fix workspace hack

Signed-off-by: Alex Chi Z <chi@neon.tech>

.dockerignore

@@ -19,7 +19,7 @@
 !pageserver/
 !pgxn/
 !proxy/
-!endpoint_storage/
+!object_storage/
 !storage_scrubber/
 !safekeeper/
 !storage_broker/

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

@@ -113,6 +113,8 @@ runs:
         TEST_OUTPUT: /tmp/test_output
         BUILD_TYPE: ${{ inputs.build_type }}
         COMPATIBILITY_SNAPSHOT_DIR: /tmp/compatibility_snapshot_pg${{ inputs.pg_version }}
+        ALLOW_BACKWARD_COMPATIBILITY_BREAKAGE: contains(github.event.pull_request.labels.*.name, 'backward compatibility breakage')
+        ALLOW_FORWARD_COMPATIBILITY_BREAKAGE: contains(github.event.pull_request.labels.*.name, 'forward compatibility breakage')
         RERUN_FAILED: ${{ inputs.rerun_failed }}
         PG_VERSION: ${{ inputs.pg_version }}
         SANITIZERS: ${{ inputs.sanitizers }}
@@ -133,7 +135,6 @@ runs:
         fi
 
         PERF_REPORT_DIR="$(realpath test_runner/perf-report-local)"
-        echo "PERF_REPORT_DIR=${PERF_REPORT_DIR}" >> ${GITHUB_ENV}
         rm -rf $PERF_REPORT_DIR
 
         TEST_SELECTION="test_runner/${{ inputs.test_selection }}"
@@ -210,12 +211,11 @@ runs:
           --verbose \
           -rA $TEST_SELECTION $EXTRA_PARAMS
 
-    - name: Upload performance report
-      if: ${{ !cancelled() && inputs.save_perf_report == 'true' }}
-      shell: bash -euxo pipefail {0}
-      run: |
-        export REPORT_FROM="${PERF_REPORT_DIR}"
-        scripts/generate_and_push_perf_report.sh
+        if [[ "${{ inputs.save_perf_report }}" == "true" ]]; then
+          export REPORT_FROM="$PERF_REPORT_DIR"
+          export REPORT_TO="$PLATFORM"
+          scripts/generate_and_push_perf_report.sh
+        fi
 
     - name: Upload compatibility snapshot
       # Note, that we use `github.base_ref` which is a target branch for a PR

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

@@ -272,13 +272,10 @@ jobs:
           # run pageserver tests with different settings
           for get_vectored_concurrent_io in sequential sidecar-task; do
             for io_engine in std-fs tokio-epoll-uring ; do
-                for io_mode in buffered direct direct-rw ; do
-                  NEON_PAGESERVER_UNIT_TEST_GET_VECTORED_CONCURRENT_IO=$get_vectored_concurrent_io \
-                  NEON_PAGESERVER_UNIT_TEST_VIRTUAL_FILE_IOENGINE=$io_engine \
-                  NEON_PAGESERVER_UNIT_TEST_VIRTUAL_FILE_IO_MODE=$io_mode \
-                  ${cov_prefix} \
-                  cargo nextest run $CARGO_FLAGS $CARGO_FEATURES  -E 'package(pageserver)'
-              done
+              NEON_PAGESERVER_UNIT_TEST_GET_VECTORED_CONCURRENT_IO=$get_vectored_concurrent_io \
+                NEON_PAGESERVER_UNIT_TEST_VIRTUAL_FILE_IOENGINE=$io_engine \
+                ${cov_prefix} \
+                cargo nextest run $CARGO_FLAGS $CARGO_FEATURES  -E 'package(pageserver)'
             done
           done
 
@@ -349,7 +346,7 @@ jobs:
       contents: read
       statuses: write
     needs: [ build-neon ]
-    runs-on: ${{ fromJSON(format('["self-hosted", "{0}"]', inputs.arch == 'arm64' && 'large-arm64' || 'large-metal')) }}
+    runs-on: ${{ fromJSON(format('["self-hosted", "{0}"]', inputs.arch == 'arm64' && 'large-arm64' || 'large')) }}
     container:
       image: ${{ inputs.build-tools-image }}
       credentials:
@@ -395,7 +392,6 @@ jobs:
           BUILD_TAG: ${{ inputs.build-tag }}
           PAGESERVER_VIRTUAL_FILE_IO_ENGINE: tokio-epoll-uring
           PAGESERVER_GET_VECTORED_CONCURRENT_IO: sidecar-task
-          PAGESERVER_VIRTUAL_FILE_IO_MODE: direct-rw
           USE_LFC: ${{ matrix.lfc_state == 'with-lfc' && 'true' || 'false' }}
 
       # Temporary disable this step until we figure out why it's so flaky

.github/workflows/_meta.yml

@@ -165,5 +165,5 @@ jobs:
           GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
           CURRENT_SHA: ${{ github.sha }}
         run: |
-          RELEASE_PR_RUN_ID=$(gh api "/repos/${GITHUB_REPOSITORY}/actions/runs?head_sha=$CURRENT_SHA" | jq '[.workflow_runs[] | select(.name == "Build and Test") | select(.head_branch | test("^rc/release.*$"; "s"))] | first | .id // ("Failed to find Build and Test run from  RC PR!" | halt_error(1))')
+          RELEASE_PR_RUN_ID=$(gh api "/repos/${GITHUB_REPOSITORY}/actions/runs?head_sha=$CURRENT_SHA" | jq '[.workflow_runs[] | select(.name == "Build and Test") | select(.head_branch | test("^rc/release(-(proxy|compute))?/[0-9]{4}-[0-9]{2}-[0-9]{2}$"; "s"))] | first | .id // ("Failed to find Build and Test run from  RC PR!" | halt_error(1))')
           echo "release-pr-run-id=$RELEASE_PR_RUN_ID" | tee -a $GITHUB_OUTPUT

.github/workflows/build_and_test.yml

@@ -323,8 +323,6 @@ jobs:
           PERF_TEST_RESULT_CONNSTR: "${{ secrets.PERF_TEST_RESULT_CONNSTR }}"
           TEST_RESULT_CONNSTR: "${{ secrets.REGRESS_TEST_RESULT_CONNSTR_NEW }}"
           PAGESERVER_VIRTUAL_FILE_IO_ENGINE: tokio-epoll-uring
-          PAGESERVER_GET_VECTORED_CONCURRENT_IO: sidecar-task
-          PAGESERVER_VIRTUAL_FILE_IO_MODE: direct-rw
           SYNC_BETWEEN_TESTS: true
       # XXX: no coverage data handling here, since benchmarks are run on release builds,
       # while coverage is currently collected for the debug ones
@@ -1238,7 +1236,7 @@ jobs:
         env:
           GH_TOKEN: ${{ secrets.CI_ACCESS_TOKEN }}
         run: |
-          TIMEOUT=5400 # 90 minutes, usually it takes ~2-3 minutes, but if runners are busy, it might take longer
+          TIMEOUT=1800 # 30 minutes, usually it takes ~2-3 minutes, but if runners are busy, it might take longer
           INTERVAL=15 # try each N seconds
 
           last_status="" # a variable to carry the last status of the "build-and-upload-extensions" context

.github/workflows/fast-forward.yml

@@ -27,17 +27,15 @@ jobs:
       - name: Fast forwarding
         uses: sequoia-pgp/fast-forward@ea7628bedcb0b0b96e94383ada458d812fca4979
         # See https://docs.github.com/en/graphql/reference/enums#mergestatestatus
-        if: ${{ contains(fromJSON('["clean", "unstable"]'), github.event.pull_request.mergeable_state) }}
+        if: ${{ github.event.pull_request.mergeable_state  == 'clean' }}
         with:
           merge: true
           comment: on-error
           github_token: ${{ secrets.CI_ACCESS_TOKEN }}
 
       - name: Comment if mergeable_state is not clean
-        if: ${{ !contains(fromJSON('["clean", "unstable"]'), github.event.pull_request.mergeable_state) }}
-        env:
-          GH_TOKEN: ${{ secrets.CI_ACCESS_TOKEN }}
+        if: ${{ github.event.pull_request.mergeable_state  != 'clean' }}
         run: |
           gh pr comment ${{ github.event.pull_request.number }} \
             --repo "${GITHUB_REPOSITORY}" \
-            --body "Not trying to forward pull-request, because \`mergeable_state\` is \`${{ github.event.pull_request.mergeable_state }}\`, not \`clean\` or \`unstable\`."
+            --body "Not trying to forward pull-request, because \`mergeable_state\` is \`${{ github.event.pull_request.mergeable_state }}\`, not \`clean\`."

.github/workflows/pg-clients.yml

@@ -30,7 +30,7 @@ permissions:
   statuses: write # require for posting a status update
 
 env:
-  DEFAULT_PG_VERSION: 17
+  DEFAULT_PG_VERSION: 16
   PLATFORM: neon-captest-new
   AWS_DEFAULT_REGION: eu-central-1
 
@@ -42,8 +42,6 @@ jobs:
       github-event-name: ${{ github.event_name }}
 
   build-build-tools-image:
-    permissions:
-      packages: write
     needs: [ check-permissions ]
     uses: ./.github/workflows/build-build-tools-image.yml
     secrets: inherit

.github/workflows/random-ops-test.yml

@@ -1,93 +0,0 @@
-name: Random Operations Test
-
-on:
-  schedule:
-    # * is a special character in YAML so you have to quote this string
-    #          ┌───────────── minute (0 - 59)
-    #          │  ┌───────────── hour (0 - 23)
-    #          │  │  ┌───────────── day of the month (1 - 31)
-    #          │  │  │ ┌───────────── month (1 - 12 or JAN-DEC)
-    #          │  │  │ │ ┌───────────── day of the week (0 - 6 or SUN-SAT)
-    - cron:  '23 */2 * * *' # runs every 2 hours
-  workflow_dispatch:
-    inputs:
-      random_seed:
-        type: number
-        description: 'The random seed'
-        required: false
-        default: 0
-      num_operations:
-        type: number
-        description: "The number of operations to test"
-        default: 250
-
-defaults:
-  run:
-    shell: bash -euxo pipefail {0}
-
-permissions: {}
-
-env:
-  DEFAULT_PG_VERSION: 16
-  PLATFORM: neon-captest-new
-  AWS_DEFAULT_REGION: eu-central-1
-
-jobs:
-  run-random-rests:
-    env:
-      POSTGRES_DISTRIB_DIR: /tmp/neon/pg_install
-    runs-on: small
-    permissions:
-      id-token: write
-      statuses: write
-
-    strategy:
-      fail-fast: false
-      matrix:
-        pg-version: [16, 17]
-
-    container:
-      image: ghcr.io/neondatabase/build-tools:pinned-bookworm
-      credentials:
-        username: ${{ github.actor }}
-        password: ${{ secrets.GITHUB_TOKEN }}
-      options: --init
-    steps:
-      - name: Harden the runner (Audit all outbound calls)
-        uses: step-security/harden-runner@4d991eb9b905ef189e4c376166672c3f2f230481 # v2.11.0
-        with:
-          egress-policy: audit
-
-      - uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2
-
-      - name: Download Neon artifact
-        uses: ./.github/actions/download
-        with:
-          name: neon-${{ runner.os }}-${{ runner.arch }}-release-artifact
-          path: /tmp/neon/
-          prefix: latest
-          aws-oicd-role-arn: ${{ vars.DEV_AWS_OIDC_ROLE_ARN }}
-
-      - name: Run tests
-        uses: ./.github/actions/run-python-test-set
-        with:
-          build_type: remote
-          test_selection: random_ops
-          run_in_parallel: false
-          extra_params: -m remote_cluster
-          pg_version: ${{ matrix.pg-version }}
-          aws-oicd-role-arn: ${{ vars.DEV_AWS_OIDC_ROLE_ARN }}
-        env:
-          NEON_API_KEY: ${{ secrets.NEON_STAGING_API_KEY }}
-          RANDOM_SEED: ${{ inputs.random_seed }}
-          NUM_OPERATIONS: ${{ inputs.num_operations }}
-
-      - name: Create Allure report
-        if: ${{ !cancelled() }}
-        id: create-allure-report
-        uses: ./.github/actions/allure-report-generate
-        with:
-          store-test-results-into-db: true
-          aws-oicd-role-arn: ${{ vars.DEV_AWS_OIDC_ROLE_ARN }}
-        env:
-          REGRESS_TEST_RESULT_CONNSTR_NEW: ${{ secrets.REGRESS_TEST_RESULT_CONNSTR_NEW }}

Cargo.lock

@@ -40,7 +40,7 @@ dependencies = [
  "getrandom 0.2.11",
  "once_cell",
  "version_check",
- "zerocopy 0.7.31",
+ "zerocopy",
 ]
 
 [[package]]
@@ -1416,7 +1416,6 @@ name = "control_plane"
 version = "0.1.0"
 dependencies = [
  "anyhow",
- "base64 0.13.1",
  "camino",
  "clap",
  "comfy-table",
@@ -1426,12 +1425,10 @@ dependencies = [
  "humantime",
  "humantime-serde",
  "hyper 0.14.30",
- "jsonwebtoken",
  "nix 0.27.1",
  "once_cell",
  "pageserver_api",
  "pageserver_client",
- "pem",
  "postgres_backend",
  "postgres_connection",
  "regex",
@@ -1440,8 +1437,6 @@ dependencies = [
  "scopeguard",
  "serde",
  "serde_json",
- "sha2",
- "spki 0.7.3",
  "storage_broker",
  "thiserror 1.0.69",
  "tokio",
@@ -2037,33 +2032,6 @@ dependencies = [
  "zeroize",
 ]
 
-[[package]]
-name = "endpoint_storage"
-version = "0.0.1"
-dependencies = [
- "anyhow",
- "axum",
- "axum-extra",
- "camino",
- "camino-tempfile",
- "futures",
- "http-body-util",
- "itertools 0.10.5",
- "jsonwebtoken",
- "prometheus",
- "rand 0.8.5",
- "remote_storage",
- "serde",
- "serde_json",
- "test-log",
- "tokio",
- "tokio-util",
- "tower 0.5.2",
- "tracing",
- "utils",
- "workspace_hack",
-]
-
 [[package]]
 name = "enum-map"
 version = "2.5.0"
@@ -2849,7 +2817,6 @@ dependencies = [
  "hyper 0.14.30",
  "itertools 0.10.5",
  "jemalloc_pprof",
- "jsonwebtoken",
  "metrics",
  "once_cell",
  "pprof",
@@ -4025,6 +3992,33 @@ dependencies = [
  "memchr",
 ]
 
+[[package]]
+name = "object_storage"
+version = "0.0.1"
+dependencies = [
+ "anyhow",
+ "axum",
+ "axum-extra",
+ "camino",
+ "camino-tempfile",
+ "futures",
+ "http-body-util",
+ "itertools 0.10.5",
+ "jsonwebtoken",
+ "prometheus",
+ "rand 0.8.5",
+ "remote_storage",
+ "serde",
+ "serde_json",
+ "test-log",
+ "tokio",
+ "tokio-util",
+ "tower 0.5.2",
+ "tracing",
+ "utils",
+ "workspace_hack",
+]
+
 [[package]]
 name = "once_cell"
 version = "1.20.2"
@@ -4275,7 +4269,6 @@ dependencies = [
  "hyper 0.14.30",
  "indoc",
  "itertools 0.10.5",
- "jsonwebtoken",
  "md5",
  "metrics",
  "nix 0.27.1",
@@ -4285,7 +4278,6 @@ dependencies = [
  "pageserver_api",
  "pageserver_client",
  "pageserver_compaction",
- "pem",
  "pin-project-lite",
  "postgres-protocol",
  "postgres-types",
@@ -4302,7 +4294,6 @@ dependencies = [
  "remote_storage",
  "reqwest",
  "rpds",
- "rstest",
  "rustls 0.23.18",
  "scopeguard",
  "send-future",
@@ -4354,7 +4345,6 @@ dependencies = [
  "humantime-serde",
  "itertools 0.10.5",
  "nix 0.27.1",
- "once_cell",
  "postgres_backend",
  "postgres_ffi",
  "rand 0.8.5",
@@ -4416,9 +4406,9 @@ dependencies = [
 
 [[package]]
 name = "papaya"
-version = "0.2.1"
+version = "0.2.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "6827e3fc394523c21d4464d02c0bb1c19966ea4a58a9844ad6d746214179d2bc"
+checksum = "aab21828b6b5952fdadd6c377728ffae53ec3a21b2febc47319ab65741f7e2fd"
 dependencies = [
  "equivalent",
  "seize",
@@ -5205,7 +5195,7 @@ dependencies = [
  "walkdir",
  "workspace_hack",
  "x509-cert",
- "zerocopy 0.8.24",
+ "zerocopy",
 ]
 
 [[package]]
@@ -5505,6 +5495,16 @@ version = "1.9.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "c707298afce11da2efef2f600116fa93ffa7a032b5d7b628aa17711ec81383ca"
 
+[[package]]
+name = "remote_keys"
+version = "0.1.0"
+dependencies = [
+ "anyhow",
+ "rand 0.8.5",
+ "utils",
+ "workspace_hack",
+]
+
 [[package]]
 name = "remote_storage"
 version = "0.1.0"
@@ -5520,6 +5520,7 @@ dependencies = [
  "azure_identity",
  "azure_storage",
  "azure_storage_blobs",
+ "base64 0.13.1",
  "bytes",
  "camino",
  "camino-tempfile",
@@ -5530,6 +5531,7 @@ dependencies = [
  "humantime-serde",
  "hyper 1.4.1",
  "itertools 0.10.5",
+ "md5",
  "metrics",
  "once_cell",
  "pin-project-lite",
@@ -5595,7 +5597,7 @@ dependencies = [
  "wasm-bindgen-futures",
  "wasm-streams",
  "web-sys",
- "webpki-roots",
+ "webpki-roots 0.26.1",
  "winreg",
 ]
 
@@ -5695,9 +5697,9 @@ dependencies = [
 
 [[package]]
 name = "ring"
-version = "0.17.14"
+version = "0.17.13"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "a4689e6c2294d81e88dc6261c768b63bc4fcdb852be6d1352498b114f61383b7"
+checksum = "70ac5d832aa16abd7d1def883a8545280c20a60f523a370aa3a9617c2b8550ee"
 dependencies = [
  "cc",
  "cfg-if",
@@ -5998,12 +6000,10 @@ dependencies = [
  "humantime",
  "hyper 0.14.30",
  "itertools 0.10.5",
- "jsonwebtoken",
  "metrics",
  "once_cell",
  "pageserver_api",
  "parking_lot 0.12.1",
- "pem",
  "postgres-protocol",
  "postgres_backend",
  "postgres_ffi",
@@ -6196,13 +6196,13 @@ checksum = "224e328af6e080cddbab3c770b1cf50f0351ba0577091ef2410c3951d835ff87"
 
 [[package]]
 name = "sentry"
-version = "0.37.0"
+version = "0.32.3"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "255914a8e53822abd946e2ce8baa41d4cded6b8e938913b7f7b9da5b7ab44335"
+checksum = "00421ed8fa0c995f07cde48ba6c89e80f2b312f74ff637326f392fbfd23abe02"
 dependencies = [
  "httpdate",
  "reqwest",
- "rustls 0.23.18",
+ "rustls 0.21.12",
  "sentry-backtrace",
  "sentry-contexts",
  "sentry-core",
@@ -6210,14 +6210,14 @@ dependencies = [
  "sentry-tracing",
  "tokio",
  "ureq",
- "webpki-roots",
+ "webpki-roots 0.25.2",
 ]
 
 [[package]]
 name = "sentry-backtrace"
-version = "0.37.0"
+version = "0.32.3"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "00293cd332a859961f24fd69258f7e92af736feaeb91020cff84dac4188a4302"
+checksum = "a79194074f34b0cbe5dd33896e5928bbc6ab63a889bd9df2264af5acb186921e"
 dependencies = [
  "backtrace",
  "once_cell",
@@ -6227,9 +6227,9 @@ dependencies = [
 
 [[package]]
 name = "sentry-contexts"
-version = "0.37.0"
+version = "0.32.3"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "961990f9caa76476c481de130ada05614cd7f5aa70fb57c2142f0e09ad3fb2aa"
+checksum = "eba8870c5dba2bfd9db25c75574a11429f6b95957b0a78ac02e2970dd7a5249a"
 dependencies = [
  "hostname",
  "libc",
@@ -6241,9 +6241,9 @@ dependencies = [
 
 [[package]]
 name = "sentry-core"
-version = "0.37.0"
+version = "0.32.3"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "1a6409d845707d82415c800290a5d63be5e3df3c2e417b0997c60531dfbd35ef"
+checksum = "46a75011ea1c0d5c46e9e57df03ce81f5c7f0a9e199086334a1f9c0a541e0826"
 dependencies = [
  "once_cell",
  "rand 0.8.5",
@@ -6254,9 +6254,9 @@ dependencies = [
 
 [[package]]
 name = "sentry-panic"
-version = "0.37.0"
+version = "0.32.3"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "609b1a12340495ce17baeec9e08ff8ed423c337c1a84dffae36a178c783623f3"
+checksum = "2eaa3ecfa3c8750c78dcfd4637cfa2598b95b52897ed184b4dc77fcf7d95060d"
 dependencies = [
  "sentry-backtrace",
  "sentry-core",
@@ -6264,9 +6264,9 @@ dependencies = [
 
 [[package]]
 name = "sentry-tracing"
-version = "0.37.0"
+version = "0.32.3"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "49f4e86402d5c50239dc7d8fd3f6d5e048221d5fcb4e026d8d50ab57fe4644cb"
+checksum = "f715932bf369a61b7256687c6f0554141b7ce097287e30e3f7ed6e9de82498fe"
 dependencies = [
  "sentry-backtrace",
  "sentry-core",
@@ -6276,9 +6276,9 @@ dependencies = [
 
 [[package]]
 name = "sentry-types"
-version = "0.37.0"
+version = "0.32.3"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "3d3f117b8755dbede8260952de2aeb029e20f432e72634e8969af34324591631"
+checksum = "4519c900ce734f7a0eb7aba0869dfb225a7af8820634a7dd51449e3b093cfb7c"
 dependencies = [
  "debugid",
  "hex",
@@ -6712,6 +6712,8 @@ version = "0.1.0"
 dependencies = [
  "anyhow",
  "async-stream",
+ "aws-config",
+ "aws-sdk-s3",
  "camino",
  "chrono",
  "clap",
@@ -7800,7 +7802,7 @@ dependencies = [
  "rustls 0.23.18",
  "rustls-pki-types",
  "url",
- "webpki-roots",
+ "webpki-roots 0.26.1",
 ]
 
 [[package]]
@@ -7882,7 +7884,6 @@ dependencies = [
  "metrics",
  "nix 0.27.1",
  "once_cell",
- "pem",
  "pin-project-lite",
  "postgres_connection",
  "pprof",
@@ -8168,6 +8169,12 @@ dependencies = [
  "wasm-bindgen",
 ]
 
+[[package]]
+name = "webpki-roots"
+version = "0.25.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "14247bb57be4f377dfb94c72830b8ce8fc6beac03cf4bf7b9732eadd414123fc"
+
 [[package]]
 name = "webpki-roots"
 version = "0.26.1"
@@ -8475,8 +8482,6 @@ dependencies = [
  "regex-syntax 0.8.2",
  "reqwest",
  "rustls 0.23.18",
- "rustls-pki-types",
- "rustls-webpki 0.102.8",
  "scopeguard",
  "sec1 0.7.3",
  "serde",
@@ -8505,6 +8510,7 @@ dependencies = [
  "tracing-log",
  "url",
  "uuid",
+ "zerocopy",
  "zeroize",
  "zstd",
  "zstd-safe",
@@ -8608,16 +8614,8 @@ version = "0.7.31"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "1c4061bedbb353041c12f413700357bec76df2c7e2ca8e4df8bac24c6bf68e3d"
 dependencies = [
- "zerocopy-derive 0.7.31",
-]
-
-[[package]]
-name = "zerocopy"
-version = "0.8.24"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "2586fea28e186957ef732a5f8b3be2da217d65c5969d4b1e17f973ebbe876879"
-dependencies = [
- "zerocopy-derive 0.8.24",
+ "byteorder",
+ "zerocopy-derive",
 ]
 
 [[package]]
@@ -8631,17 +8629,6 @@ dependencies = [
  "syn 2.0.100",
 ]
 
-[[package]]
-name = "zerocopy-derive"
-version = "0.8.24"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "a996a8f63c5c4448cd959ac1bab0aaa3306ccfd060472f85943ee0750f0169be"
-dependencies = [
- "proc-macro2",
- "quote",
- "syn 2.0.100",
-]
-
 [[package]]
 name = "zerofrom"
 version = "0.1.5"

Cargo.toml

@@ -30,6 +30,7 @@ members = [
     "libs/tenant_size_model",
     "libs/metrics",
     "libs/postgres_connection",
+    "libs/remote_keys",
     "libs/remote_storage",
     "libs/tracing-utils",
     "libs/postgres_ffi/wal_craft",
@@ -40,7 +41,7 @@ members = [
     "libs/proxy/postgres-protocol2",
     "libs/proxy/postgres-types2",
     "libs/proxy/tokio-postgres2",
-    "endpoint_storage",
+    "object_storage",
 ]
 
 [workspace.package]
@@ -141,7 +142,6 @@ parking_lot = "0.12"
 parquet = { version = "53", default-features = false, features = ["zstd"] }
 parquet_derive = "53"
 pbkdf2 = { version = "0.12.1", features = ["simple", "std"] }
-pem = "3.0.3"
 pin-project-lite = "0.2"
 pprof = { version = "0.14", features = ["criterion", "flamegraph", "frame-pointer", "prost-codec"] }
 procfs = "0.16"
@@ -164,7 +164,7 @@ scopeguard = "1.1"
 sysinfo = "0.29.2"
 sd-notify = "0.4.1"
 send-future = "0.1.0"
-sentry = { version = "0.37", default-features = false, features = ["backtrace", "contexts", "panic", "rustls", "reqwest" ] }
+sentry = { version = "0.32", default-features = false, features = ["backtrace", "contexts", "panic", "rustls", "reqwest" ] }
 serde = { version = "1.0", features = ["derive"] }
 serde_json = "1"
 serde_path_to_error = "0.1"
@@ -175,7 +175,6 @@ signal-hook = "0.3"
 smallvec = "1.11"
 smol_str = { version = "0.2.0", features = ["serde"] }
 socket2 = "0.5"
-spki = "0.7.3"
 strum = "0.26"
 strum_macros = "0.26"
 "subtle"  = "2.5.0"
@@ -220,7 +219,7 @@ uuid = { version = "1.6.1", features = ["v4", "v7", "serde"] }
 walkdir = "2.3.2"
 rustls-native-certs = "0.8"
 whoami = "1.5.1"
-zerocopy = { version = "0.8", features = ["derive", "simd"] }
+zerocopy = { version = "0.7", features = ["derive"] }
 json-structural-diff = { version = "0.2.0" }
 x509-cert = { version = "0.2.5" }
 

Dockerfile

@@ -89,7 +89,7 @@ RUN set -e \
       --bin storage_broker  \
       --bin storage_controller  \
       --bin proxy  \
-      --bin endpoint_storage \
+      --bin object_storage \
       --bin neon_local \
       --bin storage_scrubber \
       --locked --release
@@ -122,7 +122,7 @@ COPY --from=build --chown=neon:neon /home/nonroot/target/release/safekeeper
 COPY --from=build --chown=neon:neon /home/nonroot/target/release/storage_broker      /usr/local/bin
 COPY --from=build --chown=neon:neon /home/nonroot/target/release/storage_controller  /usr/local/bin
 COPY --from=build --chown=neon:neon /home/nonroot/target/release/proxy               /usr/local/bin
-COPY --from=build --chown=neon:neon /home/nonroot/target/release/endpoint_storage    /usr/local/bin
+COPY --from=build --chown=neon:neon /home/nonroot/target/release/object_storage      /usr/local/bin
 COPY --from=build --chown=neon:neon /home/nonroot/target/release/neon_local          /usr/local/bin
 COPY --from=build --chown=neon:neon /home/nonroot/target/release/storage_scrubber    /usr/local/bin
 

README.md

@@ -270,7 +270,7 @@ By default, this runs both debug and release modes, and all supported postgres v
 testing locally, it is convenient to run just one set of permutations, like this:
 
 ```sh
-DEFAULT_PG_VERSION=17 BUILD_TYPE=release ./scripts/pytest
+DEFAULT_PG_VERSION=16 BUILD_TYPE=release ./scripts/pytest
 ```
 
 ## Flamegraphs

build-tools.Dockerfile

@@ -173,7 +173,7 @@ RUN curl -fsSL "https://github.com/protocolbuffers/protobuf/releases/download/v$
     && rm -rf protoc.zip protoc
 
 # s5cmd
-ENV S5CMD_VERSION=2.3.0
+ENV S5CMD_VERSION=2.2.2
 RUN curl -sL "https://github.com/peak/s5cmd/releases/download/v${S5CMD_VERSION}/s5cmd_${S5CMD_VERSION}_Linux-$(uname -m | sed 's/x86_64/64bit/g' | sed 's/aarch64/arm64/g').tar.gz" | tar zxvf - s5cmd \
     && chmod +x s5cmd \
     && mv s5cmd /usr/local/bin/s5cmd
@@ -206,7 +206,7 @@ RUN curl "https://awscli.amazonaws.com/awscli-exe-linux-$(uname -m).zip" -o "aws
     && rm awscliv2.zip
 
 # Mold: A Modern Linker
-ENV MOLD_VERSION=v2.37.1
+ENV MOLD_VERSION=v2.34.1
 RUN set -e \
     && git clone https://github.com/rui314/mold.git \
     && mkdir mold/build \
@@ -268,7 +268,7 @@ WORKDIR /home/nonroot
 RUN echo -e "--retry-connrefused\n--connect-timeout 15\n--retry 5\n--max-time 300\n" > /home/nonroot/.curlrc
 
 # Python
-ENV PYTHON_VERSION=3.11.12 \
+ENV PYTHON_VERSION=3.11.10 \
     PYENV_ROOT=/home/nonroot/.pyenv \
     PATH=/home/nonroot/.pyenv/shims:/home/nonroot/.pyenv/bin:/home/nonroot/.poetry/bin:$PATH
 RUN set -e \
@@ -296,12 +296,12 @@ ENV RUSTC_VERSION=1.86.0
 ENV RUSTUP_HOME="/home/nonroot/.rustup"
 ENV PATH="/home/nonroot/.cargo/bin:${PATH}"
 ARG RUSTFILT_VERSION=0.2.1
-ARG CARGO_HAKARI_VERSION=0.9.36
-ARG CARGO_DENY_VERSION=0.18.2
-ARG CARGO_HACK_VERSION=0.6.36
-ARG CARGO_NEXTEST_VERSION=0.9.94
+ARG CARGO_HAKARI_VERSION=0.9.33
+ARG CARGO_DENY_VERSION=0.16.2
+ARG CARGO_HACK_VERSION=0.6.33
+ARG CARGO_NEXTEST_VERSION=0.9.85
 ARG CARGO_CHEF_VERSION=0.1.71
-ARG CARGO_DIESEL_CLI_VERSION=2.2.9
+ARG CARGO_DIESEL_CLI_VERSION=2.2.6
 RUN curl -sSO https://static.rust-lang.org/rustup/dist/$(uname -m)-unknown-linux-gnu/rustup-init && whoami && \
 	chmod +x rustup-init && \
 	./rustup-init -y --default-toolchain ${RUSTC_VERSION} && \

clippy.toml

@@ -12,5 +12,3 @@ disallowed-macros = [
     # cannot disallow this, because clippy finds used from tokio macros
     #"tokio::pin",
 ]
-
-allow-unwrap-in-tests = true

compute/compute-node.Dockerfile

@@ -1677,7 +1677,7 @@ RUN set -e \
     && apt clean && rm -rf /var/lib/apt/lists/*
 
 # Use `dist_man_MANS=` to skip manpage generation (which requires python3/pandoc)
-ENV PGBOUNCER_TAG=pgbouncer_1_24_1
+ENV PGBOUNCER_TAG=pgbouncer_1_22_1
 RUN set -e \
     && git clone --recurse-submodules --depth 1 --branch ${PGBOUNCER_TAG} https://github.com/pgbouncer/pgbouncer.git pgbouncer \
     && cd pgbouncer \

compute/patches/pg_anon.patch

@@ -0,0 +1,265 @@
+commit 00aa659afc9c7336ab81036edec3017168aabf40
+Author: Heikki Linnakangas <heikki@neon.tech>
+Date:   Tue Nov 12 16:59:19 2024 +0200
+
+    Temporarily disable test that depends on timezone
+
+diff --git a/tests/expected/generalization.out b/tests/expected/generalization.out
+index 23ef5fa..9e60deb 100644
+--- a/ext-src/pg_anon-src/tests/expected/generalization.out
++++ b/ext-src/pg_anon-src/tests/expected/generalization.out
+@@ -284,12 +284,9 @@ SELECT anon.generalize_tstzrange('19041107','century');
+  ["Tue Jan 01 00:00:00 1901 PST","Mon Jan 01 00:00:00 2001 PST")
+ (1 row)
+ 
+-SELECT anon.generalize_tstzrange('19041107','millennium');
+-                      generalize_tstzrange                       
+------------------------------------------------------------------
+- ["Thu Jan 01 00:00:00 1001 PST","Mon Jan 01 00:00:00 2001 PST")
+-(1 row)
+-
++-- temporarily disabled, see:
++-- https://gitlab.com/dalibo/postgresql_anonymizer/-/commit/199f0a392b37c59d92ae441fb8f037e094a11a52#note_2148017485
++--SELECT anon.generalize_tstzrange('19041107','millennium');
+ -- generalize_daterange
+ SELECT anon.generalize_daterange('19041107');
+   generalize_daterange   
+diff --git a/tests/sql/generalization.sql b/tests/sql/generalization.sql
+index b868344..b4fc977 100644
+--- a/ext-src/pg_anon-src/tests/sql/generalization.sql
++++ b/ext-src/pg_anon-src/tests/sql/generalization.sql
+@@ -61,7 +61,9 @@ SELECT anon.generalize_tstzrange('19041107','month');
+ SELECT anon.generalize_tstzrange('19041107','year');
+ SELECT anon.generalize_tstzrange('19041107','decade');
+ SELECT anon.generalize_tstzrange('19041107','century');
+-SELECT anon.generalize_tstzrange('19041107','millennium');
++-- temporarily disabled, see:
++-- https://gitlab.com/dalibo/postgresql_anonymizer/-/commit/199f0a392b37c59d92ae441fb8f037e094a11a52#note_2148017485
++--SELECT anon.generalize_tstzrange('19041107','millennium');
+ 
+ -- generalize_daterange
+ SELECT anon.generalize_daterange('19041107');
+
+commit 7dd414ee75f2875cffb1d6ba474df1f135a6fc6f
+Author: Alexey Masterov <alexeymasterov@neon.tech>
+Date:   Fri May 31 06:34:26 2024 +0000
+
+    These alternative expected files were added to consider the neon features
+
+diff --git a/ext-src/pg_anon-src/tests/expected/permissions_masked_role_1.out b/ext-src/pg_anon-src/tests/expected/permissions_masked_role_1.out
+new file mode 100644
+index 0000000..2539cfd
+--- /dev/null
++++ b/ext-src/pg_anon-src/tests/expected/permissions_masked_role_1.out
+@@ -0,0 +1,101 @@
++BEGIN;
++CREATE EXTENSION anon CASCADE;
++NOTICE:  installing required extension "pgcrypto"
++SELECT anon.init();
++ init 
++------
++ t
++(1 row)
++
++CREATE ROLE mallory_the_masked_user;
++SECURITY LABEL FOR anon ON ROLE mallory_the_masked_user IS 'MASKED';
++CREATE TABLE t1(i INT);
++ALTER TABLE t1 ADD COLUMN t TEXT;
++SECURITY LABEL FOR anon ON COLUMN t1.t
++IS 'MASKED WITH VALUE NULL';
++INSERT INTO t1 VALUES (1,'test');
++--
++-- We're checking the owner's permissions
++--
++-- see
++-- https://postgresql-anonymizer.readthedocs.io/en/latest/SECURITY/#permissions
++--
++SET ROLE mallory_the_masked_user;
++SELECT anon.pseudo_first_name(0) IS NOT NULL;
++ ?column? 
++----------
++ t
++(1 row)
++
++-- SHOULD FAIL
++DO $$
++BEGIN
++  PERFORM anon.init();
++  EXCEPTION WHEN insufficient_privilege
++  THEN RAISE NOTICE 'insufficient_privilege';
++END$$;
++NOTICE:  insufficient_privilege
++-- SHOULD FAIL
++DO $$
++BEGIN
++  PERFORM anon.anonymize_table('t1');
++  EXCEPTION WHEN insufficient_privilege
++  THEN RAISE NOTICE 'insufficient_privilege';
++END$$;
++NOTICE:  insufficient_privilege
++-- SHOULD FAIL
++SAVEPOINT fail_start_engine;
++SELECT anon.start_dynamic_masking();
++ERROR:  Only supersusers can start the dynamic masking engine.
++CONTEXT:  PL/pgSQL function anon.start_dynamic_masking(boolean) line 18 at RAISE
++ROLLBACK TO fail_start_engine;
++RESET ROLE;
++SELECT anon.start_dynamic_masking();
++ start_dynamic_masking 
++-----------------------
++ t
++(1 row)
++
++SET ROLE mallory_the_masked_user;
++SELECT * FROM mask.t1;
++ i | t 
++---+---
++ 1 | 
++(1 row)
++
++-- SHOULD FAIL
++DO $$
++BEGIN
++  SELECT * FROM public.t1;
++  EXCEPTION WHEN insufficient_privilege
++  THEN RAISE NOTICE 'insufficient_privilege';
++END$$;
++NOTICE:  insufficient_privilege
++-- SHOULD FAIL
++SAVEPOINT fail_stop_engine;
++SELECT anon.stop_dynamic_masking();
++ERROR:  Only supersusers can stop the dynamic masking engine.
++CONTEXT:  PL/pgSQL function anon.stop_dynamic_masking() line 18 at RAISE
++ROLLBACK TO fail_stop_engine;
++RESET ROLE;
++SELECT anon.stop_dynamic_masking();
++NOTICE:  The previous priviledges of 'mallory_the_masked_user' are not restored. You need to grant them manually.
++ stop_dynamic_masking 
++----------------------
++ t
++(1 row)
++
++SET ROLE mallory_the_masked_user;
++SELECT COUNT(*)=1 FROM anon.pg_masking_rules;
++ ?column? 
++----------
++ t
++(1 row)
++
++-- SHOULD FAIL
++SAVEPOINT fail_seclabel_on_role;
++SECURITY LABEL FOR anon ON ROLE mallory_the_masked_user IS NULL;
++ERROR:  permission denied
++DETAIL:  The current user must have the CREATEROLE attribute.
++ROLLBACK TO fail_seclabel_on_role;
++ROLLBACK;
+diff --git a/ext-src/pg_anon-src/tests/expected/permissions_owner_1.out b/ext-src/pg_anon-src/tests/expected/permissions_owner_1.out
+new file mode 100644
+index 0000000..8b090fe
+--- /dev/null
++++ b/ext-src/pg_anon-src/tests/expected/permissions_owner_1.out
+@@ -0,0 +1,104 @@
++BEGIN;
++CREATE EXTENSION anon CASCADE;
++NOTICE:  installing required extension "pgcrypto"
++SELECT anon.init();
++ init 
++------
++ t
++(1 row)
++
++CREATE ROLE oscar_the_owner;
++ALTER DATABASE :DBNAME OWNER TO oscar_the_owner;
++CREATE ROLE mallory_the_masked_user;
++SECURITY LABEL FOR anon ON ROLE mallory_the_masked_user IS 'MASKED';
++--
++-- We're checking the owner's permissions
++--
++-- see
++-- https://postgresql-anonymizer.readthedocs.io/en/latest/SECURITY/#permissions
++--
++SET ROLE oscar_the_owner;
++SELECT anon.pseudo_first_name(0) IS NOT NULL;
++ ?column? 
++----------
++ t
++(1 row)
++
++-- SHOULD FAIL
++DO $$
++BEGIN
++  PERFORM anon.init();
++  EXCEPTION WHEN insufficient_privilege
++  THEN RAISE NOTICE 'insufficient_privilege';
++END$$;
++NOTICE:  insufficient_privilege
++CREATE TABLE t1(i INT);
++ALTER TABLE t1 ADD COLUMN t TEXT;
++SECURITY LABEL FOR anon ON COLUMN t1.t
++IS 'MASKED WITH VALUE NULL';
++INSERT INTO t1 VALUES (1,'test');
++SELECT anon.anonymize_table('t1');
++ anonymize_table 
++-----------------
++ t
++(1 row)
++
++SELECT * FROM t1;
++ i | t 
++---+---
++ 1 | 
++(1 row)
++
++UPDATE t1 SET t='test' WHERE i=1;
++-- SHOULD FAIL
++SAVEPOINT fail_start_engine;
++SELECT anon.start_dynamic_masking();
++ start_dynamic_masking 
++-----------------------
++ t
++(1 row)
++
++ROLLBACK TO fail_start_engine;
++RESET ROLE;
++SELECT anon.start_dynamic_masking();
++ start_dynamic_masking 
++-----------------------
++ t
++(1 row)
++
++SET ROLE oscar_the_owner;
++SELECT * FROM t1;
++ i |  t   
++---+------
++ 1 | test
++(1 row)
++
++--SELECT * FROM mask.t1;
++-- SHOULD FAIL
++SAVEPOINT fail_stop_engine;
++SELECT anon.stop_dynamic_masking();
++ERROR:  permission denied for schema mask
++CONTEXT:  SQL statement "DROP VIEW mask.t1;"
++PL/pgSQL function anon.mask_drop_view(oid) line 3 at EXECUTE
++SQL statement "SELECT anon.mask_drop_view(oid)
++  FROM pg_catalog.pg_class
++  WHERE relnamespace=quote_ident(pg_catalog.current_setting('anon.sourceschema'))::REGNAMESPACE
++  AND relkind IN ('r','p','f')"
++PL/pgSQL function anon.stop_dynamic_masking() line 22 at PERFORM
++ROLLBACK TO fail_stop_engine;
++RESET ROLE;
++SELECT anon.stop_dynamic_masking();
++NOTICE:  The previous priviledges of 'mallory_the_masked_user' are not restored. You need to grant them manually.
++ stop_dynamic_masking 
++----------------------
++ t
++(1 row)
++
++SET ROLE oscar_the_owner;
++-- SHOULD FAIL
++SAVEPOINT fail_seclabel_on_role;
++SECURITY LABEL FOR anon ON ROLE mallory_the_masked_user IS NULL;
++ERROR:  permission denied
++DETAIL:  The current user must have the CREATEROLE attribute.
++ROLLBACK TO fail_seclabel_on_role;
++ROLLBACK;

compute/patches/pg_repack.patch

@@ -11,14 +11,6 @@ index bf6edcb..89b4c7f 100644
  
  USE_PGXS = 1	# use pgxs if not in contrib directory
  PGXS := $(shell $(PG_CONFIG) --pgxs)
-diff --git a/regress/expected/init-extension.out b/regress/expected/init-extension.out
-index 9f2e171..f6e4f8d 100644
---- a/regress/expected/init-extension.out
-+++ b/regress/expected/init-extension.out
-@@ -1,3 +1,2 @@
- SET client_min_messages = warning;
- CREATE EXTENSION pg_repack;
--RESET client_min_messages;
 diff --git a/regress/expected/nosuper.out b/regress/expected/nosuper.out
 index 8d0a94e..63b68bf 100644
 --- a/regress/expected/nosuper.out
@@ -50,14 +42,6 @@ index 8d0a94e..63b68bf 100644
  INFO: repacking table "public.tbl_cluster"
  ERROR: query failed: ERROR:  current transaction is aborted, commands ignored until end of transaction block
  DETAIL: query was: RESET lock_timeout
-diff --git a/regress/sql/init-extension.sql b/regress/sql/init-extension.sql
-index 9f2e171..f6e4f8d 100644
---- a/regress/sql/init-extension.sql
-+++ b/regress/sql/init-extension.sql
-@@ -1,3 +1,2 @@
- SET client_min_messages = warning;
- CREATE EXTENSION pg_repack;
--RESET client_min_messages;
 diff --git a/regress/sql/nosuper.sql b/regress/sql/nosuper.sql
 index 072f0fa..dbe60f8 100644
 --- a/regress/sql/nosuper.sql

compute/patches/pgvector.patch

@@ -15,7 +15,7 @@ index 7a4b88c..56678af 100644
  HEADERS = src/halfvec.h src/sparsevec.h src/vector.h
  
 diff --git a/src/hnswbuild.c b/src/hnswbuild.c
-index b667478..1298aa1 100644
+index b667478..dc95d89 100644
 --- a/src/hnswbuild.c
 +++ b/src/hnswbuild.c
 @@ -843,9 +843,17 @@ HnswParallelBuildMain(dsm_segment *seg, shm_toc *toc)
@@ -36,7 +36,7 @@ index b667478..1298aa1 100644
  	/* Close relations within worker */
  	index_close(indexRel, indexLockmode);
  	table_close(heapRel, heapLockmode);
-@@ -1100,13 +1108,25 @@ BuildIndex(Relation heap, Relation index, IndexInfo *indexInfo,
+@@ -1100,12 +1108,39 @@ BuildIndex(Relation heap, Relation index, IndexInfo *indexInfo,
  	SeedRandom(42);
  #endif
  
