 +155,87 @@ jobs:
       uses: slackapi/slack-github-action@v1
       with:
         channel-id: "C033QLM5P7D" # dev-staging-stream
-        slack-message: "Periodic perf testing: ${{ job.status }}\n${{ github.server_url }}/${{ github.repository }}/actions/runs/${{ github.run_id }}"
+        slack-message: |
+          Periodic perf testing: ${{ job.status }}
+          <${{ github.server_url }}/${{ github.repository }}/actions/runs/${{ github.run_id }}|GitHub Run>
+          <${{ steps.create-allure-report.outputs.report-url }}|Allure report>
+      env:
+        SLACK_BOT_TOKEN: ${{ secrets.SLACK_BOT_TOKEN }}
+
+  replication-tests:
+    if: ${{ github.event.inputs.run_only_pgvector_tests == 'false' || github.event.inputs.run_only_pgvector_tests == null }}
+    env:
+      POSTGRES_DISTRIB_DIR: /tmp/neon/pg_install
+      DEFAULT_PG_VERSION: 16
+      TEST_OUTPUT: /tmp/test_output
+      BUILD_TYPE: remote
+      SAVE_PERF_REPORT: ${{ github.event.inputs.save_perf_report || ( github.ref_name == 'main' ) }}
+      PLATFORM: "neon-staging"
+
+    runs-on: [ self-hosted, us-east-2, x64 ]
+    container:
+      image: 369495373322.dkr.ecr.eu-central-1.amazonaws.com/build-tools:pinned
+      options: --init
+
+    steps:
+    - uses: actions/checkout@v4
+
+
+    - name: Download Neon artifact
+      uses: ./.github/actions/download
+      with:
+        name: neon-${{ runner.os }}-${{ runner.arch }}-release-artifact
+        path: /tmp/neon/
+        prefix: latest
+
+    - name: Run Logical Replication benchmarks
+      uses: ./.github/actions/run-python-test-set
+      with:
+        build_type: ${{ env.BUILD_TYPE }}
+        test_selection: performance/test_logical_replication.py
+        run_in_parallel: false
+        save_perf_report: ${{ env.SAVE_PERF_REPORT }}
+        extra_params: -m remote_cluster --timeout 5400
+        pg_version: ${{ env.DEFAULT_PG_VERSION }}
+      env:
+        VIP_VAP_ACCESS_TOKEN: "${{ secrets.VIP_VAP_ACCESS_TOKEN }}"
+        PERF_TEST_RESULT_CONNSTR: "${{ secrets.PERF_TEST_RESULT_CONNSTR }}"
+        NEON_API_KEY: ${{ secrets.NEON_STAGING_API_KEY }}
+        BENCHMARK_PROJECT_ID_PUB: ${{ vars.BENCHMARK_PROJECT_ID_PUB }}
+        BENCHMARK_PROJECT_ID_SUB: ${{ vars.BENCHMARK_PROJECT_ID_SUB }}
+
+    - name: Run Physical Replication benchmarks
+      uses: ./.github/actions/run-python-test-set
+      with:
+        build_type: ${{ env.BUILD_TYPE }}
+        test_selection: performance/test_physical_replication.py
+        run_in_parallel: false
+        save_perf_report: ${{ env.SAVE_PERF_REPORT }}
+        extra_params: -m remote_cluster --timeout 5400
+        pg_version: ${{ env.DEFAULT_PG_VERSION }}
+      env:
+        VIP_VAP_ACCESS_TOKEN: "${{ secrets.VIP_VAP_ACCESS_TOKEN }}"
+        PERF_TEST_RESULT_CONNSTR: "${{ secrets.PERF_TEST_RESULT_CONNSTR }}"
+        NEON_API_KEY: ${{ secrets.NEON_STAGING_API_KEY }}
+
+    - name: Create Allure report
+      id: create-allure-report
+      if: ${{ !cancelled() }}
+      uses: ./.github/actions/allure-report-generate
+      with:
+        store-test-results-into-db: true
+      env:
+        REGRESS_TEST_RESULT_CONNSTR_NEW: ${{ secrets.REGRESS_TEST_RESULT_CONNSTR_NEW }}
+
+    - name: Post to a Slack channel
+      if: ${{ github.event.schedule && failure() }}
+      uses: slackapi/slack-github-action@v1
+      with:
+        channel-id: "C06T9AMNDQQ" # on-call-compute-staging-stream
+        slack-message: |
+          Periodic replication testing: ${{ job.status }}
+          <${{ github.server_url }}/${{ github.repository }}/actions/runs/${{ github.run_id }}|GitHub Run>
+          <${{ steps.create-allure-report.outputs.report-url }}|Allure report>
       env:
         SLACK_BOT_TOKEN: ${{ secrets.SLACK_BOT_TOKEN }}
 
@@ -130,13 +244,16 @@ jobs:
     # Create matrices for the benchmarking jobs, so we run benchmarks on rds only once a week (on Saturday)
     #
     # Available platforms:
-    # - neon-captest-new: Freshly created project (1 CU)
-    # - neon-captest-freetier: Use freetier-sized compute (0.25 CU)
-    # - neon-captest-reuse: Reusing existing project
+    # - neonvm-captest-new: Freshly created project (1 CU)
+    # - neonvm-captest-freetier: Use freetier-sized compute (0.25 CU)
+    # - neonvm-captest-azure-new: Freshly created project (1 CU) in azure region
+    # - neonvm-captest-azure-freetier: Use freetier-sized compute (0.25 CU) in azure region
+    # - neonvm-captest-reuse: Reusing existing project
     # - rds-aurora: Aurora Postgres Serverless v2 with autoscaling from 0.5 to 2 ACUs
     # - rds-postgres: RDS Postgres db.m5.large instance (2 vCPU, 8 GiB) with gp3 EBS storage
     env:
       RUN_AWS_RDS_AND_AURORA: ${{ github.event.inputs.run_AWS_RDS_AND_AURORA || 'false' }}
+      DEFAULT_REGION_ID: ${{ github.event.inputs.region_id || 'aws-us-east-2' }}
     runs-on: ubuntu-22.04
     outputs:
       pgbench-compare-matrix: ${{ steps.pgbench-compare-matrix.outputs.matrix }}
@@ -147,23 +264,37 @@ jobs:
     - name: Generate matrix for pgbench benchmark
       id: pgbench-compare-matrix
       run: |
+        region_id_default=${{ env.DEFAULT_REGION_ID }}
+        runner_default='["self-hosted", "us-east-2", "x64"]'
+        runner_azure='["self-hosted", "eastus2", "x64"]'
+        image_default="369495373322.dkr.ecr.eu-central-1.amazonaws.com/build-tools:pinned"
         matrix='{
+          "pg_version" : [
+            16
+          ],
+          "region_id" : [
+            "'"$region_id_default"'"
+            ],
           "platform": [
-            "neon-captest-new",
-            "neon-captest-reuse",
+            "neonvm-captest-new",
+            "neonvm-captest-reuse",
             "neonvm-captest-new"
           ],
           "db_size": [ "10gb" ],
-          "include": [{ "platform": "neon-captest-freetier",         "db_size": "3gb"  },
-                      { "platform": "neon-captest-new",              "db_size": "50gb" },
-                      { "platform": "neonvm-captest-freetier",       "db_size": "3gb"  },
-                      { "platform": "neonvm-captest-new",            "db_size": "50gb" },
-                      { "platform": "neonvm-captest-sharding-reuse", "db_size": "50gb" }]
+          "runner": ['"$runner_default"'],
+          "image": [ "'"$image_default"'" ],
+          "include": [{ "pg_version": 16, "region_id": "'"$region_id_default"'", "platform": "neonvm-captest-freetier",       "db_size": "3gb" ,"runner": '"$runner_default"', "image": "'"$image_default"'" },
+                      { "pg_version": 16, "region_id": "'"$region_id_default"'", "platform": "neonvm-captest-new",            "db_size": "10gb","runner": '"$runner_default"', "image": "'"$image_default"'" },
+                      { "pg_version": 16, "region_id": "'"$region_id_default"'", "platform": "neonvm-captest-new",            "db_size": "50gb","runner": '"$runner_default"', "image": "'"$image_default"'" },
+                      { "pg_version": 16, "region_id": "azure-eastus2",          "platform": "neonvm-azure-captest-freetier", "db_size": "3gb" ,"runner": '"$runner_azure"',   "image": "neondatabase/build-tools:pinned" },
+                      { "pg_version": 16, "region_id": "azure-eastus2",          "platform": "neonvm-azure-captest-new",      "db_size": "10gb","runner": '"$runner_azure"',   "image": "neondatabase/build-tools:pinned" },
+                      { "pg_version": 16, "region_id": "azure-eastus2",          "platform": "neonvm-azure-captest-new",      "db_size": "50gb","runner": '"$runner_azure"',   "image": "neondatabase/build-tools:pinned" },
+                      { "pg_version": 16, "region_id": "'"$region_id_default"'", "platform": "neonvm-captest-sharding-reuse", "db_size": "50gb","runner": '"$runner_default"', "image": "'"$image_default"'" }]
         }'
 
-        if [ "$(date +%A)" = "Saturday" ]; then
-          matrix=$(echo "$matrix" | jq '.include += [{ "platform": "rds-postgres", "db_size": "10gb"},
-                                                     { "platform": "rds-aurora",   "db_size": "50gb"}]')
+        if [ "$(date +%A)" = "Saturday" ] || [ ${RUN_AWS_RDS_AND_AURORA} = "true" ]; then
+          matrix=$(echo "$matrix" | jq '.include += [{ "pg_version": 16, "region_id": "'"$region_id_default"'", "platform": "rds-postgres", "db_size": "10gb","runner": '"$runner_default"', "image": "'"$image_default"'" },
+                                                     { "pg_version": 16, "region_id": "'"$region_id_default"'", "platform": "rds-aurora", "db_size": "10gb","runner": '"$runner_default"', "image": "'"$image_default"'" }]')
         fi
 
         echo "matrix=$(echo "$matrix" | jq --compact-output '.')" >> $GITHUB_OUTPUT
@@ -173,7 +304,7 @@ jobs:
       run: |
         matrix='{
           "platform": [
-            "neon-captest-reuse"
+            "neonvm-captest-reuse"
           ]
         }'
 
@@ -189,7 +320,7 @@ jobs:
       run: |
         matrix='{
           "platform": [
-            "neon-captest-reuse"
+            "neonvm-captest-reuse"
           ],
           "scale": [
             "10"
@@ -203,9 +334,17 @@ jobs:
 
         echo "matrix=$(echo "$matrix" | jq --compact-output '.')" >> $GITHUB_OUTPUT
 
+  prepare_AWS_RDS_databases:
+    uses: ./.github/workflows/_benchmarking_preparation.yml
+    secrets: inherit
+
   pgbench-compare:
     if: ${{ github.event.inputs.run_only_pgvector_tests == 'false' || github.event.inputs.run_only_pgvector_tests == null }}
-    needs: [ generate-matrices ]
+    needs: [ generate-matrices, prepare_AWS_RDS_databases ]
+    permissions:
+      contents: write
+      statuses: write
+      id-token: write # Required for OIDC authentication in azure runners
 
     strategy:
       fail-fast: false
@@ -215,15 +354,15 @@ jobs:
       TEST_PG_BENCH_DURATIONS_MATRIX: "60m"
       TEST_PG_BENCH_SCALES_MATRIX: ${{ matrix.db_size }}
       POSTGRES_DISTRIB_DIR: /tmp/neon/pg_install
-      DEFAULT_PG_VERSION: 14
+      DEFAULT_PG_VERSION: ${{ matrix.pg_version }}
       TEST_OUTPUT: /tmp/test_output
       BUILD_TYPE: remote
       SAVE_PERF_REPORT: ${{ github.event.inputs.save_perf_report || ( github.ref_name == 'main' ) }}
       PLATFORM: ${{ matrix.platform }}
 
-    runs-on: [ self-hosted, us-east-2, x64 ]
+    runs-on: ${{ matrix.runner }}
     container:
-      image: 369495373322.dkr.ecr.eu-central-1.amazonaws.com/build-tools:pinned
+      image: ${{ matrix.image }}
       options: --init
 
     # Increase timeout to 8h, default timeout is 6h
@@ -232,40 +371,41 @@ jobs:
     steps:
     - uses: actions/checkout@v4
 
+    - name: Configure AWS credentials # necessary on Azure runners
+      uses: aws-actions/configure-aws-credentials@v4
+      with:
+        aws-region: eu-central-1
+        role-to-assume: ${{ vars.DEV_AWS_OIDC_ROLE_ARN }}
+        role-duration-seconds: 18000 # 5 hours
+
     - name: Download Neon artifact
       uses: ./.github/actions/download
       with:
-        name: neon-${{ runner.os }}-release-artifact
+        name: neon-${{ runner.os }}-${{ runner.arch }}-release-artifact
         path: /tmp/neon/
         prefix: latest
 
-    - name: Add Postgres binaries to PATH
-      run: |
-        ${POSTGRES_DISTRIB_DIR}/v${DEFAULT_PG_VERSION}/bin/pgbench --version
-        echo "${POSTGRES_DISTRIB_DIR}/v${DEFAULT_PG_VERSION}/bin" >> $GITHUB_PATH
-
     - name: Create Neon Project
-      if: contains(fromJson('["neon-captest-new", "neon-captest-freetier", "neonvm-captest-new", "neonvm-captest-freetier"]'), matrix.platform)
+      if: contains(fromJson('["neonvm-captest-new", "neonvm-captest-freetier", "neonvm-azure-captest-freetier", "neonvm-azure-captest-new"]'), matrix.platform)
       id: create-neon-project
       uses: ./.github/actions/neon-project-create
       with:
-        region_id: ${{ github.event.inputs.region_id || 'aws-us-east-2' }}
+        region_id: ${{ matrix.region_id }}
         postgres_version: ${{ env.DEFAULT_PG_VERSION }}
         api_key: ${{ secrets.NEON_STAGING_API_KEY }}
-        compute_units: ${{ (matrix.platform == 'neon-captest-freetier' && '[0.25, 0.25]') || '[1, 1]' }}
-        provisioner: ${{ (contains(matrix.platform, 'neonvm-') && 'k8s-neonvm') || 'k8s-pod' }}
+        compute_units: ${{ (contains(matrix.platform, 'captest-freetier') && '[0.25, 0.25]') || '[1, 1]' }}
 
     - name: Set up Connection String
       id: set-up-connstr
       run: |
         case "${PLATFORM}" in
-          neon-captest-reuse)
+          neonvm-captest-reuse)
             CONNSTR=${{ secrets.BENCHMARK_CAPTEST_CONNSTR }}
             ;;
           neonvm-captest-sharding-reuse)
             CONNSTR=${{ secrets.BENCHMARK_CAPTEST_SHARDING_CONNSTR }}
             ;;
-          neon-captest-new | neon-captest-freetier | neonvm-captest-new | neonvm-captest-freetier)
+          neonvm-captest-new | neonvm-captest-freetier | neonvm-azure-captest-new | neonvm-azure-captest-freetier)
             CONNSTR=${{ steps.create-neon-project.outputs.dsn }}
             ;;
           rds-aurora)
@@ -282,16 +422,6 @@ jobs:
 
         echo "connstr=${CONNSTR}" >> $GITHUB_OUTPUT
 
-        QUERIES=("SELECT version()")
-        if [[ "${PLATFORM}" = "neon"* ]]; then
-          QUERIES+=("SHOW neon.tenant_id")
-          QUERIES+=("SHOW neon.timeline_id")
-        fi
-
-        for q in "${QUERIES[@]}"; do
-          psql ${CONNSTR} -c "${q}"
-        done
-
     - name: Benchmark init
       uses: ./.github/actions/run-python-test-set
       with:
@@ -300,6 +430,7 @@ jobs:
         run_in_parallel: false
         save_perf_report: ${{ env.SAVE_PERF_REPORT }}
         extra_params: -m remote_cluster --timeout 21600 -k test_pgbench_remote_init
+        pg_version: ${{ env.DEFAULT_PG_VERSION }}
       env:
         BENCHMARK_CONNSTR: ${{ steps.set-up-connstr.outputs.connstr }}
         VIP_VAP_ACCESS_TOKEN: "${{ secrets.VIP_VAP_ACCESS_TOKEN }}"
@@ -313,6 +444,7 @@ jobs:
         run_in_parallel: false
         save_perf_report: ${{ env.SAVE_PERF_REPORT }}
         extra_params: -m remote_cluster --timeout 21600 -k test_pgbench_remote_simple_update
+        pg_version: ${{ env.DEFAULT_PG_VERSION }}
       env:
         BENCHMARK_CONNSTR: ${{ steps.set-up-connstr.outputs.connstr }}
         VIP_VAP_ACCESS_TOKEN: "${{ secrets.VIP_VAP_ACCESS_TOKEN }}"
@@ -326,6 +458,7 @@ jobs:
         run_in_parallel: false
         save_perf_report: ${{ env.SAVE_PERF_REPORT }}
         extra_params: -m remote_cluster --timeout 21600 -k test_pgbench_remote_select_only
+        pg_version: ${{ env.DEFAULT_PG_VERSION }}
       env:
         BENCHMARK_CONNSTR: ${{ steps.set-up-connstr.outputs.connstr }}
         VIP_VAP_ACCESS_TOKEN: "${{ secrets.VIP_VAP_ACCESS_TOKEN }}"
@@ -339,6 +472,7 @@ jobs:
         api_key: ${{ secrets.NEON_STAGING_API_KEY }}
 
     - name: Create Allure report
+      id: create-allure-report
       if: ${{ !cancelled() }}
       uses: ./.github/actions/allure-report-generate
 
@@ -347,11 +481,29 @@ jobs:
       uses: slackapi/slack-github-action@v1
       with:
         channel-id: "C033QLM5P7D" # dev-staging-stream
-        slack-message: "Periodic perf testing ${{ matrix.platform }}: ${{ job.status }}\n${{ github.server_url }}/${{ github.repository }}/actions/runs/${{ github.run_id }}"
+        slack-message: |
+          Periodic perf testing on ${{ matrix.platform }}: ${{ job.status }}
+          <${{ github.server_url }}/${{ github.repository }}/actions/runs/${{ github.run_id }}|GitHub Run>
+          <${{ steps.create-allure-report.outputs.report-url }}|Allure report>
       env:
         SLACK_BOT_TOKEN: ${{ secrets.SLACK_BOT_TOKEN }}
 
   pgbench-pgvector:
+    permissions:
+      contents: write
+      statuses: write
+      id-token: write # Required for OIDC authentication in azure runners
+    strategy:
+      fail-fast: false
+      matrix:
+        include:
+          - PLATFORM: "neonvm-captest-pgvector"
+            RUNNER: [ self-hosted, us-east-2, x64 ]
+            IMAGE: 369495373322.dkr.ecr.eu-central-1.amazonaws.com/build-tools:pinned
+          - PLATFORM: "azure-captest-pgvector"
+            RUNNER: [ self-hosted, eastus2, x64 ]
+            IMAGE: neondatabase/build-tools:pinned
+
     env:
       TEST_PG_BENCH_DURATIONS_MATRIX: "15m"
       TEST_PG_BENCH_SCALES_MATRIX: "1"
@@ -359,43 +511,60 @@ jobs:
       DEFAULT_PG_VERSION: 16
       TEST_OUTPUT: /tmp/test_output
       BUILD_TYPE: remote
+      LD_LIBRARY_PATH: /home/nonroot/pg/usr/lib/x86_64-linux-gnu
       SAVE_PERF_REPORT: ${{ github.event.inputs.save_perf_report || ( github.ref_name == 'main' ) }}
-      PLATFORM: "neon-captest-pgvector"
+      PLATFORM: ${{ matrix.PLATFORM }}
 
-    runs-on: [ self-hosted, us-east-2, x64 ]
+    runs-on: ${{ matrix.RUNNER }}
     container:
-      image: 369495373322.dkr.ecr.eu-central-1.amazonaws.com/build-tools:pinned
+      image: ${{ matrix.IMAGE }}
       options: --init
 
     steps:
     - uses: actions/checkout@v4
 
-    - name: Download Neon artifact
-      uses: ./.github/actions/download
-      with:
-        name: neon-${{ runner.os }}-release-artifact
-        path: /tmp/neon/
-        prefix: latest
-
-    - name: Add Postgres binaries to PATH
+    # until https://github.com/neondatabase/neon/issues/8275 is fixed we temporarily install postgresql-16
+    # instead of using Neon artifacts containing pgbench
+    - name: Install postgresql-16 where pytest expects it
       run: |
-        ${POSTGRES_DISTRIB_DIR}/v${DEFAULT_PG_VERSION}/bin/pgbench --version
-        echo "${POSTGRES_DISTRIB_DIR}/v${DEFAULT_PG_VERSION}/bin" >> $GITHUB_PATH
+        cd /home/nonroot
+        wget -q https://apt.postgresql.org/pub/repos/apt/pool/main/p/postgresql-16/libpq5_16.4-1.pgdg110%2B1_amd64.deb
+        wget -q https://apt.postgresql.org/pub/repos/apt/pool/main/p/postgresql-16/postgresql-client-16_16.4-1.pgdg110%2B1_amd64.deb
+        wget -q https://apt.postgresql.org/pub/repos/apt/pool/main/p/postgresql-16/postgresql-16_16.4-1.pgdg110%2B1_amd64.deb 
+        dpkg -x libpq5_16.4-1.pgdg110+1_amd64.deb pg
+        dpkg -x postgresql-client-16_16.4-1.pgdg110+1_amd64.deb pg
+        dpkg -x postgresql-16_16.4-1.pgdg110+1_amd64.deb pg
+        mkdir -p /tmp/neon/pg_install/v16/bin
+        ln -s /home/nonroot/pg/usr/lib/postgresql/16/bin/pgbench /tmp/neon/pg_install/v16/bin/pgbench  
+        ln -s /home/nonroot/pg/usr/lib/postgresql/16/bin/psql /tmp/neon/pg_install/v16/bin/psql  
+        ln -s /home/nonroot/pg/usr/lib/x86_64-linux-gnu /tmp/neon/pg_install/v16/lib 
+        /tmp/neon/pg_install/v16/bin/pgbench --version
+        /tmp/neon/pg_install/v16/bin/psql --version
 
     - name: Set up Connection String
       id: set-up-connstr
       run: |
-        CONNSTR=${{ secrets.BENCHMARK_PGVECTOR_CONNSTR }}
-        
+        case "${PLATFORM}" in
+          neonvm-captest-pgvector)
+            CONNSTR=${{ secrets.BENCHMARK_PGVECTOR_CONNSTR }}
+            ;;
+          azure-captest-pgvector)
+            CONNSTR=${{ secrets.BENCHMARK_PGVECTOR_CONNSTR_AZURE }}
+            ;;
+          *)
+            echo >&2 "Unknown PLATFORM=${PLATFORM}"
+            exit 1
+            ;;
+        esac
+
         echo "connstr=${CONNSTR}" >> $GITHUB_OUTPUT
 
-        QUERIES=("SELECT version()")
-        QUERIES+=("SHOW neon.tenant_id")
-        QUERIES+=("SHOW neon.timeline_id")
-        
-        for q in "${QUERIES[@]}"; do
-          psql ${CONNSTR} -c "${q}"
-        done
+    - name: Configure AWS credentials # necessary on Azure runners to read/write from/to S3
+      uses: aws-actions/configure-aws-credentials@v4
+      with:
+        aws-region: eu-central-1
+        role-to-assume: ${{ vars.DEV_AWS_OIDC_ROLE_ARN }}
+        role-duration-seconds: 18000 # 5 hours
 
     - name: Benchmark pgvector hnsw indexing
       uses: ./.github/actions/run-python-test-set
@@ -405,6 +574,7 @@ jobs:
         run_in_parallel: false
         save_perf_report: ${{ env.SAVE_PERF_REPORT }}
         extra_params: -m remote_cluster --timeout 21600 -k test_pgvector_indexing
+        pg_version: ${{ env.DEFAULT_PG_VERSION }}
       env:
         VIP_VAP_ACCESS_TOKEN: "${{ secrets.VIP_VAP_ACCESS_TOKEN }}"
         PERF_TEST_RESULT_CONNSTR: "${{ secrets.PERF_TEST_RESULT_CONNSTR }}"
@@ -417,13 +587,15 @@ jobs:
         test_selection: performance/test_perf_pgvector_queries.py
         run_in_parallel: false
         save_perf_report: ${{ env.SAVE_PERF_REPORT }}
-        extra_params: -m remote_cluster --timeout 21600 
+        extra_params: -m remote_cluster --timeout 21600
+        pg_version: ${{ env.DEFAULT_PG_VERSION }}
       env:
         BENCHMARK_CONNSTR: ${{ steps.set-up-connstr.outputs.connstr }}
         VIP_VAP_ACCESS_TOKEN: "${{ secrets.VIP_VAP_ACCESS_TOKEN }}"
         PERF_TEST_RESULT_CONNSTR: "${{ secrets.PERF_TEST_RESULT_CONNSTR }}"
-    
+
     - name: Create Allure report
+      id: create-allure-report
       if: ${{ !cancelled() }}
       uses: ./.github/actions/allure-report-generate
 
@@ -432,11 +604,13 @@ jobs:
       uses: slackapi/slack-github-action@v1
       with:
         channel-id: "C033QLM5P7D" # dev-staging-stream
-        slack-message: "Periodic perf testing neon-captest-pgvector: ${{ job.status }}\n${{ github.server_url }}/${{ github.repository }}/actions/runs/${{ github.run_id }}"
+        slack-message: |
+          Periodic perf testing on ${{ env.PLATFORM }}: ${{ job.status }}
+          <${{ github.server_url }}/${{ github.repository }}/actions/runs/${{ github.run_id }}|GitHub Run>
+          <${{ steps.create-allure-report.outputs.report-url }}|Allure report>
       env:
         SLACK_BOT_TOKEN: ${{ secrets.SLACK_BOT_TOKEN }}
 
-
   clickbench-compare:
     # ClichBench DB for rds-aurora and rds-Postgres deployed to the same clusters
     # we use for performance testing in pgbench-compare.
@@ -446,7 +620,7 @@ jobs:
     # *_CLICKBENCH_CONNSTR: Genuine ClickBench DB with ~100M rows
     # *_CLICKBENCH_10M_CONNSTR: DB with the first 10M rows of ClickBench DB
     if: ${{ !cancelled() && (github.event.inputs.run_only_pgvector_tests == 'false' || github.event.inputs.run_only_pgvector_tests == null) }}
-    needs: [ generate-matrices, pgbench-compare ]
+    needs: [ generate-matrices, pgbench-compare, prepare_AWS_RDS_databases ]
 
     strategy:
       fail-fast: false
@@ -454,7 +628,7 @@ jobs:
 
     env:
       POSTGRES_DISTRIB_DIR: /tmp/neon/pg_install
-      DEFAULT_PG_VERSION: 14
+      DEFAULT_PG_VERSION: 16
       TEST_OUTPUT: /tmp/test_output
       TEST_OLAP_COLLECT_EXPLAIN: ${{ github.event.inputs.collect_olap_explain }}
       TEST_OLAP_COLLECT_PG_STAT_STATEMENTS: ${{ github.event.inputs.collect_pg_stat_statements }}
@@ -473,20 +647,15 @@ jobs:
     - name: Download Neon artifact
       uses: ./.github/actions/download
       with:
-        name: neon-${{ runner.os }}-release-artifact
+        name: neon-${{ runner.os }}-${{ runner.arch }}-release-artifact
         path: /tmp/neon/
         prefix: latest
 
-    - name: Add Postgres binaries to PATH
-      run: |
-        ${POSTGRES_DISTRIB_DIR}/v${DEFAULT_PG_VERSION}/bin/pgbench --version
-        echo "${POSTGRES_DISTRIB_DIR}/v${DEFAULT_PG_VERSION}/bin" >> $GITHUB_PATH
-
     - name: Set up Connection String
       id: set-up-connstr
       run: |
         case "${PLATFORM}" in
-          neon-captest-reuse)
+          neonvm-captest-reuse)
             CONNSTR=${{ secrets.BENCHMARK_CAPTEST_CLICKBENCH_10M_CONNSTR }}
             ;;
           rds-aurora)
@@ -496,23 +665,13 @@ jobs:
             CONNSTR=${{ secrets.BENCHMARK_RDS_POSTGRES_CLICKBENCH_10M_CONNSTR }}
             ;;
           *)
-            echo >&2 "Unknown PLATFORM=${PLATFORM}. Allowed only 'neon-captest-reuse', 'rds-aurora', or 'rds-postgres'"
+            echo >&2 "Unknown PLATFORM=${PLATFORM}. Allowed only 'neonvm-captest-reuse', 'rds-aurora', or 'rds-postgres'"
             exit 1
             ;;
         esac
 
         echo "connstr=${CONNSTR}" >> $GITHUB_OUTPUT
 
-        QUERIES=("SELECT version()")
-        if [[ "${PLATFORM}" = "neon"* ]]; then
-          QUERIES+=("SHOW neon.tenant_id")
-          QUERIES+=("SHOW neon.timeline_id")
-        fi
-
-        for q in "${QUERIES[@]}"; do
-          psql ${CONNSTR} -c "${q}"
-        done
-
     - name: ClickBench benchmark
       uses: ./.github/actions/run-python-test-set
       with:
@@ -521,6 +680,7 @@ jobs:
         run_in_parallel: false
         save_perf_report: ${{ env.SAVE_PERF_REPORT }}
         extra_params: -m remote_cluster --timeout 21600 -k test_clickbench
+        pg_version: ${{ env.DEFAULT_PG_VERSION }}
       env:
         VIP_VAP_ACCESS_TOKEN: "${{ secrets.VIP_VAP_ACCESS_TOKEN }}"
         PERF_TEST_RESULT_CONNSTR: "${{ secrets.PERF_TEST_RESULT_CONNSTR }}"
@@ -530,6 +690,7 @@ jobs:
         TEST_OLAP_SCALE: 10
 
     - name: Create Allure report
+      id: create-allure-report
       if: ${{ !cancelled() }}
       uses: ./.github/actions/allure-report-generate
 
@@ -538,7 +699,10 @@ jobs:
       uses: slackapi/slack-github-action@v1
       with:
         channel-id: "C033QLM5P7D" # dev-staging-stream
-        slack-message: "Periodic OLAP perf testing ${{ matrix.platform }}: ${{ job.status }}\n${{ github.server_url }}/${{ github.repository }}/actions/runs/${{ github.run_id }}"
+        slack-message: |
+          Periodic OLAP perf testing on ${{ matrix.platform }}: ${{ job.status }}
+          <${{ github.server_url }}/${{ github.repository }}/actions/runs/${{ github.run_id }}|GitHub Run>
+          <${{ steps.create-allure-report.outputs.report-url }}|Allure report>
       env:
         SLACK_BOT_TOKEN: ${{ secrets.SLACK_BOT_TOKEN }}
 
@@ -550,7 +714,7 @@ jobs:
     #
     # *_TPCH_S10_CONNSTR: DB generated with scale factor 10 (~10 GB)
     if: ${{ !cancelled() && (github.event.inputs.run_only_pgvector_tests == 'false' || github.event.inputs.run_only_pgvector_tests == null) }}
-    needs: [ generate-matrices, clickbench-compare ]
+    needs: [ generate-matrices, clickbench-compare, prepare_AWS_RDS_databases ]
 
     strategy:
       fail-fast: false
@@ -558,7 +722,7 @@ jobs:
 
     env:
       POSTGRES_DISTRIB_DIR: /tmp/neon/pg_install
-      DEFAULT_PG_VERSION: 14
+      DEFAULT_PG_VERSION: 16
       TEST_OUTPUT: /tmp/test_output
       BUILD_TYPE: remote
       SAVE_PERF_REPORT: ${{ github.event.inputs.save_perf_report || ( github.ref_name == 'main' ) }}
@@ -576,29 +740,24 @@ jobs:
     - name: Download Neon artifact
       uses: ./.github/actions/download
       with:
-        name: neon-${{ runner.os }}-release-artifact
+        name: neon-${{ runner.os }}-${{ runner.arch }}-release-artifact
         path: /tmp/neon/
         prefix: latest
 
-    - name: Add Postgres binaries to PATH
-      run: |
-        ${POSTGRES_DISTRIB_DIR}/v${DEFAULT_PG_VERSION}/bin/pgbench --version
-        echo "${POSTGRES_DISTRIB_DIR}/v${DEFAULT_PG_VERSION}/bin" >> $GITHUB_PATH
-
     - name: Get Connstring Secret Name
       run: |
         case "${PLATFORM}" in
-          neon-captest-reuse)
+          neonvm-captest-reuse)
             ENV_PLATFORM=CAPTEST_TPCH
             ;;
           rds-aurora)
             ENV_PLATFORM=RDS_AURORA_TPCH
             ;;
           rds-postgres)
-            ENV_PLATFORM=RDS_AURORA_TPCH
+            ENV_PLATFORM=RDS_POSTGRES_TPCH
             ;;
           *)
-            echo >&2 "Unknown PLATFORM=${PLATFORM}. Allowed only 'neon-captest-reuse', 'rds-aurora', or 'rds-postgres'"
+            echo >&2 "Unknown PLATFORM=${PLATFORM}. Allowed only 'neonvm-captest-reuse', 'rds-aurora', or 'rds-postgres'"
             exit 1
             ;;
         esac
@@ -613,16 +772,6 @@ jobs:
 
         echo "connstr=${CONNSTR}" >> $GITHUB_OUTPUT
 
-        QUERIES=("SELECT version()")
-        if [[ "${PLATFORM}" = "neon"* ]]; then
-          QUERIES+=("SHOW neon.tenant_id")
-          QUERIES+=("SHOW neon.timeline_id")
-        fi
-
-        for q in "${QUERIES[@]}"; do
-          psql ${CONNSTR} -c "${q}"
-        done
-
     - name: Run TPC-H benchmark
       uses: ./.github/actions/run-python-test-set
       with:
@@ -631,6 +780,7 @@ jobs:
         run_in_parallel: false
         save_perf_report: ${{ env.SAVE_PERF_REPORT }}
         extra_params: -m remote_cluster --timeout 21600 -k test_tpch
+        pg_version: ${{ env.DEFAULT_PG_VERSION }}
       env:
         VIP_VAP_ACCESS_TOKEN: "${{ secrets.VIP_VAP_ACCESS_TOKEN }}"
         PERF_TEST_RESULT_CONNSTR: "${{ secrets.PERF_TEST_RESULT_CONNSTR }}"
@@ -638,6 +788,7 @@ jobs:
         TEST_OLAP_SCALE: ${{ matrix.scale }}
 
     - name: Create Allure report
+      id: create-allure-report
       if: ${{ !cancelled() }}
       uses: ./.github/actions/allure-report-generate
 
@@ -646,13 +797,16 @@ jobs:
       uses: slackapi/slack-github-action@v1
       with:
         channel-id: "C033QLM5P7D" # dev-staging-stream
-        slack-message: "Periodic TPC-H perf testing ${{ matrix.platform }}: ${{ job.status }}\n${{ github.server_url }}/${{ github.repository }}/actions/runs/${{ github.run_id }}"
+        slack-message: |
+          Periodic TPC-H perf testing on ${{ matrix.platform }}: ${{ job.status }}
+          <${{ github.server_url }}/${{ github.repository }}/actions/runs/${{ github.run_id }}|GitHub Run>
+          <${{ steps.create-allure-report.outputs.report-url }}|Allure report>
       env:
         SLACK_BOT_TOKEN: ${{ secrets.SLACK_BOT_TOKEN }}
 
   user-examples-compare:
     if: ${{ !cancelled() && (github.event.inputs.run_only_pgvector_tests == 'false' || github.event.inputs.run_only_pgvector_tests == null) }}
-    needs: [ generate-matrices, tpch-compare ]
+    needs: [ generate-matrices, tpch-compare, prepare_AWS_RDS_databases ]
 
     strategy:
       fail-fast: false
@@ -660,7 +814,7 @@ jobs:
 
     env:
       POSTGRES_DISTRIB_DIR: /tmp/neon/pg_install
-      DEFAULT_PG_VERSION: 14
+      DEFAULT_PG_VERSION: 16
       TEST_OUTPUT: /tmp/test_output
       BUILD_TYPE: remote
       SAVE_PERF_REPORT: ${{ github.event.inputs.save_perf_report || ( github.ref_name == 'main' ) }}
@@ -677,20 +831,15 @@ jobs:
     - name: Download Neon artifact
       uses: ./.github/actions/download
       with:
-        name: neon-${{ runner.os }}-release-artifact
+        name: neon-${{ runner.os }}-${{ runner.arch }}-release-artifact
         path: /tmp/neon/
         prefix: latest
 
-    - name: Add Postgres binaries to PATH
-      run: |
-        ${POSTGRES_DISTRIB_DIR}/v${DEFAULT_PG_VERSION}/bin/pgbench --version
-        echo "${POSTGRES_DISTRIB_DIR}/v${DEFAULT_PG_VERSION}/bin" >> $GITHUB_PATH
-
     - name: Set up Connection String
       id: set-up-connstr
       run: |
         case "${PLATFORM}" in
-          neon-captest-reuse)
+          neonvm-captest-reuse)
             CONNSTR=${{ secrets.BENCHMARK_USER_EXAMPLE_CAPTEST_CONNSTR }}
             ;;
           rds-aurora)
@@ -700,23 +849,13 @@ jobs:
             CONNSTR=${{ secrets.BENCHMARK_USER_EXAMPLE_RDS_POSTGRES_CONNSTR }}
             ;;
           *)
-            echo >&2 "Unknown PLATFORM=${PLATFORM}. Allowed only 'neon-captest-reuse', 'rds-aurora', or 'rds-postgres'"
+            echo >&2 "Unknown PLATFORM=${PLATFORM}. Allowed only 'neonvm-captest-reuse', 'rds-aurora', or 'rds-postgres'"
             exit 1
             ;;
         esac
 
         echo "connstr=${CONNSTR}" >> $GITHUB_OUTPUT
 
-        QUERIES=("SELECT version()")
-        if [[ "${PLATFORM}" = "neon"* ]]; then
-          QUERIES+=("SHOW neon.tenant_id")
-          QUERIES+=("SHOW neon.timeline_id")
-        fi
-
-        for q in "${QUERIES[@]}"; do
-          psql ${CONNSTR} -c "${q}"
-        done
-
     - name: Run user examples
       uses: ./.github/actions/run-python-test-set
       with:
@@ -725,12 +864,14 @@ jobs:
         run_in_parallel: false
         save_perf_report: ${{ env.SAVE_PERF_REPORT }}
         extra_params: -m remote_cluster --timeout 21600 -k test_user_examples
+        pg_version: ${{ env.DEFAULT_PG_VERSION }}
       env:
         VIP_VAP_ACCESS_TOKEN: "${{ secrets.VIP_VAP_ACCESS_TOKEN }}"
         PERF_TEST_RESULT_CONNSTR: "${{ secrets.PERF_TEST_RESULT_CONNSTR }}"
         BENCHMARK_CONNSTR: ${{ steps.set-up-connstr.outputs.connstr }}
 
     - name: Create Allure report
+      id: create-allure-report
       if: ${{ !cancelled() }}
       uses: ./.github/actions/allure-report-generate
 
@@ -739,6 +880,10 @@ jobs:
       uses: slackapi/slack-github-action@v1
       with:
         channel-id: "C033QLM5P7D" # dev-staging-stream
-        slack-message: "Periodic User example perf testing ${{ matrix.platform }}: ${{ job.status }}\n${{ github.server_url }}/${{ github.repository }}/actions/runs/${{ github.run_id }}"
+        slack-message: |
+          Periodic TPC-H perf testing on ${{ matrix.platform }}: ${{ job.status }}
+          <${{ github.server_url }}/${{ github.repository }}/actions/runs/${{ github.run_id }}|GitHub Run>
+          <${{ steps.create-allure-report.outputs.report-url }}|Allure report>
+
       env:
         SLACK_BOT_TOKEN: ${{ secrets.SLACK_BOT_TOKEN }}

.github/workflows/build-build-tools-image.yml

@@ -38,7 +38,7 @@ jobs:
       matrix:
         arch: [ x64, arm64 ]
 
-    runs-on: ${{ fromJson(format('["self-hosted", "gen3", "{0}"]', matrix.arch == 'arm64' && 'large-arm64' || 'large')) }}
+    runs-on: ${{ fromJson(format('["self-hosted", "{0}"]', matrix.arch == 'arm64' && 'large-arm64' || 'large')) }}
 
     env:
       IMAGE_TAG: ${{ inputs.image-tag }}
@@ -56,35 +56,33 @@ jobs:
 
       - uses: actions/checkout@v4
 
-      # Use custom DOCKER_CONFIG directory to avoid conflicts with default settings
-      # The default value is ~/.docker
-      - name: Set custom docker config directory
-        run: |
-          mkdir -p /tmp/.docker-custom
-          echo DOCKER_CONFIG=/tmp/.docker-custom >> $GITHUB_ENV
+      - uses: ./.github/actions/set-docker-config-dir
+      - uses: docker/setup-buildx-action@v3
+        with:
+          cache-binary: false
 
-      - uses: docker/setup-buildx-action@v2
-
-      - uses: docker/login-action@v2
+      - uses: docker/login-action@v3
         with:
           username: ${{ secrets.NEON_DOCKERHUB_USERNAME }}
           password: ${{ secrets.NEON_DOCKERHUB_PASSWORD }}
 
-      - uses: docker/build-push-action@v4
+      - uses: docker/login-action@v3
+        with:
+          registry: cache.neon.build
+          username: ${{ secrets.NEON_CI_DOCKERCACHE_USERNAME }}
+          password: ${{ secrets.NEON_CI_DOCKERCACHE_PASSWORD }}
+
+      - uses: docker/build-push-action@v6
         with:
           context: .
           provenance: false
           push: true
           pull: true
           file: Dockerfile.build-tools
-          cache-from: type=registry,ref=neondatabase/build-tools:cache-${{ matrix.arch }}
-          cache-to: type=registry,ref=neondatabase/build-tools:cache-${{ matrix.arch }},mode=max
+          cache-from: type=registry,ref=cache.neon.build/build-tools:cache-${{ matrix.arch }}
+          cache-to: ${{ github.ref_name == 'main' && format('type=registry,ref=cache.neon.build/build-tools:cache-{0},mode=max', matrix.arch) || '' }}
           tags: neondatabase/build-tools:${{ inputs.image-tag }}-${{ matrix.arch }}
 
-      - name: Remove custom docker config directory
-        run: |
-          rm -rf /tmp/.docker-custom
-
   merge-images:
     needs: [ build-image ]
     runs-on: ubuntu-22.04

.github/workflows/build_and_test.yml

@@ -30,7 +30,7 @@ jobs:
     if: ${{ !contains(github.event.pull_request.labels.*.name, 'run-no-ci') }}
     uses: ./.github/workflows/check-permissions.yml
     with:
-      github-event-name: ${{ github.event_name}}
+      github-event-name: ${{ github.event_name }}
 
   cancel-previous-e2e-tests:
     needs: [ check-permissions ]
@@ -48,7 +48,7 @@ jobs:
 
   tag:
     needs: [ check-permissions ]
-    runs-on: [ self-hosted, gen3, small ]
+    runs-on: [ self-hosted, small ]
     container: 369495373322.dkr.ecr.eu-central-1.amazonaws.com/base:pinned
     outputs:
       build-tag: ${{steps.build-tag.outputs.tag}}
@@ -90,7 +90,7 @@ jobs:
 
   check-codestyle-python:
     needs: [ check-permissions, build-build-tools-image ]
-    runs-on: [ self-hosted, gen3, small ]
+    runs-on: [ self-hosted, small ]
     container:
       image: ${{ needs.build-build-tools-image.outputs.image }}
       credentials:
@@ -101,15 +101,12 @@ jobs:
     steps:
       - name: Checkout
         uses: actions/checkout@v4
-        with:
-          submodules: false
-          fetch-depth: 1
 
       - name: Cache poetry deps
         uses: actions/cache@v4
         with:
           path: ~/.cache/pypoetry/virtualenvs
-          key: v2-${{ runner.os }}-python-deps-${{ hashFiles('poetry.lock') }}
+          key: v2-${{ runner.os }}-${{ runner.arch }}-python-deps-${{ hashFiles('poetry.lock') }}
 
       - name: Install Python deps
         run: ./scripts/pysync
@@ -125,7 +122,11 @@ jobs:
 
   check-codestyle-rust:
     needs: [ check-permissions, build-build-tools-image ]
-    runs-on: [ self-hosted, gen3, small ]
+    strategy:
+      matrix:
+        arch: [ x64, arm64 ]
+    runs-on: ${{ fromJson(format('["self-hosted", "{0}"]', matrix.arch == 'arm64' && 'small-arm64' || 'small')) }}
+
     container:
       image: ${{ needs.build-build-tools-image.outputs.image }}
       credentials:
@@ -138,7 +139,6 @@ jobs:
         uses: actions/checkout@v4
         with:
           submodules: true
-          fetch-depth: 1
 
 #      Disabled for now
 #      - name: Restore cargo deps cache
@@ -149,7 +149,7 @@ jobs:
 #            !~/.cargo/registry/src
 #            ~/.cargo/git/
 #            target/
-#          key: v1-${{ runner.os }}-cargo-clippy-${{ hashFiles('rust-toolchain.toml') }}-${{ hashFiles('Cargo.lock') }}
+#          key: v1-${{ runner.os }}-${{ runner.arch }}-cargo-clippy-${{ hashFiles('rust-toolchain.toml') }}-${{ hashFiles('Cargo.lock') }}
 
       # Some of our rust modules use FFI and need those to be checked
       - name: Get postgres headers
@@ -193,295 +193,40 @@ jobs:
         if: ${{ !cancelled() }}
         run: cargo deny check --hide-inclusion-graph
 
-  build-neon:
-    needs: [ check-permissions, tag, build-build-tools-image ]
-    runs-on: [ self-hosted, gen3, large ]
-    container:
-      image: ${{ needs.build-build-tools-image.outputs.image }}
-      credentials:
-        username: ${{ secrets.NEON_DOCKERHUB_USERNAME }}
-        password: ${{ secrets.NEON_DOCKERHUB_PASSWORD }}
-      # Raise locked memory limit for tokio-epoll-uring.
-      # On 5.10 LTS kernels < 5.10.162 (and generally mainline kernels < 5.12),
-      # io_uring will account the memory of the CQ and SQ as locked.
-      # More details: https://github.com/neondatabase/neon/issues/6373#issuecomment-1905814391
-      options: --init --shm-size=512mb --ulimit memlock=67108864:67108864
+  build-and-test-locally:
+    needs: [ tag, build-build-tools-image ]
     strategy:
       fail-fast: false
       matrix:
-        build_type: [ debug, release ]
-    env:
-      BUILD_TYPE: ${{ matrix.build_type }}
-      GIT_VERSION: ${{ github.event.pull_request.head.sha || github.sha }}
-      BUILD_TAG: ${{ needs.tag.outputs.build-tag }}
-
-    steps:
-      - name: Fix git ownership
-        run: |
-          # Workaround for `fatal: detected dubious ownership in repository at ...`
-          #
-          # Use both ${{ github.workspace }} and ${GITHUB_WORKSPACE} because they're different on host and in containers
-          #   Ref https://github.com/actions/checkout/issues/785
-          #
-          git config --global --add safe.directory ${{ github.workspace }}
-          git config --global --add safe.directory ${GITHUB_WORKSPACE}
-          for r in 14 15 16; do
-            git config --global --add safe.directory "${{ github.workspace }}/vendor/postgres-v$r"
-            git config --global --add safe.directory "${GITHUB_WORKSPACE}/vendor/postgres-v$r"
-          done
-
-      - name: Checkout
-        uses: actions/checkout@v4
-        with:
-          submodules: true
-          fetch-depth: 1
-
-      - name: Set pg 14 revision for caching
-        id: pg_v14_rev
-        run: echo pg_rev=$(git rev-parse HEAD:vendor/postgres-v14) >> $GITHUB_OUTPUT
-
-      - name: Set pg 15 revision for caching
-        id: pg_v15_rev
-        run: echo pg_rev=$(git rev-parse HEAD:vendor/postgres-v15) >> $GITHUB_OUTPUT
-
-      - name: Set pg 16 revision for caching
-        id: pg_v16_rev
-        run: echo pg_rev=$(git rev-parse HEAD:vendor/postgres-v16) >> $GITHUB_OUTPUT
-
-      # Set some environment variables used by all the steps.
-      #
-      # CARGO_FLAGS is extra options to pass to "cargo build", "cargo test" etc.
-      #   It also includes --features, if any
-      #
-      # CARGO_FEATURES is passed to "cargo metadata". It is separate from CARGO_FLAGS,
-      #   because "cargo metadata" doesn't accept --release or --debug options
-      #
-      # We run tests with addtional features, that are turned off by default (e.g. in release builds), see
-      # corresponding Cargo.toml files for their descriptions.
-      - name: Set env variables
-        run: |
-          CARGO_FEATURES="--features testing"
-          if [[ $BUILD_TYPE == "debug" ]]; then
-            cov_prefix="scripts/coverage --profraw-prefix=$GITHUB_JOB --dir=/tmp/coverage run"
-            CARGO_FLAGS="--locked"
-          elif [[ $BUILD_TYPE == "release" ]]; then
-            cov_prefix=""
-            CARGO_FLAGS="--locked --release"
-          fi
-          {
-            echo "cov_prefix=${cov_prefix}"
-            echo "CARGO_FEATURES=${CARGO_FEATURES}"
-            echo "CARGO_FLAGS=${CARGO_FLAGS}"
-            echo "CARGO_HOME=${GITHUB_WORKSPACE}/.cargo"
-          } >> $GITHUB_ENV
-
-      # Disabled for now
-      # Don't include the ~/.cargo/registry/src directory. It contains just
-      # uncompressed versions of the crates in ~/.cargo/registry/cache
-      # directory, and it's faster to let 'cargo' to rebuild it from the
-      # compressed crates.
-#      - name: Cache cargo deps
-#        id: cache_cargo
-#        uses: actions/cache@v4
-#        with:
-#          path: |
-#            ~/.cargo/registry/
-#            !~/.cargo/registry/src
-#            ~/.cargo/git/
-#            target/
-#          # Fall back to older versions of the key, if no cache for current Cargo.lock was found
-#          key: |
-#            v1-${{ runner.os }}-${{ matrix.build_type }}-cargo-${{ hashFiles('rust-toolchain.toml') }}-${{ hashFiles('Cargo.lock') }}
-#            v1-${{ runner.os }}-${{ matrix.build_type }}-cargo-${{ hashFiles('rust-toolchain.toml') }}-
-
-      - name: Cache postgres v14 build
-        id: cache_pg_14
-        uses: actions/cache@v4
-        with:
-          path: pg_install/v14
-          key: v1-${{ runner.os }}-${{ matrix.build_type }}-pg-${{ steps.pg_v14_rev.outputs.pg_rev }}-${{ hashFiles('Makefile', 'Dockerfile.build-tools') }}
-
-      - name: Cache postgres v15 build
-        id: cache_pg_15
-        uses: actions/cache@v4
-        with:
-          path: pg_install/v15
-          key: v1-${{ runner.os }}-${{ matrix.build_type }}-pg-${{ steps.pg_v15_rev.outputs.pg_rev }}-${{ hashFiles('Makefile', 'Dockerfile.build-tools') }}
-
-      - name: Cache postgres v16 build
-        id: cache_pg_16
-        uses: actions/cache@v4
-        with:
-          path: pg_install/v16
-          key: v1-${{ runner.os }}-${{ matrix.build_type }}-pg-${{ steps.pg_v16_rev.outputs.pg_rev }}-${{ hashFiles('Makefile', 'Dockerfile.build-tools') }}
-
-      - name: Build postgres v14
-        if: steps.cache_pg_14.outputs.cache-hit != 'true'
-        run: mold -run make postgres-v14 -j$(nproc)
-
-      - name: Build postgres v15
-        if: steps.cache_pg_15.outputs.cache-hit != 'true'
-        run: mold -run make postgres-v15 -j$(nproc)
-
-      - name: Build postgres v16
-        if: steps.cache_pg_16.outputs.cache-hit != 'true'
-        run: mold -run make postgres-v16 -j$(nproc)
-
-      - name: Build neon extensions
-        run: mold -run make neon-pg-ext -j$(nproc)
-
-      - name: Build walproposer-lib
-        run: mold -run make walproposer-lib -j$(nproc)
-
-      - name: Run cargo build
-        run: |
-          ${cov_prefix} mold -run cargo build $CARGO_FLAGS $CARGO_FEATURES --bins --tests
-
-      # Do install *before* running rust tests because they might recompile the
-      # binaries with different features/flags.
-      - name: Install rust binaries
-        run: |
-          # Install target binaries
-          mkdir -p /tmp/neon/bin/
-          binaries=$(
-            ${cov_prefix} cargo metadata $CARGO_FEATURES --format-version=1 --no-deps |
-            jq -r '.packages[].targets[] | select(.kind | index("bin")) | .name'
-          )
-          for bin in $binaries; do
-            SRC=target/$BUILD_TYPE/$bin
-            DST=/tmp/neon/bin/$bin
-            cp "$SRC" "$DST"
-          done
-
-          # Install test executables and write list of all binaries (for code coverage)
-          if [[ $BUILD_TYPE == "debug" ]]; then
-            # Keep bloated coverage data files away from the rest of the artifact
-            mkdir -p /tmp/coverage/
-
-            mkdir -p /tmp/neon/test_bin/
-
-            test_exe_paths=$(
-              ${cov_prefix} cargo test $CARGO_FLAGS $CARGO_FEATURES --message-format=json --no-run |
-              jq -r '.executable | select(. != null)'
-            )
-            for bin in $test_exe_paths; do
-              SRC=$bin
-              DST=/tmp/neon/test_bin/$(basename $bin)
-
-              # We don't need debug symbols for code coverage, so strip them out to make
-              # the artifact smaller.
-              strip "$SRC" -o "$DST"
-              echo "$DST" >> /tmp/coverage/binaries.list
-            done
-
-            for bin in $binaries; do
-              echo "/tmp/neon/bin/$bin" >> /tmp/coverage/binaries.list
-            done
-          fi
-
-      - name: Run rust tests
-        env:
-          NEXTEST_RETRIES: 3
-        run: |
-          #nextest does not yet support running doctests
-          cargo test --doc $CARGO_FLAGS $CARGO_FEATURES
-
-          for io_engine in std-fs tokio-epoll-uring ; do
-            NEON_PAGESERVER_UNIT_TEST_VIRTUAL_FILE_IOENGINE=$io_engine ${cov_prefix} cargo nextest run $CARGO_FLAGS $CARGO_FEATURES
-          done
-
-          # Run separate tests for real S3
-          export ENABLE_REAL_S3_REMOTE_STORAGE=nonempty
-          export REMOTE_STORAGE_S3_BUCKET=neon-github-ci-tests
-          export REMOTE_STORAGE_S3_REGION=eu-central-1
-          ${cov_prefix} cargo nextest run $CARGO_FLAGS $CARGO_FEATURES -E 'package(remote_storage)' -E 'test(test_real_s3)'
-
-          # Run separate tests for real Azure Blob Storage
-          # XXX: replace region with `eu-central-1`-like region
-          export ENABLE_REAL_AZURE_REMOTE_STORAGE=y
-          export AZURE_STORAGE_ACCOUNT="${{ secrets.AZURE_STORAGE_ACCOUNT_DEV }}"
-          export AZURE_STORAGE_ACCESS_KEY="${{ secrets.AZURE_STORAGE_ACCESS_KEY_DEV }}"
-          export REMOTE_STORAGE_AZURE_CONTAINER="${{ vars.REMOTE_STORAGE_AZURE_CONTAINER }}"
-          export REMOTE_STORAGE_AZURE_REGION="${{ vars.REMOTE_STORAGE_AZURE_REGION }}"
-          ${cov_prefix} cargo nextest run $CARGO_FLAGS $CARGO_FEATURES -E 'package(remote_storage)' -E 'test(test_real_azure)'
-
-      - name: Install postgres binaries
-        run: cp -a pg_install /tmp/neon/pg_install
-
-      - name: Upload Neon artifact
-        uses: ./.github/actions/upload
-        with:
-          name: neon-${{ runner.os }}-${{ matrix.build_type }}-artifact
-          path: /tmp/neon
-
-      # XXX: keep this after the binaries.list is formed, so the coverage can properly work later
-      - name: Merge and upload coverage data
-        if: matrix.build_type == 'debug'
-        uses: ./.github/actions/save-coverage-data
-
-  regress-tests:
-    needs: [ check-permissions, build-neon, build-build-tools-image, tag ]
-    runs-on: [ self-hosted, gen3, large ]
-    container:
-      image: ${{ needs.build-build-tools-image.outputs.image }}
-      credentials:
-        username: ${{ secrets.NEON_DOCKERHUB_USERNAME }}
-        password: ${{ secrets.NEON_DOCKERHUB_PASSWORD }}
-      # for changed limits, see comments on `options:` earlier in this file
-      options: --init --shm-size=512mb --ulimit memlock=67108864:67108864
-    strategy:
-      fail-fast: false
-      matrix:
-        build_type: [ debug, release ]
-        pg_version: [ v14, v15, v16 ]
-    steps:
-      - name: Checkout
-        uses: actions/checkout@v4
-        with:
-          submodules: true
-          fetch-depth: 1
-
-      - name: Pytest regression tests
-        uses: ./.github/actions/run-python-test-set
-        timeout-minutes: 60
-        with:
-          build_type: ${{ matrix.build_type }}
-          test_selection: regress
-          needs_postgres_source: true
-          run_with_real_s3: true
-          real_s3_bucket: neon-github-ci-tests
-          real_s3_region: eu-central-1
-          rerun_flaky: true
-          pg_version: ${{ matrix.pg_version }}
-        env:
-          TEST_RESULT_CONNSTR: ${{ secrets.REGRESS_TEST_RESULT_CONNSTR_NEW }}
-          CHECK_ONDISK_DATA_COMPATIBILITY: nonempty
-          BUILD_TAG: ${{ needs.tag.outputs.build-tag }}
-          PAGESERVER_VIRTUAL_FILE_IO_ENGINE: tokio-epoll-uring
-          PAGESERVER_GET_VECTORED_IMPL: vectored
-          PAGESERVER_GET_IMPL: vectored
-          PAGESERVER_VALIDATE_VEC_GET: true
-
-      # Temporary disable this step until we figure out why it's so flaky
-      # Ref https://github.com/neondatabase/neon/issues/4540
-      - name: Merge and upload coverage data
-        if: |
-          false &&
-          matrix.build_type == 'debug' && matrix.pg_version == 'v14'
-        uses: ./.github/actions/save-coverage-data
+        arch: [ x64 ]
+        # Do not build or run tests in debug for release branches
+        build-type: ${{ fromJson((startsWith(github.ref_name, 'release') && github.event_name == 'push') && '["release"]' || '["debug", "release"]') }}
+        include:
+          - build-type: release
+            arch: arm64
+    uses: ./.github/workflows/_build-and-test-locally.yml
+    with:
+      arch: ${{ matrix.arch }}
+      build-tools-image: ${{ needs.build-build-tools-image.outputs.image }}
+      build-tag: ${{ needs.tag.outputs.build-tag }}
+      build-type: ${{ matrix.build-type }}
+      # Run tests on all Postgres versions in release builds and only on the latest version in debug builds
+      pg-versions: ${{ matrix.build-type == 'release' && '["v14", "v15", "v16"]' || '["v16"]' }}
+    secrets: inherit
 
