Add a test for generating pprof from collapsed

This provides the binaries and libraries required for the continuous
profiling at runtime.

.config/hakari.toml

@@ -21,14 +21,13 @@ platforms = [
     # "x86_64-apple-darwin",
     # "x86_64-pc-windows-msvc",
 ]
+
 [final-excludes]
 workspace-members = [
     # vm_monitor benefits from the same Cargo.lock as the rest of our artifacts, but
     # it is built primarly in separate repo neondatabase/autoscaling and thus is excluded
     # from depending on workspace-hack because most of the dependencies are not used.
     "vm_monitor",
-    # subzero-core is a stub crate that should be excluded from workspace-hack
-    "subzero-core",
     # All of these exist in libs and are not usually built independently.
     # Putting workspace hack there adds a bottleneck for cargo builds.
     "compute_api",

.dockerignore

@@ -27,4 +27,4 @@
 !storage_controller/
 !vendor/postgres-*/
 !workspace_hack/
-!build-tools/patches
+!build_tools/patches

.github/actionlint.yml

@@ -31,7 +31,6 @@ config-variables:
   - NEON_PROD_AWS_ACCOUNT_ID
   - PGREGRESS_PG16_PROJECT_ID
   - PGREGRESS_PG17_PROJECT_ID
-  - PREWARM_PGBENCH_SIZE
   - REMOTE_STORAGE_AZURE_CONTAINER
   - REMOTE_STORAGE_AZURE_REGION
   - SLACK_CICD_CHANNEL_ID

.github/actions/prepare-for-subzero/action.yml

@@ -1,28 +0,0 @@
-name: 'Prepare current job for subzero'
-description: >
-  Set git token to access `neondatabase/subzero` from cargo build,
-  and set `CARGO_NET_GIT_FETCH_WITH_CLI=true` env variable to use git CLI
-
-inputs:
-  token:
-    description: 'GitHub token with access to neondatabase/subzero'
-    required: true
-
-runs:
-  using: "composite"
-
-  steps:
-    - name: Set git token for neondatabase/subzero
-      uses: pyTooling/Actions/with-post-step@2307b526df64d55e95884e072e49aac2a00a9afa # v5.1.0
-      env:
-        SUBZERO_ACCESS_TOKEN: ${{ inputs.token }}
-      with:
-        main: |
-          git config --global url."https://x-access-token:${SUBZERO_ACCESS_TOKEN}@github.com/neondatabase/subzero".insteadOf "https://github.com/neondatabase/subzero"
-          cargo add -p proxy subzero-core --git https://github.com/neondatabase/subzero --rev 396264617e78e8be428682f87469bb25429af88a
-        post: |
-          git config --global --unset url."https://x-access-token:${SUBZERO_ACCESS_TOKEN}@github.com/neondatabase/subzero".insteadOf "https://github.com/neondatabase/subzero"
-
-    - name: Set `CARGO_NET_GIT_FETCH_WITH_CLI=true` env variable
-      shell: bash -euxo pipefail {0}
-      run: echo "CARGO_NET_GIT_FETCH_WITH_CLI=true" >> ${GITHUB_ENV}

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

@@ -176,13 +176,7 @@ runs:
         fi
 
         if [[ $BUILD_TYPE == "debug" && $RUNNER_ARCH == 'X64' ]]; then
-          # We don't use code coverage for regression tests (the step is disabled),
-          # so there's no need to collect it.
-          # Ref https://github.com/neondatabase/neon/issues/4540
-          # cov_prefix=(scripts/coverage "--profraw-prefix=$GITHUB_JOB" --dir=/tmp/coverage run)
-          cov_prefix=()
-          # Explicitly set LLVM_PROFILE_FILE to /dev/null to avoid writing *.profraw files
-          export LLVM_PROFILE_FILE=/dev/null
+          cov_prefix=(scripts/coverage "--profraw-prefix=$GITHUB_JOB" --dir=/tmp/coverage run)
         else
           cov_prefix=()
         fi

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

@@ -86,10 +86,6 @@ jobs:
         with:
           submodules: true
 
-      - uses: ./.github/actions/prepare-for-subzero
-        with:
-          token: ${{ secrets.CI_ACCESS_TOKEN }}
-
       - name: Set pg 14 revision for caching
         id: pg_v14_rev
         run: echo pg_rev=$(git rev-parse HEAD:vendor/postgres-v14) >> $GITHUB_OUTPUT
@@ -120,7 +116,7 @@ jobs:
           ARCH: ${{ inputs.arch }}
           SANITIZERS: ${{ inputs.sanitizers }}
         run: |
-          CARGO_FLAGS="--locked --features testing,rest_broker"
+          CARGO_FLAGS="--locked --features testing"
           if [[ $BUILD_TYPE == "debug" && $ARCH == 'x64' ]]; then
             cov_prefix="scripts/coverage --profraw-prefix=$GITHUB_JOB --dir=/tmp/coverage run"
             CARGO_PROFILE=""
@@ -154,7 +150,7 @@ jobs:
           secretKey: ${{ secrets.HETZNER_CACHE_SECRET_KEY }}
           use-fallback: false
           path: pg_install/v14
-          key: v1-${{ runner.os }}-${{ runner.arch }}-${{ inputs.build-type }}-pg-${{ steps.pg_v14_rev.outputs.pg_rev }}-bookworm-${{ hashFiles('Makefile', 'build-tools/Dockerfile') }}
+          key: v1-${{ runner.os }}-${{ runner.arch }}-${{ inputs.build-type }}-pg-${{ steps.pg_v14_rev.outputs.pg_rev }}-bookworm-${{ hashFiles('Makefile', 'build-tools.Dockerfile') }}
 
       - name: Cache postgres v15 build
         id: cache_pg_15
@@ -166,7 +162,7 @@ jobs:
           secretKey: ${{ secrets.HETZNER_CACHE_SECRET_KEY }}
           use-fallback: false
           path: pg_install/v15
-          key: v1-${{ runner.os }}-${{ runner.arch }}-${{ inputs.build-type }}-pg-${{ steps.pg_v15_rev.outputs.pg_rev }}-bookworm-${{ hashFiles('Makefile', 'build-tools/Dockerfile') }}
+          key: v1-${{ runner.os }}-${{ runner.arch }}-${{ inputs.build-type }}-pg-${{ steps.pg_v15_rev.outputs.pg_rev }}-bookworm-${{ hashFiles('Makefile', 'build-tools.Dockerfile') }}
 
       - name: Cache postgres v16 build
         id: cache_pg_16
@@ -178,7 +174,7 @@ jobs:
           secretKey: ${{ secrets.HETZNER_CACHE_SECRET_KEY }}
           use-fallback: false
           path: pg_install/v16
-          key: v1-${{ runner.os }}-${{ runner.arch }}-${{ inputs.build-type }}-pg-${{ steps.pg_v16_rev.outputs.pg_rev }}-bookworm-${{ hashFiles('Makefile', 'build-tools/Dockerfile') }}
+          key: v1-${{ runner.os }}-${{ runner.arch }}-${{ inputs.build-type }}-pg-${{ steps.pg_v16_rev.outputs.pg_rev }}-bookworm-${{ hashFiles('Makefile', 'build-tools.Dockerfile') }}
 
       - name: Cache postgres v17 build
         id: cache_pg_17
@@ -190,7 +186,7 @@ jobs:
           secretKey: ${{ secrets.HETZNER_CACHE_SECRET_KEY }}
           use-fallback: false
           path: pg_install/v17
-          key: v1-${{ runner.os }}-${{ runner.arch }}-${{ inputs.build-type }}-pg-${{ steps.pg_v17_rev.outputs.pg_rev }}-bookworm-${{ hashFiles('Makefile', 'build-tools/Dockerfile') }}
+          key: v1-${{ runner.os }}-${{ runner.arch }}-${{ inputs.build-type }}-pg-${{ steps.pg_v17_rev.outputs.pg_rev }}-bookworm-${{ hashFiles('Makefile', 'build-tools.Dockerfile') }}
 
       - name: Build all
         # Note: the Makefile picks up BUILD_TYPE and CARGO_PROFILE from the env variables

.github/workflows/_check-codestyle-rust.yml

@@ -46,10 +46,6 @@ jobs:
         uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2
         with:
           submodules: true
-      
-      - uses: ./.github/actions/prepare-for-subzero
-        with:
-          token: ${{ secrets.CI_ACCESS_TOKEN }}
 
       - name: Cache cargo deps
         uses: tespkg/actions-cache@b7bf5fcc2f98a52ac6080eb0fd282c2f752074b1  # v1.8.0

.github/workflows/benchmarking.yml

@@ -219,7 +219,6 @@ jobs:
           --ignore test_runner/performance/test_cumulative_statistics_persistence.py
           --ignore test_runner/performance/test_perf_many_relations.py
           --ignore test_runner/performance/test_perf_oltp_large_tenant.py
-          --ignore test_runner/performance/test_lfc_prewarm.py
       env:
         BENCHMARK_CONNSTR: ${{ steps.create-neon-project.outputs.dsn }}
         VIP_VAP_ACCESS_TOKEN: "${{ secrets.VIP_VAP_ACCESS_TOKEN }}"
@@ -411,77 +410,6 @@ jobs:
       env:
         SLACK_BOT_TOKEN: ${{ secrets.SLACK_BOT_TOKEN }}
 
-  prewarm-test:
-    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 # aws-actions/configure-aws-credentials
-    env:
-      PGBENCH_SIZE: ${{ vars.PREWARM_PGBENCH_SIZE }}
-      POSTGRES_DISTRIB_DIR: /tmp/neon/pg_install
-      DEFAULT_PG_VERSION: 17
-      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: ghcr.io/neondatabase/build-tools:pinned-bookworm
-      credentials:
-        username: ${{ github.actor }}
-        password: ${{ secrets.GITHUB_TOKEN }}
-      options: --init
-
-    steps:
-    - name: Harden the runner (Audit all outbound calls)
-      uses: step-security/harden-runner@4d991eb9b905ef189e4c376166672c3f2f230481 # v2.11.0
-      with:
-        egress-policy: audit
-
-    - uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2
-
-    - name: Configure AWS credentials
-      uses: aws-actions/configure-aws-credentials@e3dd6a429d7300a6a4c196c26e071d42e0343502 # v4.0.2
-      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 }}-${{ runner.arch }}-release-artifact
-        path: /tmp/neon/
-        prefix: latest
-        aws-oidc-role-arn: ${{ vars.DEV_AWS_OIDC_ROLE_ARN }}
-
-    - name: Run prewarm benchmark
-      uses: ./.github/actions/run-python-test-set
-      with:
-        build_type: ${{ env.BUILD_TYPE }}
-        test_selection: performance/test_lfc_prewarm.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 }}
-        aws-oidc-role-arn: ${{ vars.DEV_AWS_OIDC_ROLE_ARN }}
-      env:
-        VIP_VAP_ACCESS_TOKEN: "${{ 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
-        aws-oidc-role-arn: ${{ vars.DEV_AWS_OIDC_ROLE_ARN }}
-      env:
-        REGRESS_TEST_RESULT_CONNSTR_NEW: ${{ secrets.REGRESS_TEST_RESULT_CONNSTR_NEW }}
-
   generate-matrices:
     if: ${{ github.event.inputs.run_only_pgvector_tests == 'false' || github.event.inputs.run_only_pgvector_tests == null }}
     # Create matrices for the benchmarking jobs, so we run benchmarks on rds only once a week (on Saturday)

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

@@ -72,7 +72,7 @@ jobs:
           ARCHS: ${{ inputs.archs || '["x64","arm64"]' }}
           DEBIANS: ${{ inputs.debians || '["bullseye","bookworm"]' }}
           IMAGE_TAG: |
-            ${{ hashFiles('build-tools/Dockerfile',
+            ${{ hashFiles('build-tools.Dockerfile',
                           '.github/workflows/build-build-tools-image.yml') }}
         run: |
           echo "archs=${ARCHS}"           | tee -a ${GITHUB_OUTPUT}
@@ -144,7 +144,7 @@ jobs:
 
       - uses: docker/build-push-action@471d1dc4e07e5cdedd4c2171150001c434f0b7a4 # v6.15.0
         with:
-          file: build-tools/Dockerfile
+          file: build-tools.Dockerfile
           context: .
           provenance: false
           push: true

.github/workflows/build-macos.yml

@@ -54,10 +54,6 @@ jobs:
         uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2
         with:
           submodules: true
-      
-      - uses: ./.github/actions/prepare-for-subzero
-        with:
-          token: ${{ secrets.CI_ACCESS_TOKEN }}
 
       - name: Install build dependencies
         run: |

.github/workflows/build_and_test.yml

@@ -87,27 +87,22 @@ jobs:
     uses: ./.github/workflows/build-build-tools-image.yml
     secrets: inherit
 
-  lint-yamls:
-    needs: [ meta, check-permissions, build-build-tools-image ]
+  lint-openapi-spec:
+    runs-on: ubuntu-22.04
+    needs: [ meta, check-permissions ]
     # We do need to run this in `.*-rc-pr` because of hotfixes.
     if: ${{ contains(fromJSON('["pr", "push-main", "storage-rc-pr", "proxy-rc-pr", "compute-rc-pr"]'), needs.meta.outputs.run-kind) }}
-    runs-on: [ self-hosted, small ]
-    container:
-      image: ${{ needs.build-build-tools-image.outputs.image }}
-      credentials:
-        username: ${{ github.actor }}
-        password: ${{ secrets.GITHUB_TOKEN }}
-      options: --init
-
     steps:
       - name: Harden the runner (Audit all outbound calls)
         uses: step-security/harden-runner@4d991eb9b905ef189e4c376166672c3f2f230481 # v2.11.0
         with:
           egress-policy: audit
-
       - uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2
-
-      - run: make -C compute manifest-schema-validation
+      - uses: docker/login-action@74a5d142397b4f367a81961eba4e8cd7edddf772 # v3.4.0
+        with:
+          registry: ghcr.io
+          username: ${{ github.actor }}
+          password: ${{ secrets.GITHUB_TOKEN }}
       - run: make lint-openapi-spec
 
   check-codestyle-python:
@@ -222,6 +217,28 @@ jobs:
       build-tools-image: ${{ needs.build-build-tools-image.outputs.image }}-bookworm
     secrets: inherit
 
+  validate-compute-manifest:
+    runs-on: ubuntu-22.04
+    needs: [ meta, check-permissions ]
+    # We do need to run this in `.*-rc-pr` because of hotfixes.
+    if: ${{ contains(fromJSON('["pr", "push-main", "storage-rc-pr", "proxy-rc-pr", "compute-rc-pr"]'), needs.meta.outputs.run-kind) }}
+    steps:
+      - name: Harden the runner (Audit all outbound calls)
+        uses: step-security/harden-runner@4d991eb9b905ef189e4c376166672c3f2f230481 # v2.11.0
+        with:
+          egress-policy: audit
+
+      - uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2
+
+      - name: Set up Node.js
+        uses: actions/setup-node@49933ea5288caeca8642d1e84afbd3f7d6820020 # v4.4.0
+        with:
+          node-version: '24'
+
+      - name: Validate manifest against schema
+        run: |
+          make -C compute manifest-schema-validation
+
   build-and-test-locally:
     needs: [ meta, build-build-tools-image ]
     # We do need to run this in `.*-rc-pr` because of hotfixes.
@@ -632,8 +649,6 @@ jobs:
             BUILD_TAG=${{ needs.meta.outputs.release-tag || needs.meta.outputs.build-tag }}
             TAG=${{ needs.build-build-tools-image.outputs.image-tag }}-bookworm
             DEBIAN_VERSION=bookworm
-          secrets: |
-            SUBZERO_ACCESS_TOKEN=${{ secrets.CI_ACCESS_TOKEN }}
           provenance: false
           push: true
           pull: true

.github/workflows/neon_extra_builds.yml

@@ -72,7 +72,6 @@ jobs:
   check-macos-build:
     needs: [ check-permissions, files-changed ]
     uses: ./.github/workflows/build-macos.yml
-    secrets: inherit
     with:
       pg_versions: ${{ needs.files-changed.outputs.postgres_changes }}
       rebuild_rust_code: ${{ fromJSON(needs.files-changed.outputs.rebuild_rust_code) }}

.gitignore

@@ -26,14 +26,6 @@ docker-compose/docker-compose-parallel.yml
 *.o
 *.so
 *.Po
-*.pid
 
 # pgindent typedef lists
 *.list
-
-# Node
-**/node_modules/
-
-# various files for local testing
-/proxy/.subzero
-local_proxy.json

.gitmodules

@@ -1,16 +1,16 @@
 [submodule "vendor/postgres-v14"]
 	path = vendor/postgres-v14
-	url = ../postgres.git
+	url = https://github.com/neondatabase/postgres.git
 	branch = REL_14_STABLE_neon
 [submodule "vendor/postgres-v15"]
 	path = vendor/postgres-v15
-	url = ../postgres.git
+	url = https://github.com/neondatabase/postgres.git
 	branch = REL_15_STABLE_neon
 [submodule "vendor/postgres-v16"]
 	path = vendor/postgres-v16
-	url = ../postgres.git
+	url = https://github.com/neondatabase/postgres.git
 	branch = REL_16_STABLE_neon
 [submodule "vendor/postgres-v17"]
 	path = vendor/postgres-v17
-	url = ../postgres.git
+	url = https://github.com/neondatabase/postgres.git
 	branch = REL_17_STABLE_neon

Cargo.toml

@@ -49,7 +49,6 @@ members = [
     "libs/proxy/tokio-postgres2",
     "endpoint_storage",
     "pgxn/neon/communicator",
-    "proxy/subzero_core",
 ]
 
 [workspace.package]
@@ -131,7 +130,7 @@ jemalloc_pprof = { version = "0.7", features = ["symbolize", "flamegraph"] }
 jsonwebtoken = "9"
 lasso = "0.7"
 libc = "0.2"
-lock_api = "0.4.13"
+libproc = "0.14"
 md5 = "0.7.0"
 measured = { version = "0.0.22", features=["lasso"] }
 measured-process = { version = "0.0.22" }
@@ -143,10 +142,10 @@ notify = "6.0.0"
 num_cpus = "1.15"
 num-traits = "0.2.19"
 once_cell = "1.13"
-opentelemetry = "0.30"
-opentelemetry_sdk = "0.30"
-opentelemetry-otlp = { version = "0.30", default-features = false, features = ["http-proto", "trace", "http", "reqwest-client"] }
-opentelemetry-semantic-conventions = "0.30"
+opentelemetry = "0.27"
+opentelemetry_sdk = "0.27"
+opentelemetry-otlp = { version = "0.27", default-features = false, features = ["http-proto", "trace", "http", "reqwest-client"] }
+opentelemetry-semantic-conventions = "0.27"
 parking_lot = "0.12"
 parquet = { version = "53", default-features = false, features = ["zstd"] }
 parquet_derive = "53"
@@ -158,18 +157,16 @@ procfs = "0.16"
 prometheus = {version = "0.13", default-features=false, features = ["process"]} # removes protobuf dependency
 prost = "0.13.5"
 prost-types = "0.13.5"
-rand = "0.9"
-# Remove after p256 is updated to 0.14.
-rand_core = "=0.6"
+rand = "0.8"
 redis = { version = "0.29.2", features = ["tokio-rustls-comp", "keep-alive"] }
 regex = "1.10.2"
 reqwest = { version = "0.12", default-features = false, features = ["rustls-tls"] }
-reqwest-tracing = { version = "0.5", features = ["opentelemetry_0_30"] }
+reqwest-tracing = { version = "0.5", features = ["opentelemetry_0_27"] }
 reqwest-middleware = "0.4"
 reqwest-retry = "0.7"
 routerify = "3"
 rpds = "0.13"
-rustc-hash = "2.1.1"
+rustc-hash = "1.1.0"
 rustls = { version = "0.23.16", default-features = false }
 rustls-pemfile = "2"
 rustls-pki-types = "1.11"
@@ -205,7 +202,7 @@ tokio-epoll-uring = { git = "https://github.com/neondatabase/tokio-epoll-uring.g
 tokio-io-timeout = "1.2.0"
 tokio-postgres-rustls = "0.12.0"
 tokio-rustls = { version = "0.26.0", default-features = false, features = ["tls12", "ring"]}
-tokio-stream = { version = "0.1", features = ["sync"] }
+tokio-stream = "0.1"
 tokio-tar = "0.3"
 tokio-util = { version = "0.7.10", features = ["io", "io-util", "rt"] }
 toml = "0.8"
@@ -214,12 +211,15 @@ tonic = { version = "0.13.1", default-features = false, features = ["channel", "
 tonic-reflection = { version = "0.13.1", features = ["server"] }
 tower = { version = "0.5.2", default-features = false }
 tower-http = { version = "0.6.2", features = ["auth", "request-id", "trace"] }
-tower-otel = { version = "0.6", features = ["axum"] }
+
+# This revision uses opentelemetry 0.27. There's no tag for it.
+tower-otel = { git = "https://github.com/mattiapenati/tower-otel", rev = "56a7321053bcb72443888257b622ba0d43a11fcd" }
+
 tower-service = "0.3.3"
 tracing = "0.1"
 tracing-error = "0.2"
 tracing-log = "0.2"
-tracing-opentelemetry = "0.31"
+tracing-opentelemetry = "0.28"
 tracing-serde = "0.2.0"
 tracing-subscriber = { version = "0.3", default-features = false, features = ["smallvec", "fmt", "tracing-log", "std", "env-filter", "json"] }
 try-lock = "0.2.5"
@@ -279,6 +279,7 @@ safekeeper_api = { version = "0.1", path = "./libs/safekeeper_api" }
 safekeeper_client = { path = "./safekeeper/client" }
 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" }
+tempfile = "3"
 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/" }

Dockerfile

@@ -63,14 +63,7 @@ WORKDIR /home/nonroot
 
 COPY --chown=nonroot . .
 