@@ -48,17 +48,32 @@ index b667478..1298aa1 100644
  
  	BuildGraph(buildstate, forkNum);
  
+-	if (RelationNeedsWAL(index) || forkNum == INIT_FORKNUM)
 +#ifdef NEON_SMGR
 +	smgr_finish_unlogged_build_phase_1(RelationGetSmgr(index));
 +#endif
 +
- 	if (RelationNeedsWAL(index) || forkNum == INIT_FORKNUM)
++	if (RelationNeedsWAL(index) || forkNum == INIT_FORKNUM) {
  		log_newpage_range(index, forkNum, 0, RelationGetNumberOfBlocksInFork(index, forkNum), true);
- 
++#ifdef NEON_SMGR
++		{
++#if PG_VERSION_NUM >= 160000
++			RelFileLocator rlocator = RelationGetSmgr(index)->smgr_rlocator.locator;
++#else
++			RelFileNode rlocator = RelationGetSmgr(index)->smgr_rnode.node;
++#endif
++			if (set_lwlsn_block_range_hook)
++				set_lwlsn_block_range_hook(XactLastRecEnd, rlocator,
++										   MAIN_FORKNUM, 0, RelationGetNumberOfBlocks(index));
++			if (set_lwlsn_relation_hook)
++				set_lwlsn_relation_hook(XactLastRecEnd, rlocator, MAIN_FORKNUM);
++		}
++#endif
++	}
++
 +#ifdef NEON_SMGR
 +	smgr_end_unlogged_build(RelationGetSmgr(index));
 +#endif
-+
+ 
  	FreeBuildState(buildstate);
  }
- 

compute/patches/rum.patch

@@ -1,5 +1,5 @@
 diff --git a/src/ruminsert.c b/src/ruminsert.c
-index 255e616..1c6edb7 100644
+index 255e616..7a2240f 100644
 --- a/src/ruminsert.c
 +++ b/src/ruminsert.c
 @@ -628,6 +628,10 @@ rumbuild(Relation heap, Relation index, struct IndexInfo *indexInfo)
@@ -24,12 +24,24 @@ index 255e616..1c6edb7 100644
  	/*
  	 * Write index to xlog
  	 */
-@@ -713,6 +721,10 @@ rumbuild(Relation heap, Relation index, struct IndexInfo *indexInfo)
+@@ -713,6 +721,22 @@ rumbuild(Relation heap, Relation index, struct IndexInfo *indexInfo)
  		UnlockReleaseBuffer(buffer);
  	}
  
 +#ifdef NEON_SMGR
-+	smgr_end_unlogged_build(index->rd_smgr);
++	{
++#if PG_VERSION_NUM >= 160000
++		RelFileLocator rlocator = RelationGetSmgr(index)->smgr_rlocator.locator;
++#else
++		RelFileNode rlocator = RelationGetSmgr(index)->smgr_rnode.node;
++#endif
++		if (set_lwlsn_block_range_hook)
++			set_lwlsn_block_range_hook(XactLastRecEnd, rlocator, MAIN_FORKNUM, 0, RelationGetNumberOfBlocks(index));
++		if (set_lwlsn_relation_hook)
++			set_lwlsn_relation_hook(XactLastRecEnd, rlocator, MAIN_FORKNUM);
++
++		smgr_end_unlogged_build(index->rd_smgr);
++	}
 +#endif
 +
  	/*

compute/vm-image-spec-bookworm.yaml

@@ -22,7 +22,7 @@ commands:
   - name: local_proxy
     user: postgres
     sysvInitAction: respawn
-    shell: 'RUST_LOG="info,proxy::serverless::sql_over_http=warn" /usr/local/bin/local_proxy --config-path /etc/local_proxy/config.json --pid-path /etc/local_proxy/pid --http 0.0.0.0:10432'
+    shell: '/usr/local/bin/local_proxy --config-path /etc/local_proxy/config.json --pid-path /etc/local_proxy/pid --http 0.0.0.0:10432'
   - name: postgres-exporter
     user: nobody
     sysvInitAction: respawn

compute/vm-image-spec-bullseye.yaml

@@ -22,7 +22,7 @@ commands:
   - name: local_proxy
     user: postgres
     sysvInitAction: respawn
-    shell: 'RUST_LOG="info,proxy::serverless::sql_over_http=warn" /usr/local/bin/local_proxy --config-path /etc/local_proxy/config.json --pid-path /etc/local_proxy/pid --http 0.0.0.0:10432'
+    shell: '/usr/local/bin/local_proxy --config-path /etc/local_proxy/config.json --pid-path /etc/local_proxy/pid --http 0.0.0.0:10432'
   - name: postgres-exporter
     user: nobody
     sysvInitAction: respawn

compute_tools/src/bin/compute_ctl.rs

@@ -116,7 +116,9 @@ struct Cli {
     #[arg(long)]
     pub set_disk_quota_for_fs: Option<String>,
 
-    #[arg(short = 'c', long)]
+    // TODO(tristan957): remove alias after compatibility tests are no longer
+    // an issue
+    #[arg(short = 'c', long, alias = "spec-path")]
     pub config: Option<OsString>,
 
     #[arg(short = 'i', long, group = "compute-id")]

compute_tools/src/compute.rs

@@ -641,26 +641,7 @@ impl ComputeNode {
 
                 let log_directory_path = Path::new(&self.params.pgdata).join("log");
                 let log_directory_path = log_directory_path.to_string_lossy().to_string();
-
-                // Add project_id,endpoint_id tag to identify the logs.
-                //
-                // These ids are passed from cplane,
-                // for backwards compatibility (old computes that don't have them),
-                // we set them to None.
-                // TODO: Clean up this code when all computes have them.
-                let tag: Option<String> = match (
-                    pspec.spec.project_id.as_deref(),
-                    pspec.spec.endpoint_id.as_deref(),
-                ) {
-                    (Some(project_id), Some(endpoint_id)) => {
-                        Some(format!("{project_id}/{endpoint_id}"))
-                    }
-                    (Some(project_id), None) => Some(format!("{project_id}/None")),
-                    (None, Some(endpoint_id)) => Some(format!("None,{endpoint_id}")),
-                    (None, None) => None,
-                };
-
-                configure_audit_rsyslog(log_directory_path.clone(), tag, &remote_endpoint)?;
+                configure_audit_rsyslog(log_directory_path.clone(), "hipaa", &remote_endpoint)?;
 
                 // Launch a background task to clean up the audit logs
                 launch_pgaudit_gc(log_directory_path);

compute_tools/src/http/extract/mod.rs

@@ -6,5 +6,4 @@ pub(crate) mod request_id;
 pub(crate) use json::Json;
 pub(crate) use path::Path;
 pub(crate) use query::Query;
-#[allow(unused)]
 pub(crate) use request_id::RequestId;

compute_tools/src/http/middleware/authorize.rs

@@ -1,7 +1,7 @@
-use std::collections::HashSet;
+use std::{collections::HashSet, net::SocketAddr};
 
 use anyhow::{Result, anyhow};
-use axum::{RequestExt, body::Body};
+use axum::{RequestExt, body::Body, extract::ConnectInfo};
 use axum_extra::{
     TypedHeader,
     headers::{Authorization, authorization::Bearer},
@@ -13,7 +13,7 @@ use jsonwebtoken::{Algorithm, DecodingKey, TokenData, Validation, jwk::JwkSet};
 use tower_http::auth::AsyncAuthorizeRequest;
 use tracing::{debug, warn};
 
-use crate::http::JsonResponse;
+use crate::http::{JsonResponse, extract::RequestId};
 
 #[derive(Clone, Debug)]
 pub(in crate::http) struct Authorize {
@@ -52,6 +52,31 @@ impl AsyncAuthorizeRequest<Body> for Authorize {
         let validation = self.validation.clone();
 
         Box::pin(async move {
+            let request_id = request.extract_parts::<RequestId>().await.unwrap();
+
+            // TODO: Remove this stanza after teaching neon_local and the
+            // regression tests to use a JWT + JWKS.
+            //
+            // https://github.com/neondatabase/neon/issues/11316
+            if cfg!(feature = "testing") {
+                warn!(%request_id, "Skipping compute_ctl authorization check");
+
+                return Ok(request);
+            }
+
+            let connect_info = request
+                .extract_parts::<ConnectInfo<SocketAddr>>()
+                .await
+                .unwrap();
+
+            // In the event the request is coming from the loopback interface,
+            // allow all requests
+            if connect_info.ip().is_loopback() {
+                warn!(%request_id, "Bypassed authorization because request is coming from the loopback interface");
+
+                return Ok(request);
+            }
+
             let TypedHeader(Authorization(bearer)) = request
                 .extract_parts::<TypedHeader<Authorization<Bearer>>>()
                 .await
@@ -87,8 +112,6 @@ impl Authorize {
         token: &str,
         validation: &Validation,
     ) -> Result<TokenData<ComputeClaims>> {
-        debug_assert!(!jwks.keys.is_empty());
-
         debug!("verifying token {}", token);
 
         for jwk in jwks.keys.iter() {

compute_tools/src/metrics.rs

@@ -1,8 +1,8 @@
-use metrics::core::{AtomicF64, AtomicU64, Collector, GenericCounter, GenericGauge};
+use metrics::core::{AtomicF64, Collector, GenericGauge};
 use metrics::proto::MetricFamily;
 use metrics::{
-    IntCounterVec, IntGaugeVec, UIntGaugeVec, register_gauge, register_int_counter,
-    register_int_counter_vec, register_int_gauge_vec, register_uint_gauge_vec,
+    IntCounterVec, IntGaugeVec, UIntGaugeVec, register_gauge, register_int_counter_vec,
+    register_int_gauge_vec, register_uint_gauge_vec,
 };
 use once_cell::sync::Lazy;
 
@@ -81,22 +81,6 @@ pub(crate) static COMPUTE_CTL_UP: Lazy<IntGaugeVec> = Lazy::new(|| {
     .expect("failed to define a metric")
 });
 
-pub(crate) static PG_CURR_DOWNTIME_MS: Lazy<GenericGauge<AtomicF64>> = Lazy::new(|| {
-    register_gauge!(
-        "compute_pg_current_downtime_ms",
-        "Non-cumulative duration of Postgres downtime in ms; resets after successful check",
-    )
-    .expect("failed to define a metric")
-});
-
-pub(crate) static PG_TOTAL_DOWNTIME_MS: Lazy<GenericCounter<AtomicU64>> = Lazy::new(|| {
-    register_int_counter!(
-        "compute_pg_downtime_ms_total",
-        "Cumulative duration of Postgres downtime in ms",
-    )
-    .expect("failed to define a metric")
-});
-
 pub fn collect() -> Vec<MetricFamily> {
     let mut metrics = COMPUTE_CTL_UP.collect();
     metrics.extend(INSTALLED_EXTENSIONS.collect());
@@ -104,7 +88,5 @@ pub fn collect() -> Vec<MetricFamily> {
     metrics.extend(REMOTE_EXT_REQUESTS_TOTAL.collect());
     metrics.extend(DB_MIGRATION_FAILED.collect());
     metrics.extend(AUDIT_LOG_DIR_SIZE.collect());
-    metrics.extend(PG_CURR_DOWNTIME_MS.collect());
-    metrics.extend(PG_TOTAL_DOWNTIME_MS.collect());
     metrics
 }

compute_tools/src/monitor.rs

@@ -6,294 +6,197 @@ use chrono::{DateTime, Utc};
 use compute_api::responses::ComputeStatus;
 use compute_api::spec::ComputeFeature;
 use postgres::{Client, NoTls};
-use tracing::{Level, error, info, instrument, span};
+use tracing::{debug, error, info, warn};
 
 use crate::compute::ComputeNode;
-use crate::metrics::{PG_CURR_DOWNTIME_MS, PG_TOTAL_DOWNTIME_MS};
 
 const MONITOR_CHECK_INTERVAL: Duration = Duration::from_millis(500);
 
-struct ComputeMonitor {
-    compute: Arc<ComputeNode>,
+// Spin in a loop and figure out the last activity time in the Postgres.
+// Then update it in the shared state. This function never errors out.
+// NB: the only expected panic is at `Mutex` unwrap(), all other errors
+// should be handled gracefully.
+fn watch_compute_activity(compute: &ComputeNode) {
+    // Suppose that `connstr` doesn't change
+    let connstr = compute.params.connstr.clone();
+    let conf = compute.get_conn_conf(Some("compute_ctl:activity_monitor"));
 
-    /// The moment when Postgres had some activity,
-    /// that should prevent compute from being suspended.
-    last_active: Option<DateTime<Utc>>,
+    // During startup and configuration we connect to every Postgres database,
+    // but we don't want to count this as some user activity. So wait until
+    // the compute fully started before monitoring activity.
+    wait_for_postgres_start(compute);
 
-    /// The moment when we last tried to check Postgres.
-    last_checked: DateTime<Utc>,
-    /// The last moment we did a successful Postgres check.
-    last_up: DateTime<Utc>,
+    // Define `client` outside of the loop to reuse existing connection if it's active.
+    let mut client = conf.connect(NoTls);
 
-    /// Only used for internal statistics change tracking
-    /// between monitor runs and can be outdated.
-    active_time: Option<f64>,
-    /// Only used for internal statistics change tracking
-    /// between monitor runs and can be outdated.
-    sessions: Option<i64>,
+    let mut sleep = false;
+    let mut prev_active_time: Option<f64> = None;
+    let mut prev_sessions: Option<i64> = None;
 
-    /// Use experimental statistics-based activity monitor. It's no longer
-    /// 'experimental' per se, as it's enabled for everyone, but we still
-    /// keep the flag as an option to turn it off in some cases if it will
-    /// misbehave.
-    experimental: bool,
-}
-
-impl ComputeMonitor {
-    fn report_down(&self) {
-        let now = Utc::now();
-
-        // Calculate and report current downtime
-        // (since the last time Postgres was up)
-        let downtime = now.signed_duration_since(self.last_up);
-        PG_CURR_DOWNTIME_MS.set(downtime.num_milliseconds() as f64);
-
-        // Calculate and update total downtime
-        // (cumulative duration of Postgres downtime in ms)
-        let inc = now
-            .signed_duration_since(self.last_checked)
-            .num_milliseconds();
-        PG_TOTAL_DOWNTIME_MS.inc_by(inc as u64);
+    if compute.has_feature(ComputeFeature::ActivityMonitorExperimental) {
+        info!("starting experimental activity monitor for {}", connstr);
+    } else {
+        info!("starting activity monitor for {}", connstr);
     }
 
-    fn report_up(&mut self) {
-        self.last_up = Utc::now();
-        PG_CURR_DOWNTIME_MS.set(0.0);
-    }
-
-    fn downtime_info(&self) -> String {
-        format!(
-            "total_ms: {}, current_ms: {}, last_up: {}",
-            PG_TOTAL_DOWNTIME_MS.get(),
-            PG_CURR_DOWNTIME_MS.get(),
-            self.last_up
-        )
-    }
-
-    /// Spin in a loop and figure out the last activity time in the Postgres.
-    /// Then update it in the shared state. This function never errors out.
-    /// NB: the only expected panic is at `Mutex` unwrap(), all other errors
-    /// should be handled gracefully.
-    #[instrument(skip_all)]
-    pub fn run(&mut self) {
-        // Suppose that `connstr` doesn't change
-        let connstr = self.compute.params.connstr.clone();
-        let conf = self
-            .compute
-            .get_conn_conf(Some("compute_ctl:compute_monitor"));
-
-        // During startup and configuration we connect to every Postgres database,
-        // but we don't want to count this as some user activity. So wait until
-        // the compute fully started before monitoring activity.
-        wait_for_postgres_start(&self.compute);
-
-        // Define `client` outside of the loop to reuse existing connection if it's active.
-        let mut client = conf.connect(NoTls);
-
-        info!("starting compute monitor for {}", connstr);
-
-        loop {
-            match &mut client {
-                Ok(cli) => {
-                    if cli.is_closed() {
-                        info!(
-                            downtime_info = self.downtime_info(),
-                            "connection to Postgres is closed, trying to reconnect"
-                        );
-                        self.report_down();
-
-                        // Connection is closed, reconnect and try again.
-                        client = conf.connect(NoTls);
-                    } else {
-                        match self.check(cli) {
-                            Ok(_) => {
-                                self.report_up();
-                                self.compute.update_last_active(self.last_active);
-                            }
-                            Err(e) => {
-                                // Although we have many places where we can return errors in `check()`,
-                                // normally it shouldn't happen. I.e., we will likely return error if
-                                // connection got broken, query timed out, Postgres returned invalid data, etc.
-                                // In all such cases it's suspicious, so let's report this as downtime.
-                                self.report_down();
-                                error!(
-                                    downtime_info = self.downtime_info(),
-                                    "could not check Postgres: {}", e
-                                );
-
-                                // Reconnect to Postgres just in case. During tests, I noticed
-                                // that queries in `check()` can fail with `connection closed`,
-                                // but `cli.is_closed()` above doesn't detect it. Even if old
-                                // connection is still alive, it will be dropped when we reassign
-                                // `client` to a new connection.
-                                client = conf.connect(NoTls);
-                            }
-                        }
-                    }
-                }
-                Err(e) => {
-                    info!(
-                        downtime_info = self.downtime_info(),
-                        "could not connect to Postgres: {}, retrying", e
-                    );
-                    self.report_down();
-
-                    // Establish a new connection and try again.
-                    client = conf.connect(NoTls);
-                }
-            }
-
-            // Reset the `last_checked` timestamp and sleep before the next iteration.
-            self.last_checked = Utc::now();
+    loop {
+        // We use `continue` a lot, so it's more convenient to sleep at the top of the loop.
+        // But skip the first sleep, so we can connect to Postgres immediately.
+        if sleep {
+            // Should be outside of the mutex lock to allow others to read while we sleep.
             thread::sleep(MONITOR_CHECK_INTERVAL);
+        } else {
+            sleep = true;
         }
-    }
 
-    #[instrument(skip_all)]
-    fn check(&mut self, cli: &mut Client) -> anyhow::Result<()> {
-        // This is new logic, only enable if the feature flag is set.
-        // TODO: remove this once we are sure that it works OR drop it altogether.
-        if self.experimental {
-            // Check if the total active time or sessions across all databases has changed.
-            // If it did, it means that user executed some queries. In theory, it can even go down if
-            // some databases were dropped, but it's still user activity.
-            match get_database_stats(cli) {
-                Ok((active_time, sessions)) => {
-                    let mut detected_activity = false;
+        match &mut client {
+            Ok(cli) => {
+                if cli.is_closed() {
+                    info!("connection to Postgres is closed, trying to reconnect");
 
-                    if let Some(prev_active_time) = self.active_time {
-                        if active_time != prev_active_time {
-                            detected_activity = true;
+                    // Connection is closed, reconnect and try again.
+                    client = conf.connect(NoTls);
+                    continue;
+                }
+
+                // This is a new logic, only enable if the feature flag is set.
+                // TODO: remove this once we are sure that it works OR drop it altogether.
+                if compute.has_feature(ComputeFeature::ActivityMonitorExperimental) {
+                    // First, check if the total active time or sessions across all databases has changed.
+                    // If it did, it means that user executed some queries. In theory, it can even go down if
+                    // some databases were dropped, but it's still a user activity.
+                    match get_database_stats(cli) {
+                        Ok((active_time, sessions)) => {
+                            let mut detected_activity = false;
+
+                            prev_active_time = match prev_active_time {
+                                Some(prev_active_time) => {
+                                    if active_time != prev_active_time {
+                                        detected_activity = true;
+                                    }
+                                    Some(active_time)
+                                }
+                                None => Some(active_time),
+                            };
+                            prev_sessions = match prev_sessions {
+                                Some(prev_sessions) => {
+                                    if sessions != prev_sessions {
+                                        detected_activity = true;
+                                    }
+                                    Some(sessions)
+                                }
+                                None => Some(sessions),
+                            };
+
+                            if detected_activity {
+                                // Update the last active time and continue, we don't need to
+                                // check backends state change.
+                                compute.update_last_active(Some(Utc::now()));
+                                continue;
+                            }
+                        }
+                        Err(e) => {
+                            error!("could not get database statistics: {}", e);
+                            continue;
                         }
                     }
-                    self.active_time = Some(active_time);
+                }
 
-                    if let Some(prev_sessions) = self.sessions {
-                        if sessions != prev_sessions {
-                            detected_activity = true;
+                // Second, if database statistics is the same, check all backends state change,
+                // maybe there is some with more recent activity. `get_backends_state_change()`
+                // can return None or stale timestamp, so it's `compute.update_last_active()`
+                // responsibility to check if the new timestamp is more recent than the current one.
+                // This helps us to discover new sessions, that did nothing yet.
+                match get_backends_state_change(cli) {
+                    Ok(last_active) => {
+                        compute.update_last_active(last_active);
+                    }
+                    Err(e) => {
+                        error!("could not get backends state change: {}", e);
+                    }
+                }
+
+                // Finally, if there are existing (logical) walsenders, do not suspend.
+                //
+                // walproposer doesn't currently show up in pg_stat_replication,
+                // but protect if it will be
+                let ws_count_query = "select count(*) from pg_stat_replication where application_name != 'walproposer';";
+                match cli.query_one(ws_count_query, &[]) {
+                    Ok(r) => match r.try_get::<&str, i64>("count") {
+                        Ok(num_ws) => {
+                            if num_ws > 0 {
+                                compute.update_last_active(Some(Utc::now()));
+                                continue;
+                            }
                         }
-                    }
-                    self.sessions = Some(sessions);
-
-                    if detected_activity {
-                        // Update the last active time and continue, we don't need to
-                        // check backends state change.
-                        self.last_active = Some(Utc::now());
-                        return Ok(());
+                        Err(e) => {
+                            warn!("failed to parse walsenders count: {:?}", e);
+                            continue;
+                        }
+                    },
+                    Err(e) => {
+                        warn!("failed to get list of walsenders: {:?}", e);
+                        continue;
                     }
                 }
-                Err(e) => {
-                    return Err(anyhow::anyhow!("could not get database statistics: {}", e));
+                //
+                // Don't suspend compute if there is an active logical replication subscription
+                //
+                // `where pid is not null` – to filter out read only computes and subscription on branches
+                //
+                let logical_subscriptions_query =
+                    "select count(*) from pg_stat_subscription where pid is not null;";
+                match cli.query_one(logical_subscriptions_query, &[]) {
+                    Ok(row) => match row.try_get::<&str, i64>("count") {
+                        Ok(num_subscribers) => {
+                            if num_subscribers > 0 {
+                                compute.update_last_active(Some(Utc::now()));
+                                continue;
+                            }
+                        }
+                        Err(e) => {
+                            warn!("failed to parse `pg_stat_subscription` count: {:?}", e);
+                            continue;
+                        }
+                    },
+                    Err(e) => {
+                        warn!(
+                            "failed to get list of active logical replication subscriptions: {:?}",
+                            e
+                        );
+                        continue;
+                    }
+                }
+                //
+                // Do not suspend compute if autovacuum is running
+                //
+                let autovacuum_count_query = "select count(*) from pg_stat_activity where backend_type = 'autovacuum worker'";
+                match cli.query_one(autovacuum_count_query, &[]) {
+                    Ok(r) => match r.try_get::<&str, i64>("count") {
+                        Ok(num_workers) => {
+                            if num_workers > 0 {
+                                compute.update_last_active(Some(Utc::now()));
+                                continue;
+                            }
+                        }
+                        Err(e) => {
+                            warn!("failed to parse autovacuum workers count: {:?}", e);
+                            continue;
+                        }
+                    },
+                    Err(e) => {
+                        warn!("failed to get list of autovacuum workers: {:?}", e);
+                        continue;
+                    }
                 }
             }
-        }
-
-        // If database statistics are the same, check all backends for state changes.
-        // Maybe there are some with more recent activity. `get_backends_state_change()`
-        // can return None or stale timestamp, so it's `compute.update_last_active()`
-        // responsibility to check if the new timestamp is more recent than the current one.
-        // This helps us to discover new sessions that have not done anything yet.
-        match get_backends_state_change(cli) {
-            Ok(last_active) => match (last_active, self.last_active) {
-                (Some(last_active), Some(prev_last_active)) => {
-                    if last_active > prev_last_active {
-                        self.last_active = Some(last_active);
-                        return Ok(());
-                    }
-                }
-                (Some(last_active), None) => {
-                    self.last_active = Some(last_active);
-                    return Ok(());
-                }
-                _ => {}
-            },
             Err(e) => {
-                return Err(anyhow::anyhow!(
-                    "could not get backends state change: {}",
-                    e
-                ));
+                debug!("could not connect to Postgres: {}, retrying", e);
+
+                // Establish a new connection and try again.
+                client = conf.connect(NoTls);
             }
         }
-
-        // If there are existing (logical) walsenders, do not suspend.
-        //
-        // N.B. walproposer doesn't currently show up in pg_stat_replication,
-        // but protect if it will.
-        const WS_COUNT_QUERY: &str =
-            "select count(*) from pg_stat_replication where application_name != 'walproposer';";
-        match cli.query_one(WS_COUNT_QUERY, &[]) {
-            Ok(r) => match r.try_get::<&str, i64>("count") {
-                Ok(num_ws) => {
-                    if num_ws > 0 {
-                        self.last_active = Some(Utc::now());
-                        return Ok(());
-                    }
-                }
-                Err(e) => {
-                    let err: anyhow::Error = e.into();
-                    return Err(err.context("failed to parse walsenders count"));
-                }
-            },
-            Err(e) => {
-                return Err(anyhow::anyhow!("failed to get list of walsenders: {}", e));
-            }
-        }
-
-        // Don't suspend compute if there is an active logical replication subscription
-        //
-        // `where pid is not null` – to filter out read only computes and subscription on branches
-        const LOGICAL_SUBSCRIPTIONS_QUERY: &str =
-            "select count(*) from pg_stat_subscription where pid is not null;";
-        match cli.query_one(LOGICAL_SUBSCRIPTIONS_QUERY, &[]) {
-            Ok(row) => match row.try_get::<&str, i64>("count") {
-                Ok(num_subscribers) => {
-                    if num_subscribers > 0 {
-                        self.last_active = Some(Utc::now());
-                        return Ok(());
-                    }
-                }
-                Err(e) => {
-                    return Err(anyhow::anyhow!(
-                        "failed to parse 'pg_stat_subscription' count: {}",
-                        e
-                    ));
-                }
-            },
-            Err(e) => {
-                return Err(anyhow::anyhow!(
-                    "failed to get list of active logical replication subscriptions: {}",
-                    e
-                ));
-            }
-        }
-
-        // Do not suspend compute if autovacuum is running
-        const AUTOVACUUM_COUNT_QUERY: &str =
-            "select count(*) from pg_stat_activity where backend_type = 'autovacuum worker'";
-        match cli.query_one(AUTOVACUUM_COUNT_QUERY, &[]) {
-            Ok(r) => match r.try_get::<&str, i64>("count") {
-                Ok(num_workers) => {
-                    if num_workers > 0 {
-                        self.last_active = Some(Utc::now());
-                        return Ok(());
-                    };
-                }
-                Err(e) => {
-                    return Err(anyhow::anyhow!(
-                        "failed to parse autovacuum workers count: {}",
-                        e
-                    ));
-                }
-            },
-            Err(e) => {
-                return Err(anyhow::anyhow!(
-                    "failed to get list of autovacuum workers: {}",
-                    e
-                ));
-            }
-        }
-
-        Ok(())
     }
 }
 
@@ -412,24 +315,9 @@ fn get_backends_state_change(cli: &mut Client) -> anyhow::Result<Option<DateTime
 /// Launch a separate compute monitor thread and return its `JoinHandle`.
 pub fn launch_monitor(compute: &Arc<ComputeNode>) -> thread::JoinHandle<()> {
     let compute = Arc::clone(compute);
-    let experimental = compute.has_feature(ComputeFeature::ActivityMonitorExperimental);
-    let now = Utc::now();
-    let mut monitor = ComputeMonitor {
-        compute,
-        last_active: None,
-        last_checked: now,
-        last_up: now,
-        active_time: None,
-        sessions: None,
-        experimental,
-    };
 
-    let span = span!(Level::INFO, "compute_monitor");
     thread::Builder::new()
         .name("compute-monitor".into())
-        .spawn(move || {
-            let _enter = span.enter();
-            monitor.run();
-        })
+        .spawn(move || watch_compute_activity(&compute))
         .expect("cannot launch compute monitor thread")
 }

compute_tools/src/rsyslog.rs

@@ -50,13 +50,13 @@ fn restart_rsyslog() -> Result<()> {
 
 pub fn configure_audit_rsyslog(
     log_directory: String,
-    tag: Option<String>,
+    tag: &str,
     remote_endpoint: &str,
 ) -> Result<()> {
     let config_content: String = format!(
         include_str!("config_template/compute_audit_rsyslog_template.conf"),
         log_directory = log_directory,
-        tag = tag.unwrap_or("".to_string()),
+        tag = tag,
         remote_endpoint = remote_endpoint
     );
 

control_plane/Cargo.toml

@@ -6,16 +6,13 @@ license.workspace = true
 
 [dependencies]
 anyhow.workspace = true
-base64.workspace = true
 camino.workspace = true
 clap.workspace = true
 comfy-table.workspace = true
 futures.workspace = true
 humantime.workspace = true
-jsonwebtoken.workspace = true
 nix.workspace = true
 once_cell.workspace = true
-pem.workspace = true
 humantime-serde.workspace = true
 hyper0.workspace = true
 regex.workspace = true
@@ -23,8 +20,6 @@ reqwest = { workspace = true, features = ["blocking", "json"] }
 scopeguard.workspace = true
 serde.workspace = true
 serde_json.workspace = true
-sha2.workspace = true
-spki.workspace = true
 thiserror.workspace = true
 toml.workspace = true
 toml_edit.workspace = true

control_plane/src/bin/neon_local.rs

@@ -18,11 +18,12 @@ use anyhow::{Context, Result, anyhow, bail};
 use clap::Parser;
 use compute_api::spec::ComputeMode;
 use control_plane::endpoint::ComputeControlPlane;
-use control_plane::endpoint_storage::{ENDPOINT_STORAGE_DEFAULT_PORT, EndpointStorage};
 use control_plane::local_env::{
-    EndpointStorageConf, InitForceMode, LocalEnv, NeonBroker, NeonLocalInitConf,
-    NeonLocalInitPageserverConf, SafekeeperConf,
+    InitForceMode, LocalEnv, NeonBroker, NeonLocalInitConf, NeonLocalInitPageserverConf,
+    ObjectStorageConf, SafekeeperConf,
 };
+use control_plane::object_storage::OBJECT_STORAGE_DEFAULT_PORT;
+use control_plane::object_storage::ObjectStorage;
 use control_plane::pageserver::PageServerNode;
 use control_plane::safekeeper::SafekeeperNode;
 use control_plane::storage_controller::{
@@ -62,7 +63,7 @@ const DEFAULT_PAGESERVER_ID: NodeId = NodeId(1);
 const DEFAULT_BRANCH_NAME: &str = "main";
 project_git_version!(GIT_VERSION);
 
-const DEFAULT_PG_VERSION: u32 = 17;
+const DEFAULT_PG_VERSION: u32 = 16;
 
 const DEFAULT_PAGESERVER_CONTROL_PLANE_API: &str = "http://127.0.0.1:1234/upcall/v1/";
 
@@ -92,7 +93,7 @@ enum NeonLocalCmd {
     #[command(subcommand)]
     Safekeeper(SafekeeperCmd),
     #[command(subcommand)]
-    EndpointStorage(EndpointStorageCmd),
+    ObjectStorage(ObjectStorageCmd),
     #[command(subcommand)]
     Endpoint(EndpointCmd),
     #[command(subcommand)]
@@ -459,14 +460,14 @@ enum SafekeeperCmd {
 
 #[derive(clap::Subcommand)]
 #[clap(about = "Manage object storage")]
-enum EndpointStorageCmd {
-    Start(EndpointStorageStartCmd),
-    Stop(EndpointStorageStopCmd),
+enum ObjectStorageCmd {
+    Start(ObjectStorageStartCmd),
+    Stop(ObjectStorageStopCmd),
 }
 
 #[derive(clap::Args)]
 #[clap(about = "Start object storage")]
-struct EndpointStorageStartCmd {
+struct ObjectStorageStartCmd {
     #[clap(short = 't', long, help = "timeout until we fail the command")]
     #[arg(default_value = "10s")]
     start_timeout: humantime::Duration,
@@ -474,7 +475,7 @@ struct EndpointStorageStartCmd {
 
 #[derive(clap::Args)]
 #[clap(about = "Stop object storage")]
-struct EndpointStorageStopCmd {
+struct ObjectStorageStopCmd {
     #[arg(value_enum, default_value = "fast")]
     #[clap(
         short = 'm',
@@ -551,7 +552,6 @@ enum EndpointCmd {
     Start(EndpointStartCmdArgs),
     Reconfigure(EndpointReconfigureCmdArgs),
     Stop(EndpointStopCmdArgs),
-    GenerateJwt(EndpointGenerateJwtCmdArgs),
 }
 
 #[derive(clap::Args)]
@@ -699,13 +699,6 @@ struct EndpointStopCmdArgs {
     mode: String,
 }
 
-#[derive(clap::Args)]
-#[clap(about = "Generate a JWT for an endpoint")]
-struct EndpointGenerateJwtCmdArgs {
-    #[clap(help = "Postgres endpoint id")]
-    endpoint_id: String,
-}
-
 #[derive(clap::Subcommand)]
 #[clap(about = "Manage neon_local branch name mappings")]
 enum MappingsCmd {
@@ -796,9 +789,7 @@ fn main() -> Result<()> {
             }
             NeonLocalCmd::StorageBroker(subcmd) => rt.block_on(handle_storage_broker(&subcmd, env)),
             NeonLocalCmd::Safekeeper(subcmd) => rt.block_on(handle_safekeeper(&subcmd, env)),
-            NeonLocalCmd::EndpointStorage(subcmd) => {
-                rt.block_on(handle_endpoint_storage(&subcmd, env))
-            }
+            NeonLocalCmd::ObjectStorage(subcmd) => rt.block_on(handle_object_storage(&subcmd, env)),
             NeonLocalCmd::Endpoint(subcmd) => rt.block_on(handle_endpoint(&subcmd, env)),
             NeonLocalCmd::Mappings(subcmd) => handle_mappings(&subcmd, env),
         };
@@ -1015,8 +1006,8 @@ fn handle_init(args: &InitCmdArgs) -> anyhow::Result<LocalEnv> {
                     }
                 })
                 .collect(),
-            endpoint_storage: EndpointStorageConf {
-                port: ENDPOINT_STORAGE_DEFAULT_PORT,
+            object_storage: ObjectStorageConf {
+                port: OBJECT_STORAGE_DEFAULT_PORT,
             },
             pg_distrib_dir: None,
             neon_distrib_dir: None,
@@ -1537,16 +1528,6 @@ async fn handle_endpoint(subcmd: &EndpointCmd, env: &local_env::LocalEnv) -> Res
                 .with_context(|| format!("postgres endpoint {endpoint_id} is not found"))?;
             endpoint.stop(&args.mode, args.destroy)?;
         }
-        EndpointCmd::GenerateJwt(args) => {
-            let endpoint_id = &args.endpoint_id;
-            let endpoint = cplane
-                .endpoints
-                .get(endpoint_id)
-                .with_context(|| format!("postgres endpoint {endpoint_id} is not found"))?;
-            let jwt = endpoint.generate_jwt()?;
-
-            print!("{jwt}");
-        }
     }
 
     Ok(())
@@ -1736,15 +1717,12 @@ async fn handle_safekeeper(subcmd: &SafekeeperCmd, env: &local_env::LocalEnv) ->
     Ok(())
 }
 
-async fn handle_endpoint_storage(
-    subcmd: &EndpointStorageCmd,
-    env: &local_env::LocalEnv,
-) -> Result<()> {
-    use EndpointStorageCmd::*;
-    let storage = EndpointStorage::from_env(env);
+async fn handle_object_storage(subcmd: &ObjectStorageCmd, env: &local_env::LocalEnv) -> Result<()> {
+    use ObjectStorageCmd::*;
+    let storage = ObjectStorage::from_env(env);
 
     // In tests like test_forward_compatibility or test_graceful_cluster_restart
-    // old neon binaries (without endpoint_storage) are present
+    // old neon binaries (without object_storage) are present
     if !storage.bin.exists() {
         eprintln!(
             "{} binary not found. Ignore if this is a compatibility test",
@@ -1754,13 +1732,13 @@ async fn handle_endpoint_storage(
     }
 
     match subcmd {
-        Start(EndpointStorageStartCmd { start_timeout }) => {
+        Start(ObjectStorageStartCmd { start_timeout }) => {
             if let Err(e) = storage.start(start_timeout).await {
-                eprintln!("endpoint_storage start failed: {e}");
+                eprintln!("object_storage start failed: {e}");
                 exit(1);
             }
         }
-        Stop(EndpointStorageStopCmd { stop_mode }) => {
+        Stop(ObjectStorageStopCmd { stop_mode }) => {
             let immediate = match stop_mode {
                 StopMode::Fast => false,
                 StopMode::Immediate => true,
@@ -1870,10 +1848,10 @@ async fn handle_start_all_impl(
         }
 
         js.spawn(async move {
-            EndpointStorage::from_env(env)
+            ObjectStorage::from_env(env)
                 .start(&retry_timeout)
                 .await
-                .map_err(|e| e.context("start endpoint_storage"))
+                .map_err(|e| e.context("start object_storage"))
         });
     })();
 
@@ -1972,9 +1950,9 @@ async fn try_stop_all(env: &local_env::LocalEnv, immediate: bool) {
         }
     }
 
-    let storage = EndpointStorage::from_env(env);
+    let storage = ObjectStorage::from_env(env);
     if let Err(e) = storage.stop(immediate) {
-        eprintln!("endpoint_storage stop failed: {:#}", e);
+        eprintln!("object_storage stop failed: {:#}", e);
     }
 
     for ps_conf in &env.pageservers {

control_plane/src/endpoint.rs

@@ -42,30 +42,22 @@ use std::path::PathBuf;
 use std::process::Command;
 use std::str::FromStr;
 use std::sync::Arc;
-use std::time::{Duration, Instant};
+use std::time::{Duration, Instant, SystemTime, UNIX_EPOCH};
 
 use anyhow::{Context, Result, anyhow, bail};
-use compute_api::requests::{ComputeClaims, ConfigurationRequest};
+use compute_api::requests::ConfigurationRequest;
 use compute_api::responses::{
-    ComputeConfig, ComputeCtlConfig, ComputeStatus, ComputeStatusResponse, TlsConfig,
+    ComputeConfig, ComputeCtlConfig, ComputeStatus, ComputeStatusResponse,
 };
 use compute_api::spec::{
     Cluster, ComputeAudit, ComputeFeature, ComputeMode, ComputeSpec, Database, PgIdent,
     RemoteExtSpec, Role,
 };
-use jsonwebtoken::jwk::{
-    AlgorithmParameters, CommonParameters, EllipticCurve, Jwk, JwkSet, KeyAlgorithm, KeyOperations,
-    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::membership::SafekeeperGeneration;
 use serde::{Deserialize, Serialize};
-use sha2::{Digest, Sha256};
-use spki::der::Decode;
-use spki::{SubjectPublicKeyInfo, SubjectPublicKeyInfoRef};
 use tracing::debug;
 use url::Host;
 use utils::id::{NodeId, TenantId, TimelineId};
@@ -90,7 +82,6 @@ pub struct EndpointConf {
     drop_subscriptions_before_start: bool,
     features: Vec<ComputeFeature>,
     cluster: Option<Cluster>,
-    compute_ctl_config: ComputeCtlConfig,
 }
 
 //
@@ -146,37 +137,6 @@ impl ComputeControlPlane {
             .unwrap_or(self.base_port)
     }
 
-    /// Create a JSON Web Key Set. This ideally matches the way we create a JWKS
-    /// from the production control plane.
-    fn create_jwks_from_pem(pem: &Pem) -> Result<JwkSet> {
-        let spki: SubjectPublicKeyInfoRef = SubjectPublicKeyInfo::from_der(pem.contents())?;
-        let public_key = spki.subject_public_key.raw_bytes();
-
-        let mut hasher = Sha256::new();
-        hasher.update(public_key);
-        let key_hash = hasher.finalize();
-
-        Ok(JwkSet {
-            keys: vec![Jwk {
-                common: CommonParameters {
-                    public_key_use: Some(PublicKeyUse::Signature),
-                    key_operations: Some(vec![KeyOperations::Verify]),
-                    key_algorithm: Some(KeyAlgorithm::EdDSA),
-                    key_id: Some(base64::encode_config(key_hash, base64::URL_SAFE_NO_PAD)),
-                    x509_url: None::<String>,
-                    x509_chain: None::<Vec<String>>,
-                    x509_sha1_fingerprint: None::<String>,
-                    x509_sha256_fingerprint: None::<String>,
-                },
-                algorithm: AlgorithmParameters::OctetKeyPair(OctetKeyPairParameters {
-                    key_type: OctetKeyPairType::OctetKeyPair,
-                    curve: EllipticCurve::Ed25519,
-                    x: base64::encode_config(public_key, base64::URL_SAFE_NO_PAD),
-                }),
-            }],
-        })
-    }
-
     #[allow(clippy::too_many_arguments)]
     pub fn new_endpoint(
         &mut self,
@@ -194,10 +154,6 @@ impl ComputeControlPlane {
         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);
         let internal_http_port = internal_http_port.unwrap_or_else(|| external_http_port + 1);
-        let compute_ctl_config = ComputeCtlConfig {
-            jwks: Self::create_jwks_from_pem(&self.env.read_public_key()?)?,
-            tls: None::<TlsConfig>,
-        };
         let ep = Arc::new(Endpoint {
             endpoint_id: endpoint_id.to_owned(),
             pg_address: SocketAddr::new(IpAddr::from(Ipv4Addr::LOCALHOST), pg_port),
@@ -225,7 +181,6 @@ impl ComputeControlPlane {
             reconfigure_concurrency: 1,
             features: vec![],
             cluster: None,
-            compute_ctl_config: compute_ctl_config.clone(),
         });
 
         ep.create_endpoint_dir()?;
@@ -245,7 +200,6 @@ impl ComputeControlPlane {
                 reconfigure_concurrency: 1,
                 features: vec![],
                 cluster: None,
-                compute_ctl_config,
             })?,
         )?;
         std::fs::write(
@@ -288,6 +242,7 @@ impl ComputeControlPlane {
 
 ///////////////////////////////////////////////////////////////////////////////
 
+#[derive(Debug)]
 pub struct Endpoint {
     /// used as the directory name
     endpoint_id: String,
@@ -316,9 +271,6 @@ pub struct Endpoint {
     features: Vec<ComputeFeature>,
     // Cluster settings
     cluster: Option<Cluster>,
-
-    /// The compute_ctl config for the endpoint's compute.
-    compute_ctl_config: ComputeCtlConfig,
 }
 
 #[derive(PartialEq, Eq)]
@@ -381,7 +333,6 @@ impl Endpoint {
             drop_subscriptions_before_start: conf.drop_subscriptions_before_start,
             features: conf.features,
             cluster: conf.cluster,
-            compute_ctl_config: conf.compute_ctl_config,
         })
     }
 
@@ -629,13 +580,6 @@ impl Endpoint {
         Ok(safekeeper_connstrings)
     }
 
-    /// Generate a JWT with the correct claims.
-    pub fn generate_jwt(&self) -> Result<String> {
-        self.env.generate_auth_token(&ComputeClaims {
-            compute_id: self.endpoint_id.clone(),
-        })
-    }
-
     #[allow(clippy::too_many_arguments)]
     pub async fn start(
         &self,
@@ -762,10 +706,14 @@ impl Endpoint {
 
             ComputeConfig {
                 spec: Some(spec),
-                compute_ctl_config: self.compute_ctl_config.clone(),
+                compute_ctl_config: ComputeCtlConfig::default(),
             }
         };
 
+        // TODO(tristan957): Remove the write to spec.json after compatibility
+        // tests work themselves out
+        let spec_path = self.endpoint_path().join("spec.json");
+        std::fs::write(spec_path, serde_json::to_string_pretty(&config.spec)?)?;
         let config_path = self.endpoint_path().join("config.json");
         std::fs::write(config_path, serde_json::to_string_pretty(&config)?)?;
 
@@ -775,6 +723,16 @@ impl Endpoint {
             .append(true)
             .open(self.endpoint_path().join("compute.log"))?;
 
+        // TODO(tristan957): Remove when compatibility tests are no longer an
+        // issue
+        let old_compute_ctl = {
+            let mut cmd = Command::new(self.env.neon_distrib_dir.join("compute_ctl"));
+            let help_output = cmd.arg("--help").output()?;
+            let help_output = String::from_utf8_lossy(&help_output.stdout);
+
+            !help_output.contains("--config")
+        };
+
         // Launch compute_ctl
         let conn_str = self.connstr("cloud_admin", "postgres");
         println!("Starting postgres node at '{}'", conn_str);
@@ -793,8 +751,19 @@ impl Endpoint {
         ])
         .args(["--pgdata", self.pgdata().to_str().unwrap()])
         .args(["--connstr", &conn_str])
-        .arg("--config")
-        .arg(self.endpoint_path().join("config.json").as_os_str())
+        // TODO(tristan957): Change this to --config when compatibility tests
+        // are no longer an issue
+        .args([
+            "--spec-path",
+            self.endpoint_path()
+                .join(if old_compute_ctl {
+                    "spec.json"
+                } else {
+                    "config.json"
+                })
+                .to_str()
+                .unwrap(),
+        ])
         .args([
             "--pgbin",
             self.env
@@ -805,7 +774,16 @@ impl Endpoint {
         ])
         // TODO: It would be nice if we generated compute IDs with the same
         // algorithm as the real control plane.
-        .args(["--compute-id", &self.endpoint_id])
+        .args([
+            "--compute-id",
+            &format!(
+                "compute-{}",
+                SystemTime::now()
+                    .duration_since(UNIX_EPOCH)
+                    .unwrap()
+                    .as_secs()
+            ),
+        ])
         .stdin(std::process::Stdio::null())
         .stderr(logfile.try_clone()?)
         .stdout(logfile);
@@ -903,7 +881,6 @@ impl Endpoint {
                     self.external_http_address.port()
                 ),
             )
-            .bearer_auth(self.generate_jwt()?)
             .send()
             .await?;
 