+  # Keep `benchmarks` job outside of `build-and-test-locally` workflow to make job failures non-blocking
   get-benchmarks-durations:
+    if: github.ref_name == 'main' || contains(github.event.pull_request.labels.*.name, 'run-benchmarks')
     outputs:
       json: ${{ steps.get-benchmark-durations.outputs.json }}
     needs: [ check-permissions, build-build-tools-image ]
-    runs-on: [ self-hosted, gen3, small ]
+    runs-on: [ self-hosted, small ]
     container:
       image: ${{ needs.build-build-tools-image.outputs.image }}
       credentials:
         username: ${{ secrets.NEON_DOCKERHUB_USERNAME }}
         password: ${{ secrets.NEON_DOCKERHUB_PASSWORD }}
       options: --init
-    if: github.ref_name == 'main' || contains(github.event.pull_request.labels.*.name, 'run-benchmarks')
     steps:
       - name: Checkout
         uses: actions/checkout@v4
@@ -490,7 +235,7 @@ jobs:
         uses: actions/cache@v4
         with:
           path: ~/.cache/pypoetry/virtualenvs
-          key: v1-${{ runner.os }}-python-deps-${{ hashFiles('poetry.lock') }}
+          key: v1-${{ runner.os }}-${{ runner.arch }}-python-deps-${{ hashFiles('poetry.lock') }}
 
       - name: Install Python deps
         run: ./scripts/pysync
@@ -506,8 +251,9 @@ jobs:
           echo "json=$(jq --compact-output '.' /tmp/benchmark_durations.json)" >> $GITHUB_OUTPUT
 
   benchmarks:
-    needs: [ check-permissions, build-neon, build-build-tools-image, get-benchmarks-durations ]
-    runs-on: [ self-hosted, gen3, small ]
+    if: github.ref_name == 'main' || contains(github.event.pull_request.labels.*.name, 'run-benchmarks')
+    needs: [ check-permissions, build-and-test-locally, build-build-tools-image, get-benchmarks-durations ]
+    runs-on: [ self-hosted, small ]
     container:
       image: ${{ needs.build-build-tools-image.outputs.image }}
       credentials:
@@ -515,7 +261,6 @@ jobs:
         password: ${{ secrets.NEON_DOCKERHUB_PASSWORD }}
       # for changed limits, see comments on `options:` earlier in this file
       options: --init --shm-size=512mb --ulimit memlock=67108864:67108864
-    if: github.ref_name == 'main' || contains(github.event.pull_request.labels.*.name, 'run-benchmarks')
     strategy:
       fail-fast: false
       matrix:
@@ -535,14 +280,12 @@ jobs:
           save_perf_report: ${{ github.ref_name == 'main' }}
           extra_params: --splits 5 --group ${{ matrix.pytest_split_group }}
           benchmark_durations: ${{ needs.get-benchmarks-durations.outputs.json }}
+          pg_version: v16
         env:
           VIP_VAP_ACCESS_TOKEN: "${{ secrets.VIP_VAP_ACCESS_TOKEN }}"
           PERF_TEST_RESULT_CONNSTR: "${{ secrets.PERF_TEST_RESULT_CONNSTR }}"
           TEST_RESULT_CONNSTR: "${{ secrets.REGRESS_TEST_RESULT_CONNSTR_NEW }}"
           PAGESERVER_VIRTUAL_FILE_IO_ENGINE: tokio-epoll-uring
-          PAGESERVER_GET_VECTORED_IMPL: vectored
-          PAGESERVER_GET_IMPL: vectored
-          PAGESERVER_VALIDATE_VEC_GET: false
       # XXX: no coverage data handling here, since benchmarks are run on release builds,
       # while coverage is currently collected for the debug ones
 
@@ -556,19 +299,18 @@ jobs:
       with:
         channel-id: C060CNA47S9 # on-call-staging-storage-stream
         slack-message: |
-          Benchmarks failed on main: ${{ github.event.head_commit.url }}
-
-          Allure report: ${{ needs.create-test-report.outputs.report-url }}
+          Benchmarks failed on main <${{ github.event.head_commit.url }}|${{ github.sha }}>
+          <${{ needs.create-test-report.outputs.report-url }}|Allure report>
       env:
         SLACK_BOT_TOKEN: ${{ secrets.SLACK_BOT_TOKEN }}
 
   create-test-report:
-    needs: [ check-permissions, regress-tests, coverage-report, benchmarks, build-build-tools-image ]
+    needs: [ check-permissions, build-and-test-locally, coverage-report, build-build-tools-image, benchmarks ]
     if: ${{ !cancelled() && contains(fromJSON('["skipped", "success"]'), needs.check-permissions.result) }}
     outputs:
       report-url: ${{ steps.create-allure-report.outputs.report-url }}
 
-    runs-on: [ self-hosted, gen3, small ]
+    runs-on: [ self-hosted, small ]
     container:
       image: ${{ needs.build-build-tools-image.outputs.image }}
       credentials:
@@ -614,8 +356,8 @@ jobs:
             })
 
   coverage-report:
-    needs: [ check-permissions, regress-tests, build-build-tools-image ]
-    runs-on: [ self-hosted, gen3, small ]
+    needs: [ check-permissions, build-build-tools-image, build-and-test-locally ]
+    runs-on: [ self-hosted, small ]
     container:
       image: ${{ needs.build-build-tools-image.outputs.image }}
       credentials:
@@ -639,7 +381,7 @@ jobs:
       - name: Get Neon artifact
         uses: ./.github/actions/download
         with:
-          name: neon-${{ runner.os }}-${{ matrix.build_type }}-artifact
+          name: neon-${{ runner.os }}-${{ runner.arch }}-${{ matrix.build_type }}-artifact
           path: /tmp/neon
 
       - name: Get coverage artifact
@@ -729,7 +471,7 @@ jobs:
       matrix:
         arch: [ x64, arm64 ]
 
-    runs-on: ${{ fromJson(format('["self-hosted", "gen3", "{0}"]', matrix.arch == 'arm64' && 'large-arm64' || 'large')) }}
+    runs-on: ${{ fromJson(format('["self-hosted", "{0}"]', matrix.arch == 'arm64' && 'large-arm64' || 'large')) }}
 
     steps:
       - name: Checkout
@@ -738,23 +480,29 @@ jobs:
           submodules: true
           fetch-depth: 0
 
-      # Use custom DOCKER_CONFIG directory to avoid conflicts with default settings
-      # The default value is ~/.docker
-      - name: Set custom docker config directory
-        run: |
-          mkdir -p .docker-custom
-          echo DOCKER_CONFIG=$(pwd)/.docker-custom >> $GITHUB_ENV
-      - uses: docker/setup-buildx-action@v2
+      - uses: ./.github/actions/set-docker-config-dir
+      - uses: docker/setup-buildx-action@v3
+        with:
+          cache-binary: false
 
       - uses: docker/login-action@v3
         with:
           username: ${{ secrets.NEON_DOCKERHUB_USERNAME }}
           password: ${{ secrets.NEON_DOCKERHUB_PASSWORD }}
 
-      - uses: docker/build-push-action@v5
+      - uses: docker/login-action@v3
+        with:
+          registry: cache.neon.build
+          username: ${{ secrets.NEON_CI_DOCKERCACHE_USERNAME }}
+          password: ${{ secrets.NEON_CI_DOCKERCACHE_PASSWORD }}
+
+      - uses: docker/build-push-action@v6
         with:
           context: .
+          # ARM-specific flags are recommended for Graviton ≥ 2, these flags are also supported by Ampere Altra (Azure)
+          # https://github.com/aws/aws-graviton-getting-started/blob/57dc813626d0266f1cc12ef83474745bb1f31fb4/rust.md
           build-args: |
+            ADDITIONAL_RUSTFLAGS=${{ matrix.arch == 'arm64' && '-Ctarget-feature=+lse -Ctarget-cpu=neoverse-n1' || '' }}
             GIT_VERSION=${{ github.event.pull_request.head.sha || github.sha }}
             BUILD_TAG=${{ needs.tag.outputs.build-tag }}
             TAG=${{ needs.build-build-tools-image.outputs.image-tag }}
@@ -762,16 +510,11 @@ jobs:
           push: true
           pull: true
           file: Dockerfile
-          cache-from: type=registry,ref=neondatabase/neon:cache-${{ matrix.arch }}
-          cache-to: type=registry,ref=neondatabase/neon:cache-${{ matrix.arch }},mode=max
+          cache-from: type=registry,ref=cache.neon.build/neon:cache-${{ matrix.arch }}
+          cache-to: ${{ github.ref_name == 'main' && format('type=registry,ref=cache.neon.build/neon:cache-{0},mode=max', matrix.arch) || '' }}
           tags: |
             neondatabase/neon:${{ needs.tag.outputs.build-tag }}-${{ matrix.arch }}
 
-      - name: Remove custom docker config directory
-        if: always()
-        run: |
-          rm -rf .docker-custom
-
   neon-image:
     needs: [ neon-image-arch, tag ]
     runs-on: ubuntu-22.04
@@ -807,7 +550,7 @@ jobs:
         version: [ v14, v15, v16 ]
         arch: [ x64, arm64 ]
 
-    runs-on: ${{ fromJson(format('["self-hosted", "gen3", "{0}"]', matrix.arch == 'arm64' && 'large-arm64' || 'large')) }}
+    runs-on: ${{ fromJson(format('["self-hosted", "{0}"]', matrix.arch == 'arm64' && 'large-arm64' || 'large')) }}
 
     steps:
       - name: Checkout
@@ -816,17 +559,13 @@ jobs:
           submodules: true
           fetch-depth: 0
 
-      # Use custom DOCKER_CONFIG directory to avoid conflicts with default settings
-      # The default value is ~/.docker
-      - name: Set custom docker config directory
-        run: |
-          mkdir -p .docker-custom
-          echo DOCKER_CONFIG=$(pwd)/.docker-custom >> $GITHUB_ENV
-      - uses: docker/setup-buildx-action@v2
+      - uses: ./.github/actions/set-docker-config-dir
+      - uses: docker/setup-buildx-action@v3
         with:
+          cache-binary: false
           # Disable parallelism for docker buildkit.
           # As we already build everything with `make -j$(nproc)`, running it in additional level of parallelisam blows up the Runner.
-          config-inline: |
+          buildkitd-config-inline: |
             [worker.oci]
               max-parallelism = 1
 
@@ -841,8 +580,14 @@ jobs:
           username: ${{ secrets.AWS_ACCESS_KEY_DEV }}
           password: ${{ secrets.AWS_SECRET_KEY_DEV }}
 
+      - uses: docker/login-action@v3
+        with:
+          registry: cache.neon.build
+          username: ${{ secrets.NEON_CI_DOCKERCACHE_USERNAME }}
+          password: ${{ secrets.NEON_CI_DOCKERCACHE_PASSWORD }}
+
       - name: Build compute-node image
-        uses: docker/build-push-action@v5
+        uses: docker/build-push-action@v6
         with:
           context: .
           build-args: |
@@ -854,14 +599,14 @@ jobs:
           push: true
           pull: true
           file: Dockerfile.compute-node
-          cache-from: type=registry,ref=neondatabase/compute-node-${{ matrix.version }}:cache-${{ matrix.arch }}
-          cache-to: type=registry,ref=neondatabase/compute-node-${{ matrix.version }}:cache-${{ matrix.arch }},mode=max
+          cache-from: type=registry,ref=cache.neon.build/compute-node-${{ matrix.version }}:cache-${{ matrix.arch }}
+          cache-to: ${{ github.ref_name == 'main' && format('type=registry,ref=cache.neon.build/compute-node-{0}:cache-{1},mode=max', matrix.version, matrix.arch) || '' }}
           tags: |
             neondatabase/compute-node-${{ matrix.version }}:${{ needs.tag.outputs.build-tag }}-${{ matrix.arch }}
 
       - name: Build neon extensions test image
         if: matrix.version == 'v16'
-        uses: docker/build-push-action@v5
+        uses: docker/build-push-action@v6
         with:
           context: .
           build-args: |
@@ -874,15 +619,15 @@ jobs:
           pull: true
           file: Dockerfile.compute-node
           target: neon-pg-ext-test
-          cache-from: type=registry,ref=neondatabase/neon-test-extensions-${{ matrix.version }}:cache-${{ matrix.arch }}
-          cache-to: type=registry,ref=neondatabase/neon-test-extensions-${{ matrix.version }}:cache-${{ matrix.arch }},mode=max
+          cache-from: type=registry,ref=cache.neon.build/neon-test-extensions-${{ matrix.version }}:cache-${{ matrix.arch }}
+          cache-to: ${{ github.ref_name == 'main' && format('type=registry,ref=cache.neon.build/neon-test-extensions-{0}:cache-{1},mode=max', matrix.version, matrix.arch) || '' }}
           tags: |
             neondatabase/neon-test-extensions-${{ matrix.version }}:${{needs.tag.outputs.build-tag}}-${{ matrix.arch }}
 
       - name: Build compute-tools image
         # compute-tools are Postgres independent, so build it only once
         if: matrix.version == 'v16'
-        uses: docker/build-push-action@v5
+        uses: docker/build-push-action@v6
         with:
           target: compute-tools-image
           context: .
@@ -897,11 +642,6 @@ jobs:
           tags: |
             neondatabase/compute-tools:${{ needs.tag.outputs.build-tag }}-${{ matrix.arch }}
 
-      - name: Remove custom docker config directory
-        if: always()
-        run: |
-          rm -rf .docker-custom
-
   compute-node-image:
     needs: [ compute-node-image-arch, tag ]
     runs-on: ubuntu-22.04
@@ -955,7 +695,7 @@ jobs:
 
   vm-compute-node-image:
     needs: [ check-permissions, tag, compute-node-image ]
-    runs-on: [ self-hosted, gen3, large ]
+    runs-on: [ self-hosted, large ]
     strategy:
       fail-fast: false
       matrix:
@@ -974,13 +714,7 @@ jobs:
           curl -fL https://github.com/neondatabase/autoscaling/releases/download/$VM_BUILDER_VERSION/vm-builder -o vm-builder
           chmod +x vm-builder
 
-      # Use custom DOCKER_CONFIG directory to avoid conflicts with default settings
-      # The default value is ~/.docker
-      - name: Set custom docker config directory
-        run: |
-          mkdir -p .docker-custom
-          echo DOCKER_CONFIG=$(pwd)/.docker-custom >> $GITHUB_ENV
-
+      - uses: ./.github/actions/set-docker-config-dir
       - uses: docker/login-action@v3
         with:
           username: ${{ secrets.NEON_DOCKERHUB_USERNAME }}
@@ -1003,11 +737,6 @@ jobs:
         run: |
           docker push neondatabase/vm-compute-node-${{ matrix.version }}:${{ needs.tag.outputs.build-tag }}
 
-      - name: Remove custom docker config directory
-        if: always()
-        run: |
-          rm -rf .docker-custom
-
   test-images:
     needs: [ check-permissions, tag, neon-image, compute-node-image ]
     strategy:
@@ -1015,7 +744,7 @@ jobs:
       matrix:
         arch: [ x64, arm64 ]
 
-    runs-on: ${{ fromJson(format('["self-hosted", "gen3", "{0}"]', matrix.arch == 'arm64' && 'small-arm64' || 'small')) }}
+    runs-on: ${{ fromJson(format('["self-hosted", "{0}"]', matrix.arch == 'arm64' && 'small-arm64' || 'small')) }}
 
     steps:
       - name: Checkout
@@ -1023,6 +752,12 @@ jobs:
         with:
           fetch-depth: 0
 
+      - uses: ./.github/actions/set-docker-config-dir
+      - uses: docker/login-action@v3
+        with:
+          username: ${{ secrets.NEON_DOCKERHUB_USERNAME }}
+          password: ${{ secrets.NEON_DOCKERHUB_PASSWORD }}
+
       # `neondatabase/neon` contains multiple binaries, all of them use the same input for the version into the same version formatting library.
       # Pick pageserver as currently the only binary with extra "version" features printed in the string to verify.
       # Regular pageserver version string looks like
@@ -1058,6 +793,9 @@ jobs:
           docker compose -f ./docker-compose/docker-compose.yml down
 
   promote-images:
+    permissions:
+      contents: read  # This is required for actions/checkout
+      id-token: write # This is required for Azure Login to work.
     needs: [ check-permissions, tag, test-images, vm-compute-node-image ]
     runs-on: ubuntu-22.04
 
@@ -1070,7 +808,8 @@ jobs:
           username: ${{ secrets.NEON_DOCKERHUB_USERNAME }}
           password: ${{ secrets.NEON_DOCKERHUB_PASSWORD }}
 
-      - uses: docker/login-action@v3
+      - name: Login to dev ECR
+        uses: docker/login-action@v3
         with:
           registry: 369495373322.dkr.ecr.eu-central-1.amazonaws.com
           username: ${{ secrets.AWS_ACCESS_KEY_DEV }}
@@ -1083,6 +822,28 @@ jobs:
                                                neondatabase/vm-compute-node-${version}:${{ needs.tag.outputs.build-tag }}
           done
 
+      - name: Azure login
+        if: github.ref_name == 'main'
+        uses: azure/login@6c251865b4e6290e7b78be643ea2d005bc51f69a  # @v2.1.1
+        with:
+          client-id: ${{ secrets.AZURE_DEV_CLIENT_ID }}
+          tenant-id: ${{ secrets.AZURE_TENANT_ID }}
+          subscription-id: ${{ secrets.AZURE_DEV_SUBSCRIPTION_ID }}
+
+      - name: Login to ACR
+        if: github.ref_name == 'main'
+        run: |
+          az acr login --name=neoneastus2
+
+      - name: Copy docker images to ACR-dev
+        if: github.ref_name == 'main'
+        run: |
+          for image in neon compute-tools {vm-,}compute-node-{v14,v15,v16}; do
+            docker buildx imagetools create \
+              -t neoneastus2.azurecr.io/neondatabase/${image}:${{ needs.tag.outputs.build-tag }} \
+                                        neondatabase/${image}:${{ needs.tag.outputs.build-tag }}
+          done
+
       - name: Add latest tag to images
         if: github.ref_name == 'main'
         run: |
@@ -1104,6 +865,22 @@ jobs:
           docker buildx imagetools create -t neondatabase/neon-test-extensions-v16:latest \
                                              neondatabase/neon-test-extensions-v16:${{ needs.tag.outputs.build-tag }}
 
+      - name: Login to prod ECR
+        uses: docker/login-action@v3
+        if: github.ref_name == 'release'|| github.ref_name == 'release-proxy'
+        with:
+          registry: 093970136003.dkr.ecr.eu-central-1.amazonaws.com
+          username: ${{ secrets.PROD_GHA_RUNNER_LIMITED_AWS_ACCESS_KEY_ID }}
+          password: ${{ secrets.PROD_GHA_RUNNER_LIMITED_AWS_SECRET_ACCESS_KEY }}
+
+      - name: Copy all images to prod ECR
+        if: github.ref_name == 'release'|| github.ref_name == 'release-proxy'
+        run: |
+          for image in neon compute-tools {vm-,}compute-node-{v14,v15,v16}; do
+            docker buildx imagetools create -t 093970136003.dkr.ecr.eu-central-1.amazonaws.com/${image}:${{ needs.tag.outputs.build-tag }} \
+                                               369495373322.dkr.ecr.eu-central-1.amazonaws.com/${image}:${{ needs.tag.outputs.build-tag }}
+          done
+
   trigger-custom-extensions-build-and-wait:
     needs: [ check-permissions, tag ]
     runs-on: ubuntu-22.04
@@ -1179,10 +956,10 @@ jobs:
           exit 1
 
   deploy:
-    needs: [ check-permissions, promote-images, tag, regress-tests, trigger-custom-extensions-build-and-wait ]
+    needs: [ check-permissions, promote-images, tag, build-and-test-locally, trigger-custom-extensions-build-and-wait ]
     if: github.ref_name == 'main' || github.ref_name == 'release'|| github.ref_name == 'release-proxy'
 
-    runs-on: [ self-hosted, gen3, small ]
+    runs-on: [ self-hosted, small ]
     container: 369495373322.dkr.ecr.eu-central-1.amazonaws.com/ansible:latest
     steps:
       - name: Fix git ownership
@@ -1202,7 +979,6 @@ jobs:
       - name: Checkout
         uses: actions/checkout@v4
         with:
-          submodules: false
           fetch-depth: 0
 
       - name: Trigger deploy workflow
@@ -1210,9 +986,10 @@ jobs:
           GH_TOKEN: ${{ secrets.CI_ACCESS_TOKEN }}
         run: |
           if [[ "$GITHUB_REF_NAME" == "main" ]]; then
-            gh workflow --repo neondatabase/aws run deploy-dev.yml --ref main -f branch=main -f dockerTag=${{needs.tag.outputs.build-tag}} -f deployPreprodRegion=false
+            gh workflow --repo neondatabase/infra run deploy-dev.yml --ref main -f branch=main -f dockerTag=${{needs.tag.outputs.build-tag}} -f deployPreprodRegion=false
+            gh workflow --repo neondatabase/azure run deploy.yml -f dockerTag=${{needs.tag.outputs.build-tag}}
           elif [[ "$GITHUB_REF_NAME" == "release" ]]; then
-            gh workflow --repo neondatabase/aws run deploy-dev.yml --ref main \
+            gh workflow --repo neondatabase/infra run deploy-dev.yml --ref main \
               -f deployPgSniRouter=false \
               -f deployProxy=false \
               -f deployStorage=true \
@@ -1222,14 +999,14 @@ jobs:
               -f dockerTag=${{needs.tag.outputs.build-tag}} \
               -f deployPreprodRegion=true
 
-            gh workflow --repo neondatabase/aws run deploy-prod.yml --ref main \
+            gh workflow --repo neondatabase/infra run deploy-prod.yml --ref main \
               -f deployStorage=true \
               -f deployStorageBroker=true \
               -f deployStorageController=true \
               -f branch=main \
               -f dockerTag=${{needs.tag.outputs.build-tag}}
           elif [[ "$GITHUB_REF_NAME" == "release-proxy" ]]; then
-            gh workflow --repo neondatabase/aws run deploy-dev.yml --ref main \
+            gh workflow --repo neondatabase/infra run deploy-dev.yml --ref main \
               -f deployPgSniRouter=true \
               -f deployProxy=true \
               -f deployStorage=false \
@@ -1239,7 +1016,7 @@ jobs:
               -f dockerTag=${{needs.tag.outputs.build-tag}} \
               -f deployPreprodRegion=true
 
-            gh workflow --repo neondatabase/aws run deploy-proxy-prod.yml --ref main \
+            gh workflow --repo neondatabase/infra run deploy-proxy-prod.yml --ref main \
               -f deployPgSniRouter=true \
               -f deployProxy=true \
               -f branch=main \
@@ -1279,10 +1056,10 @@ jobs:
             })
 
   promote-compatibility-data:
-    needs: [ check-permissions, promote-images, tag, regress-tests ]
+    needs: [ check-permissions, promote-images, tag, build-and-test-locally ]
     if: github.ref_name == 'release'
 
-    runs-on: [ self-hosted, gen3, small ]
+    runs-on: [ self-hosted, small ]
     container:
       image: 369495373322.dkr.ecr.eu-central-1.amazonaws.com/base:pinned
       options: --init
@@ -1291,6 +1068,7 @@ jobs:
         env:
           BUCKET: neon-github-public-dev
           PREFIX: artifacts/latest
+          COMMIT_SHA: ${{ github.event.pull_request.head.sha || github.sha }}
         run: |
           # Update compatibility snapshot for the release
           for pg_version in v14 v15 v16; do
@@ -1304,8 +1082,8 @@ jobs:
 
           # Update Neon artifact for the release (reuse already uploaded artifact)
           for build_type in debug release; do
-            OLD_PREFIX=artifacts/${GITHUB_RUN_ID}
-            FILENAME=neon-${{ runner.os }}-${build_type}-artifact.tar.zst
+            OLD_PREFIX=artifacts/${COMMIT_SHA}/${GITHUB_RUN_ID}
+            FILENAME=neon-${{ runner.os }}-${{ runner.arch }}-${build_type}-artifact.tar.zst
 
             S3_KEY=$(aws s3api list-objects-v2 --bucket ${BUCKET} --prefix ${OLD_PREFIX} | jq -r '.Contents[]?.Key' | grep ${FILENAME} | sort --version-sort | tail -1 || true)
             if [ -z "${S3_KEY}" ]; then
@@ -1317,9 +1095,39 @@ jobs:
           done
 
   pin-build-tools-image:
-    needs: [ build-build-tools-image, promote-images, regress-tests ]
+    needs: [ build-build-tools-image, promote-images, build-and-test-locally ]
     if: github.ref_name == 'main'
     uses: ./.github/workflows/pin-build-tools-image.yml
     with:
       from-tag: ${{ needs.build-build-tools-image.outputs.image-tag }}
     secrets: inherit
+
+  # This job simplifies setting branch protection rules (in GitHub UI)
+  # by allowing to set only this job instead of listing many others.
+  # It also makes it easier to rename or parametrise jobs (using matrix)
+  # which requires changes in branch protection rules
+  #
+  # Note, that we can't add external check (like `neon-cloud-e2e`) we still need to use GitHub UI for that.
+  #
+  # https://github.com/neondatabase/neon/settings/branch_protection_rules
+  conclusion:
+    if: always()
+    # Format `needs` differently to make the list more readable.
+    # Usually we do `needs: [...]`
+    needs:
+      - build-and-test-locally
+      - check-codestyle-python
+      - check-codestyle-rust
+      - promote-images
+      - test-images
+      - trigger-custom-extensions-build-and-wait
+    runs-on: ubuntu-22.04
+    steps:
+      # The list of possible results:
+      # https://docs.github.com/en/actions/learn-github-actions/contexts#needs-context
+      - name: Fail the job if any of the dependencies do not succeed
+        run: exit 1
+        if: |
+          contains(needs.*.result, 'failure')
+          || contains(needs.*.result, 'cancelled')
+          || contains(needs.*.result, 'skipped')

.github/workflows/label-for-external-users.yml

@@ -0,0 +1,54 @@
+name: Add `external` label to issues and PRs created by external users
+
+on:
+  issues:
+    types:
+      - opened
+  pull_request_target:
+    types:
+      - opened
+
+# No permission for GITHUB_TOKEN by default; the **minimal required** set of permissions should be granted in each job.
+permissions: {}
+
+env:
+  LABEL: external
+
+jobs:
+  check-user:
+    runs-on: ubuntu-22.04
+
+    outputs:
+      is-member: ${{ steps.check-user.outputs.is-member }}
+
+    steps:
+    - name: Check whether `${{ github.actor }}` is a member of `${{ github.repository_owner }}`
+      id: check-user
+      env:
+        GH_TOKEN: ${{ secrets.CI_ACCESS_TOKEN }}
+      run: |
+        if gh api -H "Accept: application/vnd.github+json" -H "X-GitHub-Api-Version: 2022-11-28" "/orgs/${GITHUB_REPOSITORY_OWNER}/members/${GITHUB_ACTOR}"; then
+          is_member=true
+        else
+          is_member=false
+        fi
+
+        echo "is-member=${is_member}" | tee -a ${GITHUB_OUTPUT}
+
+  add-label:
+    if: needs.check-user.outputs.is-member == 'false'
+    needs: [ check-user ]
+
+    runs-on: ubuntu-22.04
+    permissions:
+      pull-requests: write # for `gh pr edit`
+      issues: write        # for `gh issue edit`
+
+    steps:
+    - name: Add `${{ env.LABEL }}` label
+      env:
+        GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+        ITEM_NUMBER: ${{ github.event[github.event_name == 'pull_request_target' && 'pull_request' || 'issue'].number }}
+        GH_CLI_COMMAND: ${{ github.event_name == 'pull_request_target' && 'pr' || 'issue' }}
+      run: |
+        gh ${GH_CLI_COMMAND} --repo ${GITHUB_REPOSITORY} edit --add-label=${LABEL} ${ITEM_NUMBER}

.github/workflows/neon_extra_builds.yml

@@ -56,7 +56,6 @@ jobs:
         uses: actions/checkout@v4
         with:
           submodules: true
-          fetch-depth: 1
 
       - name: Install macOS postgres dependencies
         run: brew install flex bison openssl protobuf icu4c pkg-config
@@ -133,214 +132,6 @@ jobs:
       - name: Check that no warnings are produced
         run: ./run_clippy.sh
 
-  check-linux-arm-build:
-    needs: [ check-permissions, build-build-tools-image ]
-    timeout-minutes: 90
-    runs-on: [ self-hosted, small-arm64 ]
-
-    env:
-      # Use release build only, to have less debug info around
-      # Hence keeping target/ (and general cache size) smaller
-      BUILD_TYPE: release
-      CARGO_FEATURES: --features testing
-      CARGO_FLAGS: --release
-      AWS_ACCESS_KEY_ID: ${{ secrets.AWS_ACCESS_KEY_DEV }}
-      AWS_SECRET_ACCESS_KEY: ${{ secrets.AWS_SECRET_KEY_DEV }}
-
-    container:
-      image: ${{ needs.build-build-tools-image.outputs.image }}
-      credentials:
-        username: ${{ secrets.NEON_DOCKERHUB_USERNAME }}
-        password: ${{ secrets.NEON_DOCKERHUB_PASSWORD }}
-      options: --init
-
-    steps:
-      - name: Fix git ownership
-        run: |
-          # Workaround for `fatal: detected dubious ownership in repository at ...`
-          #
-          # Use both ${{ github.workspace }} and ${GITHUB_WORKSPACE} because they're different on host and in containers
-          #   Ref https://github.com/actions/checkout/issues/785
-          #
-          git config --global --add safe.directory ${{ github.workspace }}
-          git config --global --add safe.directory ${GITHUB_WORKSPACE}
-          for r in 14 15 16; do
-            git config --global --add safe.directory "${{ github.workspace }}/vendor/postgres-v$r"
-            git config --global --add safe.directory "${GITHUB_WORKSPACE}/vendor/postgres-v$r"
-          done
-
-      - name: Checkout
-        uses: actions/checkout@v4
-        with:
-          submodules: true
-          fetch-depth: 1
-
-      - name: Set pg 14 revision for caching
-        id: pg_v14_rev
-        run: echo pg_rev=$(git rev-parse HEAD:vendor/postgres-v14) >> $GITHUB_OUTPUT
-
-      - name: Set pg 15 revision for caching
-        id: pg_v15_rev
-        run: echo pg_rev=$(git rev-parse HEAD:vendor/postgres-v15) >> $GITHUB_OUTPUT
-
-      - name: Set pg 16 revision for caching
-        id: pg_v16_rev
-        run: echo pg_rev=$(git rev-parse HEAD:vendor/postgres-v16) >> $GITHUB_OUTPUT
-
-      - name: Set env variables
-        run: |
-          echo "CARGO_HOME=${GITHUB_WORKSPACE}/.cargo" >> $GITHUB_ENV
-
-      - name: Cache postgres v14 build
-        id: cache_pg_14
-        uses: actions/cache@v4
-        with:
-          path: pg_install/v14
-          key: v1-${{ runner.os }}-${{ runner.arch }}-${{ env.BUILD_TYPE }}-pg-${{ steps.pg_v14_rev.outputs.pg_rev }}-${{ hashFiles('Makefile') }}
-
-      - name: Cache postgres v15 build
-        id: cache_pg_15
-        uses: actions/cache@v4
-        with:
-          path: pg_install/v15
-          key: v1-${{ runner.os }}-${{ runner.arch }}-${{ env.BUILD_TYPE }}-pg-${{ steps.pg_v15_rev.outputs.pg_rev }}-${{ hashFiles('Makefile') }}
-
-      - name: Cache postgres v16 build
-        id: cache_pg_16
-        uses: actions/cache@v4
-        with:
-          path: pg_install/v16
-          key: v1-${{ runner.os }}-${{ runner.arch }}-${{ env.BUILD_TYPE }}-pg-${{ steps.pg_v16_rev.outputs.pg_rev }}-${{ hashFiles('Makefile') }}
-
-      - name: Build postgres v14
-        if: steps.cache_pg_14.outputs.cache-hit != 'true'
-        run: mold -run make postgres-v14 -j$(nproc)
-
-      - name: Build postgres v15
-        if: steps.cache_pg_15.outputs.cache-hit != 'true'
-        run: mold -run make postgres-v15 -j$(nproc)
-
-      - name: Build postgres v16
-        if: steps.cache_pg_16.outputs.cache-hit != 'true'
-        run: mold -run make postgres-v16 -j$(nproc)
-
-      - name: Build neon extensions
-        run: mold -run make neon-pg-ext -j$(nproc)
-
-      - name: Build walproposer-lib
-        run: mold -run make walproposer-lib -j$(nproc)
-
-      - name: Run cargo build
-        run: |
-          mold -run cargo build --locked $CARGO_FLAGS $CARGO_FEATURES --bins --tests -j$(nproc)
-
-      - name: Run cargo test
-        env:
-          NEXTEST_RETRIES: 3
-        run: |
-          cargo nextest run $CARGO_FEATURES -j$(nproc)
-
-          # Run separate tests for real S3
-          export ENABLE_REAL_S3_REMOTE_STORAGE=nonempty
-          export REMOTE_STORAGE_S3_BUCKET=neon-github-ci-tests
-          export REMOTE_STORAGE_S3_REGION=eu-central-1
-          # Avoid `$CARGO_FEATURES` since there's no `testing` feature in the e2e tests now
-          cargo nextest run --package remote_storage --test test_real_s3 -j$(nproc)
-
-          # Run separate tests for real Azure Blob Storage
-          # XXX: replace region with `eu-central-1`-like region
-          export ENABLE_REAL_AZURE_REMOTE_STORAGE=y
-          export AZURE_STORAGE_ACCOUNT="${{ secrets.AZURE_STORAGE_ACCOUNT_DEV }}"
-          export AZURE_STORAGE_ACCESS_KEY="${{ secrets.AZURE_STORAGE_ACCESS_KEY_DEV }}"
-          export REMOTE_STORAGE_AZURE_CONTAINER="${{ vars.REMOTE_STORAGE_AZURE_CONTAINER }}"
-          export REMOTE_STORAGE_AZURE_REGION="${{ vars.REMOTE_STORAGE_AZURE_REGION }}"
-          # Avoid `$CARGO_FEATURES` since there's no `testing` feature in the e2e tests now
-          cargo nextest run --package remote_storage --test test_real_azure -j$(nproc)
-
-  check-codestyle-rust-arm:
-    needs: [ check-permissions, build-build-tools-image ]
-    timeout-minutes: 90
-    runs-on: [ self-hosted, small-arm64 ]
-
-    container:
-      image: ${{ needs.build-build-tools-image.outputs.image }}
-      credentials:
-        username: ${{ secrets.NEON_DOCKERHUB_USERNAME }}
-        password: ${{ secrets.NEON_DOCKERHUB_PASSWORD }}
-      options: --init
-
-    strategy:
-      fail-fast: false
-      matrix:
-        build_type: [ debug, release ]
-
-    steps:
-      - name: Fix git ownership
-        run: |
-          # Workaround for `fatal: detected dubious ownership in repository at ...`
-          #
-          # Use both ${{ github.workspace }} and ${GITHUB_WORKSPACE} because they're different on host and in containers
-          #   Ref https://github.com/actions/checkout/issues/785
-          #
-          git config --global --add safe.directory ${{ github.workspace }}
-          git config --global --add safe.directory ${GITHUB_WORKSPACE}
-          for r in 14 15 16; do
-            git config --global --add safe.directory "${{ github.workspace }}/vendor/postgres-v$r"
-            git config --global --add safe.directory "${GITHUB_WORKSPACE}/vendor/postgres-v$r"
-          done
-
-      - name: Checkout
-        uses: actions/checkout@v4
-        with:
-          submodules: true
-          fetch-depth: 1
-
-      # Some of our rust modules use FFI and need those to be checked
-      - name: Get postgres headers
-        run: make postgres-headers -j$(nproc)
-
-      # cargo hack runs the given cargo subcommand (clippy in this case) for all feature combinations.
-      # This will catch compiler & clippy warnings in all feature combinations.
-      # TODO: use cargo hack for build and test as well, but, that's quite expensive.
-      # NB: keep clippy args in sync with ./run_clippy.sh
-      - run: |
-          CLIPPY_COMMON_ARGS="$( source .neon_clippy_args; echo "$CLIPPY_COMMON_ARGS")"
-          if [ "$CLIPPY_COMMON_ARGS" = "" ]; then
-            echo "No clippy args found in .neon_clippy_args"
-            exit 1
-          fi
-          echo "CLIPPY_COMMON_ARGS=${CLIPPY_COMMON_ARGS}" >> $GITHUB_ENV
-
-      - name: Run cargo clippy (debug)
-        if: matrix.build_type == 'debug'
-        run: cargo hack --feature-powerset clippy $CLIPPY_COMMON_ARGS
-      - name: Run cargo clippy (release)
-        if: matrix.build_type == 'release'
-        run: cargo hack --feature-powerset clippy --release $CLIPPY_COMMON_ARGS
-
-      - name: Check documentation generation
-        if: matrix.build_type == 'release'
-        run: cargo doc --workspace --no-deps --document-private-items -j$(nproc)
-        env:
-            RUSTDOCFLAGS: "-Dwarnings -Arustdoc::private_intra_doc_links"
-
-      # Use `${{ !cancelled() }}` to run quck tests after the longer clippy run
-      - name: Check formatting
-        if: ${{ !cancelled() && matrix.build_type == 'release' }}
-        run: cargo fmt --all -- --check
-
-      # https://github.com/facebookincubator/cargo-guppy/tree/bec4e0eb29dcd1faac70b1b5360267fc02bf830e/tools/cargo-hakari#2-keep-the-workspace-hack-up-to-date-in-ci
-      - name: Check rust dependencies
-        if: ${{ !cancelled() && matrix.build_type == 'release' }}
-        run: |
-          cargo hakari generate --diff  # workspace-hack Cargo.toml is up-to-date
-          cargo hakari manage-deps --dry-run  # all workspace crates depend on workspace-hack
-
-      # https://github.com/EmbarkStudios/cargo-deny
-      - name: Check rust licenses/bans/advisories/sources
-        if: ${{ !cancelled() && matrix.build_type == 'release' }}
-        run: cargo deny check
-
   gather-rust-build-stats:
     needs: [ check-permissions, build-build-tools-image ]
     if: |
@@ -357,8 +148,6 @@ jobs:
 
     env:
       BUILD_TYPE: release
-      # remove the cachepot wrapper and build without crate caches
-      RUSTC_WRAPPER: ""
       # build with incremental compilation produce partial results
       # so do not attempt to cache this build, also disable the incremental compilation
       CARGO_INCREMENTAL: 0
@@ -368,7 +157,6 @@ jobs:
         uses: actions/checkout@v4
         with:
           submodules: true
-          fetch-depth: 1
 
       # Some of our rust modules use FFI and need those to be checked
       - name: Get postgres headers
@@ -378,7 +166,7 @@ jobs:
         run: make walproposer-lib -j$(nproc)
 
       - name: Produce the build stats
-        run: cargo build --all --release --timings -j$(nproc)
+        run: PQ_LIB_DIR=$(pwd)/pg_install/v16/lib cargo build --all --release --timings -j$(nproc)
 
       - name: Upload the build stats
         id: upload-stats

.github/workflows/periodic_pagebench.yml

@@ -0,0 +1,155 @@
+name: Periodic pagebench performance test on dedicated EC2 machine in eu-central-1 region
+
+on:
+  schedule:
+    # * is a special character in YAML so you have to quote this string
+    #          ┌───────────── minute (0 - 59)
+    #          │ ┌───────────── hour (0 - 23)
+    #          │ │ ┌───────────── day of the month (1 - 31)
+    #          │ │ │ ┌───────────── month (1 - 12 or JAN-DEC)
+    #          │ │ │ │ ┌───────────── day of the week (0 - 6 or SUN-SAT)
+    - cron:  '0 18 * * *' # Runs at 6 PM UTC every day
+  workflow_dispatch: # Allows manual triggering of the workflow
+    inputs:
+      commit_hash:
+        type: string
+        description: 'The long neon repo commit hash for the system under test (pageserver) to be tested.'
+        required: false
+        default: ''
+
+defaults:
+  run:
+    shell: bash -euo pipefail {0}
+
+concurrency:
+  group: ${{ github.workflow }}
+  cancel-in-progress: false
+
+jobs:
+  trigger_bench_on_ec2_machine_in_eu_central_1:
+    runs-on: [ self-hosted, small ]
+    container:
+      image: neondatabase/build-tools:pinned
+      credentials:
+        username: ${{ secrets.NEON_DOCKERHUB_USERNAME }}
+        password: ${{ secrets.NEON_DOCKERHUB_PASSWORD }}
+      options: --init
+    timeout-minutes: 360  # Set the timeout to 6 hours
+    env:
+      API_KEY: ${{ secrets.PERIODIC_PAGEBENCH_EC2_RUNNER_API_KEY }}
+      RUN_ID: ${{ github.run_id }}
+      AWS_ACCESS_KEY_ID: ${{ secrets.AWS_EC2_US_TEST_RUNNER_ACCESS_KEY_ID }}
+      AWS_SECRET_ACCESS_KEY : ${{ secrets.AWS_EC2_US_TEST_RUNNER_ACCESS_KEY_SECRET }}
+      AWS_DEFAULT_REGION : "eu-central-1"
+      AWS_INSTANCE_ID : "i-02a59a3bf86bc7e74"
+    steps:
+    # we don't need the neon source code because we run everything remotely
+    # however we still need the local github actions to run the allure step below
+    - uses: actions/checkout@v4
+
+    - name: Show my own (github runner) external IP address - usefull for IP allowlisting
+      run: curl https://ifconfig.me
+
+    - name: Start EC2 instance and wait for the instance to boot up
+      run: |
+        aws ec2 start-instances --instance-ids $AWS_INSTANCE_ID
+        aws ec2 wait instance-running --instance-ids $AWS_INSTANCE_ID
+        sleep 60 # sleep some time to allow cloudinit and our API server to start up
+
+    - name: Determine public IP of the EC2 instance and set env variable EC2_MACHINE_URL_US
+      run: |
+        public_ip=$(aws ec2 describe-instances --instance-ids $AWS_INSTANCE_ID --query 'Reservations[*].Instances[*].PublicIpAddress' --output text)
+        echo "Public IP of the EC2 instance: $public_ip"
+        echo "EC2_MACHINE_URL_US=https://${public_ip}:8443" >> $GITHUB_ENV
+
+    - name: Determine commit hash
+      env:
+        INPUT_COMMIT_HASH: ${{ github.event.inputs.commit_hash }}
+      run: |
+        if [ -z "$INPUT_COMMIT_HASH" ]; then
+          echo "COMMIT_HASH=$(curl -s https://api.github.com/repos/neondatabase/neon/commits/main | jq -r '.sha')" >> $GITHUB_ENV
+        else
+          echo "COMMIT_HASH=$INPUT_COMMIT_HASH" >> $GITHUB_ENV
+        fi
+
+    - name: Start Bench with run_id   
+      run: |
+        curl -k -X 'POST' \
+        "${EC2_MACHINE_URL_US}/start_test/${GITHUB_RUN_ID}" \
+        -H 'accept: application/json' \
+        -H 'Content-Type: application/json' \
+        -H "Authorization: Bearer $API_KEY" \
+        -d "{\"neonRepoCommitHash\": \"${COMMIT_HASH}\"}"
+
+    - name: Poll Test Status
+      id: poll_step
+      run: |
+        status=""
+        while [[ "$status" != "failure" && "$status" != "success" ]]; do
+          response=$(curl -k -X 'GET' \
+          "${EC2_MACHINE_URL_US}/test_status/${GITHUB_RUN_ID}" \
+          -H 'accept: application/json' \
+          -H "Authorization: Bearer $API_KEY")
+          echo "Response: $response"
+          set +x
+          status=$(echo $response | jq -r '.status')
+          echo "Test status: $status"
+          if [[ "$status" == "failure" ]]; then
+            echo "Test failed"
+            exit 1 # Fail the job step if status is failure
+          elif [[ "$status" == "success" || "$status" == "null" ]]; then
+            break
+          elif [[ "$status" == "too_many_runs" ]]; then
+            echo "Too many runs already running"
+            echo "too_many_runs=true" >> "$GITHUB_OUTPUT"
+            exit 1
+          fi
+
+          sleep 60 # Poll every 60 seconds
+        done
+
+    - name: Retrieve Test Logs
+      if: always() && steps.poll_step.outputs.too_many_runs != 'true'
+      run: |
+        curl -k -X 'GET' \
+        "${EC2_MACHINE_URL_US}/test_log/${GITHUB_RUN_ID}" \
+        -H 'accept: application/gzip' \
+        -H "Authorization: Bearer $API_KEY" \
+        --output "test_log_${GITHUB_RUN_ID}.gz"
+    
+    - name: Unzip Test Log and Print it into this job's log
+      if: always() && steps.poll_step.outputs.too_many_runs != 'true'
+      run: |
+        gzip -d "test_log_${GITHUB_RUN_ID}.gz"
+        cat "test_log_${GITHUB_RUN_ID}"
+
+    - name: Create Allure report
+      env:
+        AWS_ACCESS_KEY_ID: ${{ secrets.AWS_ACCESS_KEY_DEV }}
+        AWS_SECRET_ACCESS_KEY: ${{ secrets.AWS_SECRET_KEY_DEV }}
+      if: ${{ !cancelled() }}
+      uses: ./.github/actions/allure-report-generate
+
+    - name: Post to a Slack channel
+      if: ${{ github.event.schedule && failure() }}
+      uses: slackapi/slack-github-action@v1
+      with:
+        channel-id: "C033QLM5P7D" # dev-staging-stream
+        slack-message: "Periodic pagebench testing on dedicated hardware: ${{ job.status }}\n${{ github.server_url }}/${{ github.repository }}/actions/runs/${{ github.run_id }}"
+      env:
+        SLACK_BOT_TOKEN: ${{ secrets.SLACK_BOT_TOKEN }}
+
+    - name: Cleanup Test Resources
+      if: always() 
+      run: |
+        curl -k -X 'POST' \
+        "${EC2_MACHINE_URL_US}/cleanup_test/${GITHUB_RUN_ID}" \
+        -H 'accept: application/json' \
+        -H "Authorization: Bearer $API_KEY" \
+        -d ''
+
+    - name: Stop EC2 instance and wait for the instance to be stopped
+      if: always() && steps.poll_step.outputs.too_many_runs != 'true'
+      run: |
+        aws ec2 stop-instances --instance-ids $AWS_INSTANCE_ID
+        aws ec2 wait instance-stopped --instance-ids $AWS_INSTANCE_ID

.github/workflows/pg-clients.yml

@@ -0,0 +1,211 @@
+name: Test Postgres client libraries
+
+on:
+  schedule:
+    # * is a special character in YAML so you have to quote this string
+    #          ┌───────────── minute (0 - 59)
+    #          │ ┌───────────── hour (0 - 23)
+    #          │ │ ┌───────────── day of the month (1 - 31)
+    #          │ │ │ ┌───────────── month (1 - 12 or JAN-DEC)
+    #          │ │ │ │ ┌───────────── day of the week (0 - 6 or SUN-SAT)
+    - cron:  '23 02 * * *' # run once a day, timezone is utc
+  pull_request:
+    paths:
+      - '.github/workflows/pg-clients.yml'
+      - 'test_runner/pg_clients/**'
+      - 'test_runner/logical_repl/**'
+      - 'poetry.lock'
+  workflow_dispatch:
+
+concurrency:
+  group: ${{ github.workflow }}-${{ github.ref_name }}
+  cancel-in-progress: ${{ github.event_name == 'pull_request' }}
+
+defaults:
+  run:
+    shell: bash -euxo pipefail {0}
+
+env:
+  DEFAULT_PG_VERSION: 16
+  PLATFORM: neon-captest-new
+  AWS_ACCESS_KEY_ID: ${{ secrets.AWS_ACCESS_KEY_DEV }}
+  AWS_SECRET_ACCESS_KEY: ${{ secrets.AWS_SECRET_KEY_DEV }}
+  AWS_DEFAULT_REGION: eu-central-1
+
+jobs:
+  check-permissions:
+    if: ${{ !contains(github.event.pull_request.labels.*.name, 'run-no-ci') }}
+    uses: ./.github/workflows/check-permissions.yml
+    with:
+      github-event-name: ${{ github.event_name }}
+
+  check-build-tools-image:
+    needs: [ check-permissions ]
+    uses: ./.github/workflows/check-build-tools-image.yml
+
+  build-build-tools-image:
+    needs: [ check-build-tools-image ]
+    uses: ./.github/workflows/build-build-tools-image.yml
+    with:
+      image-tag: ${{ needs.check-build-tools-image.outputs.image-tag }}
+    secrets: inherit
+
+  test-logical-replication:
+    needs: [ build-build-tools-image ]
+    runs-on: ubuntu-22.04
+
+    container:
+      image: ${{ needs.build-build-tools-image.outputs.image }}
+      credentials:
+        username: ${{ secrets.NEON_DOCKERHUB_USERNAME }}
+        password: ${{ secrets.NEON_DOCKERHUB_PASSWORD }}
+      options: --init --user root
+    services:
+      clickhouse:
+        image: clickhouse/clickhouse-server:24.6.3.64
+        ports:
+          - 9000:9000
+          - 8123:8123
+      zookeeper:
+        image: quay.io/debezium/zookeeper:2.7
+        ports:
+          - 2181:2181
+      kafka:
+        image: quay.io/debezium/kafka:2.7
+        env:
+          ZOOKEEPER_CONNECT: "zookeeper:2181"
+          KAFKA_ADVERTISED_LISTENERS: PLAINTEXT://kafka:9092
+          KAFKA_BROKER_ID: 1
+          KAFKA_OFFSETS_TOPIC_REPLICATION_FACTOR: 1
+          KAFKA_JMX_PORT: 9991
+        ports:
+          - 9092:9092
+      debezium:
+        image: quay.io/debezium/connect:2.7
+        env:
+          BOOTSTRAP_SERVERS: kafka:9092
+          GROUP_ID: 1
+          CONFIG_STORAGE_TOPIC: debezium-config
+          OFFSET_STORAGE_TOPIC: debezium-offset
+          STATUS_STORAGE_TOPIC: debezium-status
+          DEBEZIUM_CONFIG_CONNECTOR_CLASS: io.debezium.connector.postgresql.PostgresConnector
+        ports:
+          - 8083:8083
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Download Neon artifact
+        uses: ./.github/actions/download
+        with:
+          name: neon-${{ runner.os }}-${{ runner.arch }}-release-artifact
+          path: /tmp/neon/
+          prefix: latest
+
+      - name: Create Neon Project
+        id: create-neon-project
+        uses: ./.github/actions/neon-project-create
+        with:
+          api_key: ${{ secrets.NEON_STAGING_API_KEY }}
+          postgres_version: ${{ env.DEFAULT_PG_VERSION }}
+
+      - name: Run tests
+        uses: ./.github/actions/run-python-test-set
+        with:
+          build_type: remote
+          test_selection: logical_repl
+          run_in_parallel: false
+          extra_params: -m remote_cluster
+          pg_version: ${{ env.DEFAULT_PG_VERSION }}
+        env:
+          BENCHMARK_CONNSTR: ${{ steps.create-neon-project.outputs.dsn }}
+
+      - name: Delete Neon Project
+        if: always()
+        uses: ./.github/actions/neon-project-delete
+        with:
+          project_id: ${{ steps.create-neon-project.outputs.project_id }}
+          api_key: ${{ secrets.NEON_STAGING_API_KEY }}
+
+      - name: Create Allure report
+        if: ${{ !cancelled() }}
+        id: create-allure-report
+        uses: ./.github/actions/allure-report-generate
+        with:
+          store-test-results-into-db: true
+        env:
+          REGRESS_TEST_RESULT_CONNSTR_NEW: ${{ secrets.REGRESS_TEST_RESULT_CONNSTR_NEW }}
+
+      - name: Post to a Slack channel
+        if: github.event.schedule && failure()
+        uses: slackapi/slack-github-action@v1
+        with:
+          channel-id: "C06KHQVQ7U3" # on-call-qa-staging-stream
+          slack-message: |
+            Testing the logical replication: <${{ github.server_url }}/${{ github.repository }}/actions/runs/${{ github.run_id }}|${{ job.status }}> (<${{ steps.create-allure-report.outputs.report-url }}|test report>)
+        env:
+          SLACK_BOT_TOKEN: ${{ secrets.SLACK_BOT_TOKEN }}
+
+  test-postgres-client-libs:
+    needs: [ build-build-tools-image ]
+    runs-on: ubuntu-22.04
+
+    container:
+      image: ${{ needs.build-build-tools-image.outputs.image }}
+      credentials:
+        username: ${{ secrets.NEON_DOCKERHUB_USERNAME }}
+        password: ${{ secrets.NEON_DOCKERHUB_PASSWORD }}
+      options: --init --user root
+
+    steps:
+    - uses: actions/checkout@v4
+
+    - name: Download Neon artifact
+      uses: ./.github/actions/download
+      with:
+        name: neon-${{ runner.os }}-${{ runner.arch }}-release-artifact
+        path: /tmp/neon/
+        prefix: latest
+
+    - name: Create Neon Project
+      id: create-neon-project
+      uses: ./.github/actions/neon-project-create
+      with:
+        api_key: ${{ secrets.NEON_STAGING_API_KEY }}
+        postgres_version: ${{ env.DEFAULT_PG_VERSION }}
+
+    - name: Run tests
+      uses: ./.github/actions/run-python-test-set
+      with:
+        build_type: remote
+        test_selection: pg_clients
+        run_in_parallel: false
+        extra_params: -m remote_cluster
+        pg_version: ${{ env.DEFAULT_PG_VERSION }}
+      env:
+        BENCHMARK_CONNSTR: ${{ steps.create-neon-project.outputs.dsn }}
+
+    - name: Delete Neon Project
+      if: always()
+      uses: ./.github/actions/neon-project-delete
+      with:
+        project_id: ${{ steps.create-neon-project.outputs.project_id }}
+        api_key: ${{ secrets.NEON_STAGING_API_KEY }}
+
+    - name: Create Allure report
+      if: ${{ !cancelled() }}
+      id: create-allure-report
+      uses: ./.github/actions/allure-report-generate
+      with:
+        store-test-results-into-db: true
+      env:
+        REGRESS_TEST_RESULT_CONNSTR_NEW: ${{ secrets.REGRESS_TEST_RESULT_CONNSTR_NEW }}
+
+    - name: Post to a Slack channel
+      if: github.event.schedule && failure()
+      uses: slackapi/slack-github-action@v1
+      with:
+        channel-id: "C06KHQVQ7U3" # on-call-qa-staging-stream
+        slack-message: |
+          Testing Postgres clients: <${{ github.server_url }}/${{ github.repository }}/actions/runs/${{ github.run_id }}|${{ job.status }}> (<${{ steps.create-allure-report.outputs.report-url }}|test report>)
+      env:
+        SLACK_BOT_TOKEN: ${{ secrets.SLACK_BOT_TOKEN }}