-RUN --mount=type=secret,uid=1000,id=SUBZERO_ACCESS_TOKEN \
-    set -e \
-    && if [ -s /run/secrets/SUBZERO_ACCESS_TOKEN ]; then \
-        export CARGO_NET_GIT_FETCH_WITH_CLI=true && \
-        git config --global url."https://$(cat /run/secrets/SUBZERO_ACCESS_TOKEN)@github.com/neondatabase/subzero".insteadOf "https://github.com/neondatabase/subzero" && \
-        cargo add -p proxy subzero-core --git https://github.com/neondatabase/subzero --rev 396264617e78e8be428682f87469bb25429af88a; \
-    fi \
-    && cargo chef prepare --recipe-path recipe.json
+RUN cargo chef prepare --recipe-path recipe.json
 
 # Main build image
 FROM $REPOSITORY/$IMAGE:$TAG AS build
@@ -78,33 +71,20 @@ WORKDIR /home/nonroot
 ARG GIT_VERSION=local
 ARG BUILD_TAG
 ARG ADDITIONAL_RUSTFLAGS=""
-ENV CARGO_FEATURES="default"
 
 # 3. Build cargo dependencies. Note that this step doesn't depend on anything else than
 # `recipe.json`, so the layer can be reused as long as none of the dependencies change.
 COPY --from=plan     /home/nonroot/recipe.json                              recipe.json
-RUN --mount=type=secret,uid=1000,id=SUBZERO_ACCESS_TOKEN \
-    set -e \
-    && if [ -s /run/secrets/SUBZERO_ACCESS_TOKEN ]; then \
-        export CARGO_NET_GIT_FETCH_WITH_CLI=true && \
-        git config --global url."https://$(cat /run/secrets/SUBZERO_ACCESS_TOKEN)@github.com/neondatabase/subzero".insteadOf "https://github.com/neondatabase/subzero"; \
-    fi \
+RUN set -e \
     && RUSTFLAGS="-Clinker=clang -Clink-arg=-fuse-ld=mold -Clink-arg=-Wl,--no-rosegment -Cforce-frame-pointers=yes ${ADDITIONAL_RUSTFLAGS}" cargo chef cook --locked --release --recipe-path recipe.json
 
 # Perform the main build. We reuse the Postgres build artifacts from the intermediate 'pg-build'
 # layer, and the cargo dependencies built in the previous step.
 COPY --chown=nonroot --from=pg-build /home/nonroot/pg_install/ pg_install
 COPY --chown=nonroot . .
-COPY --chown=nonroot --from=plan     /home/nonroot/proxy/Cargo.toml         proxy/Cargo.toml
-COPY --chown=nonroot --from=plan     /home/nonroot/Cargo.lock               Cargo.lock
 
-RUN  --mount=type=secret,uid=1000,id=SUBZERO_ACCESS_TOKEN \
-    set -e \
-    && if [ -s /run/secrets/SUBZERO_ACCESS_TOKEN ]; then \
-        export CARGO_FEATURES="rest_broker"; \
-    fi \
+RUN set -e \
     && RUSTFLAGS="-Clinker=clang -Clink-arg=-fuse-ld=mold -Clink-arg=-Wl,--no-rosegment -Cforce-frame-pointers=yes ${ADDITIONAL_RUSTFLAGS}" cargo build \
-      --features $CARGO_FEATURES \
       --bin pg_sni_router  \
       --bin pageserver  \
       --bin pagectl  \
@@ -129,6 +109,8 @@ RUN set -e \
         libreadline-dev \
         libseccomp-dev \
         ca-certificates \
+        bpfcc-tools \
+        sudo \
         openssl \
         unzip \
         curl \

Makefile

@@ -2,7 +2,7 @@ ROOT_PROJECT_DIR := $(dir $(abspath $(lastword $(MAKEFILE_LIST))))
 
 # Where to install Postgres, default is ./pg_install, maybe useful for package
 # managers.
-POSTGRES_INSTALL_DIR ?= $(ROOT_PROJECT_DIR)/pg_install
+POSTGRES_INSTALL_DIR ?= $(ROOT_PROJECT_DIR)/pg_install/
 
 # Supported PostgreSQL versions
 POSTGRES_VERSIONS = v17 v16 v15 v14
@@ -14,7 +14,7 @@ POSTGRES_VERSIONS = v17 v16 v15 v14
 # it is derived from BUILD_TYPE.
 
 # All intermediate build artifacts are stored here.
-BUILD_DIR := $(ROOT_PROJECT_DIR)/build
+BUILD_DIR := build
 
 ICU_PREFIX_DIR := /usr/local/icu
 
@@ -212,7 +212,7 @@ neon-pgindent: postgres-v17-pg-bsd-indent neon-pg-ext-v17
 		FIND_TYPEDEF=$(ROOT_PROJECT_DIR)/vendor/postgres-v17/src/tools/find_typedef \
 		INDENT=$(BUILD_DIR)/v17/src/tools/pg_bsd_indent/pg_bsd_indent \
 		PGINDENT_SCRIPT=$(ROOT_PROJECT_DIR)/vendor/postgres-v17/src/tools/pgindent/pgindent \
-		-C $(BUILD_DIR)/pgxn-v17/neon \
+		-C $(BUILD_DIR)/neon-v17 \
 		-f $(ROOT_PROJECT_DIR)/pgxn/neon/Makefile pgindent
 
 
@@ -220,15 +220,11 @@ neon-pgindent: postgres-v17-pg-bsd-indent neon-pg-ext-v17
 setup-pre-commit-hook:
 	ln -s -f $(ROOT_PROJECT_DIR)/pre-commit.py .git/hooks/pre-commit
 
-build-tools/node_modules: build-tools/package.json
-	cd build-tools && $(if $(CI),npm ci,npm install)
-	touch build-tools/node_modules
-
 .PHONY: lint-openapi-spec
-lint-openapi-spec: build-tools/node_modules
+lint-openapi-spec:
 	# operation-2xx-response: pageserver timeline delete returns 404 on success
 	find . -iname "openapi_spec.y*ml" -exec\
-		npx --prefix=build-tools/ redocly\
+		docker run --rm -v ${PWD}:/spec ghcr.io/redocly/cli:1.34.4\
 			--skip-rule=operation-operationId --skip-rule=operation-summary --extends=minimal\
 			--skip-rule=no-server-example.com --skip-rule=operation-2xx-response\
 			lint {} \+

build-tools.Dockerfile

@@ -35,7 +35,7 @@ RUN echo 'Acquire::Retries "5";' > /etc/apt/apt.conf.d/80-retries && \
     echo -e "retry_connrefused=on\ntimeout=15\ntries=5\nretry-on-host-error=on\n" > /root/.wgetrc && \
     echo -e "--retry-connrefused\n--connect-timeout 15\n--retry 5\n--max-time 300\n" > /root/.curlrc
 
-COPY build-tools/patches/pgcopydbv017.patch /pgcopydbv017.patch
+COPY build_tools/patches/pgcopydbv017.patch /pgcopydbv017.patch
 
 RUN if [ "${DEBIAN_VERSION}" = "bookworm" ]; then \
         set -e && \
@@ -61,6 +61,9 @@ RUN if [ "${DEBIAN_VERSION}" = "bookworm" ]; then \
         libpq5 \
         libpq-dev \
         libzstd-dev \
+        linux-perf \
+        bpfcc-tools \
+        linux-headers-$(case "$(uname -m)" in x86_64) echo amd64;; aarch64) echo arm64;; esac) \
         postgresql-16 \
         postgresql-server-dev-16 \
         postgresql-common  \
@@ -105,15 +108,21 @@ RUN echo 'Acquire::Retries "5";' > /etc/apt/apt.conf.d/80-retries && \
 #
 # 'gdb' is included so that we get backtraces of core dumps produced in
 # regression tests
-RUN set -e \
+RUN set -ex \
+    && KERNEL_VERSION="$(uname -r | cut -d'-' -f1 | sed 's/\.0$//')" \
+    && echo KERNEL_VERSION=${KERNEL_VERSION} >> /etc/environment \
+    && KERNEL_ARCH=$(uname -m | awk '{ if ($1 ~ /^(x86_64|i[3-6]86)$/) print "x86"; else if ($1 ~ /^(aarch64|arm.*)$/) print "aarch"; else print $1 }') \
+    && echo KERNEL_ARCH=${KERNEL_ARCH} >> /etc/environment \
     && apt update \
     && apt install -y \
         autoconf \
         automake \
+        bc \
         bison \
         build-essential \
         ca-certificates \
         cmake \
+        cpio \
         curl \
         flex \
         gdb \
@@ -122,8 +131,10 @@ RUN set -e \
         gzip \
         jq \
         jsonnet \
+        kmod \
         libcurl4-openssl-dev \
         libbz2-dev \
+        libelf-dev \
         libffi-dev \
         liblzma-dev \
         libncurses5-dev \
@@ -137,6 +148,11 @@ RUN set -e \
         libxml2-dev \
         libxmlsec1-dev \
         libxxhash-dev \
+        linux-perf \
+        bpfcc-tools \
+        libbpfcc \
+        libbpfcc-dev \
+        linux-headers-$(case "$(uname -m)" in x86_64) echo amd64;; aarch64) echo arm64;; esac) \
         lsof \
         make \
         netcat-openbsd \
@@ -144,6 +160,8 @@ RUN set -e \
         openssh-client \
         parallel \
         pkg-config \
+        rsync \
+        sudo \
         unzip \
         wget \
         xz-utils \
@@ -188,12 +206,6 @@ RUN curl -fsSL 'https://apt.llvm.org/llvm-snapshot.gpg.key' | apt-key add - \
     && 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/*
 
-# Install node
-ENV NODE_VERSION=24
-RUN curl -fsSL https://deb.nodesource.com/setup_${NODE_VERSION}.x | bash - \
-    && apt install -y nodejs \
-    && rm -rf /var/lib/apt/lists/* /tmp/* /var/tmp/*
-
 # Install docker
 RUN curl -fsSL https://download.docker.com/linux/ubuntu/gpg | gpg --dearmor -o /usr/share/keyrings/docker-archive-keyring.gpg \
     && echo "deb [arch=$(dpkg --print-architecture) signed-by=/usr/share/keyrings/docker-archive-keyring.gpg] https://download.docker.com/linux/debian ${DEBIAN_VERSION} stable" > /etc/apt/sources.list.d/docker.list \
@@ -204,6 +216,8 @@ RUN curl -fsSL https://download.docker.com/linux/ubuntu/gpg | gpg --dearmor -o /
 # Configure sudo & docker
 RUN usermod -aG sudo nonroot && \
     echo '%sudo ALL=(ALL) NOPASSWD:ALL' >> /etc/sudoers && \
+    mkdir -p /etc/sudoers.d && \
+    echo 'nonroot ALL=(ALL) NOPASSWD:ALL' > /etc/sudoers.d/nonroot && \
     usermod -aG docker nonroot
 
 # AWS CLI
@@ -317,14 +331,14 @@ 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 rustfmt clippy && \
-    cargo install rustfilt      --locked --version ${RUSTFILT_VERSION} && \
-    cargo install cargo-hakari  --locked --version ${CARGO_HAKARI_VERSION} && \
-    cargo install cargo-deny    --locked --version ${CARGO_DENY_VERSION} && \
-    cargo install cargo-hack    --locked --version ${CARGO_HACK_VERSION} && \
-    cargo install cargo-nextest --locked --version ${CARGO_NEXTEST_VERSION} && \
-    cargo install cargo-chef    --locked --version ${CARGO_CHEF_VERSION} && \
-    cargo install diesel_cli    --locked --version ${CARGO_DIESEL_CLI_VERSION} \
-                                --features postgres-bundled --no-default-features && \
+    cargo install rustfilt            --version ${RUSTFILT_VERSION} --locked && \
+    cargo install cargo-hakari        --version ${CARGO_HAKARI_VERSION} --locked && \
+    cargo install cargo-deny          --version ${CARGO_DENY_VERSION} --locked && \
+    cargo install cargo-hack          --version ${CARGO_HACK_VERSION} --locked && \
+    cargo install cargo-nextest       --version ${CARGO_NEXTEST_VERSION} --locked && \
+    cargo install cargo-chef          --version ${CARGO_CHEF_VERSION} --locked && \
+    cargo install diesel_cli          --version ${CARGO_DIESEL_CLI_VERSION} --locked \
+                                      --features postgres-bundled --no-default-features && \
     rm -rf /home/nonroot/.cargo/registry && \
     rm -rf /home/nonroot/.cargo/git
 

build-tools/package.json

@@ -1,8 +0,0 @@
-{
-  "name": "build-tools",
-  "private": true,
-  "devDependencies": {
-    "@redocly/cli": "1.34.4",
-    "@sourcemeta/jsonschema": "10.0.0"
-  }
-}

compute/Makefile

@@ -50,9 +50,9 @@ jsonnetfmt-format:
 	jsonnetfmt --in-place $(jsonnet_files)
 
 .PHONY: manifest-schema-validation
-manifest-schema-validation: ../build-tools/node_modules
-	npx --prefix=../build-tools/ jsonschema validate -d https://json-schema.org/draft/2020-12/schema manifest.schema.json manifest.yaml
+manifest-schema-validation: node_modules
+	node_modules/.bin/jsonschema validate -d https://json-schema.org/draft/2020-12/schema manifest.schema.json manifest.yaml
 
-../build-tools/node_modules: ../build-tools/package.json
-	cd ../build-tools && $(if $(CI),npm ci,npm install)
-	touch ../build-tools/node_modules
+node_modules: package.json
+	npm install
+	touch node_modules

compute/compute-node.Dockerfile

@@ -9,7 +9,7 @@
 #
 # build-tools:   This contains Rust compiler toolchain and other tools needed at compile
 #                time. This is also used for the storage builds. This image is defined in
-#                build-tools/Dockerfile.
+#                build-tools.Dockerfile.
 #
 # build-deps:    Contains C compiler, other build tools, and compile-time dependencies
 #                needed to compile PostgreSQL and most extensions. (Some extensions need
@@ -115,7 +115,7 @@ ARG EXTENSIONS=all
 FROM $BASE_IMAGE_SHA AS build-deps
 ARG DEBIAN_VERSION
 
-# Keep in sync with build-tools/Dockerfile
+# Keep in sync with build-tools.Dockerfile
 ENV PROTOC_VERSION=25.1
 
 # Use strict mode for bash to catch errors early
@@ -149,6 +149,9 @@ RUN case $DEBIAN_VERSION in \
     ninja-build git autoconf automake libtool build-essential bison flex libreadline-dev \
     zlib1g-dev libxml2-dev libcurl4-openssl-dev libossp-uuid-dev wget ca-certificates pkg-config libssl-dev \
     libicu-dev libxslt1-dev liblz4-dev libzstd-dev zstd curl unzip g++ \
+    bpfcc-tools \
+    libbpfcc \
+    libbpfcc-dev \
     libclang-dev \
     jsonnet \
     $VERSION_INSTALLS \
@@ -170,29 +173,7 @@ RUN case $DEBIAN_VERSION in \
 FROM build-deps AS pg-build
 ARG PG_VERSION
 COPY vendor/postgres-${PG_VERSION:?} postgres
-COPY compute/patches/postgres_fdw.patch .
-COPY compute/patches/pg_stat_statements_pg14-16.patch .
-COPY compute/patches/pg_stat_statements_pg17.patch .
 RUN cd postgres && \
-    # Apply patches to some contrib extensions
-    # For example, we need to grant EXECUTE on pg_stat_statements_reset() to {privileged_role_name}.
-    # In vanilla Postgres this function is limited to Postgres role superuser.
-    # In Neon we have {privileged_role_name} role that is not a superuser but replaces superuser in some cases.
-    # We could add the additional grant statements to the Postgres repository but it would be hard to maintain,
-    # whenever we need to pick up a new Postgres version and we want to limit the changes in our Postgres fork,
-    # so we do it here.
-    case "${PG_VERSION}" in \
-    "v14" | "v15" | "v16") \
-    patch -p1 < /pg_stat_statements_pg14-16.patch; \
-    ;; \
-    "v17") \
-    patch -p1 < /pg_stat_statements_pg17.patch; \
-    ;; \
-    *) \
-    # To do not forget to migrate patches to the next major version
-    echo "No contrib patches for this PostgreSQL version" && exit 1;; \
-    esac && \
-    patch -p1 < /postgres_fdw.patch && \
     export CONFIGURE_CMD="./configure CFLAGS='-O2 -g3 -fsigned-char' --enable-debug --with-openssl --with-uuid=ossp \
     --with-icu --with-libxml --with-libxslt --with-lz4" && \
     if [ "${PG_VERSION:?}" != "v14" ]; then \
@@ -206,6 +187,8 @@ RUN cd postgres && \
     echo 'trusted = true' >> /usr/local/pgsql/share/extension/autoinc.control && \
     echo 'trusted = true' >> /usr/local/pgsql/share/extension/dblink.control && \
     echo 'trusted = true' >> /usr/local/pgsql/share/extension/postgres_fdw.control && \
+    file=/usr/local/pgsql/share/extension/postgres_fdw--1.0.sql && [ -e $file ] && \
+    echo 'GRANT USAGE ON FOREIGN DATA WRAPPER postgres_fdw TO neon_superuser;' >> $file && \
     echo 'trusted = true' >> /usr/local/pgsql/share/extension/bloom.control && \
     echo 'trusted = true' >> /usr/local/pgsql/share/extension/earthdistance.control && \
     echo 'trusted = true' >> /usr/local/pgsql/share/extension/insert_username.control && \
@@ -215,7 +198,34 @@ RUN cd postgres && \
     echo 'trusted = true' >> /usr/local/pgsql/share/extension/pgrowlocks.control && \
     echo 'trusted = true' >> /usr/local/pgsql/share/extension/pgstattuple.control && \
     echo 'trusted = true' >> /usr/local/pgsql/share/extension/refint.control && \
-    echo 'trusted = true' >> /usr/local/pgsql/share/extension/xml2.control
+    echo 'trusted = true' >> /usr/local/pgsql/share/extension/xml2.control && \
+    # We need to grant EXECUTE on pg_stat_statements_reset() to neon_superuser.
+    # In vanilla postgres this function is limited to Postgres role superuser.
+    # In neon we have neon_superuser role that is not a superuser but replaces superuser in some cases.
+    # We could add the additional grant statements to the postgres repository but it would be hard to maintain,
+    # whenever we need to pick up a new postgres version and we want to limit the changes in our postgres fork,
+    # so we do it here.
+    for file in /usr/local/pgsql/share/extension/pg_stat_statements--*.sql; do \
+        filename=$(basename "$file"); \
+        # Note that there are no downgrade scripts for pg_stat_statements, so we \
+        # don't have to modify any downgrade paths or (much) older versions: we only \
+        # have to make sure every creation of the pg_stat_statements_reset function \
+        # also adds execute permissions to the neon_superuser.
+        case $filename in \
+          pg_stat_statements--1.4.sql) \
+            # pg_stat_statements_reset is first created with 1.4
+            echo 'GRANT EXECUTE ON FUNCTION pg_stat_statements_reset() TO neon_superuser;' >> $file; \
+            ;; \
+          pg_stat_statements--1.6--1.7.sql) \
+            # Then with the 1.6-1.7 migration it is re-created with a new signature, thus add the permissions back
+            echo 'GRANT EXECUTE ON FUNCTION pg_stat_statements_reset(Oid, Oid, bigint) TO neon_superuser;' >> $file; \
+            ;; \
+          pg_stat_statements--1.10--1.11.sql) \
+            # Then with the 1.10-1.11 migration it is re-created with a new signature again, thus add the permissions back
+            echo 'GRANT EXECUTE ON FUNCTION pg_stat_statements_reset(Oid, Oid, bigint, boolean) TO neon_superuser;' >> $file; \
+            ;; \
+        esac; \
+    done;
 