@@ -980,7 +957,6 @@ impl Endpoint {
                 self.external_http_address.port()
             ))
             .header(CONTENT_TYPE.as_str(), "application/json")
-            .bearer_auth(self.generate_jwt()?)
             .body(
                 serde_json::to_string(&ConfigurationRequest {
                     spec,

control_plane/src/lib.rs

@@ -9,8 +9,8 @@
 mod background_process;
 pub mod broker;
 pub mod endpoint;
-pub mod endpoint_storage;
 pub mod local_env;
+pub mod object_storage;
 pub mod pageserver;
 pub mod postgresql_conf;
 pub mod safekeeper;

control_plane/src/local_env.rs

@@ -12,18 +12,17 @@ use std::{env, fs};
 
 use anyhow::{Context, bail};
 use clap::ValueEnum;
-use pem::Pem;
 use postgres_backend::AuthType;
 use reqwest::Url;
 use serde::{Deserialize, Serialize};
 use utils::auth::encode_from_key_file;
 use utils::id::{NodeId, TenantId, TenantTimelineId, TimelineId};
 
-use crate::endpoint_storage::{ENDPOINT_STORAGE_REMOTE_STORAGE_DIR, EndpointStorage};
+use crate::object_storage::{OBJECT_STORAGE_REMOTE_STORAGE_DIR, ObjectStorage};
 use crate::pageserver::{PAGESERVER_REMOTE_STORAGE_DIR, PageServerNode};
 use crate::safekeeper::SafekeeperNode;
 
-pub const DEFAULT_PG_VERSION: u32 = 17;
+pub const DEFAULT_PG_VERSION: u32 = 16;
 
 //
 // This data structures represents neon_local CLI config
@@ -57,7 +56,6 @@ pub struct LocalEnv {
 
     // used to issue tokens during e.g pg start
     pub private_key_path: PathBuf,
-    /// Path to environment's public key
     pub public_key_path: PathBuf,
 
     pub broker: NeonBroker,
@@ -72,7 +70,7 @@ pub struct LocalEnv {
 
     pub safekeepers: Vec<SafekeeperConf>,
 
-    pub endpoint_storage: EndpointStorageConf,
+    pub object_storage: ObjectStorageConf,
 
     // Control plane upcall API for pageserver: if None, we will not run storage_controller  If set, this will
     // be propagated into each pageserver's configuration.
@@ -110,7 +108,7 @@ pub struct OnDiskConfig {
     )]
     pub pageservers: Vec<PageServerConf>,
     pub safekeepers: Vec<SafekeeperConf>,
-    pub endpoint_storage: EndpointStorageConf,
+    pub object_storage: ObjectStorageConf,
     pub control_plane_api: Option<Url>,
     pub control_plane_hooks_api: Option<Url>,
     pub control_plane_compute_hook_api: Option<Url>,
@@ -144,7 +142,7 @@ pub struct NeonLocalInitConf {
     pub storage_controller: Option<NeonStorageControllerConf>,
     pub pageservers: Vec<NeonLocalInitPageserverConf>,
     pub safekeepers: Vec<SafekeeperConf>,
-    pub endpoint_storage: EndpointStorageConf,
+    pub object_storage: ObjectStorageConf,
     pub control_plane_api: Option<Url>,
     pub control_plane_hooks_api: Option<Url>,
     pub generate_local_ssl_certs: bool,
@@ -152,7 +150,7 @@ pub struct NeonLocalInitConf {
 
 #[derive(Serialize, Default, Deserialize, PartialEq, Eq, Clone, Debug)]
 #[serde(default)]
-pub struct EndpointStorageConf {
+pub struct ObjectStorageConf {
     pub port: u16,
 }
 
@@ -413,8 +411,8 @@ impl LocalEnv {
         self.pg_dir(pg_version, "lib")
     }
 
-    pub fn endpoint_storage_bin(&self) -> PathBuf {
-        self.neon_distrib_dir.join("endpoint_storage")
+    pub fn object_storage_bin(&self) -> PathBuf {
+        self.neon_distrib_dir.join("object_storage")
     }
 
     pub fn pageserver_bin(&self) -> PathBuf {
@@ -450,8 +448,8 @@ impl LocalEnv {
         self.base_data_dir.join("safekeepers").join(data_dir_name)
     }
 
-    pub fn endpoint_storage_data_dir(&self) -> PathBuf {
-        self.base_data_dir.join("endpoint_storage")
+    pub fn object_storage_data_dir(&self) -> PathBuf {
+        self.base_data_dir.join("object_storage")
     }
 
     pub fn get_pageserver_conf(&self, id: NodeId) -> anyhow::Result<&PageServerConf> {
@@ -615,7 +613,7 @@ impl LocalEnv {
                 control_plane_compute_hook_api: _,
                 branch_name_mappings,
                 generate_local_ssl_certs,
-                endpoint_storage,
+                object_storage,
             } = on_disk_config;
             LocalEnv {
                 base_data_dir: repopath.to_owned(),
@@ -632,7 +630,7 @@ impl LocalEnv {
                 control_plane_hooks_api,
                 branch_name_mappings,
                 generate_local_ssl_certs,
-                endpoint_storage,
+                object_storage,
             }
         };
 
@@ -742,7 +740,7 @@ impl LocalEnv {
                 control_plane_compute_hook_api: None,
                 branch_name_mappings: self.branch_name_mappings.clone(),
                 generate_local_ssl_certs: self.generate_local_ssl_certs,
-                endpoint_storage: self.endpoint_storage.clone(),
+                object_storage: self.object_storage.clone(),
             },
         )
     }
@@ -760,11 +758,11 @@ impl LocalEnv {
 
     // this function is used only for testing purposes in CLI e g generate tokens during init
     pub fn generate_auth_token<S: Serialize>(&self, claims: &S) -> anyhow::Result<String> {
-        let key = self.read_private_key()?;
-        encode_from_key_file(claims, &key)
+        let private_key_path = self.get_private_key_path();
+        let key_data = fs::read(private_key_path)?;
+        encode_from_key_file(claims, &key_data)
     }
 
-    /// Get the path to the private key.
     pub fn get_private_key_path(&self) -> PathBuf {
         if self.private_key_path.is_absolute() {
             self.private_key_path.to_path_buf()
@@ -773,29 +771,6 @@ impl LocalEnv {
         }
     }
 
-    /// Get the path to the public key.
-    pub fn get_public_key_path(&self) -> PathBuf {
-        if self.public_key_path.is_absolute() {
-            self.public_key_path.to_path_buf()
-        } else {
-            self.base_data_dir.join(&self.public_key_path)
-        }
-    }
-
-    /// Read the contents of the private key file.
-    pub fn read_private_key(&self) -> anyhow::Result<Pem> {
-        let private_key_path = self.get_private_key_path();
-        let pem = pem::parse(fs::read(private_key_path)?)?;
-        Ok(pem)
-    }
-
-    /// Read the contents of the public key file.
-    pub fn read_public_key(&self) -> anyhow::Result<Pem> {
-        let public_key_path = self.get_public_key_path();
-        let pem = pem::parse(fs::read(public_key_path)?)?;
-        Ok(pem)
-    }
-
     /// Materialize the [`NeonLocalInitConf`] to disk. Called during [`neon_local init`].
     pub fn init(conf: NeonLocalInitConf, force: &InitForceMode) -> anyhow::Result<()> {
         let base_path = base_path();
@@ -849,7 +824,7 @@ impl LocalEnv {
             control_plane_api,
             generate_local_ssl_certs,
             control_plane_hooks_api,
-            endpoint_storage,
+            object_storage,
         } = conf;
 
         // Find postgres binaries.
@@ -901,7 +876,7 @@ impl LocalEnv {
             control_plane_hooks_api,
             branch_name_mappings: Default::default(),
             generate_local_ssl_certs,
-            endpoint_storage,
+            object_storage,
         };
 
         if generate_local_ssl_certs {
@@ -929,13 +904,13 @@ impl LocalEnv {
                 .context("pageserver init failed")?;
         }
 
-        EndpointStorage::from_env(&env)
+        ObjectStorage::from_env(&env)
             .init()
             .context("object storage init failed")?;
 
         // setup remote remote location for default LocalFs remote storage
         std::fs::create_dir_all(env.base_data_dir.join(PAGESERVER_REMOTE_STORAGE_DIR))?;
-        std::fs::create_dir_all(env.base_data_dir.join(ENDPOINT_STORAGE_REMOTE_STORAGE_DIR))?;
+        std::fs::create_dir_all(env.base_data_dir.join(OBJECT_STORAGE_REMOTE_STORAGE_DIR))?;
 
         env.persist_config()
     }
@@ -981,7 +956,6 @@ fn generate_auth_keys(private_key_path: &Path, public_key_path: &Path) -> anyhow
             String::from_utf8_lossy(&keygen_output.stderr)
         );
     }
-
     // Extract the public key from the private key file
     //
     // openssl pkey -in auth_private_key.pem -pubout -out auth_public_key.pem
@@ -998,7 +972,6 @@ fn generate_auth_keys(private_key_path: &Path, public_key_path: &Path) -> anyhow
             String::from_utf8_lossy(&keygen_output.stderr)
         );
     }
-
     Ok(())
 }
 

control_plane/src/object_storage.rs

@@ -1,33 +1,34 @@
 use crate::background_process::{self, start_process, stop_process};
 use crate::local_env::LocalEnv;
+use anyhow::anyhow;
 use anyhow::{Context, Result};
 use camino::Utf8PathBuf;
 use std::io::Write;
 use std::time::Duration;
 
 /// Directory within .neon which will be used by default for LocalFs remote storage.
-pub const ENDPOINT_STORAGE_REMOTE_STORAGE_DIR: &str = "local_fs_remote_storage/endpoint_storage";
-pub const ENDPOINT_STORAGE_DEFAULT_PORT: u16 = 9993;
+pub const OBJECT_STORAGE_REMOTE_STORAGE_DIR: &str = "local_fs_remote_storage/object_storage";
+pub const OBJECT_STORAGE_DEFAULT_PORT: u16 = 9993;
 
-pub struct EndpointStorage {
+pub struct ObjectStorage {
     pub bin: Utf8PathBuf,
     pub data_dir: Utf8PathBuf,
     pub pemfile: Utf8PathBuf,
     pub port: u16,
 }
 
-impl EndpointStorage {
-    pub fn from_env(env: &LocalEnv) -> EndpointStorage {
-        EndpointStorage {
-            bin: Utf8PathBuf::from_path_buf(env.endpoint_storage_bin()).unwrap(),
-            data_dir: Utf8PathBuf::from_path_buf(env.endpoint_storage_data_dir()).unwrap(),
+impl ObjectStorage {
+    pub fn from_env(env: &LocalEnv) -> ObjectStorage {
+        ObjectStorage {
+            bin: Utf8PathBuf::from_path_buf(env.object_storage_bin()).unwrap(),
+            data_dir: Utf8PathBuf::from_path_buf(env.object_storage_data_dir()).unwrap(),
             pemfile: Utf8PathBuf::from_path_buf(env.public_key_path.clone()).unwrap(),
-            port: env.endpoint_storage.port,
+            port: env.object_storage.port,
         }
     }
 
     fn config_path(&self) -> Utf8PathBuf {
-        self.data_dir.join("endpoint_storage.json")
+        self.data_dir.join("object_storage.json")
     }
 
     fn listen_addr(&self) -> Utf8PathBuf {
@@ -48,7 +49,7 @@ impl EndpointStorage {
         let cfg = Cfg {
             listen: self.listen_addr(),
             pemfile: parent.join(self.pemfile.clone()),
-            local_path: parent.join(ENDPOINT_STORAGE_REMOTE_STORAGE_DIR),
+            local_path: parent.join(OBJECT_STORAGE_REMOTE_STORAGE_DIR),
             r#type: "LocalFs".to_string(),
         };
         std::fs::create_dir_all(self.config_path().parent().unwrap())?;
@@ -58,19 +59,24 @@ impl EndpointStorage {
     }
 
     pub async fn start(&self, retry_timeout: &Duration) -> Result<()> {
-        println!("Starting endpoint_storage at {}", self.listen_addr());
+        println!("Starting s3 proxy at {}", self.listen_addr());
         std::io::stdout().flush().context("flush stdout")?;
 
         let process_status_check = || async {
-            let res = reqwest::Client::new().get(format!("http://{}/metrics", self.listen_addr()));
-            match res.send().await {
-                Ok(res) => Ok(res.status().is_success()),
-                Err(_) => Ok(false),
+            tokio::time::sleep(Duration::from_millis(500)).await;
+            let res = reqwest::Client::new()
+                .get(format!("http://{}/metrics", self.listen_addr()))
+                .send()
+                .await;
+            match res {
+                Ok(response) if response.status().is_success() => Ok(true),
+                Ok(_) => Err(anyhow!("Failed to query /metrics")),
+                Err(e) => Err(anyhow!("Failed to check node status: {e}")),
             }
         };
 
         let res = start_process(
-            "endpoint_storage",
+            "object_storage",
             &self.data_dir.clone().into_std_path_buf(),
             &self.bin.clone().into_std_path_buf(),
             vec![self.config_path().to_string()],
@@ -88,14 +94,14 @@ impl EndpointStorage {
     }
 
     pub fn stop(&self, immediate: bool) -> anyhow::Result<()> {
-        stop_process(immediate, "endpoint_storage", &self.pid_file())
+        stop_process(immediate, "object_storage", &self.pid_file())
     }
 
     fn log_file(&self) -> Utf8PathBuf {
-        self.data_dir.join("endpoint_storage.log")
+        self.data_dir.join("object_storage.log")
     }
 
     fn pid_file(&self) -> Utf8PathBuf {
-        self.data_dir.join("endpoint_storage.pid")
+        self.data_dir.join("object_storage.pid")
     }
 }

control_plane/src/pageserver.rs

@@ -413,11 +413,6 @@ impl PageServerNode {
                 .map(serde_json::from_str)
                 .transpose()
                 .context("Failed to parse 'compaction_algorithm' json")?,
-            compaction_shard_ancestor: settings
-                .remove("compaction_shard_ancestor")
-                .map(|x| x.parse::<bool>())
-                .transpose()
-                .context("Failed to parse 'compaction_shard_ancestor' as a bool")?,
             compaction_l0_first: settings
                 .remove("compaction_l0_first")
                 .map(|x| x.parse::<bool>())

control_plane/src/storage_controller.rs

@@ -18,7 +18,6 @@ use pageserver_api::models::{
 };
 use pageserver_api::shard::TenantShardId;
 use pageserver_client::mgmt_api::ResponseErrorMessageExt;
-use pem::Pem;
 use postgres_backend::AuthType;
 use reqwest::{Certificate, Method};
 use serde::de::DeserializeOwned;
@@ -35,8 +34,8 @@ use crate::local_env::{LocalEnv, NeonStorageControllerConf};
 
 pub struct StorageController {
     env: LocalEnv,
-    private_key: Option<Pem>,
-    public_key: Option<Pem>,
+    private_key: Option<Vec<u8>>,
+    public_key: Option<String>,
     client: reqwest::Client,
     config: NeonStorageControllerConf,
 
@@ -117,9 +116,7 @@ impl StorageController {
             AuthType::Trust => (None, None),
             AuthType::NeonJWT => {
                 let private_key_path = env.get_private_key_path();
-                let private_key =
-                    pem::parse(fs::read(private_key_path).expect("failed to read private key"))
-                        .expect("failed to parse PEM file");
+                let private_key = fs::read(private_key_path).expect("failed to read private key");
 
                 // If pageserver auth is enabled, this implicitly enables auth for this service,
                 // using the same credentials.
@@ -141,13 +138,9 @@ impl StorageController {
                         .expect("Empty key dir")
                         .expect("Error reading key dir");
 
-                    pem::parse(std::fs::read_to_string(dent.path()).expect("Can't read public key"))
-                        .expect("Failed to parse PEM file")
+                    std::fs::read_to_string(dent.path()).expect("Can't read public key")
                 } else {
-                    pem::parse(
-                        std::fs::read_to_string(&public_key_path).expect("Can't read public key"),
-                    )
-                    .expect("Failed to parse PEM file")
+                    std::fs::read_to_string(&public_key_path).expect("Can't read public key")
                 };
                 (Some(private_key), Some(public_key))
             }

deny.toml

@@ -45,7 +45,9 @@ allow = [
     "ISC",
     "MIT",
     "MPL-2.0",
+    "OpenSSL",
     "Unicode-3.0",
+    "Zlib",
 ]
 confidence-threshold = 0.8
 exceptions = [
@@ -54,6 +56,14 @@ exceptions = [
     { allow = ["Zlib"], name = "const_format", version = "*" },
 ]
 
+[[licenses.clarify]]
+name = "ring"
+version = "*"
+expression = "MIT AND ISC AND OpenSSL"
+license-files = [
+    { path = "LICENSE", hash = 0xbd0eed23 }
+]
+
 [licenses.private]
 ignore = true
 registries = []
@@ -106,11 +116,7 @@ name = "openssl"
 unknown-registry = "warn"
 unknown-git = "warn"
 allow-registry = ["https://github.com/rust-lang/crates.io-index"]
-allow-git = [
-    # Crate pinned to commit in origin repo due to opentelemetry version.
-    # TODO: Remove this once crate is fetched from crates.io again.
-    "https://github.com/mattiapenati/tower-otel",
-]
+allow-git = []
 
 [sources.allow-org]
 github = [

docker-compose/README.md

@@ -1,3 +1,4 @@
+
 # Example docker compose configuration
 
 The configuration in this directory is used for testing Neon docker images: it is
@@ -7,13 +8,3 @@ you can experiment with a miniature Neon system, use `cargo neon` rather than co
 This configuration does not start the storage controller, because the controller
 needs a way to reconfigure running computes, and no such thing exists in this setup.
 
-## Generating the JWKS for a compute
-
-```shell
-openssl genpkey -algorithm Ed25519 -out private-key.pem
-openssl pkey -in private-key.pem -pubout -out public-key.pem
-openssl pkey -pubin -inform pem -in public-key.pem -pubout -outform der -out public-key.der
-key="$(xxd -plain -cols 32 -s -32 public-key.der)"
-key_id="$(printf '%s' "$key" | sha256sum | awk '{ print $1 }' | basenc --base64url --wrap=0)"
-x="$(printf '%s' "$key" | basenc --base64url --wrap=0)"
-```

docker-compose/compute_wrapper/private-key.pem

@@ -1,3 +0,0 @@
------BEGIN PRIVATE KEY-----
-MC4CAQAwBQYDK2VwBCIEIOmnRbzt2AJ0d+S3aU1hiYOl/tXpvz1FmWBfwHYBgOma
------END PRIVATE KEY-----

docker-compose/compute_wrapper/public-key.pem

@@ -1,3 +0,0 @@
------BEGIN PUBLIC KEY-----
-MCowBQYDK2VwAyEADY0al/U0bgB3+9fUGk+3PKWnsck9OyxN5DjHIN6Xep0=
------END PUBLIC KEY-----

docker-compose/compute_wrapper/shell/compute.sh

@@ -81,9 +81,19 @@ sed -i "s/TIMELINE_ID/${timeline_id}/" ${CONFIG_FILE}
 
 cat ${CONFIG_FILE}
 
+# TODO(tristan957): Remove these workarounds for backwards compatibility after
+# the next compute release. That includes these next few lines and the
+# --spec-path in the compute_ctl invocation.
+if compute_ctl --help | grep --quiet -- '--config'; then
+  SPEC_PATH="$CONFIG_FILE"
+else
+  jq '.spec' < "$CONFIG_FILE" > /tmp/spec.json
+  SPEC_PATH=/tmp/spec.json
+fi
+
 echo "Start compute node"
 /usr/local/bin/compute_ctl --pgdata /var/db/postgres/compute \
      -C "postgresql://cloud_admin@localhost:55433/postgres"  \
      -b /usr/local/bin/postgres                              \
      --compute-id "compute-$RANDOM"                          \
-     --config "$CONFIG_FILE"
+     --spec-path "$SPEC_PATH"

docker-compose/compute_wrapper/var/db/postgres/configs/config.json

@@ -142,19 +142,7 @@
     },
     "compute_ctl_config": {
         "jwks": {
-            "keys": [
-                {
-                    "use": "sig",
-                    "key_ops": [
-                        "verify"
-                    ],
-                    "alg": "EdDSA",
-                    "kid": "ZGIxMzAzOGY0YWQwODk2ODU1MTk1NzMxMDFkYmUyOWU2NzZkOWNjNjMyMGRkZGJjOWY0MjdjYWVmNzE1MjUyOAo=",
-                    "kty": "OKP",
-                    "crv": "Ed25519",
-                    "x": "MGQ4ZDFhOTdmNTM0NmUwMDc3ZmJkN2Q0MWE0ZmI3M2NhNWE3YjFjOTNkM2IyYzRkZTQzOGM3MjBkZTk3N2E5ZAo="
-                }
-            ]
+            "keys": []
         }
     }
 }

libs/compute_api/src/responses.rs

@@ -160,7 +160,7 @@ pub struct CatalogObjects {
     pub databases: Vec<Database>,
 }
 
-#[derive(Clone, Debug, Deserialize, Eq, PartialEq, Serialize)]
+#[derive(Clone, Debug, Deserialize, Serialize)]
 pub struct ComputeCtlConfig {
     /// Set of JSON web keys that the compute can use to authenticate
     /// communication from the control plane.
@@ -179,7 +179,7 @@ impl Default for ComputeCtlConfig {
     }
 }
 
-#[derive(Clone, Debug, Deserialize, Eq, PartialEq, Serialize)]
+#[derive(Clone, Debug, Deserialize, Serialize)]
 pub struct TlsConfig {
     pub key_path: String,
     pub cert_path: String,

libs/compute_api/src/spec.rs

@@ -242,22 +242,13 @@ impl RemoteExtSpec {
 
         match self.extension_data.get(real_ext_name) {
             Some(_ext_data) => {
-                // We have decided to use the Go naming convention due to Kubernetes.
-
-                let arch = match std::env::consts::ARCH {
-                    "x86_64" => "amd64",
-                    "aarch64" => "arm64",
-                    arch => arch,
-                };
-
                 // Construct the path to the extension archive
                 // BUILD_TAG/PG_MAJOR_VERSION/extensions/EXTENSION_NAME.tar.zst
                 //
                 // Keep it in sync with path generation in
                 // https://github.com/neondatabase/build-custom-extensions/tree/main
-                let archive_path_str = format!(
-                    "{build_tag}/{arch}/{pg_major_version}/extensions/{real_ext_name}.tar.zst"
-                );
+                let archive_path_str =
+                    format!("{build_tag}/{pg_major_version}/extensions/{real_ext_name}.tar.zst");
                 Ok((
                     real_ext_name.to_string(),
                     RemotePath::from_string(&archive_path_str)?,

libs/http-utils/Cargo.toml

@@ -14,7 +14,6 @@ futures.workspace = true
 hyper0.workspace = true
 itertools.workspace = true
 jemalloc_pprof.workspace = true
-jsonwebtoken.workspace = true
 once_cell.workspace = true
 pprof.workspace = true
 regex.workspace = true

libs/http-utils/src/endpoint.rs

@@ -8,7 +8,6 @@ use bytes::{Bytes, BytesMut};
 use hyper::header::{AUTHORIZATION, CONTENT_DISPOSITION, CONTENT_TYPE, HeaderName};
 use hyper::http::HeaderValue;
 use hyper::{Body, Method, Request, Response};
-use jsonwebtoken::TokenData;
 use metrics::{Encoder, IntCounter, TextEncoder, register_int_counter};
 use once_cell::sync::Lazy;
 use pprof::ProfilerGuardBuilder;
@@ -619,7 +618,7 @@ pub fn auth_middleware<B: hyper::body::HttpBody + Send + Sync + 'static>(
                     })?;
                     let token = parse_token(header_value)?;
 
-                    let data: TokenData<Claims> = auth.decode(token).map_err(|err| {
+                    let data = auth.decode(token).map_err(|err| {
                         warn!("Authentication error: {err}");
                         // Rely on From<AuthError> for ApiError impl
                         err

libs/pageserver_api/Cargo.toml

@@ -35,7 +35,6 @@ nix = {workspace = true, optional = true}
 reqwest.workspace = true
 rand.workspace = true
 tracing-utils.workspace = true
-once_cell.workspace = true
 
 [dev-dependencies]
 bincode.workspace = true

libs/pageserver_api/src/config.rs

@@ -181,7 +181,6 @@ pub struct ConfigToml {
     pub generate_unarchival_heatmap: Option<bool>,
     pub tracing: Option<Tracing>,
     pub enable_tls_page_service_api: bool,
-    pub dev_mode: bool,
 }
 
 #[derive(Debug, Clone, PartialEq, Eq, serde::Serialize, serde::Deserialize)]
@@ -380,8 +379,6 @@ pub struct TenantConfigToml {
     /// size exceeds `compaction_upper_limit * checkpoint_distance`.
     pub compaction_upper_limit: usize,
     pub compaction_algorithm: crate::models::CompactionAlgorithmSettings,
-    /// If true, enable shard ancestor compaction (enabled by default).
-    pub compaction_shard_ancestor: bool,
     /// If true, compact down L0 across all tenant timelines before doing regular compaction. L0
     /// compaction must be responsive to avoid read amp during heavy ingestion. Defaults to true.
     pub compaction_l0_first: bool,
@@ -658,7 +655,6 @@ impl Default for ConfigToml {
             generate_unarchival_heatmap: None,
             tracing: None,
             enable_tls_page_service_api: false,
-            dev_mode: false,
         }
     }
 }
@@ -681,13 +677,12 @@ pub mod tenant_conf_defaults {
 
     pub const DEFAULT_COMPACTION_PERIOD: &str = "20 s";
     pub const DEFAULT_COMPACTION_THRESHOLD: usize = 10;
-    pub const DEFAULT_COMPACTION_SHARD_ANCESTOR: bool = true;
 
     // This value needs to be tuned to avoid OOM. We have 3/4*CPUs threads for L0 compaction, that's
-    // 3/4*8=6 on most of our pageservers. Compacting 10 layers requires a maximum of
-    // DEFAULT_CHECKPOINT_DISTANCE*10 memory, that's 2560MB. So with this config, we can get a maximum peak
-    // compaction usage of 15360MB.
-    pub const DEFAULT_COMPACTION_UPPER_LIMIT: usize = 10;
+    // 3/4*16=9 on most of our pageservers. Compacting 20 layers requires about 1 GB memory (could
+    // be reduced later by optimizing L0 hole calculation to avoid loading all keys into memory). So
+    // with this config, we can get a maximum peak compaction usage of 9 GB.
+    pub const DEFAULT_COMPACTION_UPPER_LIMIT: usize = 20;
     // Enable L0 compaction pass and semaphore by default. L0 compaction must be responsive to avoid
     // read amp.
     pub const DEFAULT_COMPACTION_L0_FIRST: bool = true;
@@ -704,11 +699,8 @@ pub mod tenant_conf_defaults {
     // Relevant: https://github.com/neondatabase/neon/issues/3394
     pub const DEFAULT_GC_PERIOD: &str = "1 hr";
     pub const DEFAULT_IMAGE_CREATION_THRESHOLD: usize = 3;
-    // Currently, any value other than 0 will trigger image layer creation preemption immediately with L0 backpressure
-    // without looking at the exact number of L0 layers.
-    // It was expected to have the following behavior:
-    // > If there are more than threshold * compaction_threshold (that is 3 * 10 in the default config) L0 layers, image
-    // > layer creation will end immediately. Set to 0 to disable.
+    // If there are more than threshold * compaction_threshold (that is 3 * 10 in the default config) L0 layers, image
+    // layer creation will end immediately. Set to 0 to disable.
     pub const DEFAULT_IMAGE_CREATION_PREEMPT_THRESHOLD: usize = 3;
     pub const DEFAULT_PITR_INTERVAL: &str = "7 days";
     pub const DEFAULT_WALRECEIVER_CONNECT_TIMEOUT: &str = "10 seconds";
@@ -742,7 +734,6 @@ impl Default for TenantConfigToml {
             compaction_algorithm: crate::models::CompactionAlgorithmSettings {
                 kind: DEFAULT_COMPACTION_ALGORITHM,
             },
-            compaction_shard_ancestor: DEFAULT_COMPACTION_SHARD_ANCESTOR,
             compaction_l0_first: DEFAULT_COMPACTION_L0_FIRST,
             compaction_l0_semaphore: DEFAULT_COMPACTION_L0_SEMAPHORE,
             l0_flush_delay_threshold: None,

libs/pageserver_api/src/models.rs

@@ -320,35 +320,6 @@ pub struct TimelineCreateRequest {
     pub mode: TimelineCreateRequestMode,
 }
 
-impl TimelineCreateRequest {
-    pub fn mode_tag(&self) -> &'static str {
-        match &self.mode {
-            TimelineCreateRequestMode::Branch { .. } => "branch",
-            TimelineCreateRequestMode::ImportPgdata { .. } => "import",
-            TimelineCreateRequestMode::Bootstrap { .. } => "bootstrap",
-        }
-    }
-
-    pub fn is_import(&self) -> bool {
-        matches!(self.mode, TimelineCreateRequestMode::ImportPgdata { .. })
-    }
-}
-
-#[derive(Serialize, Deserialize, Debug, Clone, PartialEq, Eq)]
-pub enum ShardImportStatus {
-    InProgress,
-    Done,
-    Error(String),
-}
-impl ShardImportStatus {
-    pub fn is_terminal(&self) -> bool {
-        match self {
-            ShardImportStatus::InProgress => false,
-            ShardImportStatus::Done | ShardImportStatus::Error(_) => true,
-        }
-    }
-}
-
 /// Storage controller specific extensions to [`TimelineInfo`].
 #[derive(Serialize, Deserialize, Clone)]
 pub struct TimelineCreateResponseStorcon {
@@ -555,8 +526,6 @@ pub struct TenantConfigPatch {
     #[serde(skip_serializing_if = "FieldPatch::is_noop")]
     pub compaction_algorithm: FieldPatch<CompactionAlgorithmSettings>,
     #[serde(skip_serializing_if = "FieldPatch::is_noop")]
-    pub compaction_shard_ancestor: FieldPatch<bool>,
-    #[serde(skip_serializing_if = "FieldPatch::is_noop")]
     pub compaction_l0_first: FieldPatch<bool>,
     #[serde(skip_serializing_if = "FieldPatch::is_noop")]
     pub compaction_l0_semaphore: FieldPatch<bool>,
@@ -646,9 +615,6 @@ pub struct TenantConfig {
     #[serde(skip_serializing_if = "Option::is_none")]
     pub compaction_algorithm: Option<CompactionAlgorithmSettings>,
 
-    #[serde(skip_serializing_if = "Option::is_none")]
-    pub compaction_shard_ancestor: Option<bool>,
-
     #[serde(skip_serializing_if = "Option::is_none")]
     pub compaction_l0_first: Option<bool>,
 
@@ -758,7 +724,6 @@ impl TenantConfig {
             mut compaction_threshold,
             mut compaction_upper_limit,
             mut compaction_algorithm,
-            mut compaction_shard_ancestor,
             mut compaction_l0_first,
             mut compaction_l0_semaphore,
             mut l0_flush_delay_threshold,
@@ -807,9 +772,6 @@ impl TenantConfig {
             .compaction_upper_limit
             .apply(&mut compaction_upper_limit);
         patch.compaction_algorithm.apply(&mut compaction_algorithm);
-        patch
-            .compaction_shard_ancestor
-            .apply(&mut compaction_shard_ancestor);
         patch.compaction_l0_first.apply(&mut compaction_l0_first);
         patch
             .compaction_l0_semaphore
@@ -898,7 +860,6 @@ impl TenantConfig {
             compaction_threshold,
             compaction_upper_limit,
             compaction_algorithm,
-            compaction_shard_ancestor,
             compaction_l0_first,
             compaction_l0_semaphore,
             l0_flush_delay_threshold,
@@ -959,9 +920,6 @@ impl TenantConfig {
                 .as_ref()
                 .unwrap_or(&global_conf.compaction_algorithm)
                 .clone(),
-            compaction_shard_ancestor: self
-                .compaction_shard_ancestor
-                .unwrap_or(global_conf.compaction_shard_ancestor),
             compaction_l0_first: self
                 .compaction_l0_first
                 .unwrap_or(global_conf.compaction_l0_first),
@@ -1803,8 +1761,6 @@ pub struct TopTenantShardsResponse {
 }
 
 pub mod virtual_file {
-    use std::sync::LazyLock;
-
     #[derive(
         Copy,
         Clone,
@@ -1842,38 +1798,14 @@ pub mod virtual_file {
     pub enum IoMode {
         /// Uses buffered IO.
         Buffered,
-        /// Uses direct IO for reads only.
+        /// Uses direct IO, error out if the operation fails.
         #[cfg(target_os = "linux")]
         Direct,
-        /// Use direct IO for reads and writes.
-        #[cfg(target_os = "linux")]
-        DirectRw,
     }
 
     impl IoMode {
-        pub fn preferred() -> Self {
-            // The default behavior when running Rust unit tests without any further
-            // flags is to use the newest behavior (DirectRw).
-            // The CI uses the following environment variable to unit tests for all
-            // different modes.
-            // NB: the Python regression & perf tests have their own defaults management
-            // that writes pageserver.toml; they do not use this variable.
-            if cfg!(test) {
-                static CACHED: LazyLock<IoMode> = LazyLock::new(|| {
-                    utils::env::var_serde_json_string(
-                        "NEON_PAGESERVER_UNIT_TEST_VIRTUAL_FILE_IO_MODE",
-                    )
-                    .unwrap_or(
-                        #[cfg(target_os = "linux")]
-                        IoMode::DirectRw,
-                        #[cfg(not(target_os = "linux"))]
-                        IoMode::Buffered,
-                    )
-                });
-                *CACHED
-            } else {
-                IoMode::Buffered
-            }
+        pub const fn preferred() -> Self {
+            Self::Buffered
         }
     }
 
@@ -1885,8 +1817,6 @@ pub mod virtual_file {
                 v if v == (IoMode::Buffered as u8) => IoMode::Buffered,
                 #[cfg(target_os = "linux")]
                 v if v == (IoMode::Direct as u8) => IoMode::Direct,
-                #[cfg(target_os = "linux")]
-                v if v == (IoMode::DirectRw as u8) => IoMode::DirectRw,
                 x => return Err(x),
             })
         }

libs/pageserver_api/src/upcall_api.rs

@@ -4,10 +4,10 @@
 //! See docs/rfcs/025-generation-numbers.md
 
 use serde::{Deserialize, Serialize};
-use utils::id::{NodeId, TimelineId};
+use utils::id::NodeId;
 
 use crate::controller_api::NodeRegisterRequest;
-use crate::models::{LocationConfigMode, ShardImportStatus};
+use crate::models::LocationConfigMode;
 use crate::shard::TenantShardId;
 
 /// Upcall message sent by the pageserver to the configured `control_plane_api` on
@@ -62,10 +62,3 @@ pub struct ValidateResponseTenant {
     pub id: TenantShardId,
     pub valid: bool,
 }
-
-#[derive(Serialize, Deserialize)]
-pub struct PutTimelineImportStatusRequest {
-    pub tenant_shard_id: TenantShardId,
-    pub timeline_id: TimelineId,
-    pub status: ShardImportStatus,
-}

libs/remote_keys/Cargo.toml

@@ -0,0 +1,13 @@
+[package]
+name = "remote_keys"
+version = "0.1.0"
+edition = "2024"
+license.workspace = true
+
+[dependencies]
+anyhow.workspace = true
+utils.workspace = true
+workspace_hack.workspace = true
+
+[dev-dependencies]
+rand.workspace = true

libs/remote_keys/src/lib.rs

@@ -0,0 +1,42 @@
+//! A module that provides a KMS implementation that generates and unwraps the keys.
+//!
+
+/// A KMS implementation that does static wrapping and unwrapping of the keys.
+pub struct NaiveKms {
+    account_id: String,
+}
+
+impl NaiveKms {
+    pub fn new(account_id: String) -> Self {
+        Self { account_id }
+    }
+
+    pub fn encrypt(&self, plain: &[u8]) -> anyhow::Result<Vec<u8>> {
+        let wrapped = [self.account_id.as_bytes(), "-wrapped-".as_bytes(), plain].concat();
+        Ok(wrapped)
+    }
+
+    pub fn decrypt(&self, wrapped: &[u8]) -> anyhow::Result<Vec<u8>> {
+        let Some(wrapped) = wrapped.strip_prefix(self.account_id.as_bytes()) else {
+            return Err(anyhow::anyhow!("invalid key"));
+        };
+        let Some(plain) = wrapped.strip_prefix(b"-wrapped-") else {
+            return Err(anyhow::anyhow!("invalid key"));
+        };
+        Ok(plain.to_vec())
+    }
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    #[test]
+    fn test_generate_key() {
+        let kms = NaiveKms::new("test-tenant".to_string());
+        let data = rand::random::<[u8; 32]>().to_vec();
+        let encrypted = kms.encrypt(&data).unwrap();
+        let decrypted = kms.decrypt(&encrypted).unwrap();
+        assert_eq!(data, decrypted);
+    }
+}

libs/remote_storage/Cargo.toml

@@ -13,6 +13,7 @@ aws-smithy-async.workspace = true
 aws-smithy-types.workspace = true
 aws-config.workspace = true
 aws-sdk-s3.workspace = true
+base64.workspace = true
 bytes.workspace = true
 camino = { workspace = true, features = ["serde1"] }
 humantime-serde.workspace = true
@@ -27,6 +28,7 @@ tokio-util = { workspace = true, features = ["compat"] }
 toml_edit.workspace = true
 tracing.workspace = true
 scopeguard.workspace = true
+md5.workspace = true
 metrics.workspace = true
 utils = { path = "../utils", default-features = false }
 pin-project-lite.workspace = true

libs/remote_storage/src/azure_blob.rs

@@ -14,9 +14,8 @@ use anyhow::{Context, Result};
 use azure_core::request_options::{IfMatchCondition, MaxResults, Metadata, Range};
 use azure_core::{Continuable, HttpClient, RetryOptions, TransportOptions};
 use azure_storage::StorageCredentials;
+use azure_storage_blobs::blob::CopyStatus;
 use azure_storage_blobs::blob::operations::GetBlobBuilder;
-use azure_storage_blobs::blob::{Blob, CopyStatus};
-use azure_storage_blobs::container::operations::ListBlobsBuilder;
 use azure_storage_blobs::prelude::{ClientBuilder, ContainerClient};
 use bytes::Bytes;
 use futures::FutureExt;
@@ -254,15 +253,53 @@ impl AzureBlobStorage {
         download
     }
 
-    fn list_streaming_for_fn<T: Default + ListingCollector>(
+    async fn permit(
+        &self,
+        kind: RequestKind,
+        cancel: &CancellationToken,
+    ) -> Result<tokio::sync::SemaphorePermit<'_>, Cancelled> {
+        let acquire = self.concurrency_limiter.acquire(kind);
+
+        tokio::select! {
+            permit = acquire => Ok(permit.expect("never closed")),
+            _ = cancel.cancelled() => Err(Cancelled),
+        }
+    }
+
+    pub fn container_name(&self) -> &str {
+        &self.container_name
+    }
+}
+
+fn to_azure_metadata(metadata: StorageMetadata) -> Metadata {
+    let mut res = Metadata::new();
+    for (k, v) in metadata.0.into_iter() {
+        res.insert(k, v);
+    }
+    res
+}
+
+fn to_download_error(error: azure_core::Error) -> DownloadError {
+    if let Some(http_err) = error.as_http_error() {
+        match http_err.status() {
+            StatusCode::NotFound => DownloadError::NotFound,
+            StatusCode::NotModified => DownloadError::Unmodified,
+            StatusCode::BadRequest => DownloadError::BadInput(anyhow::Error::new(error)),
+            _ => DownloadError::Other(anyhow::Error::new(error)),
+        }
+    } else {
+        DownloadError::Other(error.into())
+    }
+}
+
+impl RemoteStorage for AzureBlobStorage {
+    fn list_streaming(
         &self,
         prefix: Option<&RemotePath>,
         mode: ListingMode,
         max_keys: Option<NonZeroU32>,
         cancel: &CancellationToken,
-        request_kind: RequestKind,
-        customize_builder: impl Fn(ListBlobsBuilder) -> ListBlobsBuilder,
-    ) -> impl Stream<Item = Result<T, DownloadError>> {
+    ) -> impl Stream<Item = Result<Listing, DownloadError>> {
         // get the passed prefix or if it is not set use prefix_in_bucket value
         let list_prefix = prefix.map(|p| self.relative_path_to_name(p)).or_else(|| {
             self.prefix_in_container.clone().map(|mut s| {
@@ -274,7 +311,7 @@ impl AzureBlobStorage {
         });
 
         async_stream::stream! {
-            let _permit = self.permit(request_kind, cancel).await?;
+            let _permit = self.permit(RequestKind::List, cancel).await?;
 
             let mut builder = self.client.list_blobs();
 
@@ -290,8 +327,6 @@ impl AzureBlobStorage {
                 builder = builder.max_results(MaxResults::new(limit));
             }
 
-            builder = customize_builder(builder);
-
             let mut next_marker = None;
 
             let mut timeout_try_cnt = 1;
@@ -347,20 +382,26 @@ impl AzureBlobStorage {
                     break;
                 };
 
-                let mut res = T::default();
+                let mut res = Listing::default();
                 next_marker = entry.continuation();
                 let prefix_iter = entry
                     .blobs
                     .prefixes()
                     .map(|prefix| self.name_to_relative_path(&prefix.name));
-                res.add_prefixes(self, prefix_iter);
+                res.prefixes.extend(prefix_iter);
 
                 let blob_iter = entry
                     .blobs
-                    .blobs();
+                    .blobs()
+                    .map(|k| ListingObject{
+                        key: self.name_to_relative_path(&k.name),
+                        last_modified: k.properties.last_modified.into(),
+                        size: k.properties.content_length,
+                    }
+                );
 
                 for key in blob_iter {
-                    res.add_blob(self, key);
+                    res.keys.push(key);
 
                     if let Some(mut mk) = max_keys {
                         assert!(mk > 0);
@@ -382,128 +423,6 @@ impl AzureBlobStorage {
         }
     }
 