.github/workflows/pg_clients.yml

@@ -1,98 +0,0 @@
-name: Test Postgres client libraries
-
-on:
-  schedule:
-    # * is a special character in YAML so you have to quote this string
-    #          ┌───────────── minute (0 - 59)
-    #          │ ┌───────────── hour (0 - 23)
-    #          │ │ ┌───────────── day of the month (1 - 31)
-    #          │ │ │ ┌───────────── month (1 - 12 or JAN-DEC)
-    #          │ │ │ │ ┌───────────── day of the week (0 - 6 or SUN-SAT)
-    - cron:  '23 02 * * *' # run once a day, timezone is utc
-
-  workflow_dispatch:
-
-concurrency:
-  # Allow only one workflow per any non-`main` branch.
-  group: ${{ github.workflow }}-${{ github.ref_name }}-${{ github.ref_name == 'main' && github.sha || 'anysha' }}
-  cancel-in-progress: true
-
-jobs:
-  test-postgres-client-libs:
-    # TODO: switch to gen2 runner, requires docker
-    runs-on: ubuntu-22.04
-
-    env:
-      DEFAULT_PG_VERSION: 14
-      TEST_OUTPUT: /tmp/test_output
-
-    steps:
-    - name: Checkout
-      uses: actions/checkout@v4
-
-    - uses: actions/setup-python@v4
-      with:
-        python-version: 3.9
-
-    - name: Install Poetry
-      uses: snok/install-poetry@v1
-
-    - name: Cache poetry deps
-      uses: actions/cache@v4
-      with:
-        path: ~/.cache/pypoetry/virtualenvs
-        key: v2-${{ runner.os }}-python-deps-ubunutu-latest-${{ hashFiles('poetry.lock') }}
-
-    - name: Install Python deps
-      shell: bash -euxo pipefail {0}
-      run: ./scripts/pysync
-
-    - name: Create Neon Project
-      id: create-neon-project
-      uses: ./.github/actions/neon-project-create
-      with:
-        api_key: ${{ secrets.NEON_STAGING_API_KEY }}
-        postgres_version: ${{ env.DEFAULT_PG_VERSION }}
-
-    - name: Run pytest
-      env:
-        REMOTE_ENV: 1
-        BENCHMARK_CONNSTR: ${{ steps.create-neon-project.outputs.dsn }}
-        POSTGRES_DISTRIB_DIR: /tmp/neon/pg_install
-      shell: bash -euxo pipefail {0}
-      run: |
-        # Test framework expects we have psql binary;
-        # but since we don't really need it in this test, let's mock it
-        mkdir -p "$POSTGRES_DISTRIB_DIR/v${DEFAULT_PG_VERSION}/bin" && touch "$POSTGRES_DISTRIB_DIR/v${DEFAULT_PG_VERSION}/bin/psql";
-        ./scripts/pytest \
-          --junitxml=$TEST_OUTPUT/junit.xml \
-          --tb=short \
-          --verbose \
-          -m "remote_cluster" \
-          -rA "test_runner/pg_clients"
-
-    - name: Delete Neon Project
-      if: ${{ always() }}
-      uses: ./.github/actions/neon-project-delete
-      with:
-        project_id: ${{ steps.create-neon-project.outputs.project_id }}
-        api_key: ${{ secrets.NEON_STAGING_API_KEY }}
-
-    # We use GitHub's action upload-artifact because `ubuntu-latest` doesn't have configured AWS CLI.
-    # It will be fixed after switching to gen2 runner
-    - name: Upload python test logs
-      if: always()
-      uses: actions/upload-artifact@v4
-      with:
-        retention-days: 7
-        name: python-test-pg_clients-${{ runner.os }}-stage-logs
-        path: ${{ env.TEST_OUTPUT }}
-
-    - name: Post to a Slack channel
-      if: ${{ github.event.schedule && failure() }}
-      uses: slackapi/slack-github-action@v1
-      with:
-        channel-id: "C033QLM5P7D" # dev-staging-stream
-        slack-message: "Testing Postgres clients: ${{ job.status }}\n${{ github.server_url }}/${{ github.repository }}/actions/runs/${{ github.run_id }}"
-      env:
-        SLACK_BOT_TOKEN: ${{ secrets.SLACK_BOT_TOKEN }}

.github/workflows/pin-build-tools-image.yml

@@ -7,12 +7,20 @@ on:
         description: 'Source tag'
         required: true
         type: string
+      force:
+        description: 'Force the image to be pinned'
+        default: false
+        type: boolean
   workflow_call:
     inputs:
       from-tag:
         description: 'Source tag'
         required: true
         type: string
+      force:
+        description: 'Force the image to be pinned'
+        default: false
+        type: boolean
 
 defaults:
   run:
@@ -22,15 +30,18 @@ concurrency:
   group: pin-build-tools-image-${{ inputs.from-tag }}
   cancel-in-progress: false
 
+# No permission for GITHUB_TOKEN by default; the **minimal required** set of permissions should be granted in each job.
 permissions: {}
 
-jobs:
-  tag-image:
-    runs-on: ubuntu-22.04
+env:
+  FROM_TAG: ${{ inputs.from-tag }}
+  TO_TAG: pinned
 
-    env:
-      FROM_TAG: ${{ inputs.from-tag }}
-      TO_TAG: pinned
+jobs:
+  check-manifests:
+    runs-on: ubuntu-22.04
+    outputs:
+      skip: ${{ steps.check-manifests.outputs.skip }}
 
     steps:
       - name: Check if we really need to pin the image
@@ -47,27 +58,44 @@ jobs:
 
           echo "skip=${skip}" | tee -a $GITHUB_OUTPUT
 
+  tag-image:
+    needs: check-manifests
+
+    # use format(..) to catch both inputs.force = true AND inputs.force = 'true'
+    if: needs.check-manifests.outputs.skip == 'false' || format('{0}', inputs.force) == 'true'
+
+    runs-on: ubuntu-22.04
+
+    permissions:
+      id-token: write # for `azure/login`
+
+    steps:
       - uses: docker/login-action@v3
-        if: steps.check-manifests.outputs.skip == 'false'
+
         with:
           username: ${{ secrets.NEON_DOCKERHUB_USERNAME }}
           password: ${{ secrets.NEON_DOCKERHUB_PASSWORD }}
 
-      - name: Tag build-tools with `${{ env.TO_TAG }}` in Docker Hub
-        if: steps.check-manifests.outputs.skip == 'false'
-        run: |
-          docker buildx imagetools create -t neondatabase/build-tools:${TO_TAG} \
-                                             neondatabase/build-tools:${FROM_TAG}
-
       - uses: docker/login-action@v3
-        if: steps.check-manifests.outputs.skip == 'false'
         with:
           registry: 369495373322.dkr.ecr.eu-central-1.amazonaws.com
           username: ${{ secrets.AWS_ACCESS_KEY_DEV }}
           password: ${{ secrets.AWS_SECRET_KEY_DEV }}
 
-      - name: Tag build-tools with `${{ env.TO_TAG }}` in ECR
-        if: steps.check-manifests.outputs.skip == 'false'
+      - name: Azure login
+        uses: azure/login@6c251865b4e6290e7b78be643ea2d005bc51f69a  # @v2.1.1
+        with:
+          client-id: ${{ secrets.AZURE_DEV_CLIENT_ID }}
+          tenant-id: ${{ secrets.AZURE_TENANT_ID }}
+          subscription-id: ${{ secrets.AZURE_DEV_SUBSCRIPTION_ID }}
+
+      - name: Login to ACR
+        run: |
+          az acr login --name=neoneastus2
+
+      - name: Tag build-tools with `${{ env.TO_TAG }}` in Docker Hub, ECR, and ACR
         run: |
           docker buildx imagetools create -t 369495373322.dkr.ecr.eu-central-1.amazonaws.com/build-tools:${TO_TAG} \
+                                          -t neoneastus2.azurecr.io/neondatabase/build-tools:${TO_TAG} \
+                                          -t neondatabase/build-tools:${TO_TAG} \
                                              neondatabase/build-tools:${FROM_TAG}

.github/workflows/release.yml

@@ -52,13 +52,15 @@ jobs:
       env:
         GH_TOKEN: ${{ secrets.CI_ACCESS_TOKEN }}
       run: |
+        TITLE="Storage & Compute release ${RELEASE_DATE}"
+
         cat << EOF > body.md
-          ## Storage & Compute release ${RELEASE_DATE}
+          ## ${TITLE}
 
           **Please merge this Pull Request using 'Create a merge commit' button**
         EOF
 
-        gh pr create --title "Release ${RELEASE_DATE}" \
+        gh pr create --title "${TITLE}" \
                      --body-file "body.md" \
                      --head "${RELEASE_BRANCH}" \
                      --base "release"
@@ -91,13 +93,15 @@ jobs:
       env:
         GH_TOKEN: ${{ secrets.CI_ACCESS_TOKEN }}
       run: |
+        TITLE="Proxy release ${RELEASE_DATE}"
+
         cat << EOF > body.md
-          ## Proxy release ${RELEASE_DATE}
+          ## ${TITLE}
 
           **Please merge this Pull Request using 'Create a merge commit' button**
         EOF
 
-        gh pr create --title "Proxy release ${RELEASE_DATE}" \
+        gh pr create --title "${TITLE}" \
                      --body-file "body.md" \
                      --head "${RELEASE_BRANCH}" \
                      --base "release-proxy"

.github/workflows/trigger-e2e-tests.yml

@@ -13,8 +13,6 @@ defaults:
 env:
   # A concurrency group that we use for e2e-tests runs, matches `concurrency.group` above with `github.repository` as a prefix
   E2E_CONCURRENCY_GROUP: ${{ github.repository }}-e2e-tests-${{ github.ref_name }}-${{ github.ref_name == 'main' && github.sha || 'anysha' }}
-  AWS_ACCESS_KEY_ID: ${{ secrets.AWS_ACCESS_KEY_DEV }}
-  AWS_SECRET_ACCESS_KEY: ${{ secrets.AWS_SECRET_KEY_DEV }}
 
 jobs:
   cancel-previous-e2e-tests:
@@ -64,19 +62,35 @@ jobs:
     needs: [ tag ]
     runs-on: ubuntu-22.04
     env:
+      EVENT_ACTION: ${{ github.event.action }}
+      GH_TOKEN: ${{ secrets.CI_ACCESS_TOKEN }}
       TAG: ${{ needs.tag.outputs.build-tag }}
     steps:
-      - name: check if ecr image are present
-        env:
-          AWS_ACCESS_KEY_ID: ${{ secrets.AWS_ACCESS_KEY_DEV }}
-          AWS_SECRET_ACCESS_KEY: ${{ secrets.AWS_SECRET_KEY_DEV }}
+      - name: Wait for `promote-images` job to finish
+        # It's important to have a timeout here, the script in the step can run infinitely
+        timeout-minutes: 60
         run: |
-          for REPO in neon compute-tools compute-node-v14 vm-compute-node-v14 compute-node-v15 vm-compute-node-v15 compute-node-v16 vm-compute-node-v16; do
-            OUTPUT=$(aws ecr describe-images --repository-name ${REPO} --region eu-central-1 --query "imageDetails[?imageTags[?contains(@, '${TAG}')]]" --output text)
-            if [ "$OUTPUT" == "" ]; then
-              echo "$REPO with image tag $TAG not found" >> $GITHUB_OUTPUT
-              exit 1
-            fi
+          if [ "${GITHUB_EVENT_NAME}" != "pull_request" ] || [ "${EVENT_ACTION}" != "ready_for_review" ]; then
+            exit 0
+          fi
+
+          # For PRs we use the run id as the tag
+          BUILD_AND_TEST_RUN_ID=${TAG}
+          while true; do
+            conclusion=$(gh run --repo ${GITHUB_REPOSITORY} view ${BUILD_AND_TEST_RUN_ID} --json jobs --jq '.jobs[] | select(.name == "promote-images") | .conclusion')
+            case "$conclusion" in
+              success)
+                break
+                ;;
+              failure | cancelled | skipped)
+                echo "The 'promote-images' job didn't succeed: '${conclusion}'. Exiting..."
+                exit 1
+                ;;
+              *)
+                echo "The 'promote-images' hasn't succeed yet. Waiting..."
+                sleep 60
+                ;;
+            esac
           done
 
       - name: Set e2e-platforms

.neon_clippy_args

@@ -1,4 +1,5 @@
 # * `-A unknown_lints` – do not warn about unknown lint suppressions
 #                        that people with newer toolchains might use
 # * `-D warnings`      - fail on any warnings (`cargo` returns non-zero exit status)
-export CLIPPY_COMMON_ARGS="--locked --workspace --all-targets -- -A unknown_lints -D warnings"
+# * `-D clippy::todo`  - don't let `todo!()` slip into `main`
+export CLIPPY_COMMON_ARGS="--locked --workspace --all-targets -- -A unknown_lints -D warnings -D clippy::todo"

CODEOWNERS

@@ -1,13 +1,13 @@
 /compute_tools/ @neondatabase/control-plane @neondatabase/compute
 /storage_controller @neondatabase/storage
 /libs/pageserver_api/ @neondatabase/storage
-/libs/postgres_ffi/ @neondatabase/compute @neondatabase/safekeepers
+/libs/postgres_ffi/ @neondatabase/compute @neondatabase/storage
 /libs/remote_storage/ @neondatabase/storage
-/libs/safekeeper_api/ @neondatabase/safekeepers
+/libs/safekeeper_api/ @neondatabase/storage
 /libs/vm_monitor/ @neondatabase/autoscaling
 /pageserver/ @neondatabase/storage
 /pgxn/ @neondatabase/compute
-/pgxn/neon/ @neondatabase/compute @neondatabase/safekeepers
+/pgxn/neon/ @neondatabase/compute @neondatabase/storage
 /proxy/ @neondatabase/proxy
-/safekeeper/ @neondatabase/safekeepers
+/safekeeper/ @neondatabase/storage
 /vendor/ @neondatabase/compute

Cargo.toml

@@ -13,9 +13,9 @@ members = [
     "safekeeper",
     "storage_broker",
     "storage_controller",
+    "storage_controller/client",
     "storage_scrubber",
     "workspace_hack",
-    "trace",
     "libs/compute_api",
     "libs/pageserver_api",
     "libs/postgres_ffi",
@@ -77,14 +77,13 @@ const_format = "0.2"
 crc32c = "0.6"
 crossbeam-deque = "0.8.5"
 crossbeam-utils = "0.8.5"
-dashmap = { version = "6.0", features = ["raw-api"] }
+dashmap = { version = "5.5.0", features = ["raw-api"] }
 either = "1.8"
 enum-map = "2.4.2"
 enumset = "1.0.12"
 fail = "0.5.0"
 fallible-iterator = "0.2"
 framed-websockets = { version = "0.1.0", git = "https://github.com/neondatabase/framed-websockets" }
-fs2 = "0.4.3"
 futures = "0.3"
 futures-core = "0.3"
 futures-util = "0.3"
@@ -111,8 +110,8 @@ lasso = "0.7"
 leaky-bucket = "1.0.1"
 libc = "0.2"
 md5 = "0.7.0"
-measured = { version = "0.0.21", features=["lasso"] }
-measured-process = { version = "0.0.21" }
+measured = { version = "0.0.22", features=["lasso"] }
+measured-process = { version = "0.0.22" }
 memoffset = "0.8"
 nix = { version = "0.27", features = ["fs", "process", "socket", "signal", "poll"] }
 notify = "6.0.0"
@@ -127,7 +126,7 @@ parquet = { version = "51.0.0", default-features = false, features = ["zstd"] }
 parquet_derive = "51.0.0"
 pbkdf2 = { version = "0.12.1", features = ["simple", "std"] }
 pin-project-lite = "0.2"
-procfs = "0.14"
+procfs = "0.16"
 prometheus = {version = "0.13", default-features=false, features = ["process"]} # removes protobuf dependency
 prost = "0.11"
 rand = "0.8"
@@ -140,7 +139,7 @@ reqwest-retry = "0.5"
 routerify = "3"
 rpds = "0.13"
 rustc-hash = "1.1.0"
-rustls = "0.22"
+rustls = "0.23"
 rustls-pemfile = "2"
 rustls-split = "0.3"
 scopeguard = "1.1"
@@ -172,8 +171,8 @@ tikv-jemalloc-ctl = "0.5"
 tokio = { version = "1.17", features = ["macros"] }
 tokio-epoll-uring = { git = "https://github.com/neondatabase/tokio-epoll-uring.git" , branch = "main" }
 tokio-io-timeout = "1.2.0"
-tokio-postgres-rustls = "0.11.0"
-tokio-rustls = "0.25"
+tokio-postgres-rustls = "0.12.0"
+tokio-rustls = "0.26"
 tokio-stream = "0.1"
 tokio-tar = "0.3"
 tokio-util = { version = "0.7.10", features = ["io", "rt"] }
@@ -184,14 +183,17 @@ tower-service = "0.3.2"
 tracing = "0.1"
 tracing-error = "0.2.0"
 tracing-opentelemetry = "0.21.0"
-tracing-subscriber = { version = "0.3", default-features = false, features = ["smallvec", "fmt", "tracing-log", "std", "env-filter", "json", "ansi"] }
+tracing-subscriber = { version = "0.3", default-features = false, features = ["smallvec", "fmt", "tracing-log", "std", "env-filter", "json"] }
+try-lock = "0.2.5"
 twox-hash = { version = "1.6.3", default-features = false }
+typed-json = "0.1"
 url = "2.2"
 urlencoding = "2.1"
 uuid = { version = "1.6.1", features = ["v4", "v7", "serde"] }
 walkdir = "2.3.2"
 rustls-native-certs = "0.7"
 x509-parser = "0.15"
+whoami = "1.5.1"
 
 ## TODO replace this with tracing
 env_logger = "0.10"
@@ -203,9 +205,6 @@ postgres-protocol = { git = "https://github.com/neondatabase/rust-postgres.git",
 postgres-types = { git = "https://github.com/neondatabase/rust-postgres.git", branch="neon" }
 tokio-postgres = { git = "https://github.com/neondatabase/rust-postgres.git", branch="neon" }
 
-## Other git libraries
-heapless = { default-features=false, features=[], git = "https://github.com/japaric/heapless.git", rev = "644653bf3b831c6bb4963be2de24804acf5e5001" } # upstream release pending
-
 ## Local libraries
 compute_api = { version = "0.1", path = "./libs/compute_api/" }
 consumption_metrics = { version = "0.1", path = "./libs/consumption_metrics/" }
@@ -221,6 +220,7 @@ remote_storage = { version = "0.1", path = "./libs/remote_storage/" }
 safekeeper_api = { version = "0.1", path = "./libs/safekeeper_api" }
 desim = { version = "0.1", path = "./libs/desim" }
 storage_broker = { version = "0.1", path = "./storage_broker/" } # Note: main broker code is inside the binary crate, so linking with the library shouldn't be heavy.
+storage_controller_client = { path = "./storage_controller/client" }
 tenant_size_model = { version = "0.1", path = "./libs/tenant_size_model/" }
 tracing-utils = { version = "0.1", path = "./libs/tracing-utils/" }
 utils = { version = "0.1", path = "./libs/utils/" }
@@ -232,7 +232,7 @@ workspace_hack = { version = "0.1", path = "./workspace_hack/" }
 
 ## Build dependencies
 criterion = "0.5.1"
-rcgen = "0.12"
+rcgen = "0.13"
 rstest = "0.18"
 camino-tempfile = "1.0.2"
 tonic-build = "0.9"

Dockerfile

@@ -17,7 +17,7 @@ COPY --chown=nonroot pgxn pgxn
 COPY --chown=nonroot Makefile Makefile
 COPY --chown=nonroot scripts/ninstall.sh scripts/ninstall.sh
 
-ENV BUILD_TYPE release
+ENV BUILD_TYPE=release
 RUN set -e \
     && mold -run make -j $(nproc) -s neon-pg-ext \
     && rm -rf pg_install/build \
@@ -29,25 +29,15 @@ WORKDIR /home/nonroot
 ARG GIT_VERSION=local
 ARG BUILD_TAG
 
-# Enable https://github.com/paritytech/cachepot to cache Rust crates' compilation results in Docker builds.
-# Set up cachepot to use an AWS S3 bucket for cache results, to reuse it between `docker build` invocations.
-# cachepot falls back to local filesystem if S3 is misconfigured, not failing the build
-ARG RUSTC_WRAPPER=cachepot
-ENV AWS_REGION=eu-central-1
-ENV CACHEPOT_S3_KEY_PREFIX=cachepot
-ARG CACHEPOT_BUCKET=neon-github-dev
-#ARG AWS_ACCESS_KEY_ID
-#ARG AWS_SECRET_ACCESS_KEY
-
 COPY --from=pg-build /home/nonroot/pg_install/v14/include/postgresql/server pg_install/v14/include/postgresql/server
 COPY --from=pg-build /home/nonroot/pg_install/v15/include/postgresql/server pg_install/v15/include/postgresql/server
 COPY --from=pg-build /home/nonroot/pg_install/v16/include/postgresql/server pg_install/v16/include/postgresql/server
+COPY --from=pg-build /home/nonroot/pg_install/v16/lib                       pg_install/v16/lib
 COPY --chown=nonroot . .
 
-# Show build caching stats to check if it was used in the end.
-# Has to be the part of the same RUN since cachepot daemon is killed in the end of this RUN, losing the compilation stats.
+ARG ADDITIONAL_RUSTFLAGS
 RUN set -e \
-    && RUSTFLAGS="-Clinker=clang -Clink-arg=-fuse-ld=mold -Clink-arg=-Wl,--no-rosegment" cargo build  \
+    && PQ_LIB_DIR=$(pwd)/pg_install/v16/lib RUSTFLAGS="-Clinker=clang -Clink-arg=-fuse-ld=mold -Clink-arg=-Wl,--no-rosegment ${ADDITIONAL_RUSTFLAGS}" cargo build \
       --bin pg_sni_router  \
       --bin pageserver  \
       --bin pagectl  \
@@ -56,8 +46,8 @@ RUN set -e \
       --bin storage_controller  \
       --bin proxy  \
       --bin neon_local \
-      --locked --release \
-    && cachepot -s
+      --bin storage_scrubber \
+      --locked --release
 
 # Build final image
 #
@@ -82,6 +72,7 @@ COPY --from=build --chown=neon:neon /home/nonroot/target/release/storage_broker
 COPY --from=build --chown=neon:neon /home/nonroot/target/release/storage_controller  /usr/local/bin
 COPY --from=build --chown=neon:neon /home/nonroot/target/release/proxy               /usr/local/bin
 COPY --from=build --chown=neon:neon /home/nonroot/target/release/neon_local          /usr/local/bin
+COPY --from=build --chown=neon:neon /home/nonroot/target/release/storage_scrubber    /usr/local/bin
 
 COPY --from=pg-build /home/nonroot/pg_install/v14 /usr/local/v14/
 COPY --from=pg-build /home/nonroot/pg_install/v15 /usr/local/v15/
@@ -90,20 +81,24 @@ COPY --from=pg-build /home/nonroot/postgres_install.tar.gz /data/
 
 # By default, pageserver uses `.neon/` working directory in WORKDIR, so create one and fill it with the dummy config.
 # Now, when `docker run ... pageserver` is run, it can start without errors, yet will have some default dummy values.
-RUN mkdir -p /data/.neon/ && chown -R neon:neon /data/.neon/ \
-    && /usr/local/bin/pageserver -D /data/.neon/ --init \
-       -c "id=1234" \
-       -c "broker_endpoint='http://storage_broker:50051'" \
-       -c "pg_distrib_dir='/usr/local/'" \
-       -c "listen_pg_addr='0.0.0.0:6400'" \
-       -c "listen_http_addr='0.0.0.0:9898'"
+RUN mkdir -p /data/.neon/ && \
+  echo "id=1234" > "/data/.neon/identity.toml" && \
+  echo "broker_endpoint='http://storage_broker:50051'\n" \
+       "pg_distrib_dir='/usr/local/'\n" \
+       "listen_pg_addr='0.0.0.0:6400'\n" \
+       "listen_http_addr='0.0.0.0:9898'\n" \
+  > /data/.neon/pageserver.toml && \
+  chown -R neon:neon /data/.neon
 
 # When running a binary that links with libpq, default to using our most recent postgres version.  Binaries
 # that want a particular postgres version will select it explicitly: this is just a default.
-ENV LD_LIBRARY_PATH /usr/local/v16/lib
+ENV LD_LIBRARY_PATH=/usr/local/v16/lib
 
 
 VOLUME ["/data"]
 USER neon
 EXPOSE 6400
 EXPOSE 9898
+
+CMD ["/usr/local/bin/pageserver", "-D", "/data/.neon"]
+

Dockerfile.build-tools

@@ -1,5 +1,13 @@
 FROM debian:bullseye-slim
 
+# Use ARG as a build-time environment variable here to allow.
+# It's not supposed to be set outside.
+# Alternatively it can be obtained using the following command
+# ```
+# . /etc/os-release && echo "${VERSION_CODENAME}"
+# ```
+ARG DEBIAN_VERSION_CODENAME=bullseye
+
 # Add nonroot user
 RUN useradd -ms /bin/bash nonroot -b /home
 SHELL ["/bin/bash", "-c"]
@@ -26,7 +34,6 @@ RUN set -e \
         liblzma-dev \
         libncurses5-dev \
         libncursesw5-dev \
-        libpq-dev \
         libreadline-dev \
         libseccomp-dev \
         libsqlite3-dev \
@@ -51,7 +58,7 @@ RUN set -e \
     && rm -rf /var/lib/apt/lists/* /tmp/* /var/tmp/*
 
 # protobuf-compiler (protoc)
-ENV PROTOC_VERSION 25.1
+ENV PROTOC_VERSION=25.1
 RUN curl -fsSL "https://github.com/protocolbuffers/protobuf/releases/download/v${PROTOC_VERSION}/protoc-${PROTOC_VERSION}-linux-$(uname -m | sed 's/aarch64/aarch_64/g').zip" -o "protoc.zip" \
     && unzip -q protoc.zip -d protoc \
     && mv protoc/bin/protoc /usr/local/bin/protoc \
@@ -67,19 +74,24 @@ RUN curl -sL "https://github.com/peak/s5cmd/releases/download/v${S5CMD_VERSION}/
 # LLVM
 ENV LLVM_VERSION=18
 RUN curl -fsSL 'https://apt.llvm.org/llvm-snapshot.gpg.key' | apt-key add - \
-    && echo "deb http://apt.llvm.org/bullseye/ llvm-toolchain-bullseye-${LLVM_VERSION} main" > /etc/apt/sources.list.d/llvm.stable.list \
+    && echo "deb http://apt.llvm.org/${DEBIAN_VERSION_CODENAME}/ llvm-toolchain-${DEBIAN_VERSION_CODENAME}-${LLVM_VERSION} main" > /etc/apt/sources.list.d/llvm.stable.list \
     && apt update \
     && apt install -y clang-${LLVM_VERSION} llvm-${LLVM_VERSION} \
     && bash -c 'for f in /usr/bin/clang*-${LLVM_VERSION} /usr/bin/llvm*-${LLVM_VERSION}; do ln -s "${f}" "${f%-${LLVM_VERSION}}"; done' \
     && rm -rf /var/lib/apt/lists/* /tmp/* /var/tmp/*
 
-# PostgreSQL 14
-RUN curl -fsSL 'https://www.postgresql.org/media/keys/ACCC4CF8.asc' | apt-key add - \
-    && echo 'deb http://apt.postgresql.org/pub/repos/apt bullseye-pgdg main' > /etc/apt/sources.list.d/pgdg.list \
+# Install docker
+RUN curl -fsSL https://download.docker.com/linux/ubuntu/gpg | gpg --dearmor -o /usr/share/keyrings/docker-archive-keyring.gpg \
+    && echo "deb [arch=$(dpkg --print-architecture) signed-by=/usr/share/keyrings/docker-archive-keyring.gpg] https://download.docker.com/linux/debian ${DEBIAN_VERSION_CODENAME} stable" > /etc/apt/sources.list.d/docker.list \
     && apt update \
-    && apt install -y postgresql-client-14 \
+    && apt install -y docker-ce docker-ce-cli \
     && rm -rf /var/lib/apt/lists/* /tmp/* /var/tmp/*
 
+# Configure sudo & docker
+RUN usermod -aG sudo nonroot && \
+    echo '%sudo ALL=(ALL) NOPASSWD:ALL' >> /etc/sudoers && \
+    usermod -aG docker nonroot
+
 # AWS CLI
 RUN curl "https://awscli.amazonaws.com/awscli-exe-linux-$(uname -m).zip" -o "awscliv2.zip" \
     && unzip -q awscliv2.zip \
@@ -87,7 +99,7 @@ RUN curl "https://awscli.amazonaws.com/awscli-exe-linux-$(uname -m).zip" -o "aws
     && rm awscliv2.zip
 
 # Mold: A Modern Linker
-ENV MOLD_VERSION v2.31.0
+ENV MOLD_VERSION=v2.33.0
 RUN set -e \
     && git clone https://github.com/rui314/mold.git \
     && mkdir mold/build \
@@ -113,10 +125,10 @@ RUN for package in Capture::Tiny DateTime Devel::Cover Digest::MD5 File::Spec JS
     && rm -rf ../lcov.tar.gz
 
 # Compile and install the static OpenSSL library
-ENV OPENSSL_VERSION=3.2.2
+ENV OPENSSL_VERSION=1.1.1w
 ENV OPENSSL_PREFIX=/usr/local/openssl
 RUN wget -O /tmp/openssl-${OPENSSL_VERSION}.tar.gz https://www.openssl.org/source/openssl-${OPENSSL_VERSION}.tar.gz && \
-    echo "197149c18d9e9f292c43f0400acaba12e5f52cacfe050f3d199277ea738ec2e7 /tmp/openssl-${OPENSSL_VERSION}.tar.gz" | sha256sum --check && \
+    echo "cf3098950cb4d853ad95c0841f1f9c6d3dc102dccfcacd521d93925208b76ac8 /tmp/openssl-${OPENSSL_VERSION}.tar.gz" | sha256sum --check && \
     cd /tmp && \
     tar xzvf /tmp/openssl-${OPENSSL_VERSION}.tar.gz && \
     rm /tmp/openssl-${OPENSSL_VERSION}.tar.gz && \
@@ -156,7 +168,7 @@ USER nonroot:nonroot
 WORKDIR /home/nonroot
 
 # Python
-ENV PYTHON_VERSION=3.9.18 \
+ENV PYTHON_VERSION=3.9.19 \
     PYENV_ROOT=/home/nonroot/.pyenv \
     PATH=/home/nonroot/.pyenv/shims:/home/nonroot/.pyenv/bin:/home/nonroot/.poetry/bin:$PATH
 RUN set -e \
@@ -180,9 +192,14 @@ WORKDIR /home/nonroot
 
 # Rust
 # Please keep the version of llvm (installed above) in sync with rust llvm (`rustc --version --verbose | grep LLVM`)
-ENV RUSTC_VERSION=1.79.0
+ENV RUSTC_VERSION=1.80.1
 ENV RUSTUP_HOME="/home/nonroot/.rustup"
 ENV PATH="/home/nonroot/.cargo/bin:${PATH}"
+ARG RUSTFILT_VERSION=0.2.1
+ARG CARGO_HAKARI_VERSION=0.9.30
+ARG CARGO_DENY_VERSION=0.16.1
+ARG CARGO_HACK_VERSION=0.6.31
+ARG CARGO_NEXTEST_VERSION=0.9.72
 RUN curl -sSO https://static.rust-lang.org/rustup/dist/$(uname -m)-unknown-linux-gnu/rustup-init && whoami && \
 	chmod +x rustup-init && \
 	./rustup-init -y --default-toolchain ${RUSTC_VERSION} && \
@@ -191,15 +208,13 @@ RUN curl -sSO https://static.rust-lang.org/rustup/dist/$(uname -m)-unknown-linux
     . "$HOME/.cargo/env" && \
     cargo --version && rustup --version && \
     rustup component add llvm-tools-preview rustfmt clippy && \
-    cargo install --git https://github.com/paritytech/cachepot && \
-    cargo install rustfilt && \
-    cargo install cargo-hakari && \
-    cargo install cargo-deny --locked && \
-    cargo install cargo-hack && \
-    cargo install cargo-nextest && \
+    cargo install rustfilt            --version ${RUSTFILT_VERSION} && \
+    cargo install cargo-hakari        --version ${CARGO_HAKARI_VERSION} && \
+    cargo install cargo-deny --locked --version ${CARGO_DENY_VERSION} && \
+    cargo install cargo-hack          --version ${CARGO_HACK_VERSION} && \
+    cargo install cargo-nextest       --version ${CARGO_NEXTEST_VERSION} && \
     rm -rf /home/nonroot/.cargo/registry && \
     rm -rf /home/nonroot/.cargo/git
-ENV RUSTC_WRAPPER=cachepot
 
 # Show versions
 RUN whoami \

Dockerfile.compute-node

@@ -94,7 +94,7 @@ RUN wget https://gitlab.com/Oslandia/SFCGAL/-/archive/v1.3.10/SFCGAL-v1.3.10.tar
     DESTDIR=/sfcgal make install -j $(getconf _NPROCESSORS_ONLN) && \
     make clean && cp -R /sfcgal/* /
 
-ENV PATH "/usr/local/pgsql/bin:$PATH"
+ENV PATH="/usr/local/pgsql/bin:$PATH"
 
 RUN wget https://download.osgeo.org/postgis/source/postgis-3.3.3.tar.gz -O postgis.tar.gz && \
     echo "74eb356e3f85f14233791013360881b6748f78081cc688ff9d6f0f673a762d13 postgis.tar.gz" | sha256sum --check && \
@@ -311,9 +311,12 @@ RUN wget https://github.com/iCyberon/pg_hashids/archive/refs/tags/v1.2.1.tar.gz
 FROM build-deps AS rum-pg-build
 COPY --from=pg-build /usr/local/pgsql/ /usr/local/pgsql/
 
+COPY patches/rum.patch /rum.patch
+
 RUN wget https://github.com/postgrespro/rum/archive/refs/tags/1.3.13.tar.gz -O rum.tar.gz && \
     echo "6ab370532c965568df6210bd844ac6ba649f53055e48243525b0b7e5c4d69a7d rum.tar.gz" | sha256sum --check && \
     mkdir rum-src && cd rum-src && tar xzf ../rum.tar.gz --strip-components=1 -C . && \
+    patch -p1 < /rum.patch && \
     make -j $(getconf _NPROCESSORS_ONLN) PG_CONFIG=/usr/local/pgsql/bin/pg_config USE_PGXS=1 && \
     make -j $(getconf _NPROCESSORS_ONLN) install PG_CONFIG=/usr/local/pgsql/bin/pg_config USE_PGXS=1 && \
     echo 'trusted = true' >> /usr/local/pgsql/share/extension/rum.control
@@ -408,7 +411,7 @@ FROM build-deps AS timescaledb-pg-build
 COPY --from=pg-build /usr/local/pgsql/ /usr/local/pgsql/
 
 ARG PG_VERSION
-ENV PATH "/usr/local/pgsql/bin:$PATH"
+ENV PATH="/usr/local/pgsql/bin:$PATH"
 
 RUN case "${PG_VERSION}" in \
       "v14" | "v15") \
@@ -441,7 +444,7 @@ FROM build-deps AS pg-hint-plan-pg-build
 COPY --from=pg-build /usr/local/pgsql/ /usr/local/pgsql/
 
 ARG PG_VERSION
-ENV PATH "/usr/local/pgsql/bin:$PATH"
+ENV PATH="/usr/local/pgsql/bin:$PATH"
 
 RUN case "${PG_VERSION}" in \
       "v14") \
@@ -467,31 +470,6 @@ RUN case "${PG_VERSION}" in \
     make install -j $(getconf _NPROCESSORS_ONLN) && \
     echo "trusted = true" >> /usr/local/pgsql/share/extension/pg_hint_plan.control
 
-#########################################################################################
-#
-# Layer "kq-imcx-pg-build"
-# compile kq_imcx extension
-#
-#########################################################################################
-FROM build-deps AS kq-imcx-pg-build
-COPY --from=pg-build /usr/local/pgsql/ /usr/local/pgsql/
-
-ENV PATH "/usr/local/pgsql/bin/:$PATH"
-RUN apt-get update && \
-    apt-get install -y git libgtk2.0-dev libpq-dev libpam-dev libxslt-dev libkrb5-dev cmake && \
-    wget https://github.com/ketteq-neon/postgres-exts/archive/e0bd1a9d9313d7120c1b9c7bb15c48c0dede4c4e.tar.gz -O kq_imcx.tar.gz && \
-    echo "dc93a97ff32d152d32737ba7e196d9687041cda15e58ab31344c2f2de8855336 kq_imcx.tar.gz" | sha256sum --check && \
-    mkdir kq_imcx-src && cd kq_imcx-src && tar xzf ../kq_imcx.tar.gz --strip-components=1 -C . && \
-    find /usr/local/pgsql -type f | sed 's|^/usr/local/pgsql/||' > /before.txt &&\
-    mkdir build && cd build && \
-    cmake -DCMAKE_BUILD_TYPE=Release .. && \
-    make -j $(getconf _NPROCESSORS_ONLN) && \
-    make -j $(getconf _NPROCESSORS_ONLN) install && \
-    echo 'trusted = true' >> /usr/local/pgsql/share/extension/kq_imcx.control && \
-    find /usr/local/pgsql -type f | sed 's|^/usr/local/pgsql/||' > /after.txt &&\
-    mkdir -p /extensions/kq_imcx && cp /usr/local/pgsql/share/extension/kq_imcx.control /extensions/kq_imcx && \
-    sort -o /before.txt /before.txt && sort -o /after.txt /after.txt && \
-    comm -13 /before.txt /after.txt | tar --directory=/usr/local/pgsql --zstd -cf /extensions/kq_imcx.tar.zst -T -
 
 #########################################################################################
 #
@@ -502,7 +480,7 @@ RUN apt-get update && \
 FROM build-deps AS pg-cron-pg-build
 COPY --from=pg-build /usr/local/pgsql/ /usr/local/pgsql/
 
-ENV PATH "/usr/local/pgsql/bin/:$PATH"
+ENV PATH="/usr/local/pgsql/bin/:$PATH"
 RUN wget https://github.com/citusdata/pg_cron/archive/refs/tags/v1.6.0.tar.gz -O pg_cron.tar.gz && \
     echo "383a627867d730222c272bfd25cd5e151c578d73f696d32910c7db8c665cc7db pg_cron.tar.gz" | sha256sum --check && \
     mkdir pg_cron-src && cd pg_cron-src && tar xzf ../pg_cron.tar.gz --strip-components=1 -C . && \
@@ -528,7 +506,7 @@ RUN apt-get update && \
         libboost-system1.74-dev \
         libeigen3-dev
 
-ENV PATH "/usr/local/pgsql/bin/:/usr/local/pgsql/:$PATH"
+ENV PATH="/usr/local/pgsql/bin/:/usr/local/pgsql/:$PATH"
 RUN wget https://github.com/rdkit/rdkit/archive/refs/tags/Release_2023_03_3.tar.gz -O rdkit.tar.gz && \
     echo "bdbf9a2e6988526bfeb8c56ce3cdfe2998d60ac289078e2215374288185e8c8d rdkit.tar.gz" | sha256sum --check && \
     mkdir rdkit-src && cd rdkit-src && tar xzf ../rdkit.tar.gz --strip-components=1 -C . && \
@@ -568,7 +546,7 @@ RUN wget https://github.com/rdkit/rdkit/archive/refs/tags/Release_2023_03_3.tar.
 FROM build-deps AS pg-uuidv7-pg-build
 COPY --from=pg-build /usr/local/pgsql/ /usr/local/pgsql/
 
-ENV PATH "/usr/local/pgsql/bin/:$PATH"
+ENV PATH="/usr/local/pgsql/bin/:$PATH"
 RUN wget https://github.com/fboulnois/pg_uuidv7/archive/refs/tags/v1.0.1.tar.gz -O pg_uuidv7.tar.gz && \
     echo "0d0759ab01b7fb23851ecffb0bce27822e1868a4a5819bfd276101c716637a7a pg_uuidv7.tar.gz" | sha256sum --check && \
     mkdir pg_uuidv7-src && cd pg_uuidv7-src && tar xzf ../pg_uuidv7.tar.gz --strip-components=1 -C . && \
@@ -585,7 +563,7 @@ RUN wget https://github.com/fboulnois/pg_uuidv7/archive/refs/tags/v1.0.1.tar.gz
 FROM build-deps AS pg-roaringbitmap-pg-build
 COPY --from=pg-build /usr/local/pgsql/ /usr/local/pgsql/
 
-ENV PATH "/usr/local/pgsql/bin/:$PATH"
+ENV PATH="/usr/local/pgsql/bin/:$PATH"
 RUN wget https://github.com/ChenHuajun/pg_roaringbitmap/archive/refs/tags/v0.5.4.tar.gz -O pg_roaringbitmap.tar.gz && \
     echo "b75201efcb1c2d1b014ec4ae6a22769cc7a224e6e406a587f5784a37b6b5a2aa pg_roaringbitmap.tar.gz" | sha256sum --check && \
     mkdir pg_roaringbitmap-src && cd pg_roaringbitmap-src && tar xzf ../pg_roaringbitmap.tar.gz --strip-components=1 -C . && \
@@ -602,7 +580,7 @@ RUN wget https://github.com/ChenHuajun/pg_roaringbitmap/archive/refs/tags/v0.5.4
 FROM build-deps AS pg-semver-pg-build
 COPY --from=pg-build /usr/local/pgsql/ /usr/local/pgsql/
 
-ENV PATH "/usr/local/pgsql/bin/:$PATH"
+ENV PATH="/usr/local/pgsql/bin/:$PATH"
 RUN wget https://github.com/theory/pg-semver/archive/refs/tags/v0.32.1.tar.gz -O pg_semver.tar.gz && \
     echo "fbdaf7512026d62eec03fad8687c15ed509b6ba395bff140acd63d2e4fbe25d7 pg_semver.tar.gz" | sha256sum --check && \
     mkdir pg_semver-src && cd pg_semver-src && tar xzf ../pg_semver.tar.gz --strip-components=1 -C . && \
@@ -620,7 +598,7 @@ FROM build-deps AS pg-embedding-pg-build
 COPY --from=pg-build /usr/local/pgsql/ /usr/local/pgsql/
 
 ARG PG_VERSION
-ENV PATH "/usr/local/pgsql/bin/:$PATH"
+ENV PATH="/usr/local/pgsql/bin/:$PATH"
 RUN case "${PG_VERSION}" in \
       "v14" | "v15") \
         export PG_EMBEDDING_VERSION=0.3.5 \
@@ -644,7 +622,7 @@ RUN case "${PG_VERSION}" in \
 FROM build-deps AS pg-anon-pg-build
 COPY --from=pg-build /usr/local/pgsql/ /usr/local/pgsql/
 
-ENV PATH "/usr/local/pgsql/bin/:$PATH"
+ENV PATH="/usr/local/pgsql/bin/:$PATH"
 RUN wget  https://github.com/neondatabase/postgresql_anonymizer/archive/refs/tags/neon_1.1.1.tar.gz -O pg_anon.tar.gz && \
     echo "321ea8d5c1648880aafde850a2c576e4a9e7b9933a34ce272efc839328999fa9  pg_anon.tar.gz" | sha256sum --check && \
     mkdir pg_anon-src && cd pg_anon-src && tar xzf ../pg_anon.tar.gz --strip-components=1 -C . && \
@@ -679,7 +657,7 @@ RUN curl -sSO https://static.rust-lang.org/rustup/dist/$(uname -m)-unknown-linux
     chmod +x rustup-init && \
     ./rustup-init -y --no-modify-path --profile minimal --default-toolchain stable && \
     rm rustup-init && \
-    cargo install --locked --version 0.10.2 cargo-pgrx && \
+    cargo install --locked --version 0.11.3 cargo-pgrx && \
     /bin/bash -c 'cargo pgrx init --pg${PG_VERSION:1}=/usr/local/pgsql/bin/pg_config'
 
 USER root
@@ -694,10 +672,15 @@ USER root
 FROM rust-extensions-build AS pg-jsonschema-pg-build
 ARG PG_VERSION
 
-RUN wget https://github.com/supabase/pg_jsonschema/archive/refs/tags/v0.2.0.tar.gz -O pg_jsonschema.tar.gz && \
-    echo "9118fc508a6e231e7a39acaa6f066fcd79af17a5db757b47d2eefbe14f7794f0 pg_jsonschema.tar.gz" | sha256sum --check && \
+RUN wget https://github.com/supabase/pg_jsonschema/archive/refs/tags/v0.3.1.tar.gz -O pg_jsonschema.tar.gz && \
+    echo "61df3db1ed83cf24f6aa39c826f8818bfa4f0bd33b587fd6b2b1747985642297 pg_jsonschema.tar.gz" | sha256sum --check && \
     mkdir pg_jsonschema-src && cd pg_jsonschema-src && tar xzf ../pg_jsonschema.tar.gz --strip-components=1 -C . && \
-    sed -i 's/pgrx = "0.10.2"/pgrx = { version = "0.10.2", features = [ "unsafe-postgres" ] }/g' Cargo.toml && \
+    # see commit 252b3685a27a0f4c31a0f91e983c6314838e89e8
+    # `unsafe-postgres` feature allows to build pgx extensions
+    # against postgres forks that decided to change their ABI name (like us).
+    # With that we can build extensions without forking them and using stock
+    # pgx. As this feature is new few manual version bumps were required.
+    sed -i 's/pgrx = "0.11.3"/pgrx = { version = "0.11.3", features = [ "unsafe-postgres" ] }/g' Cargo.toml && \
     cargo pgrx install --release && \
     echo "trusted = true" >> /usr/local/pgsql/share/extension/pg_jsonschema.control
 
@@ -711,10 +694,10 @@ RUN wget https://github.com/supabase/pg_jsonschema/archive/refs/tags/v0.2.0.tar.
 FROM rust-extensions-build AS pg-graphql-pg-build
 ARG PG_VERSION
 
-RUN wget https://github.com/supabase/pg_graphql/archive/refs/tags/v1.4.0.tar.gz -O pg_graphql.tar.gz && \
-    echo "bd8dc7230282b3efa9ae5baf053a54151ed0e66881c7c53750e2d0c765776edc pg_graphql.tar.gz" | sha256sum --check && \
+RUN wget https://github.com/supabase/pg_graphql/archive/refs/tags/v1.5.7.tar.gz -O pg_graphql.tar.gz && \
+    echo "2b3e567a5b31019cb97ae0e33263c1bcc28580be5a444ac4c8ece5c4be2aea41 pg_graphql.tar.gz" | sha256sum --check && \
     mkdir pg_graphql-src && cd pg_graphql-src && tar xzf ../pg_graphql.tar.gz --strip-components=1 -C . && \
-    sed -i 's/pgrx = "=0.10.2"/pgrx = { version = "0.10.2", features = [ "unsafe-postgres" ] }/g' Cargo.toml && \
+    sed -i 's/pgrx = "=0.11.3"/pgrx = { version = "0.11.3", features = [ "unsafe-postgres" ] }/g' Cargo.toml && \
     cargo pgrx install --release && \
     # it's needed to enable extension because it uses untrusted C language
     sed -i 's/superuser = false/superuser = true/g' /usr/local/pgsql/share/extension/pg_graphql.control && \
@@ -734,6 +717,9 @@ ARG PG_VERSION
 RUN wget https://github.com/kelvich/pg_tiktoken/archive/26806147b17b60763039c6a6878884c41a262318.tar.gz -O pg_tiktoken.tar.gz && \
     echo "e64e55aaa38c259512d3e27c572da22c4637418cf124caba904cd50944e5004e pg_tiktoken.tar.gz" | sha256sum --check && \
     mkdir pg_tiktoken-src && cd pg_tiktoken-src && tar xzf ../pg_tiktoken.tar.gz --strip-components=1 -C . && \
+    # TODO update pgrx version in the pg_tiktoken repo and remove this line
+    sed -i 's/pgrx = { version = "=0.10.2",/pgrx = { version = "0.11.3",/g' Cargo.toml && \
+    sed -i 's/pgrx-tests = "=0.10.2"/pgrx-tests = "0.11.3"/g' Cargo.toml && \
     cargo pgrx install --release && \
     echo "trusted = true" >> /usr/local/pgsql/share/extension/pg_tiktoken.control
 
@@ -747,14 +733,10 @@ RUN wget https://github.com/kelvich/pg_tiktoken/archive/26806147b17b60763039c6a6
 FROM rust-extensions-build AS pg-pgx-ulid-build
 ARG PG_VERSION
 
-RUN wget https://github.com/pksunkara/pgx_ulid/archive/refs/tags/v0.1.3.tar.gz -O pgx_ulid.tar.gz && \
-    echo "ee5db82945d2d9f2d15597a80cf32de9dca67b897f605beb830561705f12683c pgx_ulid.tar.gz" | sha256sum --check && \
+RUN wget https://github.com/pksunkara/pgx_ulid/archive/refs/tags/v0.1.5.tar.gz -O pgx_ulid.tar.gz && \
+    echo "9d1659a2da65af0133d5451c454de31b37364e3502087dadf579f790bc8bef17 pgx_ulid.tar.gz" | sha256sum --check && \
     mkdir pgx_ulid-src && cd pgx_ulid-src && tar xzf ../pgx_ulid.tar.gz --strip-components=1 -C . && \
-    echo "******************* Apply a patch for Postgres 16 support; delete in the next release ******************" && \
-    wget https://github.com/pksunkara/pgx_ulid/commit/f84954cf63fc8c80d964ac970d9eceed3c791196.patch && \
-    patch -p1 < f84954cf63fc8c80d964ac970d9eceed3c791196.patch && \
-    echo "********************************************************************************************************" && \
-    sed -i 's/pgrx       = "=0.10.2"/pgrx = { version = "=0.10.2", features = [ "unsafe-postgres" ] }/g' Cargo.toml && \
+    sed -i 's/pgrx       = "^0.11.2"/pgrx = { version = "=0.11.3", features = [ "unsafe-postgres" ] }/g' Cargo.toml && \
     cargo pgrx install --release && \
     echo "trusted = true" >> /usr/local/pgsql/share/extension/ulid.control
 
@@ -768,7 +750,7 @@ RUN wget https://github.com/pksunkara/pgx_ulid/archive/refs/tags/v0.1.3.tar.gz -
 FROM build-deps AS wal2json-pg-build
 COPY --from=pg-build /usr/local/pgsql/ /usr/local/pgsql/
 
-ENV PATH "/usr/local/pgsql/bin/:$PATH"
+ENV PATH="/usr/local/pgsql/bin/:$PATH"
 RUN wget https://github.com/eulerto/wal2json/archive/refs/tags/wal2json_2_5.tar.gz && \
     echo "b516653575541cf221b99cf3f8be9b6821f6dbcfc125675c85f35090f824f00e wal2json_2_5.tar.gz" | sha256sum --check && \
     mkdir wal2json-src && cd wal2json-src && tar xzf ../wal2json_2_5.tar.gz --strip-components=1 -C . && \
@@ -784,7 +766,7 @@ RUN wget https://github.com/eulerto/wal2json/archive/refs/tags/wal2json_2_5.tar.
 FROM build-deps AS pg-ivm-build
 COPY --from=pg-build /usr/local/pgsql/ /usr/local/pgsql/
 
-ENV PATH "/usr/local/pgsql/bin/:$PATH"
+ENV PATH="/usr/local/pgsql/bin/:$PATH"
 RUN wget https://github.com/sraoss/pg_ivm/archive/refs/tags/v1.7.tar.gz -O pg_ivm.tar.gz && \
     echo "ebfde04f99203c7be4b0e873f91104090e2e83e5429c32ac242d00f334224d5e pg_ivm.tar.gz" | sha256sum --check && \
     mkdir pg_ivm-src && cd pg_ivm-src && tar xzf ../pg_ivm.tar.gz --strip-components=1 -C . && \
@@ -801,7 +783,7 @@ RUN wget https://github.com/sraoss/pg_ivm/archive/refs/tags/v1.7.tar.gz -O pg_iv
 FROM build-deps AS pg-partman-build
 COPY --from=pg-build /usr/local/pgsql/ /usr/local/pgsql/
 
-ENV PATH "/usr/local/pgsql/bin/:$PATH"
+ENV PATH="/usr/local/pgsql/bin/:$PATH"
 RUN wget https://github.com/pgpartman/pg_partman/archive/refs/tags/v5.0.1.tar.gz -O pg_partman.tar.gz && \
     echo "75b541733a9659a6c90dbd40fccb904a630a32880a6e3044d0c4c5f4c8a65525 pg_partman.tar.gz" | sha256sum --check && \
     mkdir pg_partman-src && cd pg_partman-src && tar xzf ../pg_partman.tar.gz --strip-components=1 -C . && \
@@ -840,7 +822,6 @@ COPY --from=hll-pg-build /usr/local/pgsql/ /usr/local/pgsql/
 COPY --from=plpgsql-check-pg-build /usr/local/pgsql/ /usr/local/pgsql/
 COPY --from=timescaledb-pg-build /usr/local/pgsql/ /usr/local/pgsql/
 COPY --from=pg-hint-plan-pg-build /usr/local/pgsql/ /usr/local/pgsql/
-COPY --from=kq-imcx-pg-build /usr/local/pgsql/ /usr/local/pgsql/
 COPY --from=pg-cron-pg-build /usr/local/pgsql/ /usr/local/pgsql/
 COPY --from=pg-pgx-ulid-build /usr/local/pgsql/ /usr/local/pgsql/
 COPY --from=rdkit-pg-build /usr/local/pgsql/ /usr/local/pgsql/
@@ -952,7 +933,8 @@ COPY --from=pgjwt-pg-build /pgjwt.tar.gz /ext-src
 #COPY --from=pg-tiktoken-pg-build /home/nonroot/pg_tiktoken.tar.gz /ext-src
 COPY --from=hypopg-pg-build /hypopg.tar.gz /ext-src
 COPY --from=pg-hashids-pg-build /pg_hashids.tar.gz /ext-src
-#COPY --from=rum-pg-build /rum.tar.gz /ext-src
+COPY --from=rum-pg-build /rum.tar.gz /ext-src
+COPY patches/rum.patch /ext-src
 #COPY --from=pgtap-pg-build /pgtap.tar.gz /ext-src
 COPY --from=ip4r-pg-build /ip4r.tar.gz /ext-src
 COPY --from=prefix-pg-build /prefix.tar.gz /ext-src
@@ -961,11 +943,10 @@ COPY --from=plpgsql-check-pg-build /plpgsql_check.tar.gz /ext-src
 #COPY --from=timescaledb-pg-build /timescaledb.tar.gz /ext-src
 COPY --from=pg-hint-plan-pg-build /pg_hint_plan.tar.gz /ext-src
 COPY patches/pg_hintplan.patch /ext-src
-#COPY --from=kq-imcx-pg-build /kq_imcx.tar.gz /ext-src
 COPY --from=pg-cron-pg-build /pg_cron.tar.gz /ext-src
 COPY patches/pg_cron.patch /ext-src
 #COPY --from=pg-pgx-ulid-build /home/nonroot/pgx_ulid.tar.gz /ext-src
-COPY --from=rdkit-pg-build /rdkit.tar.gz /ext-src
+#COPY --from=rdkit-pg-build /rdkit.tar.gz /ext-src
 COPY --from=pg-uuidv7-pg-build /pg_uuidv7.tar.gz /ext-src
 COPY --from=pg-roaringbitmap-pg-build /pg_roaringbitmap.tar.gz /ext-src
 COPY --from=pg-semver-pg-build /pg_semver.tar.gz /ext-src
@@ -980,6 +961,7 @@ RUN cd /ext-src/ && for f in *.tar.gz; \
     rm -rf $dname; mkdir $dname; tar xzf $f --strip-components=1 -C $dname \
     || exit 1; rm -f $f; done
 RUN cd /ext-src/pgvector-src && patch -p1 <../pgvector.patch
+RUN cd /ext-src/rum-src && patch -p1 <../rum.patch
 # cmake is required for the h3 test
 RUN apt-get update && apt-get install -y cmake
 RUN patch -p1 < /ext-src/pg_hintplan.patch
@@ -1052,6 +1034,6 @@ RUN apt update &&  \
     rm -rf /var/lib/apt/lists/* /tmp/* /var/tmp/* && \
     localedef -i en_US -c -f UTF-8 -A /usr/share/locale/locale.alias en_US.UTF-8
 