 # Set PATH for all the subsequent build steps
 ENV PATH="/usr/local/pgsql/bin:$PATH"
@@ -1517,7 +1527,7 @@ WORKDIR /ext-src
 COPY compute/patches/pg_duckdb_v031.patch .
 COPY compute/patches/duckdb_v120.patch .
 # pg_duckdb build requires source dir to be a git repo to get submodules
-# allow {privileged_role_name} to execute some functions that in pg_duckdb are available to superuser only:
+# allow neon_superuser to execute some functions that in pg_duckdb are available to superuser only:
 # - extension management function duckdb.install_extension()
 # - access to duckdb.extensions table and its sequence
 RUN git clone --depth 1 --branch v0.3.1 https://github.com/duckdb/pg_duckdb.git pg_duckdb-src && \
@@ -1783,7 +1793,7 @@ RUN set -e \
 #########################################################################################
 FROM build-deps AS exporters
 ARG TARGETARCH
-# Keep sql_exporter version same as in build-tools/Dockerfile and
+# Keep sql_exporter version same as in build-tools.Dockerfile and
 # test_runner/regress/test_compute_metrics.py
 # See comment on the top of the file regading `echo`, `-e` and `\n`
 RUN if [ "$TARGETARCH" = "amd64" ]; then\
@@ -1981,6 +1991,10 @@ RUN apt update && \
         locales \
         lsof \
         procps \
+        bpfcc-tools \
+        libbpfcc \
+        libbpfcc-dev \
+        libclang-dev \
         rsyslog-gnutls \
         screen \
         tcpdump \

compute/package.json

@@ -0,0 +1,7 @@
+{
+  "name": "neon-compute",
+  "private": true,
+  "dependencies": {
+    "@sourcemeta/jsonschema": "9.3.4"
+  }
+} 

compute/patches/anon_v2.patch

@@ -1,26 +1,22 @@
 diff --git a/sql/anon.sql b/sql/anon.sql
-index 0cdc769..5eab1d6 100644
+index 0cdc769..b450327 100644
 --- a/sql/anon.sql
 +++ b/sql/anon.sql
-@@ -1141,3 +1141,19 @@ $$
+@@ -1141,3 +1141,15 @@ $$
  -- TODO : https://en.wikipedia.org/wiki/L-diversity
  
  -- TODO : https://en.wikipedia.org/wiki/T-closeness
 +
 +-- NEON Patches
 +
++GRANT ALL ON SCHEMA anon to neon_superuser;
++GRANT ALL ON ALL TABLES IN SCHEMA anon TO neon_superuser;
++
 +DO $$
-+DECLARE
-+  privileged_role_name text;
 +BEGIN
-+  privileged_role_name := current_setting('neon.privileged_role_name');
-+
-+  EXECUTE format('GRANT ALL ON SCHEMA anon to %I', privileged_role_name);
-+  EXECUTE format('GRANT ALL ON ALL TABLES IN SCHEMA anon TO %I', privileged_role_name);
-+
-+  IF current_setting('server_version_num')::int >= 150000 THEN
-+    EXECUTE format('GRANT SET ON PARAMETER anon.transparent_dynamic_masking TO %I', privileged_role_name);
-+  END IF;
++    IF current_setting('server_version_num')::int >= 150000 THEN
++        GRANT SET ON PARAMETER anon.transparent_dynamic_masking TO neon_superuser;
++    END IF;
 +END $$;
 diff --git a/sql/init.sql b/sql/init.sql
 index 7da6553..9b6164b 100644

compute/patches/pg_duckdb_v031.patch

@@ -21,21 +21,13 @@ index 3235cc8..6b892bc 100644
  include Makefile.global
  
 diff --git a/sql/pg_duckdb--0.2.0--0.3.0.sql b/sql/pg_duckdb--0.2.0--0.3.0.sql
-index d777d76..3b54396 100644
+index d777d76..af60106 100644
 --- a/sql/pg_duckdb--0.2.0--0.3.0.sql
 +++ b/sql/pg_duckdb--0.2.0--0.3.0.sql
-@@ -1056,3 +1056,14 @@ GRANT ALL ON FUNCTION duckdb.cache(TEXT, TEXT) TO PUBLIC;
+@@ -1056,3 +1056,6 @@ GRANT ALL ON FUNCTION duckdb.cache(TEXT, TEXT) TO PUBLIC;
  GRANT ALL ON FUNCTION duckdb.cache_info() TO PUBLIC;
  GRANT ALL ON FUNCTION duckdb.cache_delete(TEXT) TO PUBLIC;
  GRANT ALL ON PROCEDURE duckdb.recycle_ddb() TO PUBLIC;
-+
-+DO $$
-+DECLARE
-+  privileged_role_name text;
-+BEGIN
-+  privileged_role_name := current_setting('neon.privileged_role_name');
-+
-+  EXECUTE format('GRANT ALL ON FUNCTION duckdb.install_extension(TEXT) TO %I', privileged_role_name);
-+  EXECUTE format('GRANT ALL ON TABLE duckdb.extensions TO %I', privileged_role_name);
-+  EXECUTE format('GRANT ALL ON SEQUENCE duckdb.extensions_table_seq TO %I', privileged_role_name);
-+END $$;
++GRANT ALL ON FUNCTION duckdb.install_extension(TEXT) TO neon_superuser;
++GRANT ALL ON TABLE duckdb.extensions TO neon_superuser;
++GRANT ALL ON SEQUENCE duckdb.extensions_table_seq TO neon_superuser;

compute/patches/pg_stat_statements_pg14-16.patch

@@ -1,34 +0,0 @@
-diff --git a/contrib/pg_stat_statements/pg_stat_statements--1.4.sql b/contrib/pg_stat_statements/pg_stat_statements--1.4.sql
-index 58cdf600fce..8be57a996f6 100644
---- a/contrib/pg_stat_statements/pg_stat_statements--1.4.sql
-+++ b/contrib/pg_stat_statements/pg_stat_statements--1.4.sql
-@@ -46,3 +46,12 @@ GRANT SELECT ON pg_stat_statements TO PUBLIC;
- 
- -- Don't want this to be available to non-superusers.
- REVOKE ALL ON FUNCTION pg_stat_statements_reset() FROM PUBLIC;
-+
-+DO $$
-+DECLARE
-+  privileged_role_name text;
-+BEGIN
-+  privileged_role_name := current_setting('neon.privileged_role_name');
-+
-+  EXECUTE format('GRANT EXECUTE ON FUNCTION pg_stat_statements_reset() TO %I', privileged_role_name);
-+END $$;
-diff --git a/contrib/pg_stat_statements/pg_stat_statements--1.6--1.7.sql b/contrib/pg_stat_statements/pg_stat_statements--1.6--1.7.sql
-index 6fc3fed4c93..256345a8f79 100644
---- a/contrib/pg_stat_statements/pg_stat_statements--1.6--1.7.sql
-+++ b/contrib/pg_stat_statements/pg_stat_statements--1.6--1.7.sql
-@@ -20,3 +20,12 @@ LANGUAGE C STRICT PARALLEL SAFE;
- 
- -- Don't want this to be available to non-superusers.
- REVOKE ALL ON FUNCTION pg_stat_statements_reset(Oid, Oid, bigint) FROM PUBLIC;
-+
-+DO $$
-+DECLARE
-+  privileged_role_name text;
-+BEGIN
-+  privileged_role_name := current_setting('neon.privileged_role_name');
-+
-+  EXECUTE format('GRANT EXECUTE ON FUNCTION pg_stat_statements_reset(Oid, Oid, bigint) TO %I', privileged_role_name);
-+END $$;

compute/patches/pg_stat_statements_pg17.patch

@@ -1,52 +0,0 @@
-diff --git a/contrib/pg_stat_statements/pg_stat_statements--1.10--1.11.sql b/contrib/pg_stat_statements/pg_stat_statements--1.10--1.11.sql
-index 0bb2c397711..32764db1d8b 100644
---- a/contrib/pg_stat_statements/pg_stat_statements--1.10--1.11.sql
-+++ b/contrib/pg_stat_statements/pg_stat_statements--1.10--1.11.sql
-@@ -80,3 +80,12 @@ LANGUAGE C STRICT PARALLEL SAFE;
- 
- -- Don't want this to be available to non-superusers.
- REVOKE ALL ON FUNCTION pg_stat_statements_reset(Oid, Oid, bigint, boolean) FROM PUBLIC;
-+
-+DO $$
-+DECLARE
-+  privileged_role_name text;
-+BEGIN
-+  privileged_role_name := current_setting('neon.privileged_role_name');
-+
-+  EXECUTE format('GRANT EXECUTE ON FUNCTION pg_stat_statements_reset(Oid, Oid, bigint, boolean) TO %I', privileged_role_name);
-+END $$;
-\ No newline at end of file
-diff --git a/contrib/pg_stat_statements/pg_stat_statements--1.4.sql b/contrib/pg_stat_statements/pg_stat_statements--1.4.sql
-index 58cdf600fce..8be57a996f6 100644
---- a/contrib/pg_stat_statements/pg_stat_statements--1.4.sql
-+++ b/contrib/pg_stat_statements/pg_stat_statements--1.4.sql
-@@ -46,3 +46,12 @@ GRANT SELECT ON pg_stat_statements TO PUBLIC;
- 
- -- Don't want this to be available to non-superusers.
- REVOKE ALL ON FUNCTION pg_stat_statements_reset() FROM PUBLIC;
-+
-+DO $$
-+DECLARE
-+  privileged_role_name text;
-+BEGIN
-+  privileged_role_name := current_setting('neon.privileged_role_name');
-+
-+  EXECUTE format('GRANT EXECUTE ON FUNCTION pg_stat_statements_reset() TO %I', privileged_role_name);
-+END $$;
-diff --git a/contrib/pg_stat_statements/pg_stat_statements--1.6--1.7.sql b/contrib/pg_stat_statements/pg_stat_statements--1.6--1.7.sql
-index 6fc3fed4c93..256345a8f79 100644
---- a/contrib/pg_stat_statements/pg_stat_statements--1.6--1.7.sql
-+++ b/contrib/pg_stat_statements/pg_stat_statements--1.6--1.7.sql
-@@ -20,3 +20,12 @@ LANGUAGE C STRICT PARALLEL SAFE;
- 
- -- Don't want this to be available to non-superusers.
- REVOKE ALL ON FUNCTION pg_stat_statements_reset(Oid, Oid, bigint) FROM PUBLIC;
-+
-+DO $$
-+DECLARE
-+  privileged_role_name text;
-+BEGIN
-+  privileged_role_name := current_setting('neon.privileged_role_name');
-+
-+  EXECUTE format('GRANT EXECUTE ON FUNCTION pg_stat_statements_reset(Oid, Oid, bigint) TO %I', privileged_role_name);
-+END $$;

compute/patches/postgres_fdw.patch

@@ -1,17 +0,0 @@
-diff --git a/contrib/postgres_fdw/postgres_fdw--1.0.sql b/contrib/postgres_fdw/postgres_fdw--1.0.sql
-index a0f0fc1bf45..ee077f2eea6 100644
---- a/contrib/postgres_fdw/postgres_fdw--1.0.sql
-+++ b/contrib/postgres_fdw/postgres_fdw--1.0.sql
-@@ -16,3 +16,12 @@ LANGUAGE C STRICT;
- CREATE FOREIGN DATA WRAPPER postgres_fdw
-   HANDLER postgres_fdw_handler
-   VALIDATOR postgres_fdw_validator;
-+
-+DO $$
-+DECLARE
-+  privileged_role_name text;
-+BEGIN
-+  privileged_role_name := current_setting('neon.privileged_role_name');
-+
-+  EXECUTE format('GRANT USAGE ON FOREIGN DATA WRAPPER postgres_fdw TO %I', privileged_role_name);
-+END $$;

compute/vm-image-spec-bookworm.yaml

@@ -39,6 +39,14 @@ commands:
     user: nobody
     sysvInitAction: respawn
     shell: '/bin/sql_exporter -config.file=/etc/sql_exporter_autoscaling.yml -web.listen-address=:9499'
+  - name: enable-kernel-modules
+    user: root
+    sysvInitAction: sysinit
+    shell: mkdir -p /lib/ && ln -s /neonvm/tools/lib/modules /lib/
+  - name: enable-bpfs
+    user: root
+    sysvInitAction: sysinit
+    shell: mkdir -p /sys/kernel/debug && mount -t debugfs debugfs /sys/kernel/debug && mount -t bpf bpf /sys/fs/bpf && chmod 755 /sys/fs/bpf
   # Rsyslog by default creates a unix socket under /dev/log . That's where Postgres sends logs also.
   # We run syslog with postgres user so it can't create /dev/log. Instead we configure rsyslog to
   # use a different path for the socket. The symlink actually points to our custom path.
@@ -65,7 +73,7 @@ files:
       # regardless of hostname (ALL)
       #
       # Also allow it to shut down the VM. The fast_import job does that when it's finished.
-      postgres ALL=(root) NOPASSWD: /neonvm/bin/resize-swap, /neonvm/bin/set-disk-quota, /neonvm/bin/poweroff, /usr/sbin/rsyslogd
+      postgres ALL=(root) NOPASSWD: /neonvm/bin/resize-swap, /neonvm/bin/set-disk-quota, /neonvm/bin/poweroff, /usr/sbin/rsyslogd, /neonvm/tools/bin/perf, /usr/sbin/profile-bpfcc
   - filename: cgconfig.conf
     content: |
       # Configuration for cgroups in VM compute nodes
@@ -152,6 +160,8 @@ merge: |
   RUN set -e \
       && chmod 0644 /etc/cgconfig.conf
 
+  ENV PERF_BINARY_PATH=/neonvm/tools/bin/perf
+
 
   COPY compute_rsyslog.conf /etc/compute_rsyslog.conf
   RUN chmod 0666 /etc/compute_rsyslog.conf

compute/vm-image-spec-bullseye.yaml

@@ -39,6 +39,14 @@ commands:
     user: nobody
     sysvInitAction: respawn
     shell: '/bin/sql_exporter -config.file=/etc/sql_exporter_autoscaling.yml -web.listen-address=:9499'
+  - name: enable-kernel-modules
+    user: root
+    sysvInitAction: sysinit
+    shell: mkdir -p /lib/ && ln -s /neonvm/tools/lib/modules /lib/
+  - name: enable-bpfs
+    user: root
+    sysvInitAction: sysinit
+    shell: mkdir -p /sys/kernel/debug && mount -t debugfs debugfs /sys/kernel/debug && mount -t bpf bpf /sys/fs/bpf && chmod 755 /sys/fs/bpf
   # Rsyslog by default creates a unix socket under /dev/log . That's where Postgres sends logs also.
   # We run syslog with postgres user so it can't create /dev/log. Instead we configure rsyslog to
   # use a different path for the socket. The symlink actually points to our custom path.
@@ -65,7 +73,7 @@ files:
       # regardless of hostname (ALL)
       #
       # Also allow it to shut down the VM. The fast_import job does that when it's finished.
-      postgres ALL=(root) NOPASSWD: /neonvm/bin/resize-swap, /neonvm/bin/set-disk-quota, /neonvm/bin/poweroff, /usr/sbin/rsyslogd
+      postgres ALL=(root) NOPASSWD: /neonvm/bin/resize-swap, /neonvm/bin/set-disk-quota, /neonvm/bin/poweroff, /usr/sbin/rsyslogd, /neonvm/tools/bin/perf, /usr/sbin/profile-bpfcc
   - filename: cgconfig.conf
     content: |
       # Configuration for cgroups in VM compute nodes
@@ -148,6 +156,8 @@ merge: |
   RUN set -e \
       && chmod 0644 /etc/cgconfig.conf
 
+  ENV PERF_BINARY_PATH=/neonvm/tools/bin/perf
+
   COPY compute_rsyslog.conf /etc/compute_rsyslog.conf
   RUN chmod 0666 /etc/compute_rsyslog.conf
   RUN mkdir /var/log/rsyslog && chown -R postgres /var/log/rsyslog

compute_tools/Cargo.toml

@@ -27,13 +27,11 @@ fail.workspace = true
 flate2.workspace = true
 futures.workspace = true
 http.workspace = true
-http-body-util.workspace = true
 hostname-validator = "1.1"
-hyper.workspace = true
-hyper-util.workspace = true
 indexmap.workspace = true
 itertools.workspace = true
 jsonwebtoken.workspace = true
+libproc.workspace = true
 metrics.workspace = true
 nix.workspace = true
 notify.workspace = true
@@ -47,12 +45,12 @@ postgres.workspace = true
 regex.workspace = true
 reqwest = { workspace = true, features = ["json"] }
 ring = "0.17"
-scopeguard.workspace = true
 serde.workspace = true
 serde_with.workspace = true
 serde_json.workspace = true
 signal-hook.workspace = true
 tar.workspace = true
+tempfile.workspace = true
 tower.workspace = true
 tower-http.workspace = true
 tokio = { workspace = true, features = ["rt", "rt-multi-thread"] }
@@ -82,3 +80,10 @@ zstd = "0.13"
 bytes = "1.0"
 rust-ini = "0.20.0"
 rlimit = "0.10.1"
+
+inferno = { version = "0.12", default-features = false, features = [
+    "multithreaded",
+    "nameattr",
+] }
+pprof = { version = "0.15", features = ["protobuf-codec", "flamegraph"] }
+prost = "0.12"

compute_tools/src/bin/compute_ctl.rs

@@ -87,14 +87,6 @@ struct Cli {
     #[arg(short = 'C', long, value_name = "DATABASE_URL")]
     pub connstr: String,
 