-    async fn permit(
-        &self,
-        kind: RequestKind,
-        cancel: &CancellationToken,
-    ) -> Result<tokio::sync::SemaphorePermit<'_>, Cancelled> {
-        let acquire = self.concurrency_limiter.acquire(kind);
-
-        tokio::select! {
-            permit = acquire => Ok(permit.expect("never closed")),
-            _ = cancel.cancelled() => Err(Cancelled),
-        }
-    }
-
-    pub fn container_name(&self) -> &str {
-        &self.container_name
-    }
-}
-
-trait ListingCollector {
-    fn add_prefixes(&mut self, abs: &AzureBlobStorage, prefix_it: impl Iterator<Item = RemotePath>);
-    fn add_blob(&mut self, abs: &AzureBlobStorage, blob: &Blob);
-}
-
-impl ListingCollector for Listing {
-    fn add_prefixes(
-        &mut self,
-        _abs: &AzureBlobStorage,
-        prefix_it: impl Iterator<Item = RemotePath>,
-    ) {
-        self.prefixes.extend(prefix_it);
-    }
-    fn add_blob(&mut self, abs: &AzureBlobStorage, blob: &Blob) {
-        self.keys.push(ListingObject {
-            key: abs.name_to_relative_path(&blob.name),
-            last_modified: blob.properties.last_modified.into(),
-            size: blob.properties.content_length,
-        });
-    }
-}
-
-impl ListingCollector for crate::VersionListing {
-    fn add_prefixes(
-        &mut self,
-        _abs: &AzureBlobStorage,
-        _prefix_it: impl Iterator<Item = RemotePath>,
-    ) {
-        // nothing
-    }
-    fn add_blob(&mut self, abs: &AzureBlobStorage, blob: &Blob) {
-        let id = crate::VersionId(blob.version_id.clone().expect("didn't find version ID"));
-        self.versions.push(crate::Version {
-            key: abs.name_to_relative_path(&blob.name),
-            last_modified: blob.properties.last_modified.into(),
-            kind: crate::VersionKind::Version(id),
-        });
-    }
-}
-
-fn to_azure_metadata(metadata: StorageMetadata) -> Metadata {
-    let mut res = Metadata::new();
-    for (k, v) in metadata.0.into_iter() {
-        res.insert(k, v);
-    }
-    res
-}
-
-fn to_download_error(error: azure_core::Error) -> DownloadError {
-    if let Some(http_err) = error.as_http_error() {
-        match http_err.status() {
-            StatusCode::NotFound => DownloadError::NotFound,
-            StatusCode::NotModified => DownloadError::Unmodified,
-            StatusCode::BadRequest => DownloadError::BadInput(anyhow::Error::new(error)),
-            _ => DownloadError::Other(anyhow::Error::new(error)),
-        }
-    } else {
-        DownloadError::Other(error.into())
-    }
-}
-
-impl RemoteStorage for AzureBlobStorage {
-    fn list_streaming(
-        &self,
-        prefix: Option<&RemotePath>,
-        mode: ListingMode,
-        max_keys: Option<NonZeroU32>,
-        cancel: &CancellationToken,
-    ) -> impl Stream<Item = Result<Listing, DownloadError>> {
-        let customize_builder = |builder| builder;
-        let kind = RequestKind::ListVersions;
-        self.list_streaming_for_fn(prefix, mode, max_keys, cancel, kind, customize_builder)
-    }
-
-    async fn list_versions(
-        &self,
-        prefix: Option<&RemotePath>,
-        mode: ListingMode,
-        max_keys: Option<NonZeroU32>,
-        cancel: &CancellationToken,
-    ) -> std::result::Result<crate::VersionListing, DownloadError> {
-        let customize_builder = |mut builder: ListBlobsBuilder| {
-            builder = builder.include_versions(true);
-            builder
-        };
-        let kind = RequestKind::ListVersions;
-
-        let mut stream = std::pin::pin!(self.list_streaming_for_fn(
-            prefix,
-            mode,
-            max_keys,
-            cancel,
-            kind,
-            customize_builder
-        ));
-        let mut combined: crate::VersionListing =
-            stream.next().await.expect("At least one item required")?;
-        while let Some(list) = stream.next().await {
-            let list = list?;
-            combined.versions.extend(list.versions.into_iter());
-        }
-        Ok(combined)
-    }
-
     async fn head_object(
         &self,
         key: &RemotePath,
@@ -613,12 +532,7 @@ impl RemoteStorage for AzureBlobStorage {
         let mut builder = blob_client.get();
 
         if let Some(ref etag) = opts.etag {
-            builder = builder.if_match(IfMatchCondition::NotMatch(etag.to_string()));
-        }
-
-        if let Some(ref version_id) = opts.version_id {
-            let version_id = azure_storage_blobs::prelude::VersionId::new(version_id.0.clone());
-            builder = builder.blob_versioning(version_id);
+            builder = builder.if_match(IfMatchCondition::NotMatch(etag.to_string()))
         }
 
         if let Some((start, end)) = opts.byte_range() {
@@ -636,6 +550,19 @@ impl RemoteStorage for AzureBlobStorage {
         self.download_for_builder(builder, timeout, cancel).await
     }
 
+    #[allow(unused_variables)]
+    async fn upload_with_encryption(
+        &self,
+        from: impl Stream<Item = std::io::Result<Bytes>> + Send + Sync + 'static,
+        data_size_bytes: usize,
+        to: &RemotePath,
+        metadata: Option<StorageMetadata>,
+        encryption_key: Option<&[u8]>,
+        cancel: &CancellationToken,
+    ) -> anyhow::Result<()> {
+        unimplemented!()
+    }
+
     async fn delete(&self, path: &RemotePath, cancel: &CancellationToken) -> anyhow::Result<()> {
         self.delete_objects(std::array::from_ref(path), cancel)
             .await

libs/remote_storage/src/lib.rs

@@ -176,32 +176,6 @@ pub struct Listing {
     pub keys: Vec<ListingObject>,
 }
 
-#[derive(Default)]
-pub struct VersionListing {
-    pub versions: Vec<Version>,
-}
-
-pub struct Version {
-    pub key: RemotePath,
-    pub last_modified: SystemTime,
-    pub kind: VersionKind,
-}
-
-impl Version {
-    pub fn version_id(&self) -> Option<&VersionId> {
-        match &self.kind {
-            VersionKind::Version(id) => Some(id),
-            VersionKind::DeletionMarker => None,
-        }
-    }
-}
-
-#[derive(Debug)]
-pub enum VersionKind {
-    DeletionMarker,
-    Version(VersionId),
-}
-
 /// Options for downloads. The default value is a plain GET.
 pub struct DownloadOpts {
     /// If given, returns [`DownloadError::Unmodified`] if the object still has
@@ -212,12 +186,12 @@ pub struct DownloadOpts {
     /// The end of the byte range to download, or unbounded. Must be after the
     /// start bound.
     pub byte_end: Bound<u64>,
-    /// Optionally request a specific version of a key
-    pub version_id: Option<VersionId>,
     /// Indicate whether we're downloading something small or large: this indirectly controls
     /// timeouts: for something like an index/manifest/heatmap, we should time out faster than
     /// for layer files
     pub kind: DownloadKind,
+    /// The encryption key to use for the download.
+    pub encryption_key: Option<Vec<u8>>,
 }
 
 pub enum DownloadKind {
@@ -225,17 +199,14 @@ pub enum DownloadKind {
     Small,
 }
 
-#[derive(Debug, Clone)]
-pub struct VersionId(pub String);
-
 impl Default for DownloadOpts {
     fn default() -> Self {
         Self {
             etag: Default::default(),
             byte_start: Bound::Unbounded,
             byte_end: Bound::Unbounded,
-            version_id: None,
             kind: DownloadKind::Large,
+            encryption_key: None,
         }
     }
 }
@@ -273,6 +244,15 @@ impl DownloadOpts {
                 None => format!("bytes={start}-"),
             })
     }
+
+    pub fn with_encryption_key(mut self, encryption_key: Option<impl AsRef<[u8]>>) -> Self {
+        self.encryption_key = encryption_key.map(|k| k.as_ref().to_vec());
+        self
+    }
+
+    pub fn encryption_key(&self) -> Option<&[u8]> {
+        self.encryption_key.as_deref()
+    }
 }
 
 /// Storage (potentially remote) API to manage its state.
@@ -327,14 +307,6 @@ pub trait RemoteStorage: Send + Sync + 'static {
         Ok(combined)
     }
 
-    async fn list_versions(
-        &self,
-        prefix: Option<&RemotePath>,
-        mode: ListingMode,
-        max_keys: Option<NonZeroU32>,
-        cancel: &CancellationToken,
-    ) -> Result<VersionListing, DownloadError>;
-
     /// Obtain metadata information about an object.
     async fn head_object(
         &self,
@@ -371,6 +343,19 @@ pub trait RemoteStorage: Send + Sync + 'static {
         cancel: &CancellationToken,
     ) -> Result<Download, DownloadError>;
 
+    /// Same as upload, but with remote encryption if the backend supports it (e.g. SSE-C on AWS).
+    async fn upload_with_encryption(
+        &self,
+        from: impl Stream<Item = std::io::Result<Bytes>> + Send + Sync + 'static,
+        // S3 PUT request requires the content length to be specified,
+        // otherwise it starts to fail with the concurrent connection count increasing.
+        data_size_bytes: usize,
+        to: &RemotePath,
+        metadata: Option<StorageMetadata>,
+        encryption_key: Option<&[u8]>,
+        cancel: &CancellationToken,
+    ) -> anyhow::Result<()>;
+
     /// Delete a single path from remote storage.
     ///
     /// If the operation fails because of timeout or cancellation, the root cause of the error will be
@@ -515,22 +500,6 @@ impl<Other: RemoteStorage> GenericRemoteStorage<Arc<Other>> {
         }
     }
 
-    // See [`RemoteStorage::list_versions`].
-    pub async fn list_versions<'a>(
-        &'a self,
-        prefix: Option<&'a RemotePath>,
-        mode: ListingMode,
-        max_keys: Option<NonZeroU32>,
-        cancel: &'a CancellationToken,
-    ) -> Result<VersionListing, DownloadError> {
-        match self {
-            Self::LocalFs(s) => s.list_versions(prefix, mode, max_keys, cancel).await,
-            Self::AwsS3(s) => s.list_versions(prefix, mode, max_keys, cancel).await,
-            Self::AzureBlob(s) => s.list_versions(prefix, mode, max_keys, cancel).await,
-            Self::Unreliable(s) => s.list_versions(prefix, mode, max_keys, cancel).await,
-        }
-    }
-
     // See [`RemoteStorage::head_object`].
     pub async fn head_object(
         &self,
@@ -671,6 +640,63 @@ impl<Other: RemoteStorage> GenericRemoteStorage<Arc<Other>> {
             }
         }
     }
+
+    pub async fn upload_with_encryption(
+        &self,
+        from: impl Stream<Item = std::io::Result<Bytes>> + Send + Sync + 'static,
+        data_size_bytes: usize,
+        to: &RemotePath,
+        metadata: Option<StorageMetadata>,
+        encryption_key: Option<&[u8]>,
+        cancel: &CancellationToken,
+    ) -> anyhow::Result<()> {
+        match self {
+            Self::LocalFs(s) => {
+                s.upload_with_encryption(
+                    from,
+                    data_size_bytes,
+                    to,
+                    metadata,
+                    encryption_key,
+                    cancel,
+                )
+                .await
+            }
+            Self::AwsS3(s) => {
+                s.upload_with_encryption(
+                    from,
+                    data_size_bytes,
+                    to,
+                    metadata,
+                    encryption_key,
+                    cancel,
+                )
+                .await
+            }
+            Self::AzureBlob(s) => {
+                s.upload_with_encryption(
+                    from,
+                    data_size_bytes,
+                    to,
+                    metadata,
+                    encryption_key,
+                    cancel,
+                )
+                .await
+            }
+            Self::Unreliable(s) => {
+                s.upload_with_encryption(
+                    from,
+                    data_size_bytes,
+                    to,
+                    metadata,
+                    encryption_key,
+                    cancel,
+                )
+                .await
+            }
+        }
+    }
 }
 
 impl GenericRemoteStorage {
@@ -783,7 +809,6 @@ impl ConcurrencyLimiter {
             RequestKind::Copy => &self.write,
             RequestKind::TimeTravel => &self.write,
             RequestKind::Head => &self.read,
-            RequestKind::ListVersions => &self.read,
         }
     }
 

libs/remote_storage/src/local_fs.rs

@@ -445,16 +445,6 @@ impl RemoteStorage for LocalFs {
         }
     }
 
-    async fn list_versions(
-        &self,
-        _prefix: Option<&RemotePath>,
-        _mode: ListingMode,
-        _max_keys: Option<NonZeroU32>,
-        _cancel: &CancellationToken,
-    ) -> Result<crate::VersionListing, DownloadError> {
-        unimplemented!()
-    }
-
     async fn head_object(
         &self,
         key: &RemotePath,
@@ -570,6 +560,19 @@ impl RemoteStorage for LocalFs {
         }
     }
 
+    #[allow(unused_variables)]
+    async fn upload_with_encryption(
+        &self,
+        from: impl Stream<Item = std::io::Result<Bytes>> + Send + Sync + 'static,
+        data_size_bytes: usize,
+        to: &RemotePath,
+        metadata: Option<StorageMetadata>,
+        encryption_key: Option<&[u8]>,
+        cancel: &CancellationToken,
+    ) -> anyhow::Result<()> {
+        unimplemented!()
+    }
+
     async fn delete_objects(
         &self,
         paths: &[RemotePath],

libs/remote_storage/src/metrics.rs

@@ -14,7 +14,6 @@ pub(crate) enum RequestKind {
     Copy = 4,
     TimeTravel = 5,
     Head = 6,
-    ListVersions = 7,
 }
 
 use RequestKind::*;
@@ -30,7 +29,6 @@ impl RequestKind {
             Copy => "copy_object",
             TimeTravel => "time_travel_recover",
             Head => "head_object",
-            ListVersions => "list_versions",
         }
     }
     const fn as_index(&self) -> usize {
@@ -38,10 +36,7 @@ impl RequestKind {
     }
 }
 
-const REQUEST_KIND_LIST: &[RequestKind] =
-    &[Get, Put, Delete, List, Copy, TimeTravel, Head, ListVersions];
-
-const REQUEST_KIND_COUNT: usize = REQUEST_KIND_LIST.len();
+const REQUEST_KIND_COUNT: usize = 7;
 pub(crate) struct RequestTyped<C>([C; REQUEST_KIND_COUNT]);
 
 impl<C> RequestTyped<C> {
@@ -50,11 +45,12 @@ impl<C> RequestTyped<C> {
     }
 
     fn build_with(mut f: impl FnMut(RequestKind) -> C) -> Self {
-        let mut it = REQUEST_KIND_LIST.iter();
+        use RequestKind::*;
+        let mut it = [Get, Put, Delete, List, Copy, TimeTravel, Head].into_iter();
         let arr = std::array::from_fn::<C, REQUEST_KIND_COUNT, _>(|index| {
             let next = it.next().unwrap();
             assert_eq!(index, next.as_index());
-            f(*next)
+            f(next)
         });
 
         if let Some(next) = it.next() {

libs/remote_storage/src/s3_bucket.rs

@@ -21,8 +21,9 @@ use aws_sdk_s3::config::{AsyncSleep, IdentityCache, Region, SharedAsyncSleep};
 use aws_sdk_s3::error::SdkError;
 use aws_sdk_s3::operation::get_object::GetObjectError;
 use aws_sdk_s3::operation::head_object::HeadObjectError;
-use aws_sdk_s3::types::{Delete, ObjectIdentifier, StorageClass};
+use aws_sdk_s3::types::{Delete, DeleteMarkerEntry, ObjectIdentifier, ObjectVersion, StorageClass};
 use aws_smithy_async::rt::sleep::TokioSleep;
+use aws_smithy_types::DateTime;
 use aws_smithy_types::body::SdkBody;
 use aws_smithy_types::byte_stream::ByteStream;
 use aws_smithy_types::date_time::ConversionError;
@@ -45,7 +46,7 @@ use crate::support::PermitCarrying;
 use crate::{
     ConcurrencyLimiter, Download, DownloadError, DownloadOpts, Listing, ListingMode, ListingObject,
     MAX_KEYS_PER_DELETE_S3, REMOTE_STORAGE_PREFIX_SEPARATOR, RemotePath, RemoteStorage,
-    TimeTravelError, TimeoutOrCancel, Version, VersionId, VersionKind, VersionListing,
+    TimeTravelError, TimeoutOrCancel,
 };
 
 /// AWS S3 storage.
@@ -65,8 +66,10 @@ struct GetObjectRequest {
     key: String,
     etag: Option<String>,
     range: Option<String>,
-    version_id: Option<String>,
+    /// Base64 encoded SSE-C key for server-side encryption.
+    sse_c_key: Option<Vec<u8>>,
 }
+
 impl S3Bucket {
     /// Creates the S3 storage, errors if incorrect AWS S3 configuration provided.
     pub async fn new(remote_storage_config: &S3Config, timeout: Duration) -> anyhow::Result<Self> {
@@ -251,13 +254,19 @@ impl S3Bucket {
             .get_object()
             .bucket(request.bucket)
             .key(request.key)
-            .set_version_id(request.version_id)
             .set_range(request.range);
 
         if let Some(etag) = request.etag {
             builder = builder.if_none_match(etag);
         }
 
+        if let Some(encryption_key) = request.sse_c_key {
+            builder = builder.sse_customer_algorithm("AES256");
+            builder = builder.sse_customer_key(base64::encode(&encryption_key));
+            builder = builder
+                .sse_customer_key_md5(base64::encode(md5::compute(&encryption_key).as_slice()));
+        }
+
         let get_object = builder.send();
 
         let get_object = tokio::select! {
@@ -406,124 +415,6 @@ impl S3Bucket {
         Ok(())
     }
 
-    async fn list_versions_with_permit(
-        &self,
-        _permit: &tokio::sync::SemaphorePermit<'_>,
-        prefix: Option<&RemotePath>,
-        mode: ListingMode,
-        max_keys: Option<NonZeroU32>,
-        cancel: &CancellationToken,
-    ) -> Result<crate::VersionListing, DownloadError> {
-        // get the passed prefix or if it is not set use prefix_in_bucket value
-        let prefix = prefix
-            .map(|p| self.relative_path_to_s3_object(p))
-            .or_else(|| self.prefix_in_bucket.clone());
-
-        let warn_threshold = 3;
-        let max_retries = 10;
-        let is_permanent = |e: &_| matches!(e, DownloadError::Cancelled);
-
-        let mut key_marker = None;
-        let mut version_id_marker = None;
-        let mut versions_and_deletes = Vec::new();
-
-        loop {
-            let response = backoff::retry(
-                || async {
-                    let mut request = self
-                        .client
-                        .list_object_versions()
-                        .bucket(self.bucket_name.clone())
-                        .set_prefix(prefix.clone())
-                        .set_key_marker(key_marker.clone())
-                        .set_version_id_marker(version_id_marker.clone());
-
-                    if let ListingMode::WithDelimiter = mode {
-                        request = request.delimiter(REMOTE_STORAGE_PREFIX_SEPARATOR.to_string());
-                    }
-
-                    let op = request.send();
-
-                    tokio::select! {
-                        res = op => res.map_err(|e| DownloadError::Other(e.into())),
-                        _ = cancel.cancelled() => Err(DownloadError::Cancelled),
-                    }
-                },
-                is_permanent,
-                warn_threshold,
-                max_retries,
-                "listing object versions",
-                cancel,
-            )
-            .await
-            .ok_or_else(|| DownloadError::Cancelled)
-            .and_then(|x| x)?;
-
-            tracing::trace!(
-                "  Got List response version_id_marker={:?}, key_marker={:?}",
-                response.version_id_marker,
-                response.key_marker
-            );
-            let versions = response
-                .versions
-                .unwrap_or_default()
-                .into_iter()
-                .map(|version| {
-                    let key = version.key.expect("response does not contain a key");
-                    let key = self.s3_object_to_relative_path(&key);
-                    let version_id = VersionId(version.version_id.expect("needing version id"));
-                    let last_modified =
-                        SystemTime::try_from(version.last_modified.expect("no last_modified"))?;
-                    Ok(Version {
-                        key,
-                        last_modified,
-                        kind: crate::VersionKind::Version(version_id),
-                    })
-                });
-            let deletes = response
-                .delete_markers
-                .unwrap_or_default()
-                .into_iter()
-                .map(|version| {
-                    let key = version.key.expect("response does not contain a key");
-                    let key = self.s3_object_to_relative_path(&key);
-                    let last_modified =
-                        SystemTime::try_from(version.last_modified.expect("no last_modified"))?;
-                    Ok(Version {
-                        key,
-                        last_modified,
-                        kind: crate::VersionKind::DeletionMarker,
-                    })
-                });
-            itertools::process_results(versions.chain(deletes), |n_vds| {
-                versions_and_deletes.extend(n_vds)
-            })
-            .map_err(DownloadError::Other)?;
-            fn none_if_empty(v: Option<String>) -> Option<String> {
-                v.filter(|v| !v.is_empty())
-            }
-            version_id_marker = none_if_empty(response.next_version_id_marker);
-            key_marker = none_if_empty(response.next_key_marker);
-            if version_id_marker.is_none() {
-                // The final response is not supposed to be truncated
-                if response.is_truncated.unwrap_or_default() {
-                    return Err(DownloadError::Other(anyhow::anyhow!(
-                        "Received truncated ListObjectVersions response for prefix={prefix:?}"
-                    )));
-                }
-                break;
-            }
-            if let Some(max_keys) = max_keys {
-                if versions_and_deletes.len() >= max_keys.get().try_into().unwrap() {
-                    return Err(DownloadError::Other(anyhow::anyhow!("too many versions")));
-                }
-            }
-        }
-        Ok(VersionListing {
-            versions: versions_and_deletes,
-        })
-    }
-
     pub fn bucket_name(&self) -> &str {
         &self.bucket_name
     }
@@ -740,19 +631,6 @@ impl RemoteStorage for S3Bucket {
         }
     }
 
-    async fn list_versions(
-        &self,
-        prefix: Option<&RemotePath>,
-        mode: ListingMode,
-        max_keys: Option<NonZeroU32>,
-        cancel: &CancellationToken,
-    ) -> Result<crate::VersionListing, DownloadError> {
-        let kind = RequestKind::ListVersions;
-        let permit = self.permit(kind, cancel).await?;
-        self.list_versions_with_permit(&permit, prefix, mode, max_keys, cancel)
-            .await
-    }
-
     async fn head_object(
         &self,
         key: &RemotePath,
@@ -825,12 +703,13 @@ impl RemoteStorage for S3Bucket {
         })
     }
 
-    async fn upload(
+    async fn upload_with_encryption(
         &self,
         from: impl Stream<Item = std::io::Result<Bytes>> + Send + Sync + 'static,
         from_size_bytes: usize,
         to: &RemotePath,
         metadata: Option<StorageMetadata>,
+        encryption_key: Option<&[u8]>,
         cancel: &CancellationToken,
     ) -> anyhow::Result<()> {
         let kind = RequestKind::Put;
@@ -841,7 +720,7 @@ impl RemoteStorage for S3Bucket {
         let body = StreamBody::new(from.map(|x| x.map(Frame::data)));
         let bytes_stream = ByteStream::new(SdkBody::from_body_1_x(body));
 
-        let upload = self
+        let mut upload = self
             .client
             .put_object()
             .bucket(self.bucket_name.clone())
@@ -849,8 +728,17 @@ impl RemoteStorage for S3Bucket {
             .set_metadata(metadata.map(|m| m.0))
             .set_storage_class(self.upload_storage_class.clone())
             .content_length(from_size_bytes.try_into()?)
-            .body(bytes_stream)
-            .send();
+            .body(bytes_stream);
+
+        if let Some(encryption_key) = encryption_key {
+            upload = upload.sse_customer_algorithm("AES256");
+            let base64_key = base64::encode(encryption_key);
+            upload = upload.sse_customer_key(&base64_key);
+            upload = upload
+                .sse_customer_key_md5(base64::encode(md5::compute(encryption_key).as_slice()));
+        }
+
+        let upload = upload.send();
 
         let upload = tokio::time::timeout(self.timeout, upload);
 
@@ -874,6 +762,18 @@ impl RemoteStorage for S3Bucket {
         }
     }
 
+    async fn upload(
+        &self,
+        from: impl Stream<Item = std::io::Result<Bytes>> + Send + Sync + 'static,
+        data_size_bytes: usize,
+        to: &RemotePath,
+        metadata: Option<StorageMetadata>,
+        cancel: &CancellationToken,
+    ) -> anyhow::Result<()> {
+        self.upload_with_encryption(from, data_size_bytes, to, metadata, None, cancel)
+            .await
+    }
+
     async fn copy(
         &self,
         from: &RemotePath,
@@ -933,7 +833,7 @@ impl RemoteStorage for S3Bucket {
                 key: self.relative_path_to_s3_object(from),
                 etag: opts.etag.as_ref().map(|e| e.to_string()),
                 range: opts.byte_range_header(),
-                version_id: opts.version_id.as_ref().map(|v| v.0.to_owned()),
+                sse_c_key: opts.encryption_key.clone(),
             },
             cancel,
         )
@@ -978,25 +878,94 @@ impl RemoteStorage for S3Bucket {
         let kind = RequestKind::TimeTravel;
         let permit = self.permit(kind, cancel).await?;
 
+        let timestamp = DateTime::from(timestamp);
+        let done_if_after = DateTime::from(done_if_after);
+
         tracing::trace!("Target time: {timestamp:?}, done_if_after {done_if_after:?}");
 
-        // Limit the number of versions deletions, mostly so that we don't
-        // keep requesting forever if the list is too long, as we'd put the
-        // list in RAM.
-        // Building a list of 100k entries that reaches the limit roughly takes
-        // 40 seconds, and roughly corresponds to tenants of 2 TiB physical size.
-        const COMPLEXITY_LIMIT: Option<NonZeroU32> = NonZeroU32::new(100_000);
+        // get the passed prefix or if it is not set use prefix_in_bucket value
+        let prefix = prefix
+            .map(|p| self.relative_path_to_s3_object(p))
+            .or_else(|| self.prefix_in_bucket.clone());
 
-        let mode = ListingMode::NoDelimiter;
-        let version_listing = self
-            .list_versions_with_permit(&permit, prefix, mode, COMPLEXITY_LIMIT, cancel)
+        let warn_threshold = 3;
+        let max_retries = 10;
+        let is_permanent = |e: &_| matches!(e, TimeTravelError::Cancelled);
+
+        let mut key_marker = None;
+        let mut version_id_marker = None;
+        let mut versions_and_deletes = Vec::new();
+
+        loop {
+            let response = backoff::retry(
+                || async {
+                    let op = self
+                        .client
+                        .list_object_versions()
+                        .bucket(self.bucket_name.clone())
+                        .set_prefix(prefix.clone())
+                        .set_key_marker(key_marker.clone())
+                        .set_version_id_marker(version_id_marker.clone())
+                        .send();
+
+                    tokio::select! {
+                        res = op => res.map_err(|e| TimeTravelError::Other(e.into())),
+                        _ = cancel.cancelled() => Err(TimeTravelError::Cancelled),
+                    }
+                },
+                is_permanent,
+                warn_threshold,
+                max_retries,
+                "listing object versions for time_travel_recover",
+                cancel,
+            )
             .await
-            .map_err(|err| match err {
-                DownloadError::Other(e) => TimeTravelError::Other(e),
-                DownloadError::Cancelled => TimeTravelError::Cancelled,
-                other => TimeTravelError::Other(other.into()),
-            })?;
-        let versions_and_deletes = version_listing.versions;
+            .ok_or_else(|| TimeTravelError::Cancelled)
+            .and_then(|x| x)?;
+
+            tracing::trace!(
+                "  Got List response version_id_marker={:?}, key_marker={:?}",
+                response.version_id_marker,
+                response.key_marker
+            );
+            let versions = response
+                .versions
+                .unwrap_or_default()
+                .into_iter()
+                .map(VerOrDelete::from_version);
+            let deletes = response
+                .delete_markers
+                .unwrap_or_default()
+                .into_iter()
+                .map(VerOrDelete::from_delete_marker);
+            itertools::process_results(versions.chain(deletes), |n_vds| {
+                versions_and_deletes.extend(n_vds)
+            })
+            .map_err(TimeTravelError::Other)?;
+            fn none_if_empty(v: Option<String>) -> Option<String> {
+                v.filter(|v| !v.is_empty())
+            }
+            version_id_marker = none_if_empty(response.next_version_id_marker);
+            key_marker = none_if_empty(response.next_key_marker);
+            if version_id_marker.is_none() {
+                // The final response is not supposed to be truncated
+                if response.is_truncated.unwrap_or_default() {
+                    return Err(TimeTravelError::Other(anyhow::anyhow!(
+                        "Received truncated ListObjectVersions response for prefix={prefix:?}"
+                    )));
+                }
+                break;
+            }
+            // Limit the number of versions deletions, mostly so that we don't
+            // keep requesting forever if the list is too long, as we'd put the
+            // list in RAM.
+            // Building a list of 100k entries that reaches the limit roughly takes
+            // 40 seconds, and roughly corresponds to tenants of 2 TiB physical size.
+            const COMPLEXITY_LIMIT: usize = 100_000;
+            if versions_and_deletes.len() >= COMPLEXITY_LIMIT {
+                return Err(TimeTravelError::TooManyVersions);
+            }
+        }
 
         tracing::info!(
             "Built list for time travel with {} versions and deletions",
@@ -1012,26 +981,24 @@ impl RemoteStorage for S3Bucket {
         let mut vds_for_key = HashMap::<_, Vec<_>>::new();
 
         for vd in &versions_and_deletes {
-            let Version { key, .. } = &vd;
-            let version_id = vd.version_id().map(|v| v.0.as_str());
-            if version_id == Some("null") {
+            let VerOrDelete {
+                version_id, key, ..
+            } = &vd;
+            if version_id == "null" {
                 return Err(TimeTravelError::Other(anyhow!(
                     "Received ListVersions response for key={key} with version_id='null', \
                     indicating either disabled versioning, or legacy objects with null version id values"
                 )));
             }
-            tracing::trace!("Parsing version key={key} kind={:?}", vd.kind);
+            tracing::trace!(
+                "Parsing version key={key} version_id={version_id} kind={:?}",
+                vd.kind
+            );
 
             vds_for_key.entry(key).or_default().push(vd);
         }
-
-        let warn_threshold = 3;
-        let max_retries = 10;
-        let is_permanent = |e: &_| matches!(e, TimeTravelError::Cancelled);
-
         for (key, versions) in vds_for_key {
             let last_vd = versions.last().unwrap();
-            let key = self.relative_path_to_s3_object(key);
             if last_vd.last_modified > done_if_after {
                 tracing::trace!("Key {key} has version later than done_if_after, skipping");
                 continue;
@@ -1056,11 +1023,11 @@ impl RemoteStorage for S3Bucket {
                 do_delete = true;
             } else {
                 match &versions[version_to_restore_to - 1] {
-                    Version {
-                        kind: VersionKind::Version(version_id),
+                    VerOrDelete {
+                        kind: VerOrDeleteKind::Version,
+                        version_id,
                         ..
                     } => {
-                        let version_id = &version_id.0;
                         tracing::trace!("Copying old version {version_id} for {key}...");
                         // Restore the state to the last version by copying
                         let source_id =
@@ -1072,7 +1039,7 @@ impl RemoteStorage for S3Bucket {
                                     .client
                                     .copy_object()
                                     .bucket(self.bucket_name.clone())
-                                    .key(&key)
+                                    .key(key)
                                     .set_storage_class(self.upload_storage_class.clone())
                                     .copy_source(&source_id)
                                     .send();
@@ -1093,8 +1060,8 @@ impl RemoteStorage for S3Bucket {
                         .and_then(|x| x)?;
                         tracing::info!(%version_id, %key, "Copied old version in S3");
                     }
-                    Version {
-                        kind: VersionKind::DeletionMarker,
+                    VerOrDelete {
+                        kind: VerOrDeleteKind::DeleteMarker,
                         ..
                     } => {
                         do_delete = true;
@@ -1102,7 +1069,7 @@ impl RemoteStorage for S3Bucket {
                 }
             };
             if do_delete {
-                if matches!(last_vd.kind, VersionKind::DeletionMarker) {
+                if matches!(last_vd.kind, VerOrDeleteKind::DeleteMarker) {
                     // Key has since been deleted (but there was some history), no need to do anything
                     tracing::trace!("Key {key} already deleted, skipping.");
                 } else {
@@ -1130,6 +1097,62 @@ impl RemoteStorage for S3Bucket {
     }
 }
 
+// Save RAM and only store the needed data instead of the entire ObjectVersion/DeleteMarkerEntry
+struct VerOrDelete {
+    kind: VerOrDeleteKind,
+    last_modified: DateTime,
+    version_id: String,
+    key: String,
+}
+
+#[derive(Debug)]
+enum VerOrDeleteKind {
+    Version,
+    DeleteMarker,
+}
+
+impl VerOrDelete {
+    fn with_kind(
+        kind: VerOrDeleteKind,
+        last_modified: Option<DateTime>,
+        version_id: Option<String>,
+        key: Option<String>,
+    ) -> anyhow::Result<Self> {
+        let lvk = (last_modified, version_id, key);
+        let (Some(last_modified), Some(version_id), Some(key)) = lvk else {
+            anyhow::bail!(
+                "One (or more) of last_modified, key, and id is None. \
+            Is versioning enabled in the bucket? last_modified={:?}, version_id={:?}, key={:?}",
+                lvk.0,
+                lvk.1,
+                lvk.2,
+            );
+        };
+        Ok(Self {
+            kind,
+            last_modified,
+            version_id,
+            key,
+        })
+    }
+    fn from_version(v: ObjectVersion) -> anyhow::Result<Self> {
+        Self::with_kind(
+            VerOrDeleteKind::Version,
+            v.last_modified,
+            v.version_id,
+            v.key,
+        )
+    }
+    fn from_delete_marker(v: DeleteMarkerEntry) -> anyhow::Result<Self> {
+        Self::with_kind(
+            VerOrDeleteKind::DeleteMarker,
+            v.last_modified,
+            v.version_id,
+            v.key,
+        )
+    }
+}
+
 #[cfg(test)]
 mod tests {
     use std::num::NonZeroUsize;

libs/remote_storage/src/simulate_failures.rs

@@ -139,20 +139,6 @@ impl RemoteStorage for UnreliableWrapper {
         self.inner.list(prefix, mode, max_keys, cancel).await
     }
 
-    async fn list_versions(
-        &self,
-        prefix: Option<&RemotePath>,
-        mode: ListingMode,
-        max_keys: Option<NonZeroU32>,
-        cancel: &CancellationToken,
-    ) -> Result<crate::VersionListing, DownloadError> {
-        self.attempt(RemoteOp::ListPrefixes(prefix.cloned()))
-            .map_err(DownloadError::Other)?;
-        self.inner
-            .list_versions(prefix, mode, max_keys, cancel)
-            .await
-    }
-
     async fn head_object(
         &self,
         key: &RemotePath,
@@ -192,6 +178,19 @@ impl RemoteStorage for UnreliableWrapper {
         self.inner.download(from, opts, cancel).await
     }
 
+    #[allow(unused_variables)]
+    async fn upload_with_encryption(
+        &self,
+        from: impl Stream<Item = std::io::Result<Bytes>> + Send + Sync + 'static,
+        data_size_bytes: usize,
+        to: &RemotePath,
+        metadata: Option<StorageMetadata>,
+        encryption_key: Option<&[u8]>,
+        cancel: &CancellationToken,
+    ) -> anyhow::Result<()> {
+        unimplemented!()
+    }
+
     async fn delete(&self, path: &RemotePath, cancel: &CancellationToken) -> anyhow::Result<()> {
         self.delete_inner(path, true, cancel).await
     }

libs/remote_storage/tests/test_real_s3.rs

@@ -421,7 +421,7 @@ async fn download_is_timeouted(ctx: &mut MaybeEnabledStorage) {
     ))
     .unwrap();
 
-    let len = upload_large_enough_file(&ctx.client, &path, &cancel).await;
+    let len = upload_large_enough_file(&ctx.client, &path, &cancel, None).await;
 
     let timeout = std::time::Duration::from_secs(5);
 
@@ -500,7 +500,7 @@ async fn download_is_cancelled(ctx: &mut MaybeEnabledStorage) {
     ))
     .unwrap();
 
-    let file_len = upload_large_enough_file(&ctx.client, &path, &cancel).await;
+    let file_len = upload_large_enough_file(&ctx.client, &path, &cancel, None).await;
 
     {
         let stream = ctx
@@ -555,6 +555,7 @@ async fn upload_large_enough_file(
     client: &GenericRemoteStorage,
     path: &RemotePath,
     cancel: &CancellationToken,
+    encryption_key: Option<&[u8]>,
 ) -> usize {
     let header = bytes::Bytes::from_static("remote blob data content".as_bytes());
     let body = bytes::Bytes::from(vec![0u8; 1024]);
@@ -565,9 +566,54 @@ async fn upload_large_enough_file(
     let contents = futures::stream::iter(contents.map(std::io::Result::Ok));
 
     client
-        .upload(contents, len, path, None, cancel)
+        .upload_with_encryption(contents, len, path, None, encryption_key, cancel)
         .await
         .expect("upload succeeds");
 
     len
 }
+
+#[test_context(MaybeEnabledStorage)]
+#[tokio::test]
+async fn encryption_works(ctx: &mut MaybeEnabledStorage) {
+    let MaybeEnabledStorage::Enabled(ctx) = ctx else {
+        return;
+    };
+
+    let cancel = CancellationToken::new();
+
+    let path = RemotePath::new(Utf8Path::new(
+        format!("{}/file_to_copy", ctx.base_prefix).as_str(),
+    ))
+    .unwrap();
+
+    let key = rand::random::<[u8; 32]>();
+    let file_len = upload_large_enough_file(&ctx.client, &path, &cancel, Some(&key)).await;
+
+    {
+        let download = ctx
+            .client
+            .download(
+                &path,
+                &DownloadOpts::default().with_encryption_key(Some(&key)),
+                &cancel,
+            )
+            .await
+            .expect("should succeed");
+        let vec = download_to_vec(download).await.expect("should succeed");
+        assert_eq!(vec.len(), file_len);
+    }
+
+    {
+        // Download without encryption key should fail
+        let download = ctx
+            .client
+            .download(&path, &DownloadOpts::default(), &cancel)
+            .await;
+        assert!(download.is_err());
+    }
+
+    let cancel = CancellationToken::new();
+
+    ctx.client.delete_objects(&[path], &cancel).await.unwrap();
+}

libs/utils/Cargo.toml

@@ -29,7 +29,6 @@ futures = { workspace = true }
 jsonwebtoken.workspace = true
 nix = { workspace = true, features = ["ioctl"] }
 once_cell.workspace = true
-pem.workspace = true
 pin-project-lite.workspace = true
 regex.workspace = true
 serde.workspace = true

libs/utils/src/auth.rs

@@ -11,8 +11,7 @@ use camino::Utf8Path;
 use jsonwebtoken::{
     Algorithm, DecodingKey, EncodingKey, Header, TokenData, Validation, decode, encode,
 };
-use pem::Pem;
-use serde::{Deserialize, Serialize, de::DeserializeOwned};
+use serde::{Deserialize, Serialize};
 
 use crate::id::TenantId;
 
@@ -74,10 +73,7 @@ impl SwappableJwtAuth {
     pub fn swap(&self, jwt_auth: JwtAuth) {
         self.0.swap(Arc::new(jwt_auth));
     }
-    pub fn decode<D: DeserializeOwned>(
-        &self,
-        token: &str,
-    ) -> std::result::Result<TokenData<D>, AuthError> {
+    pub fn decode(&self, token: &str) -> std::result::Result<TokenData<Claims>, AuthError> {
         self.0.load().decode(token)
     }
 }
@@ -152,10 +148,7 @@ impl JwtAuth {
     /// The function tries the stored decoding keys in succession,
     /// and returns the first yielding a successful result.
     /// If there is no working decoding key, it returns the last error.
-    pub fn decode<D: DeserializeOwned>(
-        &self,
-        token: &str,
-    ) -> std::result::Result<TokenData<D>, AuthError> {
+    pub fn decode(&self, token: &str) -> std::result::Result<TokenData<Claims>, AuthError> {
         let mut res = None;
         for decoding_key in &self.decoding_keys {
             res = Some(decode(token, decoding_key, &self.validation));
@@ -180,8 +173,8 @@ impl std::fmt::Debug for JwtAuth {
 }
 
 // this function is used only for testing purposes in CLI e g generate tokens during init
-pub fn encode_from_key_file<S: Serialize>(claims: &S, pem: &Pem) -> Result<String> {
-    let key = EncodingKey::from_ed_der(pem.contents());
+pub fn encode_from_key_file<S: Serialize>(claims: &S, key_data: &[u8]) -> Result<String> {
+    let key = EncodingKey::from_ed_pem(key_data)?;
     Ok(encode(&Header::new(STORAGE_TOKEN_ALGORITHM), claims, &key)?)
 }
 
@@ -195,13 +188,13 @@ mod tests {
     //
     // openssl genpkey -algorithm ed25519 -out ed25519-priv.pem
     // openssl pkey -in ed25519-priv.pem -pubout -out ed25519-pub.pem
-    const TEST_PUB_KEY_ED25519: &str = r#"
+    const TEST_PUB_KEY_ED25519: &[u8] = br#"
 -----BEGIN PUBLIC KEY-----
 MCowBQYDK2VwAyEARYwaNBayR+eGI0iXB4s3QxE3Nl2g1iWbr6KtLWeVD/w=
 -----END PUBLIC KEY-----
 "#;
 
-    const TEST_PRIV_KEY_ED25519: &str = r#"
+    const TEST_PRIV_KEY_ED25519: &[u8] = br#"
 -----BEGIN PRIVATE KEY-----
 MC4CAQAwBQYDK2VwBCIEID/Drmc1AA6U/znNRWpF3zEGegOATQxfkdWxitcOMsIH
 -----END PRIVATE KEY-----
@@ -229,9 +222,9 @@ MC4CAQAwBQYDK2VwBCIEID/Drmc1AA6U/znNRWpF3zEGegOATQxfkdWxitcOMsIH
 
         // Check it can be validated with the public key
         let auth = JwtAuth::new(vec![
-            DecodingKey::from_ed_pem(TEST_PUB_KEY_ED25519.as_bytes()).unwrap(),
+            DecodingKey::from_ed_pem(TEST_PUB_KEY_ED25519).unwrap(),
         ]);
-        let claims_from_token: Claims = auth.decode(encoded_eddsa).unwrap().claims;
+        let claims_from_token = auth.decode(encoded_eddsa).unwrap().claims;
         assert_eq!(claims_from_token, expected_claims);
     }
 
@@ -242,14 +235,13 @@ MC4CAQAwBQYDK2VwBCIEID/Drmc1AA6U/znNRWpF3zEGegOATQxfkdWxitcOMsIH
             scope: Scope::Tenant,
         };
 
-        let pem = pem::parse(TEST_PRIV_KEY_ED25519).unwrap();
-        let encoded = encode_from_key_file(&claims, &pem).unwrap();
+        let encoded = encode_from_key_file(&claims, TEST_PRIV_KEY_ED25519).unwrap();
 
         // decode it back
         let auth = JwtAuth::new(vec![
-            DecodingKey::from_ed_pem(TEST_PUB_KEY_ED25519.as_bytes()).unwrap(),
+            DecodingKey::from_ed_pem(TEST_PUB_KEY_ED25519).unwrap(),
         ]);
-        let decoded: TokenData<Claims> = auth.decode(&encoded).unwrap();
+        let decoded = auth.decode(&encoded).unwrap();
 
         assert_eq!(decoded.claims, claims);
     }

object_storage/Cargo.toml

@@ -1,5 +1,5 @@
 [package]
-name = "endpoint_storage"
+name = "object_storage"
 version = "0.0.1"
 edition.workspace = true
 license.workspace = true

object_storage/src/app.rs

@@ -2,7 +2,7 @@ use anyhow::anyhow;
 use axum::body::{Body, Bytes};
 use axum::response::{IntoResponse, Response};
 use axum::{Router, http::StatusCode};
-use endpoint_storage::{PrefixS3Path, S3Path, Storage, bad_request, internal_error, not_found, ok};
+use object_storage::{PrefixS3Path, S3Path, Storage, bad_request, internal_error, not_found, ok};
 use remote_storage::TimeoutOrCancel;
 use remote_storage::{DownloadError, DownloadOpts, GenericRemoteStorage, RemotePath};
 use std::{sync::Arc, time::SystemTime, time::UNIX_EPOCH};
@@ -46,12 +46,12 @@ async fn metrics() -> Result {
 
 async fn get(S3Path { path }: S3Path, state: State) -> Result {
     info!(%path, "downloading");
-    let download_err = |err| {
-        if let DownloadError::NotFound = err {
-            info!(%path, %err, "downloading"); // 404 is not an issue of _this_ service
+    let download_err = |e| {
+        if let DownloadError::NotFound = e {
+            info!(%path, %e, "downloading"); // 404 is not an issue of _this_ service
             return not_found(&path);
         }
-        internal_error(err, &path, "downloading")
+        internal_error(e, &path, "downloading")
     };
     let cancel = state.cancel.clone();
     let opts = &DownloadOpts::default();
@@ -249,7 +249,7 @@ mod tests {
         };
 
         let proxy = Storage {
-            auth: endpoint_storage::JwtAuth::new(TEST_PUB_KEY_ED25519).unwrap(),
+            auth: object_storage::JwtAuth::new(TEST_PUB_KEY_ED25519).unwrap(),
             storage,
             cancel: cancel.clone(),
             max_upload_file_limit: usize::MAX,
@@ -343,14 +343,14 @@ MC4CAQAwBQYDK2VwBCIEID/Drmc1AA6U/znNRWpF3zEGegOATQxfkdWxitcOMsIH
         TimelineId::from_array([0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 1, 2, 3, 4, 5, 7]);
     const ENDPOINT_ID: &str = "ep-winter-frost-a662z3vg";
     fn token() -> String {
-        let claims = endpoint_storage::Claims {
+        let claims = object_storage::Claims {
             tenant_id: TENANT_ID,
             timeline_id: TIMELINE_ID,
             endpoint_id: ENDPOINT_ID.into(),
             exp: u64::MAX,
         };
         let key = jsonwebtoken::EncodingKey::from_ed_pem(TEST_PRIV_KEY_ED25519).unwrap();
-        let header = jsonwebtoken::Header::new(endpoint_storage::VALIDATION_ALGO);
+        let header = jsonwebtoken::Header::new(object_storage::VALIDATION_ALGO);
         jsonwebtoken::encode(&header, &claims, &key).unwrap()
     }
 
@@ -364,10 +364,7 @@ MC4CAQAwBQYDK2VwBCIEID/Drmc1AA6U/znNRWpF3zEGegOATQxfkdWxitcOMsIH
             vec![TIMELINE_ID.to_string(), TimelineId::generate().to_string()],
             vec![ENDPOINT_ID, "ep-ololo"]
         )
-        // first one is fully valid path, second path is valid for GET as
-        // read paths may have different endpoint if tenant and timeline matches
-        // (needed for prewarming RO->RW replica)
-        .skip(2);
+        .skip(1);
 
         for ((uri, method), (tenant, timeline, endpoint)) in iproduct!(routes(), args) {
             info!(%uri, %method, %tenant, %timeline, %endpoint);
@@ -478,16 +475,6 @@ MC4CAQAwBQYDK2VwBCIEID/Drmc1AA6U/znNRWpF3zEGegOATQxfkdWxitcOMsIH
         requests_chain(chain.into_iter(), |_| token()).await;
     }
 