-ENV LANG en_US.utf8
+ENV LANG=en_US.utf8
 USER postgres
 ENTRYPOINT ["/usr/local/bin/compute_ctl"]

Makefile

@@ -69,6 +69,8 @@ CARGO_CMD_PREFIX += CARGO_TERM_PROGRESS_WHEN=never CI=1
 # Set PQ_LIB_DIR to make sure `storage_controller` get linked with bundled libpq (through diesel)
 CARGO_CMD_PREFIX += PQ_LIB_DIR=$(POSTGRES_INSTALL_DIR)/v16/lib
 
+CACHEDIR_TAG_CONTENTS := "Signature: 8a477f597d28d172789f06886806bc55"
+
 #
 # Top level Makefile to build Neon and PostgreSQL
 #
@@ -79,15 +81,24 @@ all: neon postgres neon-pg-ext
 #
 # The 'postgres_ffi' depends on the Postgres headers.
 .PHONY: neon
-neon: postgres-headers walproposer-lib
+neon: postgres-headers walproposer-lib cargo-target-dir
 	+@echo "Compiling Neon"
 	$(CARGO_CMD_PREFIX) cargo build $(CARGO_BUILD_FLAGS)
+.PHONY: cargo-target-dir
+cargo-target-dir:
+	# https://github.com/rust-lang/cargo/issues/14281
+	mkdir -p target
+	test -e target/CACHEDIR.TAG || echo "$(CACHEDIR_TAG_CONTENTS)" > target/CACHEDIR.TAG
 
 ### PostgreSQL parts
 # Some rules are duplicated for Postgres v14 and 15. We may want to refactor
 # to avoid the duplication in the future, but it's tolerable for now.
 #
 $(POSTGRES_INSTALL_DIR)/build/%/config.status:
+
+	mkdir -p $(POSTGRES_INSTALL_DIR)
+	test -e $(POSTGRES_INSTALL_DIR)/CACHEDIR.TAG || echo "$(CACHEDIR_TAG_CONTENTS)" > $(POSTGRES_INSTALL_DIR)/CACHEDIR.TAG
+
 	+@echo "Configuring Postgres $* build"
 	@test -s $(ROOT_PROJECT_DIR)/vendor/postgres-$*/configure || { \
 		echo "\nPostgres submodule not found in $(ROOT_PROJECT_DIR)/vendor/postgres-$*/, execute "; \

README.md

@@ -262,7 +262,7 @@ By default, this runs both debug and release modes, and all supported postgres v
 testing locally, it is convenient to run just one set of permutations, like this:
 
 ```sh
-DEFAULT_PG_VERSION=15 BUILD_TYPE=release ./scripts/pytest
+DEFAULT_PG_VERSION=16 BUILD_TYPE=release ./scripts/pytest
 ```
 
 ## Flamegraphs

compute_tools/Cargo.toml

@@ -4,6 +4,11 @@ version = "0.1.0"
 edition.workspace = true
 license.workspace = true
 
+[features]
+default = []
+# Enables test specific features.
+testing = []
+
 [dependencies]
 anyhow.workspace = true
 async-compression.workspace = true
@@ -44,3 +49,4 @@ vm_monitor = { version = "0.1", path = "../libs/vm_monitor/" }
 zstd = "0.13"
 bytes = "1.0"
 rust-ini = "0.20.0"
+rlimit = "0.10.1"

compute_tools/src/bin/compute_ctl.rs

@@ -6,7 +6,7 @@
 //! - Every start is a fresh start, so the data directory is removed and
 //!   initialized again on each run.
 //! - If remote_extension_config is provided, it will be used to fetch extensions list
-//!  and download `shared_preload_libraries` from the remote storage.
+//!   and download `shared_preload_libraries` from the remote storage.
 //! - Next it will put configuration files into the `PGDATA` directory.
 //! - Sync safekeepers and get commit LSN.
 //! - Get `basebackup` from pageserver using the returned on the previous step LSN.
@@ -33,7 +33,6 @@
 //!             -b /usr/local/bin/postgres \
 //!             -r http://pg-ext-s3-gateway \
 //! ```
-//!
 use std::collections::HashMap;
 use std::fs::File;
 use std::path::Path;
@@ -64,6 +63,7 @@ use compute_tools::monitor::launch_monitor;
 use compute_tools::params::*;
 use compute_tools::spec::*;
 use compute_tools::swap::resize_swap;
+use rlimit::{setrlimit, Resource};
 
 // this is an arbitrary build tag. Fine as a default / for testing purposes
 // in-case of not-set environment var
@@ -72,6 +72,9 @@ const BUILD_TAG_DEFAULT: &str = "latest";
 fn main() -> Result<()> {
     let (build_tag, clap_args) = init()?;
 
+    // enable core dumping for all child processes
+    setrlimit(Resource::CORE, rlimit::INFINITY, rlimit::INFINITY)?;
+
     let (pg_handle, start_pg_result) = {
         // Enter startup tracing context
         let _startup_context_guard = startup_context_from_env();

compute_tools/src/compute.rs

@@ -56,6 +56,7 @@ pub struct ComputeNode {
     /// - we push new spec and it does reconfiguration
     /// - but then something happens and compute pod / VM is destroyed,
     ///   so k8s controller starts it again with the **old** spec
+    ///
     /// and the same for empty computes:
     /// - we started compute without any spec
     /// - we push spec and it does configuration
@@ -399,7 +400,15 @@ impl ComputeNode {
     pub fn get_basebackup(&self, compute_state: &ComputeState, lsn: Lsn) -> Result<()> {
         let mut retry_period_ms = 500.0;
         let mut attempts = 0;
-        let max_attempts = 10;
+        const DEFAULT_ATTEMPTS: u16 = 10;
+        #[cfg(feature = "testing")]
+        let max_attempts = if let Ok(v) = env::var("NEON_COMPUTE_TESTING_BASEBACKUP_RETRIES") {
+            u16::from_str(&v).unwrap()
+        } else {
+            DEFAULT_ATTEMPTS
+        };
+        #[cfg(not(feature = "testing"))]
+        let max_attempts = DEFAULT_ATTEMPTS;
         loop {
             let result = self.try_get_basebackup(compute_state, lsn);
             match result {
@@ -798,7 +807,11 @@ impl ComputeNode {
         // In this case we need to connect with old `zenith_admin` name
         // and create new user. We cannot simply rename connected user,
         // but we can create a new one and grant it all privileges.
-        let connstr = self.connstr.clone();
+        let mut connstr = self.connstr.clone();
+        connstr
+            .query_pairs_mut()
+            .append_pair("application_name", "apply_config");
+
         let mut client = match Client::connect(connstr.as_str(), NoTls) {
             Err(e) => match e.code() {
                 Some(&SqlState::INVALID_PASSWORD)
@@ -867,15 +880,19 @@ impl ComputeNode {
 
         // Run migrations separately to not hold up cold starts
         thread::spawn(move || {
+            let mut connstr = connstr.clone();
+            connstr
+                .query_pairs_mut()
+                .append_pair("application_name", "migrations");
+
             let mut client = Client::connect(connstr.as_str(), NoTls)?;
             handle_migrations(&mut client).context("apply_config handle_migrations")
         });
         Ok(())
     }
 
-    // We could've wrapped this around `pg_ctl reload`, but right now we don't use
-    // `pg_ctl` for start / stop, so this just seems much easier to do as we already
-    // have opened connection to Postgres and superuser access.
+    // Wrapped this around `pg_ctl reload`, but right now we don't use
+    // `pg_ctl` for start / stop.
     #[instrument(skip_all)]
     fn pg_reload_conf(&self) -> Result<()> {
         let pgctl_bin = Path::new(&self.pgbin).parent().unwrap().join("pg_ctl");
@@ -1108,7 +1125,7 @@ impl ComputeNode {
     // EKS worker nodes have following core dump settings:
     //   /proc/sys/kernel/core_pattern -> core
     //   /proc/sys/kernel/core_uses_pid -> 1
-    //   ulimint -c -> unlimited
+    //   ulimit -c -> unlimited
     // which results in core dumps being written to postgres data directory as core.<pid>.
     //
     // Use that as a default location and pattern, except macos where core dumps are written
@@ -1387,7 +1404,9 @@ pub fn forward_termination_signal() {
     let pg_pid = PG_PID.load(Ordering::SeqCst);
     if pg_pid != 0 {
         let pg_pid = nix::unistd::Pid::from_raw(pg_pid as i32);
-        // use 'immediate' shutdown (SIGQUIT): https://www.postgresql.org/docs/current/server-shutdown.html
-        kill(pg_pid, Signal::SIGQUIT).ok();
+        // Use 'fast' shutdown (SIGINT) because it also creates a shutdown checkpoint, which is important for
+        // ROs to get a list of running xacts faster instead of going through the CLOG.
+        // See https://www.postgresql.org/docs/current/server-shutdown.html for the list of modes and signals.
+        kill(pg_pid, Signal::SIGINT).ok();
     }
 }

compute_tools/src/config.rs

@@ -83,12 +83,6 @@ pub fn write_postgres_conf(
         ComputeMode::Replica => {
             // hot_standby is 'on' by default, but let's be explicit
             writeln!(file, "hot_standby=on")?;
-
-            // Inform the replica about the primary state
-            // Default is 'false'
-            if let Some(primary_is_running) = spec.primary_is_running {
-                writeln!(file, "neon.primary_is_running={}", primary_is_running)?;
-            }
         }
     }
 

compute_tools/src/lib.rs

@@ -11,6 +11,7 @@ pub mod logger;
 pub mod catalog;
 pub mod compute;
 pub mod extension_server;
+mod migration;
 pub mod monitor;
 pub mod params;
 pub mod pg_helpers;

compute_tools/src/migration.rs

@@ -0,0 +1,105 @@
+use anyhow::{Context, Result};
+use postgres::Client;
+use tracing::info;
+
+pub(crate) struct MigrationRunner<'m> {
+    client: &'m mut Client,
+    migrations: &'m [&'m str],
+}
+
+impl<'m> MigrationRunner<'m> {
+    pub fn new(client: &'m mut Client, migrations: &'m [&'m str]) -> Self {
+        // The neon_migration.migration_id::id column is a bigint, which is equivalent to an i64
+        assert!(migrations.len() + 1 < i64::MAX as usize);
+
+        Self { client, migrations }
+    }
+
+    fn get_migration_id(&mut self) -> Result<i64> {
+        let query = "SELECT id FROM neon_migration.migration_id";
+        let row = self
+            .client
+            .query_one(query, &[])
+            .context("run_migrations get migration_id")?;
+
+        Ok(row.get::<&str, i64>("id"))
+    }
+
+    fn update_migration_id(&mut self, migration_id: i64) -> Result<()> {
+        let setval = format!("UPDATE neon_migration.migration_id SET id={}", migration_id);
+
+        self.client
+            .simple_query(&setval)
+            .context("run_migrations update id")?;
+
+        Ok(())
+    }
+
+    fn prepare_migrations(&mut self) -> Result<()> {
+        let query = "CREATE SCHEMA IF NOT EXISTS neon_migration";
+        self.client.simple_query(query)?;
+
+        let query = "CREATE TABLE IF NOT EXISTS neon_migration.migration_id (key INT NOT NULL PRIMARY KEY, id bigint NOT NULL DEFAULT 0)";
+        self.client.simple_query(query)?;
+
+        let query = "INSERT INTO neon_migration.migration_id VALUES (0, 0) ON CONFLICT DO NOTHING";
+        self.client.simple_query(query)?;
+
+        let query = "ALTER SCHEMA neon_migration OWNER TO cloud_admin";
+        self.client.simple_query(query)?;
+
+        let query = "REVOKE ALL ON SCHEMA neon_migration FROM PUBLIC";
+        self.client.simple_query(query)?;
+
+        Ok(())
+    }
+
+    pub fn run_migrations(mut self) -> Result<()> {
+        self.prepare_migrations()?;
+
+        let mut current_migration = self.get_migration_id()? as usize;
+        while current_migration < self.migrations.len() {
+            macro_rules! migration_id {
+                ($cm:expr) => {
+                    ($cm + 1) as i64
+                };
+            }
+
+            let migration = self.migrations[current_migration];
+
+            if migration.starts_with("-- SKIP") {
+                info!("Skipping migration id={}", migration_id!(current_migration));
+            } else {
+                info!(
+                    "Running migration id={}:\n{}\n",
+                    migration_id!(current_migration),
+                    migration
+                );
+
+                self.client
+                    .simple_query("BEGIN")
+                    .context("begin migration")?;
+
+                self.client.simple_query(migration).with_context(|| {
+                    format!(
+                        "run_migrations migration id={}",
+                        migration_id!(current_migration)
+                    )
+                })?;
+
+                // Migration IDs start at 1
+                self.update_migration_id(migration_id!(current_migration))?;
+
+                self.client
+                    .simple_query("COMMIT")
+                    .context("commit migration")?;
+
+                info!("Finished migration id={}", migration_id!(current_migration));
+            }
+
+            current_migration += 1;
+        }
+
+        Ok(())
+    }
+}

compute_tools/src/migrations/0010-grant_snapshot_synchronization_funcs_to_neon_superuser.sql

@@ -0,0 +1,7 @@
+DO $$
+BEGIN
+    IF (SELECT setting::numeric >= 160000 FROM pg_settings WHERE name = 'server_version_num') THEN
+       EXECUTE 'GRANT EXECUTE ON FUNCTION pg_export_snapshot TO neon_superuser';
+       EXECUTE 'GRANT EXECUTE ON FUNCTION pg_log_standby_snapshot TO neon_superuser';
+    END IF;
+END $$;

compute_tools/src/monitor.rs

@@ -17,7 +17,11 @@ const MONITOR_CHECK_INTERVAL: Duration = Duration::from_millis(500);
 // should be handled gracefully.
 fn watch_compute_activity(compute: &ComputeNode) {
     // Suppose that `connstr` doesn't change
-    let connstr = compute.connstr.as_str();
+    let mut connstr = compute.connstr.clone();
+    connstr
+        .query_pairs_mut()
+        .append_pair("application_name", "compute_activity_monitor");
+    let connstr = connstr.as_str();
 
     // During startup and configuration we connect to every Postgres database,
     // but we don't want to count this as some user activity. So wait until

compute_tools/src/pg_helpers.rs

@@ -489,7 +489,7 @@ pub fn handle_postgres_logs(stderr: std::process::ChildStderr) -> JoinHandle<()>
 /// Read Postgres logs from `stderr` until EOF. Buffer is flushed on one of the following conditions:
 /// - next line starts with timestamp
 /// - EOF
-/// - no new lines were written for the last second
+/// - no new lines were written for the last 100 milliseconds
 async fn handle_postgres_logs_async(stderr: tokio::process::ChildStderr) -> Result<()> {
     let mut lines = tokio::io::BufReader::new(stderr).lines();
     let timeout_duration = Duration::from_millis(100);

compute_tools/src/spec.rs

@@ -10,6 +10,7 @@ use tracing::{error, info, info_span, instrument, span_enabled, warn, Level};
 
 use crate::config;
 use crate::logger::inlinify;
+use crate::migration::MigrationRunner;
 use crate::params::PG_HBA_ALL_MD5;
 use crate::pg_helpers::*;
 
@@ -776,84 +777,25 @@ pub fn handle_migrations(client: &mut Client) -> Result<()> {
 
     // Add new migrations in numerical order.
     let migrations = [
-        include_str!("./migrations/0000-neon_superuser_bypass_rls.sql"),
-        include_str!("./migrations/0001-alter_roles.sql"),
-        include_str!("./migrations/0002-grant_pg_create_subscription_to_neon_superuser.sql"),
-        include_str!("./migrations/0003-grant_pg_monitor_to_neon_superuser.sql"),
-        include_str!("./migrations/0004-grant_all_on_tables_to_neon_superuser.sql"),
-        include_str!("./migrations/0005-grant_all_on_sequences_to_neon_superuser.sql"),
+        include_str!("./migrations/0001-neon_superuser_bypass_rls.sql"),
+        include_str!("./migrations/0002-alter_roles.sql"),
+        include_str!("./migrations/0003-grant_pg_create_subscription_to_neon_superuser.sql"),
+        include_str!("./migrations/0004-grant_pg_monitor_to_neon_superuser.sql"),
+        include_str!("./migrations/0005-grant_all_on_tables_to_neon_superuser.sql"),
+        include_str!("./migrations/0006-grant_all_on_sequences_to_neon_superuser.sql"),
         include_str!(
-            "./migrations/0006-grant_all_on_tables_to_neon_superuser_with_grant_option.sql"
+            "./migrations/0007-grant_all_on_tables_to_neon_superuser_with_grant_option.sql"
         ),
         include_str!(
-            "./migrations/0007-grant_all_on_sequences_to_neon_superuser_with_grant_option.sql"
+            "./migrations/0008-grant_all_on_sequences_to_neon_superuser_with_grant_option.sql"
+        ),
+        include_str!("./migrations/0009-revoke_replication_for_previously_allowed_roles.sql"),
+        include_str!(
+            "./migrations/0010-grant_snapshot_synchronization_funcs_to_neon_superuser.sql"
         ),
-        include_str!("./migrations/0008-revoke_replication_for_previously_allowed_roles.sql"),
     ];
 
-    let mut func = || {
-        let query = "CREATE SCHEMA IF NOT EXISTS neon_migration";
-        client.simple_query(query)?;
-
-        let query = "CREATE TABLE IF NOT EXISTS neon_migration.migration_id (key INT NOT NULL PRIMARY KEY, id bigint NOT NULL DEFAULT 0)";
-        client.simple_query(query)?;
-
-        let query = "INSERT INTO neon_migration.migration_id VALUES (0, 0) ON CONFLICT DO NOTHING";
-        client.simple_query(query)?;
-
-        let query = "ALTER SCHEMA neon_migration OWNER TO cloud_admin";
-        client.simple_query(query)?;
-
-        let query = "REVOKE ALL ON SCHEMA neon_migration FROM PUBLIC";
-        client.simple_query(query)?;
-        Ok::<_, anyhow::Error>(())
-    };
-    func().context("handle_migrations prepare")?;
-
-    let query = "SELECT id FROM neon_migration.migration_id";
-    let row = client
-        .query_one(query, &[])
-        .context("handle_migrations get migration_id")?;
-    let mut current_migration: usize = row.get::<&str, i64>("id") as usize;
-    let starting_migration_id = current_migration;
-
-    let query = "BEGIN";
-    client
-        .simple_query(query)
-        .context("handle_migrations begin")?;
-
-    while current_migration < migrations.len() {
-        let migration = &migrations[current_migration];
-        if migration.starts_with("-- SKIP") {
-            info!("Skipping migration id={}", current_migration);
-        } else {
-            info!(
-                "Running migration id={}:\n{}\n",
-                current_migration, migration
-            );
-            client.simple_query(migration).with_context(|| {
-                format!("handle_migrations current_migration={}", current_migration)
-            })?;
-        }
-        current_migration += 1;
-    }
-    let setval = format!(
-        "UPDATE neon_migration.migration_id SET id={}",
-        migrations.len()
-    );
-    client
-        .simple_query(&setval)
-        .context("handle_migrations update id")?;
-
-    let query = "COMMIT";
-    client
-        .simple_query(query)
-        .context("handle_migrations commit")?;
-
-    info!(
-        "Ran {} migrations",
-        (migrations.len() - starting_migration_id)
-    );
+    MigrationRunner::new(client, &migrations).run_migrations()?;
 
     Ok(())
 }

control_plane/Cargo.toml

@@ -40,6 +40,7 @@ safekeeper_api.workspace = true
 postgres_connection.workspace = true
 storage_broker.workspace = true
 utils.workspace = true
+whoami.workspace = true
 
 compute_api.workspace = true
 workspace_hack.workspace = true

control_plane/src/background_process.rs

@@ -36,11 +36,11 @@ use utils::pid_file::{self, PidFileRead};
 // it's waiting. If the process hasn't started/stopped after 5 seconds,
 // it prints a notice that it's taking long, but keeps waiting.
 //
-const RETRY_UNTIL_SECS: u64 = 10;
-const RETRIES: u64 = (RETRY_UNTIL_SECS * 1000) / RETRY_INTERVAL_MILLIS;
-const RETRY_INTERVAL_MILLIS: u64 = 100;
-const DOT_EVERY_RETRIES: u64 = 10;
-const NOTICE_AFTER_RETRIES: u64 = 50;
+const STOP_RETRY_TIMEOUT: Duration = Duration::from_secs(10);
+const STOP_RETRIES: u128 = STOP_RETRY_TIMEOUT.as_millis() / RETRY_INTERVAL.as_millis();
+const RETRY_INTERVAL: Duration = Duration::from_millis(100);
+const DOT_EVERY_RETRIES: u128 = 10;
+const NOTICE_AFTER_RETRIES: u128 = 50;
 
 /// Argument to `start_process`, to indicate whether it should create pidfile or if the process creates
 /// it itself.
@@ -52,6 +52,7 @@ pub enum InitialPidFile {
 }
 
 /// Start a background child process using the parameters given.
+#[allow(clippy::too_many_arguments)]
 pub async fn start_process<F, Fut, AI, A, EI>(
     process_name: &str,
     datadir: &Path,
@@ -59,6 +60,7 @@ pub async fn start_process<F, Fut, AI, A, EI>(
     args: AI,
     envs: EI,
     initial_pid_file: InitialPidFile,
+    retry_timeout: &Duration,
     process_status_check: F,
 ) -> anyhow::Result<()>
 where
@@ -69,6 +71,10 @@ where
     // Not generic AsRef<OsStr>, otherwise empty `envs` prevents type inference
     EI: IntoIterator<Item = (String, String)>,
 {
+    let retries: u128 = retry_timeout.as_millis() / RETRY_INTERVAL.as_millis();
+    if !datadir.metadata().context("stat datadir")?.is_dir() {
+        anyhow::bail!("`datadir` must be a directory when calling this function: {datadir:?}");
+    }
     let log_path = datadir.join(format!("{process_name}.log"));
     let process_log_file = fs::OpenOptions::new()
         .create(true)
@@ -85,7 +91,13 @@ where
     let background_command = command
         .stdout(process_log_file)
         .stderr(same_file_for_stderr)
-        .args(args);
+        .args(args)
+        // spawn all child processes in their datadir, useful for all kinds of things,
+        // not least cleaning up child processes e.g. after an unclean exit from the test suite:
+        // ```
+        // lsof  -d cwd -a +D  Users/cs/src/neon/test_output
+        // ```
+        .current_dir(datadir);
 
     let filled_cmd = fill_env_vars_prefixed_neon(fill_remote_storage_secrets_vars(
         fill_rust_env_vars(background_command),
@@ -121,7 +133,7 @@ where
         .unwrap();
     });
 
-    for retries in 0..RETRIES {
+    for retries in 0..retries {
         match process_started(pid, pid_file_to_check, &process_status_check).await {
             Ok(true) => {
                 println!("\n{process_name} started and passed status check, pid: {pid}");
@@ -139,7 +151,7 @@ where
                     print!(".");
                     io::stdout().flush().unwrap();
                 }
-                thread::sleep(Duration::from_millis(RETRY_INTERVAL_MILLIS));
+                thread::sleep(RETRY_INTERVAL);
             }
             Err(e) => {
                 println!("error starting process {process_name:?}: {e:#}");
@@ -148,9 +160,10 @@ where
         }
     }
     println!();
-    anyhow::bail!(
-        "{process_name} did not start+pass status checks within {RETRY_UNTIL_SECS} seconds"
-    );
+    anyhow::bail!(format!(
+        "{} did not start+pass status checks within {:?} seconds",
+        process_name, retry_timeout
+    ));
 }
 
 /// Stops the process, using the pid file given. Returns Ok also if the process is already not running.
@@ -206,7 +219,7 @@ pub fn stop_process(
 }
 
 pub fn wait_until_stopped(process_name: &str, pid: Pid) -> anyhow::Result<()> {
-    for retries in 0..RETRIES {
+    for retries in 0..STOP_RETRIES {
         match process_has_stopped(pid) {
             Ok(true) => {
                 println!("\n{process_name} stopped");
@@ -222,7 +235,7 @@ pub fn wait_until_stopped(process_name: &str, pid: Pid) -> anyhow::Result<()> {
                     print!(".");
                     io::stdout().flush().unwrap();
                 }
-                thread::sleep(Duration::from_millis(RETRY_INTERVAL_MILLIS));
+                thread::sleep(RETRY_INTERVAL);
             }
             Err(e) => {
                 println!("{process_name} with pid {pid} failed to stop: {e:#}");
@@ -231,7 +244,10 @@ pub fn wait_until_stopped(process_name: &str, pid: Pid) -> anyhow::Result<()> {
         }
     }
     println!();
-    anyhow::bail!("{process_name} with pid {pid} did not stop in {RETRY_UNTIL_SECS} seconds");
+    anyhow::bail!(format!(
+        "{} with pid {} did not stop in {:?} seconds",
+        process_name, pid, STOP_RETRY_TIMEOUT
+    ));
 }
 
 fn fill_rust_env_vars(cmd: &mut Command) -> &mut Command {
@@ -273,7 +289,7 @@ fn fill_remote_storage_secrets_vars(mut cmd: &mut Command) -> &mut Command {
 
 fn fill_env_vars_prefixed_neon(mut cmd: &mut Command) -> &mut Command {
     for (var, val) in std::env::vars() {
-        if var.starts_with("NEON_PAGESERVER_") {
+        if var.starts_with("NEON_") {
             cmd = cmd.env(var, val);
         }
     }
@@ -363,7 +379,7 @@ where
     }
 }
 
-fn process_has_stopped(pid: Pid) -> anyhow::Result<bool> {
+pub(crate) fn process_has_stopped(pid: Pid) -> anyhow::Result<bool> {
     match kill(pid, None) {
         // Process exists, keep waiting
         Ok(_) => Ok(false),

control_plane/src/bin/neon_local.rs

@@ -15,16 +15,18 @@ use control_plane::local_env::{
 };
 use control_plane::pageserver::PageServerNode;
 use control_plane::safekeeper::SafekeeperNode;
-use control_plane::storage_controller::StorageController;
+use control_plane::storage_controller::{
+    NeonStorageControllerStartArgs, NeonStorageControllerStopArgs, StorageController,
+};
 use control_plane::{broker, local_env};
 use pageserver_api::config::{
     DEFAULT_HTTP_LISTEN_PORT as DEFAULT_PAGESERVER_HTTP_PORT,
     DEFAULT_PG_LISTEN_PORT as DEFAULT_PAGESERVER_PG_PORT,
 };
-use pageserver_api::controller_api::PlacementPolicy;
-use pageserver_api::models::{
-    ShardParameters, TenantCreateRequest, TimelineCreateRequest, TimelineInfo,
+use pageserver_api::controller_api::{
+    NodeAvailabilityWrapper, PlacementPolicy, TenantCreateRequest,
 };
+use pageserver_api::models::{ShardParameters, TimelineCreateRequest, TimelineInfo};
 use pageserver_api::shard::{ShardCount, ShardStripeSize, TenantShardId};
 use postgres_backend::AuthType;
 use postgres_connection::parse_host_port;
@@ -36,6 +38,7 @@ use std::collections::{BTreeSet, HashMap};
 use std::path::PathBuf;
 use std::process::exit;
 use std::str::FromStr;
+use std::time::Duration;
 use storage_broker::DEFAULT_LISTEN_ADDR as DEFAULT_BROKER_ADDR;
 use url::Host;
 use utils::{
@@ -51,7 +54,7 @@ const DEFAULT_PAGESERVER_ID: NodeId = NodeId(1);
 const DEFAULT_BRANCH_NAME: &str = "main";
 project_git_version!(GIT_VERSION);
 
-const DEFAULT_PG_VERSION: &str = "15";
+const DEFAULT_PG_VERSION: &str = "16";
 
 const DEFAULT_PAGESERVER_CONTROL_PLANE_API: &str = "http://127.0.0.1:1234/upcall/v1/";
 
@@ -87,7 +90,8 @@ fn main() -> Result<()> {
         handle_init(sub_args).map(Some)
     } else {
         // all other commands need an existing config
-        let mut env = LocalEnv::load_config().context("Error loading config")?;
+        let mut env =
+            LocalEnv::load_config(&local_env::base_path()).context("Error loading config")?;
         let original_env = env.clone();
 
         let rt = tokio::runtime::Builder::new_current_thread()
@@ -98,7 +102,7 @@ fn main() -> Result<()> {
         let subcommand_result = match sub_name {
             "tenant" => rt.block_on(handle_tenant(sub_args, &mut env)),
             "timeline" => rt.block_on(handle_timeline(sub_args, &mut env)),
-            "start" => rt.block_on(handle_start_all(&env)),
+            "start" => rt.block_on(handle_start_all(&env, get_start_timeout(sub_args))),
             "stop" => rt.block_on(handle_stop_all(sub_args, &env)),
             "pageserver" => rt.block_on(handle_pageserver(sub_args, &env)),
             "storage_controller" => rt.block_on(handle_storage_controller(sub_args, &env)),
@@ -364,7 +368,8 @@ fn handle_init(init_match: &ArgMatches) -> anyhow::Result<LocalEnv> {
 
     LocalEnv::init(init_conf, force)
         .context("materialize initial neon_local environment on disk")?;
-    Ok(LocalEnv::load_config().expect("freshly written config should be loadable"))
+    Ok(LocalEnv::load_config(&local_env::base_path())
+        .expect("freshly written config should be loadable"))
 }
 
 /// The default pageserver is the one where CLI tenant/timeline operations are sent by default.
@@ -597,13 +602,9 @@ async fn handle_timeline(timeline_match: &ArgMatches, env: &mut local_env::Local
         Some(("import", import_match)) => {
             let tenant_id = get_tenant_id(import_match, env)?;
             let timeline_id = parse_timeline_id(import_match)?.expect("No timeline id provided");
-            let name = import_match
-                .get_one::<String>("node-name")
-                .ok_or_else(|| anyhow!("No node name provided"))?;
-            let update_catalog = import_match
-                .get_one::<bool>("update-catalog")
-                .cloned()
-                .unwrap_or_default();
+            let branch_name = import_match
+                .get_one::<String>("branch-name")
+                .ok_or_else(|| anyhow!("No branch name provided"))?;
 
             // Parse base inputs
             let base_tarfile = import_match
@@ -630,24 +631,11 @@ async fn handle_timeline(timeline_match: &ArgMatches, env: &mut local_env::Local
                 .copied()
                 .context("Failed to parse postgres version from the argument string")?;
 
-            let mut cplane = ComputeControlPlane::load(env.clone())?;
             println!("Importing timeline into pageserver ...");
             pageserver
                 .timeline_import(tenant_id, timeline_id, base, pg_wal, pg_version)
                 .await?;
-            env.register_branch_mapping(name.to_string(), tenant_id, timeline_id)?;
-
-            println!("Creating endpoint for imported timeline ...");
-            cplane.new_endpoint(
-                name,
-                tenant_id,
-                timeline_id,
-                None,
-                None,
-                pg_version,
-                ComputeMode::Primary,
-                !update_catalog,
-            )?;
+            env.register_branch_mapping(branch_name.to_string(), tenant_id, timeline_id)?;
             println!("Done");
         }
         Some(("branch", branch_match)) => {
@@ -862,20 +850,13 @@ async fn handle_endpoint(ep_match: &ArgMatches, env: &local_env::LocalEnv) -> Re
 
             let allow_multiple = sub_args.get_flag("allow-multiple");
 
-            // If --safekeepers argument is given, use only the listed safekeeper nodes.
-            let safekeepers =
-                if let Some(safekeepers_str) = sub_args.get_one::<String>("safekeepers") {
-                    let mut safekeepers: Vec<NodeId> = Vec::new();
-                    for sk_id in safekeepers_str.split(',').map(str::trim) {
-                        let sk_id = NodeId(u64::from_str(sk_id).map_err(|_| {
-                            anyhow!("invalid node ID \"{sk_id}\" in --safekeepers list")
-                        })?);
-                        safekeepers.push(sk_id);
-                    }
-                    safekeepers
-                } else {
-                    env.safekeepers.iter().map(|sk| sk.id).collect()
-                };
+            // If --safekeepers argument is given, use only the listed
+            // safekeeper nodes; otherwise all from the env.
+            let safekeepers = if let Some(safekeepers) = parse_safekeepers(sub_args)? {
+                safekeepers
+            } else {
+                env.safekeepers.iter().map(|sk| sk.id).collect()
+            };
 
             let endpoint = cplane
                 .endpoints
@@ -979,7 +960,10 @@ async fn handle_endpoint(ep_match: &ArgMatches, env: &local_env::LocalEnv) -> Re
                         })
                         .collect::<Vec<_>>()
                 };
-            endpoint.reconfigure(pageservers, None).await?;
+            // If --safekeepers argument is given, use only the listed
+            // safekeeper nodes; otherwise all from the env.
+            let safekeepers = parse_safekeepers(sub_args)?;
+            endpoint.reconfigure(pageservers, None, safekeepers).await?;
         }
         "stop" => {
             let endpoint_id = sub_args
@@ -1001,6 +985,23 @@ async fn handle_endpoint(ep_match: &ArgMatches, env: &local_env::LocalEnv) -> Re
     Ok(())
 }
 
+/// Parse --safekeepers as list of safekeeper ids.
+fn parse_safekeepers(sub_args: &ArgMatches) -> Result<Option<Vec<NodeId>>> {
+    if let Some(safekeepers_str) = sub_args.get_one::<String>("safekeepers") {
+        let mut safekeepers: Vec<NodeId> = Vec::new();
+        for sk_id in safekeepers_str.split(',').map(str::trim) {
+            let sk_id = NodeId(
+                u64::from_str(sk_id)
+                    .map_err(|_| anyhow!("invalid node ID \"{sk_id}\" in --safekeepers list"))?,
+            );
+            safekeepers.push(sk_id);
+        }
+        Ok(Some(safekeepers))
+    } else {
+        Ok(None)
+    }
+}
+
 fn handle_mappings(sub_match: &ArgMatches, env: &mut local_env::LocalEnv) -> Result<()> {
     let (sub_name, sub_args) = match sub_match.subcommand() {
         Some(ep_subcommand_data) => ep_subcommand_data,
@@ -1046,10 +1047,50 @@ fn get_pageserver(env: &local_env::LocalEnv, args: &ArgMatches) -> Result<PageSe
     ))
 }
 
+fn get_start_timeout(args: &ArgMatches) -> &Duration {
+    let humantime_duration = args
+        .get_one::<humantime::Duration>("start-timeout")
+        .expect("invalid value for start-timeout");
+    humantime_duration.as_ref()
+}
+
+fn storage_controller_start_args(args: &ArgMatches) -> NeonStorageControllerStartArgs {
+    let maybe_instance_id = args.get_one::<u8>("instance-id");
+
+    let base_port = args.get_one::<u16>("base-port");
+
+    if maybe_instance_id.is_some() && base_port.is_none() {
+        panic!("storage-controller start specificied instance-id but did not provide base-port");
+    }
+
+    let start_timeout = args
+        .get_one::<humantime::Duration>("start-timeout")
+        .expect("invalid value for start-timeout");
+
+    NeonStorageControllerStartArgs {
+        instance_id: maybe_instance_id.copied().unwrap_or(1),
+        base_port: base_port.copied(),
+        start_timeout: *start_timeout,
+    }
+}
+
+fn storage_controller_stop_args(args: &ArgMatches) -> NeonStorageControllerStopArgs {
+    let maybe_instance_id = args.get_one::<u8>("instance-id");
+    let immediate = args.get_one::<String>("stop-mode").map(|s| s.as_str()) == Some("immediate");
+
+    NeonStorageControllerStopArgs {
+        instance_id: maybe_instance_id.copied().unwrap_or(1),
+        immediate,
+    }
+}
+
 async fn handle_pageserver(sub_match: &ArgMatches, env: &local_env::LocalEnv) -> Result<()> {
     match sub_match.subcommand() {
         Some(("start", subcommand_args)) => {
-            if let Err(e) = get_pageserver(env, subcommand_args)?.start().await {
+            if let Err(e) = get_pageserver(env, subcommand_args)?
+                .start(get_start_timeout(subcommand_args))
+                .await
+            {
                 eprintln!("pageserver start failed: {e}");
                 exit(1);
             }
@@ -1075,7 +1116,7 @@ async fn handle_pageserver(sub_match: &ArgMatches, env: &local_env::LocalEnv) ->
                 exit(1);
             }
 
-            if let Err(e) = pageserver.start().await {
+            if let Err(e) = pageserver.start(get_start_timeout(sub_match)).await {
                 eprintln!("pageserver start failed: {e}");
                 exit(1);
             }
@@ -1103,20 +1144,15 @@ async fn handle_storage_controller(
 ) -> Result<()> {
     let svc = StorageController::from_env(env);
     match sub_match.subcommand() {
-        Some(("start", _start_match)) => {
-            if let Err(e) = svc.start().await {
+        Some(("start", start_match)) => {
+            if let Err(e) = svc.start(storage_controller_start_args(start_match)).await {
                 eprintln!("start failed: {e}");
                 exit(1);
             }
         }
 
         Some(("stop", stop_match)) => {
-            let immediate = stop_match
-                .get_one::<String>("stop-mode")
-                .map(|s| s.as_str())
-                == Some("immediate");
-
-            if let Err(e) = svc.stop(immediate).await {
+            if let Err(e) = svc.stop(storage_controller_stop_args(stop_match)).await {
                 eprintln!("stop failed: {}", e);
                 exit(1);
             }
@@ -1163,7 +1199,10 @@ async fn handle_safekeeper(sub_match: &ArgMatches, env: &local_env::LocalEnv) ->
         "start" => {
             let extra_opts = safekeeper_extra_opts(sub_args);
 
-            if let Err(e) = safekeeper.start(extra_opts).await {
+            if let Err(e) = safekeeper
+                .start(extra_opts, get_start_timeout(sub_args))
+                .await
+            {
                 eprintln!("safekeeper start failed: {}", e);
                 exit(1);
             }
@@ -1189,7 +1228,10 @@ async fn handle_safekeeper(sub_match: &ArgMatches, env: &local_env::LocalEnv) ->
             }
 
             let extra_opts = safekeeper_extra_opts(sub_args);
-            if let Err(e) = safekeeper.start(extra_opts).await {
+            if let Err(e) = safekeeper
+                .start(extra_opts, get_start_timeout(sub_args))
+                .await
+            {
                 eprintln!("safekeeper start failed: {}", e);
                 exit(1);
             }
@@ -1202,15 +1244,23 @@ async fn handle_safekeeper(sub_match: &ArgMatches, env: &local_env::LocalEnv) ->
     Ok(())
 }
 
-async fn handle_start_all(env: &local_env::LocalEnv) -> anyhow::Result<()> {
+async fn handle_start_all(
+    env: &local_env::LocalEnv,
+    retry_timeout: &Duration,
+) -> anyhow::Result<()> {
     // Endpoints are not started automatically
 
-    broker::start_broker_process(env).await?;
+    broker::start_broker_process(env, retry_timeout).await?;
 
     // Only start the storage controller if the pageserver is configured to need it
     if env.control_plane_api.is_some() {
         let storage_controller = StorageController::from_env(env);
-        if let Err(e) = storage_controller.start().await {
+        if let Err(e) = storage_controller
+            .start(NeonStorageControllerStartArgs::with_default_instance_id(
+                (*retry_timeout).into(),
+            ))
+            .await
+        {
             eprintln!("storage_controller start failed: {:#}", e);
             try_stop_all(env, true).await;
             exit(1);
@@ -1219,7 +1269,7 @@ async fn handle_start_all(env: &local_env::LocalEnv) -> anyhow::Result<()> {
 
     for ps_conf in &env.pageservers {
         let pageserver = PageServerNode::from_env(env, ps_conf);
-        if let Err(e) = pageserver.start().await {
+        if let Err(e) = pageserver.start(retry_timeout).await {
             eprintln!("pageserver {} start failed: {:#}", ps_conf.id, e);
             try_stop_all(env, true).await;
             exit(1);
@@ -1228,15 +1278,76 @@ async fn handle_start_all(env: &local_env::LocalEnv) -> anyhow::Result<()> {
 
     for node in env.safekeepers.iter() {
         let safekeeper = SafekeeperNode::from_env(env, node);
-        if let Err(e) = safekeeper.start(vec![]).await {
+        if let Err(e) = safekeeper.start(vec![], retry_timeout).await {
             eprintln!("safekeeper {} start failed: {:#}", safekeeper.id, e);
             try_stop_all(env, false).await;
             exit(1);
         }
     }
+
+    neon_start_status_check(env, retry_timeout).await?;
+
     Ok(())
 }
 
+async fn neon_start_status_check(
+    env: &local_env::LocalEnv,
+    retry_timeout: &Duration,
+) -> anyhow::Result<()> {
+    const RETRY_INTERVAL: Duration = Duration::from_millis(100);
+    const NOTICE_AFTER_RETRIES: Duration = Duration::from_secs(5);
+
+    if env.control_plane_api.is_none() {
+        return Ok(());
+    }
+
+    let storcon = StorageController::from_env(env);
+
+    let retries = retry_timeout.as_millis() / RETRY_INTERVAL.as_millis();
+    let notice_after_retries = retry_timeout.as_millis() / NOTICE_AFTER_RETRIES.as_millis();
+
+    println!("\nRunning neon status check");
+
+    for retry in 0..retries {
+        if retry == notice_after_retries {
+            println!("\nNeon status check has not passed yet, continuing to wait")
+        }
+
+        let mut passed = true;
+        let mut nodes = storcon.node_list().await?;
+        let mut pageservers = env.pageservers.clone();
+
+        if nodes.len() != pageservers.len() {
+            continue;
+        }
+
+        nodes.sort_by_key(|ps| ps.id);
+        pageservers.sort_by_key(|ps| ps.id);
+
+        for (idx, pageserver) in pageservers.iter().enumerate() {
+            let node = &nodes[idx];
+            if node.id != pageserver.id {
+                passed = false;
+                break;
+            }
+
+            if !matches!(node.availability, NodeAvailabilityWrapper::Active) {
+                passed = false;
+                break;
+            }
+        }
+
+        if passed {
+            println!("\nNeon started and passed status check");
+            return Ok(());
+        }
+
+        tokio::time::sleep(RETRY_INTERVAL).await;
+    }
+
+    anyhow::bail!("\nNeon passed status check")
+}
+
 async fn handle_stop_all(sub_match: &ArgMatches, env: &local_env::LocalEnv) -> Result<()> {
     let immediate =
         sub_match.get_one::<String>("stop-mode").map(|s| s.as_str()) == Some("immediate");
@@ -1279,15 +1390,35 @@ async fn try_stop_all(env: &local_env::LocalEnv, immediate: bool) {
         eprintln!("neon broker stop failed: {e:#}");
     }
 
-    if env.control_plane_api.is_some() {
+    // Stop all storage controller instances. In the most common case there's only one,
+    // but iterate though the base data directory in order to discover the instances.
+    let storcon_instances = env
+        .storage_controller_instances()
+        .await
+        .expect("Must inspect data dir");
+    for (instance_id, _instance_dir_path) in storcon_instances {
         let storage_controller = StorageController::from_env(env);
-        if let Err(e) = storage_controller.stop(immediate).await {
-            eprintln!("storage controller stop failed: {e:#}");
+        let stop_args = NeonStorageControllerStopArgs {
+            instance_id,
+            immediate,
+        };
+
+        if let Err(e) = storage_controller.stop(stop_args).await {
+            eprintln!("Storage controller instance {instance_id} stop failed: {e:#}");
         }
     }
 }
 
 fn cli() -> Command {
+    let timeout_arg = Arg::new("start-timeout")
+        .long("start-timeout")
+        .short('t')
+        .global(true)
+        .help("timeout until we fail the command, e.g. 30s")
+        .value_parser(value_parser!(humantime::Duration))
+        .default_value("10s")
+        .required(false);
+
     let branch_name_arg = Arg::new("branch-name")
         .long("branch-name")
         .help("Name of the branch to be created or used as an alias for other services")
@@ -1413,6 +1544,18 @@ fn cli() -> Command {
         .action(ArgAction::SetTrue)
         .required(false);
 
+    let instance_id = Arg::new("instance-id")
+        .long("instance-id")
+        .help("Identifier used to distinguish storage controller instances (default 1)")
+        .value_parser(value_parser!(u8))
+        .required(false);
+
+    let base_port = Arg::new("base-port")
+        .long("base-port")
+        .help("Base port for the storage controller instance idenfified by instance-id (defaults to pagserver cplane api)")
+        .value_parser(value_parser!(u16))
+        .required(false);
+
     Command::new("Neon CLI")
         .arg_required_else_help(true)
         .version(GIT_VERSION)
@@ -1456,8 +1599,7 @@ fn cli() -> Command {
                 .about("Import timeline from basebackup directory")
                 .arg(tenant_id_arg.clone())
                 .arg(timeline_id_arg.clone())
-                .arg(Arg::new("node-name").long("node-name")
-                    .help("Name to assign to the imported timeline"))
+                .arg(branch_name_arg.clone())
                 .arg(Arg::new("base-tarfile")
                     .long("base-tarfile")
                     .value_parser(value_parser!(PathBuf))
@@ -1473,7 +1615,6 @@ fn cli() -> Command {
                 .arg(Arg::new("end-lsn").long("end-lsn")
                     .help("Lsn the basebackup ends at"))
                 .arg(pg_version_arg.clone())
-                .arg(update_catalog.clone())
             )
         ).subcommand(
             Command::new("tenant")
@@ -1507,6 +1648,7 @@ fn cli() -> Command {
                 .subcommand(Command::new("status"))
                 .subcommand(Command::new("start")
                     .about("Start local pageserver")
+                    .arg(timeout_arg.clone())
                 )
                 .subcommand(Command::new("stop")
                     .about("Stop local pageserver")
@@ -1514,15 +1656,20 @@ fn cli() -> Command {
                 )
                 .subcommand(Command::new("restart")
                     .about("Restart local pageserver")
+                    .arg(timeout_arg.clone())
                 )
         )
         .subcommand(
             Command::new("storage_controller")
                 .arg_required_else_help(true)
                 .about("Manage storage_controller")
-                .subcommand(Command::new("start").about("Start storage controller"))
+                .subcommand(Command::new("start").about("Start storage controller")
+                            .arg(timeout_arg.clone())
+                            .arg(instance_id.clone())
+                            .arg(base_port))
                 .subcommand(Command::new("stop").about("Stop storage controller")
-                            .arg(stop_mode_arg.clone()))
+                            .arg(stop_mode_arg.clone())
+                            .arg(instance_id))
         )
         .subcommand(
             Command::new("safekeeper")
@@ -1532,6 +1679,7 @@ fn cli() -> Command {
                             .about("Start local safekeeper")
                             .arg(safekeeper_id_arg.clone())
                             .arg(safekeeper_extra_opt_arg.clone())
+                            .arg(timeout_arg.clone())
                 )
                 .subcommand(Command::new("stop")
                             .about("Stop local safekeeper")
@@ -1543,6 +1691,7 @@ fn cli() -> Command {
                             .arg(safekeeper_id_arg)
                             .arg(stop_mode_arg.clone())
                             .arg(safekeeper_extra_opt_arg)
+                            .arg(timeout_arg.clone())
                 )
         )
         .subcommand(
@@ -1573,14 +1722,16 @@ fn cli() -> Command {
                     .about("Start postgres.\n If the endpoint doesn't exist yet, it is created.")
                     .arg(endpoint_id_arg.clone())
                     .arg(endpoint_pageserver_id_arg.clone())
-                    .arg(safekeepers_arg)
+                    .arg(safekeepers_arg.clone())
                     .arg(remote_ext_config_args)
                     .arg(create_test_user)
                     .arg(allow_multiple.clone())
+                    .arg(timeout_arg.clone())
                 )
                 .subcommand(Command::new("reconfigure")
                             .about("Reconfigure the endpoint")
                             .arg(endpoint_pageserver_id_arg)
+                            .arg(safekeepers_arg)
                             .arg(endpoint_id_arg.clone())
                             .arg(tenant_id_arg.clone())
                 )
@@ -1628,6 +1779,7 @@ fn cli() -> Command {
         .subcommand(
             Command::new("start")
                 .about("Start page server and safekeepers")
+                .arg(timeout_arg.clone())
         )
         .subcommand(
             Command::new("stop")

control_plane/src/broker.rs

@@ -1,17 +1,22 @@
 //! Code to manage the storage broker
 //!
-//! In the local test environment, the data for each safekeeper is stored in
+//! In the local test environment, the storage broker stores its data directly in
 //!
 //! ```text
-//!   .neon/safekeepers/<safekeeper id>
+//!   .neon
 //! ```
+use std::time::Duration;
+
 use anyhow::Context;
 
 use camino::Utf8PathBuf;
 
 use crate::{background_process, local_env};
 
-pub async fn start_broker_process(env: &local_env::LocalEnv) -> anyhow::Result<()> {
+pub async fn start_broker_process(
+    env: &local_env::LocalEnv,
+    retry_timeout: &Duration,
+) -> anyhow::Result<()> {
     let broker = &env.broker;
     let listen_addr = &broker.listen_addr;
 
@@ -27,6 +32,7 @@ pub async fn start_broker_process(env: &local_env::LocalEnv) -> anyhow::Result<(
         args,
         [],
         background_process::InitialPidFile::Create(storage_broker_pid_file_path(env)),
+        retry_timeout,
         || async {
             let url = broker.client_url();
             let status_url = url.join("status").with_context(|| {

control_plane/src/endpoint.rs

@@ -499,6 +499,23 @@ impl Endpoint {
             .join(",")
     }
 
+    /// Map safekeepers ids to the actual connection strings.
+    fn build_safekeepers_connstrs(&self, sk_ids: Vec<NodeId>) -> Result<Vec<String>> {
+        let mut safekeeper_connstrings = Vec::new();
+        if self.mode == ComputeMode::Primary {
+            for sk_id in sk_ids {
+                let sk = self
+                    .env
+                    .safekeepers
+                    .iter()
+                    .find(|node| node.id == sk_id)
+                    .ok_or_else(|| anyhow!("safekeeper {sk_id} does not exist"))?;
+                safekeeper_connstrings.push(format!("127.0.0.1:{}", sk.get_compute_port()));
+            }
+        }
+        Ok(safekeeper_connstrings)
+    }
+
     pub async fn start(
         &self,
         auth_token: &Option<String>,
@@ -523,18 +540,7 @@ impl Endpoint {
         let pageserver_connstring = Self::build_pageserver_connstr(&pageservers);
         assert!(!pageserver_connstring.is_empty());
 
-        let mut safekeeper_connstrings = Vec::new();
-        if self.mode == ComputeMode::Primary {
-            for sk_id in safekeepers {
-                let sk = self
-                    .env
-                    .safekeepers
-                    .iter()
-                    .find(|node| node.id == sk_id)
-                    .ok_or_else(|| anyhow!("safekeeper {sk_id} does not exist"))?;
-                safekeeper_connstrings.push(format!("127.0.0.1:{}", sk.get_compute_port()));
-            }
-        }
+        let safekeeper_connstrings = self.build_safekeepers_connstrs(safekeepers)?;
 
         // check for file remote_extensions_spec.json
         // if it is present, read it and pass to compute_ctl
@@ -592,7 +598,6 @@ impl Endpoint {
             remote_extensions,
             pgbouncer_settings: None,
             shard_stripe_size: Some(shard_stripe_size),
-            primary_is_running: None,
         };
         let spec_path = self.endpoint_path().join("spec.json");
         std::fs::write(spec_path, serde_json::to_string_pretty(&spec)?)?;
@@ -741,6 +746,7 @@ impl Endpoint {
         &self,
         mut pageservers: Vec<(Host, u16)>,
         stripe_size: Option<ShardStripeSize>,
+        safekeepers: Option<Vec<NodeId>>,
     ) -> Result<()> {
         let mut spec: ComputeSpec = {
             let spec_path = self.endpoint_path().join("spec.json");
@@ -775,6 +781,12 @@ impl Endpoint {
             spec.shard_stripe_size = stripe_size.map(|s| s.0 as usize);
         }
 
+        // If safekeepers are not specified, don't change them.
+        if let Some(safekeepers) = safekeepers {
+            let safekeeper_connstrings = self.build_safekeepers_connstrs(safekeepers)?;
+            spec.safekeeper_connstrings = safekeeper_connstrings;
+        }
+
         let client = reqwest::Client::builder()
             .timeout(Duration::from_secs(30))
             .build()
@@ -812,11 +824,12 @@ impl Endpoint {
         // cleanup work to do after postgres stops, like syncing safekeepers,
         // etc.
         //
-        // If destroying, send it SIGTERM before waiting. Sometimes we do *not*
-        // want this cleanup: tests intentionally do stop when majority of
-        // safekeepers is down, so sync-safekeepers would hang otherwise. This
-        // could be a separate flag though.
-        self.wait_for_compute_ctl_to_exit(destroy)?;
+        // If destroying or stop mode is immediate, send it SIGTERM before
+        // waiting. Sometimes we do *not* want this cleanup: tests intentionally
+        // do stop when majority of safekeepers is down, so sync-safekeepers
+        // would hang otherwise. This could be a separate flag though.
+        let send_sigterm = destroy || mode == "immediate";
+        self.wait_for_compute_ctl_to_exit(send_sigterm)?;
         if destroy {
             println!(
                 "Destroying postgres data directory '{}'",

control_plane/src/local_env.rs

@@ -27,7 +27,7 @@ use crate::pageserver::PageServerNode;
 use crate::pageserver::PAGESERVER_REMOTE_STORAGE_DIR;
 use crate::safekeeper::SafekeeperNode;
 
-pub const DEFAULT_PG_VERSION: u32 = 15;
+pub const DEFAULT_PG_VERSION: u32 = 16;
 
 //
 // This data structures represents neon_local CLI config
@@ -42,8 +42,8 @@ pub struct LocalEnv {
     // compute endpoints).
     //
     // This is not stored in the config file. Rather, this is the path where the
-    // config file itself is. It is read from the NEON_REPO_DIR env variable or
-    // '.neon' if not given.
+    // config file itself is. It is read from the NEON_REPO_DIR env variable which
+    // must be an absolute path. If the env var is not set, $PWD/.neon is used.
     pub base_data_dir: PathBuf,
 
     // Path to postgres distribution. It's expected that "bin", "include",
@@ -151,23 +151,38 @@ pub struct NeonBroker {
 pub struct NeonStorageControllerConf {
     /// Heartbeat timeout before marking a node offline
     #[serde(with = "humantime_serde")]
-    pub max_unavailable: Duration,
+    pub max_offline: Duration,
+
+    #[serde(with = "humantime_serde")]
+    pub max_warming_up: Duration,
+
+    pub start_as_candidate: bool,
+
+    /// Database url used when running multiple storage controller instances
+    pub database_url: Option<SocketAddr>,
 
     /// Threshold for auto-splitting a tenant into shards
     pub split_threshold: Option<u64>,
+
+    pub max_secondary_lag_bytes: Option<u64>,
 }
 
 impl NeonStorageControllerConf {
     // Use a shorter pageserver unavailability interval than the default to speed up tests.
-    const DEFAULT_MAX_UNAVAILABLE_INTERVAL: std::time::Duration =
-        std::time::Duration::from_secs(10);
+    const DEFAULT_MAX_OFFLINE_INTERVAL: std::time::Duration = std::time::Duration::from_secs(10);
+
+    const DEFAULT_MAX_WARMING_UP_INTERVAL: std::time::Duration = std::time::Duration::from_secs(30);
 }
 
 impl Default for NeonStorageControllerConf {
     fn default() -> Self {
         Self {
-            max_unavailable: Self::DEFAULT_MAX_UNAVAILABLE_INTERVAL,
+            max_offline: Self::DEFAULT_MAX_OFFLINE_INTERVAL,
+            max_warming_up: Self::DEFAULT_MAX_WARMING_UP_INTERVAL,
+            start_as_candidate: false,
+            database_url: None,
             split_threshold: None,
+            max_secondary_lag_bytes: None,
         }
     }
 }
@@ -325,11 +340,16 @@ impl LocalEnv {
         }
     }
 
-    pub fn pg_bin_dir(&self, pg_version: u32) -> anyhow::Result<PathBuf> {
-        Ok(self.pg_distrib_dir(pg_version)?.join("bin"))
+    pub fn pg_dir(&self, pg_version: u32, dir_name: &str) -> anyhow::Result<PathBuf> {
+        Ok(self.pg_distrib_dir(pg_version)?.join(dir_name))
     }
+
+    pub fn pg_bin_dir(&self, pg_version: u32) -> anyhow::Result<PathBuf> {
+        self.pg_dir(pg_version, "bin")
+    }
+
     pub fn pg_lib_dir(&self, pg_version: u32) -> anyhow::Result<PathBuf> {
-        Ok(self.pg_distrib_dir(pg_version)?.join("lib"))
+        self.pg_dir(pg_version, "lib")
     }
 
     pub fn pageserver_bin(&self) -> PathBuf {
@@ -379,6 +399,36 @@ impl LocalEnv {
         }
     }
 