-    #[arg(
-        long,
-        default_value = "neon_superuser",
-        value_name = "PRIVILEGED_ROLE_NAME",
-        value_parser = Self::parse_privileged_role_name
-    )]
-    pub privileged_role_name: String,
-
     #[cfg(target_os = "linux")]
     #[arg(long, default_value = "neon-postgres")]
     pub cgroup: String,
@@ -138,12 +130,6 @@ struct Cli {
     /// Run in development mode, skipping VM-specific operations like process termination
     #[arg(long, action = clap::ArgAction::SetTrue)]
     pub dev: bool,
-
-    #[arg(long)]
-    pub pg_init_timeout: Option<u64>,
-
-    #[arg(long, default_value_t = false, action = clap::ArgAction::Set)]
-    pub lakebase_mode: bool,
 }
 
 impl Cli {
@@ -163,21 +149,6 @@ impl Cli {
 
         Ok(url)
     }
-
-    /// For simplicity, we do not escape `privileged_role_name` anywhere in the code.
-    /// Since it's a system role, which we fully control, that's fine. Still, let's
-    /// validate it to avoid any surprises.
-    fn parse_privileged_role_name(value: &str) -> Result<String> {
-        use regex::Regex;
-
-        let pattern = Regex::new(r"^[a-z_]+$").unwrap();
-
-        if !pattern.is_match(value) {
-            bail!("--privileged-role-name can only contain lowercase letters and underscores")
-        }
-
-        Ok(value.to_string())
-    }
 }
 
 fn main() -> Result<()> {
@@ -194,7 +165,7 @@ fn main() -> Result<()> {
         .build()?;
     let _rt_guard = runtime.enter();
 
-    let tracing_provider = init(cli.dev)?;
+    runtime.block_on(init(cli.dev))?;
 
     // enable core dumping for all child processes
     setrlimit(Resource::CORE, rlimit::INFINITY, rlimit::INFINITY)?;
@@ -207,7 +178,6 @@ fn main() -> Result<()> {
         ComputeNodeParams {
             compute_id: cli.compute_id,
             connstr,
-            privileged_role_name: cli.privileged_role_name.clone(),
             pgdata: cli.pgdata.clone(),
             pgbin: cli.pgbin.clone(),
             pgversion: get_pg_version_string(&cli.pgbin),
@@ -225,8 +195,6 @@ fn main() -> Result<()> {
             installed_extensions_collection_interval: Arc::new(AtomicU64::new(
                 cli.installed_extensions_collection_interval,
             )),
-            pg_init_timeout: cli.pg_init_timeout.map(Duration::from_secs),
-            lakebase_mode: cli.lakebase_mode,
         },
         config,
     )?;
@@ -235,11 +203,11 @@ fn main() -> Result<()> {
 
     scenario.teardown();
 
-    deinit_and_exit(tracing_provider, exit_code);
+    deinit_and_exit(exit_code);
 }
 
-fn init(dev_mode: bool) -> Result<Option<tracing_utils::Provider>> {
-    let provider = init_tracing_and_logging(DEFAULT_LOG_LEVEL)?;
+async fn init(dev_mode: bool) -> Result<()> {
+    init_tracing_and_logging(DEFAULT_LOG_LEVEL).await?;
 
     let mut signals = Signals::new([SIGINT, SIGTERM, SIGQUIT])?;
     thread::spawn(move || {
@@ -250,7 +218,7 @@ fn init(dev_mode: bool) -> Result<Option<tracing_utils::Provider>> {
 
     info!("compute build_tag: {}", &BUILD_TAG.to_string());
 
-    Ok(provider)
+    Ok(())
 }
 
 fn get_config(cli: &Cli) -> Result<ComputeConfig> {
@@ -275,27 +243,25 @@ fn get_config(cli: &Cli) -> Result<ComputeConfig> {
     }
 }
 
-fn deinit_and_exit(tracing_provider: Option<tracing_utils::Provider>, exit_code: Option<i32>) -> ! {
-    if let Some(p) = tracing_provider {
-        // Shutdown trace pipeline gracefully, so that it has a chance to send any
-        // pending traces before we exit. Shutting down OTEL tracing provider may
-        // hang for quite some time, see, for example:
-        // - https://github.com/open-telemetry/opentelemetry-rust/issues/868
-        // - and our problems with staging https://github.com/neondatabase/cloud/issues/3707#issuecomment-1493983636
-        //
-        // Yet, we want computes to shut down fast enough, as we may need a new one
-        // for the same timeline ASAP. So wait no longer than 2s for the shutdown to
-        // complete, then just error out and exit the main thread.
-        info!("shutting down tracing");
-        let (sender, receiver) = mpsc::channel();
-        let _ = thread::spawn(move || {
-            _ = p.shutdown();
-            sender.send(()).ok()
-        });
-        let shutdown_res = receiver.recv_timeout(Duration::from_millis(2000));
-        if shutdown_res.is_err() {
-            error!("timed out while shutting down tracing, exiting anyway");
-        }
+fn deinit_and_exit(exit_code: Option<i32>) -> ! {
+    // Shutdown trace pipeline gracefully, so that it has a chance to send any
+    // pending traces before we exit. Shutting down OTEL tracing provider may
+    // hang for quite some time, see, for example:
+    // - https://github.com/open-telemetry/opentelemetry-rust/issues/868
+    // - and our problems with staging https://github.com/neondatabase/cloud/issues/3707#issuecomment-1493983636
+    //
+    // Yet, we want computes to shut down fast enough, as we may need a new one
+    // for the same timeline ASAP. So wait no longer than 2s for the shutdown to
+    // complete, then just error out and exit the main thread.
+    info!("shutting down tracing");
+    let (sender, receiver) = mpsc::channel();
+    let _ = thread::spawn(move || {
+        tracing_utils::shutdown_tracing();
+        sender.send(()).ok()
+    });
+    let shutdown_res = receiver.recv_timeout(Duration::from_millis(2000));
+    if shutdown_res.is_err() {
+        error!("timed out while shutting down tracing, exiting anyway");
     }
 
     info!("shutting down");
@@ -361,49 +327,4 @@ mod test {
         ])
         .expect_err("URL parameters are not allowed");
     }
-
-    #[test]
-    fn verify_privileged_role_name() {
-        // Valid name
-        let cli = Cli::parse_from([
-            "compute_ctl",
-            "--pgdata=test",
-            "--connstr=test",
-            "--compute-id=test",
-            "--privileged-role-name",
-            "my_superuser",
-        ]);
-        assert_eq!(cli.privileged_role_name, "my_superuser");
-
-        // Invalid names
-        Cli::try_parse_from([
-            "compute_ctl",
-            "--pgdata=test",
-            "--connstr=test",
-            "--compute-id=test",
-            "--privileged-role-name",
-            "NeonSuperuser",
-        ])
-        .expect_err("uppercase letters are not allowed");
-
-        Cli::try_parse_from([
-            "compute_ctl",
-            "--pgdata=test",
-            "--connstr=test",
-            "--compute-id=test",
-            "--privileged-role-name",
-            "$'neon_superuser",
-        ])
-        .expect_err("special characters are not allowed");
-
-        Cli::try_parse_from([
-            "compute_ctl",
-            "--pgdata=test",
-            "--connstr=test",
-            "--compute-id=test",
-            "--privileged-role-name",
-            "",
-        ])
-        .expect_err("empty name is not allowed");
-    }
 }

compute_tools/src/communicator_socket_client.rs

@@ -1,98 +0,0 @@
-//! Client for making request to a running Postgres server's communicator control socket.
-//!
-//! The storage communicator process that runs inside Postgres exposes an HTTP endpoint in
-//! a Unix Domain Socket in the Postgres data directory. This provides access to it.
-
-use std::path::Path;
-
-use anyhow::Context;
-use hyper::client::conn::http1::SendRequest;
-use hyper_util::rt::TokioIo;
-
-/// Name of the socket within the Postgres data directory. This better match that in
-/// `pgxn/neon/communicator/src/lib.rs`.
-const NEON_COMMUNICATOR_SOCKET_NAME: &str = "neon-communicator.socket";
-
-/// Open a connection to the communicator's control socket, prepare to send requests to it
-/// with hyper.
-pub async fn connect_communicator_socket<B>(pgdata: &Path) -> anyhow::Result<SendRequest<B>>
-where
-    B: hyper::body::Body + 'static + Send,
-    B::Data: Send,
-    B::Error: Into<Box<dyn std::error::Error + Send + Sync>>,
-{
-    let socket_path = pgdata.join(NEON_COMMUNICATOR_SOCKET_NAME);
-    let socket_path_len = socket_path.display().to_string().len();
-
-    // There is a limit of around 100 bytes (108 on Linux?) on the length of the path to a
-    // Unix Domain socket. The limit is on the connect(2) function used to open the
-    // socket, not on the absolute path itself. Postgres changes the current directory to
-    // the data directory and uses a relative path to bind to the socket, and the relative
-    // path "./neon-communicator.socket" is always short, but when compute_ctl needs to
-    // open the socket, we need to use a full path, which can be arbitrarily long.
-    //
-    // There are a few ways we could work around this:
-    //
-    // 1. Change the current directory to the Postgres data directory and use a relative
-    //    path in the connect(2) call. That's problematic because the current directory
-    //    applies to the whole process. We could change the current directory early in
-    //    compute_ctl startup, and that might be a good idea anyway for other reasons too:
-    //    it would be more robust if the data directory is moved around or unlinked for
-    //    some reason, and you would be less likely to accidentally litter other parts of
-    //    the filesystem with e.g. temporary files. However, that's a pretty invasive
-    //    change.
-    //
-    // 2. On Linux, you could open() the data directory, and refer to the the socket
-    //    inside it as "/proc/self/fd/<fd>/neon-communicator.socket". But that's
-    //    Linux-only.
-    //
-    // 3. Create a symbolic link to the socket with a shorter path, and use that.
-    //
-    // We use the symbolic link approach here. Hopefully the paths we use in production
-    // are shorter, so that we can open the socket directly, so that this hack is needed
-    // only in development.
-    let connect_result = if socket_path_len < 100 {
-        // We can open the path directly with no hacks.
-        tokio::net::UnixStream::connect(socket_path).await
-    } else {
-        // The path to the socket is too long. Create a symlink to it with a shorter path.
-        let short_path = std::env::temp_dir().join(format!(
-            "compute_ctl.short-socket.{}.{}",
-            std::process::id(),
-            tokio::task::id()
-        ));
-        std::os::unix::fs::symlink(&socket_path, &short_path)?;
-
-        // Delete the symlink as soon as we have connected to it. There's a small chance
-        // of leaking if the process dies before we remove it, so try to keep that window
-        // as small as possible.
-        scopeguard::defer! {
-            if let Err(err) = std::fs::remove_file(&short_path) {
-                tracing::warn!("could not remove symlink \"{}\" created for socket: {}",
-                               short_path.display(), err);
-            }
-        }
-
-        tracing::info!(
-            "created symlink \"{}\" for socket \"{}\", opening it now",
-            short_path.display(),
-            socket_path.display()
-        );
-
-        tokio::net::UnixStream::connect(&short_path).await
-    };
-
-    let stream = connect_result.context("connecting to communicator control socket")?;
-
-    let io = TokioIo::new(stream);
-    let (request_sender, connection) = hyper::client::conn::http1::handshake(io).await?;
-
-    // spawn a task to poll the connection and drive the HTTP state
-    tokio::spawn(async move {
-        if let Err(err) = connection.await {
-            eprintln!("Error in connection: {err}");
-        }
-    });
-
-    Ok(request_sender)
-}

compute_tools/src/compute.rs

@@ -74,20 +74,12 @@ const DEFAULT_INSTALLED_EXTENSIONS_COLLECTION_INTERVAL: u64 = 3600;
 
 /// Static configuration params that don't change after startup. These mostly
 /// come from the CLI args, or are derived from them.
-#[derive(Clone, Debug)]
 pub struct ComputeNodeParams {
     /// The ID of the compute
     pub compute_id: String,
-
-    /// Url type maintains proper escaping
+    // Url type maintains proper escaping
     pub connstr: url::Url,
 
-    /// The name of the 'weak' superuser role, which we give to the users.
-    /// It follows the allow list approach, i.e., we take a standard role
-    /// and grant it extra permissions with explicit GRANTs here and there,
-    /// and core patches.
-    pub privileged_role_name: String,
-
     pub resize_swap_on_bind: bool,
     pub set_disk_quota_for_fs: Option<String>,
 
@@ -113,11 +105,6 @@ pub struct ComputeNodeParams {
 
     /// Interval for installed extensions collection
     pub installed_extensions_collection_interval: Arc<AtomicU64>,
-
-    /// Timeout of PG compute startup in the Init state.
-    pub pg_init_timeout: Option<Duration>,
-
-    pub lakebase_mode: bool,
 }
 
 type TaskHandle = Mutex<Option<JoinHandle<()>>>;
@@ -159,7 +146,6 @@ pub struct RemoteExtensionMetrics {
 #[derive(Clone, Debug)]
 pub struct ComputeState {
     pub start_time: DateTime<Utc>,
-    pub pg_start_time: Option<DateTime<Utc>>,
     pub status: ComputeStatus,
     /// Timestamp of the last Postgres activity. It could be `None` if
     /// compute wasn't used since start.
@@ -197,7 +183,6 @@ impl ComputeState {
     pub fn new() -> Self {
         Self {
             start_time: Utc::now(),
-            pg_start_time: None,
             status: ComputeStatus::Empty,
             last_active: None,
             error: None,
@@ -386,7 +371,9 @@ fn maybe_cgexec(cmd: &str) -> Command {
     }
 }
 
-struct PostgresHandle {
+/// A handle to the Postgres process that is running in the compute
+/// node.
+pub struct PostgresHandle {
     postgres: std::process::Child,
     log_collector: JoinHandle<Result<()>>,
 }
@@ -655,9 +642,6 @@ impl ComputeNode {
             };
             _this_entered = start_compute_span.enter();
 
-            // Hadron: Record postgres start time (used to enforce pg_init_timeout).
-            state_guard.pg_start_time.replace(Utc::now());
-
             state_guard.set_status(ComputeStatus::Init, &self.state_changed);
             compute_state = state_guard.clone()
         }
@@ -1058,8 +1042,6 @@ impl ComputeNode {
             PageserverProtocol::Grpc => self.try_get_basebackup_grpc(spec, lsn)?,
         };
 
-        self.fix_zenith_signal_neon_signal()?;
-
         let mut state = self.state.lock().unwrap();
         state.metrics.pageserver_connect_micros =
             connected.duration_since(started).as_micros() as u64;
@@ -1069,27 +1051,6 @@ impl ComputeNode {
         Ok(())
     }
 
-    /// Move the Zenith signal file to Neon signal file location.
-    /// This makes Compute compatible with older PageServers that don't yet
-    /// know about the Zenith->Neon rename.
-    fn fix_zenith_signal_neon_signal(&self) -> Result<()> {
-        let datadir = Path::new(&self.params.pgdata);
-
-        let neonsig = datadir.join("neon.signal");
-
-        if neonsig.is_file() {
-            return Ok(());
-        }
-
-        let zenithsig = datadir.join("zenith.signal");
-
-        if zenithsig.is_file() {
-            fs::copy(zenithsig, neonsig)?;
-        }
-
-        Ok(())
-    }
-
     /// Fetches a basebackup via gRPC. The connstring must use grpc://. Returns the timestamp when
     /// the connection was established, and the (compressed) size of the basebackup.
     fn try_get_basebackup_grpc(&self, spec: &ParsedSpec, lsn: Lsn) -> Result<(Instant, usize)> {
@@ -1304,7 +1265,9 @@ impl ComputeNode {
 
         // In case of error, log and fail the check, but don't crash.
         // We're playing it safe because these errors could be transient
-        // and we don't yet retry.
+        // and we don't yet retry. Also being careful here allows us to
+        // be backwards compatible with safekeepers that don't have the
+        // TIMELINE_STATUS API yet.
         if responses.len() < quorum {
             error!(
                 "failed sync safekeepers check {:?} {:?} {:?}",
@@ -1407,7 +1370,6 @@ impl ComputeNode {
         self.create_pgdata()?;
         config::write_postgres_conf(
             pgdata_path,
-            &self.params,
             &pspec.spec,
             self.params.internal_http_port,
             tls_config,
@@ -1451,7 +1413,7 @@ impl ComputeNode {
         })?;
 
         // Update pg_hba.conf received with basebackup.
-        update_pg_hba(pgdata_path, None)?;
+        update_pg_hba(pgdata_path)?;
 
         // Place pg_dynshmem under /dev/shm. This allows us to use
         // 'dynamic_shared_memory_type = mmap' so that the files are placed in
@@ -1756,8 +1718,6 @@ impl ComputeNode {
         }
 
         // Run migrations separately to not hold up cold starts
-        let lakebase_mode = self.params.lakebase_mode;
-        let params = self.params.clone();
         tokio::spawn(async move {
             let mut conf = conf.as_ref().clone();
             conf.application_name("compute_ctl:migrations");
@@ -1769,7 +1729,7 @@ impl ComputeNode {
                             eprintln!("connection error: {e}");
                         }
                     });
-                    if let Err(e) = handle_migrations(params, &mut client, lakebase_mode).await {
+                    if let Err(e) = handle_migrations(&mut client).await {
                         error!("Failed to run migrations: {}", e);
                     }
                 }
@@ -1848,7 +1808,6 @@ impl ComputeNode {
         let pgdata_path = Path::new(&self.params.pgdata);
         config::write_postgres_conf(
             pgdata_path,
-            &self.params,
             &spec,
             self.params.internal_http_port,
             tls_config,
@@ -2461,31 +2420,14 @@ LIMIT 100",
     pub fn spawn_lfc_offload_task(self: &Arc<Self>, interval: Duration) {
         self.terminate_lfc_offload_task();
         let secs = interval.as_secs();
+        info!("spawning lfc offload worker with {secs}s interval");
         let this = self.clone();
-
-        info!("spawning LFC offload worker with {secs}s interval");
         let handle = spawn(async move {
             let mut interval = time::interval(interval);
             interval.tick().await; // returns immediately
             loop {
                 interval.tick().await;
-
-                let prewarm_state = this.state.lock().unwrap().lfc_prewarm_state.clone();
-                // Do not offload LFC state if we are currently prewarming or any issue occurred.
-                // If we'd do that, we might override the LFC state in endpoint storage with some
-                // incomplete state. Imagine a situation:
-                // 1. Endpoint started with `autoprewarm: true`
-                // 2. While prewarming is not completed, we upload the new incomplete state
-                // 3. Compute gets interrupted and restarts
-                // 4. We start again and try to prewarm with the state from 2. instead of the previous complete state
-                if matches!(
-                    prewarm_state,
-                    LfcPrewarmState::Completed
-                        | LfcPrewarmState::NotPrewarmed
-                        | LfcPrewarmState::Skipped
-                ) {
-                    this.offload_lfc_async().await;
-                }
+                this.offload_lfc_async().await;
             }
         });
         *self.lfc_offload_task.lock().unwrap() = Some(handle);
@@ -2524,7 +2466,7 @@ pub async fn installed_extensions(conf: tokio_postgres::Config) -> Result<()> {
                 serde_json::to_string(&extensions).expect("failed to serialize extensions list")
             );
         }
-        Err(err) => error!("could not get installed extensions: {err}"),
+        Err(err) => error!("could not get installed extensions: {err:?}"),
     }
     Ok(())
 }

compute_tools/src/compute_prewarm.rs

@@ -89,7 +89,7 @@ impl ComputeNode {
         self.state.lock().unwrap().lfc_offload_state.clone()
     }
 