-    #[testlog(tokio::test)]
-    async fn read_other_endpoint_data() {
-        let uri = format!("/{TENANT_ID}/{TIMELINE_ID}/other_endpoint/key");
-        let chain = vec![
-            (uri.clone(), "GET", "", StatusCode::NOT_FOUND, false),
-            (uri.clone(), "PUT", "", StatusCode::UNAUTHORIZED, false),
-        ];
-        requests_chain(chain.into_iter(), |_| token()).await;
-    }
-
     fn delete_prefix_token(uri: &str) -> String {
         use serde::Serialize;
         let parts = uri.split("/").collect::<Vec<&str>>();
@@ -495,7 +482,7 @@ MC4CAQAwBQYDK2VwBCIEID/Drmc1AA6U/znNRWpF3zEGegOATQxfkdWxitcOMsIH
         struct PrefixClaims {
             tenant_id: TenantId,
             timeline_id: Option<TimelineId>,
-            endpoint_id: Option<endpoint_storage::EndpointId>,
+            endpoint_id: Option<object_storage::EndpointId>,
             exp: u64,
         }
         let claims = PrefixClaims {
@@ -505,7 +492,7 @@ MC4CAQAwBQYDK2VwBCIEID/Drmc1AA6U/znNRWpF3zEGegOATQxfkdWxitcOMsIH
             exp: u64::MAX,
         };
         let key = jsonwebtoken::EncodingKey::from_ed_pem(TEST_PRIV_KEY_ED25519).unwrap();
-        let header = jsonwebtoken::Header::new(endpoint_storage::VALIDATION_ALGO);
+        let header = jsonwebtoken::Header::new(object_storage::VALIDATION_ALGO);
         jsonwebtoken::encode(&header, &claims, &key).unwrap()
     }
 

object_storage/src/lib.rs

@@ -169,19 +169,10 @@ impl FromRequestParts<Arc<Storage>> for S3Path {
             .auth
             .decode(bearer.token())
             .map_err(|e| bad_request(e, "decoding token"))?;
-
-        // Read paths may have different endpoint ids. For readonly -> readwrite replica
-        // prewarming, endpoint must read other endpoint's data.
-        let endpoint_id = if parts.method == axum::http::Method::GET {
-            claims.endpoint_id.clone()
-        } else {
-            path.endpoint_id.clone()
-        };
-
         let route = Claims {
             tenant_id: path.tenant_id,
             timeline_id: path.timeline_id,
-            endpoint_id,
+            endpoint_id: path.endpoint_id.clone(),
             exp: claims.exp,
         };
         if route != claims {

object_storage/src/main.rs

@@ -1,4 +1,4 @@
-//! `endpoint_storage` is a service which provides API for uploading and downloading
+//! `object_storage` is a service which provides API for uploading and downloading
 //! files. It is used by compute and control plane for accessing LFC prewarm data.
 //! This service is deployed either as a separate component or as part of compute image
 //! for large computes.
@@ -33,7 +33,7 @@ async fn main() -> anyhow::Result<()> {
 
     let config: String = std::env::args().skip(1).take(1).collect();
     if config.is_empty() {
-        anyhow::bail!("Usage: endpoint_storage config.json")
+        anyhow::bail!("Usage: object_storage config.json")
     }
     info!("Reading config from {config}");
     let config = std::fs::read_to_string(config.clone())?;
@@ -41,7 +41,7 @@ async fn main() -> anyhow::Result<()> {
     info!("Reading pemfile from {}", config.pemfile.clone());
     let pemfile = std::fs::read(config.pemfile.clone())?;
     info!("Loading public key from {}", config.pemfile.clone());
-    let auth = endpoint_storage::JwtAuth::new(&pemfile)?;
+    let auth = object_storage::JwtAuth::new(&pemfile)?;
 
     let listener = tokio::net::TcpListener::bind(config.listen).await.unwrap();
     info!("listening on {}", listener.local_addr().unwrap());
@@ -50,7 +50,7 @@ async fn main() -> anyhow::Result<()> {
     let cancel = tokio_util::sync::CancellationToken::new();
     app::check_storage_permissions(&storage, cancel.clone()).await?;
 
-    let proxy = std::sync::Arc::new(endpoint_storage::Storage {
+    let proxy = std::sync::Arc::new(object_storage::Storage {
         auth,
         storage,
         cancel: cancel.clone(),

pageserver/Cargo.toml

@@ -35,7 +35,6 @@ humantime.workspace = true
 humantime-serde.workspace = true
 hyper0.workspace = true
 itertools.workspace = true
-jsonwebtoken.workspace = true
 md5.workspace = true
 nix.workspace = true
 # hack to get the number of worker threads tokio uses
@@ -78,7 +77,6 @@ metrics.workspace = true
 pageserver_api.workspace = true
 pageserver_client.workspace = true # for ResponseErrorMessageExt TOOD refactor that
 pageserver_compaction.workspace = true
-pem.workspace = true
 postgres_connection.workspace = true
 postgres_ffi.workspace = true
 pq_proto.workspace = true
@@ -106,7 +104,6 @@ hex-literal.workspace = true
 tokio = { workspace = true, features = ["process", "sync", "fs", "rt", "io-util", "time", "test-util"] }
 indoc.workspace = true
 uuid.workspace = true
-rstest.workspace = true
 
 [[bench]]
 name = "bench_layer_map"

pageserver/benches/bench_ingest.rs

@@ -11,7 +11,6 @@ use pageserver::task_mgr::TaskKind;
 use pageserver::tenant::storage_layer::InMemoryLayer;
 use pageserver::{page_cache, virtual_file};
 use pageserver_api::key::Key;
-use pageserver_api::models::virtual_file::IoMode;
 use pageserver_api::shard::TenantShardId;
 use pageserver_api::value::Value;
 use tokio_util::sync::CancellationToken;
@@ -29,7 +28,6 @@ fn murmurhash32(mut h: u32) -> u32 {
     h
 }
 
-#[derive(serde::Serialize, Clone, Copy, Debug)]
 enum KeyLayout {
     /// Sequential unique keys
     Sequential,
@@ -39,7 +37,6 @@ enum KeyLayout {
     RandomReuse(u32),
 }
 
-#[derive(serde::Serialize, Clone, Copy, Debug)]
 enum WriteDelta {
     Yes,
     No,
@@ -141,15 +138,12 @@ async fn ingest(
 /// Wrapper to instantiate a tokio runtime
 fn ingest_main(
     conf: &'static PageServerConf,
-    io_mode: IoMode,
     put_size: usize,
     put_count: usize,
     key_layout: KeyLayout,
     write_delta: WriteDelta,
 ) {
-    pageserver::virtual_file::set_io_mode(io_mode);
-
-    let runtime = tokio::runtime::Builder::new_multi_thread()
+    let runtime = tokio::runtime::Builder::new_current_thread()
         .enable_all()
         .build()
         .unwrap();
@@ -180,245 +174,93 @@ fn criterion_benchmark(c: &mut Criterion) {
     virtual_file::init(
         16384,
         virtual_file::io_engine_for_bench(),
-        // immaterial, each `ingest_main` invocation below overrides this
         conf.virtual_file_io_mode,
-        // without actually doing syncs, buffered writes have an unfair advantage over direct IO writes
         virtual_file::SyncMode::Sync,
     );
     page_cache::init(conf.page_cache_size);
 
-    #[derive(serde::Serialize)]
-    struct ExplodedParameters {
-        io_mode: IoMode,
-        volume_mib: usize,
-        key_size: usize,
-        key_layout: KeyLayout,
-        write_delta: WriteDelta,
-    }
-    #[derive(Clone)]
-    struct HandPickedParameters {
-        volume_mib: usize,
-        key_size: usize,
-        key_layout: KeyLayout,
-        write_delta: WriteDelta,
-    }
-    let expect = vec![
-        // Small values (100b) tests
-        HandPickedParameters {
-            volume_mib: 128,
-            key_size: 100,
-            key_layout: KeyLayout::Sequential,
-            write_delta: WriteDelta::Yes,
-        },
-        HandPickedParameters {
-            volume_mib: 128,
-            key_size: 100,
-            key_layout: KeyLayout::Random,
-            write_delta: WriteDelta::Yes,
-        },
-        HandPickedParameters {
-            volume_mib: 128,
-            key_size: 100,
-            key_layout: KeyLayout::RandomReuse(0x3ff),
-            write_delta: WriteDelta::Yes,
-        },
-        HandPickedParameters {
-            volume_mib: 128,
-            key_size: 100,
-            key_layout: KeyLayout::Sequential,
-            write_delta: WriteDelta::No,
-        },
-        // Large values (8k) tests
-        HandPickedParameters {
-            volume_mib: 128,
-            key_size: 8192,
-            key_layout: KeyLayout::Sequential,
-            write_delta: WriteDelta::Yes,
-        },
-        HandPickedParameters {
-            volume_mib: 128,
-            key_size: 8192,
-            key_layout: KeyLayout::Sequential,
-            write_delta: WriteDelta::No,
-        },
-    ];
-    let exploded_parameters = {
-        let mut out = Vec::new();
-        for io_mode in [
-            IoMode::Buffered,
-            #[cfg(target_os = "linux")]
-            IoMode::Direct,
-            #[cfg(target_os = "linux")]
-            IoMode::DirectRw,
-        ] {
-            for param in expect.clone() {
-                let HandPickedParameters {
-                    volume_mib,
-                    key_size,
-                    key_layout,
-                    write_delta,
-                } = param;
-                out.push(ExplodedParameters {
-                    io_mode,
-                    volume_mib,
-                    key_size,
-                    key_layout,
-                    write_delta,
-                });
-            }
-        }
-        out
-    };
-    impl ExplodedParameters {
-        fn benchmark_id(&self) -> String {
-            let ExplodedParameters {
-                io_mode,
-                volume_mib,
-                key_size,
-                key_layout,
-                write_delta,
-            } = self;
-            format!(
-                "io_mode={io_mode:?} volume_mib={volume_mib:?} key_size_bytes={key_size:?} key_layout={key_layout:?} write_delta={write_delta:?}"
-            )
-        }
-    }
-    let mut group = c.benchmark_group("ingest");
-    for params in exploded_parameters {
-        let id = params.benchmark_id();
-        let ExplodedParameters {
-            io_mode,
-            volume_mib,
-            key_size,
-            key_layout,
-            write_delta,
-        } = params;
-        let put_count = volume_mib * 1024 * 1024 / key_size;
-        group.throughput(criterion::Throughput::Bytes((key_size * put_count) as u64));
+    {
+        let mut group = c.benchmark_group("ingest-small-values");
+        let put_size = 100usize;
+        let put_count = 128 * 1024 * 1024 / put_size;
+        group.throughput(criterion::Throughput::Bytes((put_size * put_count) as u64));
         group.sample_size(10);
-        group.bench_function(id, |b| {
-            b.iter(|| ingest_main(conf, io_mode, key_size, put_count, key_layout, write_delta))
+        group.bench_function("ingest 128MB/100b seq", |b| {
+            b.iter(|| {
+                ingest_main(
+                    conf,
+                    put_size,
+                    put_count,
+                    KeyLayout::Sequential,
+                    WriteDelta::Yes,
+                )
+            })
+        });
+        group.bench_function("ingest 128MB/100b rand", |b| {
+            b.iter(|| {
+                ingest_main(
+                    conf,
+                    put_size,
+                    put_count,
+                    KeyLayout::Random,
+                    WriteDelta::Yes,
+                )
+            })
+        });
+        group.bench_function("ingest 128MB/100b rand-1024keys", |b| {
+            b.iter(|| {
+                ingest_main(
+                    conf,
+                    put_size,
+                    put_count,
+                    KeyLayout::RandomReuse(0x3ff),
+                    WriteDelta::Yes,
+                )
+            })
+        });
+        group.bench_function("ingest 128MB/100b seq, no delta", |b| {
+            b.iter(|| {
+                ingest_main(
+                    conf,
+                    put_size,
+                    put_count,
+                    KeyLayout::Sequential,
+                    WriteDelta::No,
+                )
+            })
+        });
+    }
+
+    {
+        let mut group = c.benchmark_group("ingest-big-values");
+        let put_size = 8192usize;
+        let put_count = 128 * 1024 * 1024 / put_size;
+        group.throughput(criterion::Throughput::Bytes((put_size * put_count) as u64));
+        group.sample_size(10);
+        group.bench_function("ingest 128MB/8k seq", |b| {
+            b.iter(|| {
+                ingest_main(
+                    conf,
+                    put_size,
+                    put_count,
+                    KeyLayout::Sequential,
+                    WriteDelta::Yes,
+                )
+            })
+        });
+        group.bench_function("ingest 128MB/8k seq, no delta", |b| {
+            b.iter(|| {
+                ingest_main(
+                    conf,
+                    put_size,
+                    put_count,
+                    KeyLayout::Sequential,
+                    WriteDelta::No,
+                )
+            })
         });
     }
 }
 
 criterion_group!(benches, criterion_benchmark);
 criterion_main!(benches);
-
-/*
-cargo bench --bench bench_ingest
-
-im4gn.2xlarge:
-
-ingest/io_mode=Buffered volume_mib=128 key_size_bytes=100 key_layout=Sequential write_delta=Yes
-                        time:   [1.2901 s 1.2943 s 1.2991 s]
-                        thrpt:  [98.533 MiB/s 98.892 MiB/s 99.220 MiB/s]
-ingest/io_mode=Buffered volume_mib=128 key_size_bytes=100 key_layout=Random write_delta=Yes
-                        time:   [2.1387 s 2.1623 s 2.1845 s]
-                        thrpt:  [58.595 MiB/s 59.197 MiB/s 59.851 MiB/s]
-ingest/io_mode=Buffered volume_mib=128 key_size_bytes=100 key_layout=RandomReuse(1023) write_delta=Y...
-                        time:   [1.2036 s 1.2074 s 1.2122 s]
-                        thrpt:  [105.60 MiB/s 106.01 MiB/s 106.35 MiB/s]
-ingest/io_mode=Buffered volume_mib=128 key_size_bytes=100 key_layout=Sequential write_delta=No
-                        time:   [520.55 ms 521.46 ms 522.57 ms]
-                        thrpt:  [244.94 MiB/s 245.47 MiB/s 245.89 MiB/s]
-ingest/io_mode=Buffered volume_mib=128 key_size_bytes=8192 key_layout=Sequential write_delta=Yes
-                        time:   [440.33 ms 442.24 ms 444.10 ms]
-                        thrpt:  [288.22 MiB/s 289.43 MiB/s 290.69 MiB/s]
-ingest/io_mode=Buffered volume_mib=128 key_size_bytes=8192 key_layout=Sequential write_delta=No
-                        time:   [168.78 ms 169.42 ms 170.18 ms]
-                        thrpt:  [752.16 MiB/s 755.52 MiB/s 758.40 MiB/s]
-ingest/io_mode=Direct volume_mib=128 key_size_bytes=100 key_layout=Sequential write_delta=Yes
-                        time:   [1.2978 s 1.3094 s 1.3227 s]
-                        thrpt:  [96.775 MiB/s 97.758 MiB/s 98.632 MiB/s]
-ingest/io_mode=Direct volume_mib=128 key_size_bytes=100 key_layout=Random write_delta=Yes
-                        time:   [2.1976 s 2.2067 s 2.2154 s]
-                        thrpt:  [57.777 MiB/s 58.006 MiB/s 58.245 MiB/s]
-ingest/io_mode=Direct volume_mib=128 key_size_bytes=100 key_layout=RandomReuse(1023) write_delta=Yes
-                        time:   [1.2103 s 1.2160 s 1.2233 s]
-                        thrpt:  [104.64 MiB/s 105.26 MiB/s 105.76 MiB/s]
-ingest/io_mode=Direct volume_mib=128 key_size_bytes=100 key_layout=Sequential write_delta=No
-                        time:   [525.05 ms 526.37 ms 527.79 ms]
-                        thrpt:  [242.52 MiB/s 243.17 MiB/s 243.79 MiB/s]
-ingest/io_mode=Direct volume_mib=128 key_size_bytes=8192 key_layout=Sequential write_delta=Yes
-                        time:   [443.06 ms 444.88 ms 447.15 ms]
-                        thrpt:  [286.26 MiB/s 287.72 MiB/s 288.90 MiB/s]
-ingest/io_mode=Direct volume_mib=128 key_size_bytes=8192 key_layout=Sequential write_delta=No
-                        time:   [169.40 ms 169.80 ms 170.17 ms]
-                        thrpt:  [752.21 MiB/s 753.81 MiB/s 755.60 MiB/s]
-ingest/io_mode=DirectRw volume_mib=128 key_size_bytes=100 key_layout=Sequential write_delta=Yes
-                        time:   [1.2844 s 1.2915 s 1.2990 s]
-                        thrpt:  [98.536 MiB/s 99.112 MiB/s 99.657 MiB/s]
-ingest/io_mode=DirectRw volume_mib=128 key_size_bytes=100 key_layout=Random write_delta=Yes
-                        time:   [2.1431 s 2.1663 s 2.1900 s]
-                        thrpt:  [58.446 MiB/s 59.087 MiB/s 59.726 MiB/s]
-ingest/io_mode=DirectRw volume_mib=128 key_size_bytes=100 key_layout=RandomReuse(1023) write_delta=Y...
-                        time:   [1.1906 s 1.1926 s 1.1947 s]
-                        thrpt:  [107.14 MiB/s 107.33 MiB/s 107.51 MiB/s]
-ingest/io_mode=DirectRw volume_mib=128 key_size_bytes=100 key_layout=Sequential write_delta=No
-                        time:   [516.86 ms 518.25 ms 519.47 ms]
-                        thrpt:  [246.40 MiB/s 246.98 MiB/s 247.65 MiB/s]
-ingest/io_mode=DirectRw volume_mib=128 key_size_bytes=8192 key_layout=Sequential write_delta=Yes
-                        time:   [536.50 ms 536.53 ms 536.60 ms]
-                        thrpt:  [238.54 MiB/s 238.57 MiB/s 238.59 MiB/s]
-ingest/io_mode=DirectRw volume_mib=128 key_size_bytes=8192 key_layout=Sequential write_delta=No
-                        time:   [267.77 ms 267.90 ms 268.04 ms]
-                        thrpt:  [477.53 MiB/s 477.79 MiB/s 478.02 MiB/s]
-
-Hetzner AX102:
-
-ingest/io_mode=Buffered volume_mib=128 key_size_bytes=100 key_layout=Sequential write_delta=Yes
-                        time:   [836.58 ms 861.93 ms 886.57 ms]
-                        thrpt:  [144.38 MiB/s 148.50 MiB/s 153.00 MiB/s]
-ingest/io_mode=Buffered volume_mib=128 key_size_bytes=100 key_layout=Random write_delta=Yes
-                        time:   [1.2782 s 1.3191 s 1.3665 s]
-                        thrpt:  [93.668 MiB/s 97.037 MiB/s 100.14 MiB/s]
-ingest/io_mode=Buffered volume_mib=128 key_size_bytes=100 key_layout=RandomReuse(1023) write_delta=Y...
-                        time:   [791.27 ms 807.08 ms 822.95 ms]
-                        thrpt:  [155.54 MiB/s 158.60 MiB/s 161.77 MiB/s]
-ingest/io_mode=Buffered volume_mib=128 key_size_bytes=100 key_layout=Sequential write_delta=No
-                        time:   [310.78 ms 314.66 ms 318.47 ms]
-                        thrpt:  [401.92 MiB/s 406.79 MiB/s 411.87 MiB/s]
-ingest/io_mode=Buffered volume_mib=128 key_size_bytes=8192 key_layout=Sequential write_delta=Yes
-                        time:   [377.11 ms 387.77 ms 399.21 ms]
-                        thrpt:  [320.63 MiB/s 330.10 MiB/s 339.42 MiB/s]
-ingest/io_mode=Buffered volume_mib=128 key_size_bytes=8192 key_layout=Sequential write_delta=No
-                        time:   [128.37 ms 132.96 ms 138.55 ms]
-                        thrpt:  [923.83 MiB/s 962.69 MiB/s 997.11 MiB/s]
-ingest/io_mode=Direct volume_mib=128 key_size_bytes=100 key_layout=Sequential write_delta=Yes
-                        time:   [900.38 ms 914.88 ms 928.86 ms]
-                        thrpt:  [137.80 MiB/s 139.91 MiB/s 142.16 MiB/s]
-ingest/io_mode=Direct volume_mib=128 key_size_bytes=100 key_layout=Random write_delta=Yes
-                        time:   [1.2538 s 1.2936 s 1.3313 s]
-                        thrpt:  [96.149 MiB/s 98.946 MiB/s 102.09 MiB/s]
-ingest/io_mode=Direct volume_mib=128 key_size_bytes=100 key_layout=RandomReuse(1023) write_delta=Yes
-                        time:   [787.17 ms 803.89 ms 820.63 ms]
-                        thrpt:  [155.98 MiB/s 159.23 MiB/s 162.61 MiB/s]
-ingest/io_mode=Direct volume_mib=128 key_size_bytes=100 key_layout=Sequential write_delta=No
-                        time:   [318.78 ms 321.89 ms 324.74 ms]
-                        thrpt:  [394.16 MiB/s 397.65 MiB/s 401.53 MiB/s]
-ingest/io_mode=Direct volume_mib=128 key_size_bytes=8192 key_layout=Sequential write_delta=Yes
-                        time:   [374.01 ms 383.45 ms 393.20 ms]
-                        thrpt:  [325.53 MiB/s 333.81 MiB/s 342.24 MiB/s]
-ingest/io_mode=Direct volume_mib=128 key_size_bytes=8192 key_layout=Sequential write_delta=No
-                        time:   [137.98 ms 141.31 ms 143.57 ms]
-                        thrpt:  [891.58 MiB/s 905.79 MiB/s 927.66 MiB/s]
-ingest/io_mode=DirectRw volume_mib=128 key_size_bytes=100 key_layout=Sequential write_delta=Yes
-                        time:   [613.69 ms 622.48 ms 630.97 ms]
-                        thrpt:  [202.86 MiB/s 205.63 MiB/s 208.57 MiB/s]
-ingest/io_mode=DirectRw volume_mib=128 key_size_bytes=100 key_layout=Random write_delta=Yes
-                        time:   [1.0299 s 1.0766 s 1.1273 s]
-                        thrpt:  [113.55 MiB/s 118.90 MiB/s 124.29 MiB/s]
-ingest/io_mode=DirectRw volume_mib=128 key_size_bytes=100 key_layout=RandomReuse(1023) write_delta=Y...
-                        time:   [637.80 ms 647.78 ms 658.01 ms]
-                        thrpt:  [194.53 MiB/s 197.60 MiB/s 200.69 MiB/s]
-ingest/io_mode=DirectRw volume_mib=128 key_size_bytes=100 key_layout=Sequential write_delta=No
-                        time:   [266.09 ms 267.20 ms 268.31 ms]
-                        thrpt:  [477.06 MiB/s 479.04 MiB/s 481.04 MiB/s]
-ingest/io_mode=DirectRw volume_mib=128 key_size_bytes=8192 key_layout=Sequential write_delta=Yes
-                        time:   [269.34 ms 273.27 ms 277.69 ms]
-                        thrpt:  [460.95 MiB/s 468.40 MiB/s 475.24 MiB/s]
-ingest/io_mode=DirectRw volume_mib=128 key_size_bytes=8192 key_layout=Sequential write_delta=No
-                        time:   [123.18 ms 124.24 ms 125.15 ms]
-                        thrpt:  [1022.8 MiB/s 1.0061 GiB/s 1.0148 GiB/s]
-*/

pageserver/client/src/mgmt_api.rs

@@ -419,23 +419,6 @@ impl Client {
         }
     }
 
-    pub async fn timeline_detail(
-        &self,
-        tenant_shard_id: TenantShardId,
-        timeline_id: TimelineId,
-    ) -> Result<TimelineInfo> {
-        let uri = format!(
-            "{}/v1/tenant/{tenant_shard_id}/timeline/{timeline_id}",
-            self.mgmt_api_endpoint
-        );
-
-        self.request(Method::GET, &uri, ())
-            .await?
-            .json()
-            .await
-            .map_err(Error::ReceiveBody)
-    }
-
     pub async fn timeline_archival_config(
         &self,
         tenant_shard_id: TenantShardId,

pageserver/pagebench/src/cmd/getpage_latest_lsn.rs

@@ -68,13 +68,6 @@ pub(crate) struct Args {
     targets: Option<Vec<TenantTimelineId>>,
 }
 
-/// State shared by all clients
-#[derive(Debug)]
-struct SharedState {
-    start_work_barrier: tokio::sync::Barrier,
-    live_stats: LiveStats,
-}
-
 #[derive(Debug, Default)]
 struct LiveStats {
     completed_requests: AtomicU64,
@@ -247,26 +240,24 @@ async fn main_impl(
         all_ranges
     };
 
+    let live_stats = Arc::new(LiveStats::default());
+
     let num_live_stats_dump = 1;
     let num_work_sender_tasks = args.num_clients.get() * timelines.len();
     let num_main_impl = 1;
 
-    let shared_state = Arc::new(SharedState {
-        start_work_barrier: tokio::sync::Barrier::new(
-            num_live_stats_dump + num_work_sender_tasks + num_main_impl,
-        ),
-        live_stats: LiveStats::default(),
-    });
-    let cancel = CancellationToken::new();
+    let start_work_barrier = Arc::new(tokio::sync::Barrier::new(
+        num_live_stats_dump + num_work_sender_tasks + num_main_impl,
+    ));
 
-    let ss = shared_state.clone();
     tokio::spawn({
+        let stats = Arc::clone(&live_stats);
+        let start_work_barrier = Arc::clone(&start_work_barrier);
         async move {
-            ss.start_work_barrier.wait().await;
+            start_work_barrier.wait().await;
             loop {
                 let start = std::time::Instant::now();
                 tokio::time::sleep(std::time::Duration::from_secs(1)).await;
-                let stats = &ss.live_stats;
                 let completed_requests = stats.completed_requests.swap(0, Ordering::Relaxed);
                 let missed = stats.missed.swap(0, Ordering::Relaxed);
                 let elapsed = start.elapsed();
@@ -279,12 +270,14 @@ async fn main_impl(
         }
     });
 
+    let cancel = CancellationToken::new();
+
     let rps_period = args
         .per_client_rate
         .map(|rps_limit| Duration::from_secs_f64(1.0 / (rps_limit as f64)));
     let make_worker: &dyn Fn(WorkerId) -> Pin<Box<dyn Send + Future<Output = ()>>> = &|worker_id| {
-        let ss = shared_state.clone();
-        let cancel = cancel.clone();
+        let live_stats = live_stats.clone();
+        let start_work_barrier = start_work_barrier.clone();
         let ranges: Vec<KeyRange> = all_ranges
             .iter()
             .filter(|r| r.timeline == worker_id.timeline)
@@ -294,8 +287,85 @@ async fn main_impl(
             rand::distributions::weighted::WeightedIndex::new(ranges.iter().map(|v| v.len()))
                 .unwrap();
 
+        let cancel = cancel.clone();
         Box::pin(async move {
-            client_libpq(args, worker_id, ss, cancel, rps_period, ranges, weights).await
+            let client =
+                pageserver_client::page_service::Client::new(args.page_service_connstring.clone())
+                    .await
+                    .unwrap();
+            let mut client = client
+                .pagestream(worker_id.timeline.tenant_id, worker_id.timeline.timeline_id)
+                .await
+                .unwrap();
+
+            start_work_barrier.wait().await;
+            let client_start = Instant::now();
+            let mut ticks_processed = 0;
+            let mut inflight = VecDeque::new();
+            while !cancel.is_cancelled() {
+                // Detect if a request took longer than the RPS rate
+                if let Some(period) = &rps_period {
+                    let periods_passed_until_now =
+                        usize::try_from(client_start.elapsed().as_micros() / period.as_micros())
+                            .unwrap();
+
+                    if periods_passed_until_now > ticks_processed {
+                        live_stats.missed((periods_passed_until_now - ticks_processed) as u64);
+                    }
+                    ticks_processed = periods_passed_until_now;
+                }
+
+                while inflight.len() < args.queue_depth.get() {
+                    let start = Instant::now();
+                    let req = {
+                        let mut rng = rand::thread_rng();
+                        let r = &ranges[weights.sample(&mut rng)];
+                        let key: i128 = rng.gen_range(r.start..r.end);
+                        let key = Key::from_i128(key);
+                        assert!(key.is_rel_block_key());
+                        let (rel_tag, block_no) = key
+                            .to_rel_block()
+                            .expect("we filter non-rel-block keys out above");
+                        PagestreamGetPageRequest {
+                            hdr: PagestreamRequest {
+                                reqid: 0,
+                                request_lsn: if rng.gen_bool(args.req_latest_probability) {
+                                    Lsn::MAX
+                                } else {
+                                    r.timeline_lsn
+                                },
+                                not_modified_since: r.timeline_lsn,
+                            },
+                            rel: rel_tag,
+                            blkno: block_no,
+                        }
+                    };
+                    client.getpage_send(req).await.unwrap();
+                    inflight.push_back(start);
+                }
+
+                let start = inflight.pop_front().unwrap();
+                client.getpage_recv().await.unwrap();
+                let end = Instant::now();
+                live_stats.request_done();
+                ticks_processed += 1;
+                STATS.with(|stats| {
+                    stats
+                        .borrow()
+                        .lock()
+                        .unwrap()
+                        .observe(end.duration_since(start))
+                        .unwrap();
+                });
+
+                if let Some(period) = &rps_period {
+                    let next_at = client_start
+                        + Duration::from_micros(
+                            (ticks_processed) as u64 * u64::try_from(period.as_micros()).unwrap(),
+                        );
+                    tokio::time::sleep_until(next_at.into()).await;
+                }
+            }
         })
     };
 
@@ -317,7 +387,7 @@ async fn main_impl(
     };
 
     info!("waiting for everything to become ready");
-    shared_state.start_work_barrier.wait().await;
+    start_work_barrier.wait().await;
     info!("work started");
     if let Some(runtime) = args.runtime {
         tokio::time::sleep(runtime.into()).await;
@@ -346,91 +416,3 @@ async fn main_impl(
 
     anyhow::Ok(())
 }
-
-async fn client_libpq(
-    args: &Args,
-    worker_id: WorkerId,
-    shared_state: Arc<SharedState>,
-    cancel: CancellationToken,
-    rps_period: Option<Duration>,
-    ranges: Vec<KeyRange>,
-    weights: rand::distributions::weighted::WeightedIndex<i128>,
-) {
-    let client = pageserver_client::page_service::Client::new(args.page_service_connstring.clone())
-        .await
-        .unwrap();
-    let mut client = client
-        .pagestream(worker_id.timeline.tenant_id, worker_id.timeline.timeline_id)
-        .await
-        .unwrap();
-
-    shared_state.start_work_barrier.wait().await;
-    let client_start = Instant::now();
-    let mut ticks_processed = 0;
-    let mut inflight = VecDeque::new();
-    while !cancel.is_cancelled() {
-        // Detect if a request took longer than the RPS rate
-        if let Some(period) = &rps_period {
-            let periods_passed_until_now =
-                usize::try_from(client_start.elapsed().as_micros() / period.as_micros()).unwrap();
-
-            if periods_passed_until_now > ticks_processed {
-                shared_state
-                    .live_stats
-                    .missed((periods_passed_until_now - ticks_processed) as u64);
-            }
-            ticks_processed = periods_passed_until_now;
-        }
-
-        while inflight.len() < args.queue_depth.get() {
-            let start = Instant::now();
-            let req = {
-                let mut rng = rand::thread_rng();
-                let r = &ranges[weights.sample(&mut rng)];
-                let key: i128 = rng.gen_range(r.start..r.end);
-                let key = Key::from_i128(key);
-                assert!(key.is_rel_block_key());
-                let (rel_tag, block_no) = key
-                    .to_rel_block()
-                    .expect("we filter non-rel-block keys out above");
-                PagestreamGetPageRequest {
-                    hdr: PagestreamRequest {
-                        reqid: 0,
-                        request_lsn: if rng.gen_bool(args.req_latest_probability) {
-                            Lsn::MAX
-                        } else {
-                            r.timeline_lsn
-                        },
-                        not_modified_since: r.timeline_lsn,
-                    },
-                    rel: rel_tag,
-                    blkno: block_no,
-                }
-            };
-            client.getpage_send(req).await.unwrap();
-            inflight.push_back(start);
-        }
-
-        let start = inflight.pop_front().unwrap();
-        client.getpage_recv().await.unwrap();
-        let end = Instant::now();
-        shared_state.live_stats.request_done();
-        ticks_processed += 1;
-        STATS.with(|stats| {
-            stats
-                .borrow()
-                .lock()
-                .unwrap()
-                .observe(end.duration_since(start))
-                .unwrap();
-        });
-
-        if let Some(period) = &rps_period {
-            let next_at = client_start
-                + Duration::from_micros(
-                    (ticks_processed) as u64 * u64::try_from(period.as_micros()).unwrap(),
-                );
-            tokio::time::sleep_until(next_at.into()).await;
-        }
-    }
-}

pageserver/src/bin/pageserver.rs

@@ -416,18 +416,8 @@ fn start_pageserver(
     // The storage_broker::connect call needs to happen inside a tokio runtime thread.
     let broker_client = WALRECEIVER_RUNTIME
         .block_on(async {
-            let tls_config = storage_broker::ClientTlsConfig::new().ca_certificates(
-                conf.ssl_ca_certs
-                    .iter()
-                    .map(pem::encode)
-                    .map(storage_broker::Certificate::from_pem),
-            );
             // Note: we do not attempt connecting here (but validate endpoints sanity).
-            storage_broker::connect(
-                conf.broker_endpoint.clone(),
-                conf.broker_keepalive_interval,
-                tls_config,
-            )
+            storage_broker::connect(conf.broker_endpoint.clone(), conf.broker_keepalive_interval)
         })
         .with_context(|| {
             format!(

pageserver/src/config.rs

@@ -17,10 +17,9 @@ use once_cell::sync::OnceCell;
 use pageserver_api::config::{DiskUsageEvictionTaskConfig, MaxVectoredReadBytes};
 use pageserver_api::models::ImageCompressionAlgorithm;
 use pageserver_api::shard::TenantShardId;
-use pem::Pem;
 use postgres_backend::AuthType;
 use remote_storage::{RemotePath, RemoteStorageConfig};
-use reqwest::Url;
+use reqwest::{Certificate, Url};
 use storage_broker::Uri;
 use utils::id::{NodeId, TimelineId};
 use utils::logging::{LogFormat, SecretString};
@@ -68,8 +67,8 @@ pub struct PageServerConf {
     /// Period to reload certificate and private key from files.
     /// Default: 60s.
     pub ssl_cert_reload_period: Duration,
-    /// Trusted root CA certificates to use in https APIs in PEM format.
-    pub ssl_ca_certs: Vec<Pem>,
+    /// Trusted root CA certificates to use in https APIs.
+    pub ssl_ca_certs: Vec<Certificate>,
 
     /// Current availability zone. Used for traffic metrics.
     pub availability_zone: Option<String>,
@@ -119,13 +118,13 @@ pub struct PageServerConf {
     /// A lower value implicitly deprioritizes loading such tenants, vs. other work in the system.
     pub concurrent_tenant_warmup: ConfigurableSemaphore,
 
-    /// Number of concurrent [`TenantShard::gather_size_inputs`](crate::tenant::TenantShard::gather_size_inputs) allowed.
+    /// Number of concurrent [`Tenant::gather_size_inputs`](crate::tenant::Tenant::gather_size_inputs) allowed.
     pub concurrent_tenant_size_logical_size_queries: ConfigurableSemaphore,
-    /// Limit of concurrent [`TenantShard::gather_size_inputs`] issued by module `eviction_task`.
+    /// Limit of concurrent [`Tenant::gather_size_inputs`] issued by module `eviction_task`.
     /// The number of permits is the same as `concurrent_tenant_size_logical_size_queries`.
     /// See the comment in `eviction_task` for details.
     ///
-    /// [`TenantShard::gather_size_inputs`]: crate::tenant::TenantShard::gather_size_inputs
+    /// [`Tenant::gather_size_inputs`]: crate::tenant::Tenant::gather_size_inputs
     pub eviction_task_immitated_concurrent_logical_size_queries: ConfigurableSemaphore,
 
     // How often to collect metrics and send them to the metrics endpoint.
@@ -225,11 +224,6 @@ pub struct PageServerConf {
     /// Does not force TLS: the client negotiates TLS usage during the handshake.
     /// Uses key and certificate from ssl_key_file/ssl_cert_file.
     pub enable_tls_page_service_api: bool,
-
-    /// Run in development mode, which disables certain safety checks
-    /// such as authentication requirements for HTTP and PostgreSQL APIs.
-    /// This is insecure and should only be used in development environments.
-    pub dev_mode: bool,
 }
 
 /// Token for authentication to safekeepers
@@ -403,7 +397,6 @@ impl PageServerConf {
             generate_unarchival_heatmap,
             tracing,
             enable_tls_page_service_api,
-            dev_mode,
         } = config_toml;
 
         let mut conf = PageServerConf {
@@ -455,7 +448,6 @@ impl PageServerConf {
             get_vectored_concurrent_io,
             tracing,
             enable_tls_page_service_api,
-            dev_mode,
 
             // ------------------------------------------------------------
             // fields that require additional validation or custom handling
@@ -505,10 +497,7 @@ impl PageServerConf {
             ssl_ca_certs: match ssl_ca_file {
                 Some(ssl_ca_file) => {
                     let buf = std::fs::read(ssl_ca_file)?;
-                    pem::parse_many(&buf)?
-                        .into_iter()
-                        .filter(|pem| pem.tag() == "CERTIFICATE")
-                        .collect()
+                    Certificate::from_pem_bundle(&buf)?
                 }
                 None => Vec::new(),
             },
@@ -599,10 +588,10 @@ impl ConfigurableSemaphore {
     /// Initializse using a non-zero amount of permits.
     ///
     /// Require a non-zero initial permits, because using permits == 0 is a crude way to disable a
-    /// feature such as [`TenantShard::gather_size_inputs`]. Otherwise any semaphore using future will
+    /// feature such as [`Tenant::gather_size_inputs`]. Otherwise any semaphore using future will
     /// behave like [`futures::future::pending`], just waiting until new permits are added.
     ///
-    /// [`TenantShard::gather_size_inputs`]: crate::tenant::TenantShard::gather_size_inputs
+    /// [`Tenant::gather_size_inputs`]: crate::tenant::Tenant::gather_size_inputs
     pub fn new(initial_permits: NonZeroUsize) -> Self {
         ConfigurableSemaphore {
             initial_permits,

pageserver/src/consumption_metrics.rs

@@ -24,7 +24,7 @@ use crate::task_mgr::{self, BACKGROUND_RUNTIME, TaskKind};
 use crate::tenant::mgr::TenantManager;
 use crate::tenant::size::CalculateSyntheticSizeError;
 use crate::tenant::tasks::BackgroundLoopKind;
-use crate::tenant::{LogicalSizeCalculationCause, TenantShard};
+use crate::tenant::{LogicalSizeCalculationCause, Tenant};
 
 mod disk_cache;
 mod metrics;
@@ -428,7 +428,7 @@ async fn calculate_synthetic_size_worker(
     }
 }
 
-async fn calculate_and_log(tenant: &TenantShard, cancel: &CancellationToken, ctx: &RequestContext) {
+async fn calculate_and_log(tenant: &Tenant, cancel: &CancellationToken, ctx: &RequestContext) {
     const CAUSE: LogicalSizeCalculationCause =
         LogicalSizeCalculationCause::ConsumptionMetricsSyntheticSize;
 

pageserver/src/consumption_metrics/metrics.rs

@@ -175,9 +175,9 @@ impl MetricsKey {
         .absolute_values()
     }
 
-    /// [`TenantShard::remote_size`]
+    /// [`Tenant::remote_size`]
     ///
-    /// [`TenantShard::remote_size`]: crate::tenant::TenantShard::remote_size
+    /// [`Tenant::remote_size`]: crate::tenant::Tenant::remote_size
     const fn remote_storage_size(tenant_id: TenantId) -> AbsoluteValueFactory {
         MetricsKey {
             tenant_id,
@@ -199,9 +199,9 @@ impl MetricsKey {
         .absolute_values()
     }
 
-    /// [`TenantShard::cached_synthetic_size`] as refreshed by [`calculate_synthetic_size_worker`].
+    /// [`Tenant::cached_synthetic_size`] as refreshed by [`calculate_synthetic_size_worker`].
     ///
-    /// [`TenantShard::cached_synthetic_size`]: crate::tenant::TenantShard::cached_synthetic_size
+    /// [`Tenant::cached_synthetic_size`]: crate::tenant::Tenant::cached_synthetic_size
     /// [`calculate_synthetic_size_worker`]: super::calculate_synthetic_size_worker
     const fn synthetic_size(tenant_id: TenantId) -> AbsoluteValueFactory {
         MetricsKey {
@@ -254,7 +254,7 @@ pub(super) async fn collect_all_metrics(
 
 async fn collect<S>(tenants: S, cache: &Cache, ctx: &RequestContext) -> Vec<NewRawMetric>
 where
-    S: futures::stream::Stream<Item = (TenantId, Arc<crate::tenant::TenantShard>)>,
+    S: futures::stream::Stream<Item = (TenantId, Arc<crate::tenant::Tenant>)>,
 {
     let mut current_metrics: Vec<NewRawMetric> = Vec::new();
 
@@ -263,9 +263,7 @@ where
     while let Some((tenant_id, tenant)) = tenants.next().await {
         let mut tenant_resident_size = 0;
 
-        let timelines = tenant.list_timelines();
-        let timelines_len = timelines.len();
-        for timeline in timelines {
+        for timeline in tenant.list_timelines() {
             let timeline_id = timeline.timeline_id;
 
             match TimelineSnapshot::collect(&timeline, ctx) {
@@ -291,11 +289,6 @@ where
             tenant_resident_size += timeline.resident_physical_size();
         }
 
-        if timelines_len == 0 {
-            // Force set it to 1 byte to avoid not being reported -- all timelines are offloaded.
-            tenant_resident_size = 1;
-        }
-
         let snap = TenantSnapshot::collect(&tenant, tenant_resident_size);
         snap.to_metrics(tenant_id, Utc::now(), cache, &mut current_metrics);
     }
@@ -315,7 +308,7 @@ impl TenantSnapshot {
     ///
     /// `resident_size` is calculated of the timelines we had access to for other metrics, so we
     /// cannot just list timelines here.
-    fn collect(t: &Arc<crate::tenant::TenantShard>, resident_size: u64) -> Self {
+    fn collect(t: &Arc<crate::tenant::Tenant>, resident_size: u64) -> Self {
         TenantSnapshot {
             resident_size,
             remote_size: t.remote_size(),

pageserver/src/controller_upcall_client.rs

@@ -3,19 +3,17 @@ use std::collections::HashMap;
 use futures::Future;
 use pageserver_api::config::NodeMetadata;
 use pageserver_api::controller_api::{AvailabilityZone, NodeRegisterRequest};
-use pageserver_api::models::ShardImportStatus;
 use pageserver_api::shard::TenantShardId;
 use pageserver_api::upcall_api::{
-    PutTimelineImportStatusRequest, ReAttachRequest, ReAttachResponse, ReAttachResponseTenant,
-    ValidateRequest, ValidateRequestTenant, ValidateResponse,
+    ReAttachRequest, ReAttachResponse, ReAttachResponseTenant, ValidateRequest,
+    ValidateRequestTenant, ValidateResponse,
 };
-use reqwest::Certificate;
 use serde::Serialize;
 use serde::de::DeserializeOwned;
 use tokio_util::sync::CancellationToken;
 use url::Url;
 use utils::generation::Generation;
-use utils::id::{NodeId, TimelineId};
+use utils::id::NodeId;
 use utils::{backoff, failpoint_support};
 
 use crate::config::PageServerConf;
@@ -47,12 +45,6 @@ pub trait StorageControllerUpcallApi {
         &self,
         tenants: Vec<(TenantShardId, Generation)>,
     ) -> impl Future<Output = Result<HashMap<TenantShardId, bool>, RetryForeverError>> + Send;
-    fn put_timeline_import_status(
-        &self,
-        tenant_shard_id: TenantShardId,
-        timeline_id: TimelineId,
-        status: ShardImportStatus,
-    ) -> impl Future<Output = Result<(), RetryForeverError>> + Send;
 }
 
 impl StorageControllerUpcallClient {
@@ -84,8 +76,8 @@ impl StorageControllerUpcallClient {
             client = client.default_headers(headers);
         }
 
-        for cert in &conf.ssl_ca_certs {
-            client = client.add_root_certificate(Certificate::from_der(cert.contents())?);
+        for ssl_ca_cert in &conf.ssl_ca_certs {
+            client = client.add_root_certificate(ssl_ca_cert.clone());
         }
 
         Ok(Some(Self {
@@ -280,30 +272,4 @@ impl StorageControllerUpcallApi for StorageControllerUpcallClient {
 