+    /// Inspect the base data directory and extract the instance id and instance directory path
+    /// for all storage controller instances
+    pub async fn storage_controller_instances(&self) -> std::io::Result<Vec<(u8, PathBuf)>> {
+        let mut instances = Vec::default();
+
+        let dir = std::fs::read_dir(self.base_data_dir.clone())?;
+        for dentry in dir {
+            let dentry = dentry?;
+            let is_dir = dentry.metadata()?.is_dir();
+            let filename = dentry.file_name().into_string().unwrap();
+            let parsed_instance_id = match filename.strip_prefix("storage_controller_") {
+                Some(suffix) => suffix.parse::<u8>().ok(),
+                None => None,
+            };
+
+            let is_instance_dir = is_dir && parsed_instance_id.is_some();
+
+            if !is_instance_dir {
+                continue;
+            }
+
+            instances.push((
+                parsed_instance_id.expect("Checked previously"),
+                dentry.path(),
+            ));
+        }
+
+        Ok(instances)
+    }
+
     pub fn register_branch_mapping(
         &mut self,
         branch_name: String,
@@ -431,9 +481,7 @@ impl LocalEnv {
     }
 
     ///  Construct `Self` from on-disk state.
-    pub fn load_config() -> anyhow::Result<Self> {
-        let repopath = base_path();
-
+    pub fn load_config(repopath: &Path) -> anyhow::Result<Self> {
         if !repopath.exists() {
             bail!(
                 "Neon config is not found in {}. You need to run 'neon_local init' first",
@@ -461,7 +509,7 @@ impl LocalEnv {
                 branch_name_mappings,
             } = on_disk_config;
             LocalEnv {
-                base_data_dir: repopath.clone(),
+                base_data_dir: repopath.to_owned(),
                 pg_distrib_dir,
                 neon_distrib_dir,
                 default_tenant_id,
@@ -482,7 +530,7 @@ impl LocalEnv {
             "we ensure this during deserialization"
         );
         env.pageservers = {
-            let iter = std::fs::read_dir(&repopath).context("open dir")?;
+            let iter = std::fs::read_dir(repopath).context("open dir")?;
             let mut pageservers = Vec::new();
             for res in iter {
                 let dentry = res?;
@@ -506,7 +554,6 @@ impl LocalEnv {
                 #[derive(serde::Serialize, serde::Deserialize)]
                 // (allow unknown fields, unlike PageServerConf)
                 struct PageserverConfigTomlSubset {
-                    id: NodeId,
                     listen_pg_addr: String,
                     listen_http_addr: String,
                     pg_auth_type: AuthType,
@@ -518,18 +565,30 @@ impl LocalEnv {
                         .with_context(|| format!("read {:?}", config_toml_path))?,
                 )
                 .context("parse pageserver.toml")?;
+                let identity_toml_path = dentry.path().join("identity.toml");
+                #[derive(serde::Serialize, serde::Deserialize)]
+                struct IdentityTomlSubset {
+                    id: NodeId,
+                }
+                let identity_toml: IdentityTomlSubset = toml_edit::de::from_str(
+                    &std::fs::read_to_string(&identity_toml_path)
+                        .with_context(|| format!("read {:?}", identity_toml_path))?,
+                )
+                .context("parse identity.toml")?;
                 let PageserverConfigTomlSubset {
-                    id: config_toml_id,
                     listen_pg_addr,
                     listen_http_addr,
                     pg_auth_type,
                     http_auth_type,
                 } = config_toml;
+                let IdentityTomlSubset {
+                    id: identity_toml_id,
+                } = identity_toml;
                 let conf = PageServerConf {
                     id: {
                         anyhow::ensure!(
-                            config_toml_id == id,
-                            "id mismatch: config_toml.id={config_toml_id} id={id}",
+                            identity_toml_id == id,
+                            "id mismatch: identity.toml:id={identity_toml_id} pageserver_(.*) id={id}",
                         );
                         id
                     },
@@ -719,10 +778,25 @@ impl LocalEnv {
 }
 
 pub fn base_path() -> PathBuf {
-    match std::env::var_os("NEON_REPO_DIR") {
-        Some(val) => PathBuf::from(val),
-        None => PathBuf::from(".neon"),
-    }
+    let path = match std::env::var_os("NEON_REPO_DIR") {
+        Some(val) => {
+            let path = PathBuf::from(val);
+            if !path.is_absolute() {
+                // repeat the env var in the error because our default is always absolute
+                panic!("NEON_REPO_DIR must be an absolute path, got {path:?}");
+            }
+            path
+        }
+        None => {
+            let pwd = std::env::current_dir()
+                // technically this can fail but it's quite unlikeley
+                .expect("determine current directory");
+            let pwd_abs = pwd.canonicalize().expect("canonicalize current directory");
+            pwd_abs.join(".neon")
+        }
+    };
+    assert!(path.is_absolute());
+    path
 }
 
 /// Generate a public/private key pair for JWT authentication

control_plane/src/pageserver.rs

@@ -1,8 +1,10 @@
 //! Code to manage pageservers
 //!
-//! In the local test environment, the pageserver stores its data directly in
+//! In the local test environment, the data for each pageserver is stored in
 //!
-//!   .neon/
+//! ```text
+//!   .neon/pageserver_<pageserver_id>
+//! ```
 //!
 use std::collections::HashMap;
 
@@ -15,16 +17,15 @@ use std::time::Duration;
 
 use anyhow::{bail, Context};
 use camino::Utf8PathBuf;
-use futures::SinkExt;
 use pageserver_api::models::{
-    self, AuxFilePolicy, LocationConfig, ShardParameters, TenantHistorySize, TenantInfo,
-    TimelineInfo,
+    self, AuxFilePolicy, LocationConfig, TenantHistorySize, TenantInfo, TimelineInfo,
 };
 use pageserver_api::shard::TenantShardId;
 use pageserver_client::mgmt_api;
 use postgres_backend::AuthType;
 use postgres_connection::{parse_host_port, PgConnectionConfig};
 use utils::auth::{Claims, Scope};
+use utils::id::NodeId;
 use utils::{
     id::{TenantId, TimelineId},
     lsn::Lsn,
@@ -74,6 +75,10 @@ impl PageServerNode {
         }
     }
 
+    fn pageserver_make_identity_toml(&self, node_id: NodeId) -> toml_edit::Document {
+        toml_edit::Document::from_str(&format!("id={node_id}")).unwrap()
+    }
+
     fn pageserver_init_make_toml(
         &self,
         conf: NeonLocalInitPageserverConf,
@@ -122,10 +127,13 @@ impl PageServerNode {
         }
 
         // Apply the user-provided overrides
-        overrides.push(
-            toml_edit::ser::to_string_pretty(&conf)
-                .expect("we deserialized this from toml earlier"),
-        );
+        overrides.push({
+            let mut doc =
+                toml_edit::ser::to_document(&conf).expect("we deserialized this from toml earlier");
+            // `id` is written out to `identity.toml` instead of `pageserver.toml`
+            doc.remove("id").expect("it's part of the struct");
+            doc.to_string()
+        });
 
         // Turn `overrides` into a toml document.
         // TODO: above code is legacy code, it should be refactored to use toml_edit directly.
@@ -158,8 +166,8 @@ impl PageServerNode {
             .expect("non-Unicode path")
     }
 
-    pub async fn start(&self) -> anyhow::Result<()> {
-        self.start_node().await
+    pub async fn start(&self, retry_timeout: &Duration) -> anyhow::Result<()> {
+        self.start_node(retry_timeout).await
     }
 
     fn pageserver_init(&self, conf: NeonLocalInitPageserverConf) -> anyhow::Result<()> {
@@ -186,6 +194,19 @@ impl PageServerNode {
             .write_all(config.to_string().as_bytes())
             .context("write pageserver toml")?;
         drop(config_file);
+
+        let identity_file_path = datadir.join("identity.toml");
+        let mut identity_file = std::fs::OpenOptions::new()
+            .create_new(true)
+            .write(true)
+            .open(identity_file_path)
+            .with_context(|| format!("open identity toml for write: {config_file_path:?}"))?;
+        let identity_toml = self.pageserver_make_identity_toml(node_id);
+        identity_file
+            .write_all(identity_toml.to_string().as_bytes())
+            .context("write identity toml")?;
+        drop(identity_toml);
+
         // TODO: invoke a TBD config-check command to validate that pageserver will start with the written config
 
         // Write metadata file, used by pageserver on startup to register itself with
@@ -214,14 +235,15 @@ impl PageServerNode {
         Ok(())
     }
 
-    async fn start_node(&self) -> anyhow::Result<()> {
+    async fn start_node(&self, retry_timeout: &Duration) -> anyhow::Result<()> {
         // TODO: using a thread here because start_process() is not async but we need to call check_status()
         let datadir = self.repo_path();
         print!(
-            "Starting pageserver node {} at '{}' in {:?}",
+            "Starting pageserver node {} at '{}' in {:?}, retrying for {:?}",
             self.conf.id,
             self.pg_connection_config.raw_address(),
-            datadir
+            datadir,
+            retry_timeout
         );
         io::stdout().flush().context("flush stdout")?;
 
@@ -239,6 +261,7 @@ impl PageServerNode {
             args,
             self.pageserver_env_variables()?,
             background_process::InitialPidFile::Expect(self.pid_file()),
+            retry_timeout,
             || async {
                 let st = self.check_status().await;
                 match st {
@@ -349,11 +372,6 @@ impl PageServerNode {
                 .map(|x| x.parse::<NonZeroU64>())
                 .transpose()
                 .context("Failed to parse 'max_lsn_wal_lag' as non zero integer")?,
-            trace_read_requests: settings
-                .remove("trace_read_requests")
-                .map(|x| x.parse::<bool>())
-                .transpose()
-                .context("Failed to parse 'trace_read_requests' as bool")?,
             eviction_policy: settings
                 .remove("eviction_policy")
                 .map(serde_json::from_str)
@@ -395,28 +413,6 @@ impl PageServerNode {
         }
     }
 
-    pub async fn tenant_create(
-        &self,
-        new_tenant_id: TenantId,
-        generation: Option<u32>,
-        settings: HashMap<&str, &str>,
-    ) -> anyhow::Result<TenantId> {
-        let config = Self::parse_config(settings.clone())?;
-
-        let request = models::TenantCreateRequest {
-            new_tenant_id: TenantShardId::unsharded(new_tenant_id),
-            generation,
-            config,
-            shard_parameters: ShardParameters::default(),
-            // Placement policy is not meaningful for creations not done via storage controller
-            placement_policy: None,
-        };
-        if !settings.is_empty() {
-            bail!("Unrecognized tenant settings: {settings:?}")
-        }
-        Ok(self.http_client.tenant_create(&request).await?)
-    }
-
     pub async fn tenant_config(
         &self,
         tenant_id: TenantId,
@@ -476,11 +472,6 @@ impl PageServerNode {
                     .map(|x| x.parse::<NonZeroU64>())
                     .transpose()
                     .context("Failed to parse 'max_lsn_wal_lag' as non zero integer")?,
-                trace_read_requests: settings
-                    .remove("trace_read_requests")
-                    .map(|x| x.parse::<bool>())
-                    .transpose()
-                    .context("Failed to parse 'trace_read_requests' as bool")?,
                 eviction_policy: settings
                     .remove("eviction_policy")
                     .map(serde_json::from_str)
@@ -587,60 +578,39 @@ impl PageServerNode {
         pg_wal: Option<(Lsn, PathBuf)>,
         pg_version: u32,
     ) -> anyhow::Result<()> {
-        let (client, conn) = self.page_server_psql_client().await?;
-        // The connection object performs the actual communication with the database,
-        // so spawn it off to run on its own.
-        tokio::spawn(async move {
-            if let Err(e) = conn.await {
-                eprintln!("connection error: {}", e);
-            }
-        });
-        let client = std::pin::pin!(client);
-
         // Init base reader
         let (start_lsn, base_tarfile_path) = base;
         let base_tarfile = tokio::fs::File::open(base_tarfile_path).await?;
-        let base_tarfile = tokio_util::io::ReaderStream::new(base_tarfile);
+        let base_tarfile =
+            mgmt_api::ReqwestBody::wrap_stream(tokio_util::io::ReaderStream::new(base_tarfile));
 
         // Init wal reader if necessary
         let (end_lsn, wal_reader) = if let Some((end_lsn, wal_tarfile_path)) = pg_wal {
             let wal_tarfile = tokio::fs::File::open(wal_tarfile_path).await?;
-            let wal_reader = tokio_util::io::ReaderStream::new(wal_tarfile);
+            let wal_reader =
+                mgmt_api::ReqwestBody::wrap_stream(tokio_util::io::ReaderStream::new(wal_tarfile));
             (end_lsn, Some(wal_reader))
         } else {
             (start_lsn, None)
         };
 
-        let copy_in = |reader, cmd| {
-            let client = &client;
-            async move {
-                let writer = client.copy_in(&cmd).await?;
-                let writer = std::pin::pin!(writer);
-                let mut writer = writer.sink_map_err(|e| {
-                    std::io::Error::new(std::io::ErrorKind::Other, format!("{e}"))
-                });
-                let mut reader = std::pin::pin!(reader);
-                writer.send_all(&mut reader).await?;
-                writer.into_inner().finish().await?;
-                anyhow::Ok(())
-            }
-        };
-
         // Import base
-        copy_in(
-            base_tarfile,
-            format!(
-                "import basebackup {tenant_id} {timeline_id} {start_lsn} {end_lsn} {pg_version}"
-            ),
-        )
-        .await?;
-        // Import wal if necessary
-        if let Some(wal_reader) = wal_reader {
-            copy_in(
-                wal_reader,
-                format!("import wal {tenant_id} {timeline_id} {start_lsn} {end_lsn}"),
+        self.http_client
+            .import_basebackup(
+                tenant_id,
+                timeline_id,
+                start_lsn,
+                end_lsn,
+                pg_version,
+                base_tarfile,
             )
             .await?;
+
+        // Import wal if necessary
+        if let Some(wal_reader) = wal_reader {
+            self.http_client
+                .import_wal(tenant_id, timeline_id, start_lsn, end_lsn, wal_reader)
+                .await?;
         }
 
         Ok(())

control_plane/src/safekeeper.rs

@@ -7,6 +7,7 @@
 //! ```
 use std::io::Write;
 use std::path::PathBuf;
+use std::time::Duration;
 use std::{io, result};
 
 use anyhow::Context;
@@ -111,11 +112,16 @@ impl SafekeeperNode {
             .expect("non-Unicode path")
     }
 
-    pub async fn start(&self, extra_opts: Vec<String>) -> anyhow::Result<()> {
+    pub async fn start(
+        &self,
+        extra_opts: Vec<String>,
+        retry_timeout: &Duration,
+    ) -> anyhow::Result<()> {
         print!(
-            "Starting safekeeper at '{}' in '{}'",
+            "Starting safekeeper at '{}' in '{}', retrying for {:?}",
             self.pg_connection_config.raw_address(),
-            self.datadir_path().display()
+            self.datadir_path().display(),
+            retry_timeout,
         );
         io::stdout().flush().unwrap();
 
@@ -200,6 +206,7 @@ impl SafekeeperNode {
             &args,
             self.safekeeper_env_variables()?,
             background_process::InitialPidFile::Expect(self.pid_file()),
+            retry_timeout,
             || async {
                 match self.check_status().await {
                     Ok(()) => Ok(true),

control_plane/src/storage_controller.rs

@@ -3,14 +3,16 @@ use crate::{
     local_env::{LocalEnv, NeonStorageControllerConf},
 };
 use camino::{Utf8Path, Utf8PathBuf};
+use hyper::Uri;
+use nix::unistd::Pid;
 use pageserver_api::{
     controller_api::{
-        NodeConfigureRequest, NodeRegisterRequest, TenantCreateResponse, TenantLocateResponse,
-        TenantShardMigrateRequest, TenantShardMigrateResponse,
+        NodeConfigureRequest, NodeDescribeResponse, NodeRegisterRequest, TenantCreateRequest,
+        TenantCreateResponse, TenantLocateResponse, TenantShardMigrateRequest,
+        TenantShardMigrateResponse,
     },
     models::{
-        TenantCreateRequest, TenantShardSplitRequest, TenantShardSplitResponse,
-        TimelineCreateRequest, TimelineInfo,
+        TenantShardSplitRequest, TenantShardSplitResponse, TimelineCreateRequest, TimelineInfo,
     },
     shard::{ShardStripeSize, TenantShardId},
 };
@@ -18,7 +20,7 @@ use pageserver_client::mgmt_api::ResponseErrorMessageExt;
 use postgres_backend::AuthType;
 use reqwest::Method;
 use serde::{de::DeserializeOwned, Deserialize, Serialize};
-use std::{fs, str::FromStr};
+use std::{fs, net::SocketAddr, path::PathBuf, str::FromStr, sync::OnceLock};
 use tokio::process::Command;
 use tracing::instrument;
 use url::Url;
@@ -29,23 +31,57 @@ use utils::{
 
 pub struct StorageController {
     env: LocalEnv,
-    listen: String,
-    path: Utf8PathBuf,
     private_key: Option<Vec<u8>>,
     public_key: Option<String>,
-    postgres_port: u16,
     client: reqwest::Client,
     config: NeonStorageControllerConf,
+
+    // The listen addresses is learned when starting the storage controller,
+    // hence the use of OnceLock to init it at the right time.
+    listen: OnceLock<SocketAddr>,
 }
 
 const COMMAND: &str = "storage_controller";
 
 const STORAGE_CONTROLLER_POSTGRES_VERSION: u32 = 16;
 
+const DB_NAME: &str = "storage_controller";
+
+pub struct NeonStorageControllerStartArgs {
+    pub instance_id: u8,
+    pub base_port: Option<u16>,
+    pub start_timeout: humantime::Duration,
+}
+
+impl NeonStorageControllerStartArgs {
+    pub fn with_default_instance_id(start_timeout: humantime::Duration) -> Self {
+        Self {
+            instance_id: 1,
+            base_port: None,
+            start_timeout,
+        }
+    }
+}
+
+pub struct NeonStorageControllerStopArgs {
+    pub instance_id: u8,
+    pub immediate: bool,
+}
+
+impl NeonStorageControllerStopArgs {
+    pub fn with_default_instance_id(immediate: bool) -> Self {
+        Self {
+            instance_id: 1,
+            immediate,
+        }
+    }
+}
+
 #[derive(Serialize, Deserialize)]
 pub struct AttachHookRequest {
     pub tenant_shard_id: TenantShardId,
     pub node_id: Option<NodeId>,
+    pub generation_override: Option<i32>,
 }
 
 #[derive(Serialize, Deserialize)]
@@ -65,27 +101,6 @@ pub struct InspectResponse {
 
 impl StorageController {
     pub fn from_env(env: &LocalEnv) -> Self {
-        let path = Utf8PathBuf::from_path_buf(env.base_data_dir.clone())
-            .unwrap()
-            .join("attachments.json");
-
-        // Makes no sense to construct this if pageservers aren't going to use it: assume
-        // pageservers have control plane API set
-        let listen_url = env.control_plane_api.clone().unwrap();
-
-        let listen = format!(
-            "{}:{}",
-            listen_url.host_str().unwrap(),
-            listen_url.port().unwrap()
-        );
-
-        // Convention: NeonEnv in python tests reserves the next port after the control_plane_api
-        // port, for use by our captive postgres.
-        let postgres_port = listen_url
-            .port()
-            .expect("Control plane API setting should always have a port")
-            + 1;
-
         // Assume all pageservers have symmetric auth configuration: this service
         // expects to use one JWT token to talk to all of them.
         let ps_conf = env
@@ -128,21 +143,28 @@ impl StorageController {
 
         Self {
             env: env.clone(),
-            path,
-            listen,
             private_key,
             public_key,
-            postgres_port,
             client: reqwest::ClientBuilder::new()
                 .build()
                 .expect("Failed to construct http client"),
             config: env.storage_controller.clone(),
+            listen: OnceLock::default(),
         }
     }
 
-    fn pid_file(&self) -> Utf8PathBuf {
-        Utf8PathBuf::from_path_buf(self.env.base_data_dir.join("storage_controller.pid"))
-            .expect("non-Unicode path")
+    fn storage_controller_instance_dir(&self, instance_id: u8) -> PathBuf {
+        self.env
+            .base_data_dir
+            .join(format!("storage_controller_{}", instance_id))
+    }
+
+    fn pid_file(&self, instance_id: u8) -> Utf8PathBuf {
+        Utf8PathBuf::from_path_buf(
+            self.storage_controller_instance_dir(instance_id)
+                .join("storage_controller.pid"),
+        )
+        .expect("non-Unicode path")
     }
 
     /// PIDFile for the postgres instance used to store storage controller state
@@ -155,16 +177,16 @@ impl StorageController {
         .expect("non-Unicode path")
     }
 
-    /// Find the directory containing postgres binaries, such as `initdb` and `pg_ctl`
+    /// Find the directory containing postgres subdirectories, such `bin` and `lib`
     ///
     /// This usually uses STORAGE_CONTROLLER_POSTGRES_VERSION of postgres, but will fall back
     /// to other versions if that one isn't found.  Some automated tests create circumstances
     /// where only one version is available in pg_distrib_dir, such as `test_remote_extensions`.
-    pub async fn get_pg_bin_dir(&self) -> anyhow::Result<Utf8PathBuf> {
+    async fn get_pg_dir(&self, dir_name: &str) -> anyhow::Result<Utf8PathBuf> {
         let prefer_versions = [STORAGE_CONTROLLER_POSTGRES_VERSION, 15, 14];
 
         for v in prefer_versions {
-            let path = Utf8PathBuf::from_path_buf(self.env.pg_bin_dir(v)?).unwrap();
+            let path = Utf8PathBuf::from_path_buf(self.env.pg_dir(v, dir_name)?).unwrap();
             if tokio::fs::try_exists(&path).await? {
                 return Ok(path);
             }
@@ -172,30 +194,38 @@ impl StorageController {
 
         // Fall through
         anyhow::bail!(
-            "Postgres binaries not found in {}",
-            self.env.pg_distrib_dir.display()
+            "Postgres directory '{}' not found in {}",
+            dir_name,
+            self.env.pg_distrib_dir.display(),
         );
     }
 
+    pub async fn get_pg_bin_dir(&self) -> anyhow::Result<Utf8PathBuf> {
+        self.get_pg_dir("bin").await
+    }
+
+    pub async fn get_pg_lib_dir(&self) -> anyhow::Result<Utf8PathBuf> {
+        self.get_pg_dir("lib").await
+    }
+
     /// Readiness check for our postgres process
-    async fn pg_isready(&self, pg_bin_dir: &Utf8Path) -> anyhow::Result<bool> {
+    async fn pg_isready(&self, pg_bin_dir: &Utf8Path, postgres_port: u16) -> anyhow::Result<bool> {
         let bin_path = pg_bin_dir.join("pg_isready");
-        let args = ["-h", "localhost", "-p", &format!("{}", self.postgres_port)];
+        let args = ["-h", "localhost", "-p", &format!("{}", postgres_port)];
         let exitcode = Command::new(bin_path).args(args).spawn()?.wait().await?;
 
         Ok(exitcode.success())
     }
 
-    /// Create our database if it doesn't exist, and run migrations.
+    /// Create our database if it doesn't exist
     ///
     /// This function is equivalent to the `diesel setup` command in the diesel CLI.  We implement
     /// the same steps by hand to avoid imposing a dependency on installing diesel-cli for developers
     /// who just want to run `cargo neon_local` without knowing about diesel.
     ///
     /// Returns the database url
-    pub async fn setup_database(&self) -> anyhow::Result<String> {
-        const DB_NAME: &str = "storage_controller";
-        let database_url = format!("postgresql://localhost:{}/{DB_NAME}", self.postgres_port);
+    pub async fn setup_database(&self, postgres_port: u16) -> anyhow::Result<String> {
+        let database_url = format!("postgresql://localhost:{}/{DB_NAME}", postgres_port);
 
         let pg_bin_dir = self.get_pg_bin_dir().await?;
         let createdb_path = pg_bin_dir.join("createdb");
@@ -204,7 +234,7 @@ impl StorageController {
                 "-h",
                 "localhost",
                 "-p",
-                &format!("{}", self.postgres_port),
+                &format!("{}", postgres_port),
                 DB_NAME,
             ])
             .output()
@@ -223,80 +253,211 @@ impl StorageController {
         Ok(database_url)
     }
 
-    pub async fn start(&self) -> anyhow::Result<()> {
-        // Start a vanilla Postgres process used by the storage controller for persistence.
-        let pg_data_path = Utf8PathBuf::from_path_buf(self.env.base_data_dir.clone())
-            .unwrap()
-            .join("storage_controller_db");
-        let pg_bin_dir = self.get_pg_bin_dir().await?;
-        let pg_log_path = pg_data_path.join("postgres.log");
+    pub async fn connect_to_database(
+        &self,
+        postgres_port: u16,
+    ) -> anyhow::Result<(
+        tokio_postgres::Client,
+        tokio_postgres::Connection<tokio_postgres::Socket, tokio_postgres::tls::NoTlsStream>,
+    )> {
+        tokio_postgres::Config::new()
+            .host("localhost")
+            .port(postgres_port)
+            // The user is the ambient operating system user name.
+            // That is an impurity which we want to fix in => TODO https://github.com/neondatabase/neon/issues/8400
+            //
+            // Until we get there, use the ambient operating system user name.
+            // Recent tokio-postgres versions default to this if the user isn't specified.
+            // But tokio-postgres fork doesn't have this upstream commit:
+            // https://github.com/sfackler/rust-postgres/commit/cb609be758f3fb5af537f04b584a2ee0cebd5e79
+            // => we should rebase our fork => TODO https://github.com/neondatabase/neon/issues/8399
+            .user(&whoami::username())
+            .dbname(DB_NAME)
+            .connect(tokio_postgres::NoTls)
+            .await
+            .map_err(anyhow::Error::new)
+    }
 
-        if !tokio::fs::try_exists(&pg_data_path).await? {
-            // Initialize empty database
-            let initdb_path = pg_bin_dir.join("initdb");
-            let mut child = Command::new(&initdb_path)
-                .args(["-D", pg_data_path.as_ref()])
-                .spawn()
-                .expect("Failed to spawn initdb");
-            let status = child.wait().await?;
-            if !status.success() {
-                anyhow::bail!("initdb failed with status {status}");
+    pub async fn start(&self, start_args: NeonStorageControllerStartArgs) -> anyhow::Result<()> {
+        let instance_dir = self.storage_controller_instance_dir(start_args.instance_id);
+        if let Err(err) = tokio::fs::create_dir(&instance_dir).await {
+            if err.kind() != std::io::ErrorKind::AlreadyExists {
+                panic!("Failed to create instance dir {instance_dir:?}");
             }
+        }
+
+        let (listen, postgres_port) = {
+            if let Some(base_port) = start_args.base_port {
+                (
+                    format!("127.0.0.1:{base_port}"),
+                    self.config
+                        .database_url
+                        .expect("--base-port requires NeonStorageControllerConf::database_url")
+                        .port(),
+                )
+            } else {
+                let listen_url = self.env.control_plane_api.clone().unwrap();
+
+                let listen = format!(
+                    "{}:{}",
+                    listen_url.host_str().unwrap(),
+                    listen_url.port().unwrap()
+                );
+
+                (listen, listen_url.port().unwrap() + 1)
+            }
+        };
+
+        let socket_addr = listen
+            .parse()
+            .expect("listen address is a valid socket address");
+        self.listen
+            .set(socket_addr)
+            .expect("StorageController::listen is only set here");
+
+        // Do we remove the pid file on stop?
+        let pg_started = self.is_postgres_running().await?;
+        let pg_lib_dir = self.get_pg_lib_dir().await?;
+
+        if !pg_started {
+            // Start a vanilla Postgres process used by the storage controller for persistence.
+            let pg_data_path = Utf8PathBuf::from_path_buf(self.env.base_data_dir.clone())
+                .unwrap()
+                .join("storage_controller_db");
+            let pg_bin_dir = self.get_pg_bin_dir().await?;
+            let pg_log_path = pg_data_path.join("postgres.log");
+
+            if !tokio::fs::try_exists(&pg_data_path).await? {
+                // Initialize empty database
+                let initdb_path = pg_bin_dir.join("initdb");
+                let mut child = Command::new(&initdb_path)
+                    .envs(vec![
+                        ("LD_LIBRARY_PATH".to_owned(), pg_lib_dir.to_string()),
+                        ("DYLD_LIBRARY_PATH".to_owned(), pg_lib_dir.to_string()),
+                    ])
+                    .args(["-D", pg_data_path.as_ref()])
+                    .spawn()
+                    .expect("Failed to spawn initdb");
+                let status = child.wait().await?;
+                if !status.success() {
+                    anyhow::bail!("initdb failed with status {status}");
+                }
+            };
 
             // Write a minimal config file:
             // - Specify the port, since this is chosen dynamically
             // - Switch off fsync, since we're running on lightweight test environments and when e.g. scale testing
             //   the storage controller we don't want a slow local disk to interfere with that.
+            //
+            // NB: it's important that we rewrite this file on each start command so we propagate changes
+            // from `LocalEnv`'s config file (`.neon/config`).
             tokio::fs::write(
                 &pg_data_path.join("postgresql.conf"),
-                format!("port = {}\nfsync=off\n", self.postgres_port),
+                format!("port = {}\nfsync=off\n", postgres_port),
             )
             .await?;
+
+            println!("Starting storage controller database...");
+            let db_start_args = [
+                "-w",
+                "-D",
+                pg_data_path.as_ref(),
+                "-l",
+                pg_log_path.as_ref(),
+                "start",
+            ];
+
+            background_process::start_process(
+                "storage_controller_db",
+                &self.env.base_data_dir,
+                pg_bin_dir.join("pg_ctl").as_std_path(),
+                db_start_args,
+                vec![
+                    ("LD_LIBRARY_PATH".to_owned(), pg_lib_dir.to_string()),
+                    ("DYLD_LIBRARY_PATH".to_owned(), pg_lib_dir.to_string()),
+                ],
+                background_process::InitialPidFile::Create(self.postgres_pid_file()),
+                &start_args.start_timeout,
+                || self.pg_isready(&pg_bin_dir, postgres_port),
+            )
+            .await?;
+
+            self.setup_database(postgres_port).await?;
+        }
+
+        let database_url = format!("postgresql://localhost:{}/{DB_NAME}", postgres_port);
+
+        // We support running a startup SQL script to fiddle with the database before we launch storcon.
+        // This is used by the test suite.
+        let startup_script_path = self
+            .env
+            .base_data_dir
+            .join("storage_controller_db.startup.sql");
+        let startup_script = match tokio::fs::read_to_string(&startup_script_path).await {
+            Ok(script) => {
+                tokio::fs::remove_file(startup_script_path).await?;
+                script
+            }
+            Err(e) => {
+                if e.kind() == std::io::ErrorKind::NotFound {
+                    // always run some startup script so that this code path doesn't bit rot
+                    "BEGIN; COMMIT;".to_string()
+                } else {
+                    anyhow::bail!("Failed to read startup script: {e}")
+                }
+            }
         };
+        let (mut client, conn) = self.connect_to_database(postgres_port).await?;
+        let conn = tokio::spawn(conn);
+        let tx = client.build_transaction();
+        let tx = tx.start().await?;
+        tx.batch_execute(&startup_script).await?;
+        tx.commit().await?;
+        drop(client);
+        conn.await??;
 
-        println!("Starting storage controller database...");
-        let db_start_args = [
-            "-w",
-            "-D",
-            pg_data_path.as_ref(),
-            "-l",
-            pg_log_path.as_ref(),
-            "start",
-        ];
-
-        background_process::start_process(
-            "storage_controller_db",
-            &self.env.base_data_dir,
-            pg_bin_dir.join("pg_ctl").as_std_path(),
-            db_start_args,
-            [],
-            background_process::InitialPidFile::Create(self.postgres_pid_file()),
-            || self.pg_isready(&pg_bin_dir),
-        )
-        .await?;
-
-        // Run migrations on every startup, in case something changed.
-        let database_url = self.setup_database().await?;
+        let listen = self
+            .listen
+            .get()
+            .expect("cell is set earlier in this function");
+        let address_for_peers = Uri::builder()
+            .scheme("http")
+            .authority(format!("{}:{}", listen.ip(), listen.port()))
+            .path_and_query("")
+            .build()
+            .unwrap();
 
         let mut args = vec![
             "-l",
-            &self.listen,
-            "-p",
-            self.path.as_ref(),
+            &listen.to_string(),
             "--dev",
             "--database-url",
             &database_url,
-            "--max-unavailable-interval",
-            &humantime::Duration::from(self.config.max_unavailable).to_string(),
+            "--max-offline-interval",
+            &humantime::Duration::from(self.config.max_offline).to_string(),
+            "--max-warming-up-interval",
+            &humantime::Duration::from(self.config.max_warming_up).to_string(),
+            "--address-for-peers",
+            &address_for_peers.to_string(),
         ]
         .into_iter()
         .map(|s| s.to_string())
         .collect::<Vec<_>>();
+
+        if self.config.start_as_candidate {
+            args.push("--start-as-candidate".to_string());
+        }
+
         if let Some(private_key) = &self.private_key {
             let claims = Claims::new(None, Scope::PageServerApi);
             let jwt_token =
                 encode_from_key_file(&claims, private_key).expect("failed to generate jwt token");
             args.push(format!("--jwt-token={jwt_token}"));
+
+            let peer_claims = Claims::new(None, Scope::Admin);
+            let peer_jwt_token = encode_from_key_file(&peer_claims, private_key)
+                .expect("failed to generate jwt token");
+            args.push(format!("--peer-jwt-token={peer_jwt_token}"));
         }
 
         if let Some(public_key) = &self.public_key {
@@ -313,16 +474,26 @@ impl StorageController {
             args.push(format!("--split-threshold={split_threshold}"))
         }
 
+        if let Some(lag) = self.config.max_secondary_lag_bytes.as_ref() {
+            args.push(format!("--max-secondary-lag-bytes={lag}"))
+        }
+
+        args.push(format!(
+            "--neon-local-repo-dir={}",
+            self.env.base_data_dir.display()
+        ));
+
         background_process::start_process(
             COMMAND,
-            &self.env.base_data_dir,
+            &instance_dir,
             &self.env.storage_controller_bin(),
             args,
-            [(
-                "NEON_REPO_DIR".to_string(),
-                self.env.base_data_dir.to_string_lossy().to_string(),
-            )],
-            background_process::InitialPidFile::Create(self.pid_file()),
+            vec![
+                ("LD_LIBRARY_PATH".to_owned(), pg_lib_dir.to_string()),
+                ("DYLD_LIBRARY_PATH".to_owned(), pg_lib_dir.to_string()),
+            ],
+            background_process::InitialPidFile::Create(self.pid_file(start_args.instance_id)),
+            &start_args.start_timeout,
             || async {
                 match self.ready().await {
                     Ok(_) => Ok(true),
@@ -335,8 +506,35 @@ impl StorageController {
         Ok(())
     }
 
-    pub async fn stop(&self, immediate: bool) -> anyhow::Result<()> {
-        background_process::stop_process(immediate, COMMAND, &self.pid_file())?;
+    pub async fn stop(&self, stop_args: NeonStorageControllerStopArgs) -> anyhow::Result<()> {
+        background_process::stop_process(
+            stop_args.immediate,
+            COMMAND,
+            &self.pid_file(stop_args.instance_id),
+        )?;
+
+        let storcon_instances = self.env.storage_controller_instances().await?;
+        for (instance_id, instanced_dir_path) in storcon_instances {
+            if instance_id == stop_args.instance_id {
+                continue;
+            }
+
+            let pid_file = instanced_dir_path.join("storage_controller.pid");
+            let pid = tokio::fs::read_to_string(&pid_file)
+                .await
+                .map_err(|err| {
+                    anyhow::anyhow!("Failed to read storcon pid file at {pid_file:?}: {err}")
+                })?
+                .parse::<i32>()
+                .expect("pid is valid i32");
+
+            let other_proc_alive = !background_process::process_has_stopped(Pid::from_raw(pid))?;
+            if other_proc_alive {
+                // There is another storage controller instance running, so we return
+                // and leave the database running.
+                return Ok(());
+            }
+        }
 
         let pg_data_path = self.env.base_data_dir.join("storage_controller_db");
         let pg_bin_dir = self.get_pg_bin_dir().await?;
@@ -349,27 +547,51 @@ impl StorageController {
             .wait()
             .await?;
         if !stop_status.success() {
-            let pg_status_args = ["-D", &pg_data_path.to_string_lossy(), "status"];
-            let status_exitcode = Command::new(pg_bin_dir.join("pg_ctl"))
-                .args(pg_status_args)
-                .spawn()?
-                .wait()
-                .await?;
-
-            // pg_ctl status returns this exit code if postgres is not running: in this case it is
-            // fine that stop failed.  Otherwise it is an error that stop failed.
-            const PG_STATUS_NOT_RUNNING: i32 = 3;
-            if Some(PG_STATUS_NOT_RUNNING) == status_exitcode.code() {
-                println!("Storage controller database is already stopped");
-                return Ok(());
-            } else {
-                anyhow::bail!("Failed to stop storage controller database: {stop_status}")
+            match self.is_postgres_running().await {
+                Ok(false) => {
+                    println!("Storage controller database is already stopped");
+                    return Ok(());
+                }
+                Ok(true) => {
+                    anyhow::bail!("Failed to stop storage controller database");
+                }
+                Err(err) => {
+                    anyhow::bail!("Failed to stop storage controller database: {err}");
+                }
             }
         }
 
         Ok(())
     }
 
+    async fn is_postgres_running(&self) -> anyhow::Result<bool> {
+        let pg_data_path = self.env.base_data_dir.join("storage_controller_db");
+        let pg_bin_dir = self.get_pg_bin_dir().await?;
+
+        let pg_status_args = ["-D", &pg_data_path.to_string_lossy(), "status"];
+        let status_exitcode = Command::new(pg_bin_dir.join("pg_ctl"))
+            .args(pg_status_args)
+            .spawn()?
+            .wait()
+            .await?;
+
+        // pg_ctl status returns this exit code if postgres is not running: in this case it is
+        // fine that stop failed.  Otherwise it is an error that stop failed.
+        const PG_STATUS_NOT_RUNNING: i32 = 3;
+        const PG_NO_DATA_DIR: i32 = 4;
+        const PG_STATUS_RUNNING: i32 = 0;
+        match status_exitcode.code() {
+            Some(PG_STATUS_NOT_RUNNING) => Ok(false),
+            Some(PG_NO_DATA_DIR) => Ok(false),
+            Some(PG_STATUS_RUNNING) => Ok(true),
+            Some(code) => Err(anyhow::anyhow!(
+                "pg_ctl status returned unexpected status code: {:?}",
+                code
+            )),
+            None => Err(anyhow::anyhow!("pg_ctl status returned no status code")),
+        }
+    }
+
     fn get_claims_for_path(path: &str) -> anyhow::Result<Option<Claims>> {
         let category = match path.find('/') {
             Some(idx) => &path[..idx],
@@ -395,15 +617,31 @@ impl StorageController {
         RQ: Serialize + Sized,
         RS: DeserializeOwned + Sized,
     {
-        // The configured URL has the /upcall path prefix for pageservers to use: we will strip that out
-        // for general purpose API access.
-        let listen_url = self.env.control_plane_api.clone().unwrap();
-        let url = Url::from_str(&format!(
-            "http://{}:{}/{path}",
-            listen_url.host_str().unwrap(),
-            listen_url.port().unwrap()
-        ))
-        .unwrap();
+        // In the special case of the `storage_controller start` subcommand, we wish
+        // to use the API endpoint of the newly started storage controller in order
+        // to pass the readiness check. In this scenario [`Self::listen`] will be set
+        // (see [`Self::start`]).
+        //
+        // Otherwise, we infer the storage controller api endpoint from the configured
+        // control plane API.
+        let url = if let Some(socket_addr) = self.listen.get() {
+            Url::from_str(&format!(
+                "http://{}:{}/{path}",
+                socket_addr.ip().to_canonical(),
+                socket_addr.port()
+            ))
+            .unwrap()
+        } else {
+            // The configured URL has the /upcall path prefix for pageservers to use: we will strip that out
+            // for general purpose API access.
+            let listen_url = self.env.control_plane_api.clone().unwrap();
+            Url::from_str(&format!(
+                "http://{}:{}/{path}",
+                listen_url.host_str().unwrap(),
+                listen_url.port().unwrap()
+            ))
+            .unwrap()
+        };
 
         let mut builder = self.client.request(method, url);
         if let Some(body) = body {
@@ -440,6 +678,7 @@ impl StorageController {
         let request = AttachHookRequest {
             tenant_shard_id,
             node_id: Some(pageserver_id),
+            generation_override: None,
         };
 
         let response = self
@@ -551,6 +790,15 @@ impl StorageController {
         .await
     }
 
+    pub async fn node_list(&self) -> anyhow::Result<Vec<NodeDescribeResponse>> {
+        self.dispatch::<(), Vec<NodeDescribeResponse>>(
+            Method::GET,
+            "control/v1/node".to_string(),
+            None,
+        )
+        .await
+    }
+
     #[instrument(skip(self))]
     pub async fn ready(&self) -> anyhow::Result<()> {
         self.dispatch::<(), ()>(Method::GET, "ready".to_string(), None)

control_plane/storcon_cli/Cargo.toml

@@ -17,6 +17,7 @@ pageserver_client.workspace = true
 reqwest.workspace = true
 serde.workspace = true
 serde_json = { workspace = true, features = ["raw_value"] }
+storage_controller_client.workspace = true
 thiserror.workspace = true
 tokio.workspace = true
 tracing.workspace = true

control_plane/storcon_cli/src/main.rs

@@ -1,28 +1,28 @@
 use futures::StreamExt;
-use std::{collections::HashMap, str::FromStr, time::Duration};
+use std::{str::FromStr, time::Duration};
 
 use clap::{Parser, Subcommand};
 use pageserver_api::{
     controller_api::{
-        NodeAvailabilityWrapper, NodeDescribeResponse, ShardSchedulingPolicy,
+        NodeAvailabilityWrapper, NodeDescribeResponse, ShardSchedulingPolicy, TenantCreateRequest,
         TenantDescribeResponse, TenantPolicyRequest,
     },
     models::{
         EvictionPolicy, EvictionPolicyLayerAccessThreshold, LocationConfigSecondary,
-        ShardParameters, TenantConfig, TenantConfigRequest, TenantCreateRequest,
-        TenantShardSplitRequest, TenantShardSplitResponse,
+        ShardParameters, TenantConfig, TenantConfigRequest, TenantShardSplitRequest,
+        TenantShardSplitResponse,
     },
     shard::{ShardStripeSize, TenantShardId},
 };
-use pageserver_client::mgmt_api::{self, ResponseErrorMessageExt};
+use pageserver_client::mgmt_api::{self};
 use reqwest::{Method, StatusCode, Url};
-use serde::{de::DeserializeOwned, Serialize};
 use utils::id::{NodeId, TenantId};
 
 use pageserver_api::controller_api::{
     NodeConfigureRequest, NodeRegisterRequest, NodeSchedulingPolicy, PlacementPolicy,
-    TenantLocateResponse, TenantShardMigrateRequest, TenantShardMigrateResponse,
+    TenantShardMigrateRequest, TenantShardMigrateResponse,
 };
+use storage_controller_client::control_api::Client;
 
 #[derive(Subcommand, Debug)]
 enum Command {
@@ -56,6 +56,10 @@ enum Command {
         #[arg(long)]
         scheduling: Option<NodeSchedulingPolicy>,
     },
+    NodeDelete {
+        #[arg(long)]
+        node_id: NodeId,
+    },
     /// Modify a tenant's policies in the storage controller
     TenantPolicy {
         #[arg(long)]
@@ -110,12 +114,6 @@ enum Command {
         #[arg(long)]
         config: String,
     },
-    /// Attempt to balance the locations for a tenant across pageservers.  This is a client-side
-    /// alternative to the storage controller's scheduling optimization behavior.
-    TenantScatter {
-        #[arg(long)]
-        tenant_id: TenantId,
-    },
     /// Print details about a particular tenant, including all its shards' states.
     TenantDescribe {
         #[arg(long)]
@@ -251,64 +249,6 @@ impl FromStr for NodeAvailabilityArg {
     }
 }
 
-struct Client {
-    base_url: Url,
-    jwt_token: Option<String>,
-    client: reqwest::Client,
-}
-
-impl Client {
-    fn new(base_url: Url, jwt_token: Option<String>) -> Self {
-        Self {
-            base_url,
-            jwt_token,
-            client: reqwest::ClientBuilder::new()
-                .build()
-                .expect("Failed to construct http client"),
-        }
-    }
-
-    /// Simple HTTP request wrapper for calling into storage controller
-    async fn dispatch<RQ, RS>(
-        &self,
-        method: Method,
-        path: String,
-        body: Option<RQ>,
-    ) -> mgmt_api::Result<RS>
-    where
-        RQ: Serialize + Sized,
-        RS: DeserializeOwned + Sized,
-    {
-        // The configured URL has the /upcall path prefix for pageservers to use: we will strip that out
-        // for general purpose API access.
-        let url = Url::from_str(&format!(
-            "http://{}:{}/{path}",
-            self.base_url.host_str().unwrap(),
-            self.base_url.port().unwrap()
-        ))
-        .unwrap();
-
-        let mut builder = self.client.request(method, url);
-        if let Some(body) = body {
-            builder = builder.json(&body)
-        }
-        if let Some(jwt_token) = &self.jwt_token {
-            builder = builder.header(
-                reqwest::header::AUTHORIZATION,
-                format!("Bearer {jwt_token}"),
-            );
-        }
-
-        let response = builder.send().await.map_err(mgmt_api::Error::ReceiveBody)?;
-        let response = response.error_from_body().await?;
-
-        response
-            .json()
-            .await
-            .map_err(pageserver_client::mgmt_api::Error::ReceiveBody)
-    }
-}
-
 #[tokio::main]
 async fn main() -> anyhow::Result<()> {
     let cli = Cli::parse();
@@ -342,14 +282,18 @@ async fn main() -> anyhow::Result<()> {
                 .await?;
         }
         Command::TenantCreate { tenant_id } => {
-            vps_client
-                .tenant_create(&TenantCreateRequest {
-                    new_tenant_id: TenantShardId::unsharded(tenant_id),
-                    generation: None,
-                    shard_parameters: ShardParameters::default(),
-                    placement_policy: Some(PlacementPolicy::Attached(1)),
-                    config: TenantConfig::default(),
-                })
+            storcon_client
+                .dispatch::<_, ()>(
+                    Method::POST,
+                    "v1/tenant".to_string(),
+                    Some(TenantCreateRequest {
+                        new_tenant_id: TenantShardId::unsharded(tenant_id),
+                        generation: None,
+                        shard_parameters: ShardParameters::default(),
+                        placement_policy: Some(PlacementPolicy::Attached(1)),
+                        config: TenantConfig::default(),
+                    }),
+                )
                 .await?;
         }
         Command::TenantDelete { tenant_id } => {
@@ -359,13 +303,16 @@ async fn main() -> anyhow::Result<()> {
             tracing::info!("Delete status: {}", status);
         }
         Command::Nodes {} => {
-            let resp = storcon_client
+            let mut resp = storcon_client
                 .dispatch::<(), Vec<NodeDescribeResponse>>(
                     Method::GET,
                     "control/v1/node".to_string(),
                     None,
                 )
                 .await?;
+
+            resp.sort_by(|a, b| a.listen_http_addr.cmp(&b.listen_http_addr));
+
             let mut table = comfy_table::Table::new();
             table.set_header(["Id", "Hostname", "Scheduling", "Availability"]);
             for node in resp {
@@ -397,13 +344,16 @@ async fn main() -> anyhow::Result<()> {
                 .await?;
         }
         Command::Tenants {} => {
-            let resp = storcon_client
+            let mut resp = storcon_client
                 .dispatch::<(), Vec<TenantDescribeResponse>>(
                     Method::GET,
                     "control/v1/tenant".to_string(),
                     None,
                 )
                 .await?;
+
+            resp.sort_by(|a, b| a.tenant_id.cmp(&b.tenant_id));
+
             let mut table = comfy_table::Table::new();
             table.set_header([
                 "TenantId",
@@ -498,88 +448,6 @@ async fn main() -> anyhow::Result<()> {
                 })
                 .await?;
         }
-        Command::TenantScatter { tenant_id } => {
-            // Find the shards
-            let locate_response = storcon_client
-                .dispatch::<(), TenantLocateResponse>(
-                    Method::GET,
-                    format!("control/v1/tenant/{tenant_id}/locate"),
-                    None,
-                )
-                .await?;
-            let shards = locate_response.shards;
-
-            let mut node_to_shards: HashMap<NodeId, Vec<TenantShardId>> = HashMap::new();
-            let shard_count = shards.len();
-            for s in shards {
-                let entry = node_to_shards.entry(s.node_id).or_default();
-                entry.push(s.shard_id);
-            }
-
-            // Load list of available nodes
-            let nodes_resp = storcon_client
-                .dispatch::<(), Vec<NodeDescribeResponse>>(
-                    Method::GET,
-                    "control/v1/node".to_string(),
-                    None,
-                )
-                .await?;
-
-            for node in nodes_resp {
-                if matches!(node.availability, NodeAvailabilityWrapper::Active) {
-                    node_to_shards.entry(node.id).or_default();
-                }
-            }
-
-            let max_shard_per_node = shard_count / node_to_shards.len();
-
-            loop {
-                let mut migrate_shard = None;
-                for shards in node_to_shards.values_mut() {
-                    if shards.len() > max_shard_per_node {
-                        // Pick the emptiest
-                        migrate_shard = Some(shards.pop().unwrap());
-                    }
-                }
-                let Some(migrate_shard) = migrate_shard else {
-                    break;
-                };
-
-                // Pick the emptiest node to migrate to
-                let mut destinations = node_to_shards
-                    .iter()
-                    .map(|(k, v)| (k, v.len()))
-                    .collect::<Vec<_>>();
-                destinations.sort_by_key(|i| i.1);
-                let (destination_node, destination_count) = *destinations.first().unwrap();
-                if destination_count + 1 > max_shard_per_node {
-                    // Even the emptiest destination doesn't have space: we're done
-                    break;
-                }
-                let destination_node = *destination_node;
-
-                node_to_shards
-                    .get_mut(&destination_node)
-                    .unwrap()
-                    .push(migrate_shard);
-
-                println!("Migrate {} -> {} ...", migrate_shard, destination_node);
-
-                storcon_client
-                    .dispatch::<TenantShardMigrateRequest, TenantShardMigrateResponse>(
-                        Method::PUT,
-                        format!("control/v1/tenant/{migrate_shard}/migrate"),
-                        Some(TenantShardMigrateRequest {
-                            tenant_shard_id: migrate_shard,
-                            node_id: destination_node,
-                        }),
-                    )
-                    .await?;
-                println!("Migrate {} -> {} OK", migrate_shard, destination_node);
-            }
-
-            // Spread the shards across the nodes
-        }
         Command::TenantDescribe { tenant_id } => {
             let describe_response = storcon_client
                 .dispatch::<(), TenantDescribeResponse>(
@@ -734,6 +602,11 @@ async fn main() -> anyhow::Result<()> {
                 .dispatch::<(), ()>(Method::POST, format!("debug/v1/node/{node_id}/drop"), None)
                 .await?;
         }
+        Command::NodeDelete { node_id } => {
+            storcon_client
+                .dispatch::<(), ()>(Method::DELETE, format!("control/v1/node/{node_id}"), None)
+                .await?;
+        }
         Command::TenantSetTimeBasedEviction {
             tenant_id,
             period,
@@ -749,6 +622,7 @@ async fn main() -> anyhow::Result<()> {
                                 threshold: threshold.into(),
                             },
                         )),
+                        heatmap_period: Some("300s".to_string()),
                         ..Default::default()
                     },
                 })

deny.toml

@@ -4,6 +4,7 @@
 # to your expectations and requirements.
 