-    /// If there is a prewarm request ongoing, return `false`, `true` otherwise.
+    /// If there is a prewarm request ongoing, return false, true otherwise
     pub fn prewarm_lfc(self: &Arc<Self>, from_endpoint: Option<String>) -> bool {
         {
             let state = &mut self.state.lock().unwrap().lfc_prewarm_state;
@@ -101,25 +101,15 @@ impl ComputeNode {
 
         let cloned = self.clone();
         spawn(async move {
-            let state = match cloned.prewarm_impl(from_endpoint).await {
-                Ok(true) => LfcPrewarmState::Completed,
-                Ok(false) => {
-                    info!(
-                        "skipping LFC prewarm because LFC state is not found in endpoint storage"
-                    );
-                    LfcPrewarmState::Skipped
-                }
-                Err(err) => {
-                    crate::metrics::LFC_PREWARM_ERRORS.inc();
-                    error!(%err, "could not prewarm LFC");
-
-                    LfcPrewarmState::Failed {
-                        error: err.to_string(),
-                    }
-                }
+            let Err(err) = cloned.prewarm_impl(from_endpoint).await else {
+                cloned.state.lock().unwrap().lfc_prewarm_state = LfcPrewarmState::Completed;
+                return;
+            };
+            crate::metrics::LFC_PREWARM_ERRORS.inc();
+            error!(%err, "prewarming lfc");
+            cloned.state.lock().unwrap().lfc_prewarm_state = LfcPrewarmState::Failed {
+                error: err.to_string(),
             };
-
-            cloned.state.lock().unwrap().lfc_prewarm_state = state;
         });
         true
     }
@@ -130,21 +120,15 @@ impl ComputeNode {
         EndpointStoragePair::from_spec_and_endpoint(state.pspec.as_ref().unwrap(), from_endpoint)
     }
 
-    /// Request LFC state from endpoint storage and load corresponding pages into Postgres.
-    /// Returns a result with `false` if the LFC state is not found in endpoint storage.
-    async fn prewarm_impl(&self, from_endpoint: Option<String>) -> Result<bool> {
+    async fn prewarm_impl(&self, from_endpoint: Option<String>) -> Result<()> {
         let EndpointStoragePair { url, token } = self.endpoint_storage_pair(from_endpoint)?;
-
         info!(%url, "requesting LFC state from endpoint storage");
+
         let request = Client::new().get(&url).bearer_auth(token);
         let res = request.send().await.context("querying endpoint storage")?;
         let status = res.status();
-        match status {
-            StatusCode::OK => (),
-            StatusCode::NOT_FOUND => {
-                return Ok(false);
-            }
-            _ => bail!("{status} querying endpoint storage"),
+        if status != StatusCode::OK {
+            bail!("{status} querying endpoint storage")
         }
 
         let mut uncompressed = Vec::new();
@@ -157,8 +141,7 @@ impl ComputeNode {
             .await
             .context("decoding LFC state")?;
         let uncompressed_len = uncompressed.len();
-
-        info!(%url, "downloaded LFC state, uncompressed size {uncompressed_len}, loading into Postgres");
+        info!(%url, "downloaded LFC state, uncompressed size {uncompressed_len}, loading into postgres");
 
         ComputeNode::get_maintenance_client(&self.tokio_conn_conf)
             .await
@@ -166,9 +149,7 @@ impl ComputeNode {
             .query_one("select neon.prewarm_local_cache($1)", &[&uncompressed])
             .await
             .context("loading LFC state into postgres")
-            .map(|_| ())?;
-
-        Ok(true)
+            .map(|_| ())
     }
 
     /// If offload request is ongoing, return false, true otherwise
@@ -196,14 +177,12 @@ impl ComputeNode {
 
     async fn offload_lfc_with_state_update(&self) {
         crate::metrics::LFC_OFFLOADS.inc();
-
         let Err(err) = self.offload_lfc_impl().await else {
             self.state.lock().unwrap().lfc_offload_state = LfcOffloadState::Completed;
             return;
         };
-
         crate::metrics::LFC_OFFLOAD_ERRORS.inc();
-        error!(%err, "could not offload LFC state to endpoint storage");
+        error!(%err, "offloading lfc");
         self.state.lock().unwrap().lfc_offload_state = LfcOffloadState::Failed {
             error: err.to_string(),
         };
@@ -211,7 +190,7 @@ impl ComputeNode {
 
     async fn offload_lfc_impl(&self) -> Result<()> {
         let EndpointStoragePair { url, token } = self.endpoint_storage_pair(None)?;
-        info!(%url, "requesting LFC state from Postgres");
+        info!(%url, "requesting LFC state from postgres");
 
         let mut compressed = Vec::new();
         ComputeNode::get_maintenance_client(&self.tokio_conn_conf)
@@ -226,17 +205,13 @@ impl ComputeNode {
             .read_to_end(&mut compressed)
             .await
             .context("compressing LFC state")?;
-
         let compressed_len = compressed.len();
         info!(%url, "downloaded LFC state, compressed size {compressed_len}, writing to endpoint storage");
 
         let request = Client::new().put(url).bearer_auth(token).body(compressed);
         match request.send().await {
             Ok(res) if res.status() == StatusCode::OK => Ok(()),
-            Ok(res) => bail!(
-                "Request to endpoint storage failed with status: {}",
-                res.status()
-            ),
+            Ok(res) => bail!("Error writing to endpoint storage: {}", res.status()),
             Err(err) => Err(err).context("writing to endpoint storage"),
         }
     }

compute_tools/src/config.rs

@@ -9,7 +9,6 @@ use std::path::Path;
 use compute_api::responses::TlsConfig;
 use compute_api::spec::{ComputeAudit, ComputeMode, ComputeSpec, GenericOption};
 
-use crate::compute::ComputeNodeParams;
 use crate::pg_helpers::{
     GenericOptionExt, GenericOptionsSearch, PgOptionsSerialize, escape_conf_value,
 };
@@ -42,7 +41,6 @@ pub fn line_in_file(path: &Path, line: &str) -> Result<bool> {
 /// Create or completely rewrite configuration file specified by `path`
 pub fn write_postgres_conf(
     pgdata_path: &Path,
-    params: &ComputeNodeParams,
     spec: &ComputeSpec,
     extension_server_port: u16,
     tls_config: &Option<TlsConfig>,
@@ -56,15 +54,14 @@ pub fn write_postgres_conf(
         writeln!(file, "{conf}")?;
     }
 
-    // Stripe size GUC should be defined prior to connection string
-    if let Some(stripe_size) = spec.shard_stripe_size {
-        writeln!(file, "neon.stripe_size={stripe_size}")?;
-    }
     // Add options for connecting to storage
     writeln!(file, "# Neon storage settings")?;
     if let Some(s) = &spec.pageserver_connstring {
         writeln!(file, "neon.pageserver_connstring={}", escape_conf_value(s))?;
     }
+    if let Some(stripe_size) = spec.shard_stripe_size {
+        writeln!(file, "neon.stripe_size={stripe_size}")?;
+    }
     if !spec.safekeeper_connstrings.is_empty() {
         let mut neon_safekeepers_value = String::new();
         tracing::info!(
@@ -164,12 +161,6 @@ pub fn write_postgres_conf(
         }
     }
 
-    writeln!(
-        file,
-        "neon.privileged_role_name={}",
-        escape_conf_value(params.privileged_role_name.as_str())
-    )?;
-
     // If there are any extra options in the 'settings' field, append those
     if spec.cluster.settings.is_some() {
         writeln!(file, "# Managed by compute_ctl: begin")?;

compute_tools/src/hadron_metrics.rs

@@ -1,60 +0,0 @@
-use metrics::{
-    IntCounter, IntGaugeVec, core::Collector, proto::MetricFamily, register_int_counter,
-    register_int_gauge_vec,
-};
-use once_cell::sync::Lazy;
-
-// Counter keeping track of the number of PageStream request errors reported by Postgres.
-// An error is registered every time Postgres calls compute_ctl's /refresh_configuration API.
-// Postgres will invoke this API if it detected trouble with PageStream requests (get_page@lsn,
-// get_base_backup, etc.) it sends to any pageserver. An increase in this counter value typically
-// indicates Postgres downtime, as PageStream requests are critical for Postgres to function.
-pub static POSTGRES_PAGESTREAM_REQUEST_ERRORS: Lazy<IntCounter> = Lazy::new(|| {
-    register_int_counter!(
-        "pg_cctl_pagestream_request_errors_total",
-        "Number of PageStream request errors reported by the postgres process"
-    )
-    .expect("failed to define a metric")
-});
-
-// Counter keeping track of the number of compute configuration errors due to Postgres statement
-// timeouts. An error is registered every time `ComputeNode::reconfigure()` fails due to Postgres
-// error code 57014 (query cancelled). This statement timeout typically occurs when postgres is
-// stuck in a problematic retry loop when the PS is reject its connection requests (usually due
-// to PG pointing at the wrong PS). We should investigate the root cause when this counter value
-// increases by checking PG and PS logs.
-pub static COMPUTE_CONFIGURE_STATEMENT_TIMEOUT_ERRORS: Lazy<IntCounter> = Lazy::new(|| {
-    register_int_counter!(
-        "pg_cctl_configure_statement_timeout_errors_total",
-        "Number of compute configuration errors due to Postgres statement timeouts."
-    )
-    .expect("failed to define a metric")
-});
-
-pub static COMPUTE_ATTACHED: Lazy<IntGaugeVec> = Lazy::new(|| {
-    register_int_gauge_vec!(
-        "pg_cctl_attached",
-        "Compute node attached status (1 if attached)",
-        &[
-            "pg_compute_id",
-            "pg_instance_id",
-            "tenant_id",
-            "timeline_id"
-        ]
-    )
-    .expect("failed to define a metric")
-});
-
-pub fn collect() -> Vec<MetricFamily> {
-    let mut metrics = Vec::new();
-    metrics.extend(POSTGRES_PAGESTREAM_REQUEST_ERRORS.collect());
-    metrics.extend(COMPUTE_CONFIGURE_STATEMENT_TIMEOUT_ERRORS.collect());
-    metrics.extend(COMPUTE_ATTACHED.collect());
-    metrics
-}
-
-pub fn initialize_metrics() {
-    Lazy::force(&POSTGRES_PAGESTREAM_REQUEST_ERRORS);
-    Lazy::force(&COMPUTE_CONFIGURE_STATEMENT_TIMEOUT_ERRORS);
-    Lazy::force(&COMPUTE_ATTACHED);
-}

compute_tools/src/http/openapi_spec.yaml

@@ -613,11 +613,11 @@ components:
         - skipped
       properties:
         status:
-          description: LFC prewarm status
-          enum: [not_prewarmed, prewarming, completed, failed, skipped]
+          description: Lfc prewarm status
+          enum: [not_prewarmed, prewarming, completed, failed]
           type: string
         error:
-          description: LFC prewarm error, if any
+          description: Lfc prewarm error, if any
           type: string
         total:
           description: Total pages processed
@@ -635,11 +635,11 @@ components:
         - status
       properties:
         status:
-          description: LFC offload status
+          description: Lfc offload status
           enum: [not_offloaded, offloading, completed, failed]
           type: string
         error:
-          description: LFC offload error, if any
+          description: Lfc offload error, if any
           type: string
 
     PromoteState:

compute_tools/src/http/routes/metrics.rs

@@ -1,18 +1,10 @@
-use std::path::Path;
-use std::sync::Arc;
-
-use anyhow::Context;
 use axum::body::Body;
-use axum::extract::State;
 use axum::response::Response;
+use http::StatusCode;
 use http::header::CONTENT_TYPE;
-use http_body_util::BodyExt;
-use hyper::{Request, StatusCode};
 use metrics::proto::MetricFamily;
 use metrics::{Encoder, TextEncoder};
 
-use crate::communicator_socket_client::connect_communicator_socket;
-use crate::compute::ComputeNode;
 use crate::http::JsonResponse;
 use crate::metrics::collect;
 
@@ -39,42 +31,3 @@ pub(in crate::http) async fn get_metrics() -> Response {
         .body(Body::from(buffer))
         .unwrap()
 }
-
-/// Fetch and forward metrics from the Postgres neon extension's metrics
-/// exporter that are used by autoscaling-agent.
-///
-/// The neon extension exposes these metrics over a Unix domain socket
-/// in the data directory. That's not accessible directly from the outside
-/// world, so we have this endpoint in compute_ctl to expose it
-pub(in crate::http) async fn get_autoscaling_metrics(
-    State(compute): State<Arc<ComputeNode>>,
-) -> Result<Response, Response> {
-    let pgdata = Path::new(&compute.params.pgdata);
-
-    // Connect to the communicator process's metrics socket
-    let mut metrics_client = connect_communicator_socket(pgdata)
-        .await
-        .map_err(|e| JsonResponse::error(StatusCode::INTERNAL_SERVER_ERROR, format!("{e:#}")))?;
-
-    // Make a request for /autoscaling_metrics
-    let request = Request::builder()
-        .method("GET")
-        .uri("/autoscaling_metrics")
-        .header("Host", "localhost") // hyper requires Host, even though the server won't care
-        .body(Body::from(""))
-        .unwrap();
-    let resp = metrics_client
-        .send_request(request)
-        .await
-        .context("fetching metrics from Postgres metrics service")
-        .map_err(|e| JsonResponse::error(StatusCode::INTERNAL_SERVER_ERROR, format!("{e:#}")))?;
-
-    // Build a response that just forwards the response we got.
-    let mut response = Response::builder();
-    response = response.status(resp.status());
-    if let Some(content_type) = resp.headers().get(CONTENT_TYPE) {
-        response = response.header(CONTENT_TYPE, content_type);
-    }
-    let body = tonic::service::AxumBody::from_stream(resp.into_body().into_data_stream());
-    Ok(response.body(body).unwrap())
-}

compute_tools/src/http/routes/mod.rs

@@ -15,6 +15,7 @@ pub(in crate::http) mod lfc;
 pub(in crate::http) mod metrics;
 pub(in crate::http) mod metrics_json;
 pub(in crate::http) mod promote;
+pub(in crate::http) mod profile;
 pub(in crate::http) mod status;
 pub(in crate::http) mod terminate;
 

compute_tools/src/http/routes/profile.rs

@@ -0,0 +1,217 @@
+//! Contains the route for profiling the compute.
+//!
+//! Profiling the compute means generating a pprof profile of the
+//! postgres processes.
+//!
+//! The profiling is done using the `perf` tool, which is expected to be
+//! available somewhere in `$PATH`.
+use std::sync::atomic::Ordering;
+
+use axum::Json;
+use axum::response::IntoResponse;
+use http::StatusCode;
+use nix::unistd::Pid;
+use once_cell::sync::Lazy;
+use tokio::sync::Mutex;
+
+use crate::http::JsonResponse;
+
+static CANCEL_CHANNEL: Lazy<Mutex<Option<tokio::sync::broadcast::Sender<()>>>> =
+    Lazy::new(|| Mutex::new(None));
+
+fn default_sampling_frequency() -> u16 {
+    100
+}
+
+fn default_timeout_seconds() -> u8 {
+    5
+}
+
+fn deserialize_sampling_frequency<'de, D>(deserializer: D) -> Result<u16, D::Error>
+where
+    D: serde::Deserializer<'de>,
+{
+    use serde::Deserialize;
+
+    const MIN_SAMPLING_FREQUENCY: u16 = 1;
+    const MAX_SAMPLING_FREQUENCY: u16 = 1000;
+
+    let value = u16::deserialize(deserializer)?;
+
+    if !(MIN_SAMPLING_FREQUENCY..=MAX_SAMPLING_FREQUENCY).contains(&value) {
+        return Err(serde::de::Error::custom(format!(
+            "sampling_frequency must be between {MIN_SAMPLING_FREQUENCY} and {MAX_SAMPLING_FREQUENCY}, got {value}"
+        )));
+    }
+    Ok(value)
+}
+
+fn deserialize_profiling_timeout<'de, D>(deserializer: D) -> Result<u8, D::Error>
+where
+    D: serde::Deserializer<'de>,
+{
+    use serde::Deserialize;
+
+    const MIN_TIMEOUT_SECONDS: u8 = 1;
+    const MAX_TIMEOUT_SECONDS: u8 = 60;
+
+    let value = u8::deserialize(deserializer)?;
+
+    if !(MIN_TIMEOUT_SECONDS..=MAX_TIMEOUT_SECONDS).contains(&value) {
+        return Err(serde::de::Error::custom(format!(
+            "timeout_seconds must be between {MIN_TIMEOUT_SECONDS} and {MAX_TIMEOUT_SECONDS}, got {value}"
+        )));
+    }
+    Ok(value)
+}
+
+/// Request parameters for profiling the compute.
+#[derive(Debug, Clone, serde::Deserialize)]
+pub(in crate::http) struct ProfileRequest {
+    /// The profiling tool to use, currently only `perf` is supported.
+    profiler: crate::profiling::ProfileGenerator,
+    #[serde(default = "default_sampling_frequency")]
+    #[serde(deserialize_with = "deserialize_sampling_frequency")]
+    sampling_frequency: u16,
+    #[serde(default = "default_timeout_seconds")]
+    #[serde(deserialize_with = "deserialize_profiling_timeout")]
+    timeout_seconds: u8,
+    #[serde(default)]
+    archive: bool,
+}
+
+/// The HTTP request handler for reporting the profiling status of
+/// the compute.
+pub(in crate::http) async fn profile_status() -> impl IntoResponse {
+    tracing::info!("Profile status request received.");
+
+    let cancel_channel = CANCEL_CHANNEL.lock().await;
+
+    if let Some(tx) = cancel_channel.as_ref() {
+        if tx.receiver_count() > 0 {
+            return JsonResponse::create_response(
+                StatusCode::OK,
+                "Profiling is currently in progress.",
+            );
+        }
+    }
+
+    JsonResponse::create_response(StatusCode::NO_CONTENT, "Profiling is not in progress.")
+}
+
+/// The HTTP request handler for stopping profiling the compute.
+pub(in crate::http) async fn profile_stop() -> impl IntoResponse {
+    tracing::info!("Profile stop request received.");
+
+    match CANCEL_CHANNEL.lock().await.take() {
+        Some(tx) => {
+            if tx.send(()).is_err() {
+                tracing::error!("Failed to send cancellation signal.");
+                return JsonResponse::create_response(
+                    StatusCode::INTERNAL_SERVER_ERROR,
+                    "Failed to send cancellation signal",
+                );
+            }
+            JsonResponse::create_response(StatusCode::OK, "Profiling stopped successfully.")
+        }
+        None => JsonResponse::create_response(
+            StatusCode::PRECONDITION_FAILED,
+            "Profiling is not in progress, there is nothing to stop.",
+        ),
+    }
+}
+
+/// The HTTP request handler for starting profiling the compute.
+pub(in crate::http) async fn profile_start(
+    Json(request): Json<ProfileRequest>,
+) -> impl IntoResponse {
+    tracing::info!("Profile start request received: {request:?}");
+
+    let tx = tokio::sync::broadcast::Sender::<()>::new(1);
+
+    {
+        let mut cancel_channel = CANCEL_CHANNEL.lock().await;
+
+        if cancel_channel.is_some() {
+            return JsonResponse::create_response(
+                StatusCode::CONFLICT,
+                "Profiling is already in progress.",
+            );
+        }
+        *cancel_channel = Some(tx.clone());
+    }
+
+    tracing::info!("Profiling will start with parameters: {request:?}");
+    let pg_pid = Pid::from_raw(crate::compute::PG_PID.load(Ordering::SeqCst) as _);
+
+    let run_with_sudo = !cfg!(feature = "testing");
+
+    let options = crate::profiling::ProfileGenerationOptions {
+        profiler: request.profiler,
+        run_with_sudo,
+        pids: [pg_pid].into_iter().collect(),
+        follow_forks: true,
+        sampling_frequency: request.sampling_frequency as u32,
+        blocklist_symbols: vec![
+            "libc".to_owned(),
+            "libgcc".to_owned(),
+            "pthread".to_owned(),
+            "vdso".to_owned(),
+        ],
+        archive: request.archive,
+    };
+
+    let options = crate::profiling::ProfileGenerationTaskOptions {
+        options,
+        timeout: std::time::Duration::from_secs(request.timeout_seconds as u64),
+        should_stop: Some(tx),
+    };
+
+    let pprof_data = crate::profiling::generate_pprof_profile(options).await;
+
+    if CANCEL_CHANNEL.lock().await.take().is_none() {
+        tracing::error!("Profiling was cancelled from another request.");
+
+        return JsonResponse::create_response(
+            StatusCode::NO_CONTENT,
+            "Profiling was cancelled from another request.",
+        );
+    }
+
+    let pprof_data = match pprof_data {
+        Ok(data) => data,
+        Err(e) => {
+            tracing::error!(error = ?e, "failed to generate pprof data");
+            return JsonResponse::create_response(
+                StatusCode::INTERNAL_SERVER_ERROR,
+                format!("Failed to generate pprof data: {e:?}"),
+            );
+        }
+    };
+
+    tracing::info!("Profiling has completed successfully.");
+
+    let mut headers = http::HeaderMap::new();
+
+    if request.archive {
+        headers.insert(
+            http::header::CONTENT_TYPE,
+            http::HeaderValue::from_static("application/gzip"),
+        );
+        headers.insert(
+            http::header::CONTENT_DISPOSITION,
+            http::HeaderValue::from_static("attachment; filename=\"profile.pb.gz\""),
+        );
+    } else {
+        headers.insert(
+            http::header::CONTENT_TYPE,
+            http::HeaderValue::from_static("application/octet-stream"),
+        );
+        headers.insert(
+            http::header::CONTENT_DISPOSITION,
+            http::HeaderValue::from_static("attachment; filename=\"profile.pb\""),
+        );
+    }
+
+    (headers, pprof_data.0).into_response()
+}

compute_tools/src/http/server.rs

@@ -27,6 +27,7 @@ use super::{
     },
 };
 use crate::compute::ComputeNode;
+use crate::http::routes::profile;
 