         Ok(result.into_iter().collect())
     }
-
-    /// Send a shard import status to the storage controller
-    ///
-    /// The implementation must have at-least-once delivery semantics.
-    /// To this end, we retry the request until it succeeds. If the pageserver
-    /// restarts or crashes, the shard import will start again from the beggining.
-    #[tracing::instrument(skip_all)] // so that warning logs from retry_http_forever have context
-    async fn put_timeline_import_status(
-        &self,
-        tenant_shard_id: TenantShardId,
-        timeline_id: TimelineId,
-        status: ShardImportStatus,
-    ) -> Result<(), RetryForeverError> {
-        let url = self
-            .base_url
-            .join("timeline_import_status")
-            .expect("Failed to build path");
-
-        let request = PutTimelineImportStatusRequest {
-            tenant_shard_id,
-            timeline_id,
-            status,
-        };
-
-        self.retry_http_forever(&url, request).await
-    }
 }

pageserver/src/deletion_queue.rs

@@ -787,15 +787,6 @@ mod test {
 
             Ok(result)
         }
-
-        async fn put_timeline_import_status(
-            &self,
-            _tenant_shard_id: TenantShardId,
-            _timeline_id: TimelineId,
-            _status: pageserver_api::models::ShardImportStatus,
-        ) -> Result<(), RetryForeverError> {
-            unimplemented!()
-        }
     }
 
     async fn setup(test_name: &str) -> anyhow::Result<TestSetup> {

pageserver/src/http/routes.rs

@@ -1873,7 +1873,7 @@ async fn update_tenant_config_handler(
         &ShardParameters::default(),
     );
 
-    crate::tenant::TenantShard::persist_tenant_config(state.conf, &tenant_shard_id, &location_conf)
+    crate::tenant::Tenant::persist_tenant_config(state.conf, &tenant_shard_id, &location_conf)
         .await
         .map_err(|e| ApiError::InternalServerError(anyhow::anyhow!(e)))?;
 
@@ -1917,7 +1917,7 @@ async fn patch_tenant_config_handler(
         &ShardParameters::default(),
     );
 
-    crate::tenant::TenantShard::persist_tenant_config(state.conf, &tenant_shard_id, &location_conf)
+    crate::tenant::Tenant::persist_tenant_config(state.conf, &tenant_shard_id, &location_conf)
         .await
         .map_err(|e| ApiError::InternalServerError(anyhow::anyhow!(e)))?;
 

pageserver/src/lib.rs

@@ -49,7 +49,7 @@ use tracing::{info, info_span};
 /// backwards-compatible changes to the metadata format.
 pub const STORAGE_FORMAT_VERSION: u16 = 3;
 
-pub const DEFAULT_PG_VERSION: u32 = 17;
+pub const DEFAULT_PG_VERSION: u32 = 16;
 
 // Magic constants used to identify different kinds of files
 pub const IMAGE_FILE_MAGIC: u16 = 0x5A60;

pageserver/src/metrics.rs

@@ -1086,7 +1086,7 @@ pub(crate) static TIMELINE_EPHEMERAL_BYTES: Lazy<UIntGauge> = Lazy::new(|| {
     .expect("Failed to register metric")
 });
 
-/// Metrics related to the lifecycle of a [`crate::tenant::TenantShard`] object: things
+/// Metrics related to the lifecycle of a [`crate::tenant::Tenant`] object: things
 /// like how long it took to load.
 ///
 /// Note that these are process-global metrics, _not_ per-tenant metrics.  Per-tenant
@@ -1289,7 +1289,6 @@ pub(crate) enum StorageIoOperation {
     Seek,
     Fsync,
     Metadata,
-    SetLen,
 }
 
 impl StorageIoOperation {
@@ -1304,7 +1303,6 @@ impl StorageIoOperation {
             StorageIoOperation::Seek => "seek",
             StorageIoOperation::Fsync => "fsync",
             StorageIoOperation::Metadata => "metadata",
-            StorageIoOperation::SetLen => "set_len",
         }
     }
 }

pageserver/src/page_service.rs

@@ -15,7 +15,6 @@ use async_compression::tokio::write::GzipEncoder;
 use bytes::Buf;
 use futures::FutureExt;
 use itertools::Itertools;
-use jsonwebtoken::TokenData;
 use once_cell::sync::OnceCell;
 use pageserver_api::config::{
     PageServicePipeliningConfig, PageServicePipeliningConfigPipelined,
@@ -76,7 +75,7 @@ use crate::tenant::timeline::{self, WaitLsnError};
 use crate::tenant::{GetTimelineError, PageReconstructError, Timeline};
 use crate::{basebackup, timed_after_cancellation};
 
-/// How long we may wait for a [`crate::tenant::mgr::TenantSlot::InProgress`]` and/or a [`crate::tenant::TenantShard`] which
+/// How long we may wait for a [`crate::tenant::mgr::TenantSlot::InProgress`]` and/or a [`crate::tenant::Tenant`] which
 /// is not yet in state [`TenantState::Active`].
 ///
 /// NB: this is a different value than [`crate::http::routes::ACTIVE_TENANT_TIMEOUT`].
@@ -2838,7 +2837,7 @@ where
     ) -> Result<(), QueryError> {
         // this unwrap is never triggered, because check_auth_jwt only called when auth_type is NeonJWT
         // which requires auth to be present
-        let data: TokenData<Claims> = self
+        let data = self
             .auth
             .as_ref()
             .unwrap()

pageserver/src/tenant.rs

@@ -158,7 +158,7 @@ pub struct TenantSharedResources {
     pub l0_flush_global_state: L0FlushGlobalState,
 }
 
-/// A [`TenantShard`] is really an _attached_ tenant.  The configuration
+/// A [`Tenant`] is really an _attached_ tenant.  The configuration
 /// for an attached tenant is a subset of the [`LocationConf`], represented
 /// in this struct.
 #[derive(Clone)]
@@ -245,7 +245,7 @@ pub(crate) enum SpawnMode {
 ///
 /// Tenant consists of multiple timelines. Keep them in a hash table.
 ///
-pub struct TenantShard {
+pub struct Tenant {
     // Global pageserver config parameters
     pub conf: &'static PageServerConf,
 
@@ -267,7 +267,7 @@ pub struct TenantShard {
     shard_identity: ShardIdentity,
 
     /// The remote storage generation, used to protect S3 objects from split-brain.
-    /// Does not change over the lifetime of the [`TenantShard`] object.
+    /// Does not change over the lifetime of the [`Tenant`] object.
     ///
     /// This duplicates the generation stored in LocationConf, but that structure is mutable:
     /// this copy enforces the invariant that generatio doesn't change during a Tenant's lifetime.
@@ -309,7 +309,7 @@ pub struct TenantShard {
     // Access to global deletion queue for when this tenant wants to schedule a deletion
     deletion_queue_client: DeletionQueueClient,
 
-    /// Cached logical sizes updated updated on each [`TenantShard::gather_size_inputs`].
+    /// Cached logical sizes updated updated on each [`Tenant::gather_size_inputs`].
     cached_logical_sizes: tokio::sync::Mutex<HashMap<(TimelineId, Lsn), u64>>,
     cached_synthetic_tenant_size: Arc<AtomicU64>,
 
@@ -337,12 +337,12 @@ pub struct TenantShard {
     // Timelines' cancellation token.
     pub(crate) cancel: CancellationToken,
 
-    // Users of the TenantShard such as the page service must take this Gate to avoid
-    // trying to use a TenantShard which is shutting down.
+    // Users of the Tenant such as the page service must take this Gate to avoid
+    // trying to use a Tenant which is shutting down.
     pub(crate) gate: Gate,
 
     /// Throttle applied at the top of [`Timeline::get`].
-    /// All [`TenantShard::timelines`] of a given [`TenantShard`] instance share the same [`throttle::Throttle`] instance.
+    /// All [`Tenant::timelines`] of a given [`Tenant`] instance share the same [`throttle::Throttle`] instance.
     pub(crate) pagestream_throttle: Arc<throttle::Throttle>,
 
     pub(crate) pagestream_throttle_metrics: Arc<crate::metrics::tenant_throttling::Pagestream>,
@@ -362,7 +362,7 @@ pub struct TenantShard {
 
     l0_flush_global_state: L0FlushGlobalState,
 }
-impl std::fmt::Debug for TenantShard {
+impl std::fmt::Debug for Tenant {
     fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
         write!(f, "{} ({})", self.tenant_shard_id, self.current_state())
     }
@@ -841,7 +841,7 @@ impl Debug for SetStoppingError {
     }
 }
 
-/// Arguments to [`TenantShard::create_timeline`].
+/// Arguments to [`Tenant::create_timeline`].
 ///
 /// Not usable as an idempotency key for timeline creation because if [`CreateTimelineParamsBranch::ancestor_start_lsn`]
 /// is `None`, the result of the timeline create call is not deterministic.
@@ -876,7 +876,7 @@ pub(crate) struct CreateTimelineParamsImportPgdata {
     pub(crate) idempotency_key: import_pgdata::index_part_format::IdempotencyKey,
 }
 
-/// What is used to determine idempotency of a [`TenantShard::create_timeline`] call in  [`TenantShard::start_creating_timeline`] in  [`TenantShard::start_creating_timeline`].
+/// What is used to determine idempotency of a [`Tenant::create_timeline`] call in  [`Tenant::start_creating_timeline`] in  [`Tenant::start_creating_timeline`].
 ///
 /// Each [`Timeline`] object holds [`Self`] as an immutable property in [`Timeline::create_idempotency`].
 ///
@@ -914,7 +914,7 @@ pub(crate) struct CreatingTimelineIdempotencyImportPgdata {
     idempotency_key: import_pgdata::index_part_format::IdempotencyKey,
 }
 
-/// What is returned by [`TenantShard::start_creating_timeline`].
+/// What is returned by [`Tenant::start_creating_timeline`].
 #[must_use]
 enum StartCreatingTimelineResult {
     CreateGuard(TimelineCreateGuard),
@@ -943,13 +943,13 @@ struct TimelineInitAndSyncNeedsSpawnImportPgdata {
     guard: TimelineCreateGuard,
 }
 
-/// What is returned by [`TenantShard::create_timeline`].
+/// What is returned by [`Tenant::create_timeline`].
 enum CreateTimelineResult {
     Created(Arc<Timeline>),
     Idempotent(Arc<Timeline>),
-    /// IMPORTANT: This [`Arc<Timeline>`] object is not in [`TenantShard::timelines`] when
+    /// IMPORTANT: This [`Arc<Timeline>`] object is not in [`Tenant::timelines`] when
     /// we return this result, nor will this concrete object ever be added there.
-    /// Cf method comment on [`TenantShard::create_timeline_import_pgdata`].
+    /// Cf method comment on [`Tenant::create_timeline_import_pgdata`].
     ImportSpawned(Arc<Timeline>),
 }
 
@@ -1082,7 +1082,7 @@ pub(crate) enum LoadConfigError {
     NotFound(Utf8PathBuf),
 }
 
-impl TenantShard {
+impl Tenant {
     /// Yet another helper for timeline initialization.
     ///
     /// - Initializes the Timeline struct and inserts it into the tenant's hash map
@@ -1303,7 +1303,7 @@ impl TenantShard {
         init_order: Option<InitializationOrder>,
         mode: SpawnMode,
         ctx: &RequestContext,
-    ) -> Result<Arc<TenantShard>, GlobalShutDown> {
+    ) -> Result<Arc<Tenant>, GlobalShutDown> {
         let wal_redo_manager =
             WalRedoManager::new(PostgresRedoManager::new(conf, tenant_shard_id))?;
 
@@ -1317,7 +1317,7 @@ impl TenantShard {
         let attach_mode = attached_conf.location.attach_mode;
         let generation = attached_conf.location.generation;
 
-        let tenant = Arc::new(TenantShard::new(
+        let tenant = Arc::new(Tenant::new(
             TenantState::Attaching,
             conf,
             attached_conf,
@@ -1334,7 +1334,7 @@ impl TenantShard {
         let attach_gate_guard = tenant
             .gate
             .enter()
-            .expect("We just created the TenantShard: nothing else can have shut it down yet");
+            .expect("We just created the Tenant: nothing else can have shut it down yet");
 
         // Do all the hard work in the background
         let tenant_clone = Arc::clone(&tenant);
@@ -1362,7 +1362,7 @@ impl TenantShard {
                     }
                 }
 
-                fn make_broken_or_stopping(t: &TenantShard, err: anyhow::Error) {
+                fn make_broken_or_stopping(t: &Tenant, err: anyhow::Error) {
                     t.state.send_modify(|state| match state {
                         // TODO: the old code alluded to DeleteTenantFlow sometimes setting
                         // TenantState::Stopping before we get here, but this may be outdated.
@@ -1627,7 +1627,7 @@ impl TenantShard {
     /// No background tasks are started as part of this routine.
     ///
     async fn attach(
-        self: &Arc<TenantShard>,
+        self: &Arc<Tenant>,
         preload: Option<TenantPreload>,
         ctx: &RequestContext,
     ) -> anyhow::Result<()> {
@@ -1957,7 +1957,7 @@ impl TenantShard {
     }
 
     async fn load_timelines_metadata(
-        self: &Arc<TenantShard>,
+        self: &Arc<Tenant>,
         timeline_ids: HashSet<TimelineId>,
         remote_storage: &GenericRemoteStorage,
         heatmap: Option<(HeatMapTenant, std::time::Instant)>,
@@ -2028,7 +2028,7 @@ impl TenantShard {
     }
 
     fn load_timeline_metadata(
-        self: &Arc<TenantShard>,
+        self: &Arc<Tenant>,
         timeline_id: TimelineId,
         remote_storage: GenericRemoteStorage,
         previous_heatmap: Option<PreviousHeatmap>,
@@ -2429,14 +2429,14 @@ impl TenantShard {
     /// This is used by tests & import-from-basebackup.
     ///
     /// The returned [`UninitializedTimeline`] contains no data nor metadata and it is in
-    /// a state that will fail [`TenantShard::load_remote_timeline`] because `disk_consistent_lsn=Lsn(0)`.
+    /// a state that will fail [`Tenant::load_remote_timeline`] because `disk_consistent_lsn=Lsn(0)`.
     ///
     /// The caller is responsible for getting the timeline into a state that will be accepted
-    /// by [`TenantShard::load_remote_timeline`] / [`TenantShard::attach`].
+    /// by [`Tenant::load_remote_timeline`] / [`Tenant::attach`].
     /// Then they may call [`UninitializedTimeline::finish_creation`] to add the timeline
-    /// to the [`TenantShard::timelines`].
+    /// to the [`Tenant::timelines`].
     ///
-    /// Tests should use `TenantShard::create_test_timeline` to set up the minimum required metadata keys.
+    /// Tests should use `Tenant::create_test_timeline` to set up the minimum required metadata keys.
     pub(crate) async fn create_empty_timeline(
         self: &Arc<Self>,
         new_timeline_id: TimelineId,
@@ -2584,7 +2584,7 @@ impl TenantShard {
     /// the same timeline ID already exists, returns CreateTimelineError::AlreadyExists.
     #[allow(clippy::too_many_arguments)]
     pub(crate) async fn create_timeline(
-        self: &Arc<TenantShard>,
+        self: &Arc<Tenant>,
         params: CreateTimelineParams,
         broker_client: storage_broker::BrokerClientChannel,
         ctx: &RequestContext,
@@ -2751,13 +2751,13 @@ impl TenantShard {
         Ok(activated_timeline)
     }
 
-    /// The returned [`Arc<Timeline>`] is NOT in the [`TenantShard::timelines`] map until the import
+    /// The returned [`Arc<Timeline>`] is NOT in the [`Tenant::timelines`] map until the import
     /// completes in the background. A DIFFERENT [`Arc<Timeline>`] will be inserted into the
-    /// [`TenantShard::timelines`] map when the import completes.
+    /// [`Tenant::timelines`] map when the import completes.
     /// We only return an [`Arc<Timeline>`] here so the API handler can create a [`pageserver_api::models::TimelineInfo`]
     /// for the response.
     async fn create_timeline_import_pgdata(
-        self: &Arc<Self>,
+        self: &Arc<Tenant>,
         params: CreateTimelineParamsImportPgdata,
         activate: ActivateTimelineArgs,
         ctx: &RequestContext,
@@ -2854,7 +2854,7 @@ impl TenantShard {
 
     #[instrument(skip_all, fields(tenant_id=%self.tenant_shard_id.tenant_id, shard_id=%self.tenant_shard_id.shard_slug(), timeline_id=%timeline.timeline_id))]
     async fn create_timeline_import_pgdata_task(
-        self: Arc<TenantShard>,
+        self: Arc<Tenant>,
         timeline: Arc<Timeline>,
         index_part: import_pgdata::index_part_format::Root,
         activate: ActivateTimelineArgs,
@@ -2882,7 +2882,7 @@ impl TenantShard {
     }
 
     async fn create_timeline_import_pgdata_task_impl(
-        self: Arc<TenantShard>,
+        self: Arc<Tenant>,
         timeline: Arc<Timeline>,
         index_part: import_pgdata::index_part_format::Root,
         activate: ActivateTimelineArgs,
@@ -2899,10 +2899,10 @@ impl TenantShard {
         // Reload timeline from remote.
         // This proves that the remote state is attachable, and it reuses the code.
         //
-        // TODO: think about whether this is safe to do with concurrent TenantShard::shutdown.
+        // TODO: think about whether this is safe to do with concurrent Tenant::shutdown.
         // timeline_create_guard hols the tenant gate open, so, shutdown cannot _complete_ until we exit.
-        // But our activate() call might launch new background tasks after TenantShard::shutdown
-        // already went past shutting down the TenantShard::timelines, which this timeline here is no part of.
+        // But our activate() call might launch new background tasks after Tenant::shutdown
+        // already went past shutting down the Tenant::timelines, which this timeline here is no part of.
         // I think the same problem exists with the bootstrap & branch mgmt API tasks (tenant shutting
         // down while bootstrapping/branching + activating), but, the race condition is much more likely
         // to manifest because of the long runtime of this import task.
@@ -2917,7 +2917,7 @@ impl TenantShard {
         // };
         let timeline_id = timeline.timeline_id;
 
-        // load from object storage like TenantShard::attach does
+        // load from object storage like Tenant::attach does
         let resources = self.build_timeline_resources(timeline_id);
         let index_part = resources
             .remote_client
@@ -3938,7 +3938,7 @@ enum ActivateTimelineArgs {
     No,
 }
 
-impl TenantShard {
+impl Tenant {
     pub fn tenant_specific_overrides(&self) -> pageserver_api::models::TenantConfig {
         self.tenant_conf.load().tenant_conf.clone()
     }
@@ -4096,7 +4096,7 @@ impl TenantShard {
         update: F,
     ) -> anyhow::Result<pageserver_api::models::TenantConfig> {
         // Use read-copy-update in order to avoid overwriting the location config
-        // state if this races with [`TenantShard::set_new_location_config`]. Note that
+        // state if this races with [`Tenant::set_new_location_config`]. Note that
         // this race is not possible if both request types come from the storage
         // controller (as they should!) because an exclusive op lock is required
         // on the storage controller side.
@@ -4219,7 +4219,7 @@ impl TenantShard {
         Ok((timeline, timeline_ctx))
     }
 
-    /// [`TenantShard::shutdown`] must be called before dropping the returned [`TenantShard`] object
+    /// [`Tenant::shutdown`] must be called before dropping the returned [`Tenant`] object
     /// to ensure proper cleanup of background tasks and metrics.
     //
     // Allow too_many_arguments because a constructor's argument list naturally grows with the
@@ -4235,7 +4235,7 @@ impl TenantShard {
         remote_storage: GenericRemoteStorage,
         deletion_queue_client: DeletionQueueClient,
         l0_flush_global_state: L0FlushGlobalState,
-    ) -> TenantShard {
+    ) -> Tenant {
         debug_assert!(
             !attached_conf.location.generation.is_none() || conf.control_plane_api.is_none()
         );
@@ -4295,7 +4295,7 @@ impl TenantShard {
             }
         });
 
-        TenantShard {
+        Tenant {
             tenant_shard_id,
             shard_identity,
             generation: attached_conf.location.generation,
@@ -4330,7 +4330,7 @@ impl TenantShard {
             cancel: CancellationToken::default(),
             gate: Gate::default(),
             pagestream_throttle: Arc::new(throttle::Throttle::new(
-                TenantShard::get_pagestream_throttle_config(conf, &attached_conf.tenant_conf),
+                Tenant::get_pagestream_throttle_config(conf, &attached_conf.tenant_conf),
             )),
             pagestream_throttle_metrics: Arc::new(
                 crate::metrics::tenant_throttling::Pagestream::new(&tenant_shard_id),
@@ -4466,11 +4466,11 @@ impl TenantShard {
 
         // Perform GC for each timeline.
         //
-        // Note that we don't hold the `TenantShard::gc_cs` lock here because we don't want to delay the
+        // Note that we don't hold the `Tenant::gc_cs` lock here because we don't want to delay the
         // branch creation task, which requires the GC lock. A GC iteration can run concurrently
         // with branch creation.
         //
-        // See comments in [`TenantShard::branch_timeline`] for more information about why branch
+        // See comments in [`Tenant::branch_timeline`] for more information about why branch
         // creation task can run concurrently with timeline's GC iteration.
         for timeline in gc_timelines {
             if cancel.is_cancelled() {
@@ -4500,7 +4500,7 @@ impl TenantShard {
 
     /// Refreshes the Timeline::gc_info for all timelines, returning the
     /// vector of timelines which have [`Timeline::get_last_record_lsn`] past
-    /// [`TenantShard::get_gc_horizon`].
+    /// [`Tenant::get_gc_horizon`].
     ///
     /// This is usually executed as part of periodic gc, but can now be triggered more often.
     pub(crate) async fn refresh_gc_info(
@@ -5499,7 +5499,7 @@ impl TenantShard {
             }
         }
 
-        // The flushes we did above were just writes, but the TenantShard might have had
+        // The flushes we did above were just writes, but the Tenant might have had
         // pending deletions as well from recent compaction/gc: we want to flush those
         // as well.  This requires flushing the global delete queue.  This is cheap
         // because it's typically a no-op.
@@ -5517,7 +5517,7 @@ impl TenantShard {
 
     /// How much local storage would this tenant like to have?  It can cope with
     /// less than this (via eviction and on-demand downloads), but this function enables
-    /// the TenantShard to advertise how much storage it would prefer to have to provide fast I/O
+    /// the Tenant to advertise how much storage it would prefer to have to provide fast I/O
     /// by keeping important things on local disk.
     ///
     /// This is a heuristic, not a guarantee: tenants that are long-idle will actually use less
@@ -5540,11 +5540,11 @@ impl TenantShard {
     /// manifest in `Self::remote_tenant_manifest`.
     ///
     /// TODO: instead of requiring callers to remember to call `maybe_upload_tenant_manifest` after
-    /// changing any `TenantShard` state that's included in the manifest, consider making the manifest
+    /// changing any `Tenant` state that's included in the manifest, consider making the manifest
     /// the authoritative source of data with an API that automatically uploads on changes. Revisit
     /// this when the manifest is more widely used and we have a better idea of the data model.
     pub(crate) async fn maybe_upload_tenant_manifest(&self) -> Result<(), TenantManifestError> {
-        // Multiple tasks may call this function concurrently after mutating the TenantShard runtime
+        // Multiple tasks may call this function concurrently after mutating the Tenant runtime
         // state, affecting the manifest generated by `build_tenant_manifest`. We use an async mutex
         // to serialize these callers. `eq_ignoring_version` acts as a slightly inefficient but
         // simple coalescing mechanism.
@@ -5812,7 +5812,7 @@ pub(crate) mod harness {
             info_span!("TenantHarness", tenant_id=%self.tenant_shard_id.tenant_id, shard_id=%self.tenant_shard_id.shard_slug())
         }
 
-        pub(crate) async fn load(&self) -> (Arc<TenantShard>, RequestContext) {
+        pub(crate) async fn load(&self) -> (Arc<Tenant>, RequestContext) {
             let ctx = RequestContext::new(TaskKind::UnitTest, DownloadBehavior::Error)
                 .with_scope_unit_test();
             (
@@ -5827,10 +5827,10 @@ pub(crate) mod harness {
         pub(crate) async fn do_try_load(
             &self,
             ctx: &RequestContext,
-        ) -> anyhow::Result<Arc<TenantShard>> {
+        ) -> anyhow::Result<Arc<Tenant>> {
             let walredo_mgr = Arc::new(WalRedoManager::from(TestRedoManager));
 
-            let tenant = Arc::new(TenantShard::new(
+            let tenant = Arc::new(Tenant::new(
                 TenantState::Attaching,
                 self.conf,
                 AttachedTenantConf::try_from(LocationConf::attached_single(
@@ -6046,7 +6046,7 @@ mod tests {
     #[cfg(feature = "testing")]
     #[allow(clippy::too_many_arguments)]
     async fn randomize_timeline(
-        tenant: &Arc<TenantShard>,
+        tenant: &Arc<Tenant>,
         new_timeline_id: TimelineId,
         pg_version: u32,
         spec: TestTimelineSpecification,
@@ -6936,7 +6936,7 @@ mod tests {
     }
 
     async fn bulk_insert_compact_gc(
-        tenant: &TenantShard,
+        tenant: &Tenant,
         timeline: &Arc<Timeline>,
         ctx: &RequestContext,
         lsn: Lsn,
@@ -6948,7 +6948,7 @@ mod tests {
     }
 
     async fn bulk_insert_maybe_compact_gc(
-        tenant: &TenantShard,
+        tenant: &Tenant,
         timeline: &Arc<Timeline>,
         ctx: &RequestContext,
         mut lsn: Lsn,
@@ -7858,7 +7858,7 @@ mod tests {
             let (tline, _ctx) = tenant
                 .create_empty_timeline(TIMELINE_ID, Lsn(0), DEFAULT_PG_VERSION, &ctx)
                 .await?;
-            // Leave the timeline ID in [`TenantShard::timelines_creating`] to exclude attempting to create it again
+            // Leave the timeline ID in [`Tenant::timelines_creating`] to exclude attempting to create it again
             let raw_tline = tline.raw_timeline().unwrap();
             raw_tline
                 .shutdown(super::timeline::ShutdownMode::Hard)

pageserver/src/tenant/blob_io.rs

@@ -15,23 +15,21 @@
 //! len >= 128: 1CCCXXXX XXXXXXXX XXXXXXXX XXXXXXXX
 //!
 use std::cmp::min;
+use std::io::Error;
 
-use anyhow::Context;
 use async_compression::Level;
 use bytes::{BufMut, BytesMut};
 use pageserver_api::models::ImageCompressionAlgorithm;
 use tokio::io::AsyncWriteExt;
-use tokio_epoll_uring::IoBuf;
+use tokio_epoll_uring::{BoundedBuf, IoBuf, Slice};
 use tokio_util::sync::CancellationToken;
 use tracing::warn;
 
 use crate::context::RequestContext;
 use crate::page_cache::PAGE_SZ;
 use crate::tenant::block_io::BlockCursor;
-use crate::virtual_file::IoBufferMut;
+use crate::virtual_file::VirtualFile;
 use crate::virtual_file::owned_buffers_io::io_buf_ext::{FullSlice, IoBufExt};
-use crate::virtual_file::owned_buffers_io::write::{BufferedWriter, FlushTaskError};
-use crate::virtual_file::owned_buffers_io::write::{BufferedWriterShutdownMode, OwnedAsyncWriter};
 
 #[derive(Copy, Clone, Debug)]
 pub struct CompressionInfo {
@@ -39,67 +37,6 @@ pub struct CompressionInfo {
     pub compressed_size: Option<usize>,
 }
 
-/// A blob header, with header+data length and compression info.
-///
-/// TODO: use this more widely, and add an encode() method too.
-/// TODO: document the header format.
-#[derive(Clone, Copy, Default)]
-pub struct Header {
-    pub header_len: usize,
-    pub data_len: usize,
-    pub compression_bits: u8,
-}
-
-impl Header {
-    /// Decodes a header from a byte slice.
-    pub fn decode(bytes: &[u8]) -> anyhow::Result<Self> {
-        let Some(&first_header_byte) = bytes.first() else {
-            anyhow::bail!("zero-length blob header");
-        };
-
-        // If the first bit is 0, this is just a 1-byte length prefix up to 128 bytes.
-        if first_header_byte < 0x80 {
-            return Ok(Self {
-                header_len: 1, // by definition
-                data_len: first_header_byte as usize,
-                compression_bits: BYTE_UNCOMPRESSED,
-            });
-        }
-
-        // Otherwise, this is a 4-byte header containing compression information and length.
-        const HEADER_LEN: usize = 4;
-        let mut header_buf: [u8; HEADER_LEN] = bytes[0..HEADER_LEN]
-            .try_into()
-            .map_err(|_| anyhow::anyhow!("blob header too short: {bytes:?}"))?;
-
-        // TODO: verify the compression bits and convert to an enum.
-        let compression_bits = header_buf[0] & LEN_COMPRESSION_BIT_MASK;
-        header_buf[0] &= !LEN_COMPRESSION_BIT_MASK;
-        let data_len = u32::from_be_bytes(header_buf) as usize;
-
-        Ok(Self {
-            header_len: HEADER_LEN,
-            data_len,
-            compression_bits,
-        })
-    }
-
-    /// Returns the total header+data length.
-    pub fn total_len(&self) -> usize {
-        self.header_len + self.data_len
-    }
-}
-
-#[derive(Debug, thiserror::Error)]
-pub enum WriteBlobError {
-    #[error(transparent)]
-    Flush(FlushTaskError),
-    #[error("blob too large ({len} bytes)")]
-    BlobTooLarge { len: usize },
-    #[error(transparent)]
-    WriteBlobRaw(anyhow::Error),
-}
-
 impl BlockCursor<'_> {
     /// Read a blob into a new buffer.
     pub async fn read_blob(
@@ -219,64 +156,143 @@ pub(super) const BYTE_UNCOMPRESSED: u8 = 0x80;
 pub(super) const BYTE_ZSTD: u8 = BYTE_UNCOMPRESSED | 0x10;
 
 /// A wrapper of `VirtualFile` that allows users to write blobs.
-pub struct BlobWriter<W> {
+///
+/// If a `BlobWriter` is dropped, the internal buffer will be
+/// discarded. You need to call [`flush_buffer`](Self::flush_buffer)
+/// manually before dropping.
+pub struct BlobWriter<const BUFFERED: bool> {
+    inner: VirtualFile,
+    offset: u64,
+    /// A buffer to save on write calls, only used if BUFFERED=true
+    buf: Vec<u8>,
     /// We do tiny writes for the length headers; they need to be in an owned buffer;
     io_buf: Option<BytesMut>,
-    writer: BufferedWriter<IoBufferMut, W>,
-    offset: u64,
 }
 
-impl<W> BlobWriter<W>
-where
-    W: OwnedAsyncWriter + std::fmt::Debug + Send + Sync + 'static,
-{
-    /// See [`BufferedWriter`] struct-level doc comment for semantics of `start_offset`.
+impl<const BUFFERED: bool> BlobWriter<BUFFERED> {
     pub fn new(
-        file: W,
+        inner: VirtualFile,
         start_offset: u64,
-        gate: &utils::sync::gate::Gate,
-        cancel: CancellationToken,
-        ctx: &RequestContext,
-        flush_task_span: tracing::Span,
-    ) -> anyhow::Result<Self> {
-        Ok(Self {
-            io_buf: Some(BytesMut::new()),
-            writer: BufferedWriter::new(
-                file,
-                start_offset,
-                || IoBufferMut::with_capacity(Self::CAPACITY),
-                gate.enter()?,
-                cancel,
-                ctx,
-                flush_task_span,
-            ),
+        _gate: &utils::sync::gate::Gate,
+        _cancel: CancellationToken,
+        _ctx: &RequestContext,
+    ) -> Self {
+        Self {
+            inner,
             offset: start_offset,
-        })
+            buf: Vec::with_capacity(Self::CAPACITY),
+            io_buf: Some(BytesMut::new()),
+        }
     }
 
     pub fn size(&self) -> u64 {
         self.offset
     }
 
-    const CAPACITY: usize = 64 * 1024;
+    const CAPACITY: usize = if BUFFERED { 64 * 1024 } else { 0 };
 
-    /// Writes `src_buf` to the file at the current offset.
+    /// Writes the given buffer directly to the underlying `VirtualFile`.
+    /// You need to make sure that the internal buffer is empty, otherwise
+    /// data will be written in wrong order.
+    #[inline(always)]
+    async fn write_all_unbuffered<Buf: IoBuf + Send>(
+        &mut self,
+        src_buf: FullSlice<Buf>,
+        ctx: &RequestContext,
+    ) -> (FullSlice<Buf>, Result<(), Error>) {
+        let (src_buf, res) = self.inner.write_all(src_buf, ctx).await;
+        let nbytes = match res {
+            Ok(nbytes) => nbytes,
+            Err(e) => return (src_buf, Err(e)),
+        };
+        self.offset += nbytes as u64;
+        (src_buf, Ok(()))
+    }
+
+    #[inline(always)]
+    /// Flushes the internal buffer to the underlying `VirtualFile`.
+    pub async fn flush_buffer(&mut self, ctx: &RequestContext) -> Result<(), Error> {
+        let buf = std::mem::take(&mut self.buf);
+        let (slice, res) = self.inner.write_all(buf.slice_len(), ctx).await;
+        res?;
+        let mut buf = slice.into_raw_slice().into_inner();
+        buf.clear();
+        self.buf = buf;
+        Ok(())
+    }
+
+    #[inline(always)]
+    /// Writes as much of `src_buf` into the internal buffer as it fits
+    fn write_into_buffer(&mut self, src_buf: &[u8]) -> usize {
+        let remaining = Self::CAPACITY - self.buf.len();
+        let to_copy = src_buf.len().min(remaining);
+        self.buf.extend_from_slice(&src_buf[..to_copy]);
+        self.offset += to_copy as u64;
+        to_copy
+    }
+
+    /// Internal, possibly buffered, write function
     async fn write_all<Buf: IoBuf + Send>(
         &mut self,
         src_buf: FullSlice<Buf>,
         ctx: &RequestContext,
-    ) -> (FullSlice<Buf>, Result<(), FlushTaskError>) {
-        let res = self
-            .writer
-            // TODO: why are we taking a FullSlice if we're going to pass a borrow downstack?
-            // Can remove all the complexity around owned buffers upstack
-            .write_buffered_borrowed(&src_buf, ctx)
-            .await
-            .map(|len| {
-                self.offset += len as u64;
-            });
+    ) -> (FullSlice<Buf>, Result<(), Error>) {
+        let src_buf = src_buf.into_raw_slice();
+        let src_buf_bounds = src_buf.bounds();
+        let restore = move |src_buf_slice: Slice<_>| {
+            FullSlice::must_new(Slice::from_buf_bounds(
+                src_buf_slice.into_inner(),
+                src_buf_bounds,
+            ))
+        };
 
-        (src_buf, res)
+        if !BUFFERED {
+            assert!(self.buf.is_empty());
+            return self
+                .write_all_unbuffered(FullSlice::must_new(src_buf), ctx)
+                .await;
+        }
+        let remaining = Self::CAPACITY - self.buf.len();
+        let src_buf_len = src_buf.bytes_init();
+        if src_buf_len == 0 {
+            return (restore(src_buf), Ok(()));
+        }
+        let mut src_buf = src_buf.slice(0..src_buf_len);
+        // First try to copy as much as we can into the buffer
+        if remaining > 0 {
+            let copied = self.write_into_buffer(&src_buf);
+            src_buf = src_buf.slice(copied..);
+        }
+        // Then, if the buffer is full, flush it out
+        if self.buf.len() == Self::CAPACITY {
+            if let Err(e) = self.flush_buffer(ctx).await {
+                return (restore(src_buf), Err(e));
+            }
+        }
+        // Finally, write the tail of src_buf:
+        // If it wholly fits into the buffer without
+        // completely filling it, then put it there.
+        // If not, write it out directly.
+        let src_buf = if !src_buf.is_empty() {
+            assert_eq!(self.buf.len(), 0);
+            if src_buf.len() < Self::CAPACITY {
+                let copied = self.write_into_buffer(&src_buf);
+                // We just verified above that src_buf fits into our internal buffer.
+                assert_eq!(copied, src_buf.len());
+                restore(src_buf)
+            } else {
+                let (src_buf, res) = self
+                    .write_all_unbuffered(FullSlice::must_new(src_buf), ctx)
+                    .await;
+                if let Err(e) = res {
+                    return (src_buf, Err(e));
+                }
+                src_buf
+            }
+        } else {
+            restore(src_buf)
+        };
+        (src_buf, Ok(()))
     }
 
     /// Write a blob of data. Returns the offset that it was written to,
@@ -285,7 +301,7 @@ where
         &mut self,
         srcbuf: FullSlice<Buf>,
         ctx: &RequestContext,
-    ) -> (FullSlice<Buf>, Result<u64, WriteBlobError>) {
+    ) -> (FullSlice<Buf>, Result<u64, Error>) {
         let (buf, res) = self
             .write_blob_maybe_compressed(srcbuf, ctx, ImageCompressionAlgorithm::Disabled)
             .await;
@@ -299,10 +315,7 @@ where
         srcbuf: FullSlice<Buf>,
         ctx: &RequestContext,
         algorithm: ImageCompressionAlgorithm,
-    ) -> (
-        FullSlice<Buf>,
-        Result<(u64, CompressionInfo), WriteBlobError>,
-    ) {
+    ) -> (FullSlice<Buf>, Result<(u64, CompressionInfo), Error>) {
         let offset = self.offset;
         let mut compression_info = CompressionInfo {
             written_compressed: false,
@@ -318,16 +331,14 @@ where
             if len < 128 {
                 // Short blob. Write a 1-byte length header
                 io_buf.put_u8(len as u8);
-                let (slice, res) = self.write_all(io_buf.slice_len(), ctx).await;
-                let res = res.map_err(WriteBlobError::Flush);
-                ((slice, res), srcbuf)
+                (self.write_all(io_buf.slice_len(), ctx).await, srcbuf)
             } else {
                 // Write a 4-byte length header
                 if len > MAX_SUPPORTED_BLOB_LEN {
                     return (
                         (
                             io_buf.slice_len(),
-                            Err(WriteBlobError::BlobTooLarge { len }),
+                            Err(Error::other(format!("blob too large ({len} bytes)"))),
                         ),
                         srcbuf,
                     );
@@ -361,9 +372,7 @@ where
                 assert_eq!(len_buf[0] & 0xf0, 0);
                 len_buf[0] |= high_bit_mask;
                 io_buf.extend_from_slice(&len_buf[..]);
-                let (slice, res) = self.write_all(io_buf.slice_len(), ctx).await;
-                let res = res.map_err(WriteBlobError::Flush);
-                ((slice, res), srcbuf)
+                (self.write_all(io_buf.slice_len(), ctx).await, srcbuf)
             }
         }
         .await;
@@ -378,49 +387,33 @@ where
         } else {
             self.write_all(srcbuf, ctx).await
         };
-        let res = res.map_err(WriteBlobError::Flush);
         (srcbuf, res.map(|_| (offset, compression_info)))
     }
+}
 
-    /// Writes a raw blob containing both header and data, returning its offset.
-    pub(crate) async fn write_blob_raw<Buf: IoBuf + Send>(
-        &mut self,
-        raw_with_header: FullSlice<Buf>,
-        ctx: &RequestContext,
-    ) -> (FullSlice<Buf>, Result<u64, WriteBlobError>) {
-        // Verify the header, to ensure we don't write invalid/corrupt data.
-        let header = match Header::decode(&raw_with_header)
-            .context("decoding blob header")
-            .map_err(WriteBlobError::WriteBlobRaw)
-        {
-            Ok(header) => header,
-            Err(err) => return (raw_with_header, Err(err)),
-        };
-        if raw_with_header.len() != header.total_len() {
-            let header_total_len = header.total_len();
-            let raw_len = raw_with_header.len();
-            return (
-                raw_with_header,
-                Err(WriteBlobError::WriteBlobRaw(anyhow::anyhow!(
-                    "header length mismatch: {header_total_len} != {raw_len}"
-                ))),
-            );
-        }
-
-        let offset = self.offset;
-        let (raw_with_header, result) = self.write_all(raw_with_header, ctx).await;
-        let result = result.map_err(WriteBlobError::Flush);
-        (raw_with_header, result.map(|_| offset))
+impl BlobWriter<true> {
+    /// Access the underlying `VirtualFile`.
+    ///
+    /// This function flushes the internal buffer before giving access
+    /// to the underlying `VirtualFile`.
+    pub async fn into_inner(mut self, ctx: &RequestContext) -> Result<VirtualFile, Error> {
+        self.flush_buffer(ctx).await?;
+        Ok(self.inner)
     }
 
-    /// Finish this blob writer and return the underlying `W`.
-    pub async fn shutdown(
-        self,
-        mode: BufferedWriterShutdownMode,
-        ctx: &RequestContext,
-    ) -> Result<W, FlushTaskError> {
-        let (_, file) = self.writer.shutdown(mode, ctx).await?;
-        Ok(file)
+    /// Access the underlying `VirtualFile`.
+    ///
+    /// Unlike [`into_inner`](Self::into_inner), this doesn't flush
+    /// the internal buffer before giving access.
+    pub fn into_inner_no_flush(self) -> VirtualFile {
+        self.inner
+    }
+}
+
+impl BlobWriter<false> {
+    /// Access the underlying `VirtualFile`.
+    pub fn into_inner(self) -> VirtualFile {
+        self.inner
     }
 }
 
@@ -429,25 +422,21 @@ pub(crate) mod tests {
     use camino::Utf8PathBuf;
     use camino_tempfile::Utf8TempDir;
     use rand::{Rng, SeedableRng};
-    use tracing::info_span;
 
     use super::*;
     use crate::context::DownloadBehavior;
     use crate::task_mgr::TaskKind;
     use crate::tenant::block_io::BlockReaderRef;
-    use crate::virtual_file;
-    use crate::virtual_file::TempVirtualFile;
-    use crate::virtual_file::VirtualFile;
 
-    async fn round_trip_test(blobs: &[Vec<u8>]) -> anyhow::Result<()> {
-        round_trip_test_compressed(blobs, false).await
+    async fn round_trip_test<const BUFFERED: bool>(blobs: &[Vec<u8>]) -> Result<(), Error> {
+        round_trip_test_compressed::<BUFFERED>(blobs, false).await
     }
 
-    pub(crate) async fn write_maybe_compressed(
+    pub(crate) async fn write_maybe_compressed<const BUFFERED: bool>(
         blobs: &[Vec<u8>],
         compression: bool,
         ctx: &RequestContext,
-    ) -> anyhow::Result<(Utf8TempDir, Utf8PathBuf, Vec<u64>)> {
+    ) -> Result<(Utf8TempDir, Utf8PathBuf, Vec<u64>), Error> {
         let temp_dir = camino_tempfile::tempdir()?;
         let pathbuf = temp_dir.path().join("file");
         let gate = utils::sync::gate::Gate::default();
@@ -456,19 +445,8 @@ pub(crate) mod tests {
         // Write part (in block to drop the file)
         let mut offsets = Vec::new();
         {
-            let file = TempVirtualFile::new(
-                VirtualFile::open_with_options_v2(
-                    pathbuf.as_path(),
-                    virtual_file::OpenOptions::new()
-                        .create_new(true)
-                        .write(true),
-                    ctx,
-                )
-                .await?,
-                gate.enter()?,
-            );
-            let mut wtr =
-                BlobWriter::new(file, 0, &gate, cancel.clone(), ctx, info_span!("test")).unwrap();
+            let file = VirtualFile::create(pathbuf.as_path(), ctx).await?;
+            let mut wtr = BlobWriter::<BUFFERED>::new(file, 0, &gate, cancel.clone(), ctx);
             for blob in blobs.iter() {
                 let (_, res) = if compression {
                     let res = wtr
@@ -485,28 +463,26 @@ pub(crate) mod tests {
                 let offs = res?;
                 offsets.push(offs);
             }
-            let file = wtr
-                .shutdown(
-                    BufferedWriterShutdownMode::ZeroPadToNextMultiple(PAGE_SZ),
-                    ctx,
-                )
-                .await?;
-            file.disarm_into_inner()
-        };
+            // Write out one page worth of zeros so that we can
+            // read again with read_blk
+            let (_, res) = wtr.write_blob(vec![0; PAGE_SZ].slice_len(), ctx).await;
+            let offs = res?;
+            println!("Writing final blob at offs={offs}");
+            wtr.flush_buffer(ctx).await?;
+        }
         Ok((temp_dir, pathbuf, offsets))
     }
 
-    async fn round_trip_test_compressed(
+    async fn round_trip_test_compressed<const BUFFERED: bool>(
         blobs: &[Vec<u8>],
         compression: bool,
-    ) -> anyhow::Result<()> {
+    ) -> Result<(), Error> {
         let ctx =
             RequestContext::new(TaskKind::UnitTest, DownloadBehavior::Error).with_scope_unit_test();
         let (_temp_dir, pathbuf, offsets) =
-            write_maybe_compressed(blobs, compression, &ctx).await?;
+            write_maybe_compressed::<BUFFERED>(blobs, compression, &ctx).await?;
 
-        println!("Done writing!");
-        let file = VirtualFile::open_v2(pathbuf, &ctx).await?;
+        let file = VirtualFile::open(pathbuf, &ctx).await?;
         let rdr = BlockReaderRef::VirtualFile(&file);
         let rdr = BlockCursor::new_with_compression(rdr, compression);
         for (idx, (blob, offset)) in blobs.iter().zip(offsets.iter()).enumerate() {
@@ -525,27 +501,30 @@ pub(crate) mod tests {
     }
 
     #[tokio::test]
-    async fn test_one() -> anyhow::Result<()> {
+    async fn test_one() -> Result<(), Error> {
         let blobs = &[vec![12, 21, 22]];
-        round_trip_test(blobs).await?;
+        round_trip_test::<false>(blobs).await?;
+        round_trip_test::<true>(blobs).await?;
         Ok(())
     }
 