 # Root options
+[graph]
 targets = [
     { triple = "x86_64-unknown-linux-gnu" },
     { triple = "aarch64-unknown-linux-gnu" },
@@ -12,6 +13,7 @@ targets = [
 ]
 all-features = false
 no-default-features = false
+[output]
 feature-depth = 1
 
 # This section is considered when running `cargo deny check advisories`
@@ -19,17 +21,16 @@ feature-depth = 1
 # https://embarkstudios.github.io/cargo-deny/checks/advisories/cfg.html
 [advisories]
 db-urls = ["https://github.com/rustsec/advisory-db"]
-vulnerability = "deny"
-unmaintained = "warn"
 yanked = "warn"
-notice = "warn"
-ignore = []
+
+[[advisories.ignore]]
+id = "RUSTSEC-2023-0071"
+reason = "the marvin attack only affects private key decryption, not public key signature verification"
 
 # This section is considered when running `cargo deny check licenses`
 # More documentation for the licenses section can be found here:
 # https://embarkstudios.github.io/cargo-deny/checks/licenses/cfg.html
 [licenses]
-unlicensed = "deny"
 allow = [
     "Apache-2.0",
     "Artistic-2.0",
@@ -42,10 +43,6 @@ allow = [
     "OpenSSL",
     "Unicode-DFS-2016",
 ]
-deny = []
-copyleft = "warn"
-allow-osi-fsf-free = "neither"
-default = "deny"
 confidence-threshold = 0.8
 exceptions = [
     # Zlib license has some restrictions if we decide to change sth

docker-compose/README.md

@@ -0,0 +1,10 @@
+
+# Example docker compose configuration
+
+The configuration in this directory is used for testing Neon docker images: it is
+not intended for deploying a usable system.  To run a development environment where
+you can experiment with a minature Neon system, use `cargo neon` rather than container images.
+
+This configuration does not start the storage controller, because the controller
+needs a way to reconfigure running computes, and no such thing exists in this setup.
+

docker-compose/compute_wrapper/shell/compute.sh

@@ -23,18 +23,17 @@ echo "Page server is ready."
 echo "Create a tenant and timeline"
 generate_id tenant_id
 PARAMS=(
-     -sb 
-     -X POST
+     -X PUT
      -H "Content-Type: application/json"
-     -d "{\"new_tenant_id\": \"${tenant_id}\"}"
-     http://pageserver:9898/v1/tenant/
+     -d "{\"mode\": \"AttachedSingle\", \"generation\": 1, \"tenant_conf\": {}}"
+     "http://pageserver:9898/v1/tenant/${tenant_id}/location_config"
 )
 result=$(curl "${PARAMS[@]}")
 echo $result | jq .
 
 generate_id timeline_id
 PARAMS=(
-     -sb 
+     -sbf
      -X POST
      -H "Content-Type: application/json"
      -d "{\"new_timeline_id\": \"${timeline_id}\", \"pg_version\": ${PG_VERSION}}"

docker-compose/docker-compose.yml

@@ -31,25 +31,14 @@ services:
     restart: always
     image: ${REPOSITORY:-neondatabase}/neon:${TAG:-latest}
     environment:
-      - BROKER_ENDPOINT='http://storage_broker:50051'
       - AWS_ACCESS_KEY_ID=minio
       - AWS_SECRET_ACCESS_KEY=password
       #- RUST_BACKTRACE=1
     ports:
        #- 6400:6400  # pg protocol handler
        - 9898:9898 # http endpoints
-    entrypoint:
-      - "/bin/sh"
-      - "-c"
-    command:
-      - "/usr/local/bin/pageserver -D /data/.neon/
-                                   -c \"broker_endpoint=$$BROKER_ENDPOINT\"
-                                   -c \"listen_pg_addr='0.0.0.0:6400'\"
-                                   -c \"listen_http_addr='0.0.0.0:9898'\"
-                                   -c \"remote_storage={endpoint='http://minio:9000',
-                                                        bucket_name='neon',
-                                                        bucket_region='eu-north-1',
-                                                        prefix_in_bucket='/pageserver/'}\""
+    volumes:
+      - ./pageserver_config:/data/.neon/
     depends_on:
       - storage_broker
       - minio_create_buckets

docker-compose/docker_compose_test.sh

@@ -78,7 +78,7 @@ for pg_version in 14 15 16; do
         docker cp $TMPDIR/data $COMPUTE_CONTAINER_NAME:/ext-src/pg_hint_plan-src/
         rm -rf $TMPDIR
         # We are running tests now
-        if docker exec -e SKIP=rum-src,timescaledb-src,rdkit-src,postgis-src,pgx_ulid-src,pgtap-src,pg_tiktoken-src,pg_jsonschema-src,pg_graphql-src,kq_imcx-src,wal2json_2_5-src \
+        if docker exec -e SKIP=timescaledb-src,rdkit-src,postgis-src,pgx_ulid-src,pgtap-src,pg_tiktoken-src,pg_jsonschema-src,pg_graphql-src,kq_imcx-src,wal2json_2_5-src \
             $TEST_CONTAINER_NAME /run-tests.sh | tee testout.txt
         then
             cleanup

docker-compose/pageserver_config/identity.toml

@@ -0,0 +1 @@
+id=1234

docker-compose/pageserver_config/pageserver.toml

@@ -0,0 +1,5 @@
+broker_endpoint='http://storage_broker:50051'
+pg_distrib_dir='/usr/local/'
+listen_pg_addr='0.0.0.0:6400'
+listen_http_addr='0.0.0.0:9898'
+remote_storage={ endpoint='http://minio:9000', bucket_name='neon', bucket_region='eu-north-1', prefix_in_bucket='/pageserver' }

docker-compose/run-tests.sh

@@ -1,15 +1,15 @@
 #!/bin/bash
 set -x
 
-cd /ext-src
+cd /ext-src || exit 2
 FAILED=
-LIST=$((echo ${SKIP} | sed 's/,/\n/g'; ls -d *-src) | sort | uniq -u)
+LIST=$( (echo "${SKIP//","/"\n"}"; ls -d -- *-src) | sort | uniq -u)
 for d in ${LIST}
 do
-       [ -d ${d} ] || continue
+       [ -d "${d}" ] || continue
     psql -c "select 1" >/dev/null || break
-       make -C ${d} installcheck || FAILED="${d} ${FAILED}"
+       USE_PGXS=1 make -C "${d}" installcheck || FAILED="${d} ${FAILED}"
 done
 [ -z "${FAILED}" ] && exit 0
-echo ${FAILED}
+echo "${FAILED}"
 exit 1

docs/SUMMARY.md

@@ -1,13 +1,18 @@
 # Summary
 
+# Looking for `neon.tech` docs?
+
+This page linkes to a selection of technical content about the open source code in this repository.
+
+Please visit https://neon.tech/docs for documentation about using the Neon service, which is based on the code
+in this repository.
+
+# Architecture
+
 [Introduction]()
 - [Separation of Compute and Storage](./separation-compute-storage.md)
 
-# Architecture
-
 - [Compute]()
-  - [WAL proposer]()
-  - [WAL Backpressure]()
   - [Postgres changes](./core_changes.md)
 
 - [Pageserver](./pageserver.md)
@@ -16,33 +21,15 @@
     - [WAL Redo](./pageserver-walredo.md)
     - [Page cache](./pageserver-pagecache.md)
     - [Storage](./pageserver-storage.md)
-        - [Datadir mapping]()
-        - [Layer files]()
-        - [Branching]()
-        - [Garbage collection]()
-    - [Cloud Storage]()
     - [Processing a GetPage request](./pageserver-processing-getpage.md)
     - [Processing WAL](./pageserver-processing-wal.md)
-	- [Management API]()
-	- [Tenant Rebalancing]()
 
 - [WAL Service](walservice.md)
   - [Consensus protocol](safekeeper-protocol.md)
-  - [Management API]()
-  - [Rebalancing]()
-
-- [Control Plane]()
-
-- [Proxy]()
 
 - [Source view](./sourcetree.md)
   - [docker.md](./docker.md) — Docker images and building pipeline.
   - [Error handling and logging](./error-handling.md)
-  - [Testing]()
-    - [Unit testing]()
-    - [Integration testing]()
-    - [Benchmarks]()
-
 
 - [Glossary](./glossary.md)
 
@@ -58,28 +45,6 @@
 
 # RFCs
 
-- [RFCs](./rfcs/README.md)
-
-- [002-storage](rfcs/002-storage.md)
-- [003-laptop-cli](rfcs/003-laptop-cli.md)
-- [004-durability](rfcs/004-durability.md)
-- [005-zenith_local](rfcs/005-zenith_local.md)
-- [006-laptop-cli-v2-CLI](rfcs/006-laptop-cli-v2-CLI.md)
-- [006-laptop-cli-v2-repository-structure](rfcs/006-laptop-cli-v2-repository-structure.md)
-- [007-serverless-on-laptop](rfcs/007-serverless-on-laptop.md)
-- [008-push-pull](rfcs/008-push-pull.md)
-- [009-snapshot-first-storage-cli](rfcs/009-snapshot-first-storage-cli.md)
-- [009-snapshot-first-storage](rfcs/009-snapshot-first-storage.md)
-- [009-snapshot-first-storage-pitr](rfcs/009-snapshot-first-storage-pitr.md)
-- [010-storage_details](rfcs/010-storage_details.md)
-- [011-retention-policy](rfcs/011-retention-policy.md)
-- [012-background-tasks](rfcs/012-background-tasks.md)
-- [013-term-history](rfcs/013-term-history.md)
-- [014-safekeepers-gossip](rfcs/014-safekeepers-gossip.md)
-- [014-storage-lsm](rfcs/014-storage-lsm.md)
-- [015-storage-messaging](rfcs/015-storage-messaging.md)
-- [016-connection-routing](rfcs/016-connection-routing.md)
-- [017-timeline-data-management](rfcs/017-timeline-data-management.md)
-- [018-storage-messaging-2](rfcs/018-storage-messaging-2.md)
-- [019-tenant-timeline-lifecycles](rfcs/019-tenant-timeline-lifecycles.md)
-- [cluster-size-limits](rfcs/cluster-size-limits.md)
+Major changes are documented in RFCS:
+- See [RFCs](./rfcs/README.md) for more information
+- view the RFCs at https://github.com/neondatabase/neon/tree/main/docs/rfcs

docs/pageserver-pagecache.md

@@ -5,4 +5,3 @@ TODO:
 - shared across tenants
 - store pages from layer files
 - store pages from "in-memory layer"
-- store materialized pages

docs/rfcs/033-storage-controller-drain-and-fill.md

@@ -0,0 +1,345 @@
+# Graceful Restarts of Storage Controller Managed Clusters
+
+## Summary
+This RFC describes new storage controller APIs for draining and filling tenant shards from/on pageserver nodes.
+It also covers how these new APIs should be used by an orchestrator (e.g. Ansible) in order to implement
+graceful cluster restarts.
+
+## Motivation
+
+Pageserver restarts cause read availablity downtime for tenants.
+
+For example pageserver-3 @ us-east-1 was unavailable for a randomly
+picked tenant (which requested on-demand activation) for around 30 seconds
+during the restart at 2024-04-03 16:37 UTC.
+
+Note that lots of shutdowns on loaded pageservers do not finish within the
+[10 second systemd enforced timeout](https://github.com/neondatabase/infra/blob/0a5280b383e43c063d43cbf87fa026543f6d6ad4/.github/ansible/systemd/pageserver.service#L16). This means we are shutting down without flushing ephemeral layers
+and have to reingest data in order to serve requests after restarting, potentially making first request latencies worse.
+
+This problem is not yet very acutely felt in storage controller managed pageservers since
+tenant density is much lower there. However, we are planning on eventually migrating all
+pageservers to storage controller management, so it makes sense to solve the issue proactively.
+
+## Requirements
+
+- Pageserver re-deployments cause minimal downtime for tenants
+- The storage controller exposes HTTP API hooks for draining and filling tenant shards
+from a given pageserver. Said hooks can be used by an orchestrator proces or a human operator.
+- The storage controller exposes some HTTP API to cancel draining and filling background operations.
+- Failures to drain or fill the node should not be fatal. In such cases, cluster restarts should proceed
+as usual (with downtime).
+- Progress of draining/filling is visible through metrics
+
+## Non Goals
+
+- Integration with the control plane
+- Graceful restarts for large non-HA tenants.
+
+## Impacted Components
+
+- storage controller
+- deployment orchestrator (i.e. Ansible)
+- pageserver (indirectly)
+
+## Terminology
+
+** Draining ** is the process through which all tenant shards that can be migrated from a given pageserver
+are distributed across the rest of the cluster.
+
+** Filling ** is the symmetric opposite of draining. In this process tenant shards are migrated onto a given
+pageserver until the cluster reaches a resonable, quiescent distribution of tenant shards across pageservers.
+
+** Node scheduling policies ** act as constraints to the scheduler. For instance, when a
+node is set in the `Paused` policy, no further shards will be scheduled on it.
+
+** Node ** is a pageserver. Term is used interchangeably in this RFC.
+
+** Deployment orchestrator ** is a generic term for whatever drives our deployments.
+Currently, it's an Ansible playbook.
+
+## Background
+
+### Storage Controller Basics (skip if already familiar)
+
+Fundamentally, the storage controller is a reconciler which aims to move from the observed mapping between pageservers and tenant shards to an intended mapping. Pageserver nodes and tenant shards metadata is durably persisted in a database, but note that the mapping between the two entities is not durably persisted. Instead, this mapping (*observed state*) is constructed at startup by sending `GET location_config` requests to registered pageservers.
+
+An internal scheduler maps tenant shards to pageservers while respecting certain constraints. The result of scheduling is the *intent state*. When the intent state changes, a *reconciliation* will inform pageservers about the new assigment via `PUT location_config` requests and will notify the compute via the configured hook.
+
+### Background Optimizations
+
+The storage controller performs scheduling optimizations in the background. It will
+migrate attachments to warm secondaries and replace secondaries in order to balance
+the cluster out.
+
+### Reconciliations Concurrency Limiting
+
+There's a hard limit on the number of reconciles that the storage controller
+can have in flight at any given time. To get an idea of scales, the limit is
+128 at the time of writing.
+
+## Implementation
+
+Note: this section focuses on the core functionality of the graceful restart process.
+It doesn't neccesarily describe the most efficient approach. Optimizations are described
+separately in a later section.
+
+### Overall Flow
+
+This section describes how to implement graceful restarts from the perspective
+of Ansible, the deployment orchestrator. Pageservers are already restarted sequentially.
+The orchestrator shall implement the following epilogue and prologue steps for each
+pageserver restart:
+
+#### Prologue
+
+The orchestrator shall first fetch the pageserver node id from the control plane or
+the pageserver it aims to restart directly. Next, it issues an HTTP request
+to the storage controller in order to start the drain of said pageserver node.
+All error responses are retried with a short back-off. When a 202 (Accepted)
+HTTP code is returned, the drain has started. Now the orchestrator polls the
+node status endpoint exposed by the storage controller in order to await the
+end of the drain process. When the `policy` field of the node status response
+becomes `PauseForRestart`, the drain has completed and the orchestrator can
+proceed with restarting the pageserver.
+
+The prologue is subject to an overall timeout. It will have a value in the ballpark
+of minutes. As storage controller managed pageservers become more loaded this timeout
+will likely have to increase.
+
+#### Epilogue
+
+After restarting the pageserver, the orchestrator issues an HTTP request
+to the storage controller to kick off the filling process. This API call
+may be retried for all error codes with a short backoff. This also serves
+as a synchronization primitive as the fill will be refused if the pageserver
+has not yet re-attached to the storage controller. When a 202(Accepted) HTTP
+code is returned, the fill has started. Now the orchestrator polls the node
+status endpoint exposed by the storage controller in order to await the end of
+the filling process. When the `policy` field of the node status response becomes
+`Active`, the fill has completed and the orchestrator may proceed to the next pageserver.
+
+Again, the epilogue is subject to an overall timeout. We can start off with
+using the same timeout as for the prologue, but can also consider relying on
+the storage controller's background optimizations with a shorter timeout.
+
+In the case that the deployment orchestrator times out, it attempts to cancel
+the fill. This operation shall be retried with a short back-off. If it ultimately
+fails it will require manual intervention to set the nodes scheduling policy to
+`NodeSchedulingPolicy::Active`. Not doing that is not immediately problematic,
+but it constrains the scheduler as mentioned previously.
+
+### Node Scheduling Policy State Machine
+
+The state machine below encodes the behaviours discussed above and
+the various failover situations described in a later section.
+
+Assuming no failures and/or timeouts the flow should be:
+`Active -> Draining -> PauseForRestart -> Active -> Filling -> Active`
+
+```
+                          Operator requested drain
+               +-----------------------------------------+
+               |                                         |
+       +-------+-------+                         +-------v-------+
+       |               |                         |               |
+       |     Pause     |             +----------->    Draining   +----------+
+       |               |             |           |               |          |
+       +---------------+             |           +-------+-------+          |
+                                     |                   |                  |
+                                     |                   |                  |
+                      Drain requested|                   |                  |
+                                     |                   |Drain complete    | Drain failed
+                                     |                   |                  | Cancelled/PS reattach/Storcon restart
+                                     |                   |                  |
+                             +-------+-------+           |                  |
+                             |               |           |                  |
+               +-------------+    Active     <-----------+------------------+
+               |             |               |           |
+Fill requested |             +---^---^-------+           |
+               |                 |   |                   |
+               |                 |   |                   |
+               |                 |   |                   |
+               |   Fill completed|   |                   |
+               |                 |   |PS reattach        |
+               |                 |   |after restart      |
+       +-------v-------+         |   |           +-------v-------+
+       |               |         |   |           |               |
+       |    Filling    +---------+   +-----------+PauseForRestart|
+       |               |                         |               |
+       +---------------+                         +---------------+
+```
+
+### Draining/Filling APIs
+
+The storage controller API to trigger the draining of a given node is:
+`PUT /v1/control/node/:node_id/{drain,fill}`.
+
+The following HTTP non-success return codes are used.
+All of them are safely retriable from the perspective of the storage controller.
+- 404: Requested node was not found
+- 503: Requested node is known to the storage controller, but unavailable
+- 412: Drain precondition failed: there is no other node to drain to or the node's schedulling policy forbids draining
+- 409: A {drain, fill} is already in progress. Only one such background operation
+is allowed per node.
+
+When the drain is accepted and commenced a 202 HTTP code is returned.
+
+Drains and fills shall be cancellable by the deployment orchestrator or a
+human operator via: `DELETE /v1/control/node/:node_id/{drain,fill}`. A 200
+response is returned when the cancelation is successful. Errors are retriable.
+
+### Drain Process
+
+Before accpeting a drain request the following validations is applied:
+* Ensure that the node is known the storage controller
+* Ensure that the schedulling policy is `NodeSchedulingPolicy::Active` or `NodeSchedulingPolicy::Pause`
+* Ensure that another drain or fill is not already running on the node
+* Ensure that a drain is possible (i.e. check that there is at least one
+schedulable node to drain to)
+
+After accepting the drain, the scheduling policy of the node is set to
+`NodeSchedulingPolicy::Draining` and persisted in both memory and the database.
+This disallows the optimizer from adding or removing shards from the node which
+is desirable to avoid them racing.
+
+Next, a separate Tokio task is spawned to manage the draining. For each tenant
+shard attached to the node being drained, demote the node to a secondary and
+attempt to schedule the node away. Scheduling might fail due to unsatisfiable
+constraints, but that is fine. Draining is a best effort process since it might
+not always be possible to cut over all shards.
+
+Importantly, this task manages the concurrency of issued reconciles in order to
+avoid drowning out the target pageservers and to allow other important reconciles
+to proceed.
+
+Once the triggered reconciles have finished or timed out, set the node's scheduling
+policy to `NodeSchedulingPolicy::PauseForRestart` to signal the end of the drain.
+
+A note on non HA tenants: These tenants do not have secondaries, so by the description
+above, they would not be migrated. It makes sense to skip them (especially the large ones)
+since, depending on tenant size, this might be more disruptive than the restart since the
+pageserver we've moved to do will need to on-demand download the entire working set for the tenant.
+We can consider expanding to small non-HA tenants in the future.
+
+### Fill Process
+
+Before accpeting a fill request the following validations is applied:
+* Ensure that the node is known the storage controller
+* Ensure that the schedulling policy is `NodeSchedulingPolicy::Active`.
+This is the only acceptable policy for the fill starting state. When a node re-attaches,
+it set the scheduling policy to `NodeSchedulingPolicy::Active` if it was equal to
+`NodeSchedulingPolicy::PauseForRestart` or `NodeSchedulingPolicy::Draining` (possible end states for a node drain).
+* Ensure that another drain or fill is not already running on the node
+
+After accepting the drain, the scheduling policy of the node is set to
+`NodeSchedulingPolicy::Filling` and persisted in both memory and the database.
+This disallows the optimizer from adding or removing shards from the node which
+is desirable to avoid them racing.
+
+Next, a separate Tokio task is spawned to manage the draining. For each tenant
+shard where the filled node is a secondary, promote the secondary. This is done
+until we run out of shards or the counts of attached shards become balanced across
+the cluster.
+
+Like for draining, the concurrency of spawned reconciles is limited.
+
+### Failure Modes & Handling
+
+Failures are generally handled by transition back into the `Active`
+(neutral) state. This simplifies the implementation greatly at the
+cost of adding transitions to the state machine. For example, we
+could detect the `Draining` state upon restart and proceed with a drain,
+but how should the storage controller know that's what the orchestrator
+needs still?
+
+#### Storage Controller Crash
+
+When the storage controller starts up reset the node scheduling policy
+of all nodes in states `Draining`, `Filling` or `PauseForRestart` to
+`Active`. The rationale is that when the storage controller restarts,
+we have lost context of what the deployment orchestrator wants. It also
+has the benefit of making things easier to reason about.
+
+#### Pageserver Crash During Drain
+
+The pageserver will attempt to re-attach during restart at which
+point the node scheduling policy will be set back to `Active`, thus
+reenabling the scheduler to use the node.
+
+#### Non-drained Pageserver Crash During Drain
+
+What should happen when a pageserver we are draining to crashes during the
+process. Two reasonable options are: cancel the drain and focus on the failover
+*or* do both, but prioritise failover. Since the number of concurrent reconciles
+produced by drains/fills are limited, we get the later behaviour for free.
+My suggestion is we take this approach, but the cancellation option is trivial
+to implement as well.
+
+#### Pageserver Crash During Fill
+
+The pageserver will attempt to re-attach during restart at which
+point the node scheduling policy will be set back to `Active`, thus
+reenabling the scheduler to use the node.
+
+#### Pageserver Goes unavailable During Drain/Fill
+
+The drain and fill jobs handle this by stopping early. When the pageserver
+is detected as online by storage controller heartbeats, reset its scheduling
+policy to `Active`. If a restart happens instead, see the pageserver crash
+failure mode.
+
+#### Orchestrator Drain Times Out
+
+Orchestrator will still proceed with the restart.
+When the pageserver re-attaches, the scheduling policy is set back to
+`Active`.
+
+#### Orchestrator Fill Times Out
+
+Orchestrator will attempt to cancel the fill operation. If that fails,
+the fill will continue until it quiesces and the node will be left
+in the `Filling` scheduling policy. This hinders the scheduler, but is
+otherwise harmless. A human operator can handle this by setting the scheduling
+policy to `Active`, or we can bake in a fill timeout into the storage controller.
+
+## Optimizations
+
+### Location Warmth
+
+When cutting over to a secondary, the storage controller will wait for it to
+become "warm" (i.e. download enough of the tenants data). This means that some
+reconciliations can take significantly longer than others and hold up precious
+reconciliations units. As an optimization, the drain stage can only cut over
+tenants that are already "warm". Similarly, the fill stage can prioritise the
+"warmest" tenants in the fill.
+
+Given that the number of tenants by the storage controller will be fairly low
+for the foreseable future, the first implementation could simply query the tenants
+for secondary status. This doesn't scale well with increasing tenant counts, so
+eventually we will need new pageserver API endpoints to report the sets of
+"warm" and "cold" nodes.
+
+## Alternatives Considered
+
+### Draining and Filling Purely as Scheduling Constraints
+
+At its core, the storage controller is a big background loop that detects changes
+in the environment and reacts on them. One could express draining and filling
+of nodes purely in terms of constraining the scheduler (as opposed to having
+such background tasks).
+
+While theoretically nice, I think that's harder to implement and more importantly operate and reason about.
+Consider cancellation of a drain/fill operation. We would have to update the scheduler state, create
+an entirely new schedule (intent state) and start work on applying that. It gets trickier if we wish
+to cancel the reconciliation tasks spawned by drain/fill nodes. How would we know which ones belong
+to the conceptual drain/fill? One could add labels to reconciliations, but it gets messy in my opinion.
+
+It would also mean that reconciliations themselves have side effects that persist in the database
+(persist something to the databse when the drain is done), which I'm not conceptually fond of.
+
+## Proof of Concept
+
+This RFC is accompanied by a POC which implements nearly everything mentioned here
+apart from the optimizations and some of the failure handling:
+https://github.com/neondatabase/neon/pull/7682

docs/rfcs/034-ancestor-deletion.md

@@ -0,0 +1,252 @@
+# Ancestor Timeline Deletion
+
+Created on: 2024-02-23
+
+Author: John Spray
+
+# Summary
+
+When a tenant creates a new timeline that they will treat as their 'main' history,
+it is awkward to permanently retain an 'old main' timeline as its ancestor. Currently
+this is necessary because it is forbidden to delete a timeline which has descendents.
+
+A new pageserver API is proposed to 'adopt' data from a parent timeline into
+one of its children, such that the link between ancestor and child can be severed,
+leaving the parent in a state where it may then be deleted.
+
+# Motivation
+
+Retaining parent timelines currently has two costs:
+
+- Cognitive load on users, who have to remember which is the "real" main timeline.
+- Storage capacity cost, as the parent timeline will retain layers up to the
+  child's timeline point, even if the child fully covers its keyspace with image
+  layers and will never actually read from the parent.
+
+# Solution
+
+A new pageserver API `PUT /v1/tenant/:tenant_id/timeline/:timeline_id/detach_ancestor`
+will be added. The `timeline_id` in this URL is that of the _child_ timeline that we
+wish to detach from its parent.
+
+On success, this API will leave the following state:
+
+- The detached child timeline will no longer have an ancestor, and will contain all
+  the data needed to service reads without recursing into an ancestor.
+- Any other children of the parent whose timeline points were at a lower LSN than
+  the detached child timeline will be modified to have the child timeline as their
+  new parent.
+- The parent timeline will still exist, but the child will no longer have it as an
+  ancestor. If this was the last timeline that depended on the parent, then the
+  parent will become deletable.
+
+This API's implementation will consist of a series of retryable steps, such that
+on failures/timeout it can safely be called again to reach the target state.
+
+## Example
+
+### Before
+
+The user has "rolled back" their project to LSN X, resulting in a "new main"
+timeline. The parent "old main" timeline still exists, and they would like
+to clean it up.
+
+They have two other timelines A and B. A is from before the rollback point,
+and B is from after the rollback point.
+
+```
+----"old main" timeline-------X-------------------------------------------->
+                |             |                         |
+                |-> child A   |                         |
+                              |-> "new main" timeline   |
+                                                        -> child B
+
+```
+
+### After calling detach ancestor API
+
+The "new main" timeline is no longer dependent on old main, and neither
+is child A, because it had a branch point before X.
+
+The user may now choose to delete child B and "old main" to get to
+a pristine state. Child B is likely to be unwanted since the user
+chose to roll back to X, and it branches from after X. However, we
+don't assume this in the API; it is up to the user to delete it.
+
+```
+|----"old main" timeline---------------------------------------------------->
+                                                         |
+                                                         |
+                                                         |
+                                                         -> child B
+
+|----"new main" timeline--------->
+                 |
+                 |-> child A
+
+
+```
+
+### After removing timelines
+
+We end up with a totally clean state that leaves no trace that a rollback
+ever happened: there is only one root timeline.
+
+```
+| ----"new main" timeline----------->
+                |
+                |-> child A
+
+
+```
+
+## Caveats
+
+Important things for API users to bear in mind:
+
+- this API does not delete the parent timeline: you must still do that explicitly.
+- if there are other child timelines ahead of the branch point of the detached
+  child, the parent won't be deletable: you must either delete or detach those
+  children.
+- do _not_ simply loop over all children and detach them all: this can have an
+  extremely high storage cost. The detach ancestor API is intended for use on a single
+  timeline to make it the new "main".
+- The detach ancestor API should also not be
+  exposed directly to the user as button/API, because they might decide
+  to click it for all the children and thereby generate many copies of the
+  parent's data -- the detach ancestor API should be used as part
+  of a high level "clean up after rollback" feature.
+
+## `detach_ancestor` API implementation
+
+Terms used in the following sections:
+
+- "the child": the timeline whose ID is specified in the detach ancestor API URL, also
+  called "new main" in the example.
+- "the parent": the parent of "the child". Also called "old main" in the example.
+- "the branch point" the ancestor_lsn of "the child"
+
+### Phase 1: write out adopted layers to S3
+
+The child will "adopt" layers from the parent, such that its end state contains
+all the parent's history as well as its own.
+
+For all layers in the parent's layer map whose high LSN is below the branch
+point, issue S3 CopyObject requests to duplicate them into the child timeline's
+prefix. Do not add them to the child's layer map yet.
+
+For delta layers in the parent's layer map which straddle the branch point, read them
+and write out only content up to the branch point into new layer objects.
+
+This is a long running operation if the parent has many layers: it should be
+implemented in a way that resumes rather than restarting from scratch, if the API
+times out and is called again.
+
+As an optimization, if there are no other timelines that will be adopted into
+the child, _and_ the child's image layers already full cover the branch LSN,
+then we may skip adopting layers.
+
+### Phase 2: update the child's index
+
+Having written out all needed layers in phase 1, atomically link them all
+into the child's IndexPart and upload to S3. This may be done while the
+child Timeline is still running.
+
+### Phase 3: modify timelines ancestry
+
+Modify the child's ancestor to None, and upload its IndexPart to persist the change.
+
+For all timelines which have the same parent as the child, and have a branch
+point lower than our branch point, switch their ancestor_timeline to the child,
+and upload their IndexPart to persist the change.
+
+## Alternatives considered
+
+### Generate full image layer on child, rather than adopting parent deltas
+
+This would work for the case of a single child, but would prevent re-targeting
+other timelines that depended on the parent. If we detached many children this
+way, the storage cost would become prohibitive (consider a 1TB database with
+100 child timelines: it would cost 100TiB if they all generated their own image layers).
+
+### Don't rewrite anything: just fake it in the API
+
+We could add a layer of indirection that let a child "pretend" that it had no
+ancestor, when in reality it still had the parent. The pageserver API could
+accept deletion of ancestor timelines, and just update child metadata to make
+them look like they have no ancestor.
+
+This would not achieve the desired reduction in storage cost, and may well be more
+complex to maintain than simply implementing the API described in this RFC.
+
+### Avoid copying objects: enable child index to use parent layers directly
+
+We could teach IndexPart to store a TimelineId for each layer, such that a child
+timeline could reference a parent's layers directly, rather than copying them
+into the child's prefix.
+
+This would impose a cost for the normal case of indices that only target the
+timeline's own layers, add complexity, and break the useful simplifying
+invariant that timelines "own" their own path. If child timelines were
+referencing layers from the parent, we would have to ensure that the parent
+never runs GC/compaction again, which would make the API less flexible (the
+proposal in this RFC enables deletion of the parent but doesn't require it.)
+
+## Performance
+
+### Adopting layers
+
+- CopyObject is a relatively cheap operation, but we may need to issue tens of thousands
+  of such requests: this can take up to tens of seconds and will compete for RemoteStorage
+  semaphore units with other activity on the pageserver.
+- If we are running on storage backend that doesn't implement CopyObject, then
+  this part will be much more expensive as we would stream all layer content
+  through the pageserver. This is no different to issuing a lot
+  of reads to a timeline that does not have a warm local cache: it will move
+  a lot of gigabytes, but that shouldn't break anything.
+- Generating truncated layers for delta that straddle the branch point will
+  require streaming read/write of all the layers in question.
+
+### Updating timeline ancestry
+
+The simplest way to update timeline ancestry will probably be to stop and start
+all the Timeline objects: this is preferable to the complexity of making their
+ancestry mutable at runtime.
+
+There will be a corresponding "stutter" in the availability of the timelines,
+of the order 10-100ms, which is the time taken to upload their IndexPart, and
+restart the Timeline.
+
+# Interaction with other features
+
+## Concurrent timeline creation
+
+If new historic timelines are created using the parent as an ancestor while the
+detach ancestor API is running, they will not be re-parented to the child. This
+doesn't break anything, but it leaves the parent in a state where it might not
+be possible to delete it.
+
+Since timeline creations are an explicit user action, this is not something we need to
+worry about as the storage layer: a user who wants to delete their parent timeline will not create
+new children, and if they do, they can choose to delete those children to
+enable deleting the parent.
+
+For the least surprise to the user, before starting the detach ancestor branch
+operation, the control plane should wait until all branches are created and not
+allow any branches to be created before the branch point on the ancestor branch
+while the operation is ongoing.
+
+## WAL based disaster recovery
+
+WAL based disaster recovery currently supports only restoring of the main
+branch. Enabling WAL based disaster recovery in the future requires that we
+keep a record which timeline generated the WAL and at which LSN was a parent
+detached. Keep a list of timeline ids and the LSN in which they were detached in
+the `index_part.json`. Limit the size of the list to 100 first entries, after
+which the WAL disaster recovery will not be possible.
+
+## Sharded tenants
+
+For sharded tenants, calls to the detach ancestor API will pass through the storage
+controller, which will handle them the same as timeline creations: invoke first
+on shard zero, and then on all the other shards.

docs/rfcs/035-safekeeper-dynamic-membership-change.md

@@ -0,0 +1,495 @@
+# Safekeeper dynamic membership change
+
+To quickly recover from safekeeper node failures and do rebalancing we need to
+be able to change set of safekeepers the timeline resides on. The procedure must
+be safe (not lose committed log) regardless of safekeepers and compute state. It
+should be able to progress if any majority of old safekeeper set, any majority
+of new safekeeper set and compute are up and connected. This is known as a
+consensus membership change. It always involves two phases: 1) switch old
+majority to old + new configuration, preventing commits without acknowledge from
+the new set 2) bootstrap the new set by ensuring majority of the new set has all
+data which ever could have been committed before the first phase completed;
+after that switch is safe to finish. Without two phases switch to the new set
+which quorum might not intersect with quorum of the old set (and typical case of
+ABC -> ABD switch is an example of that, because quorums AC and BD don't
+intersect). Furthermore, procedure is typically carried out by the consensus
+leader, and so enumeration of configurations which establishes order between
+them is done through consensus log.
+
+In our case consensus leader is compute (walproposer), and we don't want to wake
+up all computes for the change. Neither we want to fully reimplement the leader
+logic second time outside compute. Because of that the proposed algorithm relies
+for issuing configurations on the external fault tolerant (distributed) strongly
+consisent storage with simple API: CAS (compare-and-swap) on the single key.
+Properly configured postgres suits this.
+
+In the system consensus is implemented at the timeline level, so algorithm below
+applies to the single timeline.
+
+## Algorithm
+
+### Definitions
+
+A configuration is
+
+```
+struct Configuration {
+    generation: Generation, // a number uniquely identifying configuration
+    sk_set: Vec<NodeId>, // current safekeeper set
+    new_sk_set: Optional<Vec<NodeId>>,
+}
+```
+
+Configuration with `new_set` present is used for the intermediate step during
+the change and called joint configuration. Generations establish order of
+generations: we say `c1` is higher than `c2` if `c1.generation` >
+`c2.generation`.
+
+### Persistently stored data changes
+
+Safekeeper starts storing its current configuration in the control file. Update
+of is atomic, so in-memory value always matches the persistent one.
+
+External CAS providing storage (let's call it configuration storage here) also
+stores configuration for each timeline. It is initialized with generation 1 and
+initial set of safekeepers during timeline creation. Executed CAS on it must
+never be lost.
+
+### Compute <-> safekeeper protocol changes
+
+`ProposerGreeting` message carries walproposer's configuration if it is already
+established (see below), else null.  `AcceptorGreeting` message carries
+safekeeper's current `Configuration`. All further messages (`VoteRequest`,
+`VoteResponse`, `ProposerElected`, `AppendRequest`, `AppendResponse`) carry
+generation number, of walproposer in case of wp->sk message or of safekeeper in
+case of sk->wp message.
+
+### Safekeeper changes
+
+Basic rule: once safekeeper observes configuration higher than his own it
+immediately switches to it. It must refuse all messages with lower generation
+that his. It also refuses messages if it is not member of the current generation
+(that is, of either `sk_set` of `sk_new_set`), though it is likely not unsafe to
+process them (walproposer should ignore result anyway).
+
+If there is non null configuration in `ProposerGreeting` and it is higher than
+current safekeeper one, safekeeper switches to it.
+
+Safekeeper sends its current configuration in its first message to walproposer
+`AcceptorGreeting`. It refuses all other walproposer messages if the
+configuration generation in them is less than its current one. Namely, it
+refuses to vote, to truncate WAL in `handle_elected` and to accept WAL. In
+response it sends its current configuration generation to let walproposer know.
+
+Safekeeper gets `PUT /v1/tenants/{tenant_id}/timelines/{timeline_id}/configuration` 
+accepting `Configuration`. Safekeeper switches to the given conf it is higher than its
+current one and ignores it otherwise. In any case it replies with
+```
+struct ConfigurationSwitchResponse {
+    conf: Configuration,
+    term: Term,
+    last_log_term: Term,
+    flush_lsn: Lsn,
+}
+```
+
+### Compute (walproposer) changes
+
+Basic rule is that joint configuration requires votes from majorities in the
+both `set` and `new_sk_set`.
+
+Compute receives list of safekeepers to connect to from the control plane as
+currently and tries to communicate with all of them. However, the list does not
+define consensus members. Instead, on start walproposer tracks highest
+configuration it receives from `AcceptorGreeting`s. Once it assembles greetings
+from majority of `sk_set` and majority of `new_sk_set` (if it is present), it
+establishes this configuration as its own and moves to voting. 
+
+It should stop talking to safekeepers not listed in the configuration at this
+point, though it is not unsafe to continue doing so.
+
+To be elected it must receive votes from both majorites if `new_sk_set` is present.
+Similarly, to commit WAL it must receive flush acknowledge from both majorities.
+
+If walproposer hears from safekeeper configuration higher than his own (i.e.
+refusal to accept due to configuration change) it simply restarts.
+
+### Change algorithm
+
+The following algorithm can be executed anywhere having access to configuration
+storage and safekeepers. It is safe to interrupt / restart it and run multiple
+instances of it concurrently, though likely one of them won't make
+progress then. It accepts `desired_set: Vec<NodeId>` as input. 
+
+Algorithm will refuse to make the change if it encounters previous interrupted
+change attempt, but in this case it will try to finish it.
+
+It will eventually converge if old majority, new majority and configuration
+storage are reachable.
+
+1) Fetch current timeline configuration from the configuration storage.
+2) If it is already joint one and `new_set` is different from `desired_set`
+   refuse to change. However, assign join conf to (in memory) var
+   `join_conf` and proceed to step 4 to finish the ongoing change.
+3) Else, create joint `joint_conf: Configuration`: increment current conf number
+   `n` and put `desired_set` to `new_sk_set`. Persist it in the configuration
+   storage by doing CAS on the current generation: change happens only if
+   current configuration number is still `n`. Apart from guaranteeing uniqueness
+   of configurations, CAS linearizes them, ensuring that new configuration is
+   created only following the previous one when we know that the transition is
+   safe. Failed CAS aborts the procedure.
+4) Call `PUT` `configuration` on safekeepers from the current set,
+   delivering them `joint_conf`. Collecting responses from majority is required
+   to proceed. If any response returned generation higher than 
+   `joint_conf.generation`, abort (another switch raced us). Otherwise, choose
+   max `<last_log_term, flush_lsn>` among responses and establish it as
+   (in memory) `sync_position`. Also choose max `term` and establish it as (in
+   memory) `sync_term`. We can't finish the switch until majority of the new set
+   catches up to this `sync_position` because data before it could be committed
+   without ack from the new set. Similarly, we'll bump term on new majority
+   to `sync_term` so that two computes with the same term are never elected.
+4) Initialize timeline on safekeeper(s) from `new_sk_set` where it
+   doesn't exist yet by doing `pull_timeline` from the majority of the 
+   current set. Doing that on majority of `new_sk_set` is enough to
+   proceed, but it is reasonable to ensure that all `new_sk_set` members
+   are initialized -- if some of them are down why are we migrating there?
+5) Call `POST` `bump_term(sync_term)` on safekeepers from the new set. 
+   Success on majority is enough.
+6) Repeatedly call `PUT` `configuration` on safekeepers from the new set,
+   delivering them `joint_conf` and collecting their positions. This will
+   switch them to the `joint_conf` which generally won't be needed 
+   because `pull_timeline` already includes it and plus additionally would be
+   broadcast by compute. More importantly, we may proceed to the next step
+   only when `<last_log_term, flush_lsn>` on the majority of the new set reached 
+   `sync_position`. Similarly, on the happy path no waiting is not needed because 
+   `pull_timeline` already includes it. However, we should double
+    check to be safe. For example, timeline could have been created earlier e.g.
+    manually or after try-to-migrate, abort, try-to-migrate-again sequence. 
+7) Create `new_conf: Configuration` incrementing `join_conf` generation and having new 
+   safekeeper set as `sk_set` and None `new_sk_set`. Write it to configuration 
+   storage under one more CAS.
+8) Call `PUT` `configuration` on safekeepers from the new set,
+   delivering them `new_conf`. It is enough to deliver it to the majority 
+   of the new set; the rest can be updated by compute.
+
+I haven't put huge effort to make the description above very precise, because it
+is natural language prone to interpretations anyway. Instead I'd like to make TLA+
+spec of it.
+
+Description above focuses on safety. To make the flow practical and live, here a few more 
+considerations.
+1) It makes sense to ping new set to ensure it we are migrating to live node(s) before 
+  step 3.
+2) If e.g. accidentally wrong new sk set has been specified, before CAS in step `6` is completed 
+   it is safe to rollback to the old conf with one more CAS.
+3) On step 4 timeline might be already created on members of the new set for various reasons; 
+   the simplest is the procedure restart. There are more complicated scenarious like mentioned
+   in step 5. Deleting and re-doing `pull_timeline` is generally unsafe without involving 
+   generations, so seems simpler to treat existing timeline as success. However, this also 
+   has a disadvantage: you might imagine an surpassingly unlikely schedule where condition in
+   the step 5 is never reached until compute is (re)awaken up to synchronize new member(s).
+   I don't think we'll observe this in practice, but can add waking up compute if needed.
+4) In the end timeline should be locally deleted on the safekeeper(s) which are
+   in the old set but not in the new one, unless they are unreachable. To be
+   safe this also should be done under generation number (deletion proceeds only if 
+   current configuration is <= than one in request and safekeeper is not memeber of it).
+5) If current conf fetched on step 1 is already not joint and members equal to `desired_set`,
+   jump to step 7, using it as `new_conf`.
+
+## Implementation
+
+The procedure ought to be driven from somewhere. Obvious candidates are control
+plane and storage_controller; and as each of them already has db we don't want
+yet another storage. I propose to manage safekeepers in storage_controller
+because 1) since it is in rust it simplifies simulation testing (more on this
+below) 2) it already manages pageservers. 
+
+This assumes that migration will be fully usable only after we migrate all
+tenants/timelines to storage_controller. It is discussible whether we want also
+to manage pageserver attachments for all of these, but likely we do.
+
+This requires us to define storcon <-> cplane interface.
+
+### storage_controller <-> control plane interface
+
+First of all, control plane should
+[change](https://neondb.slack.com/archives/C03438W3FLZ/p1719226543199829)
+storing safekeepers per timeline instead of per tenant because we can't migrate
+tenants atomically. 
+
+The important question is how updated configuration is delivered from
+storage_controller to control plane to provide it to computes. As always, there
+are two options, pull and push. Let's do it the same push as with pageserver
+`/notify-attach` because 1) it keeps storage_controller out of critical compute
+start path 2) provides easier upgrade: there won't be such a thing as 'timeline
+managed by control plane / storcon', cplane just takes the value out of its db
+when needed 3) uniformity. It makes storage_controller responsible for retrying notifying
+control plane until it succeeds.
+
+So, cplane `/notify-safekeepers` for the timeline accepts `Configuration` and
+updates it in the db if the provided conf generation is higher (the cplane db
+should also store generations for this). Similarly to [`/notify-attach`](https://www.notion.so/neondatabase/Storage-Controller-Control-Plane-interface-6de56dd310a043bfa5c2f5564fa98365), it
+should update db which makes the call successful, and then try to schedule
+`apply_config` if possible, it is ok if not. storage_controller 
+should rate limit calling the endpoint, but likely this won't be needed, as migration
+throughput is limited by `pull_timeline`.
+
+Timeline (branch) creation in cplane should call storage_controller POST
+`tenant/:tenant_id/timeline` like it currently does for sharded tenants.
+Response should be augmented with `safekeeper_conf: Configuration`. The call
+should be retried until succeeds.
+
+Timeline deletion and tenant deletion in cplane should call appropriate
+storage_controller endpoints like it currently does for sharded tenants. The
+calls should be retried until they succeed.
+
+### storage_controller implementation
+
+Current 'load everything on startup and keep in memory' easy design is fine.
+Single timeline shouldn't take more than 100 bytes (it's 16 byte tenant_id, 16
+byte timeline_id, int generation, vec of ~3 safekeeper ids plus some flags), so
+10^6 of timelines shouldn't take more than 100MB.
+
+Similar to pageserver attachment Intents storage_controller would have in-memory
+`MigrationRequest` (or its absense) for each timeline and pool of tasks trying
+to make these request reality; this ensures one instance of storage_controller
+won't do several migrations on the same timeline concurrently. In the first
+version it is simpler to have more manual control and no retries, i.e. migration
+failure removes the request. Later we can build retries and automatic
+scheduling/migration. `MigrationRequest` is
+```
+enum MigrationRequest {
+    To(Vec<NodeId>),
+    FinishPending,
+}
+```
+
+`FinishPending` requests to run the procedure to ensure state is clean: current
+configuration is not joint and majority of safekeepers are aware of it, but do
+not attempt to migrate anywhere. If current configuration fetched on step 1 is
+not joint it jumps to step 7. It should be run at startup for all timelines (but
+similarly, in the first version it is ok to trigger it manually).
+
+#### Schema
+
+`safekeepers` table mirroring current `nodes` should be added, except that for
+`scheduling_policy` field (seems like `status` is a better name for it): it is enough
+to have at least in the beginning only 3 fields: 1) `active` 2) `offline` 3)
+`decomissioned`.
+
+`timelines` table:
+```
+table! {
+    // timeline_id is primary key
+    timelines (tenant_id, timeline_id) {
+        timeline_id -> Varchar,
+        tenant_id -> Varchar,
+        generation -> Int4,
+        sk_set -> Array<Int4>, // list of safekeeper ids
+        new_sk_set -> Nullable<Array<Int4>>, // list of safekeeper ids, null if not joint conf
+        cplane_notified_generation -> Int4,
+    }
+}
+```
+
+#### API
+
+Node management is similar to pageserver:
+1) POST `/control/v1/safekeepers` upserts safekeeper.
+2) GET `/control/v1/safekeepers` lists safekeepers.
+3) GET `/control/v1/safekeepers/:node_id` gets safekeeper.
+4) PUT `/control/v1/safekepers/:node_id/status` changes status to e.g.
+   `offline` or `decomissioned`. Initially it is simpler not to schedule any
+    migrations here.
+
+Safekeeper deploy scripts should register safekeeper at storage_contorller as
+they currently do with cplane, under the same id.
+
+Timeline creation/deletion: already existing POST `tenant/:tenant_id/timeline`
+would 1) choose initial set of safekeepers; 2) write to the db initial
+`Configuration` with `INSERT ON CONFLICT DO NOTHING` returning existing row in
+case of conflict; 3) create timeline on the majority of safekeepers (already
+created is ok).
+
+We don't want to block timeline creation when one safekeeper is down. Currently
+this is solved by compute implicitly creating timeline on any safekeeper it is
+connected to. This creates ugly timeline state on safekeeper when timeline is
+created, but start LSN is not defined yet. It would be nice to remove this; to
+do that, controller can in the background retry to create timeline on
+safekeeper(s) which missed that during initial creation call. It can do that
+through `pull_timeline` from majority so it doesn't need to remember
+`parent_lsn` in its db.
+
+Timeline deletion removes the row from the db and forwards deletion to the
+current configuration members. Without additional actions deletions might leak,
+see below on this; initially let's ignore these, reporting to cplane success if
+at least one safekeeper deleted the timeline (this will remove s3 data).
+
+Tenant deletion repeats timeline deletion for all timelines.
+
+Migration API: the first version is the simplest and the most imperative:
+1) PUT `/control/v1/safekeepers/migrate` schedules `MigrationRequest`s to move
+all timelines from one safekeeper to another. It accepts json
+```
+{
+    "src_sk": u32,
+    "dst_sk": u32,
+    "limit": Optional<u32>,
+}
+```
+
+Returns list of scheduled requests.
+
+2) PUT `/control/v1/tenant/:tenant_id/timeline/:timeline_id/safekeeper_migrate` schedules `MigrationRequest`
+   to move single timeline to given set of safekeepers:
+```
+{
+    "desired_set": Vec<u32>,
+}
+```
+
+Returns scheduled request.
+
+Similar call should be added for the tenant.
+
+It would be great to have some way of subscribing to the results (apart from
+looking at logs/metrics).
+
+Migration is executed as described above. One subtlety is that (local) deletion on
+source safekeeper might fail, which is not a problem if we are going to
+decomission the node but leaves garbage otherwise. I'd propose in the first version
+1) Don't attempt deletion at all if node status is `offline`.
+2) If it failed, just issue warning.
+And add PUT `/control/v1/safekeepers/:node_id/scrub` endpoint which would find and 
+remove garbage timelines for manual use. It will 1) list all timelines on the 
+safekeeper 2) compare each one against configuration storage: if timeline 
+doesn't exist at all (had been deleted), it can be deleted. Otherwise, it can 
+be deleted under generation number if node is not member of current generation.
+
+Automating this is untrivial; we'd need to register all potential missing
+deletions <tenant_id, timeline_id, generation, node_id> in the same transaction
+which switches configurations. Similarly when timeline is fully deleted to
+prevent cplane operation from blocking when some safekeeper is not available
+deletion should be also registered.
+
+One more task pool should infinitely retry notifying control plane about changed
+safekeeper sets.
+
+3) GET `/control/v1/tenant/:tenant_id/timeline/:timeline_id/` should return
+   current in memory state of the timeline and pending `MigrationRequest`,
+   if any.
+
+4) PUT `/control/v1/tenant/:tenant_id/timeline/:timeline_id/safekeeper_migrate_abort` tries to abort the
+   migration by switching configuration from the joint to the one with (previous) `sk_set` under CAS
+   (incrementing generation as always).
+
+#### Dealing with multiple instances of storage_controller
+
+Operations described above executed concurrently might create some errors but do
+not prevent progress, so while we normally don't want to run multiple instances
+of storage_controller it is fine to have it temporarily, e.g. during redeploy.
+
+Any interactions with db update in-memory controller state, e.g. if migration
+request failed because different one is in progress, controller remembers that
+and tries to finish it.
+
+## Testing
+
+`neon_local` should be switched to use storage_controller, playing role of
+control plane.
+
+There should be following layers of tests:
+1) Model checked TLA+ spec specifies the algorithm and verifies its basic safety.
+
+2) To cover real code and at the same time test many schedules we should have
+   simulation tests. For that, configuration storage, storage_controller <->
+   safekeeper communication and pull_timeline need to be mocked and main switch
+   procedure wrapped to as a node (thread) in simulation tests, using these
+   mocks. Test would inject migrations like it currently injects
+   safekeeper/walproposer restars. Main assert is the same -- committed WAL must
+   not be lost.
+
+3) Since simulation testing injects at relatively high level points (not
+   syscalls), it omits some code, in particular `pull_timeline`. Thus it is
+   better to have basic tests covering whole system as well. Extended version of
+   `test_restarts_under_load` would do: start background load and do migration 
+   under it, then restart endpoint and check that no reported commits 
+   had been lost. I'd also add one more creating classic network split scenario, with
+   one compute talking to AC and another to BD while migration from nodes ABC to ABD
+   happens.
+
+4) Simple e2e test should ensure that full flow including cplane notification works.
+
+## Order of implementation and rollout
+
+Note that 
+- Control plane parts and integration with it is fully independent from everything else
+  (tests would use simulation and neon_local).
+- There is a lot of infra work making storage_controller aware of timelines and safekeepers
+  and its impl/rollout should be separate from migration itself.
+- Initially walproposer can just stop working while it observers joint configuration.
+  Such window would be typically very short anyway.
+
+To rollout smoothly, both walproposer and safekeeper should have flag
+`configurations_enabled`; when set to false, they would work as currently, i.e.
+walproposer is able to commit on whatever safekeeper set it is provided. Until
+all timelines are managed by storcon we'd need to use current script to migrate
+and update/drop entries in the storage_controller database if it has any.
+
+Safekeepers would need to be able to talk both current and new protocol version
+with compute to reduce number of computes restarted in prod once v2 protocol is
+deployed (though before completely switching we'd need to force this).
+
+Let's have the following rollout order:
+- storage_controller becomes aware of safekeepers;
+- storage_controller gets timeline creation for new timelines and deletion requests, but
+  doesn't manage all timelines yet. Migration can be tested on these new timelines.
+  To keep control plane and storage_controller databases in sync while control 
+  plane still chooses the safekeepers initially (until all timelines are imported
+  it can choose better), `TimelineCreateRequest` can get optional safekeepers
+  field with safekeepers chosen by cplane.
+- Then we can import all existing timelines from control plane to
+  storage_controller and gradually enable configurations region by region.
+
+
+Very rough implementation order:
+- Add concept of configurations to safekeepers (including control file),
+  implement v3 protocol.
+- Implement walproposer changes, including protocol.
+- Implement storconn part. Use it in neon_local (and pytest).
+- Make cplane store safekeepers per timeline instead of per tenant.
+- Implement cplane/storcon integration. Route branch creation/deletion 
+  through storcon. Then we can test migration of new branches.
+- Finally import existing branches. Then we can drop cplane 
+  safekeeper selection code. Gradually enable configurations at 
+  computes and safekeepers. Before that, all computes must talk only
+  v3 protocol version.
+
+## Integration with evicted timelines
+
+Currently, `pull_timeline` doesn't work correctly with evicted timelines because
+copy would point to original partial file. To fix let's just do s3 copy of the
+file. It is a bit stupid as generally unnecessary work, but it makes sense to
+implement proper migration before doing smarter timeline archival. [Issue](https://github.com/neondatabase/neon/issues/8542)
+
+## Possible optimizations
+
+Steps above suggest walproposer restart (with re-election) and thus reconnection
+to safekeepers. Since by bumping term on new majority we ensure that leader
+terms are unique even across generation switches it is possible to preserve
+connections. However, it is more complicated, reconnection is very fast and it
+is much more important to avoid compute restart than millisecond order of write
+stall.
+
+Multiple joint consensus: algorithm above rejects attempt to change membership
+while another attempt is in progress. It is possible to overlay them and AFAIK
+Aurora does this but similarly I don't think this is needed.
+
+## Misc
+
+We should use Compute <-> safekeeper protocol change to include other (long
+yearned) modifications:
+- send data in network order to make arm work.
+- remove term_start_lsn from AppendRequest
+- add horizon to TermHistory
+- add to ProposerGreeting number of connection from this wp to sk