 /// `compute_ctl` has two servers: internal and external. The internal server
 /// binds to the loopback interface and handles communication from clients on
@@ -84,8 +85,10 @@ impl From<&Server> for Router<Arc<ComputeNode>> {
                 let unauthenticated_router = Router::<Arc<ComputeNode>>::new()
                     .route("/metrics", get(metrics::get_metrics))
                     .route(
-                        "/autoscaling_metrics",
-                        get(metrics::get_autoscaling_metrics),
+                        "/profile/cpu",
+                        get(profile::profile_status)
+                            .post(profile::profile_start)
+                            .delete(profile::profile_stop),
                     );
 
                 let authenticated_router = Router::<Arc<ComputeNode>>::new()

compute_tools/src/installed_extensions.rs

@@ -2,7 +2,6 @@ use std::collections::HashMap;
 
 use anyhow::Result;
 use compute_api::responses::{InstalledExtension, InstalledExtensions};
-use tokio_postgres::error::Error as PostgresError;
 use tokio_postgres::{Client, Config, NoTls};
 
 use crate::metrics::INSTALLED_EXTENSIONS;
@@ -11,7 +10,7 @@ use crate::metrics::INSTALLED_EXTENSIONS;
 /// and to make database listing query here more explicit.
 ///
 /// Limit the number of databases to 500 to avoid excessive load.
-async fn list_dbs(client: &mut Client) -> Result<Vec<String>, PostgresError> {
+async fn list_dbs(client: &mut Client) -> Result<Vec<String>> {
     // `pg_database.datconnlimit = -2` means that the database is in the
     // invalid state
     let databases = client
@@ -38,9 +37,7 @@ async fn list_dbs(client: &mut Client) -> Result<Vec<String>, PostgresError> {
 /// Same extension can be installed in multiple databases with different versions,
 /// so we report a separate metric (number of databases where it is installed)
 /// for each extension version.
-pub async fn get_installed_extensions(
-    mut conf: Config,
-) -> Result<InstalledExtensions, PostgresError> {
+pub async fn get_installed_extensions(mut conf: Config) -> Result<InstalledExtensions> {
     conf.application_name("compute_ctl:get_installed_extensions");
     let databases: Vec<String> = {
         let (mut client, connection) = conf.connect(NoTls).await?;

compute_tools/src/lib.rs

@@ -4,7 +4,6 @@
 #![deny(clippy::undocumented_unsafe_blocks)]
 
 pub mod checker;
-pub mod communicator_socket_client;
 pub mod config;
 pub mod configurator;
 pub mod http;
@@ -16,7 +15,6 @@ pub mod compute_prewarm;
 pub mod compute_promote;
 pub mod disk_quota;
 pub mod extension_server;
-pub mod hadron_metrics;
 pub mod installed_extensions;
 pub mod local_proxy;
 pub mod lsn_lease;
@@ -26,6 +24,7 @@ pub mod monitor;
 pub mod params;
 pub mod pg_helpers;
 pub mod pgbouncer;
+pub mod profiling;
 pub mod rsyslog;
 pub mod spec;
 mod spec_apply;

compute_tools/src/logger.rs

@@ -13,9 +13,7 @@ use tracing_subscriber::prelude::*;
 /// set `OTEL_EXPORTER_OTLP_ENDPOINT=http://jaeger:4318`. See
 /// `tracing-utils` package description.
 ///
-pub fn init_tracing_and_logging(
-    default_log_level: &str,
-) -> anyhow::Result<Option<tracing_utils::Provider>> {
+pub async fn init_tracing_and_logging(default_log_level: &str) -> anyhow::Result<()> {
     // Initialize Logging
     let env_filter = tracing_subscriber::EnvFilter::try_from_default_env()
         .unwrap_or_else(|_| tracing_subscriber::EnvFilter::new(default_log_level));
@@ -26,9 +24,8 @@ pub fn init_tracing_and_logging(
         .with_writer(std::io::stderr);
 
     // Initialize OpenTelemetry
-    let provider =
-        tracing_utils::init_tracing("compute_ctl", tracing_utils::ExportConfig::default());
-    let otlp_layer = provider.as_ref().map(tracing_utils::layer);
+    let otlp_layer =
+        tracing_utils::init_tracing("compute_ctl", tracing_utils::ExportConfig::default()).await;
 
     // Put it all together
     tracing_subscriber::registry()
@@ -40,7 +37,7 @@ pub fn init_tracing_and_logging(
 
     utils::logging::replace_panic_hook_with_tracing_panic_hook().forget();
 
-    Ok(provider)
+    Ok(())
 }
 
 /// Replace all newline characters with a special character to make it

compute_tools/src/migration.rs

@@ -9,20 +9,15 @@ use crate::metrics::DB_MIGRATION_FAILED;
 pub(crate) struct MigrationRunner<'m> {
     client: &'m mut Client,
     migrations: &'m [&'m str],
-    lakebase_mode: bool,
 }
 
 impl<'m> MigrationRunner<'m> {
     /// Create a new migration runner
-    pub fn new(client: &'m mut Client, migrations: &'m [&'m str], lakebase_mode: bool) -> Self {
+    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,
-            lakebase_mode,
-        }
+        Self { client, migrations }
     }
 
     /// Get the current value neon_migration.migration_id
@@ -135,13 +130,8 @@ impl<'m> MigrationRunner<'m> {
             // ID is also the next index
             let migration_id = (current_migration + 1) as i64;
             let migration = self.migrations[current_migration];
-            let migration = if self.lakebase_mode {
-                migration.replace("neon_superuser", "databricks_superuser")
-            } else {
-                migration.to_string()
-            };
 
-            match Self::run_migration(self.client, migration_id, &migration).await {
+            match Self::run_migration(self.client, migration_id, migration).await {
                 Ok(_) => {
                     info!("Finished migration id={}", migration_id);
                 }

compute_tools/src/migrations/0001-add_bypass_rls_to_privileged_role.sql

@@ -1 +0,0 @@
-ALTER ROLE {privileged_role_name} BYPASSRLS;

compute_tools/src/migrations/0001-neon_superuser_bypass_rls.sql

@@ -0,0 +1 @@
+ALTER ROLE neon_superuser BYPASSRLS;

compute_tools/src/migrations/0002-alter_roles.sql

@@ -15,7 +15,7 @@ DO $$
 DECLARE
     role_name text;
 BEGIN
-    FOR role_name IN SELECT rolname FROM pg_roles WHERE pg_has_role(rolname, '{privileged_role_name}', 'member')
+    FOR role_name IN SELECT rolname FROM pg_roles WHERE pg_has_role(rolname, 'neon_superuser', 'member')
     LOOP
         RAISE NOTICE 'EXECUTING ALTER ROLE % INHERIT', quote_ident(role_name);
         EXECUTE 'ALTER ROLE ' || quote_ident(role_name) || ' INHERIT';
@@ -23,7 +23,7 @@ BEGIN
 
     FOR role_name IN SELECT rolname FROM pg_roles
         WHERE
-            NOT pg_has_role(rolname, '{privileged_role_name}', 'member') AND NOT starts_with(rolname, 'pg_')
+            NOT pg_has_role(rolname, 'neon_superuser', 'member') AND NOT starts_with(rolname, 'pg_')
     LOOP
         RAISE NOTICE 'EXECUTING ALTER ROLE % NOBYPASSRLS', quote_ident(role_name);
         EXECUTE 'ALTER ROLE ' || quote_ident(role_name) || ' NOBYPASSRLS';

compute_tools/src/migrations/0003-grant_pg_create_subscription_to_neon_superuser.sql

@@ -1,6 +1,6 @@
 DO $$
 BEGIN
     IF (SELECT setting::numeric >= 160000 FROM pg_settings WHERE name = 'server_version_num') THEN
-        EXECUTE 'GRANT pg_create_subscription TO {privileged_role_name}';
+        EXECUTE 'GRANT pg_create_subscription TO neon_superuser';
     END IF;
 END $$;

compute_tools/src/migrations/0004-grant_pg_monitor_to_neon_superuser.sql

@@ -0,0 +1 @@
+GRANT pg_monitor TO neon_superuser WITH ADMIN OPTION;

compute_tools/src/migrations/0004-grant_pg_monitor_to_privileged_role.sql

@@ -1 +0,0 @@
-GRANT pg_monitor TO {privileged_role_name} WITH ADMIN OPTION;

compute_tools/src/migrations/0005-grant_all_on_tables_to_neon_superuser.sql

@@ -1,4 +1,4 @@
 -- SKIP: Deemed insufficient for allowing relations created by extensions to be
---       interacted with by {privileged_role_name} without permission issues.
+--       interacted with by neon_superuser without permission issues.
 
-ALTER DEFAULT PRIVILEGES IN SCHEMA public GRANT ALL ON TABLES TO {privileged_role_name};
+ALTER DEFAULT PRIVILEGES IN SCHEMA public GRANT ALL ON TABLES TO neon_superuser;

compute_tools/src/migrations/0006-grant_all_on_sequences_to_neon_superuser.sql

@@ -1,4 +1,4 @@
 -- SKIP: Deemed insufficient for allowing relations created by extensions to be
---       interacted with by {privileged_role_name} without permission issues.
+--       interacted with by neon_superuser without permission issues.
 
-ALTER DEFAULT PRIVILEGES IN SCHEMA public GRANT ALL ON SEQUENCES TO {privileged_role_name};
+ALTER DEFAULT PRIVILEGES IN SCHEMA public GRANT ALL ON SEQUENCES TO neon_superuser;

compute_tools/src/migrations/0007-grant_all_on_tables_to_neon_superuser_with_grant_option.sql

@@ -1,3 +1,3 @@
 -- SKIP: Moved inline to the handle_grants() functions.
 
-ALTER DEFAULT PRIVILEGES IN SCHEMA public GRANT ALL ON TABLES TO {privileged_role_name} WITH GRANT OPTION;
+ALTER DEFAULT PRIVILEGES IN SCHEMA public GRANT ALL ON TABLES TO neon_superuser WITH GRANT OPTION;

compute_tools/src/migrations/0008-grant_all_on_sequences_to_neon_superuser_with_grant_option.sql

@@ -1,3 +1,3 @@
 -- SKIP: Moved inline to the handle_grants() functions.
 
-ALTER DEFAULT PRIVILEGES IN SCHEMA public GRANT ALL ON SEQUENCES TO {privileged_role_name} WITH GRANT OPTION;
+ALTER DEFAULT PRIVILEGES IN SCHEMA public GRANT ALL ON SEQUENCES TO neon_superuser WITH GRANT OPTION;

compute_tools/src/migrations/0010-grant_snapshot_synchronization_funcs_to_neon_superuser.sql

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

compute_tools/src/migrations/0011-grant_pg_show_replication_origin_status_to_neon_superuser.sql

@@ -0,0 +1 @@
+GRANT EXECUTE ON FUNCTION pg_show_replication_origin_status TO neon_superuser;

compute_tools/src/migrations/0011-grant_pg_show_replication_origin_status_to_privileged_role.sql

@@ -1 +0,0 @@
-GRANT EXECUTE ON FUNCTION pg_show_replication_origin_status TO {privileged_role_name};

compute_tools/src/migrations/0012-grant_pg_signal_backend_to_neon_superuser.sql

@@ -0,0 +1 @@
+GRANT pg_signal_backend TO neon_superuser WITH ADMIN OPTION;

compute_tools/src/migrations/0012-grant_pg_signal_backend_to_privileged_role.sql

@@ -1 +0,0 @@
-GRANT pg_signal_backend TO {privileged_role_name} WITH ADMIN OPTION;

compute_tools/src/monitor.rs

@@ -11,7 +11,6 @@ use tracing::{Level, error, info, instrument, span};
 use crate::compute::ComputeNode;
 use crate::metrics::{PG_CURR_DOWNTIME_MS, PG_TOTAL_DOWNTIME_MS};
 
-const PG_DEFAULT_INIT_TIMEOUIT: Duration = Duration::from_secs(60);
 const MONITOR_CHECK_INTERVAL: Duration = Duration::from_millis(500);
 
 /// Struct to store runtime state of the compute monitor thread.
@@ -353,47 +352,13 @@ impl ComputeMonitor {
 // Hang on condition variable waiting until the compute status is `Running`.
 fn wait_for_postgres_start(compute: &ComputeNode) {
     let mut state = compute.state.lock().unwrap();
-    let pg_init_timeout = compute
-        .params
-        .pg_init_timeout
-        .unwrap_or(PG_DEFAULT_INIT_TIMEOUIT);
-
     while state.status != ComputeStatus::Running {
         info!("compute is not running, waiting before monitoring activity");
-        if !compute.params.lakebase_mode {
-            state = compute.state_changed.wait(state).unwrap();
+        state = compute.state_changed.wait(state).unwrap();
 
-            if state.status == ComputeStatus::Running {
-                break;
-            }
-            continue;
+        if state.status == ComputeStatus::Running {
+            break;
         }
-
-        if state.pg_start_time.is_some()
-            && Utc::now()
-                .signed_duration_since(state.pg_start_time.unwrap())
-                .to_std()
-                .unwrap_or_default()
-                > pg_init_timeout
-        {
-            // If Postgres isn't up and running with working PS/SK connections within POSTGRES_STARTUP_TIMEOUT, it is
-            // possible that we started Postgres with a wrong spec (so it is talking to the wrong PS/SK nodes). To prevent
-            // deadends we simply exit (panic) the compute node so it can restart with the latest spec.
-            //
-            // NB: We skip this check if we have not attempted to start PG yet (indicated by state.pg_start_up == None).
-            // This is to make sure the more appropriate errors are surfaced if we encounter issues before we even attempt
-            // to start PG (e.g., if we can't pull the spec, can't sync safekeepers, or can't get the basebackup).
-            error!(
-                "compute did not enter Running state in {} seconds, exiting",
-                pg_init_timeout.as_secs()
-            );
-            std::process::exit(1);
-        }
-        state = compute
-            .state_changed
-            .wait_timeout(state, Duration::from_secs(5))
-            .unwrap()
-            .0;
     }
 }
 

compute_tools/src/pg_helpers.rs

@@ -11,9 +11,7 @@ use std::time::{Duration, Instant};
 
 use anyhow::{Result, bail};
 use compute_api::responses::TlsConfig;
-use compute_api::spec::{
-    Database, DatabricksSettings, GenericOption, GenericOptions, PgIdent, Role,
-};
+use compute_api::spec::{Database, GenericOption, GenericOptions, PgIdent, Role};
 use futures::StreamExt;
 use indexmap::IndexMap;
 use ini::Ini;
@@ -186,42 +184,6 @@ impl DatabaseExt for Database {
     }
 }
 
-pub trait DatabricksSettingsExt {
-    fn as_pg_settings(&self) -> String;
-}
-
-impl DatabricksSettingsExt for DatabricksSettings {
-    fn as_pg_settings(&self) -> String {
-        // Postgres GUCs rendered from DatabricksSettings
-        vec![
-            // ssl_ca_file
-            Some(format!(
-                "ssl_ca_file = '{}'",
-                self.pg_compute_tls_settings.ca_file
-            )),
-            // [Optional] databricks.workspace_url
-            Some(format!(
-                "databricks.workspace_url = '{}'",
-                &self.databricks_workspace_host
-            )),
-            // todo(vikas.jain): these are not required anymore as they are moved to static
-            // conf but keeping these to avoid image mismatch between hcc and pg.
-            // Once hcc and pg are in sync, we can remove these.
-            //
-            // databricks.enable_databricks_identity_login
-            Some("databricks.enable_databricks_identity_login = true".to_string()),
-            // databricks.enable_sql_restrictions
-            Some("databricks.enable_sql_restrictions = true".to_string()),
-        ]
-        .into_iter()
-        // Removes `None`s
-        .flatten()
-        .collect::<Vec<String>>()
-        .join("\n")
-            + "\n"
-    }
-}
-
 /// Generic trait used to provide quoting / encoding for strings used in the
 /// Postgres SQL queries and DATABASE_URL.
 pub trait Escaping {

compute_tools/src/spec.rs

@@ -1,6 +1,4 @@
 use std::fs::File;
-use std::fs::{self, Permissions};
-use std::os::unix::fs::PermissionsExt;
 use std::path::Path;
 
 use anyhow::{Result, anyhow, bail};
@@ -11,7 +9,6 @@ use reqwest::StatusCode;
 use tokio_postgres::Client;
 use tracing::{error, info, instrument};
 
-use crate::compute::ComputeNodeParams;
 use crate::config;
 use crate::metrics::{CPLANE_REQUESTS_TOTAL, CPlaneRequestRPC, UNKNOWN_HTTP_STATUS};
 use crate::migration::MigrationRunner;
@@ -135,25 +132,10 @@ pub fn get_config_from_control_plane(base_uri: &str, compute_id: &str) -> Result
 }
 
 /// Check `pg_hba.conf` and update if needed to allow external connections.
-pub fn update_pg_hba(pgdata_path: &Path, databricks_pg_hba: Option<&String>) -> Result<()> {
+pub fn update_pg_hba(pgdata_path: &Path) -> Result<()> {
     // XXX: consider making it a part of config.json
     let pghba_path = pgdata_path.join("pg_hba.conf");
 