     #[tokio::test]
-    async fn test_hello_simple() -> anyhow::Result<()> {
+    async fn test_hello_simple() -> Result<(), Error> {
         let blobs = &[
             vec![0, 1, 2, 3],
             b"Hello, World!".to_vec(),
             Vec::new(),
             b"foobar".to_vec(),
         ];
-        round_trip_test(blobs).await?;
-        round_trip_test_compressed(blobs, true).await?;
+        round_trip_test::<false>(blobs).await?;
+        round_trip_test::<true>(blobs).await?;
+        round_trip_test_compressed::<false>(blobs, true).await?;
+        round_trip_test_compressed::<true>(blobs, true).await?;
         Ok(())
     }
 
     #[tokio::test]
-    async fn test_really_big_array() -> anyhow::Result<()> {
+    async fn test_really_big_array() -> Result<(), Error> {
         let blobs = &[
             b"test".to_vec(),
             random_array(10 * PAGE_SZ),
@@ -554,22 +533,25 @@ pub(crate) mod tests {
             vec![0xf3; 24 * PAGE_SZ],
             b"foobar".to_vec(),
         ];
-        round_trip_test(blobs).await?;
-        round_trip_test_compressed(blobs, true).await?;
+        round_trip_test::<false>(blobs).await?;
+        round_trip_test::<true>(blobs).await?;
+        round_trip_test_compressed::<false>(blobs, true).await?;
+        round_trip_test_compressed::<true>(blobs, true).await?;
         Ok(())
     }
 
     #[tokio::test]
-    async fn test_arrays_inc() -> anyhow::Result<()> {
+    async fn test_arrays_inc() -> Result<(), Error> {
         let blobs = (0..PAGE_SZ / 8)
             .map(|v| random_array(v * 16))
             .collect::<Vec<_>>();
-        round_trip_test(&blobs).await?;
+        round_trip_test::<false>(&blobs).await?;
+        round_trip_test::<true>(&blobs).await?;
         Ok(())
     }
 
     #[tokio::test]
-    async fn test_arrays_random_size() -> anyhow::Result<()> {
+    async fn test_arrays_random_size() -> Result<(), Error> {
         let mut rng = rand::rngs::StdRng::seed_from_u64(42);
         let blobs = (0..1024)
             .map(|_| {
@@ -581,18 +563,20 @@ pub(crate) mod tests {
                 random_array(sz.into())
             })
             .collect::<Vec<_>>();
-        round_trip_test(&blobs).await?;
+        round_trip_test::<false>(&blobs).await?;
+        round_trip_test::<true>(&blobs).await?;
         Ok(())
     }
 
     #[tokio::test]
-    async fn test_arrays_page_boundary() -> anyhow::Result<()> {
+    async fn test_arrays_page_boundary() -> Result<(), Error> {
         let blobs = &[
             random_array(PAGE_SZ - 4),
             random_array(PAGE_SZ - 4),
             random_array(PAGE_SZ - 4),
         ];
-        round_trip_test(blobs).await?;
+        round_trip_test::<false>(blobs).await?;
+        round_trip_test::<true>(blobs).await?;
         Ok(())
     }
 }

pageserver/src/tenant/block_io.rs

@@ -4,12 +4,14 @@
 
 use std::ops::Deref;
 
+use bytes::Bytes;
+
 use super::storage_layer::delta_layer::{Adapter, DeltaLayerInner};
 use crate::context::RequestContext;
 use crate::page_cache::{self, FileId, PAGE_SZ, PageReadGuard, PageWriteGuard, ReadBufResult};
 #[cfg(test)]
 use crate::virtual_file::IoBufferMut;
-use crate::virtual_file::{IoBuffer, VirtualFile};
+use crate::virtual_file::VirtualFile;
 
 /// This is implemented by anything that can read 8 kB (PAGE_SZ)
 /// blocks, using the page cache
@@ -245,17 +247,17 @@ pub trait BlockWriter {
     /// 'buf' must be of size PAGE_SZ. Returns the block number the page was
     /// written to.
     ///
-    fn write_blk(&mut self, buf: IoBuffer) -> Result<u32, std::io::Error>;
+    fn write_blk(&mut self, buf: Bytes) -> Result<u32, std::io::Error>;
 }
 
 ///
 /// A simple in-memory buffer of blocks.
 ///
 pub struct BlockBuf {
-    pub blocks: Vec<IoBuffer>,
+    pub blocks: Vec<Bytes>,
 }
 impl BlockWriter for BlockBuf {
-    fn write_blk(&mut self, buf: IoBuffer) -> Result<u32, std::io::Error> {
+    fn write_blk(&mut self, buf: Bytes) -> Result<u32, std::io::Error> {
         assert!(buf.len() == PAGE_SZ);
         let blknum = self.blocks.len();
         self.blocks.push(buf);

pageserver/src/tenant/disk_btree.rs

@@ -25,7 +25,7 @@ use std::{io, result};
 
 use async_stream::try_stream;
 use byteorder::{BE, ReadBytesExt};
-use bytes::BufMut;
+use bytes::{BufMut, Bytes, BytesMut};
 use either::Either;
 use futures::{Stream, StreamExt};
 use hex;
@@ -34,7 +34,6 @@ use tracing::error;
 
 use crate::context::RequestContext;
 use crate::tenant::block_io::{BlockReader, BlockWriter};
-use crate::virtual_file::{IoBuffer, IoBufferMut, owned_buffers_io::write::Buffer};
 
 // The maximum size of a value stored in the B-tree. 5 bytes is enough currently.
 pub const VALUE_SZ: usize = 5;
@@ -788,12 +787,12 @@ impl<const L: usize> BuildNode<L> {
     ///
     /// Serialize the node to on-disk format.
     ///
-    fn pack(&self) -> IoBuffer {
+    fn pack(&self) -> Bytes {
         assert!(self.keys.len() == self.num_children as usize * self.suffix_len);
         assert!(self.values.len() == self.num_children as usize * VALUE_SZ);
         assert!(self.num_children > 0);
 
-        let mut buf = IoBufferMut::with_capacity(PAGE_SZ);
+        let mut buf = BytesMut::new();
 
         buf.put_u16(self.num_children);
         buf.put_u8(self.level);
@@ -806,7 +805,7 @@ impl<const L: usize> BuildNode<L> {
         assert!(buf.len() == self.size);
 
         assert!(buf.len() <= PAGE_SZ);
-        buf.extend_with(0, PAGE_SZ - buf.len());
+        buf.resize(PAGE_SZ, 0);
         buf.freeze()
     }
 
@@ -840,7 +839,7 @@ pub(crate) mod tests {
 
     #[derive(Clone, Default)]
     pub(crate) struct TestDisk {
-        blocks: Vec<IoBuffer>,
+        blocks: Vec<Bytes>,
     }
     impl TestDisk {
         fn new() -> Self {
@@ -858,7 +857,7 @@ pub(crate) mod tests {
         }
     }
     impl BlockWriter for &mut TestDisk {
-        fn write_blk(&mut self, buf: IoBuffer) -> io::Result<u32> {
+        fn write_blk(&mut self, buf: Bytes) -> io::Result<u32> {
             let blknum = self.blocks.len();
             self.blocks.push(buf);
             Ok(blknum as u32)

pageserver/src/tenant/ephemeral_file.rs

@@ -12,7 +12,6 @@ use tokio_epoll_uring::{BoundedBuf, Slice};
 use tokio_util::sync::CancellationToken;
 use tracing::{error, info_span};
 use utils::id::TimelineId;
-use utils::sync::gate::GateGuard;
 
 use crate::assert_u64_eq_usize::{U64IsUsize, UsizeIsU64};
 use crate::config::PageServerConf;
@@ -22,33 +21,16 @@ use crate::tenant::storage_layer::inmemory_layer::vectored_dio_read::File;
 use crate::virtual_file::owned_buffers_io::io_buf_aligned::IoBufAlignedMut;
 use crate::virtual_file::owned_buffers_io::slice::SliceMutExt;
 use crate::virtual_file::owned_buffers_io::write::{Buffer, FlushTaskError};
-use crate::virtual_file::{self, IoBufferMut, TempVirtualFile, VirtualFile, owned_buffers_io};
-
-use self::owned_buffers_io::write::OwnedAsyncWriter;
+use crate::virtual_file::{self, IoBufferMut, VirtualFile, owned_buffers_io};
 
 pub struct EphemeralFile {
     _tenant_shard_id: TenantShardId,
     _timeline_id: TimelineId,
     page_cache_file_id: page_cache::FileId,
     bytes_written: u64,
-    file: TempVirtualFileCoOwnedByEphemeralFileAndBufferedWriter,
-    buffered_writer: BufferedWriter,
-}
-
-type BufferedWriter = owned_buffers_io::write::BufferedWriter<
-    IoBufferMut,
-    TempVirtualFileCoOwnedByEphemeralFileAndBufferedWriter,
->;
-
-/// A TempVirtualFile that is co-owned by the [`EphemeralFile`]` and [`BufferedWriter`].
-///
-/// (Actually [`BufferedWriter`] internally is just a client to a background flush task.
-/// The co-ownership is between [`EphemeralFile`] and that flush task.)
-///
-/// Co-ownership allows us to serve reads for data that has already been flushed by the [`BufferedWriter`].
-#[derive(Debug, Clone)]
-struct TempVirtualFileCoOwnedByEphemeralFileAndBufferedWriter {
-    inner: Arc<TempVirtualFile>,
+    buffered_writer: owned_buffers_io::write::BufferedWriter<IoBufferMut, VirtualFile>,
+    /// Gate guard is held on as long as we need to do operations in the path (delete on drop)
+    _gate_guard: utils::sync::gate::GateGuard,
 }
 
 const TAIL_SZ: usize = 64 * 1024;
@@ -62,12 +44,9 @@ impl EphemeralFile {
         cancel: &CancellationToken,
         ctx: &RequestContext,
     ) -> anyhow::Result<EphemeralFile> {
-        // TempVirtualFile requires us to never reuse a filename while an old
-        // instance of TempVirtualFile created with that filename is not done dropping yet.
-        // So, we use a monotonic counter to disambiguate the filenames.
-        static NEXT_TEMP_DISAMBIGUATOR: AtomicU64 = AtomicU64::new(1);
+        static NEXT_FILENAME: AtomicU64 = AtomicU64::new(1);
         let filename_disambiguator =
-            NEXT_TEMP_DISAMBIGUATOR.fetch_add(1, std::sync::atomic::Ordering::Relaxed);
+            NEXT_FILENAME.fetch_add(1, std::sync::atomic::Ordering::Relaxed);
 
         let filename = conf
             .timeline_path(&tenant_shard_id, &timeline_id)
@@ -75,17 +54,16 @@ impl EphemeralFile {
                 "ephemeral-{filename_disambiguator}"
             )));
 
-        let file = TempVirtualFileCoOwnedByEphemeralFileAndBufferedWriter::new(
+        let file = Arc::new(
             VirtualFile::open_with_options_v2(
                 &filename,
                 virtual_file::OpenOptions::new()
-                    .create_new(true)
                     .read(true)
-                    .write(true),
+                    .write(true)
+                    .create(true),
                 ctx,
             )
             .await?,
-            gate.enter()?,
         );
 
         let page_cache_file_id = page_cache::next_file_id(); // XXX get rid, we're not page-caching anymore
@@ -95,60 +73,37 @@ impl EphemeralFile {
             _timeline_id: timeline_id,
             page_cache_file_id,
             bytes_written: 0,
-            file: file.clone(),
-            buffered_writer: BufferedWriter::new(
+            buffered_writer: owned_buffers_io::write::BufferedWriter::new(
                 file,
-                0,
                 || IoBufferMut::with_capacity(TAIL_SZ),
                 gate.enter()?,
                 cancel.child_token(),
                 ctx,
                 info_span!(parent: None, "ephemeral_file_buffered_writer", tenant_id=%tenant_shard_id.tenant_id, shard_id=%tenant_shard_id.shard_slug(), timeline_id=%timeline_id, path = %filename),
             ),
+            _gate_guard: gate.enter()?,
         })
     }
 }
 
-impl TempVirtualFileCoOwnedByEphemeralFileAndBufferedWriter {
-    fn new(file: VirtualFile, gate_guard: GateGuard) -> Self {
-        Self {
-            inner: Arc::new(TempVirtualFile::new(file, gate_guard)),
+impl Drop for EphemeralFile {
+    fn drop(&mut self) {
+        // unlink the file
+        // we are clear to do this, because we have entered a gate
+        let path = self.buffered_writer.as_inner().path();
+        let res = std::fs::remove_file(path);
+        if let Err(e) = res {
+            if e.kind() != std::io::ErrorKind::NotFound {
+                // just never log the not found errors, we cannot do anything for them; on detach
+                // the tenant directory is already gone.
+                //
+                // not found files might also be related to https://github.com/neondatabase/neon/issues/2442
+                error!("could not remove ephemeral file '{path}': {e}");
+            }
         }
     }
 }
 
-impl OwnedAsyncWriter for TempVirtualFileCoOwnedByEphemeralFileAndBufferedWriter {
-    fn write_all_at<Buf: owned_buffers_io::io_buf_aligned::IoBufAligned + Send>(
-        &self,
-        buf: owned_buffers_io::io_buf_ext::FullSlice<Buf>,
-        offset: u64,
-        ctx: &RequestContext,
-    ) -> impl std::future::Future<
-        Output = (
-            owned_buffers_io::io_buf_ext::FullSlice<Buf>,
-            std::io::Result<()>,
-        ),
-    > + Send {
-        self.inner.write_all_at(buf, offset, ctx)
-    }
-
-    fn set_len(
-        &self,
-        len: u64,
-        ctx: &RequestContext,
-    ) -> impl Future<Output = std::io::Result<()>> + Send {
-        self.inner.set_len(len, ctx)
-    }
-}
-
-impl std::ops::Deref for TempVirtualFileCoOwnedByEphemeralFileAndBufferedWriter {
-    type Target = VirtualFile;
-
-    fn deref(&self) -> &Self::Target {
-        &self.inner
-    }
-}
-
 #[derive(Debug, thiserror::Error)]
 pub(crate) enum EphemeralFileWriteError {
     #[error("{0}")]
@@ -307,9 +262,9 @@ impl super::storage_layer::inmemory_layer::vectored_dio_read::File for Ephemeral
         let mutable_range = Range(std::cmp::max(start, submitted_offset), end);
 
         let dst = if written_range.len() > 0 {
+            let file: &VirtualFile = self.buffered_writer.as_inner();
             let bounds = dst.bounds();
-            let slice = self
-                .file
+            let slice = file
                 .read_exact_at(dst.slice(0..written_range.len().into_usize()), start, ctx)
                 .await?;
             Slice::from_buf_bounds(Slice::into_inner(slice), bounds)
@@ -501,7 +456,7 @@ mod tests {
             assert_eq!(&buf, &content[range]);
         }
 
-        let file_contents = std::fs::read(file.file.path()).unwrap();
+        let file_contents = std::fs::read(file.buffered_writer.as_inner().path()).unwrap();
         assert!(file_contents == content[0..cap * 2]);
 
         let maybe_flushed_buffer_contents = file.buffered_writer.inspect_maybe_flushed().unwrap();
@@ -534,7 +489,7 @@ mod tests {
         // assert the state is as this test expects it to be
         let load_io_buf_res = file.load_to_io_buf(&ctx).await.unwrap();
         assert_eq!(&load_io_buf_res[..], &content[0..cap * 2 + cap / 2]);
-        let md = file.file.path().metadata().unwrap();
+        let md = file.buffered_writer.as_inner().path().metadata().unwrap();
         assert_eq!(
             md.len(),
             2 * cap.into_u64(),

pageserver/src/tenant/metadata.rs

@@ -564,9 +564,8 @@ mod tests {
             Lsn(0),
             Lsn(0),
             Lsn(0),
-            // Updating this version to 17 will cause the test to fail at the
-            // next assert_eq!().
-            16,
+            // Any version will do here, so use the default
+            crate::DEFAULT_PG_VERSION,
         );
         let expected_bytes = vec![
             /* TimelineMetadataHeader */

pageserver/src/tenant/mgr.rs

@@ -52,9 +52,7 @@ use crate::tenant::config::{
 use crate::tenant::span::debug_assert_current_span_has_tenant_id;
 use crate::tenant::storage_layer::inmemory_layer;
 use crate::tenant::timeline::ShutdownMode;
-use crate::tenant::{
-    AttachedTenantConf, GcError, LoadConfigError, SpawnMode, TenantShard, TenantState,
-};
+use crate::tenant::{AttachedTenantConf, GcError, LoadConfigError, SpawnMode, Tenant, TenantState};
 use crate::virtual_file::MaybeFatalIo;
 use crate::{InitializationOrder, TEMP_FILE_SUFFIX};
 
@@ -69,7 +67,7 @@ use crate::{InitializationOrder, TEMP_FILE_SUFFIX};
 /// having a properly acquired generation (Secondary doesn't need a generation)
 #[derive(Clone)]
 pub(crate) enum TenantSlot {
-    Attached(Arc<TenantShard>),
+    Attached(Arc<Tenant>),
     Secondary(Arc<SecondaryTenant>),
     /// In this state, other administrative operations acting on the TenantId should
     /// block, or return a retry indicator equivalent to HTTP 503.
@@ -88,7 +86,7 @@ impl std::fmt::Debug for TenantSlot {
 
 impl TenantSlot {
     /// Return the `Tenant` in this slot if attached, else None
-    fn get_attached(&self) -> Option<&Arc<TenantShard>> {
+    fn get_attached(&self) -> Option<&Arc<Tenant>> {
         match self {
             Self::Attached(t) => Some(t),
             Self::Secondary(_) => None,
@@ -166,7 +164,7 @@ impl TenantStartupMode {
 /// Result type for looking up a TenantId to a specific shard
 pub(crate) enum ShardResolveResult {
     NotFound,
-    Found(Arc<TenantShard>),
+    Found(Arc<Tenant>),
     // Wait for this barrrier, then query again
     InProgress(utils::completion::Barrier),
 }
@@ -175,7 +173,7 @@ impl TenantsMap {
     /// Convenience function for typical usage, where we want to get a `Tenant` object, for
     /// working with attached tenants.  If the TenantId is in the map but in Secondary state,
     /// None is returned.
-    pub(crate) fn get(&self, tenant_shard_id: &TenantShardId) -> Option<&Arc<TenantShard>> {
+    pub(crate) fn get(&self, tenant_shard_id: &TenantShardId) -> Option<&Arc<Tenant>> {
         match self {
             TenantsMap::Initializing => None,
             TenantsMap::Open(m) | TenantsMap::ShuttingDown(m) => {
@@ -412,7 +410,7 @@ fn load_tenant_config(
         return None;
     }
 
-    Some(TenantShard::load_tenant_config(conf, &tenant_shard_id))
+    Some(Tenant::load_tenant_config(conf, &tenant_shard_id))
 }
 
 /// Initial stage of load: walk the local tenants directory, clean up any temp files,
@@ -608,8 +606,7 @@ pub async fn init_tenant_mgr(
         // Presence of a generation number implies attachment: attach the tenant
         // if it wasn't already, and apply the generation number.
         config_write_futs.push(async move {
-            let r =
-                TenantShard::persist_tenant_config(conf, &tenant_shard_id, &location_conf).await;
+            let r = Tenant::persist_tenant_config(conf, &tenant_shard_id, &location_conf).await;
             (tenant_shard_id, location_conf, r)
         });
     }
@@ -697,7 +694,7 @@ fn tenant_spawn(
     init_order: Option<InitializationOrder>,
     mode: SpawnMode,
     ctx: &RequestContext,
-) -> Result<Arc<TenantShard>, GlobalShutDown> {
+) -> Result<Arc<Tenant>, GlobalShutDown> {
     // All these conditions should have been satisfied by our caller: the tenant dir exists, is a well formed
     // path, and contains a configuration file.  Assertions that do synchronous I/O are limited to debug mode
     // to avoid impacting prod runtime performance.
@@ -709,7 +706,7 @@ fn tenant_spawn(
             .unwrap()
     );
 
-    TenantShard::spawn(
+    Tenant::spawn(
         conf,
         tenant_shard_id,
         resources,
@@ -886,12 +883,12 @@ impl TenantManager {
     /// Gets the attached tenant from the in-memory data, erroring if it's absent, in secondary mode, or currently
     /// undergoing a state change (i.e. slot is InProgress).
     ///
-    /// The return TenantShard is not guaranteed to be active: check its status after obtaing it, or
-    /// use [`TenantShard::wait_to_become_active`] before using it if you will do I/O on it.
+    /// The return Tenant is not guaranteed to be active: check its status after obtaing it, or
+    /// use [`Tenant::wait_to_become_active`] before using it if you will do I/O on it.
     pub(crate) fn get_attached_tenant_shard(
         &self,
         tenant_shard_id: TenantShardId,
-    ) -> Result<Arc<TenantShard>, GetTenantError> {
+    ) -> Result<Arc<Tenant>, GetTenantError> {
         let locked = self.tenants.read().unwrap();
 
         let peek_slot = tenant_map_peek_slot(&locked, &tenant_shard_id, TenantSlotPeekMode::Read)?;
@@ -940,12 +937,12 @@ impl TenantManager {
         flush: Option<Duration>,
         mut spawn_mode: SpawnMode,
         ctx: &RequestContext,
-    ) -> Result<Option<Arc<TenantShard>>, UpsertLocationError> {
+    ) -> Result<Option<Arc<Tenant>>, UpsertLocationError> {
         debug_assert_current_span_has_tenant_id();
         info!("configuring tenant location to state {new_location_config:?}");
 
         enum FastPathModified {
-            Attached(Arc<TenantShard>),
+            Attached(Arc<Tenant>),
             Secondary(Arc<SecondaryTenant>),
         }
 
@@ -1002,13 +999,9 @@ impl TenantManager {
         // phase of writing config and/or waiting for flush, before returning.
         match fast_path_taken {
             Some(FastPathModified::Attached(tenant)) => {
-                TenantShard::persist_tenant_config(
-                    self.conf,
-                    &tenant_shard_id,
-                    &new_location_config,
-                )
-                .await
-                .fatal_err("write tenant shard config");
+                Tenant::persist_tenant_config(self.conf, &tenant_shard_id, &new_location_config)
+                    .await
+                    .fatal_err("write tenant shard config");
 
                 // Transition to AttachedStale means we may well hold a valid generation
                 // still, and have been requested to go stale as part of a migration.  If
@@ -1037,13 +1030,9 @@ impl TenantManager {
                 return Ok(Some(tenant));
             }
             Some(FastPathModified::Secondary(_secondary_tenant)) => {
-                TenantShard::persist_tenant_config(
-                    self.conf,
-                    &tenant_shard_id,
-                    &new_location_config,
-                )
-                .await
-                .fatal_err("write tenant shard config");
+                Tenant::persist_tenant_config(self.conf, &tenant_shard_id, &new_location_config)
+                    .await
+                    .fatal_err("write tenant shard config");
 
                 return Ok(None);
             }
@@ -1133,7 +1122,7 @@ impl TenantManager {
         // Before activating either secondary or attached mode, persist the
         // configuration, so that on restart we will re-attach (or re-start
         // secondary) on the tenant.
-        TenantShard::persist_tenant_config(self.conf, &tenant_shard_id, &new_location_config)
+        Tenant::persist_tenant_config(self.conf, &tenant_shard_id, &new_location_config)
             .await
             .fatal_err("write tenant shard config");
 
@@ -1273,7 +1262,7 @@ impl TenantManager {
 
         let tenant_path = self.conf.tenant_path(&tenant_shard_id);
         let timelines_path = self.conf.timelines_path(&tenant_shard_id);
-        let config = TenantShard::load_tenant_config(self.conf, &tenant_shard_id)?;
+        let config = Tenant::load_tenant_config(self.conf, &tenant_shard_id)?;
 
         if drop_cache {
             tracing::info!("Dropping local file cache");
@@ -1308,7 +1297,7 @@ impl TenantManager {
         Ok(())
     }
 
-    pub(crate) fn get_attached_active_tenant_shards(&self) -> Vec<Arc<TenantShard>> {
+    pub(crate) fn get_attached_active_tenant_shards(&self) -> Vec<Arc<Tenant>> {
         let locked = self.tenants.read().unwrap();
         match &*locked {
             TenantsMap::Initializing => Vec::new(),
@@ -1457,7 +1446,7 @@ impl TenantManager {
     #[instrument(skip_all, fields(tenant_id=%tenant.get_tenant_shard_id().tenant_id, shard_id=%tenant.get_tenant_shard_id().shard_slug(), new_shard_count=%new_shard_count.literal()))]
     pub(crate) async fn shard_split(
         &self,
-        tenant: Arc<TenantShard>,
+        tenant: Arc<Tenant>,
         new_shard_count: ShardCount,
         new_stripe_size: Option<ShardStripeSize>,
         ctx: &RequestContext,
@@ -1487,7 +1476,7 @@ impl TenantManager {
 
     pub(crate) async fn do_shard_split(
         &self,
-        tenant: Arc<TenantShard>,
+        tenant: Arc<Tenant>,
         new_shard_count: ShardCount,
         new_stripe_size: Option<ShardStripeSize>,
         ctx: &RequestContext,
@@ -1714,7 +1703,7 @@ impl TenantManager {
     /// For each resident layer in the parent shard, we will hard link it into all of the child shards.
     async fn shard_split_hardlink(
         &self,
-        parent_shard: &TenantShard,
+        parent_shard: &Tenant,
         child_shards: Vec<TenantShardId>,
     ) -> anyhow::Result<()> {
         debug_assert_current_span_has_tenant_id();
@@ -1999,7 +1988,7 @@ impl TenantManager {
             }
 
             let tenant_path = self.conf.tenant_path(&tenant_shard_id);
-            let config = TenantShard::load_tenant_config(self.conf, &tenant_shard_id)
+            let config = Tenant::load_tenant_config(self.conf, &tenant_shard_id)
                 .map_err(|e| Error::DetachReparent(e.into()))?;
 
             let shard_identity = config.shard;

pageserver/src/tenant/remote_timeline_client.rs

@@ -133,7 +133,7 @@
 //! - Initiate upload queue with that [`IndexPart`].
 //! - Reschedule all lost operations by comparing the local filesystem state
 //!   and remote state as per [`IndexPart`]. This is done in
-//!   [`TenantShard::timeline_init_and_sync`].
+//!   [`Tenant::timeline_init_and_sync`].
 //!
 //! Note that if we crash during file deletion between the index update
 //! that removes the file from the list of files, and deleting the remote file,
@@ -171,7 +171,7 @@
 //! If no remote storage configuration is provided, the [`RemoteTimelineClient`] is
 //! not created and the uploads are skipped.
 //!
-//! [`TenantShard::timeline_init_and_sync`]: super::TenantShard::timeline_init_and_sync
+//! [`Tenant::timeline_init_and_sync`]: super::Tenant::timeline_init_and_sync
 //! [`Timeline::load_layer_map`]: super::Timeline::load_layer_map
 
 pub(crate) mod download;
@@ -2743,7 +2743,7 @@ mod tests {
     use crate::tenant::config::AttachmentMode;
     use crate::tenant::harness::{TIMELINE_ID, TenantHarness};
     use crate::tenant::storage_layer::layer::local_layer_path;
-    use crate::tenant::{TenantShard, Timeline};
+    use crate::tenant::{Tenant, Timeline};
 
     pub(super) fn dummy_contents(name: &str) -> Vec<u8> {
         format!("contents for {name}").into()
@@ -2796,7 +2796,7 @@ mod tests {
 
     struct TestSetup {
         harness: TenantHarness,
-        tenant: Arc<TenantShard>,
+        tenant: Arc<Tenant>,
         timeline: Arc<Timeline>,
         tenant_ctx: RequestContext,
     }

pageserver/src/tenant/remote_timeline_client/download.rs

@@ -6,7 +6,6 @@
 use std::collections::HashSet;
 use std::future::Future;
 use std::str::FromStr;
-use std::sync::atomic::AtomicU64;
 use std::time::SystemTime;
 
 use anyhow::{Context, anyhow};
@@ -16,7 +15,7 @@ use remote_storage::{
     DownloadError, DownloadKind, DownloadOpts, GenericRemoteStorage, ListingMode, RemotePath,
 };
 use tokio::fs::{self, File, OpenOptions};
-use tokio::io::AsyncSeekExt;
+use tokio::io::{AsyncSeekExt, AsyncWriteExt};
 use tokio_util::io::StreamReader;
 use tokio_util::sync::CancellationToken;
 use tracing::warn;
@@ -41,10 +40,7 @@ use crate::span::{
 use crate::tenant::Generation;
 use crate::tenant::remote_timeline_client::{remote_layer_path, remote_timelines_path};
 use crate::tenant::storage_layer::LayerName;
-use crate::virtual_file;
-use crate::virtual_file::owned_buffers_io::write::FlushTaskError;
-use crate::virtual_file::{IoBufferMut, MaybeFatalIo, VirtualFile};
-use crate::virtual_file::{TempVirtualFile, owned_buffers_io};
+use crate::virtual_file::{MaybeFatalIo, VirtualFile, on_fatal_io_error};
 
 ///
 /// If 'metadata' is given, we will validate that the downloaded file's size matches that
@@ -76,34 +72,21 @@ pub async fn download_layer_file<'a>(
         layer_metadata.generation,
     );
 
-    let (bytes_amount, temp_file) = download_retry(
+    // Perform a rename inspired by durable_rename from file_utils.c.
+    // The sequence:
+    //     write(tmp)
+    //     fsync(tmp)
+    //     rename(tmp, new)
+    //     fsync(new)
+    //     fsync(parent)
+    // For more context about durable_rename check this email from postgres mailing list:
+    // https://www.postgresql.org/message-id/56583BDD.9060302@2ndquadrant.com
+    // If pageserver crashes the temp file will be deleted on startup and re-downloaded.
+    let temp_file_path = path_with_suffix_extension(local_path, TEMP_DOWNLOAD_EXTENSION);
+
+    let bytes_amount = download_retry(
         || async {
-            // TempVirtualFile requires us to never reuse a filename while an old
-            // instance of TempVirtualFile created with that filename is not done dropping yet.
-            // So, we use a monotonic counter to disambiguate the filenames.
-            static NEXT_TEMP_DISAMBIGUATOR: AtomicU64 = AtomicU64::new(1);
-            let filename_disambiguator =
-                NEXT_TEMP_DISAMBIGUATOR.fetch_add(1, std::sync::atomic::Ordering::Relaxed);
-
-            let temp_file_path = path_with_suffix_extension(
-                local_path,
-                &format!("{filename_disambiguator:x}.{TEMP_DOWNLOAD_EXTENSION}"),
-            );
-
-            let temp_file = TempVirtualFile::new(
-                VirtualFile::open_with_options_v2(
-                    &temp_file_path,
-                    virtual_file::OpenOptions::new()
-                        .create_new(true)
-                        .write(true),
-                    ctx,
-                )
-                .await
-                .with_context(|| format!("create a temp file for layer download: {temp_file_path}"))
-                .map_err(DownloadError::Other)?,
-                gate.enter().map_err(|_| DownloadError::Cancelled)?,
-            );
-            download_object(storage, &remote_path, temp_file, gate, cancel, ctx).await
+            download_object(storage, &remote_path, &temp_file_path, gate, cancel, ctx).await
         },
         &format!("download {remote_path:?}"),
         cancel,
@@ -113,8 +96,7 @@ pub async fn download_layer_file<'a>(
     let expected = layer_metadata.file_size;
     if expected != bytes_amount {
         return Err(DownloadError::Other(anyhow!(
-            "According to layer file metadata should have downloaded {expected} bytes but downloaded {bytes_amount} bytes into file {:?}",
-            temp_file.path()
+            "According to layer file metadata should have downloaded {expected} bytes but downloaded {bytes_amount} bytes into file {temp_file_path:?}",
         )));
     }
 
@@ -124,28 +106,11 @@ pub async fn download_layer_file<'a>(
         )))
     });
 
-    // Try rename before disarming the temp file.
-    // That way, if rename fails for whatever reason, we clean up the temp file on the return path.
-
-    fs::rename(temp_file.path(), &local_path)
+    fs::rename(&temp_file_path, &local_path)
         .await
         .with_context(|| format!("rename download layer file to {local_path}"))
         .map_err(DownloadError::Other)?;
 
-    // The temp file's VirtualFile points to the temp_file_path which we moved above.
-    // Drop it immediately, it's invalid.
-    // This will get better in https://github.com/neondatabase/neon/issues/11692
-    let _: VirtualFile = temp_file.disarm_into_inner();
-    // NB: The gate guard that was stored in `temp_file` is dropped but we continue
-    // to operate on it and on the parent timeline directory.
-    // Those operations are safe to do because higher-level code is holding another gate guard:
-    // - attached mode: the download task spawned by struct Layer is holding the gate guard
-    // - secondary mode: The TenantDownloader::download holds the gate open
-
-    // The rename above is not durable yet.
-    // It doesn't matter for crash consistency because pageserver startup deletes temp
-    // files and we'll re-download on demand if necessary.
-
     // We use fatal_err() below because the after the rename above,
     // the in-memory state of the filesystem already has the layer file in its final place,
     // and subsequent pageserver code could think it's durable while it really isn't.
@@ -181,64 +146,147 @@ pub async fn download_layer_file<'a>(
 async fn download_object(
     storage: &GenericRemoteStorage,
     src_path: &RemotePath,
-    destination_file: TempVirtualFile,
-    gate: &utils::sync::gate::Gate,
+    dst_path: &Utf8PathBuf,
+    #[cfg_attr(target_os = "macos", allow(unused_variables))] gate: &utils::sync::gate::Gate,
     cancel: &CancellationToken,
-    ctx: &RequestContext,
-) -> Result<(u64, TempVirtualFile), DownloadError> {
-    let mut download = storage
-        .download(src_path, &DownloadOpts::default(), cancel)
-        .await?;
+    #[cfg_attr(target_os = "macos", allow(unused_variables))] ctx: &RequestContext,
+) -> Result<u64, DownloadError> {
+    let res = match crate::virtual_file::io_engine::get() {
+        crate::virtual_file::io_engine::IoEngine::NotSet => panic!("unset"),
+        crate::virtual_file::io_engine::IoEngine::StdFs => {
+            async {
+                let destination_file = tokio::fs::File::create(dst_path)
+                    .await
+                    .with_context(|| format!("create a destination file for layer '{dst_path}'"))
+                    .map_err(DownloadError::Other)?;
 
-    pausable_failpoint!("before-downloading-layer-stream-pausable");
+                let download = storage
+                    .download(src_path, &DownloadOpts::default(), cancel)
+                    .await?;
 
-    let dst_path = destination_file.path().to_owned();
-    let mut buffered = owned_buffers_io::write::BufferedWriter::<IoBufferMut, _>::new(
-        destination_file,
-        0,
-        || IoBufferMut::with_capacity(super::BUFFER_SIZE),
-        gate.enter().map_err(|_| DownloadError::Cancelled)?,
-        cancel.child_token(),
-        ctx,
-        tracing::info_span!(parent: None, "download_object_buffered_writer", %dst_path),
-    );
+                pausable_failpoint!("before-downloading-layer-stream-pausable");
 
-    // TODO: use vectored write (writev) once supported by tokio-epoll-uring.
-    // There's chunks_vectored() on the stream.
-    let (bytes_amount, destination_file) = async {
-        while let Some(res) = futures::StreamExt::next(&mut download.download_stream).await {
-            let chunk = match res {
-                Ok(chunk) => chunk,
-                Err(e) => return Err(DownloadError::from(e)),
-            };
-            buffered
-                .write_buffered_borrowed(&chunk, ctx)
-                .await
-                .map_err(|e| match e {
-                    FlushTaskError::Cancelled => DownloadError::Cancelled,
-                })?;
-        }
-        buffered
-            .shutdown(
-                owned_buffers_io::write::BufferedWriterShutdownMode::PadThenTruncate,
-                ctx,
-            )
+                let mut buf_writer =
+                    tokio::io::BufWriter::with_capacity(super::BUFFER_SIZE, destination_file);
+
+                let mut reader = tokio_util::io::StreamReader::new(download.download_stream);
+
+                let bytes_amount = tokio::io::copy_buf(&mut reader, &mut buf_writer).await?;
+                buf_writer.flush().await?;
+
+                let mut destination_file = buf_writer.into_inner();
+
+                // Tokio doc here: https://docs.rs/tokio/1.17.0/tokio/fs/struct.File.html states that:
+                // A file will not be closed immediately when it goes out of scope if there are any IO operations
+                // that have not yet completed. To ensure that a file is closed immediately when it is dropped,
+                // you should call flush before dropping it.
+                //
+                // From the tokio code I see that it waits for pending operations to complete. There shouldt be any because
+                // we assume that `destination_file` file is fully written. I e there is no pending .write(...).await operations.
+                // But for additional safety lets check/wait for any pending operations.
+                destination_file
+                    .flush()
+                    .await
+                    .maybe_fatal_err("download_object sync_all")
+                    .with_context(|| format!("flush source file at {dst_path}"))
+                    .map_err(DownloadError::Other)?;
+
+                // not using sync_data because it can lose file size update
+                destination_file
+                    .sync_all()
+                    .await
+                    .maybe_fatal_err("download_object sync_all")
+                    .with_context(|| format!("failed to fsync source file at {dst_path}"))
+                    .map_err(DownloadError::Other)?;
+
+                Ok(bytes_amount)
+            }
             .await
-            .map_err(|e| match e {
-                FlushTaskError::Cancelled => DownloadError::Cancelled,
-            })
+        }
+        #[cfg(target_os = "linux")]
+        crate::virtual_file::io_engine::IoEngine::TokioEpollUring => {
+            use crate::virtual_file::owned_buffers_io::write::FlushTaskError;
+            use std::sync::Arc;
+
+            use crate::virtual_file::{IoBufferMut, owned_buffers_io};
+            async {
+                let destination_file = Arc::new(
+                    VirtualFile::create(dst_path, ctx)
+                        .await
+                        .with_context(|| {
+                            format!("create a destination file for layer '{dst_path}'")
+                        })
+                        .map_err(DownloadError::Other)?,
+                );
+
+                let mut download = storage
+                    .download(src_path, &DownloadOpts::default(), cancel)
+                    .await?;
+
+                pausable_failpoint!("before-downloading-layer-stream-pausable");
+
+                let mut buffered = owned_buffers_io::write::BufferedWriter::<IoBufferMut, _>::new(
+                    destination_file,
+                    || IoBufferMut::with_capacity(super::BUFFER_SIZE),
+                    gate.enter().map_err(|_| DownloadError::Cancelled)?,
+                    cancel.child_token(),
+                    ctx,
+                    tracing::info_span!(parent: None, "download_object_buffered_writer", %dst_path),
+                );
+
+                // TODO: use vectored write (writev) once supported by tokio-epoll-uring.
+                // There's chunks_vectored() on the stream.
+                let (bytes_amount, destination_file) = async {
+                    while let Some(res) =
+                        futures::StreamExt::next(&mut download.download_stream).await
+                    {
+                        let chunk = match res {
+                            Ok(chunk) => chunk,
+                            Err(e) => return Err(DownloadError::from(e)),
+                        };
+                        buffered
+                            .write_buffered_borrowed(&chunk, ctx)
+                            .await
+                            .map_err(|e| match e {
+                                FlushTaskError::Cancelled => DownloadError::Cancelled,
+                            })?;
+                    }
+                    let inner = buffered
+                        .flush_and_into_inner(ctx)
+                        .await
+                        .map_err(|e| match e {
+                            FlushTaskError::Cancelled => DownloadError::Cancelled,
+                        })?;
+                    Ok(inner)
+                }
+                .await?;
+
+                // not using sync_data because it can lose file size update
+                destination_file
+                    .sync_all()
+                    .await
+                    .maybe_fatal_err("download_object sync_all")
+                    .with_context(|| format!("failed to fsync source file at {dst_path}"))
+                    .map_err(DownloadError::Other)?;
+
+                Ok(bytes_amount)
+            }
+            .await
+        }
+    };
+
+    // in case the download failed, clean up
+    match res {
+        Ok(bytes_amount) => Ok(bytes_amount),
+        Err(e) => {
+            if let Err(e) = tokio::fs::remove_file(dst_path).await {
+                if e.kind() != std::io::ErrorKind::NotFound {
+                    on_fatal_io_error(&e, &format!("Removing temporary file {dst_path}"));
+                }
+            }
+            Err(e)
+        }
     }
-    .await?;
-
-    // not using sync_data because it can lose file size update
-    destination_file
-        .sync_all()
-        .await
-        .maybe_fatal_err("download_object sync_all")
-        .with_context(|| format!("failed to fsync source file at {dst_path}"))
-        .map_err(DownloadError::Other)?;
-
-    Ok((bytes_amount, destination_file))
 }
 
 const TEMP_DOWNLOAD_EXTENSION: &str = "temp_download";
@@ -404,7 +452,7 @@ async fn do_download_index_part(
 /// generation (normal case when migrating/restarting).  Only if both of these return 404 do we fall back
 /// to listing objects.
 ///
-/// * `my_generation`: the value of `[crate::tenant::TenantShard::generation]`
+/// * `my_generation`: the value of `[crate::tenant::Tenant::generation]`
 /// * `what`: for logging, what object are we downloading
 /// * `prefix`: when listing objects, use this prefix (i.e. the part of the object path before the generation)
 /// * `do_download`: a GET of the object in a particular generation, which should **retry indefinitely** unless

pageserver/src/tenant/secondary/downloader.rs

@@ -646,7 +646,7 @@ enum UpdateError {
     NoData,
     #[error("Insufficient local storage space")]
     NoSpace,
-    #[error("Failed to download: {0}")]
+    #[error("Failed to download")]
     DownloadError(DownloadError),
     #[error(transparent)]
     Deserialize(#[from] serde_json::Error),
@@ -1521,11 +1521,12 @@ async fn load_heatmap(
     path: &Utf8PathBuf,
     ctx: &RequestContext,
 ) -> Result<Option<HeatMapTenant>, anyhow::Error> {
-    let st = match VirtualFile::read_to_string(path, ctx).await {
-        Ok(st) => st,
+    let mut file = match VirtualFile::open(path, ctx).await {
+        Ok(file) => file,
         Err(e) if e.kind() == std::io::ErrorKind::NotFound => return Ok(None),
         Err(e) => Err(e)?,
     };
+    let st = file.read_to_string(ctx).await?;
     let htm = serde_json::from_str(&st)?;
     Ok(Some(htm))
 }

pageserver/src/tenant/secondary/heatmap_uploader.rs

@@ -21,7 +21,7 @@ use super::scheduler::{
 use super::{CommandRequest, SecondaryTenantError, UploadCommand};
 use crate::TEMP_FILE_SUFFIX;
 use crate::metrics::SECONDARY_MODE;
-use crate::tenant::TenantShard;
+use crate::tenant::Tenant;
 use crate::tenant::config::AttachmentMode;
 use crate::tenant::mgr::{GetTenantError, TenantManager};
 use crate::tenant::remote_timeline_client::remote_heatmap_path;
@@ -74,7 +74,7 @@ impl RunningJob for WriteInProgress {
 }
 
 struct UploadPending {
-    tenant: Arc<TenantShard>,
+    tenant: Arc<Tenant>,
     last_upload: Option<LastUploadState>,
     target_time: Option<Instant>,
     period: Option<Duration>,
@@ -106,7 +106,7 @@ impl scheduler::Completion for WriteComplete {
 struct UploaderTenantState {
     // This Weak only exists to enable culling idle instances of this type
     // when the Tenant has been deallocated.
-    tenant: Weak<TenantShard>,
+    tenant: Weak<Tenant>,
 
     /// Digest of the serialized heatmap that we last successfully uploaded
     last_upload_state: Option<LastUploadState>,
@@ -357,7 +357,7 @@ struct LastUploadState {
 /// of the object we would have uploaded.
 async fn upload_tenant_heatmap(
     remote_storage: GenericRemoteStorage,
-    tenant: &Arc<TenantShard>,
+    tenant: &Arc<Tenant>,
     last_upload: Option<LastUploadState>,
 ) -> Result<UploadHeatmapOutcome, UploadHeatmapError> {
     debug_assert_current_span_has_tenant_id();

pageserver/src/tenant/secondary/scheduler.rs

@@ -360,7 +360,7 @@ where
 
     /// Periodic execution phase: inspect all attached tenants and schedule any work they require.
     ///
-    /// The type in `tenants` should be a tenant-like structure, e.g. [`crate::tenant::TenantShard`] or [`crate::tenant::secondary::SecondaryTenant`]
+    /// The type in `tenants` should be a tenant-like structure, e.g. [`crate::tenant::Tenant`] or [`crate::tenant::secondary::SecondaryTenant`]
     ///
     /// This function resets the pending list: it is assumed that the caller may change their mind about
     /// which tenants need work between calls to schedule_iteration.

pageserver/src/tenant/size.rs

@@ -12,7 +12,7 @@ use tracing::*;
 use utils::id::TimelineId;
 use utils::lsn::Lsn;
 