docs/rfcs/035-timeline-archive.md

@@ -0,0 +1,507 @@
+# Timeline Archival
+
+## Summary
+
+This RFC describes a mechanism for pageservers to eliminate local storage + compute work
+for timelines which are not in use, in response to external API calls to "archive" a timeline.
+
+The archived state roughly corresponds to fully offloading a timeline to object storage, such
+that its cost is purely the cost of that object storage.
+
+## Motivation
+
+Archived timelines serve multiple purposes:
+- Act as a 'snapshot' for workloads that would like to retain restorable copies of their
+  database from longer ago than their PITR window.
+- Enable users to create huge numbers of branches (e.g. one per github PR) without having
+  to diligently clean them up later to avoid overloading the pageserver (currently we support
+  up to ~500 branches per tenant).
+
+### Prior art
+
+Most storage and database systems have some form of snapshot, which can be implemented several ways:
+1. full copies of data (e.g. an EBS snapshot to S3)
+2. shallow snapshots which are CoW relative to the original version of the data, e.g. on a typical NFS appliance, or a filesystem like CephFS.
+3. a series of snapshots which are CoW or de-duplicated relative to one another.
+
+Today's Neon branches are approximately like `2.`, although due to implementation details branches
+often end up storing much more data than they really need, as parent branches assume that all data
+at the branch point is needed.  The layers pinned in the parent branch may have a much larger size
+than the physical size of a compressed image layer representing the data at the branch point.
+
+## Requirements
+
+- Enter & exit the archived state in response to external admin API calls
+- API calls to modify the archived state are atomic and durable
+- An archived timeline should eventually (once out of PITR window) use an efficient compressed
+  representation, and avoid retaining arbitrarily large data in its parent branch.
+- Remote object GETs during tenant start may be O(N) with the number of _active_ branches,
+  but must not scale with the number of _archived_ branches.
+- Background I/O for archived branches should only be done a limited number of times to evolve them
+  to a long-term-efficient state (e.g. rewriting to image layers).  There should be no ongoing "housekeeping"
+  overhead for archived branches, including operations related to calculating sizes for billing.
+- The pageserver should put no load on the safekeeper for archived branches.
+- Performance of un-archiving a branch must make good use of S3/disk bandwidth to restore the branch
+  to a performant state in a short time (linear with the branch's logical size)
+
+## Non Goals
+
+- Archived branches are not a literal `fullbackup` postgres snapshot: they are still stored
+  in Neon's internal format.
+- Compute cold starts after activating an archived branch will not have comparable performance to
+  cold starts on an active branch.
+- Archived branches will not use any new/additional compression or de-duplication beyond what
+  is already implemented for image layers (zstd per page).
+- The pageserver will not "auto start" archived branches in response to page_service API requests: they
+  are only activated explicitly via the HTTP API.
+- We will not implement a total offload of archived timelines from safekeepers: their control file (small) will
+  remain on local disk, although existing eviction mechanisms will remove any segments from local disk.
+- We will not expose any prometheus metrics for archived timelines, or make them visible in any
+  detailed HTTP APIs other than the specific API for listing archived timelines.
+- A parent branch may not be archived unless all its children are.
+
+## Impacted Components
+
+pageserver, storage controller
+
+## Terminology
+
+**Archived**: a branch is _archived_ when an HTTP API request to archive it has succeeded: the caller
+may assume that this branch is now very cheap to store, although this may not be physically so until the
+branch proceeds to the offloaded state.
+
+**Active** branches are branches which are available for use by page_service clients, and have a relatively
+high cost due to consuming local storage.
+
+**Offloaded** branches are a subset of _archived_ branches, which have had their local state removed such
+that they now consume minimal runtime resources and have a cost similar to the cost of object storage.
+
+**Activate** (verb): transition from Archived to Active
+
+**Archive** (verb): transition from Active to Archived
+
+**Offload** (verb): transition from Archived to Offloaded
+
+**Offload manifest**: an object stored in S3 that describes timelines which pageservers do not load.
+
+**Warm up** (verb): operation done on an active branch, by downloading its active layers.  Once a branch is
+warmed up, good performance will be available to page_service clients.
+
+## Implementation
+
+### High level flow
+
+We may think of a timeline which is archived and then activated as proceeding through a series of states:
+
+```mermaid
+stateDiagram
+  [*] --> Active(warm)
+  Active(warm) --> Archived
+  Archived --> Offloaded
+  Archived --> Active(warm)
+  Offloaded --> Active(cold)
+  Active(cold) --> Active(warm)
+```
+
+Note that the transition from Archived to Active(warm) is expected to be fairly rare: the most common lifecycles
+of branches will be:
+- Very frequent: Short lived branches: Active -> Deleted
+- Frequent: Long-lived branches: Active -> Archived -> Offloaded -> Deleted
+- Rare: Branches used to restore old state: Active ->Archived -> Offloaded -> Active
+
+These states are _not_ all stored as a single physical state on the timeline, but rather represent the combination
+of:
+- the timeline's lifecycle state: active or archived, stored in the timeline's index
+- its offload state: whether pageserver has chosen to drop local storage of the timeline and write it into the
+  manifest of offloaded timelines.
+- cache state (whether it's warm or cold).
+
+### Storage format changes
+
+There are two storage format changes:
+1. `index_part.json` gets a new attribute `state` that describes whether the timeline is to
+   be considered active or archived.
+2. A new tenant-level _manifest_ object `tenant_manifest-v1.json` describes which timelines a tenant does not need to load
+   at startup (and is available for storing other small, rarely changing tenant-wide attributes in future)
+
+The manifest object will have a format like this:
+```
+{
+  "offload_timelines": [
+    {
+      "timeline_id": ...
+      "last_record_lsn": ...
+      "last_record_lsn_time": ...
+      "pitr_interval": ...
+      "last_gc_lsn": ...  # equal to last_record_lsn if this branch has no history (i.e. a snapshot)
+      "logical_size": ...  # The size at last_record_lsn
+      "physical_size" ...
+      "parent": Option<{
+        "timeline_id"...
+        "lsn"... # Branch point LSN on the parent
+        "requires_data": bool # True if this branch depends on layers in its parent, identify it here
+
+      }>
+    }
+  ]
+}
+```
+
+The information about a timeline in its offload state is intentionally minimal: just enough to decide:
+- Whether it requires [archive optimization](#archive-branch-optimization) by rewriting as a set of image layers: we may infer this
+  by checking if now > last_record_lsn_time - pitr_interval, and pitr_lsn < last_record_lsn.
+- Whether a parent branch should include this offloaded branch in its GC inputs to avoid removing
+  layers that the archived branch depends on
+- Whether requests to delete this `timeline_id` should be executed (i.e. if a deletion request
+  is received for a timeline_id that isn't in the site of live `Timelines` or in the manifest, then
+  we don't need to go to S3 for the deletion.
+- How much archived space to report in consumption metrics
+
+The contents of the manifest's offload list will also be stored as an attribute of `Tenant`, such that the total
+set of timelines may be found by the union of `Tenant::timelines` (non-offloaded timelines) and `Tenant::offloaded`
+(offloaded timelines).
+
+For split-brain protection, the manifest object will be written with a generation suffix, in the same way as
+index_part objects are (see [generation numbers RFC](025-generation-numbers.md)).  This will add some complexity, but
+give us total safety against two pageservers with the same tenant attached fighting over the object.  Existing code
+for finding the latest generation and for cleaning up old generations (in the scrubber) will be generalized to cover
+the manifest file.
+
+### API & Timeline state
+
+Timelines will store a lifecycle state (enum of Active or Archived) in their IndexPart.  This will
+be controlled by a new per-timeline `configure` endpoint.  This is intentionally generic naming, which
+may be used in future to control other per-timeline attributes (e.g. in future we may make PITR interval
+a per-timeline configuration).
+
+`PUT /v1/tenants/{tenant_id}/timelines/{timeline_id}/configure`
+```
+{
+  'state': 'active|archive'
+}
+```
+
+When archiving a timeline, this API will complete as soon as the timeline's state has been set in index_part, and that index has been uploaded.
+
+When activating a timeline, this API will complete as soon as the timeline's state has been set in index_part,
+**and** the `Timeline` object has been instantiated and activated.  This will require reading the timeline's
+index, but not any data: it should be about as fast as a couple of small S3 requests.
+
+The API will be available with identical path via the storage controller: calling this on a sharded tenant
+will simply map the API call to all the shards.
+
+Archived timelines may never have descendent timelines which are active.  This will be enforced at the API level,
+such that activating a timeline requires that all its ancestors are active, and archiving a timeline requires
+that all its descendents are archived.  It is the callers responsibility to walk the hierarchy of timelines
+in the proper order if they would like to archive whole trees of branches.
+
+Because archive timelines will be excluded from the usual timeline listing APIs, a new API specifically
+for archived timelines will be added: this is for use in support/debug:
+
+```
+GET /v1/tenants/{tenant_id}/archived_timelines
+
+{
+  ...same per-timeline content as the tenant manifest...
+}
+
+```
+
+### Tenant attach changes
+
+Currently, during Tenant::spawn we list all the timelines in the S3 bucket, and then for each timeline
+we load their index_part.json.  To avoid the number of GETs scaling linearly with the number of archived
+timelines, we must have a single object that tells us which timelines do not need to be loaded.  The
+number of ListObjects requests while listing timelines will still scale O(N), but this is less problematic
+because each request covers 1000 timelines.
+
+This is **not** literally the same as the set of timelines who have state=archived.  Rather, it is
+the set of timelines which have been offloaded in the background after their state was set to archived.
+
+We may simply skip loading these timelines: there will be no special state of `Timeline`, they just won't
+exist from the perspective of an active `Tenant` apart from in deletion: timeline deletion will need
+to check for offloaded timelines as well as active timelines, to avoid wrongly returning 404 on trying
+to delete an offloaded timeline.
+
+### Warm-up API
+
+`PUT /v1/tenants/{tenant_id}/timelines/{timeline_id}/download?wait_ms=1234`
+
+This API will be similar to the existing `download_remote_layers` API, but smarter:
+- It will not download _all_ remote layers, just the visible set (i.e. layers needed for a read)
+- It will download layers in the visible set until reaching `wait_ms`, then return a struct describing progress
+  of downloads, so that the caller can poll.
+
+The _visible set_ mentioned above will be calculated by the pageserver in the background, by taking the set
+of readable LSNs (i.e. branch points and heads of branches), and walking the layer map to work out which layers
+can possibly be read from these LSNs.  This concept of layer visibility is more generally useful for cache
+eviction and heatmaps, as well as in this specific case of warming up a timeline.
+
+The caller does not have to wait for the warm up API, or call it at all.  But it is strongly advised
+to call it, because otherwise populating local contents for a timeline can take a long time when waiting
+for SQL queries to coincidentally hit all the layers, and during that time query latency remains quite
+volatile.
+
+### Background work
+
+Archived branches are not subject to normal compaction.  Instead, when the compaction loop encounters
+an archived branch, it will consider rewriting the branch to just image layers if the branch has no history
+([archive branch optimization](#archive-branch-optimization)), or offloading the timeline from local disk
+if its state permits that.
+
+Additionally, the tenant compaction task will walk the state of already offloaded timelines to consider
+optimizing their storage, e.g. if a timeline had some history when offloaded, but since then its PITR
+has elapsed and it can now be rewritten to image layers.
+
+#### Archive branch offload
+
+Recall that when we archive a timeline via the HTTP API, this only sets a state: it doesn't do
+any actual work.
+
+This work is done in the background compaction loop.  It makes sense to tag this work on to the compaction
+loop, because it is spiritually aligned: offloading data for archived branches improves storage efficiency.
+
+The condition for offload is simple:
+ - a `Timeline` object exists with state `Archived`
+ - the timeline does not have any non-offloaded children.
+ 
+ Regarding the condition that children must be offloaded, this will always be eventually true, because
+ we enforce at the API level that children of archived timelines must themselves be archived, and all
+ archived timelines will eventually be offloaded.
+
+Offloading a timeline is simple:
+- Read the timeline's attributes that we will store in its offloaded state (especially its logical size)
+- Call `shutdown()` on the timeline and remove it from the `Tenant` (as if we were about to delete it)
+- Erase all the timeline's content from local storage (`remove_dir_all` on its path)
+- Write the tenant manifest to S3 to prevent this timeline being loaded on next start.
+
+#### Archive branch optimization (flattening)
+
+When we offloaded a branch, it might have had some history that prevented rewriting it to a single
+point in time set of image layers.  For example, a branch might have several days of writes and a 7
+day PITR: when we archive it, it still has those days of history.
+
+Once the PITR has expired, we have an opportunity to reduce the physical footprint of the branch by:
+- Writing compressed image layers within the archived branch, as these are more efficient as a way of storing
+  a point in time compared with delta layers
+- Updating the branch's offload metadata to indicate that this branch no longer depends on its ancestor
+  for data, i.e. the ancestor is free to GC layers files at+below the branch point
+
+Fully compacting an archived branch into image layers at a single LSN may be thought of as *flattening* the
+branch, such that it is now a one-dimensional keyspace rather than a two-dimensional key/lsn space. It becomes
+a true snapshot at that LSN.
+
+It is not always more efficient to flatten a branch than to keep some extra history on the parent: this
+is described in more detail in [optimizations](#delaying-storage-optimization-if-retaining-parent-layers-is-cheaper)
+
+Archive branch optimization should be done _before_ background offloads during compaction, because there may
+be timelines which are ready to be offloaded but also would benefit from the optimization step before
+being offloaded.  For example, a branch which has already fallen out of PITR window and has no history
+of its own may be immediately re-written as a series of image layers before being offloaded.
+
+### Consumption metrics
+
+Archived timelines and offloaded timelines will be excluded from the synthetic size calculation, in anticipating
+that billing structures based on consumption metrics are highly likely to apply different $/GB rates to archived
+vs. ordinary content.
+
+Archived and offloaded timelines' logical size will be reported under the existing `timeline_logical_size`
+variant of `MetricsKey`: receivers are then free to bill on this metric as they please.
+
+### Secondary locations
+
+Archived timelines (including offloaded timelines) will be excluded from heatmaps, and thereby
+when a timeline is archived, after the next cycle of heatmap upload & secondary download, its contents
+will be dropped from secondary locations.
+
+### Sharding
+
+Archiving or activating a timeline will be done symmetrically across all shards in a tenant, in
+the same way that timeline creation and deletion is done.  There are no special rules about ordering:
+the storage controller may dispatch concurrent calls to all shards when archiving or activating a timeline.
+
+Since consumption metrics are only transmitted from shard zero, the state of archival on this shard
+will be authoritative for consumption metrics.
+
+## Error cases
+
+### Errors in sharded tenants
+
+If one shard in a tenant fails an operation but others succeed, the tenant may end up in a mixed
+state, where a timeline is archived on some shards but not on others.  
+
+We will not bother implementing a rollback mechanism for this: errors in archiving/activating a timeline
+are either transient (e.g. S3 unavailable, shutting down), or the fault of the caller (NotFound, BadRequest).
+In the transient case callers are expected to retry until success, or to make appropriate API calls to clear
+up their mistake.  We rely on this good behavior of callers to eventually get timelines into a consistent
+state across all shards.  If callers do leave a timeline in an inconsistent state across shards, this doesn't
+break anything, it's just "weird".
+
+This is similar to the status quo for timeline creation and deletion: callers are expected to retry
+these operations until they succeed.
+
+### Archiving/activating
+
+Archiving/activating a timeline can fail in a limited number of ways:
+1. I/O error storing/reading the timeline's updated index
+    - These errors are always retryable: a fundamental design assumption of the pageserver is that remote
+      storage errors are always transient. 
+2. NotFound if the timeline doesn't exist
+    - Callers of the API are expected to avoid calling deletion and archival APIs concurrently.
+    - The storage controller has runtime locking to prevent races such as deleting a timeline while
+      archiving it.
+3. BadRequest if the rules around ancestors/descendents of archived timelines would be violated
+    - Callers are expected to do their own checks to avoid hitting this case.  If they make
+      a mistake and encounter this error, they should give up.
+
+### Offloading
+
+Offloading can only fail if remote storage is unavailable, which would prevent us from writing the
+tenant manifest.  In such error cases, we give up in the expectation that offloading will be tried 
+again at the next iteration of the compaction loop.
+
+### Archive branch optimization
+
+Optimization is a special form of compaction, so can encounter all the same errors as regular compaction
+can: it should return Result<(), CompactionError>, and as with compaction it will be retried on
+the next iteration of the compaction loop.
+
+## Optimizations
+
+### Delaying storage optimization if retaining parent layers is cheaper
+
+Optimizing archived branches to image layers and thereby enabling parent branch GC to progress
+is a safe default: archived branches cannot over-fill a pageserver's local disk, and once they
+are offloaded to S3 they're totally safe, inert things.
+
+However, in some cases it can be advantageous to retain extra history on their parent branch rather
+than flattening the archived branch.  For example, if a 1TB parent branch is rather slow-changing (1GB
+of data per day), and archive branches are being created nightly, then writing out full 1TB image layers
+for each nightly branch is inefficient compared with just keeping more history on the main branch.
+
+Getting this right requires consideration of:
+- Compaction: if keeping more history on the main branch is going to prompt the main branch's compaction to
+  write out extra image layers, then it might make more sense to just write out the image layers on
+  the archived branch.
+- Metadata bloat: keeping extra history on a parent branch doesn't just cost GB of storage, it makes
+  the layer map (and index_part) bigger.  There are practical limits beyond which writing an indefinitely
+  large layer map can cause problems elsewhere.
+
+This optimization can probably be implemented quite cheaply with some basic heuristics like:
+- don't bother doing optimization on an archive branch if the LSN distance between
+  its branch point and the end of the PITR window is <5% of the logical size of the archive branch.
+- ...but, Don't keep more history on the main branch than double the PITR
+
+### Creating a timeline in archived state (a snapshot)
+
+Sometimes, one might want to create a branch with no history, which will not be written to
+before it is archived.  This is a snapshot, although we do not require a special snapshot API,
+since a snapshot can be represented as a timeline with no history.
+
+This can be accomplished by simply creating a timeline and then immediately archiving it, but
+that is somewhat wasteful: this timeline it will spin up various tasks and open a connection to the storage
+broker to try and ingest WAL, before being shutdown in the subsequent archival call.  To explicitly
+support this common special case, we may add a parameter to the timeline creation API which
+creates a timeline directly into the archived state.
+
+Such a timeline creation will do exactly two I/Os at creation time:
+- write the index_part object to record the timeline's existence
+- when the timeline is offloaded in the next iteration of the compaction loop (~20s later),
+  write the tenant manifest.
+
+Later, when the timeline falls off the end of the PITR interval, the usual offload logic will wake
+up the 'snapshot' branch and write out image layers.
+
+## Future Work
+
+### Enabling `fullbackup` dumps from archive branches
+
+It would be useful to be able to export an archive branch to another system, or for use in a local
+postgres database.
+
+This could be implemented as a general capability for all branches, in which case it would "just work"
+for archive branches by activating them.  However, downloading all the layers in a branch just to generate
+a fullbackup is a bit inefficient: we could implement a special case for flattened archived branches
+which streams image layers from S3 and outputs the fullbackup stream without writing the layers out to disk.
+
+Implementing `fullbackup` is a bit more complicated than this because of sharding, but solving that problem
+is unrelated to the topic of archived branches (it probably involves having each shard write out a fullbackup 
+stream to S3 in an intermediate format and, then having one node stitch them together).
+
+### Tagging layers from archived branches
+
+When we know a layer is an image layer written for an archived branch that has fallen off the PITR window,
+we may add tags to the S3 objects to enable writing lifecycle policies that transition such layers to even
+cheaper storage.
+
+This could be done for all archived layers, or it could be driven by the archival API, to give the pageserver
+external hints on which branches are likely to be reactivated, and which branches are good candidates for
+tagging for low performance storage.
+
+Tagging+lifecycles is just one mechanism: one might also directly use S3 storage classes.  Other clouds' object
+stores have similar mechanisms.
+
+### Storing sequences of archive branches as deltas
+
+When archived branches are used as scheduled snapshots, we could store them even more efficiently
+by encoding them as deltas relative to each other (i.e. for nightly snapshots, when we do the
+storage optimization for Tuesday's snapshot, we would read Monday's snapshot and store only the modified
+pages). This is the kind of encoding that many backup storage systems use.
+
+The utility of this depends a lot on the churn rate of the data, and the cost of doing the delta encoding
+vs. just writing out a simple stream of the entire database.  For smaller databases, writing out a full
+copy is pretty trivial (e.g. writing a compressed copy of a 10GiB database to S3 can take under 10 seconds,
+so the complexity tradeoff of diff-encoding it is dubious).
+
+One does not necessarily have to read-back the previous snapshot in order to encoded the next one: if the
+pageserver knows about the schedule, it can intentionally retain extra history on the main branch so that
+we can say: "A branch exists from Monday night.  I have Monday night's data still active in the main branch,
+so now I can read at the Monday LSN and the Tuesday LSN, calculate the delta, and store it as Tuesday's
+delta snapshot".
+
+Clearly this all requires careful housekeeping to retain the relationship between branches that depend on
+each other: perhaps this would be done by making the archive branches have child/parent relationships with
+each other, or perhaps we would permit them to remain children of their original parent, but additionally
+have a relationship with the snapshot they're encoded relative to.
+
+Activating a branch that is diff-encoded may require activating several earlier branches too, so figuring
+out how frequently to write a full copy is important.  This is essentially a zoomed-out version of what
+we do with delta layers and image layers within a timeline, except each "layer" is a whole timeline.
+
+
+## FAQ/Alternatives
+
+### Store all timelines in the tenant manifest
+
+Rather than special-casing offloaded timelines in the offload manifest, we could store a total
+manifest of all timelines, eliminating the need for the pageserver to list timelines in S3 on
+startup.
+
+That would be a more invasive change (require hooking in to timeline creation), and would
+generate much more I/O to this manifest for tenants that had many branches _and_ frequent
+create/delete cycles for short lived branches.  Restricting the manifest to offloaded timelines
+means that we only have to cope with the rate at which long-lived timelines are archived, rather
+than the rate at which sort lived timelines are created & destroyed.
+
+### Automatically archiving/activating timelines without external API calls
+
+We could implement TTL driven offload of timelines, waking them up when a page request
+arrives.
+
+This has downsides:
+- Opacity: if we do TTL-driven offload inside the pageserver, then the end user doesn't
+  know which of their branches are in this state, and might get a surprise when they try
+  to use such a branch.
+- Price fluctuation: if the archival of a branch is used in end user pricing, then users
+  prefer clarity & consistency.  Ideally a branch's storage should cost the same from the moment it
+  is created, rather than having a usage-dependency storage price.
+- Complexity: enabling the page service to call up into the Tenant to activate a timeline
+  would be awkward, compared with an external entry point.
+
+### Make offloaded a state of Timeline
+
+To reduce the operator-facing complexity of having some timelines APIs that only return
+non-offloaded timelines, we could build the offloaded state into the Timeline type.
+
+`timeline.rs` is already one of the most egregiously long source files in the tree, so
+this is rejected on the basis that we need to avoid making that complexity worse.

docs/rfcs/036-physical-replication.md

@@ -0,0 +1,265 @@
+# Physical Replication
+
+This RFC is a bit special in that we have already implemented physical
+replication a long time ago. However, we never properly wrote down all
+the decisions and assumptions, and in the last months when more users
+have started to use the feature, numerous issues have surfaced.
+
+This RFC documents the design decisions that have been made.
+
+## Summary
+
+PostgreSQL has a feature called streaming replication, where a replica
+streams WAL from the primary and continuously applies it. It is also
+known as "physical replication", to distinguish it from logical
+replication.  In PostgreSQL, a replica is initialized by taking a
+physical backup of the primary. In Neon, the replica is initialized
+from a slim "base backup" from the pageserver, just like a primary,
+and the primary and the replicas connect to the same pageserver,
+sharing the storage.
+
+There are two kinds of read-only replicas in Neon:
+- replicas that follow the primary, and
+- "static" replicas that are pinned at a particular LSN.
+
+A static replica is useful e.g. for performing time-travel queries and
+running one-off slow queries without affecting the primary. A replica
+that follows the primary can be used e.g. to scale out read-only
+workloads.
+
+## Motivation
+
+Read-only replicas allow offloading read-only queries. It's useful for
+isolation, if you want to make sure that read-only queries don't
+affect the primary, and it's also an easy way to provide guaranteed
+read-only access to an application, without having to mess with access
+controls.
+
+## Non Goals (if relevant)
+
+This RFC is all about WAL-based *physical* replication. Logical
+replication is a different feature.
+
+Neon also has the capability to launch "static" read-only nodes which
+do not follow the primary, but are pinned to a particular LSN. They
+can be used for long-running one-off queries, or for Point-in-time
+queries. They work similarly to read replicas that follow the primary,
+but some things are simpler: there are no concerns about cache
+invalidation when the data changes on the primary, or worrying about
+transactions that are in-progress on the primary.
+
+## Impacted components (e.g. pageserver, safekeeper, console, etc)
+
+- Control plane launches the replica
+- Replica Postgres instance connects to the safekeepers, to stream the WAL
+- The primary does not know about the standby, except for the hot standby feedback
+- The primary and replicas all connect to the same pageservers
+
+
+# Context
+
+Some useful things to know about hot standby and replicas in
+PostgreSQL.
+
+## PostgreSQL startup sequence
+
+"Running" and "start up" terms are little imprecise. PostgreSQL
+replica startup goes through several stages:
+
+1. First, the process is started up, and various initialization steps
+   are performed, like initializing shared memory. If you try to
+   connect to the server in this stage, you get an error: ERROR: the
+   database system is starting up. This stage happens very quickly, no
+
+2. Then the server reads the checpoint record from the WAL and starts
+   the WAL replay starting from the checkpoint. This works differently
+   in Neon: we start the WAL replay at the basebackup LSN, not from a
+   checkpoint! If you connect to the server in this state, you get an
+   error: ERROR: the database system is not yet accepting
+   connections. We proceed to the next stage, when the WAL replay sees
+   a running-xacts record. Or in Neon, the "CLOG scanning" mechanism
+   can allow us to move directly to next stage, with all the caveats
+   listed in this RFC.
+
+3. When the running-xacts information is established, the server
+   starts to accept connections normally.
+
+From PostgreSQL's point of view, the server is already running in
+stage 2, even though it's not accepting connections yet. Our
+`compute_ctl` does not consider it as running until stage 3. If the
+transition from stage 2 to 3 doesn't happen fast enough, the control
+plane will mark the start operation as failed.
+
+
+## Decisions, Issues
+
+### Cache invalidation in replica
+
+When a read replica follows the primary in PostgreSQL, it needs to
+stream all the WAL from the primary and apply all the records, to keep
+the local copy of the data consistent with the primary. In Neon, the
+replica can fetch the updated page versions from the pageserver, so
+it's not necessary to apply all the WAL. However, it needs to ensure
+that any pages that are currently in the Postgres buffer cache, or the
+Local File Cache, are either updated, or thrown away so that the next
+read of the page will fetch the latest version.
+
+We choose to apply the WAL records for pages that are already in the
+buffer cache, and skip records for other pages. Somewhat arbitrarily,
+we also apply records affecting catalog relations, fetching the old
+page version from the pageserver if necessary first. See
+`neon_redo_read_buffer_filter()` function.
+
+The replica wouldn't necessarily need to see all the WAL records, only
+the records that apply to cached pages. For simplicity, we do stream
+all the WAL to the replica, and the replica simply ignores WAL records
+that require no action.
+
+Like in PostgreSQL, the read replica maintains a "replay LSN", which
+is the LSN up to which the replica has received and replayed the
+WAL. The replica can lag behind the primary, if it cannot quite keep
+up with the primary, or if a long-running query conflicts with changes
+that are about to be applied, or even intentionally if the user wishes
+to see delayed data (see recovery_min_apply_delay). It's important
+that the replica sees a consistent view of the whole cluster at the
+replay LSN, when it's lagging behind.
+
+In Neon, the replica connects to a safekeeper to get the WAL
+stream. That means that the safekeepers must be able to regurgitate
+the original WAL as far back as the replay LSN of any running read
+replica. (A static read-only node that does not follow the primary
+does not require a WAL stream however). The primary does not need to
+be running, and when it is, the replicas don't incur any extra
+overhead to the primary (see hot standby feedback though).
+
+### In-progress transactions
+
+In PostgreSQL, when a hot standby server starts up, it cannot
+immediately open up for queries (see [PostgreSQL startup
+sequence]). It first needs to establish a complete list of in-progress
+transactions, including subtransactions, that are running at the
+primary, at the current replay LSN. Normally that happens quickly,
+when the replica sees a "running-xacts" WAL record, because the
+primary writes a running-xacts WAL record at every checkpoint, and in
+PostgreSQL the replica always starts the WAL replay from a checkpoint
+REDO point. (A shutdown checkpoint WAL record also implies that all
+the non-prepared transactions have ended.) If there are a lot of
+subtransactions in progress, however, the standby might need to wait
+for old transactions to complete before it can open up for queries.
+
+In Neon that problem is worse: a replica can start at any LSN, so
+there's no guarantee that it will see a running-xacts record any time
+soon. In particular, if the primary is not running when the replica is
+started, it might never see a running-xacts record.
+
+To make things worse, we initially missed this issue, and always
+started accepting queries at replica startup, even if it didn't have
+the transaction information. That could lead to incorrect query
+results and data corruption later. However, as we fixed that, we
+introduced a new problem compared to what we had before: previously
+the replica would always start up, but after fixing that bug, it might
+not. In a superficial way, the old behavior was better (but could lead
+to serious issues later!). That made fixing that bug was very hard,
+because as we fixed it, we made things (superficially) worse for
+others.
+
+See https://github.com/neondatabase/neon/pull/7288 which fixed the
+bug, and follow-up PRs https://github.com/neondatabase/neon/pull/8323
+and https://github.com/neondatabase/neon/pull/8484 to try to claw back
+the cases that started to cause trouble as fixing it. As of this
+writing, there are still cases where a replica might not immediately
+start up, causing the control plane operation to fail, the remaining
+issues are tracked in https://github.com/neondatabase/neon/issues/6211.
+
+One long-term fix for this is to switch to using so-called CSN
+snapshots in read replica. That would make it unnecessary to have the
+full in-progress transaction list in the replica at startup time. See
+https://commitfest.postgresql.org/48/4912/ for a work-in-progress
+patch to upstream to implement that.
+
+Another thing we could do is to teach the control plane about that
+distinction between "starting up" and "running but haven't received
+running-xacts information yet", so that we could keep the replica
+waiting longer in that stage, and also give any client connections the
+same `ERROR: the database system is not yet accepting connections`
+error that you get in standalone PostgreSQL in that state.
+
+
+### Recovery conflicts and Hot standby feedback
+
+It's possible that a tuple version is vacuumed away in the primary,
+even though it is still needed by a running transactions in the
+replica. This is called a "recovery conflict", and PostgreSQL provides
+various options for dealing with it. By default, the WAL replay will
+wait up to 30 s for the conflicting query to finish. After that, it
+will kill the running query, so that the WAL replay can proceed.
+
+Another way to avoid the situation is to enable the
+[`hot_standby_feedback`](https://www.postgresql.org/docs/current/runtime-config-replication.html#GUC-HOT-STANDBY-FEEDBACK)
+option. When it is enabled, the primary will refrain from vacuuming
+tuples that are still needed in the primary. That means potentially
+bloating the primary, which violates the usual rule that read replicas
+don't affect the operations on the primary, which is why it's off by
+default. We leave it to users to decide if they want to turn it on,
+same as PostgreSQL.
+
+Neon supports `hot_standby_feedback` by passing the feedback messages
+from the replica to the safekeepers, and from safekeepers to the
+primary.
+
+### Relationship of settings between primary and replica
+
+In order to enter hot standby mode, some configuration options need to
+be set to the same or larger values in the standby, compared to the
+primary.  See [explanation in the PostgreSQL
+docs](https://www.postgresql.org/docs/current/hot-standby.html#HOT-STANDBY-ADMIN)
+
+In Neon, we have this problem too. To prevent customers from hitting
+it, the control plane automatically adjusts the settings of a replica,
+so that they match or exceed the primary's settings (see
+https://github.com/neondatabase/cloud/issues/14903). However, you
+can still hit the issue if the primary is restarted with larger
+settings, while the replica is running.
+
+
+### Interaction with Pageserver GC
+
+The read replica can lag behind the primary. If there are recovery
+conflicts or the replica cannot keep up for some reason, the lag can
+in principle grow indefinitely. The replica will issue all GetPage
+requests to the pageservers at the current replay LSN, and needs to
+see the old page versions.
+
+If the retention period in the pageserver is set to be small, it may
+have already garbage collected away the old page versions. That will
+cause read errors in the compute, and can mean that the replica cannot
+make progress with the replication anymore.
+
+There is a mechanism for replica to pass information about its replay
+LSN to the pageserver, so that the pageserver refrains from GC'ing
+data that is still needed by the standby. It's called
+'standby_horizon' in the pageserver code, see
+https://github.com/neondatabase/neon/pull/7368. A separate "lease"
+mechanism also is in the works, where the replica could hold a lease
+on the old LSN, preventing the pageserver from advancing the GC
+horizon past that point. The difference is that the standby_horizon
+mechanism relies on a feedback message from replica to safekeeper,
+while the least API is exposed directly from the pageserver. A static
+read-only node is not connected to safekeepers, so it cannot use the
+standby_horizon mechanism.
+
+
+### Synchronous replication
+
+We haven't put any effort into synchronous replication yet.
+
+PostgreSQL provides multiple levels of synchronicity. In the weaker
+levels, a transaction is not acknowledged as committed to the client
+in the primary until the WAL has been streamed to a replica or flushed
+to disk there. Those modes don't make senses in Neon, because the
+safekeepers handle durability.
+
+`synchronous_commit=remote_apply` mode would make sense. In that mode,
+the commit is not acknowledged to the client until it has been
+replayed in the replica. That ensures that after commit, you can see
+the commit in the replica too (aka. read-your-write consistency).

docs/settings.md

@@ -134,7 +134,7 @@ depends on that, so if you change it, bad things will happen.
 
 #### page_cache_size
 
-Size of the page cache, to hold materialized page versions. Unit is
+Size of the page cache. Unit is
 number of 8 kB blocks. The default is 8192, which means 64 MB.
 
 #### max_file_descriptors

docs/storage_controller.md

@@ -44,7 +44,7 @@ If you need to modify the database schema, here’s how to create a migration:
 - Use `diesel migration generate <name>` to create a new migration
 - Populate the SQL files in the `migrations/` subdirectory
 - Use `DATABASE_URL=... diesel migration run` to apply the migration you just wrote: this will update the `[schema.rs](http://schema.rs)` file automatically.
-  - This requires a running database: the easiest way to do that is to just run `cargo neon init ; cargo neon start`, which will leave a database available at `postgresql://localhost:1235/attachment_service`
+  - This requires a running database: the easiest way to do that is to just run `cargo neon init ; cargo neon start`, which will leave a database available at `postgresql://localhost:1235/storage_controller`
 - Commit the migration files and the changes to schema.rs
 - If you need to iterate, you can rewind migrations with `diesel migration revert -a` and then `diesel migration run` again.
 - The migrations are build into the storage controller binary, and automatically run at startup after it is deployed, so once you’ve committed a migration no further steps are needed.

docs/synthetic-size.md

@@ -21,9 +21,9 @@ implementation where we keep more data than we would need to, do not
 change the synthetic size or incur any costs to the user.
 
 The synthetic size is calculated for the whole project. It is not
-straightforward to attribute size to individual branches. See "What is
-the size of an individual branch?" for discussion on those
-difficulties.
+straightforward to attribute size to individual branches. See [What is
+the size of an individual branch?](#what-is-the-size-of-an-individual-branch)
+for a discussion of those difficulties.
 
 The synthetic size is designed to:
 
@@ -40,8 +40,9 @@ The synthetic size is designed to:
 - logical size is the size of a branch *at a given point in
   time*. It's the total size of all tables in all databases, as you
   see with "\l+" in psql for example, plus the Postgres SLRUs and some
-  small amount of metadata. NOTE that currently, Neon does not include
-  the SLRUs and metadata in the logical size. See comment to `get_current_logical_size_non_incremental()`.
+  small amount of metadata. Note that currently, Neon does not include
+  the SLRUs and metadata in the logical size. Refer to the comment in
+  [`get_current_logical_size_non_incremental()`](/pageserver/src/pgdatadir_mapping.rs#L813-L814).
 
 - a "point in time" is defined as an LSN value. You can convert a
   timestamp to an LSN, but the storage internally works with LSNs.

libs/compute_api/src/spec.rs

@@ -96,12 +96,6 @@ pub struct ComputeSpec {
     // Stripe size for pageserver sharding, in pages
     #[serde(default)]
     pub shard_stripe_size: Option<usize>,
-
-    // When we are starting a new replica in hot standby mode,
-    // we need to know if the primary is running.
-    // This is used to determine if replica should wait for
-    // RUNNING_XACTS from primary or not.
-    pub primary_is_running: Option<bool>,
 }
 
 /// Feature flag to signal `compute_ctl` to enable certain experimental functionality.

libs/metrics/src/hll.rs

@@ -13,11 +13,7 @@ use std::{
 
 use measured::{
     label::{LabelGroupVisitor, LabelName, LabelValue, LabelVisitor},
-    metric::{
-        group::{Encoding, MetricValue},
-        name::MetricNameEncoder,
-        Metric, MetricType, MetricVec,
-    },
+    metric::{counter::CounterState, name::MetricNameEncoder, Metric, MetricType, MetricVec},
     text::TextEncoder,
     LabelGroup,
 };
@@ -144,6 +140,7 @@ impl<const N: usize> HyperLogLogState<N> {
         })
     }
 }
+
 impl<W: std::io::Write, const N: usize> measured::metric::MetricEncoding<TextEncoder<W>>
     for HyperLogLogState<N>
 {
@@ -182,12 +179,13 @@ impl<W: std::io::Write, const N: usize> measured::metric::MetricEncoding<TextEnc
             .into_iter()
             .enumerate()
             .try_for_each(|(hll_shard, val)| {
-                enc.write_metric_value(
-                    name.by_ref(),
+                CounterState::new(val as u64).collect_into(
+                    &(),
                     labels.by_ref().compose_with(HllShardLabel {
                         hll_shard: hll_shard as i64,
                     }),
-                    MetricValue::Int(val as i64),
+                    name.by_ref(),
+                    enc,
                 )
             })
     }

libs/metrics/src/lib.rs

@@ -9,7 +9,7 @@ use measured::{
     metric::{
         counter::CounterState,
         gauge::GaugeState,
-        group::{Encoding, MetricValue},
+        group::Encoding,
         name::{MetricName, MetricNameEncoder},
         MetricEncoding, MetricFamilyEncoding,
     },
@@ -103,9 +103,10 @@ static MAXRSS_KB: Lazy<IntGauge> = Lazy::new(|| {
     .expect("Failed to register maxrss_kb int gauge")
 });
 
-pub const DISK_WRITE_SECONDS_BUCKETS: &[f64] = &[
-    0.000_050, 0.000_100, 0.000_500, 0.001, 0.003, 0.005, 0.01, 0.05, 0.1, 0.3, 0.5,
-];
+/// Most common fsync latency is 50 µs - 100 µs, but it can be much higher,
+/// especially during many concurrent disk operations.
+pub const DISK_FSYNC_SECONDS_BUCKETS: &[f64] =
+    &[0.001, 0.005, 0.01, 0.05, 0.1, 0.5, 1.0, 5.0, 10.0, 30.0];
 
 pub struct BuildInfo {
     pub revision: &'static str,
@@ -170,8 +171,11 @@ fn write_gauge<Enc: Encoding>(
     labels: impl LabelGroup,
     name: impl MetricNameEncoder,
     enc: &mut Enc,
-) -> Result<(), Enc::Err> {
-    enc.write_metric_value(name, labels, MetricValue::Int(x))
+) -> Result<(), Enc::Err>
+where
+    GaugeState: MetricEncoding<Enc>,
+{
+    GaugeState::new(x).collect_into(&(), labels, name, enc)
 }
 
 #[derive(Default)]
@@ -543,15 +547,6 @@ impl<T: Encoding> Encoding for Inc<T> {
     fn write_help(&mut self, name: impl MetricNameEncoder, help: &str) -> Result<(), Self::Err> {
         self.0.write_help(name, help)
     }
-
-    fn write_metric_value(
-        &mut self,
-        name: impl MetricNameEncoder,
-        labels: impl LabelGroup,
-        value: MetricValue,
-    ) -> Result<(), Self::Err> {
-        self.0.write_metric_value(name, labels, value)
-    }
 }
 
 impl<T: Encoding> MetricEncoding<Inc<T>> for MeasuredCounterPairState
@@ -578,15 +573,6 @@ impl<T: Encoding> Encoding for Dec<T> {
     fn write_help(&mut self, name: impl MetricNameEncoder, help: &str) -> Result<(), Self::Err> {
         self.0.write_help(name, help)
     }
-
-    fn write_metric_value(
-        &mut self,
-        name: impl MetricNameEncoder,
-        labels: impl LabelGroup,
-        value: MetricValue,
-    ) -> Result<(), Self::Err> {
-        self.0.write_metric_value(name, labels, value)
-    }
 }
 
 /// Write the dec counter to the encoder

libs/pageserver_api/src/controller_api.rs

@@ -1,4 +1,6 @@
+use std::collections::HashSet;
 use std::str::FromStr;
+use std::time::{Duration, Instant};
 
 /// Request/response types for the storage controller
 /// API (`/control/v1` prefix).  Implemented by the server
@@ -11,6 +13,27 @@ use crate::{
     shard::{ShardStripeSize, TenantShardId},
 };
 
+#[derive(Serialize, Deserialize, Debug)]
+#[serde(deny_unknown_fields)]
+pub struct TenantCreateRequest {
+    pub new_tenant_id: TenantShardId,
+    #[serde(default)]
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub generation: Option<u32>,
+
+    // If omitted, create a single shard with TenantShardId::unsharded()
+    #[serde(default)]
+    #[serde(skip_serializing_if = "ShardParameters::is_unsharded")]
+    pub shard_parameters: ShardParameters,
+
+    #[serde(default)]
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub placement_policy: Option<PlacementPolicy>,
+
+    #[serde(flatten)]
+    pub config: TenantConfig, // as we have a flattened field, we should reject all unknown fields in it
+}
+
 #[derive(Serialize, Deserialize)]
 pub struct TenantCreateResponseShard {
     pub shard_id: TenantShardId,
@@ -66,7 +89,7 @@ pub struct TenantLocateResponse {
     pub shard_params: ShardParameters,
 }
 
-#[derive(Serialize, Deserialize)]
+#[derive(Serialize, Deserialize, Debug)]
 pub struct TenantDescribeResponse {
     pub tenant_id: TenantId,
     pub shards: Vec<TenantDescribeResponseShard>,
@@ -89,7 +112,7 @@ pub struct NodeDescribeResponse {
     pub listen_pg_port: u16,
 }
 
-#[derive(Serialize, Deserialize)]
+#[derive(Serialize, Deserialize, Debug)]
 pub struct TenantDescribeResponseShard {
     pub tenant_shard_id: TenantShardId,
 
@@ -129,11 +152,16 @@ impl UtilizationScore {
     }
 }
 
-#[derive(Serialize, Deserialize, Clone, Copy, Debug)]
+#[derive(Serialize, Clone, Copy, Debug)]
 #[serde(into = "NodeAvailabilityWrapper")]
 pub enum NodeAvailability {
     // Normal, happy state
     Active(UtilizationScore),
+    // Node is warming up, but we expect it to become available soon. Covers
+    // the time span between the re-attach response being composed on the storage controller
+    // and the first successful heartbeat after the processing of the re-attach response
+    // finishes on the pageserver.
+    WarmingUp(Instant),
     // Offline: Tenants shouldn't try to attach here, but they may assume that their
     // secondary locations on this node still exist.  Newly added nodes are in this
     // state until we successfully contact them.
@@ -143,7 +171,10 @@ pub enum NodeAvailability {
 impl PartialEq for NodeAvailability {
     fn eq(&self, other: &Self) -> bool {
         use NodeAvailability::*;
-        matches!((self, other), (Active(_), Active(_)) | (Offline, Offline))
+        matches!(
+            (self, other),
+            (Active(_), Active(_)) | (Offline, Offline) | (WarmingUp(_), WarmingUp(_))
+        )
     }
 }
 
@@ -155,6 +186,7 @@ impl Eq for NodeAvailability {}
 #[derive(Serialize, Deserialize, Clone, Copy, Debug)]
 pub enum NodeAvailabilityWrapper {
     Active,
+    WarmingUp,
     Offline,
 }
 
@@ -164,6 +196,7 @@ impl From<NodeAvailabilityWrapper> for NodeAvailability {
             // Assume the worst utilisation score to begin with. It will later be updated by
             // the heartbeats.
             NodeAvailabilityWrapper::Active => NodeAvailability::Active(UtilizationScore::worst()),
+            NodeAvailabilityWrapper::WarmingUp => NodeAvailability::WarmingUp(Instant::now()),
             NodeAvailabilityWrapper::Offline => NodeAvailability::Offline,
         }
     }
@@ -173,6 +206,7 @@ impl From<NodeAvailability> for NodeAvailabilityWrapper {
     fn from(val: NodeAvailability) -> Self {
         match val {
             NodeAvailability::Active(_) => NodeAvailabilityWrapper::Active,
+            NodeAvailability::WarmingUp(_) => NodeAvailabilityWrapper::WarmingUp,
             NodeAvailability::Offline => NodeAvailabilityWrapper::Offline,
         }
     }
@@ -209,6 +243,7 @@ pub enum NodeSchedulingPolicy {
     Active,
     Filling,
     Pause,
+    PauseForRestart,
     Draining,
 }
 
@@ -220,6 +255,7 @@ impl FromStr for NodeSchedulingPolicy {
             "active" => Ok(Self::Active),
             "filling" => Ok(Self::Filling),
             "pause" => Ok(Self::Pause),
+            "pause_for_restart" => Ok(Self::PauseForRestart),
             "draining" => Ok(Self::Draining),
             _ => Err(anyhow::anyhow!("Unknown scheduling state '{s}'")),
         }
@@ -233,6 +269,7 @@ impl From<NodeSchedulingPolicy> for String {
             Active => "active",
             Filling => "filling",
             Pause => "pause",
+            PauseForRestart => "pause_for_restart",
             Draining => "draining",
         }
         .to_string()
@@ -258,6 +295,39 @@ pub enum PlacementPolicy {
 #[derive(Serialize, Deserialize, Debug)]
 pub struct TenantShardMigrateResponse {}
 
+/// Metadata health record posted from scrubber.
+#[derive(Serialize, Deserialize, Debug)]
+pub struct MetadataHealthRecord {
+    pub tenant_shard_id: TenantShardId,
+    pub healthy: bool,
+    pub last_scrubbed_at: chrono::DateTime<chrono::Utc>,
+}
+
+#[derive(Serialize, Deserialize, Debug)]
+pub struct MetadataHealthUpdateRequest {
+    pub healthy_tenant_shards: HashSet<TenantShardId>,
+    pub unhealthy_tenant_shards: HashSet<TenantShardId>,
+}
+
+#[derive(Serialize, Deserialize, Debug)]
+pub struct MetadataHealthUpdateResponse {}
+
+#[derive(Serialize, Deserialize, Debug)]
+pub struct MetadataHealthListUnhealthyResponse {
+    pub unhealthy_tenant_shards: Vec<TenantShardId>,
+}
+
+#[derive(Serialize, Deserialize, Debug)]
+pub struct MetadataHealthListOutdatedRequest {
+    #[serde(with = "humantime_serde")]
+    pub not_scrubbed_for: Duration,
+}
+
+#[derive(Serialize, Deserialize, Debug)]
+pub struct MetadataHealthListOutdatedResponse {
+    pub health_records: Vec<MetadataHealthRecord>,
+}
+
 #[cfg(test)]
 mod test {
     use super::*;
@@ -277,4 +347,19 @@ mod test {
         assert_eq!(serde_json::from_str::<PlacementPolicy>(&encoded)?, v);
         Ok(())
     }
+
+    #[test]
+    fn test_reject_unknown_field() {
+        let id = TenantId::generate();
+        let create_request = serde_json::json!({
+            "new_tenant_id": id.to_string(),
+            "unknown_field": "unknown_value".to_string(),
+        });
+        let err = serde_json::from_value::<TenantCreateRequest>(create_request).unwrap_err();
+        assert!(
+            err.to_string().contains("unknown field `unknown_field`"),
+            "expect unknown field `unknown_field` error, got: {}",
+            err
+        );
+    }
 }

libs/pageserver_api/src/key.rs

@@ -22,6 +22,11 @@ pub struct Key {
     pub field6: u32,
 }
 
+/// When working with large numbers of Keys in-memory, it is more efficient to handle them as i128 than as
+/// a struct of fields.
+#[derive(Clone, Copy, Hash, PartialEq, Eq, Ord, PartialOrd)]
+pub struct CompactKey(i128);
+
 /// The storage key size.
 pub const KEY_SIZE: usize = 18;
 
@@ -29,7 +34,7 @@ pub const KEY_SIZE: usize = 18;
 /// See [`Key::to_i128`] for more information on the encoding.
 pub const METADATA_KEY_SIZE: usize = 16;
 
-/// The key prefix start range for the metadata keys. All keys with the first byte >= 0x40 is a metadata key.
+/// The key prefix start range for the metadata keys. All keys with the first byte >= 0x60 is a metadata key.
 pub const METADATA_KEY_BEGIN_PREFIX: u8 = 0x60;
 pub const METADATA_KEY_END_PREFIX: u8 = 0x7F;
 
@@ -107,7 +112,10 @@ impl Key {
     /// As long as Neon does not support tablespace (because of lack of access to local file system),
     /// we can assume that only some predefined namespace OIDs are used which can fit in u16
     pub fn to_i128(&self) -> i128 {
-        assert!(self.field2 <= 0xFFFF || self.field2 == 0xFFFFFFFF || self.field2 == 0x22222222);
+        assert!(
+            self.field2 <= 0xFFFF || self.field2 == 0xFFFFFFFF || self.field2 == 0x22222222,
+            "invalid key: {self}",
+        );
         (((self.field1 & 0x7F) as i128) << 120)
             | (((self.field2 & 0xFFFF) as i128) << 104)
             | ((self.field3 as i128) << 72)
@@ -127,6 +135,14 @@ impl Key {
         }
     }
 
+    pub fn to_compact(&self) -> CompactKey {
+        CompactKey(self.to_i128())
+    }
+
+    pub fn from_compact(k: CompactKey) -> Self {
+        Self::from_i128(k.0)
+    }
+
     pub const fn next(&self) -> Key {
         self.add(1)
     }
@@ -160,8 +176,9 @@ impl Key {
         key
     }
 
-    /// Convert a 18B slice to a key. This function should not be used for metadata keys because field2 is handled differently.
-    /// Use [`Key::from_i128`] instead if you want to handle 16B keys (i.e., metadata keys).
+    /// Convert a 18B slice to a key. This function should not be used for 16B metadata keys because `field2` is handled differently.
+    /// Use [`Key::from_i128`] instead if you want to handle 16B keys (i.e., metadata keys). There are some restrictions on `field2`,
+    /// and therefore not all 18B slices are valid page server keys.
     pub fn from_slice(b: &[u8]) -> Self {
         Key {
             field1: b[0],
@@ -173,7 +190,7 @@ impl Key {
         }
     }
 
-    /// Convert a key to a 18B slice. This function should not be used for metadata keys because field2 is handled differently.
+    /// Convert a key to a 18B slice. This function should not be used for getting a 16B metadata key because `field2` is handled differently.
     /// Use [`Key::to_i128`] instead if you want to get a 16B key (i.e., metadata keys).
     pub fn write_to_byte_slice(&self, buf: &mut [u8]) {
         buf[0] = self.field1;
@@ -195,6 +212,13 @@ impl fmt::Display for Key {
     }
 }
 
+impl fmt::Display for CompactKey {
+    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
+        let k = Key::from_compact(*self);
+        k.fmt(f)
+    }
+}
+
 impl Key {
     pub const MIN: Key = Key {
         field1: u8::MIN,

libs/pageserver_api/src/keyspace.rs

@@ -17,6 +17,16 @@ pub struct KeySpace {
     pub ranges: Vec<Range<Key>>,
 }
 
+impl std::fmt::Display for KeySpace {
+    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
+        write!(f, "[")?;
+        for range in &self.ranges {
+            write!(f, "{}..{},", range.start, range.end)?;
+        }
+        write!(f, "]")
+    }
+}
+
 /// A wrapper type for sparse keyspaces.
 #[derive(Clone, Debug, Default, PartialEq, Eq)]
 pub struct SparseKeySpace(pub KeySpace);

libs/pageserver_api/src/models.rs

@@ -5,10 +5,10 @@ pub mod utilization;
 pub use utilization::PageserverUtilization;
 
 use std::{
-    borrow::Cow,
     collections::HashMap,
     io::{BufRead, Read},
     num::{NonZeroU64, NonZeroUsize},
+    str::FromStr,
     sync::atomic::AtomicUsize,
     time::{Duration, SystemTime},
 };
@@ -19,13 +19,11 @@ use serde::{Deserialize, Serialize};
 use serde_with::serde_as;
 use utils::{
     completion,
-    history_buffer::HistoryBufferWithDropCounter,
     id::{NodeId, TenantId, TimelineId},
     lsn::Lsn,
     serde_system_time,
 };
 
-use crate::controller_api::PlacementPolicy;
 use crate::{
     reltag::RelTag,
     shard::{ShardCount, ShardStripeSize, TenantShardId},
@@ -229,6 +227,11 @@ pub struct TimelineCreateRequest {
     pub pg_version: Option<u32>,
 }
 
+#[derive(Serialize, Deserialize, Clone)]
+pub struct LsnLeaseRequest {
+    pub lsn: Lsn,
+}
+
 #[derive(Serialize, Deserialize)]
 pub struct TenantShardSplitRequest {
     pub new_shard_count: u8,
@@ -271,44 +274,6 @@ impl Default for ShardParameters {
     }
 }
 
-#[derive(Serialize, Deserialize, Debug)]
-#[serde(deny_unknown_fields)]
-pub struct TenantCreateRequest {
-    pub new_tenant_id: TenantShardId,
-    #[serde(default)]
-    #[serde(skip_serializing_if = "Option::is_none")]
-    pub generation: Option<u32>,
-
-    // If omitted, create a single shard with TenantShardId::unsharded()
-    #[serde(default)]
-    #[serde(skip_serializing_if = "ShardParameters::is_unsharded")]
-    pub shard_parameters: ShardParameters,
-
-    // This parameter is only meaningful in requests sent to the storage controller
-    #[serde(default)]
-    #[serde(skip_serializing_if = "Option::is_none")]
-    pub placement_policy: Option<PlacementPolicy>,
-
-    #[serde(flatten)]
-    pub config: TenantConfig, // as we have a flattened field, we should reject all unknown fields in it
-}
-
-#[derive(Deserialize, Debug)]
-#[serde(deny_unknown_fields)]
-pub struct TenantLoadRequest {
-    #[serde(default)]
-    #[serde(skip_serializing_if = "Option::is_none")]
-    pub generation: Option<u32>,
-}
-
-impl std::ops::Deref for TenantCreateRequest {
-    type Target = TenantConfig;
-
-    fn deref(&self) -> &Self::Target {
-        &self.config
-    }
-}
-
 /// An alternative representation of `pageserver::tenant::TenantConf` with
 /// simpler types.
 #[derive(Serialize, Deserialize, Debug, Default, Clone, Eq, PartialEq)]
@@ -327,7 +292,6 @@ pub struct TenantConfig {
     pub walreceiver_connect_timeout: Option<String>,
     pub lagging_wal_timeout: Option<String>,
     pub max_lsn_wal_lag: Option<NonZeroU64>,
-    pub trace_read_requests: Option<bool>,
     pub eviction_policy: Option<EvictionPolicy>,
     pub min_resident_size_override: Option<u64>,
     pub evictions_low_residence_duration_metric_threshold: Option<String>,
@@ -471,6 +435,41 @@ pub enum CompactionAlgorithm {
     Tiered,
 }
 
+#[derive(Debug, Clone, Copy, PartialEq, Eq, Serialize, Deserialize)]
+pub enum ImageCompressionAlgorithm {
+    // Disabled for writes, support decompressing during read path
+    Disabled,
+    /// Zstandard compression. Level 0 means and None mean the same (default level). Levels can be negative as well.
+    /// For details, see the [manual](http://facebook.github.io/zstd/zstd_manual.html).
+    Zstd {
+        level: Option<i8>,
+    },
+}
+
+impl FromStr for ImageCompressionAlgorithm {
+    type Err = anyhow::Error;
+    fn from_str(s: &str) -> Result<Self, Self::Err> {
+        let mut components = s.split(['(', ')']);
+        let first = components
+            .next()
+            .ok_or_else(|| anyhow::anyhow!("empty string"))?;
+        match first {
+            "disabled" => Ok(ImageCompressionAlgorithm::Disabled),
+            "zstd" => {
+                let level = if let Some(v) = components.next() {
+                    let v: i8 = v.parse()?;
+                    Some(v)
+                } else {
+                    None
+                };
+
+                Ok(ImageCompressionAlgorithm::Zstd { level })
+            }
+            _ => anyhow::bail!("invalid specifier '{first}'"),
+        }
+    }
+}
+
 #[derive(Eq, PartialEq, Debug, Clone, Serialize, Deserialize)]
 pub struct CompactionAlgorithmSettings {
     pub kind: CompactionAlgorithm,
@@ -563,10 +562,6 @@ pub struct LocationConfigListResponse {
     pub tenant_shards: Vec<(TenantShardId, Option<LocationConfig>)>,
 }
 
-#[derive(Serialize, Deserialize)]
-#[serde(transparent)]
-pub struct TenantCreateResponse(pub TenantId);
-
 #[derive(Serialize)]
 pub struct StatusResponse {
     pub id: NodeId,
@@ -623,31 +618,6 @@ impl TenantConfigRequest {
     }
 }
 
-#[derive(Debug, Deserialize)]
-pub struct TenantAttachRequest {
-    #[serde(default)]
-    pub config: TenantAttachConfig,
-    #[serde(default)]
-    pub generation: Option<u32>,
-}
-
-/// Newtype to enforce deny_unknown_fields on TenantConfig for
-/// its usage inside `TenantAttachRequest`.
-#[derive(Debug, Serialize, Deserialize, Default)]
-#[serde(deny_unknown_fields)]
-pub struct TenantAttachConfig {
-    #[serde(flatten)]
-    allowing_unknown_fields: TenantConfig,
-}
-
-impl std::ops::Deref for TenantAttachConfig {
-    type Target = TenantConfig;
-
-    fn deref(&self) -> &Self::Target {
-        &self.allowing_unknown_fields
-    }
-}
-
 /// See [`TenantState::attachment_status`] and the OpenAPI docs for context.
 #[derive(Serialize, Deserialize, Clone)]
 #[serde(tag = "slug", content = "data", rename_all = "snake_case")]
@@ -666,8 +636,14 @@ pub struct TenantInfo {
     /// If a layer is present in both local FS and S3, it counts only once.
     pub current_physical_size: Option<u64>, // physical size is only included in `tenant_status` endpoint
     pub attachment_status: TenantAttachmentStatus,
+    pub generation: u32,
+
+    /// Opaque explanation if gc is being blocked.
+    ///
+    /// Only looked up for the individual tenant detail, not the listing. This is purely for
+    /// debugging, not included in openapi.
     #[serde(skip_serializing_if = "Option::is_none")]
-    pub generation: Option<u32>,
+    pub gc_blocking: Option<String>,
 }
 
 #[derive(Serialize, Deserialize, Clone)]
@@ -680,6 +656,17 @@ pub struct TenantDetails {
     pub timelines: Vec<TimelineId>,
 }
 
+#[derive(Serialize, Deserialize, PartialEq, Eq, Clone, Copy, Debug)]
+pub enum TimelineArchivalState {
+    Archived,
+    Unarchived,
+}
+
+#[derive(Serialize, Deserialize, PartialEq, Eq, Clone)]
+pub struct TimelineArchivalConfigRequest {
+    pub state: TimelineArchivalState,
+}
+
 /// This represents the output of the "timeline_detail" and "timeline_list" API calls.
 #[derive(Debug, Serialize, Deserialize, Clone)]
 pub struct TimelineInfo {
@@ -712,6 +699,16 @@ pub struct TimelineInfo {
     pub current_physical_size: Option<u64>, // is None when timeline is Unloaded
     pub current_logical_size_non_incremental: Option<u64>,
 
+    /// How many bytes of WAL are within this branch's pitr_interval.  If the pitr_interval goes
+    /// beyond the branch's branch point, we only count up to the branch point.
+    pub pitr_history_size: u64,
+
+    /// Whether this branch's branch point is within its ancestor's PITR interval (i.e. any
+    /// ancestor data used by this branch would have been retained anyway).  If this is false, then
+    /// this branch may be imposing a cost on the ancestor by causing it to retain layers that it would
+    /// otherwise be able to GC.
+    pub within_ancestor_pitr: bool,
+
     pub timeline_dir_layer_file_size_sum: Option<u64>,
 
     pub wal_source_connstr: Option<String>,
@@ -734,58 +731,7 @@ pub struct LayerMapInfo {
     pub historic_layers: Vec<HistoricLayerInfo>,
 }
 