-    // Update pg_hba to contains databricks specfic settings before adding neon settings
-    // PG uses the first record that matches to perform authentication, so we need to have
-    // our rules before the default ones from neon.
-    // See https://www.postgresql.org/docs/16/auth-pg-hba-conf.html
-    if let Some(databricks_pg_hba) = databricks_pg_hba {
-        if config::line_in_file(
-            &pghba_path,
-            &format!("include_if_exists {}\n", *databricks_pg_hba),
-        )? {
-            info!("updated pg_hba.conf to include databricks_pg_hba.conf");
-        } else {
-            info!("pg_hba.conf already included databricks_pg_hba.conf");
-        }
-    }
-
     if config::line_in_file(&pghba_path, PG_HBA_ALL_MD5)? {
         info!("updated pg_hba.conf to allow external connections");
     } else {
@@ -163,59 +145,6 @@ pub fn update_pg_hba(pgdata_path: &Path, databricks_pg_hba: Option<&String>) ->
     Ok(())
 }
 
-/// Check `pg_ident.conf` and update if needed to allow databricks config.
-pub fn update_pg_ident(pgdata_path: &Path, databricks_pg_ident: Option<&String>) -> Result<()> {
-    info!("checking pg_ident.conf");
-    let pghba_path = pgdata_path.join("pg_ident.conf");
-
-    // Update pg_ident to contains databricks specfic settings
-    if let Some(databricks_pg_ident) = databricks_pg_ident {
-        if config::line_in_file(
-            &pghba_path,
-            &format!("include_if_exists {}\n", *databricks_pg_ident),
-        )? {
-            info!("updated pg_ident.conf to include databricks_pg_ident.conf");
-        } else {
-            info!("pg_ident.conf already included databricks_pg_ident.conf");
-        }
-    }
-
-    Ok(())
-}
-
-/// Copy tls key_file and cert_file from k8s secret mount directory
-/// to pgdata and set private key file permissions as expected by Postgres.
-/// See this doc for expected permission <https://www.postgresql.org/docs/current/ssl-tcp.html>
-/// K8s secrets mount on dblet does not honor permission and ownership
-/// specified in the Volume or VolumeMount. So we need to explicitly copy the file and set the permissions.
-pub fn copy_tls_certificates(
-    key_file: &String,
-    cert_file: &String,
-    pgdata_path: &Path,
-) -> Result<()> {
-    let files = [cert_file, key_file];
-    for file in files.iter() {
-        let source = Path::new(file);
-        let dest = pgdata_path.join(source.file_name().unwrap());
-        if !dest.exists() {
-            std::fs::copy(source, &dest)?;
-            info!(
-                "Copying tls file: {} to {}",
-                &source.display(),
-                &dest.display()
-            );
-        }
-        if *file == key_file {
-            // Postgres requires private key to be readable only by the owner by having
-            // chmod 600 permissions.
-            let permissions = Permissions::from_mode(0o600);
-            fs::set_permissions(&dest, permissions)?;
-            info!("Setting permission on {}.", &dest.display());
-        }
-    }
-    Ok(())
-}
-
 /// Create a standby.signal file
 pub fn add_standby_signal(pgdata_path: &Path) -> Result<()> {
     // XXX: consider making it a part of config.json
@@ -240,11 +169,7 @@ pub async fn handle_neon_extension_upgrade(client: &mut Client) -> Result<()> {
 }
 
 #[instrument(skip_all)]
-pub async fn handle_migrations(
-    params: ComputeNodeParams,
-    client: &mut Client,
-    lakebase_mode: bool,
-) -> Result<()> {
+pub async fn handle_migrations(client: &mut Client) -> Result<()> {
     info!("handle migrations");
 
     // !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
@@ -253,62 +178,29 @@ pub async fn handle_migrations(
 
     // Add new migrations in numerical order.
     let migrations = [
-        &format!(
-            include_str!("./migrations/0001-add_bypass_rls_to_privileged_role.sql"),
-            privileged_role_name = params.privileged_role_name
+        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/0007-grant_all_on_tables_to_neon_superuser_with_grant_option.sql"
         ),
-        &format!(
-            include_str!("./migrations/0002-alter_roles.sql"),
-            privileged_role_name = params.privileged_role_name
-        ),
-        &format!(
-            include_str!("./migrations/0003-grant_pg_create_subscription_to_privileged_role.sql"),
-            privileged_role_name = params.privileged_role_name
-        ),
-        &format!(
-            include_str!("./migrations/0004-grant_pg_monitor_to_privileged_role.sql"),
-            privileged_role_name = params.privileged_role_name
-        ),
-        &format!(
-            include_str!("./migrations/0005-grant_all_on_tables_to_privileged_role.sql"),
-            privileged_role_name = params.privileged_role_name
-        ),
-        &format!(
-            include_str!("./migrations/0006-grant_all_on_sequences_to_privileged_role.sql"),
-            privileged_role_name = params.privileged_role_name
-        ),
-        &format!(
-            include_str!(
-                "./migrations/0007-grant_all_on_tables_with_grant_option_to_privileged_role.sql"
-            ),
-            privileged_role_name = params.privileged_role_name
-        ),
-        &format!(
-            include_str!(
-                "./migrations/0008-grant_all_on_sequences_with_grant_option_to_privileged_role.sql"
-            ),
-            privileged_role_name = params.privileged_role_name
+        include_str!(
+            "./migrations/0008-grant_all_on_sequences_to_neon_superuser_with_grant_option.sql"
         ),
         include_str!("./migrations/0009-revoke_replication_for_previously_allowed_roles.sql"),
-        &format!(
-            include_str!(
-                "./migrations/0010-grant_snapshot_synchronization_funcs_to_privileged_role.sql"
-            ),
-            privileged_role_name = params.privileged_role_name
+        include_str!(
+            "./migrations/0010-grant_snapshot_synchronization_funcs_to_neon_superuser.sql"
         ),
-        &format!(
-            include_str!(
-                "./migrations/0011-grant_pg_show_replication_origin_status_to_privileged_role.sql"
-            ),
-            privileged_role_name = params.privileged_role_name
-        ),
-        &format!(
-            include_str!("./migrations/0012-grant_pg_signal_backend_to_privileged_role.sql"),
-            privileged_role_name = params.privileged_role_name
+        include_str!(
+            "./migrations/0011-grant_pg_show_replication_origin_status_to_neon_superuser.sql"
         ),
+        include_str!("./migrations/0012-grant_pg_signal_backend_to_neon_superuser.sql"),
     ];
 
-    MigrationRunner::new(client, &migrations, lakebase_mode)
+    MigrationRunner::new(client, &migrations)
         .run_migrations()
         .await?;
 

compute_tools/src/spec_apply.rs

@@ -13,14 +13,14 @@ use tokio_postgres::Client;
 use tokio_postgres::error::SqlState;
 use tracing::{Instrument, debug, error, info, info_span, instrument, warn};
 
-use crate::compute::{ComputeNode, ComputeNodeParams, ComputeState};
+use crate::compute::{ComputeNode, ComputeState};
 use crate::pg_helpers::{
     DatabaseExt, Escaping, GenericOptionsSearch, RoleExt, get_existing_dbs_async,
     get_existing_roles_async,
 };
 use crate::spec_apply::ApplySpecPhase::{
-    CreateAndAlterDatabases, CreateAndAlterRoles, CreateAvailabilityCheck, CreatePgauditExtension,
-    CreatePgauditlogtofileExtension, CreatePrivilegedRole, CreateSchemaNeon,
+    CreateAndAlterDatabases, CreateAndAlterRoles, CreateAvailabilityCheck, CreateNeonSuperuser,
+    CreatePgauditExtension, CreatePgauditlogtofileExtension, CreateSchemaNeon,
     DisablePostgresDBPgAudit, DropInvalidDatabases, DropRoles, FinalizeDropLogicalSubscriptions,
     HandleNeonExtension, HandleOtherExtensions, RenameAndDeleteDatabases, RenameRoles,
     RunInEachDatabase,
@@ -49,7 +49,6 @@ impl ComputeNode {
             // Proceed with post-startup configuration. Note, that order of operations is important.
             let client = Self::get_maintenance_client(&conf).await?;
             let spec = spec.clone();
-            let params = Arc::new(self.params.clone());
 
             let databases = get_existing_dbs_async(&client).await?;
             let roles = get_existing_roles_async(&client)
@@ -158,7 +157,6 @@ impl ComputeNode {
 
                     let conf = Arc::new(conf);
                     let fut = Self::apply_spec_sql_db(
-                        params.clone(),
                         spec.clone(),
                         conf,
                         ctx.clone(),
@@ -187,7 +185,7 @@ impl ComputeNode {
             }
 
             for phase in [
-                CreatePrivilegedRole,
+                CreateNeonSuperuser,
                 DropInvalidDatabases,
                 RenameRoles,
                 CreateAndAlterRoles,
@@ -197,7 +195,6 @@ impl ComputeNode {
             ] {
                 info!("Applying phase {:?}", &phase);
                 apply_operations(
-                    params.clone(),
                     spec.clone(),
                     ctx.clone(),
                     jwks_roles.clone(),
@@ -246,7 +243,6 @@ impl ComputeNode {
                     }
 
                     let fut = Self::apply_spec_sql_db(
-                        params.clone(),
                         spec.clone(),
                         conf,
                         ctx.clone(),
@@ -297,7 +293,6 @@ impl ComputeNode {
             for phase in phases {
                 debug!("Applying phase {:?}", &phase);
                 apply_operations(
-                    params.clone(),
                     spec.clone(),
                     ctx.clone(),
                     jwks_roles.clone(),
@@ -318,9 +313,7 @@ impl ComputeNode {
     /// May opt to not connect to databases that don't have any scheduled
     /// operations.  The function is concurrency-controlled with the provided
     /// semaphore.  The caller has to make sure the semaphore isn't exhausted.
-    #[allow(clippy::too_many_arguments)] // TODO: needs bigger refactoring
     async fn apply_spec_sql_db(
-        params: Arc<ComputeNodeParams>,
         spec: Arc<ComputeSpec>,
         conf: Arc<tokio_postgres::Config>,
         ctx: Arc<tokio::sync::RwLock<MutableApplyContext>>,
@@ -335,7 +328,6 @@ impl ComputeNode {
 
         for subphase in subphases {
             apply_operations(
-                params.clone(),
                 spec.clone(),
                 ctx.clone(),
                 jwks_roles.clone(),
@@ -411,8 +403,7 @@ impl ComputeNode {
             .map(|limit| match limit {
                 0..10 => limit,
                 10..30 => 10,
-                30..300 => limit / 3,
-                300.. => 100,
+                30.. => limit / 3,
             })
             // If we didn't find max_connections, default to 10 concurrent connections.
             .unwrap_or(10)
@@ -476,7 +467,7 @@ pub enum PerDatabasePhase {
 
 #[derive(Clone, Debug)]
 pub enum ApplySpecPhase {
-    CreatePrivilegedRole,
+    CreateNeonSuperuser,
     DropInvalidDatabases,
     RenameRoles,
     CreateAndAlterRoles,
@@ -519,7 +510,6 @@ pub struct MutableApplyContext {
 /// - No timeouts have (yet) been implemented.
 /// - The caller is responsible for limiting and/or applying concurrency.
 pub async fn apply_operations<'a, Fut, F>(
-    params: Arc<ComputeNodeParams>,
     spec: Arc<ComputeSpec>,
     ctx: Arc<RwLock<MutableApplyContext>>,
     jwks_roles: Arc<HashSet<String>>,
@@ -537,7 +527,7 @@ where
         debug!("Processing phase {:?}", &apply_spec_phase);
         let ctx = ctx;
 
-        let mut ops = get_operations(&params, &spec, &ctx, &jwks_roles, &apply_spec_phase)
+        let mut ops = get_operations(&spec, &ctx, &jwks_roles, &apply_spec_phase)
             .await?
             .peekable();
 
@@ -598,18 +588,14 @@ where
 /// sort/merge/batch execution, but for now this is a nice way to improve
 /// batching behavior of the commands.
 async fn get_operations<'a>(
-    params: &'a ComputeNodeParams,
     spec: &'a ComputeSpec,
     ctx: &'a RwLock<MutableApplyContext>,
     jwks_roles: &'a HashSet<String>,
     apply_spec_phase: &'a ApplySpecPhase,
 ) -> Result<Box<dyn Iterator<Item = Operation> + 'a + Send>> {
     match apply_spec_phase {
-        ApplySpecPhase::CreatePrivilegedRole => Ok(Box::new(once(Operation {
-            query: format!(
-                include_str!("sql/create_privileged_role.sql"),
-                privileged_role_name = params.privileged_role_name
-            ),
+        ApplySpecPhase::CreateNeonSuperuser => Ok(Box::new(once(Operation {
+            query: include_str!("sql/create_neon_superuser.sql").to_string(),
             comment: None,
         }))),
         ApplySpecPhase::DropInvalidDatabases => {
@@ -711,9 +697,8 @@ async fn get_operations<'a>(
                         None => {
                             let query = if !jwks_roles.contains(role.name.as_str()) {
                                 format!(
-                                    "CREATE ROLE {} INHERIT CREATEROLE CREATEDB BYPASSRLS REPLICATION IN ROLE {} {}",
+                                    "CREATE ROLE {} INHERIT CREATEROLE CREATEDB BYPASSRLS REPLICATION IN ROLE neon_superuser {}",
                                     role.name.pg_quote(),
-                                    params.privileged_role_name,
                                     role.to_pg_options(),
                                 )
                             } else {
@@ -864,9 +849,8 @@ async fn get_operations<'a>(
                                 // ALL PRIVILEGES grants CREATE, CONNECT, and TEMPORARY on the database
                                 // (see https://www.postgresql.org/docs/current/ddl-priv.html)
                                 query: format!(
-                                    "GRANT ALL PRIVILEGES ON DATABASE {} TO {}",
-                                    db.name.pg_quote(),
-                                    params.privileged_role_name
+                                    "GRANT ALL PRIVILEGES ON DATABASE {} TO neon_superuser",
+                                    db.name.pg_quote()
                                 ),
                                 comment: None,
                             },

compute_tools/src/sql/create_neon_superuser.sql

@@ -0,0 +1,8 @@
+DO $$
+    BEGIN
+        IF NOT EXISTS (SELECT FROM pg_catalog.pg_roles WHERE rolname = 'neon_superuser')
+        THEN
+            CREATE ROLE neon_superuser CREATEDB CREATEROLE NOLOGIN REPLICATION BYPASSRLS IN ROLE pg_read_all_data, pg_write_all_data;
+        END IF;
+    END
+$$;

compute_tools/src/sql/create_privileged_role.sql

@@ -1,8 +0,0 @@
-DO $$
-    BEGIN
-        IF NOT EXISTS (SELECT FROM pg_catalog.pg_roles WHERE rolname = '{privileged_role_name}')
-        THEN
-            CREATE ROLE {privileged_role_name} CREATEDB CREATEROLE NOLOGIN REPLICATION BYPASSRLS IN ROLE pg_read_all_data, pg_write_all_data;
-        END IF;
-    END
-$$;

control_plane/README.md

@@ -8,10 +8,10 @@ code changes locally, but not suitable for running production systems.
 
 ## Example: Start with Postgres 16
 
-To create and start a local development environment with Postgres 16, you will need to provide `--pg-version` flag to 2 of the start-up commands.
+To create and start a local development environment with Postgres 16, you will need to provide `--pg-version` flag to 3 of the start-up commands.
 
 ```shell
-cargo neon init
+cargo neon init --pg-version 16
 cargo neon start
 cargo neon tenant create --set-default --pg-version 16
 cargo neon endpoint create main --pg-version 16

control_plane/src/bin/neon_local.rs

@@ -407,12 +407,6 @@ struct StorageControllerStartCmdArgs {
         help = "Base port for the storage controller instance idenfified by instance-id (defaults to pageserver cplane api)"
     )]
     base_port: Option<u16>,
-
-    #[clap(
-        long,
-        help = "Whether the storage controller should handle pageserver-reported local disk loss events."
-    )]
-    handle_ps_local_disk_loss: Option<bool>,
 }
 
 #[derive(clap::Args)]
@@ -637,10 +631,6 @@ struct EndpointCreateCmdArgs {
         help = "Allow multiple primary endpoints running on the same branch. Shouldn't be used normally, but useful for tests."
     )]
     allow_multiple: bool,
-
-    /// Only allow changing it on creation
-    #[clap(long, help = "Name of the privileged role for the endpoint")]
-    privileged_role_name: Option<String>,
 }
 
 #[derive(clap::Args)]
@@ -1490,7 +1480,6 @@ async fn handle_endpoint(subcmd: &EndpointCmd, env: &local_env::LocalEnv) -> Res
                 args.grpc,
                 !args.update_catalog,
                 false,
-                args.privileged_role_name.clone(),
             )?;
         }
         EndpointCmd::Start(args) => {
@@ -1815,7 +1804,6 @@ async fn handle_storage_controller(
                 instance_id: args.instance_id,
                 base_port: args.base_port,
                 start_timeout: args.start_timeout,
-                handle_ps_local_disk_loss: args.handle_ps_local_disk_loss,
             };
 
             if let Err(e) = svc.start(start_args).await {

control_plane/src/broker.rs

@@ -36,7 +36,7 @@ impl StorageBroker {
     pub async fn start(&self, retry_timeout: &Duration) -> anyhow::Result<()> {
         let broker = &self.env.broker;
 
-        println!("Starting neon broker at {}", broker.client_url());
+        print!("Starting neon broker at {}", broker.client_url());
 
         let mut args = Vec::new();
 

control_plane/src/endpoint.rs

@@ -32,8 +32,7 @@
 //!     config.json                 - passed to `compute_ctl`
 //!     pgdata/
 //!         postgresql.conf       - copy of postgresql.conf created by `compute_ctl`
-//!         neon.signal
-//!         zenith.signal         - copy of neon.signal, for backward compatibility
+//!         zenith.signal
 //!         <other PostgreSQL files>
 //! ```
 //!
@@ -65,6 +64,7 @@ use jsonwebtoken::jwk::{
     OctetKeyPairParameters, OctetKeyPairType, PublicKeyUse,
 };
 use nix::sys::signal::{Signal, kill};
+use pageserver_api::shard::ShardStripeSize;
 use pem::Pem;
 use reqwest::header::CONTENT_TYPE;
 use safekeeper_api::PgMajorVersion;
@@ -76,7 +76,6 @@ use spki::{SubjectPublicKeyInfo, SubjectPublicKeyInfoRef};
 use tracing::debug;
 use url::Host;
 use utils::id::{NodeId, TenantId, TimelineId};
-use utils::shard::ShardStripeSize;
 
 use crate::local_env::LocalEnv;
 use crate::postgresql_conf::PostgresConf;
@@ -99,7 +98,6 @@ pub struct EndpointConf {
     features: Vec<ComputeFeature>,
     cluster: Option<Cluster>,
     compute_ctl_config: ComputeCtlConfig,
-    privileged_role_name: Option<String>,
 }
 
 //
@@ -200,7 +198,6 @@ impl ComputeControlPlane {
         grpc: bool,
         skip_pg_catalog_updates: bool,
         drop_subscriptions_before_start: bool,
-        privileged_role_name: Option<String>,
     ) -> Result<Arc<Endpoint>> {
         let pg_port = pg_port.unwrap_or_else(|| self.get_port());
         let external_http_port = external_http_port.unwrap_or_else(|| self.get_port() + 1);
@@ -238,7 +235,6 @@ impl ComputeControlPlane {
             features: vec![],
             cluster: None,
             compute_ctl_config: compute_ctl_config.clone(),
-            privileged_role_name: privileged_role_name.clone(),
         });
 
         ep.create_endpoint_dir()?;
@@ -260,7 +256,6 @@ impl ComputeControlPlane {
                 features: vec![],
                 cluster: None,
                 compute_ctl_config,
-                privileged_role_name,
             })?,
         )?;
         std::fs::write(
@@ -336,9 +331,6 @@ pub struct Endpoint {
 
     /// The compute_ctl config for the endpoint's compute.
     compute_ctl_config: ComputeCtlConfig,
-
-    /// The name of the privileged role for the endpoint.
-    privileged_role_name: Option<String>,
 }
 
 #[derive(PartialEq, Eq)]
@@ -439,7 +431,6 @@ impl Endpoint {
             features: conf.features,
             cluster: conf.cluster,
             compute_ctl_config: conf.compute_ctl_config,
-            privileged_role_name: conf.privileged_role_name,
         })
     }
 
@@ -472,7 +463,7 @@ impl Endpoint {
         conf.append("max_connections", "100");
         conf.append("wal_level", "logical");
         // wal_sender_timeout is the maximum time to wait for WAL replication.
-        // It also defines how often the walreceiver will send a feedback message to the wal sender.
+        // It also defines how often the walreciever will send a feedback message to the wal sender.
         conf.append("wal_sender_timeout", "5s");
         conf.append("listen_addresses", &self.pg_address.ip().to_string());
         conf.append("port", &self.pg_address.port().to_string());
@@ -878,10 +869,6 @@ impl Endpoint {
             cmd.arg("--dev");
         }
 
-        if let Some(privileged_role_name) = self.privileged_role_name.clone() {
-            cmd.args(["--privileged-role-name", &privileged_role_name]);
-        }
-
         let child = cmd.spawn()?;
         // set up a scopeguard to kill & wait for the child in case we panic or bail below
         let child = scopeguard::guard(child, |mut child| {

control_plane/src/local_env.rs

@@ -217,9 +217,6 @@ pub struct NeonStorageControllerConf {
     pub posthog_config: Option<PostHogConfig>,
 
     pub kick_secondary_downloads: Option<bool>,
-
-    #[serde(with = "humantime_serde")]
-    pub shard_split_request_timeout: Option<Duration>,
 }
 
 impl NeonStorageControllerConf {
@@ -253,7 +250,6 @@ impl Default for NeonStorageControllerConf {
             timeline_safekeeper_count: None,
             posthog_config: None,
             kick_secondary_downloads: None,
-            shard_split_request_timeout: None,
         }
     }
 }

control_plane/src/pageserver.rs

@@ -303,7 +303,7 @@ impl PageServerNode {
     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();
-        println!(
+        print!(
             "Starting pageserver node {} at '{}' in {:?}, retrying for {:?}",
             self.conf.id,
             self.pg_connection_config.raw_address(),

control_plane/src/safekeeper.rs

@@ -127,7 +127,7 @@ impl SafekeeperNode {
         extra_opts: &[String],
         retry_timeout: &Duration,
     ) -> anyhow::Result<()> {
-        println!(
+        print!(
             "Starting safekeeper at '{}' in '{}', retrying for {:?}",
             self.pg_connection_config.raw_address(),
             self.datadir_path().display(),

control_plane/src/storage_controller.rs

@@ -56,7 +56,6 @@ pub struct NeonStorageControllerStartArgs {
     pub instance_id: u8,
     pub base_port: Option<u16>,
     pub start_timeout: humantime::Duration,
-    pub handle_ps_local_disk_loss: Option<bool>,
 }
 
 impl NeonStorageControllerStartArgs {
@@ -65,7 +64,6 @@ impl NeonStorageControllerStartArgs {
             instance_id: 1,
             base_port: None,
             start_timeout,
-            handle_ps_local_disk_loss: None,
         }
     }
 }
@@ -650,13 +648,6 @@ impl StorageController {
             args.push(format!("--timeline-safekeeper-count={sk_cnt}"));
         }
 
-        if let Some(duration) = self.config.shard_split_request_timeout {
-            args.push(format!(
-                "--shard-split-request-timeout={}",
-                humantime::Duration::from(duration)
-            ));
-        }
-
         let mut envs = vec![
             ("LD_LIBRARY_PATH".to_owned(), pg_lib_dir.to_string()),
             ("DYLD_LIBRARY_PATH".to_owned(), pg_lib_dir.to_string()),
@@ -669,11 +660,7 @@ impl StorageController {
             ));
         }
 
-        println!("Starting storage controller at {scheme}://{host}:{listen_port}");
-
-        if start_args.handle_ps_local_disk_loss.unwrap_or_default() {
-            args.push("--handle-ps-local-disk-loss".to_string());
-        }
+        println!("Starting storage controller");
 
         background_process::start_process(
             COMMAND,

control_plane/storcon_cli/Cargo.toml

@@ -14,7 +14,6 @@ humantime.workspace = true
 pageserver_api.workspace = true
 pageserver_client.workspace = true
 reqwest.workspace = true
-safekeeper_api.workspace=true
 serde_json = { workspace = true, features = ["raw_value"] }
 storage_controller_client.workspace = true
 tokio.workspace = true

control_plane/storcon_cli/src/main.rs

@@ -11,7 +11,7 @@ use pageserver_api::controller_api::{
     PlacementPolicy, SafekeeperDescribeResponse, SafekeeperSchedulingPolicyRequest,
     ShardSchedulingPolicy, ShardsPreferredAzsRequest, ShardsPreferredAzsResponse,
     SkSchedulingPolicy, TenantCreateRequest, TenantDescribeResponse, TenantPolicyRequest,
-    TenantShardMigrateRequest, TenantShardMigrateResponse, TimelineSafekeeperMigrateRequest,
+    TenantShardMigrateRequest, TenantShardMigrateResponse,
 };
 use pageserver_api::models::{
     EvictionPolicy, EvictionPolicyLayerAccessThreshold, ShardParameters, TenantConfig,
@@ -21,7 +21,6 @@ use pageserver_api::models::{
 use pageserver_api::shard::{ShardStripeSize, TenantShardId};
 use pageserver_client::mgmt_api::{self};
 use reqwest::{Certificate, Method, StatusCode, Url};
-use safekeeper_api::models::TimelineLocateResponse;
 use storage_controller_client::control_api::Client;
 use utils::id::{NodeId, TenantId, TimelineId};
 
@@ -76,12 +75,6 @@ enum Command {
     NodeStartDelete {
         #[arg(long)]
         node_id: NodeId,
-        /// When `force` is true, skip waiting for shards to prewarm during migration.
-        /// This can significantly speed up node deletion since prewarming all shards
-        /// can take considerable time, but may result in slower initial access to
-        /// migrated shards until they warm up naturally.
-        #[arg(long)]
-        force: bool,
     },
     /// Cancel deletion of the specified pageserver and wait for `timeout`
     /// for the operation to be canceled. May be retried.
@@ -286,23 +279,6 @@ enum Command {
         #[arg(long)]
         concurrency: Option<usize>,
     },
-    /// Locate safekeepers for a timeline from the storcon DB.
-    TimelineLocate {
-        #[arg(long)]
-        tenant_id: TenantId,
-        #[arg(long)]
-        timeline_id: TimelineId,
-    },
-    /// Migrate a timeline to a new set of safekeepers
-    TimelineSafekeeperMigrate {
-        #[arg(long)]
-        tenant_id: TenantId,
-        #[arg(long)]
-        timeline_id: TimelineId,
-        /// Example: --new-sk-set 1,2,3
-        #[arg(long, required = true, value_delimiter = ',')]
-        new_sk_set: Vec<NodeId>,
-    },
 }
 
 #[derive(Parser)]
@@ -482,7 +458,6 @@ async fn main() -> anyhow::Result<()> {
                         listen_http_port,
                         listen_https_port,
                         availability_zone_id: AvailabilityZone(availability_zone_id),
-                        node_ip_addr: None,
                     }),
                 )
                 .await?;
@@ -958,14 +933,13 @@ async fn main() -> anyhow::Result<()> {
                 .dispatch::<(), ()>(Method::DELETE, format!("control/v1/node/{node_id}"), None)
                 .await?;
         }
-        Command::NodeStartDelete { node_id, force } => {
-            let query = if force {
-                format!("control/v1/node/{node_id}/delete?force=true")
-            } else {
-                format!("control/v1/node/{node_id}/delete")
-            };
+        Command::NodeStartDelete { node_id } => {
             storcon_client
-                .dispatch::<(), ()>(Method::PUT, query, None)
+                .dispatch::<(), ()>(
+                    Method::PUT,
+                    format!("control/v1/node/{node_id}/delete"),
+                    None,
+                )
                 .await?;
             println!("Delete started for {node_id}");
         }
@@ -1350,7 +1324,7 @@ async fn main() -> anyhow::Result<()> {
             concurrency,
         } => {
             let mut path = format!(
-                "v1/tenant/{tenant_shard_id}/timeline/{timeline_id}/download_heatmap_layers",
+                "/v1/tenant/{tenant_shard_id}/timeline/{timeline_id}/download_heatmap_layers",
             );
 
             if let Some(c) = concurrency {
@@ -1361,41 +1335,6 @@ async fn main() -> anyhow::Result<()> {
                 .dispatch::<(), ()>(Method::POST, path, None)
                 .await?;
         }
-        Command::TimelineLocate {
-            tenant_id,
-            timeline_id,
-        } => {
-            let path = format!("debug/v1/tenant/{tenant_id}/timeline/{timeline_id}/locate");
-
-            let resp = storcon_client
-                .dispatch::<(), TimelineLocateResponse>(Method::GET, path, None)
-                .await?;
-
-            let sk_set = resp.sk_set.iter().map(|id| id.0 as i64).collect::<Vec<_>>();
-            let new_sk_set = resp
-                .new_sk_set
-                .as_ref()
-                .map(|ids| ids.iter().map(|id| id.0 as i64).collect::<Vec<_>>());
-
-            println!("generation = {}", resp.generation);
-            println!("sk_set = {sk_set:?}");
-            println!("new_sk_set = {new_sk_set:?}");
-        }
-        Command::TimelineSafekeeperMigrate {
-            tenant_id,
-            timeline_id,
-            new_sk_set,
-        } => {
-            let path = format!("v1/tenant/{tenant_id}/timeline/{timeline_id}/safekeeper_migrate");
-
-            storcon_client
-                .dispatch::<_, ()>(
-                    Method::POST,
-                    path,
-                    Some(TimelineSafekeeperMigrateRequest { new_sk_set }),
-                )
-                .await?;
-        }
     }
 
     Ok(())

deny.toml

@@ -35,7 +35,6 @@ reason = "The paste crate is a build-only dependency with no runtime components.
 # More documentation for the licenses section can be found here:
 # https://embarkstudios.github.io/cargo-deny/checks/licenses/cfg.html
 [licenses]
-version = 2
 allow = [
     "0BSD",
     "Apache-2.0",

docs/continuous-profiling.md

@@ -0,0 +1,58 @@
+# Continuous Crofiling (Compute)
+
+The continuous profiling of the compute node is performed by `perf` or `bcc-tools`, the latter is preferred.
+
+The executables profiled are all the postgres-related ones only, excluding the actual compute code (Rust). This can be done as well but
+was not the main goal.
+
+## Tools
+
+The aforementioned tools are available within the same Docker image as
+the compute node itself, but the corresponding dependencies linux the
+linux kernel headers and the linux kernel itself are not and can't be
+for obvious reasons. To solve that, as we run the compute nodes as a
+virtual machine (qemu), we need to deliver these dependencies to it.
+This is done by the `autoscaling` part, which builds and deploys the
+kernel headers, needed modules, and the `perf` binary into an ext4-fs
+disk image, which is later attached to the VM and is symlinked to be
+made available for the compute node.
+
+## Output
+
+The output of the profiling is always a binary file in the same format
+of `pprof`. It can, however, be archived by `gzip` additionally, if the
+corresponding argument is provided in the JSON request.
+
+## REST API
+
+### Test profiling
+
+One can test the profiling after connecting to the VM and running:
+
+```sh
+curl -X POST -H "Content-Type: application/json" http://localhost:3080/profile/cpu -d '{"profiler": {"BccProfile": null}, "sampling_frequency": 99, "timeout_seconds": 5, "archive": false}' -v --output profile.pb
+```
+
+This uses the `Bcc` profiler and does not archive the output. The
+profiling data will be saved into the `profile.pb` file locally.
+
+**Only one profiling session can be run at a time.**
+
+To check the profiling status (to see whether it is already running or
+not), one can perform the `GET` request:
+
+```sh
+curl http://localhost:3080/profile/cpu -v
+```
+
+The profiling can be stopped by performing the `DELETE` request:
+
+```sh
+curl -X DELETE http://localhost:3080/profile/cpu -v
+```
+
+## Supported profiling data
+
+For now, only the CPU profiling is done and ther is no heap profiling.
+Also, only the postgres-related executables are tracked, the compute
+(Rust) part itself **is not tracked**.

docs/core_changes.md

@@ -129,10 +129,9 @@ segment to bootstrap the WAL writing, but it doesn't contain the checkpoint reco
 changes in xlog.c, to allow starting the compute node without reading the last checkpoint record
 from WAL.
 
-This includes code to read the `neon.signal` (also `zenith.signal`) file, which tells the startup 
-code the LSN to start at. When the `neon.signal` file is present, the startup uses that LSN
-instead of the last checkpoint's LSN. The system is known to be consistent at that LSN, without 
-any WAL redo.
+This includes code to read the `zenith.signal` file, which tells the startup code the LSN to start
+at. When the `zenith.signal` file is present, the startup uses that LSN instead of the last
+checkpoint's LSN. The system is known to be consistent at that LSN, without any WAL redo.
 
 
 ### How to get rid of the patch

docs/pageserver-services.md

@@ -75,7 +75,7 @@ CLI examples:
 * AWS S3  : `env AWS_ACCESS_KEY_ID='SOMEKEYAAAAASADSAH*#' AWS_SECRET_ACCESS_KEY='SOMEsEcReTsd292v' ${PAGESERVER_BIN} -c "remote_storage={bucket_name='some-sample-bucket',bucket_region='eu-north-1', prefix_in_bucket='/test_prefix/'}"`
 
 For Amazon AWS S3, a key id and secret access key could be located in `~/.aws/credentials` if awscli was ever configured to work with the desired bucket, on the AWS Settings page for a certain user. Also note, that the bucket names does not contain any protocols when used on AWS.
-For local S3 installations, refer to their documentation for name format and credentials.
+For local S3 installations, refer to the their documentation for name format and credentials.
 
 Similar to other pageserver settings, toml config file can be used to configure either of the storages as backup targets.
 Required sections are:

endpoint_storage/src/app.rs

@@ -233,7 +233,7 @@ mod tests {
                 .unwrap()
                 .as_millis();
             use rand::Rng;
-            let random = rand::rng().random::<u32>();
+            let random = rand::thread_rng().r#gen::<u32>();
 
             let s3_config = remote_storage::S3Config {
                 bucket_name: var(REAL_S3_BUCKET).unwrap(),

libs/compute_api/src/responses.rs

@@ -46,33 +46,16 @@ pub struct ExtensionInstallResponse {
     pub version: ExtVersion,
 }
 
-/// Status of the LFC prewarm process. The same state machine is reused for
-/// both autoprewarm (prewarm after compute/Postgres start using the previously
-/// stored LFC state) and explicit prewarming via API.
 #[derive(Serialize, Default, Debug, Clone, PartialEq)]
 #[serde(tag = "status", rename_all = "snake_case")]
 pub enum LfcPrewarmState {
-    /// Default value when compute boots up.
     #[default]
     NotPrewarmed,
-    /// Prewarming thread is active and loading pages into LFC.
     Prewarming,
-    /// We found requested LFC state in the endpoint storage and
-    /// completed prewarming successfully.
     Completed,
-    /// Unexpected error happened during prewarming. Note, `Not Found 404`
-    /// response from the endpoint storage is explicitly excluded here
-    /// because it can normally happen on the first compute start,
-    /// since LFC state is not available yet.
-    Failed { error: String },
-    /// We tried to fetch the corresponding LFC state from the endpoint storage,
-    /// but received `Not Found 404`. This should normally happen only during the
-    /// first endpoint start after creation with `autoprewarm: true`.
-    ///
-    /// During the orchestrated prewarm via API, when a caller explicitly
-    /// provides the LFC state key to prewarm from, it's the caller responsibility
-    /// to handle this status as an error state in this case.
-    Skipped,
+    Failed {
+        error: String,
+    },
 }
 
 impl Display for LfcPrewarmState {
@@ -81,7 +64,6 @@ impl Display for LfcPrewarmState {
             LfcPrewarmState::NotPrewarmed => f.write_str("NotPrewarmed"),
             LfcPrewarmState::Prewarming => f.write_str("Prewarming"),
             LfcPrewarmState::Completed => f.write_str("Completed"),
-            LfcPrewarmState::Skipped => f.write_str("Skipped"),
             LfcPrewarmState::Failed { error } => write!(f, "Error({error})"),
         }
     }

libs/compute_api/src/spec.rs

@@ -416,32 +416,6 @@ pub struct GenericOption {
     pub vartype: String,
 }
 
-/// Postgres compute TLS settings.
-#[derive(Clone, Debug, Deserialize, Serialize, PartialEq)]
-pub struct PgComputeTlsSettings {
-    // Absolute path to the certificate file for server-side TLS.
-    pub cert_file: String,
-    // Absolute path to the private key file for server-side TLS.
-    pub key_file: String,
-    // Absolute path to the certificate authority file for verifying client certificates.
-    pub ca_file: String,
-}
-
-/// Databricks specific options for compute instance.
-/// This is used to store any other settings that needs to be propagate to Compute
-/// but should not be persisted to ComputeSpec in the database.
-#[derive(Clone, Debug, Deserialize, Serialize, PartialEq)]
-pub struct DatabricksSettings {
-    pub pg_compute_tls_settings: PgComputeTlsSettings,
-    // Absolute file path to databricks_pg_hba.conf file.
-    pub databricks_pg_hba: String,
-    // Absolute file path to databricks_pg_ident.conf file.
-    pub databricks_pg_ident: String,
-    // Hostname portion of the Databricks workspace URL of the endpoint, or empty string if not known.
-    // A valid hostname is required for the compute instance to support PAT logins.
-    pub databricks_workspace_host: String,
-}
-
 /// Optional collection of `GenericOption`'s. Type alias allows us to
 /// declare a `trait` on it.
 pub type GenericOptions = Option<Vec<GenericOption>>;

libs/consumption_metrics/src/lib.rs

@@ -90,7 +90,7 @@ impl<'a> IdempotencyKey<'a> {
         IdempotencyKey {
             now: Utc::now(),
             node_id,
-            nonce: rand::rng().random_range(0..=9999),
+            nonce: rand::thread_rng().gen_range(0..=9999),
         }
     }
 

libs/desim/src/node_os.rs

@@ -41,7 +41,7 @@ impl NodeOs {
 
     /// Generate a random number in range [0, max).
     pub fn random(&self, max: u64) -> u64 {
-        self.internal.rng.lock().random_range(0..max)
+        self.internal.rng.lock().gen_range(0..max)
     }
 
     /// Append a new event to the world event log.

libs/desim/src/options.rs

@@ -32,10 +32,10 @@ impl Delay {
     /// Generate a random delay in range [min, max]. Return None if the
     /// message should be dropped.
     pub fn delay(&self, rng: &mut StdRng) -> Option<u64> {
-        if rng.random_bool(self.fail_prob) {
+        if rng.gen_bool(self.fail_prob) {
             return None;
         }
-        Some(rng.random_range(self.min..=self.max))
+        Some(rng.gen_range(self.min..=self.max))
     }
 }