-#[derive(Debug, Hash, PartialEq, Eq, Clone, Copy, Serialize, Deserialize, enum_map::Enum)]
-#[repr(usize)]
-pub enum LayerAccessKind {
-    GetValueReconstructData,
-    Iter,
-    KeyIter,
-    Dump,
-}
-
-#[derive(Debug, Clone, Serialize, Deserialize)]
-pub struct LayerAccessStatFullDetails {
-    pub when_millis_since_epoch: u64,
-    pub task_kind: Cow<'static, str>,
-    pub access_kind: LayerAccessKind,
-}
-
-/// An event that impacts the layer's residence status.
-#[serde_as]
-#[derive(Debug, Clone, Serialize, Deserialize)]
-pub struct LayerResidenceEvent {
-    /// The time when the event occurred.
-    /// NB: this timestamp is captured while the residence status changes.
-    /// So, it might be behind/ahead of the actual residence change by a short amount of time.
-    ///
-    #[serde(rename = "timestamp_millis_since_epoch")]
-    #[serde_as(as = "serde_with::TimestampMilliSeconds")]
-    pub timestamp: SystemTime,
-    /// The new residence status of the layer.
-    pub status: LayerResidenceStatus,
-    /// The reason why we had to record this event.
-    pub reason: LayerResidenceEventReason,
-}
-
-/// The reason for recording a given [`LayerResidenceEvent`].
-#[derive(Debug, Clone, Copy, Serialize, Deserialize)]
-pub enum LayerResidenceEventReason {
-    /// The layer map is being populated, e.g. during timeline load or attach.
-    /// This includes [`RemoteLayer`] objects created in [`reconcile_with_remote`].
-    /// We need to record such events because there is no persistent storage for the events.
-    ///
-    // https://github.com/rust-lang/rust/issues/74481
-    /// [`RemoteLayer`]: ../../tenant/storage_layer/struct.RemoteLayer.html
-    /// [`reconcile_with_remote`]: ../../tenant/struct.Timeline.html#method.reconcile_with_remote
-    LayerLoad,
-    /// We just created the layer (e.g., freeze_and_flush or compaction).
-    /// Such layers are always [`LayerResidenceStatus::Resident`].
-    LayerCreate,
-    /// We on-demand downloaded or evicted the given layer.
-    ResidenceChange,
-}
-
-/// The residence status of the layer, after the given [`LayerResidenceEvent`].
+/// The residence status of a layer
 #[derive(Debug, Clone, Copy, Serialize, Deserialize)]
 pub enum LayerResidenceStatus {
     /// Residence status for a layer file that exists locally.
@@ -795,23 +741,16 @@ pub enum LayerResidenceStatus {
     Evicted,
 }
 
-impl LayerResidenceEvent {
-    pub fn new(status: LayerResidenceStatus, reason: LayerResidenceEventReason) -> Self {
-        Self {
-            status,
-            reason,
-            timestamp: SystemTime::now(),
-        }
-    }
-}
-
+#[serde_as]
 #[derive(Debug, Clone, Serialize, Deserialize)]
 pub struct LayerAccessStats {
-    pub access_count_by_access_kind: HashMap<LayerAccessKind, u64>,
-    pub task_kind_access_flag: Vec<Cow<'static, str>>,
-    pub first: Option<LayerAccessStatFullDetails>,
-    pub accesses_history: HistoryBufferWithDropCounter<LayerAccessStatFullDetails, 16>,
-    pub residence_events_history: HistoryBufferWithDropCounter<LayerResidenceEvent, 16>,
+    #[serde_as(as = "serde_with::TimestampMilliSeconds")]
+    pub access_time: SystemTime,
+
+    #[serde_as(as = "serde_with::TimestampMilliSeconds")]
+    pub residence_time: SystemTime,
+
+    pub visible: bool,
 }
 
 #[derive(Debug, Clone, Serialize, Deserialize)]
@@ -1008,6 +947,8 @@ pub struct TopTenantShardsResponse {
 }
 
 pub mod virtual_file {
+    use std::path::PathBuf;
+
     #[derive(
         Copy,
         Clone,
@@ -1026,6 +967,53 @@ pub mod virtual_file {
         #[cfg(target_os = "linux")]
         TokioEpollUring,
     }
+
+    /// Direct IO modes for a pageserver.
+    #[derive(Debug, PartialEq, Eq, Clone, serde::Deserialize, serde::Serialize, Default)]
+    #[serde(tag = "mode", rename_all = "kebab-case", deny_unknown_fields)]
+    pub enum DirectIoMode {
+        /// Direct IO disabled (uses usual buffered IO).
+        #[default]
+        Disabled,
+        /// Direct IO disabled (performs checks and perf simulations).
+        Evaluate {
+            /// Alignment check level
+            alignment_check: DirectIoAlignmentCheckLevel,
+            /// Latency padded for performance simulation.
+            latency_padding: DirectIoLatencyPadding,
+        },
+        /// Direct IO enabled.
+        Enabled {
+            /// Actions to perform on alignment error.
+            on_alignment_error: DirectIoOnAlignmentErrorAction,
+        },
+    }
+
+    #[derive(Debug, PartialEq, Eq, Clone, serde::Deserialize, serde::Serialize, Default)]
+    #[serde(rename_all = "kebab-case")]
+    pub enum DirectIoAlignmentCheckLevel {
+        #[default]
+        Error,
+        Log,
+        None,
+    }
+
+    #[derive(Debug, PartialEq, Eq, Clone, serde::Deserialize, serde::Serialize, Default)]
+    #[serde(rename_all = "kebab-case")]
+    pub enum DirectIoOnAlignmentErrorAction {
+        Error,
+        #[default]
+        FallbackToBuffered,
+    }
+
+    #[derive(Debug, PartialEq, Eq, Clone, serde::Deserialize, serde::Serialize, Default)]
+    #[serde(tag = "type", rename_all = "kebab-case")]
+    pub enum DirectIoLatencyPadding {
+        /// Pad virtual file operations with IO to a fake file.
+        FakeFileRW { path: PathBuf },
+        #[default]
+        None,
+    }
 }
 
 // Wrapped in libpq CopyData
@@ -1494,7 +1482,8 @@ mod tests {
             state: TenantState::Active,
             current_physical_size: Some(42),
             attachment_status: TenantAttachmentStatus::Attached,
-            generation: None,
+            generation: 1,
+            gc_blocking: None,
         };
         let expected_active = json!({
             "id": original_active.id.to_string(),
@@ -1504,7 +1493,8 @@ mod tests {
             "current_physical_size": 42,
             "attachment_status": {
                 "slug":"attached",
-            }
+            },
+            "generation" : 1
         });
 
         let original_broken = TenantInfo {
@@ -1515,7 +1505,8 @@ mod tests {
             },
             current_physical_size: Some(42),
             attachment_status: TenantAttachmentStatus::Attached,
-            generation: None,
+            generation: 1,
+            gc_blocking: None,
         };
         let expected_broken = json!({
             "id": original_broken.id.to_string(),
@@ -1529,7 +1520,8 @@ mod tests {
             "current_physical_size": 42,
             "attachment_status": {
                 "slug":"attached",
-            }
+            },
+            "generation" : 1
         });
 
         assert_eq!(
@@ -1547,18 +1539,6 @@ mod tests {
 
     #[test]
     fn test_reject_unknown_field() {
-        let id = TenantId::generate();
-        let create_request = json!({
-            "new_tenant_id": id.to_string(),
-            "unknown_field": "unknown_value".to_string(),
-        });
-        let err = serde_json::from_value::<TenantCreateRequest>(create_request).unwrap_err();
-        assert!(
-            err.to_string().contains("unknown field `unknown_field`"),
-            "expect unknown field `unknown_field` error, got: {}",
-            err
-        );
-
         let id = TenantId::generate();
         let config_request = json!({
             "tenant_id": id.to_string(),
@@ -1570,18 +1550,6 @@ mod tests {
             "expect unknown field `unknown_field` error, got: {}",
             err
         );
-
-        let attach_request = json!({
-            "config": {
-                "unknown_field": "unknown_value".to_string(),
-            },
-        });
-        let err = serde_json::from_value::<TenantAttachRequest>(attach_request).unwrap_err();
-        assert!(
-            err.to_string().contains("unknown field `unknown_field`"),
-            "expect unknown field `unknown_field` error, got: {}",
-            err
-        );
     }
 
     #[test]
@@ -1705,4 +1673,25 @@ mod tests {
             AuxFilePolicy::CrossValidation
         );
     }
+
+    #[test]
+    fn test_image_compression_algorithm_parsing() {
+        use ImageCompressionAlgorithm::*;
+        assert_eq!(
+            ImageCompressionAlgorithm::from_str("disabled").unwrap(),
+            Disabled
+        );
+        assert_eq!(
+            ImageCompressionAlgorithm::from_str("zstd").unwrap(),
+            Zstd { level: None }
+        );
+        assert_eq!(
+            ImageCompressionAlgorithm::from_str("zstd(18)").unwrap(),
+            Zstd { level: Some(18) }
+        );
+        assert_eq!(
+            ImageCompressionAlgorithm::from_str("zstd(-3)").unwrap(),
+            Zstd { level: Some(-3) }
+        );
+    }
 }

libs/pageserver_api/src/models/detach_ancestor.rs

@@ -1,6 +1,8 @@
+use std::collections::HashSet;
+
 use utils::id::TimelineId;
 
-#[derive(Default, serde::Serialize)]
+#[derive(Debug, Default, PartialEq, serde::Serialize, serde::Deserialize)]
 pub struct AncestorDetached {
-    pub reparented_timelines: Vec<TimelineId>,
+    pub reparented_timelines: HashSet<TimelineId>,
 }

libs/pageserver_api/src/models/utilization.rs

@@ -1,4 +1,5 @@
-use utils::serde_system_time::SystemTime;
+use std::time::SystemTime;
+use utils::{serde_percent::Percent, serde_system_time};
 
 /// Pageserver current utilization and scoring for how good candidate the pageserver would be for
 /// the next tenant.
@@ -9,19 +10,88 @@ use utils::serde_system_time::SystemTime;
 /// not handle full u64 values properly.
 #[derive(serde::Serialize, serde::Deserialize, Debug, Clone)]
 pub struct PageserverUtilization {
-    /// Used disk space
+    /// Used disk space (physical, ground truth from statfs())
     #[serde(serialize_with = "ser_saturating_u63")]
     pub disk_usage_bytes: u64,
     /// Free disk space
     #[serde(serialize_with = "ser_saturating_u63")]
     pub free_space_bytes: u64,
-    /// Lower is better score for how good candidate for a next tenant would this pageserver be.
-    #[serde(serialize_with = "ser_saturating_u63")]
+
+    /// Wanted disk space, based on the tenant shards currently present on this pageserver: this
+    /// is like disk_usage_bytes, but it is stable and does not change with the cache state of
+    /// tenants, whereas disk_usage_bytes may reach the disk eviction `max_usage_pct` and stay
+    /// there, or may be unrealistically low if the pageserver has attached tenants which haven't
+    /// downloaded layers yet.
+    #[serde(serialize_with = "ser_saturating_u63", default)]
+    pub disk_wanted_bytes: u64,
+
+    // What proportion of total disk space will this pageserver use before it starts evicting data?
+    #[serde(default = "unity_percent")]
+    pub disk_usable_pct: Percent,
+
+    // How many shards are currently on this node?
+    #[serde(default)]
+    pub shard_count: u32,
+
+    // How many shards should this node be able to handle at most?
+    #[serde(default)]
+    pub max_shard_count: u32,
+
+    /// Cached result of [`Self::score`]
     pub utilization_score: u64,
+
     /// When was this snapshot captured, pageserver local time.
     ///
     /// Use millis to give confidence that the value is regenerated often enough.
-    pub captured_at: SystemTime,
+    pub captured_at: serde_system_time::SystemTime,
+}
+
+fn unity_percent() -> Percent {
+    Percent::new(0).unwrap()
+}
+
+impl PageserverUtilization {
+    const UTILIZATION_FULL: u64 = 1000000;
+
+    /// Calculate a utilization score.  The result is to be inrepreted as a fraction of
+    /// Self::UTILIZATION_FULL.
+    ///
+    /// Lower values are more affine to scheduling more work on this node.
+    /// - UTILIZATION_FULL represents an ideal node which is fully utilized but should not receive any more work.
+    /// - 0.0 represents an empty node.
+    /// - Negative values are forbidden
+    /// - Values over UTILIZATION_FULL indicate an overloaded node, which may show degraded performance due to
+    ///   layer eviction.
+    pub fn score(&self) -> u64 {
+        let disk_usable_capacity = ((self.disk_usage_bytes + self.free_space_bytes)
+            * self.disk_usable_pct.get() as u64)
+            / 100;
+        let disk_utilization_score =
+            self.disk_wanted_bytes * Self::UTILIZATION_FULL / disk_usable_capacity;
+
+        let shard_utilization_score =
+            self.shard_count as u64 * Self::UTILIZATION_FULL / self.max_shard_count as u64;
+        std::cmp::max(disk_utilization_score, shard_utilization_score)
+    }
+
+    pub fn refresh_score(&mut self) {
+        self.utilization_score = self.score();
+    }
+
+    /// A utilization structure that has a full utilization score: use this as a placeholder when
+    /// you need a utilization but don't have real values yet.
+    pub fn full() -> Self {
+        Self {
+            disk_usage_bytes: 1,
+            free_space_bytes: 0,
+            disk_wanted_bytes: 1,
+            disk_usable_pct: Percent::new(100).unwrap(),
+            shard_count: 1,
+            max_shard_count: 1,
+            utilization_score: Self::UTILIZATION_FULL,
+            captured_at: serde_system_time::SystemTime(SystemTime::now()),
+        }
+    }
 }
 
 /// openapi knows only `format: int64`, so avoid outputting a non-parseable value by generated clients.
@@ -49,15 +119,19 @@ mod tests {
         let doc = PageserverUtilization {
             disk_usage_bytes: u64::MAX,
             free_space_bytes: 0,
-            utilization_score: u64::MAX,
-            captured_at: SystemTime(
+            disk_wanted_bytes: u64::MAX,
+            utilization_score: 13,
+            disk_usable_pct: Percent::new(90).unwrap(),
+            shard_count: 100,
+            max_shard_count: 200,
+            captured_at: serde_system_time::SystemTime(
                 std::time::SystemTime::UNIX_EPOCH + Duration::from_secs(1708509779),
             ),
         };
 
         let s = serde_json::to_string(&doc).unwrap();
 
-        let expected = r#"{"disk_usage_bytes":9223372036854775807,"free_space_bytes":0,"utilization_score":9223372036854775807,"captured_at":"2024-02-21T10:02:59.000Z"}"#;
+        let expected = "{\"disk_usage_bytes\":9223372036854775807,\"free_space_bytes\":0,\"disk_wanted_bytes\":9223372036854775807,\"disk_usable_pct\":90,\"shard_count\":100,\"max_shard_count\":200,\"utilization_score\":13,\"captured_at\":\"2024-02-21T10:02:59.000Z\"}";
 
         assert_eq!(s, expected);
     }

libs/pageserver_api/src/shard.rs

@@ -1,59 +1,42 @@
-use std::{ops::RangeInclusive, str::FromStr};
+//! See docs/rfcs/031-sharding-static.md for an overview of sharding.
+//!
+//! This module contains a variety of types used to represent the concept of sharding
+//! a Neon tenant across multiple physical shards.  Since there are quite a few of these,
+//! we provide an summary here.
+//!
+//! Types used to describe shards:
+//! - [`ShardCount`] describes how many shards make up a tenant, plus the magic `unsharded` value
+//!   which identifies a tenant which is not shard-aware.  This means its storage paths do not include
+//!   a shard suffix.
+//! - [`ShardNumber`] is simply the zero-based index of a shard within a tenant.
+//! - [`ShardIndex`] is the 2-tuple of `ShardCount` and `ShardNumber`, it's just like a `TenantShardId`
+//!   without the tenant ID.  This is useful for things that are implicitly scoped to a particular
+//!   tenant, such as layer files.
+//! - [`ShardIdentity`]` is the full description of a particular shard's parameters, in sufficient
+//!   detail to convert a [`Key`] to a [`ShardNumber`] when deciding where to write/read.
+//! - The [`ShardSlug`] is a terse formatter for ShardCount and ShardNumber, written as
+//!   four hex digits.  An unsharded tenant is `0000`.
+//! - [`TenantShardId`] is the unique ID of a particular shard within a particular tenant
+//!
+//! Types used to describe the parameters for data distribution in a sharded tenant:
+//! - [`ShardStripeSize`] controls how long contiguous runs of [`Key`]s (stripes) are when distributed across
+//!   multiple shards.  Its value is given in 8kiB pages.
+//! - [`ShardLayout`] describes the data distribution scheme, and at time of writing is
+//!   always zero: this is provided for future upgrades that might introduce different
+//!   data distribution schemes.
+//!
+//! Examples:
+//! - A legacy unsharded tenant has one shard with ShardCount(0), ShardNumber(0), and its slug is 0000
+//! - A single sharded tenant has one shard with ShardCount(1), ShardNumber(0), and its slug is 0001
+//! - In a tenant with 4 shards, each shard has ShardCount(N), ShardNumber(i) where i in 0..N-1 (inclusive),
+//!   and their slugs are 0004, 0104, 0204, and 0304.
 
 use crate::{key::Key, models::ShardParameters};
-use hex::FromHex;
 use postgres_ffi::relfile_utils::INIT_FORKNUM;
 use serde::{Deserialize, Serialize};
-use utils::id::TenantId;
 
-/// See docs/rfcs/031-sharding-static.md for an overview of sharding.
-///
-/// This module contains a variety of types used to represent the concept of sharding
-/// a Neon tenant across multiple physical shards.  Since there are quite a few of these,
-/// we provide an summary here.
-///
-/// Types used to describe shards:
-/// - [`ShardCount`] describes how many shards make up a tenant, plus the magic `unsharded` value
-///   which identifies a tenant which is not shard-aware.  This means its storage paths do not include
-///   a shard suffix.
-/// - [`ShardNumber`] is simply the zero-based index of a shard within a tenant.
-/// - [`ShardIndex`] is the 2-tuple of `ShardCount` and `ShardNumber`, it's just like a `TenantShardId`
-///   without the tenant ID.  This is useful for things that are implicitly scoped to a particular
-///   tenant, such as layer files.
-/// - [`ShardIdentity`]` is the full description of a particular shard's parameters, in sufficient
-///   detail to convert a [`Key`] to a [`ShardNumber`] when deciding where to write/read.
-/// - The [`ShardSlug`] is a terse formatter for ShardCount and ShardNumber, written as
-///   four hex digits.  An unsharded tenant is `0000`.
-/// - [`TenantShardId`] is the unique ID of a particular shard within a particular tenant
-///
-/// Types used to describe the parameters for data distribution in a sharded tenant:
-/// - [`ShardStripeSize`] controls how long contiguous runs of [`Key`]s (stripes) are when distributed across
-///   multiple shards.  Its value is given in 8kiB pages.
-/// - [`ShardLayout`] describes the data distribution scheme, and at time of writing is
-///   always zero: this is provided for future upgrades that might introduce different
-///   data distribution schemes.
-///
-/// Examples:
-/// - A legacy unsharded tenant has one shard with ShardCount(0), ShardNumber(0), and its slug is 0000
-/// - A single sharded tenant has one shard with ShardCount(1), ShardNumber(0), and its slug is 0001
-/// - In a tenant with 4 shards, each shard has ShardCount(N), ShardNumber(i) where i in 0..N-1 (inclusive),
-///   and their slugs are 0004, 0104, 0204, and 0304.
-
-#[derive(Ord, PartialOrd, Eq, PartialEq, Clone, Copy, Serialize, Deserialize, Debug, Hash)]
-pub struct ShardNumber(pub u8);
-
-#[derive(Ord, PartialOrd, Eq, PartialEq, Clone, Copy, Serialize, Deserialize, Debug, Hash)]
-pub struct ShardCount(u8);
-
-/// Combination of ShardNumber and ShardCount.  For use within the context of a particular tenant,
-/// when we need to know which shard we're dealing with, but do not need to know the full
-/// ShardIdentity (because we won't be doing any page->shard mapping), and do not need to know
-/// the fully qualified TenantShardId.
-#[derive(Eq, PartialEq, PartialOrd, Ord, Clone, Copy, Hash)]
-pub struct ShardIndex {
-    pub shard_number: ShardNumber,
-    pub shard_count: ShardCount,
-}
+#[doc(inline)]
+pub use ::utils::shard::*;
 
 /// The ShardIdentity contains enough information to map a [`Key`] to a [`ShardNumber`],
 /// and to check whether that [`ShardNumber`] is the same as the current shard.
@@ -65,362 +48,6 @@ pub struct ShardIdentity {
     layout: ShardLayout,
 }
 
-/// Formatting helper, for generating the `shard_id` label in traces.
-struct ShardSlug<'a>(&'a TenantShardId);
-
-/// TenantShardId globally identifies a particular shard in a particular tenant.
-///
-/// These are written as `<TenantId>-<ShardSlug>`, for example:
-///   # The second shard in a two-shard tenant
-///   072f1291a5310026820b2fe4b2968934-0102
-///
-/// If the `ShardCount` is _unsharded_, the `TenantShardId` is written without
-/// a shard suffix and is equivalent to the encoding of a `TenantId`: this enables
-/// an unsharded [`TenantShardId`] to be used interchangably with a [`TenantId`].
-///
-/// The human-readable encoding of an unsharded TenantShardId, such as used in API URLs,
-/// is both forward and backward compatible with TenantId: a legacy TenantId can be
-/// decoded as a TenantShardId, and when re-encoded it will be parseable
-/// as a TenantId.
-#[derive(Eq, PartialEq, PartialOrd, Ord, Clone, Copy, Hash)]
-pub struct TenantShardId {
-    pub tenant_id: TenantId,
-    pub shard_number: ShardNumber,
-    pub shard_count: ShardCount,
-}
-
-impl ShardCount {
-    pub const MAX: Self = Self(u8::MAX);
-
-    /// The internal value of a ShardCount may be zero, which means "1 shard, but use
-    /// legacy format for TenantShardId that excludes the shard suffix", also known
-    /// as [`TenantShardId::unsharded`].
-    ///
-    /// This method returns the actual number of shards, i.e. if our internal value is
-    /// zero, we return 1 (unsharded tenants have 1 shard).
-    pub fn count(&self) -> u8 {
-        if self.0 > 0 {
-            self.0
-        } else {
-            1
-        }
-    }
-
-    /// The literal internal value: this is **not** the number of shards in the
-    /// tenant, as we have a special zero value for legacy unsharded tenants.  Use
-    /// [`Self::count`] if you want to know the cardinality of shards.
-    pub fn literal(&self) -> u8 {
-        self.0
-    }
-
-    /// Whether the `ShardCount` is for an unsharded tenant, so uses one shard but
-    /// uses the legacy format for `TenantShardId`. See also the documentation for
-    /// [`Self::count`].
-    pub fn is_unsharded(&self) -> bool {
-        self.0 == 0
-    }
-
-    /// `v` may be zero, or the number of shards in the tenant.  `v` is what
-    /// [`Self::literal`] would return.
-    pub const fn new(val: u8) -> Self {
-        Self(val)
-    }
-}
-
-impl ShardNumber {
-    pub const MAX: Self = Self(u8::MAX);
-}
-
-impl TenantShardId {
-    pub fn unsharded(tenant_id: TenantId) -> Self {
-        Self {
-            tenant_id,
-            shard_number: ShardNumber(0),
-            shard_count: ShardCount(0),
-        }
-    }
-
-    /// The range of all TenantShardId that belong to a particular TenantId.  This is useful when
-    /// you have a BTreeMap of TenantShardId, and are querying by TenantId.
-    pub fn tenant_range(tenant_id: TenantId) -> RangeInclusive<Self> {
-        RangeInclusive::new(
-            Self {
-                tenant_id,
-                shard_number: ShardNumber(0),
-                shard_count: ShardCount(0),
-            },
-            Self {
-                tenant_id,
-                shard_number: ShardNumber::MAX,
-                shard_count: ShardCount::MAX,
-            },
-        )
-    }
-
-    pub fn shard_slug(&self) -> impl std::fmt::Display + '_ {
-        ShardSlug(self)
-    }
-
-    /// Convenience for code that has special behavior on the 0th shard.
-    pub fn is_shard_zero(&self) -> bool {
-        self.shard_number == ShardNumber(0)
-    }
-
-    /// The "unsharded" value is distinct from simply having a single shard: it represents
-    /// a tenant which is not shard-aware at all, and whose storage paths will not include
-    /// a shard suffix.
-    pub fn is_unsharded(&self) -> bool {
-        self.shard_number == ShardNumber(0) && self.shard_count.is_unsharded()
-    }
-
-    /// Convenience for dropping the tenant_id and just getting the ShardIndex: this
-    /// is useful when logging from code that is already in a span that includes tenant ID, to
-    /// keep messages reasonably terse.
-    pub fn to_index(&self) -> ShardIndex {
-        ShardIndex {
-            shard_number: self.shard_number,
-            shard_count: self.shard_count,
-        }
-    }
-
-    /// Calculate the children of this TenantShardId when splitting the overall tenant into
-    /// the given number of shards.
-    pub fn split(&self, new_shard_count: ShardCount) -> Vec<TenantShardId> {
-        let effective_old_shard_count = std::cmp::max(self.shard_count.0, 1);
-        let mut child_shards = Vec::new();
-        for shard_number in 0..ShardNumber(new_shard_count.0).0 {
-            // Key mapping is based on a round robin mapping of key hash modulo shard count,
-            // so our child shards are the ones which the same keys would map to.
-            if shard_number % effective_old_shard_count == self.shard_number.0 {
-                child_shards.push(TenantShardId {
-                    tenant_id: self.tenant_id,
-                    shard_number: ShardNumber(shard_number),
-                    shard_count: new_shard_count,
-                })
-            }
-        }
-
-        child_shards
-    }
-}
-
-impl<'a> std::fmt::Display for ShardSlug<'a> {
-    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
-        write!(
-            f,
-            "{:02x}{:02x}",
-            self.0.shard_number.0, self.0.shard_count.0
-        )
-    }
-}
-
-impl std::fmt::Display for TenantShardId {
-    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
-        if self.shard_count != ShardCount(0) {
-            write!(f, "{}-{}", self.tenant_id, self.shard_slug())
-        } else {
-            // Legacy case (shard_count == 0) -- format as just the tenant id.  Note that this
-            // is distinct from the normal single shard case (shard count == 1).
-            self.tenant_id.fmt(f)
-        }
-    }
-}
-
-impl std::fmt::Debug for TenantShardId {
-    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
-        // Debug is the same as Display: the compact hex representation
-        write!(f, "{}", self)
-    }
-}
-
-impl std::str::FromStr for TenantShardId {
-    type Err = hex::FromHexError;
-
-    fn from_str(s: &str) -> Result<Self, Self::Err> {
-        // Expect format: 16 byte TenantId, '-', 1 byte shard number, 1 byte shard count
-        if s.len() == 32 {
-            // Legacy case: no shard specified
-            Ok(Self {
-                tenant_id: TenantId::from_str(s)?,
-                shard_number: ShardNumber(0),
-                shard_count: ShardCount(0),
-            })
-        } else if s.len() == 37 {
-            let bytes = s.as_bytes();
-            let tenant_id = TenantId::from_hex(&bytes[0..32])?;
-            let mut shard_parts: [u8; 2] = [0u8; 2];
-            hex::decode_to_slice(&bytes[33..37], &mut shard_parts)?;
-            Ok(Self {
-                tenant_id,
-                shard_number: ShardNumber(shard_parts[0]),
-                shard_count: ShardCount(shard_parts[1]),
-            })
-        } else {
-            Err(hex::FromHexError::InvalidStringLength)
-        }
-    }
-}
-
-impl From<[u8; 18]> for TenantShardId {
-    fn from(b: [u8; 18]) -> Self {
-        let tenant_id_bytes: [u8; 16] = b[0..16].try_into().unwrap();
-
-        Self {
-            tenant_id: TenantId::from(tenant_id_bytes),
-            shard_number: ShardNumber(b[16]),
-            shard_count: ShardCount(b[17]),
-        }
-    }
-}
-
-impl ShardIndex {
-    pub fn new(number: ShardNumber, count: ShardCount) -> Self {
-        Self {
-            shard_number: number,
-            shard_count: count,
-        }
-    }
-    pub fn unsharded() -> Self {
-        Self {
-            shard_number: ShardNumber(0),
-            shard_count: ShardCount(0),
-        }
-    }
-
-    /// The "unsharded" value is distinct from simply having a single shard: it represents
-    /// a tenant which is not shard-aware at all, and whose storage paths will not include
-    /// a shard suffix.
-    pub fn is_unsharded(&self) -> bool {
-        self.shard_number == ShardNumber(0) && self.shard_count == ShardCount(0)
-    }
-
-    /// For use in constructing remote storage paths: concatenate this with a TenantId
-    /// to get a fully qualified TenantShardId.
-    ///
-    /// Backward compat: this function returns an empty string if Self::is_unsharded, such
-    /// that the legacy pre-sharding remote key format is preserved.
-    pub fn get_suffix(&self) -> String {
-        if self.is_unsharded() {
-            "".to_string()
-        } else {
-            format!("-{:02x}{:02x}", self.shard_number.0, self.shard_count.0)
-        }
-    }
-}
-
-impl std::fmt::Display for ShardIndex {
-    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
-        write!(f, "{:02x}{:02x}", self.shard_number.0, self.shard_count.0)
-    }
-}
-
-impl std::fmt::Debug for ShardIndex {
-    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
-        // Debug is the same as Display: the compact hex representation
-        write!(f, "{}", self)
-    }
-}
-
-impl std::str::FromStr for ShardIndex {
-    type Err = hex::FromHexError;
-
-    fn from_str(s: &str) -> Result<Self, Self::Err> {
-        // Expect format: 1 byte shard number, 1 byte shard count
-        if s.len() == 4 {
-            let bytes = s.as_bytes();
-            let mut shard_parts: [u8; 2] = [0u8; 2];
-            hex::decode_to_slice(bytes, &mut shard_parts)?;
-            Ok(Self {
-                shard_number: ShardNumber(shard_parts[0]),
-                shard_count: ShardCount(shard_parts[1]),
-            })
-        } else {
-            Err(hex::FromHexError::InvalidStringLength)
-        }
-    }
-}
-
-impl From<[u8; 2]> for ShardIndex {
-    fn from(b: [u8; 2]) -> Self {
-        Self {
-            shard_number: ShardNumber(b[0]),
-            shard_count: ShardCount(b[1]),
-        }
-    }
-}
-
-impl Serialize for TenantShardId {
-    fn serialize<S>(&self, serializer: S) -> Result<S::Ok, S::Error>
-    where
-        S: serde::Serializer,
-    {
-        if serializer.is_human_readable() {
-            serializer.collect_str(self)
-        } else {
-            // Note: while human encoding of [`TenantShardId`] is backward and forward
-            // compatible, this binary encoding is not.
-            let mut packed: [u8; 18] = [0; 18];
-            packed[0..16].clone_from_slice(&self.tenant_id.as_arr());
-            packed[16] = self.shard_number.0;
-            packed[17] = self.shard_count.0;
-
-            packed.serialize(serializer)
-        }
-    }
-}
-
-impl<'de> Deserialize<'de> for TenantShardId {
-    fn deserialize<D>(deserializer: D) -> Result<Self, D::Error>
-    where
-        D: serde::Deserializer<'de>,
-    {
-        struct IdVisitor {
-            is_human_readable_deserializer: bool,
-        }
-
-        impl<'de> serde::de::Visitor<'de> for IdVisitor {
-            type Value = TenantShardId;
-
-            fn expecting(&self, formatter: &mut std::fmt::Formatter) -> std::fmt::Result {
-                if self.is_human_readable_deserializer {
-                    formatter.write_str("value in form of hex string")
-                } else {
-                    formatter.write_str("value in form of integer array([u8; 18])")
-                }
-            }
-
-            fn visit_seq<A>(self, seq: A) -> Result<Self::Value, A::Error>
-            where
-                A: serde::de::SeqAccess<'de>,
-            {
-                let s = serde::de::value::SeqAccessDeserializer::new(seq);
-                let id: [u8; 18] = Deserialize::deserialize(s)?;
-                Ok(TenantShardId::from(id))
-            }
-
-            fn visit_str<E>(self, v: &str) -> Result<Self::Value, E>
-            where
-                E: serde::de::Error,
-            {
-                TenantShardId::from_str(v).map_err(E::custom)
-            }
-        }
-
-        if deserializer.is_human_readable() {
-            deserializer.deserialize_str(IdVisitor {
-                is_human_readable_deserializer: true,
-            })
-        } else {
-            deserializer.deserialize_tuple(
-                18,
-                IdVisitor {
-                    is_human_readable_deserializer: false,
-                },
-            )
-        }
-    }
-}
-
 /// Stripe size in number of pages
 #[derive(Clone, Copy, Serialize, Deserialize, Eq, PartialEq, Debug)]
 pub struct ShardStripeSize(pub u32);
@@ -585,77 +212,6 @@ impl ShardIdentity {
     }
 }
 
-impl Serialize for ShardIndex {
-    fn serialize<S>(&self, serializer: S) -> Result<S::Ok, S::Error>
-    where
-        S: serde::Serializer,
-    {
-        if serializer.is_human_readable() {
-            serializer.collect_str(self)
-        } else {
-            // Binary encoding is not used in index_part.json, but is included in anticipation of
-            // switching various structures (e.g. inter-process communication, remote metadata) to more
-            // compact binary encodings in future.
-            let mut packed: [u8; 2] = [0; 2];
-            packed[0] = self.shard_number.0;
-            packed[1] = self.shard_count.0;
-            packed.serialize(serializer)
-        }
-    }
-}
-
-impl<'de> Deserialize<'de> for ShardIndex {
-    fn deserialize<D>(deserializer: D) -> Result<Self, D::Error>
-    where
-        D: serde::Deserializer<'de>,
-    {
-        struct IdVisitor {
-            is_human_readable_deserializer: bool,
-        }
-
-        impl<'de> serde::de::Visitor<'de> for IdVisitor {
-            type Value = ShardIndex;
-
-            fn expecting(&self, formatter: &mut std::fmt::Formatter) -> std::fmt::Result {
-                if self.is_human_readable_deserializer {
-                    formatter.write_str("value in form of hex string")
-                } else {
-                    formatter.write_str("value in form of integer array([u8; 2])")
-                }
-            }
-
-            fn visit_seq<A>(self, seq: A) -> Result<Self::Value, A::Error>
-            where
-                A: serde::de::SeqAccess<'de>,
-            {
-                let s = serde::de::value::SeqAccessDeserializer::new(seq);
-                let id: [u8; 2] = Deserialize::deserialize(s)?;
-                Ok(ShardIndex::from(id))
-            }
-
-            fn visit_str<E>(self, v: &str) -> Result<Self::Value, E>
-            where
-                E: serde::de::Error,
-            {
-                ShardIndex::from_str(v).map_err(E::custom)
-            }
-        }
-
-        if deserializer.is_human_readable() {
-            deserializer.deserialize_str(IdVisitor {
-                is_human_readable_deserializer: true,
-            })
-        } else {
-            deserializer.deserialize_tuple(
-                2,
-                IdVisitor {
-                    is_human_readable_deserializer: false,
-                },
-            )
-        }
-    }
-}
-
 /// Whether this key is always held on shard 0 (e.g. shard 0 holds all SLRU keys
 /// in order to be able to serve basebackup requests without peer communication).
 fn key_is_shard0(key: &Key) -> bool {
@@ -737,7 +293,9 @@ pub fn describe(
 
 #[cfg(test)]
 mod tests {
-    use utils::Hex;
+    use std::str::FromStr;
+
+    use utils::{id::TenantId, Hex};
 
     use super::*;
 

libs/postgres_backend/Cargo.toml

@@ -13,6 +13,7 @@ rustls.workspace = true
 serde.workspace = true
 thiserror.workspace = true
 tokio.workspace = true
+tokio-util.workspace = true
 tokio-rustls.workspace = true
 tracing.workspace = true
 
@@ -23,4 +24,4 @@ workspace_hack.workspace = true
 once_cell.workspace = true
 rustls-pemfile.workspace = true
 tokio-postgres.workspace = true
-tokio-postgres-rustls.workspace = true
+tokio-postgres-rustls.workspace = true

libs/postgres_backend/src/lib.rs

@@ -16,6 +16,7 @@ use std::{fmt, io};
 use std::{future::Future, str::FromStr};
 use tokio::io::{AsyncRead, AsyncWrite};
 use tokio_rustls::TlsAcceptor;
+use tokio_util::sync::CancellationToken;
 use tracing::{debug, error, info, trace, warn};
 
 use pq_proto::framed::{ConnectionError, Framed, FramedReader, FramedWriter};
@@ -400,21 +401,15 @@ impl<IO: AsyncRead + AsyncWrite + Unpin> PostgresBackend<IO> {
     }
 
     /// Wrapper for run_message_loop() that shuts down socket when we are done
-    pub async fn run<F, S>(
+    pub async fn run(
         mut self,
         handler: &mut impl Handler<IO>,
-        shutdown_watcher: F,
-    ) -> Result<(), QueryError>
-    where
-        F: Fn() -> S + Clone,
-        S: Future,
-    {
-        let ret = self
-            .run_message_loop(handler, shutdown_watcher.clone())
-            .await;
+        cancel: &CancellationToken,
+    ) -> Result<(), QueryError> {
+        let ret = self.run_message_loop(handler, cancel).await;
 
         tokio::select! {
-            _ = shutdown_watcher() => {
+            _ = cancel.cancelled() => {
                 // do nothing; we most likely got already stopped by shutdown and will log it next.
             }
             _ = self.framed.shutdown() => {
@@ -444,21 +439,17 @@ impl<IO: AsyncRead + AsyncWrite + Unpin> PostgresBackend<IO> {
         }
     }
 
-    async fn run_message_loop<F, S>(
+    async fn run_message_loop(
         &mut self,
         handler: &mut impl Handler<IO>,
-        shutdown_watcher: F,
-    ) -> Result<(), QueryError>
-    where
-        F: Fn() -> S,
-        S: Future,
-    {
+        cancel: &CancellationToken,
+    ) -> Result<(), QueryError> {
         trace!("postgres backend to {:?} started", self.peer_addr);
 
         tokio::select!(
             biased;
 
-            _ = shutdown_watcher() => {
+            _ = cancel.cancelled() => {
                 // We were requested to shut down.
                 tracing::info!("shutdown request received during handshake");
                 return Err(QueryError::Shutdown)
@@ -473,7 +464,7 @@ impl<IO: AsyncRead + AsyncWrite + Unpin> PostgresBackend<IO> {
         let mut query_string = Bytes::new();
         while let Some(msg) = tokio::select!(
             biased;
-            _ = shutdown_watcher() => {
+            _ = cancel.cancelled() => {
                 // We were requested to shut down.
                 tracing::info!("shutdown request received in run_message_loop");
                 return Err(QueryError::Shutdown)
@@ -485,7 +476,7 @@ impl<IO: AsyncRead + AsyncWrite + Unpin> PostgresBackend<IO> {
             let result = self.process_message(handler, msg, &mut query_string).await;
             tokio::select!(
                 biased;
-                _ = shutdown_watcher() => {
+                _ = cancel.cancelled() => {
                     // We were requested to shut down.
                     tracing::info!("shutdown request received during response flush");
 
@@ -672,11 +663,17 @@ impl<IO: AsyncRead + AsyncWrite + Unpin> PostgresBackend<IO> {
         assert!(self.state < ProtoState::Authentication);
         let have_tls = self.tls_config.is_some();
         match msg {
-            FeStartupPacket::SslRequest => {
+            FeStartupPacket::SslRequest { direct } => {
                 debug!("SSL requested");
 
-                self.write_message(&BeMessage::EncryptionResponse(have_tls))
-                    .await?;
+                if !direct {
+                    self.write_message(&BeMessage::EncryptionResponse(have_tls))
+                        .await?;
+                } else if !have_tls {
+                    return Err(QueryError::Other(anyhow::anyhow!(
+                        "direct SSL negotiation but no TLS support"
+                    )));
+                }
 
                 if have_tls {
                     self.start_tls().await?;

libs/postgres_backend/tests/simple_select.rs

@@ -3,13 +3,14 @@ use once_cell::sync::Lazy;
 use postgres_backend::{AuthType, Handler, PostgresBackend, QueryError};
 use pq_proto::{BeMessage, RowDescriptor};
 use std::io::Cursor;
-use std::{future, sync::Arc};
+use std::sync::Arc;
 use tokio::io::{AsyncRead, AsyncWrite};
 use tokio::net::{TcpListener, TcpStream};
 use tokio_postgres::config::SslMode;
 use tokio_postgres::tls::MakeTlsConnect;
 use tokio_postgres::{Config, NoTls, SimpleQueryMessage};
 use tokio_postgres_rustls::MakeRustlsConnect;
+use tokio_util::sync::CancellationToken;
 
 // generate client, server test streams
 async fn make_tcp_pair() -> (TcpStream, TcpStream) {
@@ -50,7 +51,7 @@ async fn simple_select() {
 
     tokio::spawn(async move {
         let mut handler = TestHandler {};
-        pgbackend.run(&mut handler, future::pending::<()>).await
+        pgbackend.run(&mut handler, &CancellationToken::new()).await
     });
 
     let conf = Config::new();
@@ -102,7 +103,7 @@ async fn simple_select_ssl() {
 
     tokio::spawn(async move {
         let mut handler = TestHandler {};
-        pgbackend.run(&mut handler, future::pending::<()>).await
+        pgbackend.run(&mut handler, &CancellationToken::new()).await
     });
 
     let client_cfg = rustls::ClientConfig::builder()

libs/postgres_ffi/src/controlfile_utils.rs

@@ -29,7 +29,7 @@ use anyhow::{bail, Result};
 use bytes::{Bytes, BytesMut};
 
 /// Equivalent to sizeof(ControlFileData) in C
-const SIZEOF_CONTROLDATA: usize = std::mem::size_of::<ControlFileData>();
+const SIZEOF_CONTROLDATA: usize = size_of::<ControlFileData>();
 
 impl ControlFileData {
     /// Compute the offset of the `crc` field within the `ControlFileData` struct.

libs/postgres_ffi/src/lib.rs

@@ -143,8 +143,8 @@ pub use v14::xlog_utils::XLogFileName;
 
 pub use v14::bindings::DBState_DB_SHUTDOWNED;
 
-pub fn bkpimage_is_compressed(bimg_info: u8, version: u32) -> anyhow::Result<bool> {
-    dispatch_pgversion!(version, Ok(pgv::bindings::bkpimg_is_compressed(bimg_info)))
+pub fn bkpimage_is_compressed(bimg_info: u8, version: u32) -> bool {
+    dispatch_pgversion!(version, pgv::bindings::bkpimg_is_compressed(bimg_info))
 }
 
 pub fn generate_wal_segment(

libs/postgres_ffi/src/pg_constants.rs

@@ -31,7 +31,7 @@ pub const SMGR_TRUNCATE_FSM: u32 = 0x0004;
 //
 
 // Assumes 8 byte alignment
-const SIZEOF_PAGE_HEADER_DATA: usize = std::mem::size_of::<PageHeaderData>();
+const SIZEOF_PAGE_HEADER_DATA: usize = size_of::<PageHeaderData>();
 pub const MAXALIGN_SIZE_OF_PAGE_HEADER_DATA: usize = (SIZEOF_PAGE_HEADER_DATA + 7) & !7;
 
 //
@@ -191,7 +191,7 @@ pub const XLR_RMGR_INFO_MASK: u8 = 0xF0;
 pub const XLOG_TBLSPC_CREATE: u8 = 0x00;
 pub const XLOG_TBLSPC_DROP: u8 = 0x10;
 
-pub const SIZEOF_XLOGRECORD: u32 = std::mem::size_of::<XLogRecord>() as u32;
+pub const SIZEOF_XLOGRECORD: u32 = size_of::<XLogRecord>() as u32;
 
 //
 // from xlogrecord.h

libs/postgres_ffi/src/xlog_utils.rs

@@ -42,9 +42,9 @@ pub const XLP_FIRST_IS_CONTRECORD: u16 = 0x0001;
 pub const XLP_REM_LEN_OFFS: usize = 2 + 2 + 4 + 8;
 pub const XLOG_RECORD_CRC_OFFS: usize = 4 + 4 + 8 + 1 + 1 + 2;
 
-pub const XLOG_SIZE_OF_XLOG_SHORT_PHD: usize = std::mem::size_of::<XLogPageHeaderData>();
-pub const XLOG_SIZE_OF_XLOG_LONG_PHD: usize = std::mem::size_of::<XLogLongPageHeaderData>();
-pub const XLOG_SIZE_OF_XLOG_RECORD: usize = std::mem::size_of::<XLogRecord>();
+pub const XLOG_SIZE_OF_XLOG_SHORT_PHD: usize = size_of::<XLogPageHeaderData>();
+pub const XLOG_SIZE_OF_XLOG_LONG_PHD: usize = size_of::<XLogLongPageHeaderData>();
+pub const XLOG_SIZE_OF_XLOG_RECORD: usize = size_of::<XLogRecord>();
 #[allow(clippy::identity_op)]
 pub const SIZE_OF_XLOG_RECORD_DATA_HEADER_SHORT: usize = 1 * 2;
 
@@ -311,7 +311,7 @@ impl XLogLongPageHeaderData {
     }
 }
 
-pub const SIZEOF_CHECKPOINT: usize = std::mem::size_of::<CheckPoint>();
+pub const SIZEOF_CHECKPOINT: usize = size_of::<CheckPoint>();
 
 impl CheckPoint {
     pub fn encode(&self) -> Result<Bytes, SerializeError> {
@@ -356,6 +356,28 @@ impl CheckPoint {
         }
         false
     }
+
+    /// Advance next multi-XID/offset to those given in arguments.
+    ///
+    /// It's important that this handles wraparound correctly. This should match the
+    /// MultiXactAdvanceNextMXact() logic in PostgreSQL's xlog_redo() function.
+    ///
+    /// Returns 'true' if the Checkpoint was updated.
+    pub fn update_next_multixid(&mut self, multi_xid: u32, multi_offset: u32) -> bool {
+        let mut modified = false;
+
+        if multi_xid.wrapping_sub(self.nextMulti) as i32 > 0 {
+            self.nextMulti = multi_xid;
+            modified = true;
+        }
+
+        if multi_offset.wrapping_sub(self.nextMultiOffset) as i32 > 0 {
+            self.nextMultiOffset = multi_offset;
+            modified = true;
+        }
+
+        modified
+    }
 }
 
 /// Generate new, empty WAL segment, with correct block headers at the first

libs/postgres_ffi/wal_craft/src/xlog_utils_test.rs

@@ -178,7 +178,7 @@ pub fn test_find_end_of_wal_last_crossing_segment() {
 /// currently 1024.
 #[test]
 pub fn test_update_next_xid() {
-    let checkpoint_buf = [0u8; std::mem::size_of::<CheckPoint>()];
+    let checkpoint_buf = [0u8; size_of::<CheckPoint>()];
     let mut checkpoint = CheckPoint::decode(&checkpoint_buf).unwrap();
 
     checkpoint.nextXid = FullTransactionId { value: 10 };
@@ -202,6 +202,53 @@ pub fn test_update_next_xid() {
     assert_eq!(checkpoint.nextXid.value, 2048);
 }
 
+#[test]
+pub fn test_update_next_multixid() {
+    let checkpoint_buf = [0u8; size_of::<CheckPoint>()];
+    let mut checkpoint = CheckPoint::decode(&checkpoint_buf).unwrap();
+
+    // simple case
+    checkpoint.nextMulti = 20;
+    checkpoint.nextMultiOffset = 20;
+    checkpoint.update_next_multixid(1000, 2000);
+    assert_eq!(checkpoint.nextMulti, 1000);
+    assert_eq!(checkpoint.nextMultiOffset, 2000);
+
+    // No change
+    checkpoint.update_next_multixid(500, 900);
+    assert_eq!(checkpoint.nextMulti, 1000);
+    assert_eq!(checkpoint.nextMultiOffset, 2000);
+
+    // Close to wraparound, but not wrapped around yet
+    checkpoint.nextMulti = 0xffff0000;
+    checkpoint.nextMultiOffset = 0xfffe0000;
+    checkpoint.update_next_multixid(0xffff00ff, 0xfffe00ff);
+    assert_eq!(checkpoint.nextMulti, 0xffff00ff);
+    assert_eq!(checkpoint.nextMultiOffset, 0xfffe00ff);
+
+    // Wraparound
+    checkpoint.update_next_multixid(1, 900);
+    assert_eq!(checkpoint.nextMulti, 1);
+    assert_eq!(checkpoint.nextMultiOffset, 900);
+
+    // Wraparound nextMulti to 0.
+    //
+    // It's a bit surprising that nextMulti can be 0, because that's a special value
+    // (InvalidMultiXactId). However, that's how Postgres does it at multi-xid wraparound:
+    // nextMulti wraps around to 0, but then when the next multi-xid is assigned, it skips
+    // the 0 and the next multi-xid actually assigned is 1.
+    checkpoint.nextMulti = 0xffff0000;
+    checkpoint.nextMultiOffset = 0xfffe0000;
+    checkpoint.update_next_multixid(0, 0xfffe00ff);
+    assert_eq!(checkpoint.nextMulti, 0);
+    assert_eq!(checkpoint.nextMultiOffset, 0xfffe00ff);
+
+    // Wraparound nextMultiOffset to 0
+    checkpoint.update_next_multixid(0, 0);
+    assert_eq!(checkpoint.nextMulti, 0);
+    assert_eq!(checkpoint.nextMultiOffset, 0);
+}
+
 #[test]
 pub fn test_encode_logical_message() {
     let expected = [

libs/pq_proto/src/framed.rs

@@ -44,9 +44,9 @@ impl ConnectionError {
 /// Wraps async io `stream`, providing messages to write/flush + read Postgres
 /// messages.
 pub struct Framed<S> {
-    stream: S,
-    read_buf: BytesMut,
-    write_buf: BytesMut,
+    pub stream: S,
+    pub read_buf: BytesMut,
+    pub write_buf: BytesMut,
 }
 
 impl<S> Framed<S> {

libs/pq_proto/src/lib.rs

@@ -39,14 +39,39 @@ pub enum FeMessage {
     PasswordMessage(Bytes),
 }
 
+#[derive(Clone, Copy, PartialEq, PartialOrd)]
+pub struct ProtocolVersion(u32);
+
+impl ProtocolVersion {
+    pub const fn new(major: u16, minor: u16) -> Self {
+        Self((major as u32) << 16 | minor as u32)
+    }
+    pub const fn minor(self) -> u16 {
+        self.0 as u16
+    }
+    pub const fn major(self) -> u16 {
+        (self.0 >> 16) as u16
+    }
+}
+
+impl fmt::Debug for ProtocolVersion {
+    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
+        f.debug_list()
+            .entry(&self.major())
+            .entry(&self.minor())
+            .finish()
+    }
+}
+
 #[derive(Debug)]
 pub enum FeStartupPacket {
     CancelRequest(CancelKeyData),
-    SslRequest,
+    SslRequest {
+        direct: bool,
+    },
     GssEncRequest,
     StartupMessage {
-        major_version: u32,
-        minor_version: u32,
+        version: ProtocolVersion,
         params: StartupMessageParams,
     },
 }
@@ -301,11 +326,23 @@ impl FeStartupPacket {
     /// different from [`FeMessage::parse`] because startup messages don't have
     /// message type byte; otherwise, its comments apply.
     pub fn parse(buf: &mut BytesMut) -> Result<Option<FeStartupPacket>, ProtocolError> {
+        /// <https://github.com/postgres/postgres/blob/ca481d3c9ab7bf69ff0c8d71ad3951d407f6a33c/src/include/libpq/pqcomm.h#L118>
         const MAX_STARTUP_PACKET_LENGTH: usize = 10000;
-        const RESERVED_INVALID_MAJOR_VERSION: u32 = 1234;
-        const CANCEL_REQUEST_CODE: u32 = 5678;
-        const NEGOTIATE_SSL_CODE: u32 = 5679;
-        const NEGOTIATE_GSS_CODE: u32 = 5680;
+        const RESERVED_INVALID_MAJOR_VERSION: u16 = 1234;
+        /// <https://github.com/postgres/postgres/blob/ca481d3c9ab7bf69ff0c8d71ad3951d407f6a33c/src/include/libpq/pqcomm.h#L132>
+        const CANCEL_REQUEST_CODE: ProtocolVersion = ProtocolVersion::new(1234, 5678);
+        /// <https://github.com/postgres/postgres/blob/ca481d3c9ab7bf69ff0c8d71ad3951d407f6a33c/src/include/libpq/pqcomm.h#L166>
+        const NEGOTIATE_SSL_CODE: ProtocolVersion = ProtocolVersion::new(1234, 5679);
+        /// <https://github.com/postgres/postgres/blob/ca481d3c9ab7bf69ff0c8d71ad3951d407f6a33c/src/include/libpq/pqcomm.h#L167>
+        const NEGOTIATE_GSS_CODE: ProtocolVersion = ProtocolVersion::new(1234, 5680);
+
+        // <https://github.com/postgres/postgres/blob/04bcf9e19a4261fe9c7df37c777592c2e10c32a7/src/backend/tcop/backend_startup.c#L378-L382>
+        // First byte indicates standard SSL handshake message
+        // (It can't be a Postgres startup length because in network byte order
+        // that would be a startup packet hundreds of megabytes long)
+        if buf.first() == Some(&0x16) {
+            return Ok(Some(FeStartupPacket::SslRequest { direct: true }));
+        }
 
         // need at least 4 bytes with packet len
         if buf.len() < 4 {
@@ -338,12 +375,10 @@ impl FeStartupPacket {
         let mut msg = buf.split_to(len).freeze();
         msg.advance(4); // consume len
 
-        let request_code = msg.get_u32();
-        let req_hi = request_code >> 16;
-        let req_lo = request_code & ((1 << 16) - 1);
+        let request_code = ProtocolVersion(msg.get_u32());
         // StartupMessage, CancelRequest, SSLRequest etc are differentiated by request code.
-        let message = match (req_hi, req_lo) {
-            (RESERVED_INVALID_MAJOR_VERSION, CANCEL_REQUEST_CODE) => {
+        let message = match request_code {
+            CANCEL_REQUEST_CODE => {
                 if msg.remaining() != 8 {
                     return Err(ProtocolError::BadMessage(
                         "CancelRequest message is malformed, backend PID / secret key missing"
@@ -355,21 +390,22 @@ impl FeStartupPacket {
                     cancel_key: msg.get_i32(),
                 })
             }
-            (RESERVED_INVALID_MAJOR_VERSION, NEGOTIATE_SSL_CODE) => {
+            NEGOTIATE_SSL_CODE => {
                 // Requested upgrade to SSL (aka TLS)
-                FeStartupPacket::SslRequest
+                FeStartupPacket::SslRequest { direct: false }
             }
-            (RESERVED_INVALID_MAJOR_VERSION, NEGOTIATE_GSS_CODE) => {
+            NEGOTIATE_GSS_CODE => {
                 // Requested upgrade to GSSAPI
                 FeStartupPacket::GssEncRequest
             }
-            (RESERVED_INVALID_MAJOR_VERSION, unrecognized_code) => {
+            version if version.major() == RESERVED_INVALID_MAJOR_VERSION => {
                 return Err(ProtocolError::Protocol(format!(
-                    "Unrecognized request code {unrecognized_code}"
+                    "Unrecognized request code {}",
+                    version.minor()
                 )));
             }
             // TODO bail if protocol major_version is not 3?
-            (major_version, minor_version) => {
+            version => {
                 // StartupMessage
 
                 let s = str::from_utf8(&msg).map_err(|_e| {
@@ -382,8 +418,7 @@ impl FeStartupPacket {
                 })?;
 
                 FeStartupPacket::StartupMessage {
-                    major_version,
-                    minor_version,
+                    version,
                     params: StartupMessageParams {
                         params: msg.slice_ref(s.as_bytes()),
                     },
@@ -522,6 +557,10 @@ pub enum BeMessage<'a> {
     RowDescription(&'a [RowDescriptor<'a>]),
     XLogData(XLogDataBody<'a>),
     NoticeResponse(&'a str),
+    NegotiateProtocolVersion {
+        version: ProtocolVersion,
+        options: &'a [&'a str],
+    },
     KeepAlive(WalSndKeepAlive),
 }
 
@@ -945,6 +984,18 @@ impl<'a> BeMessage<'a> {
                     buf.put_u8(u8::from(req.request_reply));
                 });
             }
+
+            BeMessage::NegotiateProtocolVersion { version, options } => {
+                buf.put_u8(b'v');
+                write_body(buf, |buf| {
+                    buf.put_u32(version.0);
+                    buf.put_u32(options.len() as u32);
+                    for option in options.iter() {
+                        write_cstr(option, buf)?;
+                    }
+                    Ok(())
+                })?
+            }
         }
         Ok(())
     }

libs/remote_storage/Cargo.toml

@@ -7,6 +7,7 @@ license.workspace = true
 [dependencies]
 anyhow.workspace = true
 async-trait.workspace = true
+async-stream.workspace = true
 once_cell.workspace = true
 aws-smithy-async.workspace = true
 aws-smithy-types.workspace = true
@@ -14,8 +15,9 @@ aws-config.workspace = true
 aws-sdk-s3.workspace = true
 aws-credential-types.workspace = true
 bytes.workspace = true
-camino.workspace = true
+camino = { workspace = true, features = ["serde1"] }
 humantime.workspace = true
+humantime-serde.workspace = true
 hyper = { workspace = true, features = ["stream"] }
 futures.workspace = true
 rand.workspace = true