-use super::{GcError, LogicalSizeCalculationCause, TenantShard};
+use super::{GcError, LogicalSizeCalculationCause, Tenant};
 use crate::context::RequestContext;
 use crate::pgdatadir_mapping::CalculateLogicalSizeError;
 use crate::tenant::{MaybeOffloaded, Timeline};
@@ -156,7 +156,7 @@ pub struct TimelineInputs {
 ///   initdb_lsn  branchpoints*  next_pitr_cutoff  latest
 /// ```
 pub(super) async fn gather_inputs(
-    tenant: &TenantShard,
+    tenant: &Tenant,
     limit: &Arc<Semaphore>,
     max_retention_period: Option<u64>,
     logical_size_cache: &mut HashMap<(TimelineId, Lsn), u64>,

pageserver/src/tenant/storage_layer/delta_layer.rs

@@ -29,11 +29,11 @@
 //!
 use std::collections::{HashMap, VecDeque};
 use std::fs::File;
+use std::io::SeekFrom;
 use std::ops::Range;
 use std::os::unix::fs::FileExt;
 use std::str::FromStr;
 use std::sync::Arc;
-use std::sync::atomic::AtomicU64;
 
 use anyhow::{Context, Result, bail, ensure};
 use camino::{Utf8Path, Utf8PathBuf};
@@ -45,13 +45,14 @@ use pageserver_api::keyspace::KeySpace;
 use pageserver_api::models::ImageCompressionAlgorithm;
 use pageserver_api::shard::TenantShardId;
 use pageserver_api::value::Value;
+use rand::Rng;
+use rand::distributions::Alphanumeric;
 use serde::{Deserialize, Serialize};
 use tokio::sync::OnceCell;
 use tokio_epoll_uring::IoBuf;
 use tokio_util::sync::CancellationToken;
 use tracing::*;
 use utils::bin_ser::BeSer;
-use utils::bin_ser::SerializeError;
 use utils::id::{TenantId, TimelineId};
 use utils::lsn::Lsn;
 
@@ -73,10 +74,8 @@ use crate::tenant::vectored_blob_io::{
     BlobFlag, BufView, StreamingVectoredReadPlanner, VectoredBlobReader, VectoredRead,
     VectoredReadPlanner,
 };
-use crate::virtual_file::TempVirtualFile;
 use crate::virtual_file::owned_buffers_io::io_buf_ext::{FullSlice, IoBufExt};
-use crate::virtual_file::owned_buffers_io::write::{Buffer, BufferedWriterShutdownMode};
-use crate::virtual_file::{self, IoBuffer, IoBufferMut, MaybeFatalIo, VirtualFile};
+use crate::virtual_file::{self, IoBufferMut, MaybeFatalIo, VirtualFile};
 use crate::{DELTA_FILE_MAGIC, STORAGE_FORMAT_VERSION, TEMP_FILE_SUFFIX};
 
 ///
@@ -114,15 +113,6 @@ impl From<&DeltaLayer> for Summary {
 }
 
 impl Summary {
-    /// Serializes the summary header into an aligned buffer of lenth `PAGE_SZ`.
-    pub fn ser_into_page(&self) -> Result<IoBuffer, SerializeError> {
-        let mut buf = IoBufferMut::with_capacity(PAGE_SZ);
-        Self::ser_into(self, &mut buf)?;
-        // Pad zeroes to the buffer so the length is a multiple of the alignment.
-        buf.extend_with(0, buf.capacity() - buf.len());
-        Ok(buf.freeze())
-    }
-
     pub(super) fn expected(
         tenant_id: TenantId,
         timeline_id: TimelineId,
@@ -298,20 +288,19 @@ impl DeltaLayer {
         key_start: Key,
         lsn_range: &Range<Lsn>,
     ) -> Utf8PathBuf {
-        // TempVirtualFile requires us to never reuse a filename while an old
-        // instance of TempVirtualFile created with that filename is not done dropping yet.
-        // So, we use a monotonic counter to disambiguate the filenames.
-        static NEXT_TEMP_DISAMBIGUATOR: AtomicU64 = AtomicU64::new(1);
-        let filename_disambiguator =
-            NEXT_TEMP_DISAMBIGUATOR.fetch_add(1, std::sync::atomic::Ordering::Relaxed);
+        let rand_string: String = rand::thread_rng()
+            .sample_iter(&Alphanumeric)
+            .take(8)
+            .map(char::from)
+            .collect();
 
         conf.timeline_path(tenant_shard_id, timeline_id)
             .join(format!(
-                "{}-XXX__{:016X}-{:016X}.{:x}.{}",
+                "{}-XXX__{:016X}-{:016X}.{}.{}",
                 key_start,
                 u64::from(lsn_range.start),
                 u64::from(lsn_range.end),
-                filename_disambiguator,
+                rand_string,
                 TEMP_FILE_SUFFIX,
             ))
     }
@@ -402,7 +391,7 @@ struct DeltaLayerWriterInner {
 
     tree: DiskBtreeBuilder<BlockBuf, DELTA_KEY_SIZE>,
 
-    blob_writer: BlobWriter<TempVirtualFile>,
+    blob_writer: BlobWriter<true>,
 
     // Number of key-lsns in the layer.
     num_keys: usize,
@@ -426,29 +415,16 @@ impl DeltaLayerWriterInner {
         // Create the file initially with a temporary filename. We don't know
         // the end key yet, so we cannot form the final filename yet. We will
         // rename it when we're done.
+        //
+        // Note: This overwrites any existing file. There shouldn't be any.
+        // FIXME: throw an error instead?
         let path =
             DeltaLayer::temp_path_for(conf, &tenant_shard_id, &timeline_id, key_start, &lsn_range);
-        let file = TempVirtualFile::new(
-            VirtualFile::open_with_options_v2(
-                &path,
-                virtual_file::OpenOptions::new()
-                    .create_new(true)
-                    .write(true),
-                ctx,
-            )
-            .await?,
-            gate.enter()?,
-        );
 
-        // Start at PAGE_SZ, make room for the header block
-        let blob_writer = BlobWriter::new(
-            file,
-            PAGE_SZ as u64,
-            gate,
-            cancel,
-            ctx,
-            info_span!(parent: None, "delta_layer_writer_flush_task", tenant_id=%tenant_shard_id.tenant_id, shard_id=%tenant_shard_id.shard_slug(), timeline_id=%timeline_id, path = %path),
-        )?;
+        let mut file = VirtualFile::create(&path, ctx).await?;
+        // make room for the header block
+        file.seek(SeekFrom::Start(PAGE_SZ as u64)).await?;
+        let blob_writer = BlobWriter::new(file, PAGE_SZ as u64, gate, cancel, ctx);
 
         // Initialize the b-tree index builder
         let block_buf = BlockBuf::new();
@@ -539,27 +515,34 @@ impl DeltaLayerWriterInner {
         self,
         key_end: Key,
         ctx: &RequestContext,
+    ) -> anyhow::Result<(PersistentLayerDesc, Utf8PathBuf)> {
+        let temp_path = self.path.clone();
+        let result = self.finish0(key_end, ctx).await;
+        if let Err(ref e) = result {
+            tracing::info!(%temp_path, "cleaning up temporary file after error during writing: {e}");
+            if let Err(e) = std::fs::remove_file(&temp_path) {
+                tracing::warn!(error=%e, %temp_path, "error cleaning up temporary layer file after error during writing");
+            }
+        }
+        result
+    }
+
+    async fn finish0(
+        self,
+        key_end: Key,
+        ctx: &RequestContext,
     ) -> anyhow::Result<(PersistentLayerDesc, Utf8PathBuf)> {
         let index_start_blk = self.blob_writer.size().div_ceil(PAGE_SZ as u64) as u32;
 
-        let file = self
-            .blob_writer
-            .shutdown(
-                BufferedWriterShutdownMode::ZeroPadToNextMultiple(PAGE_SZ),
-                ctx,
-            )
-            .await?;
+        let mut file = self.blob_writer.into_inner(ctx).await?;
 
         // Write out the index
         let (index_root_blk, block_buf) = self.tree.finish()?;
-        let mut offset = index_start_blk as u64 * PAGE_SZ as u64;
-
-        // TODO(yuchen): https://github.com/neondatabase/neon/issues/10092
-        // Should we just replace BlockBuf::blocks with one big buffer
+        file.seek(SeekFrom::Start(index_start_blk as u64 * PAGE_SZ as u64))
+            .await?;
         for buf in block_buf.blocks {
-            let (_buf, res) = file.write_all_at(buf.slice_len(), offset, ctx).await;
+            let (_buf, res) = file.write_all(buf.slice_len(), ctx).await;
             res?;
-            offset += PAGE_SZ as u64;
         }
         assert!(self.lsn_range.start < self.lsn_range.end);
         // Fill in the summary on blk 0
@@ -574,9 +557,11 @@ impl DeltaLayerWriterInner {
             index_root_blk,
         };
 
-        // Writes summary at the first block (offset 0).
-        let buf = summary.ser_into_page()?;
-        let (_buf, res) = file.write_all_at(buf.slice_len(), 0, ctx).await;
+        let mut buf = Vec::with_capacity(PAGE_SZ);
+        // TODO: could use smallvec here but it's a pain with Slice<T>
+        Summary::ser_into(&summary, &mut buf)?;
+        file.seek(SeekFrom::Start(0)).await?;
+        let (_buf, res) = file.write_all(buf.slice_len(), ctx).await;
         res?;
 
         let metadata = file
@@ -613,10 +598,6 @@ impl DeltaLayerWriterInner {
 
         trace!("created delta layer {}", self.path);
 
-        // The gate guard stored in `destination_file` is dropped. Callers (e.g.. flush loop or compaction)
-        // keep the gate open also, so that it's safe for them to rename the file to its final destination.
-        file.disarm_into_inner();
-
         Ok((desc, self.path))
     }
 }
@@ -745,6 +726,17 @@ impl DeltaLayerWriter {
     }
 }
 
+impl Drop for DeltaLayerWriter {
+    fn drop(&mut self) {
+        if let Some(inner) = self.inner.take() {
+            // We want to remove the virtual file here, so it's fine to not
+            // having completely flushed unwritten data.
+            let vfile = inner.blob_writer.into_inner_no_flush();
+            vfile.remove();
+        }
+    }
+}
+
 #[derive(thiserror::Error, Debug)]
 pub enum RewriteSummaryError {
     #[error("magic mismatch")]
@@ -768,7 +760,7 @@ impl DeltaLayer {
     where
         F: Fn(Summary) -> Summary,
     {
-        let file = VirtualFile::open_with_options_v2(
+        let mut file = VirtualFile::open_with_options(
             path,
             virtual_file::OpenOptions::new().read(true).write(true),
             ctx,
@@ -785,8 +777,11 @@ impl DeltaLayer {
 
         let new_summary = rewrite(actual_summary);
 
-        let buf = new_summary.ser_into_page().context("serialize")?;
-        let (_buf, res) = file.write_all_at(buf.slice_len(), 0, ctx).await;
+        let mut buf = Vec::with_capacity(PAGE_SZ);
+        // TODO: could use smallvec here, but it's a pain with Slice<T>
+        Summary::ser_into(&new_summary, &mut buf).context("serialize")?;
+        file.seek(SeekFrom::Start(0)).await?;
+        let (_buf, res) = file.write_all(buf.slice_len(), ctx).await;
         res?;
         Ok(())
     }
@@ -1614,8 +1609,8 @@ pub(crate) mod test {
     use bytes::Bytes;
     use itertools::MinMaxResult;
     use pageserver_api::value::Value;
+    use rand::RngCore;
     use rand::prelude::{SeedableRng, SliceRandom, StdRng};
-    use rand::{Rng, RngCore};
 
     use super::*;
     use crate::DEFAULT_PG_VERSION;
@@ -1625,7 +1620,7 @@ pub(crate) mod test {
     use crate::tenant::harness::{TIMELINE_ID, TenantHarness};
     use crate::tenant::storage_layer::{Layer, ResidentLayer};
     use crate::tenant::vectored_blob_io::StreamingVectoredReadPlanner;
-    use crate::tenant::{TenantShard, Timeline};
+    use crate::tenant::{Tenant, Timeline};
 
     /// Construct an index for a fictional delta layer and and then
     /// traverse in order to plan vectored reads for a query. Finally,
@@ -2214,7 +2209,7 @@ pub(crate) mod test {
     }
 
     pub(crate) async fn produce_delta_layer(
-        tenant: &TenantShard,
+        tenant: &Tenant,
         tline: &Arc<Timeline>,
         mut deltas: Vec<(Key, Lsn, Value)>,
         ctx: &RequestContext,

pageserver/src/tenant/storage_layer/image_layer.rs

@@ -27,11 +27,11 @@
 //! actual page images are stored in the "values" part.
 use std::collections::{HashMap, VecDeque};
 use std::fs::File;
+use std::io::SeekFrom;
 use std::ops::Range;
 use std::os::unix::prelude::FileExt;
 use std::str::FromStr;
 use std::sync::Arc;
-use std::sync::atomic::AtomicU64;
 
 use anyhow::{Context, Result, bail, ensure};
 use bytes::Bytes;
@@ -43,13 +43,14 @@ use pageserver_api::key::{DBDIR_KEY, KEY_SIZE, Key};
 use pageserver_api::keyspace::KeySpace;
 use pageserver_api::shard::{ShardIdentity, TenantShardId};
 use pageserver_api::value::Value;
+use rand::Rng;
+use rand::distributions::Alphanumeric;
 use serde::{Deserialize, Serialize};
 use tokio::sync::OnceCell;
 use tokio_stream::StreamExt;
 use tokio_util::sync::CancellationToken;
 use tracing::*;
 use utils::bin_ser::BeSer;
-use utils::bin_ser::SerializeError;
 use utils::id::{TenantId, TimelineId};
 use utils::lsn::Lsn;
 
@@ -71,10 +72,8 @@ use crate::tenant::vectored_blob_io::{
     BlobFlag, BufView, StreamingVectoredReadPlanner, VectoredBlobReader, VectoredRead,
     VectoredReadPlanner,
 };
-use crate::virtual_file::TempVirtualFile;
 use crate::virtual_file::owned_buffers_io::io_buf_ext::IoBufExt;
-use crate::virtual_file::owned_buffers_io::write::{Buffer, BufferedWriterShutdownMode};
-use crate::virtual_file::{self, IoBuffer, IoBufferMut, MaybeFatalIo, VirtualFile};
+use crate::virtual_file::{self, IoBufferMut, MaybeFatalIo, VirtualFile};
 use crate::{IMAGE_FILE_MAGIC, STORAGE_FORMAT_VERSION, TEMP_FILE_SUFFIX};
 
 ///
@@ -113,15 +112,6 @@ impl From<&ImageLayer> for Summary {
 }
 
 impl Summary {
-    /// Serializes the summary header into an aligned buffer of lenth `PAGE_SZ`.
-    pub fn ser_into_page(&self) -> Result<IoBuffer, SerializeError> {
-        let mut buf = IoBufferMut::with_capacity(PAGE_SZ);
-        Self::ser_into(self, &mut buf)?;
-        // Pad zeroes to the buffer so the length is a multiple of the alignment.
-        buf.extend_with(0, buf.capacity() - buf.len());
-        Ok(buf.freeze())
-    }
-
     pub(super) fn expected(
         tenant_id: TenantId,
         timeline_id: TimelineId,
@@ -262,18 +252,14 @@ impl ImageLayer {
         tenant_shard_id: TenantShardId,
         fname: &ImageLayerName,
     ) -> Utf8PathBuf {
-        // TempVirtualFile requires us to never reuse a filename while an old
-        // instance of TempVirtualFile created with that filename is not done dropping yet.
-        // So, we use a monotonic counter to disambiguate the filenames.
-        static NEXT_TEMP_DISAMBIGUATOR: AtomicU64 = AtomicU64::new(1);
-        let filename_disambiguator =
-            NEXT_TEMP_DISAMBIGUATOR.fetch_add(1, std::sync::atomic::Ordering::Relaxed);
+        let rand_string: String = rand::thread_rng()
+            .sample_iter(&Alphanumeric)
+            .take(8)
+            .map(char::from)
+            .collect();
 
         conf.timeline_path(&tenant_shard_id, &timeline_id)
-            .join(format!(
-                "{fname}.{:x}.{TEMP_FILE_SUFFIX}",
-                filename_disambiguator
-            ))
+            .join(format!("{fname}.{rand_string}.{TEMP_FILE_SUFFIX}"))
     }
 
     ///
@@ -363,7 +349,7 @@ impl ImageLayer {
     where
         F: Fn(Summary) -> Summary,
     {
-        let file = VirtualFile::open_with_options_v2(
+        let mut file = VirtualFile::open_with_options(
             path,
             virtual_file::OpenOptions::new().read(true).write(true),
             ctx,
@@ -380,8 +366,11 @@ impl ImageLayer {
 
         let new_summary = rewrite(actual_summary);
 
-        let buf = new_summary.ser_into_page().context("serialize")?;
-        let (_buf, res) = file.write_all_at(buf.slice_len(), 0, ctx).await;
+        let mut buf = Vec::with_capacity(PAGE_SZ);
+        // TODO: could use smallvec here but it's a pain with Slice<T>
+        Summary::ser_into(&new_summary, &mut buf).context("serialize")?;
+        file.seek(SeekFrom::Start(0)).await?;
+        let (_buf, res) = file.write_all(buf.slice_len(), ctx).await;
         res?;
         Ok(())
     }
@@ -570,12 +559,11 @@ impl ImageLayerInner {
             let view = BufView::new_slice(&blobs_buf.buf);
 
             for meta in blobs_buf.blobs.iter() {
-                // Just read the raw header+data and pass it through to the target layer, without
-                // decoding and recompressing it.
-                let raw = meta.raw_with_header(&view);
+                let img_buf = meta.read(&view).await?;
+
                 key_count += 1;
                 writer
-                    .put_image_raw(meta.meta.key, raw.into_bytes(), ctx)
+                    .put_image(meta.meta.key, img_buf.into_bytes(), ctx)
                     .await
                     .context(format!("Storing key {}", meta.meta.key))?;
             }
@@ -750,7 +738,7 @@ struct ImageLayerWriterInner {
     // Number of keys in the layer.
     num_keys: usize,
 
-    blob_writer: BlobWriter<TempVirtualFile>,
+    blob_writer: BlobWriter<false>,
     tree: DiskBtreeBuilder<BlockBuf, KEY_SIZE>,
 
     #[cfg(feature = "testing")]
@@ -784,27 +772,19 @@ impl ImageLayerWriterInner {
             },
         );
         trace!("creating image layer {}", path);
-        let file = TempVirtualFile::new(
-            VirtualFile::open_with_options_v2(
+        let mut file = {
+            VirtualFile::open_with_options(
                 &path,
                 virtual_file::OpenOptions::new()
-                    .create_new(true)
-                    .write(true),
+                    .write(true)
+                    .create_new(true),
                 ctx,
             )
-            .await?,
-            gate.enter()?,
-        );
-
-        // Start at `PAGE_SZ` to make room for the header block.
-        let blob_writer = BlobWriter::new(
-            file,
-            PAGE_SZ as u64,
-            gate,
-            cancel,
-            ctx,
-            info_span!(parent: None, "image_layer_writer_flush_task", tenant_id=%tenant_shard_id.tenant_id, shard_id=%tenant_shard_id.shard_slug(), timeline_id=%timeline_id, path = %path),
-        )?;
+            .await?
+        };
+        // make room for the header block
+        file.seek(SeekFrom::Start(PAGE_SZ as u64)).await?;
+        let blob_writer = BlobWriter::new(file, PAGE_SZ as u64, gate, cancel, ctx);
 
         // Initialize the b-tree index builder
         let block_buf = BlockBuf::new();
@@ -874,44 +854,28 @@ impl ImageLayerWriterInner {
     }
 
     ///
-    /// Write the next image to the file, as a raw blob header and data.
+    /// Finish writing the image layer.
     ///
-    /// The page versions must be appended in blknum order.
-    ///
-    async fn put_image_raw(
-        &mut self,
-        key: Key,
-        raw_with_header: Bytes,
+    async fn finish(
+        self,
         ctx: &RequestContext,
-    ) -> anyhow::Result<()> {
-        ensure!(self.key_range.contains(&key));
-
-        // NB: we don't update the (un)compressed metrics, since we can't determine them without
-        // decompressing the image. This seems okay.
-        self.num_keys += 1;
-
-        let (_, res) = self
-            .blob_writer
-            .write_blob_raw(raw_with_header.slice_len(), ctx)
-            .await;
-        let offset = res?;
-
-        let mut keybuf: [u8; KEY_SIZE] = [0u8; KEY_SIZE];
-        key.write_to_byte_slice(&mut keybuf);
-        self.tree.append(&keybuf, offset)?;
-
-        #[cfg(feature = "testing")]
-        {
-            self.last_written_key = key;
+        end_key: Option<Key>,
+    ) -> anyhow::Result<(PersistentLayerDesc, Utf8PathBuf)> {
+        let temp_path = self.path.clone();
+        let result = self.finish0(ctx, end_key).await;
+        if let Err(ref e) = result {
+            tracing::info!(%temp_path, "cleaning up temporary file after error during writing: {e}");
+            if let Err(e) = std::fs::remove_file(&temp_path) {
+                tracing::warn!(error=%e, %temp_path, "error cleaning up temporary layer file after error during writing");
+            }
         }
-
-        Ok(())
+        result
     }
 
     ///
     /// Finish writing the image layer.
     ///
-    async fn finish(
+    async fn finish0(
         self,
         ctx: &RequestContext,
         end_key: Option<Key>,
@@ -924,32 +888,17 @@ impl ImageLayerWriterInner {
         crate::metrics::COMPRESSION_IMAGE_INPUT_BYTES_CONSIDERED
             .inc_by(self.uncompressed_bytes_eligible);
         crate::metrics::COMPRESSION_IMAGE_INPUT_BYTES_CHOSEN.inc_by(self.uncompressed_bytes_chosen);
+        crate::metrics::COMPRESSION_IMAGE_OUTPUT_BYTES.inc_by(compressed_size);
 
-        // NB: filter() may pass through raw pages from a different layer, without looking at
-        // whether these are compressed or not. We don't track metrics for these, so avoid
-        // increasing `COMPRESSION_IMAGE_OUTPUT_BYTES` in this case too.
-        if self.uncompressed_bytes > 0 {
-            crate::metrics::COMPRESSION_IMAGE_OUTPUT_BYTES.inc_by(compressed_size);
-        };
-
-        let file = self
-            .blob_writer
-            .shutdown(
-                BufferedWriterShutdownMode::ZeroPadToNextMultiple(PAGE_SZ),
-                ctx,
-            )
-            .await?;
+        let mut file = self.blob_writer.into_inner();
 
         // Write out the index
-        let mut offset = index_start_blk as u64 * PAGE_SZ as u64;
+        file.seek(SeekFrom::Start(index_start_blk as u64 * PAGE_SZ as u64))
+            .await?;
         let (index_root_blk, block_buf) = self.tree.finish()?;
-
-        // TODO(yuchen): https://github.com/neondatabase/neon/issues/10092
-        // Should we just replace BlockBuf::blocks with one big buffer?
         for buf in block_buf.blocks {
-            let (_buf, res) = file.write_all_at(buf.slice_len(), offset, ctx).await;
+            let (_buf, res) = file.write_all(buf.slice_len(), ctx).await;
             res?;
-            offset += PAGE_SZ as u64;
         }
 
         let final_key_range = if let Some(end_key) = end_key {
@@ -970,9 +919,11 @@ impl ImageLayerWriterInner {
             index_root_blk,
         };
 
-        // Writes summary at the first block (offset 0).
-        let buf = summary.ser_into_page()?;
-        let (_buf, res) = file.write_all_at(buf.slice_len(), 0, ctx).await;
+        let mut buf = Vec::with_capacity(PAGE_SZ);
+        // TODO: could use smallvec here but it's a pain with Slice<T>
+        Summary::ser_into(&summary, &mut buf)?;
+        file.seek(SeekFrom::Start(0)).await?;
+        let (_buf, res) = file.write_all(buf.slice_len(), ctx).await;
         res?;
 
         let metadata = file
@@ -1007,10 +958,6 @@ impl ImageLayerWriterInner {
 
         trace!("created image layer {}", self.path);
 
-        // The gate guard stored in `destination_file` is dropped. Callers (e.g.. flush loop or compaction)
-        // keep the gate open also, so that it's safe for them to rename the file to its final destination.
-        file.disarm_into_inner();
-
         Ok((desc, self.path))
     }
 }
@@ -1087,25 +1034,6 @@ impl ImageLayerWriter {
         self.inner.as_mut().unwrap().put_image(key, img, ctx).await
     }
 
-    ///
-    /// Write the next value to the file, as a raw header and data. This allows passing through a
-    /// raw, potentially compressed image from a different layer file without recompressing it.
-    ///
-    /// The page versions must be appended in blknum order.
-    ///
-    pub async fn put_image_raw(
-        &mut self,
-        key: Key,
-        raw_with_header: Bytes,
-        ctx: &RequestContext,
-    ) -> anyhow::Result<()> {
-        self.inner
-            .as_mut()
-            .unwrap()
-            .put_image_raw(key, raw_with_header, ctx)
-            .await
-    }
-
     /// Estimated size of the image layer.
     pub(crate) fn estimated_size(&self) -> u64 {
         let inner = self.inner.as_ref().unwrap();
@@ -1136,6 +1064,14 @@ impl ImageLayerWriter {
     }
 }
 
+impl Drop for ImageLayerWriter {
+    fn drop(&mut self) {
+        if let Some(inner) = self.inner.take() {
+            inner.blob_writer.into_inner().remove();
+        }
+    }
+}
+
 pub struct ImageLayerIterator<'a> {
     image_layer: &'a ImageLayerInner,
     ctx: &'a RequestContext,
@@ -1231,7 +1167,7 @@ mod test {
     use crate::tenant::harness::{TIMELINE_ID, TenantHarness};
     use crate::tenant::storage_layer::{Layer, ResidentLayer};
     use crate::tenant::vectored_blob_io::StreamingVectoredReadPlanner;
-    use crate::tenant::{TenantShard, Timeline};
+    use crate::tenant::{Tenant, Timeline};
 
     #[tokio::test]
     async fn image_layer_rewrite() {
@@ -1413,7 +1349,7 @@ mod test {
     }
 
     async fn produce_image_layer(
-        tenant: &TenantShard,
+        tenant: &Tenant,
         tline: &Arc<Timeline>,
         mut images: Vec<(Key, Bytes)>,
         lsn: Lsn,

pageserver/src/tenant/tasks.rs

@@ -24,7 +24,7 @@ use crate::task_mgr::{self, BACKGROUND_RUNTIME, TOKIO_WORKER_THREADS, TaskKind};
 use crate::tenant::throttle::Stats;
 use crate::tenant::timeline::CompactionError;
 use crate::tenant::timeline::compaction::CompactionOutcome;
-use crate::tenant::{TenantShard, TenantState};
+use crate::tenant::{Tenant, TenantState};
 
 /// Semaphore limiting concurrent background tasks (across all tenants).
 ///
@@ -117,7 +117,7 @@ pub(crate) async fn acquire_concurrency_permit(
 }
 
 /// Start per tenant background loops: compaction, GC, and ingest housekeeping.
-pub fn start_background_loops(tenant: &Arc<TenantShard>, can_start: Option<&Barrier>) {
+pub fn start_background_loops(tenant: &Arc<Tenant>, can_start: Option<&Barrier>) {
     let tenant_shard_id = tenant.tenant_shard_id;
 
     task_mgr::spawn(
@@ -198,7 +198,7 @@ pub fn start_background_loops(tenant: &Arc<TenantShard>, can_start: Option<&Barr
 }
 
 /// Compaction task's main loop.
-async fn compaction_loop(tenant: Arc<TenantShard>, cancel: CancellationToken) {
+async fn compaction_loop(tenant: Arc<Tenant>, cancel: CancellationToken) {
     const BASE_BACKOFF_SECS: f64 = 1.0;
     const MAX_BACKOFF_SECS: f64 = 300.0;
     const RECHECK_CONFIG_INTERVAL: Duration = Duration::from_secs(10);
@@ -348,7 +348,7 @@ pub(crate) fn log_compaction_error(
 }
 
 /// GC task's main loop.
-async fn gc_loop(tenant: Arc<TenantShard>, cancel: CancellationToken) {
+async fn gc_loop(tenant: Arc<Tenant>, cancel: CancellationToken) {
     const MAX_BACKOFF_SECS: f64 = 300.0;
     let mut error_run = 0; // consecutive errors
 
@@ -432,7 +432,7 @@ async fn gc_loop(tenant: Arc<TenantShard>, cancel: CancellationToken) {
 }
 
 /// Tenant housekeeping's main loop.
-async fn tenant_housekeeping_loop(tenant: Arc<TenantShard>, cancel: CancellationToken) {
+async fn tenant_housekeeping_loop(tenant: Arc<Tenant>, cancel: CancellationToken) {
     let mut last_throttle_flag_reset_at = Instant::now();
     loop {
         if wait_for_active_tenant(&tenant, &cancel).await.is_break() {
@@ -483,7 +483,7 @@ async fn tenant_housekeeping_loop(tenant: Arc<TenantShard>, cancel: Cancellation
 
 /// Waits until the tenant becomes active, or returns `ControlFlow::Break()` to shut down.
 async fn wait_for_active_tenant(
-    tenant: &Arc<TenantShard>,
+    tenant: &Arc<Tenant>,
     cancel: &CancellationToken,
 ) -> ControlFlow<()> {
     if tenant.current_state() == TenantState::Active {

pageserver/src/tenant/timeline.rs

@@ -412,7 +412,7 @@ pub struct Timeline {
     /// Timeline deletion will acquire both compaction and gc locks in whatever order.
     gc_lock: tokio::sync::Mutex<()>,
 
-    /// Cloned from [`super::TenantShard::pagestream_throttle`] on construction.
+    /// Cloned from [`super::Tenant::pagestream_throttle`] on construction.
     pub(crate) pagestream_throttle: Arc<crate::tenant::throttle::Throttle>,
 
     /// Size estimator for aux file v2
@@ -1285,10 +1285,6 @@ impl Timeline {
         reconstruct_state: &mut ValuesReconstructState,
         ctx: &RequestContext,
     ) -> Result<BTreeMap<Key, Result<Bytes, PageReconstructError>>, GetVectoredError> {
-        if query.is_empty() {
-            return Ok(BTreeMap::default());
-        }
-
         let read_path = if self.conf.enable_read_path_debugging || ctx.read_path_debug() {
             Some(ReadPath::new(
                 query.total_keyspace(),
@@ -2069,7 +2065,7 @@ impl Timeline {
 
     pub(crate) fn activate(
         self: &Arc<Self>,
-        parent: Arc<crate::tenant::TenantShard>,
+        parent: Arc<crate::tenant::Tenant>,
         broker_client: BrokerClientChannel,
         background_jobs_can_start: Option<&completion::Barrier>,
         ctx: &RequestContext,
@@ -2706,14 +2702,6 @@ impl Timeline {
             .clone()
     }
 
-    pub fn get_compaction_shard_ancestor(&self) -> bool {
-        let tenant_conf = self.tenant_conf.load();
-        tenant_conf
-            .tenant_conf
-            .compaction_shard_ancestor
-            .unwrap_or(self.conf.default_tenant_conf.compaction_shard_ancestor)
-    }
-
     fn get_eviction_policy(&self) -> EvictionPolicy {
         let tenant_conf = self.tenant_conf.load();
         tenant_conf
@@ -3329,7 +3317,7 @@ impl Timeline {
         //     (1) and (4)
         // TODO: this is basically a no-op now, should we remove it?
         self.remote_client.schedule_barrier()?;
-        // TenantShard::create_timeline will wait for these uploads to happen before returning, or
+        // Tenant::create_timeline will wait for these uploads to happen before returning, or
         // on retry.
 
         // Now that we have the full layer map, we may calculate the visibility of layers within it (a global scan)
@@ -5758,7 +5746,7 @@ impl Timeline {
     /// from our ancestor to be branches of this timeline.
     pub(crate) async fn prepare_to_detach_from_ancestor(
         self: &Arc<Timeline>,
-        tenant: &crate::tenant::TenantShard,
+        tenant: &crate::tenant::Tenant,
         options: detach_ancestor::Options,
         behavior: DetachBehavior,
         ctx: &RequestContext,
@@ -5777,7 +5765,7 @@ impl Timeline {
     /// resetting the tenant.
     pub(crate) async fn detach_from_ancestor_and_reparent(
         self: &Arc<Timeline>,
-        tenant: &crate::tenant::TenantShard,
+        tenant: &crate::tenant::Tenant,
         prepared: detach_ancestor::PreparedTimelineDetach,
         ancestor_timeline_id: TimelineId,
         ancestor_lsn: Lsn,
@@ -5801,7 +5789,7 @@ impl Timeline {
     /// The tenant must've been reset if ancestry was modified previously (in tenant manager).
     pub(crate) async fn complete_detaching_timeline_ancestor(
         self: &Arc<Timeline>,
-        tenant: &crate::tenant::TenantShard,
+        tenant: &crate::tenant::Tenant,
         attempt: detach_ancestor::Attempt,
         ctx: &RequestContext,
     ) -> Result<(), detach_ancestor::Error> {
@@ -6863,14 +6851,14 @@ impl Timeline {
     /// Persistently blocks gc for `Manual` reason.
     ///
     /// Returns true if no such block existed before, false otherwise.
-    pub(crate) async fn block_gc(&self, tenant: &super::TenantShard) -> anyhow::Result<bool> {
+    pub(crate) async fn block_gc(&self, tenant: &super::Tenant) -> anyhow::Result<bool> {
         use crate::tenant::remote_timeline_client::index::GcBlockingReason;
         assert_eq!(self.tenant_shard_id, tenant.tenant_shard_id);
         tenant.gc_block.insert(self, GcBlockingReason::Manual).await
     }
 
     /// Persistently unblocks gc for `Manual` reason.
-    pub(crate) async fn unblock_gc(&self, tenant: &super::TenantShard) -> anyhow::Result<()> {
+    pub(crate) async fn unblock_gc(&self, tenant: &super::Tenant) -> anyhow::Result<()> {
         use crate::tenant::remote_timeline_client::index::GcBlockingReason;
         assert_eq!(self.tenant_shard_id, tenant.tenant_shard_id);
         tenant.gc_block.remove(self, GcBlockingReason::Manual).await
@@ -6888,8 +6876,8 @@ impl Timeline {
 
     /// Force create an image layer and place it into the layer map.
     ///
-    /// DO NOT use this function directly. Use [`TenantShard::branch_timeline_test_with_layers`]
-    /// or [`TenantShard::create_test_timeline_with_layers`] to ensure all these layers are
+    /// DO NOT use this function directly. Use [`Tenant::branch_timeline_test_with_layers`]
+    /// or [`Tenant::create_test_timeline_with_layers`] to ensure all these layers are
     /// placed into the layer map in one run AND be validated.
     #[cfg(test)]
     pub(super) async fn force_create_image_layer(
@@ -6945,8 +6933,8 @@ impl Timeline {
 
     /// Force create a delta layer and place it into the layer map.
     ///
-    /// DO NOT use this function directly. Use [`TenantShard::branch_timeline_test_with_layers`]
-    /// or [`TenantShard::create_test_timeline_with_layers`] to ensure all these layers are
+    /// DO NOT use this function directly. Use [`Tenant::branch_timeline_test_with_layers`]
+    /// or [`Tenant::create_test_timeline_with_layers`] to ensure all these layers are
     /// placed into the layer map in one run AND be validated.
     #[cfg(test)]
     pub(super) async fn force_create_delta_layer(

pageserver/src/tenant/timeline/compaction.rs

@@ -77,7 +77,7 @@ const COMPACTION_DELTA_THRESHOLD: usize = 5;
 /// shard split, which gets expensive for large tenants.
 const ANCESTOR_COMPACTION_REWRITE_THRESHOLD: f64 = 0.3;
 
-#[derive(Default, Debug, Clone, Copy, Hash, PartialEq, Eq, Serialize)]
+#[derive(Debug, Clone, Copy, Hash, PartialEq, Eq)]
 pub struct GcCompactionJobId(pub usize);
 
 impl std::fmt::Display for GcCompactionJobId {
@@ -105,50 +105,6 @@ pub enum GcCompactionQueueItem {
     Notify(GcCompactionJobId, Option<Lsn>),
 }
 
-/// Statistics for gc-compaction meta jobs, which contains several sub compaction jobs.
-#[derive(Debug, Clone, Serialize, Default)]
-pub struct GcCompactionMetaStatistics {
-    /// The total number of sub compaction jobs.
-    pub total_sub_compaction_jobs: usize,
-    /// The total number of sub compaction jobs that failed.
-    pub failed_sub_compaction_jobs: usize,
-    /// The total number of sub compaction jobs that succeeded.
-    pub succeeded_sub_compaction_jobs: usize,
-    /// The layer size before compaction.
-    pub before_compaction_layer_size: u64,
-    /// The layer size after compaction.
-    pub after_compaction_layer_size: u64,
-    /// The start time of the meta job.
-    pub start_time: Option<chrono::DateTime<chrono::Utc>>,
-    /// The end time of the meta job.
-    pub end_time: Option<chrono::DateTime<chrono::Utc>>,
-    /// The duration of the meta job.
-    pub duration_secs: f64,
-    /// The id of the meta job.
-    pub meta_job_id: GcCompactionJobId,
-    /// The LSN below which the layers are compacted, used to compute the statistics.
-    pub below_lsn: Lsn,
-    /// The retention ratio of the meta job (after_compaction_layer_size / before_compaction_layer_size)
-    pub retention_ratio: f64,
-}
-
-impl GcCompactionMetaStatistics {
-    fn finalize(&mut self) {
-        let end_time = chrono::Utc::now();
-        if let Some(start_time) = self.start_time {
-            if end_time > start_time {
-                let delta = end_time - start_time;
-                if let Ok(std_dur) = delta.to_std() {
-                    self.duration_secs = std_dur.as_secs_f64();
-                }
-            }
-        }
-        self.retention_ratio = self.after_compaction_layer_size as f64
-            / (self.before_compaction_layer_size as f64 + 1.0);
-        self.end_time = Some(end_time);
-    }
-}
-
 impl GcCompactionQueueItem {
     pub fn into_compact_info_resp(
         self,
@@ -186,7 +142,6 @@ struct GcCompactionQueueInner {
     queued: VecDeque<(GcCompactionJobId, GcCompactionQueueItem)>,
     guards: HashMap<GcCompactionJobId, GcCompactionGuardItems>,
     last_id: GcCompactionJobId,
-    meta_statistics: Option<GcCompactionMetaStatistics>,
 }
 
 impl GcCompactionQueueInner {
@@ -218,7 +173,6 @@ impl GcCompactionQueue {
                 queued: VecDeque::new(),
                 guards: HashMap::new(),
                 last_id: GcCompactionJobId(0),
-                meta_statistics: None,
             }),
             consumer_lock: tokio::sync::Mutex::new(()),
         }
@@ -403,23 +357,6 @@ impl GcCompactionQueue {
         Ok(())
     }
 
-    async fn collect_layer_below_lsn(
-        &self,
-        timeline: &Arc<Timeline>,
-        lsn: Lsn,
-    ) -> Result<u64, CompactionError> {
-        let guard = timeline.layers.read().await;
-        let layer_map = guard.layer_map()?;
-        let layers = layer_map.iter_historic_layers().collect_vec();
-        let mut size = 0;
-        for layer in layers {
-            if layer.lsn_range.start <= lsn {
-                size += layer.file_size();
-            }
-        }
-        Ok(size)
-    }
-
     /// Notify the caller the job has finished and unblock GC.
     fn notify_and_unblock(&self, id: GcCompactionJobId) {
         info!("compaction job id={} finished", id);
@@ -429,16 +366,6 @@ impl GcCompactionQueue {
                 let _ = tx.send(());
             }
         }
-        if let Some(ref meta_statistics) = guard.meta_statistics {
-            if meta_statistics.meta_job_id == id {
-                if let Ok(stats) = serde_json::to_string(&meta_statistics) {
-                    info!(
-                        "gc-compaction meta statistics for job id = {}: {}",
-                        id, stats
-                    );
-                }
-            }
-        }
     }
 
     fn clear_running_job(&self) {
@@ -478,11 +405,7 @@ impl GcCompactionQueue {
             let mut pending_tasks = Vec::new();
             // gc-compaction might pick more layers or fewer layers to compact. The L2 LSN does not need to be accurate.
             // And therefore, we simply assume the maximum LSN of all jobs is the expected L2 LSN.
-            let expected_l2_lsn = jobs
-                .iter()
-                .map(|job| job.compact_lsn_range.end)
-                .max()
-                .unwrap();
+            let expected_l2_lsn = jobs.iter().map(|job| job.compact_lsn_range.end).max();
             for job in jobs {
                 // Unfortunately we need to convert the `GcCompactJob` back to `CompactionOptions`
                 // until we do further refactors to allow directly call `compact_with_gc`.
@@ -507,13 +430,9 @@ impl GcCompactionQueue {
             if !auto {
                 pending_tasks.push(GcCompactionQueueItem::Notify(id, None));
             } else {
-                pending_tasks.push(GcCompactionQueueItem::Notify(id, Some(expected_l2_lsn)));
+                pending_tasks.push(GcCompactionQueueItem::Notify(id, expected_l2_lsn));
             }
 
-            let layer_size = self
-                .collect_layer_below_lsn(timeline, expected_l2_lsn)
-                .await?;
-
             {
                 let mut guard = self.inner.lock().unwrap();
                 let mut tasks = Vec::new();
@@ -525,16 +444,7 @@ impl GcCompactionQueue {
                 for item in tasks {
                     guard.queued.push_front(item);
                 }
-                guard.meta_statistics = Some(GcCompactionMetaStatistics {
-                    meta_job_id: id,
-                    start_time: Some(chrono::Utc::now()),
-                    before_compaction_layer_size: layer_size,
-                    below_lsn: expected_l2_lsn,
-                    total_sub_compaction_jobs: jobs_len,
-                    ..Default::default()
-                });
             }
-
             info!(
                 "scheduled enhanced gc bottom-most compaction with sub-compaction, split into {} jobs",
                 jobs_len
@@ -663,10 +573,6 @@ impl GcCompactionQueue {
                     Err(err) => {
                         warn!(%err, "failed to run gc-compaction subcompaction job");
                         self.clear_running_job();
-                        let mut guard = self.inner.lock().unwrap();
-                        if let Some(ref mut meta_statistics) = guard.meta_statistics {
-                            meta_statistics.failed_sub_compaction_jobs += 1;
-                        }
                         return Err(err);
                     }
                 };
@@ -676,34 +582,8 @@ impl GcCompactionQueue {
                     // we need to clean things up before returning from the function.
                     yield_for_l0 = true;
                 }
-                {
-                    let mut guard = self.inner.lock().unwrap();
-                    if let Some(ref mut meta_statistics) = guard.meta_statistics {
-                        meta_statistics.succeeded_sub_compaction_jobs += 1;
-                    }
-                }
             }
             GcCompactionQueueItem::Notify(id, l2_lsn) => {
-                let below_lsn = {
-                    let mut guard = self.inner.lock().unwrap();
-                    if let Some(ref mut meta_statistics) = guard.meta_statistics {
-                        meta_statistics.below_lsn
-                    } else {
-                        Lsn::INVALID
-                    }
-                };
-                let layer_size = if below_lsn != Lsn::INVALID {
-                    self.collect_layer_below_lsn(timeline, below_lsn).await?
-                } else {
-                    0
-                };
-                {
-                    let mut guard = self.inner.lock().unwrap();
-                    if let Some(ref mut meta_statistics) = guard.meta_statistics {
-                        meta_statistics.after_compaction_layer_size = layer_size;
-                        meta_statistics.finalize();
-                    }
-                }
                 self.notify_and_unblock(id);
                 if let Some(l2_lsn) = l2_lsn {
                     let current_l2_lsn = timeline
@@ -1359,7 +1239,8 @@ impl Timeline {
         let partition_count = self.partitioning.read().0.0.parts.len();
 
         // 4. Shard ancestor compaction
-        if self.get_compaction_shard_ancestor() && self.shard_identity.count >= ShardCount::new(2) {
+
+        if self.shard_identity.count >= ShardCount::new(2) {
             // Limit the number of layer rewrites to the number of partitions: this means its
             // runtime should be comparable to a full round of image layer creations, rather than
             // being potentially much longer.

pageserver/src/tenant/timeline/delete.rs

@@ -18,8 +18,8 @@ use crate::tenant::remote_timeline_client::{
     PersistIndexPartWithDeletedFlagError, RemoteTimelineClient,
 };
 use crate::tenant::{
-    CreateTimelineCause, DeleteTimelineError, MaybeDeletedIndexPart, TenantManifestError,
-    TenantShard, Timeline, TimelineOrOffloaded,
+    CreateTimelineCause, DeleteTimelineError, MaybeDeletedIndexPart, Tenant, TenantManifestError,
+    Timeline, TimelineOrOffloaded,
 };
 use crate::virtual_file::MaybeFatalIo;
 
@@ -113,7 +113,7 @@ pub(super) async fn delete_local_timeline_directory(
 /// It is important that this gets called when DeletionGuard is being held.
 /// For more context see comments in [`make_timeline_delete_guard`]
 async fn remove_maybe_offloaded_timeline_from_tenant(
-    tenant: &TenantShard,
+    tenant: &Tenant,
     timeline: &TimelineOrOffloaded,
     _: &DeletionGuard, // using it as a witness
 ) -> anyhow::Result<()> {
@@ -192,7 +192,7 @@ impl DeleteTimelineFlow {
     // error out if some of the shutdown tasks have already been completed!
     #[instrument(skip_all)]
     pub async fn run(
-        tenant: &Arc<TenantShard>,
+        tenant: &Arc<Tenant>,
         timeline_id: TimelineId,
     ) -> Result<(), DeleteTimelineError> {
         super::debug_assert_current_span_has_tenant_and_timeline_id();
@@ -288,7 +288,7 @@ impl DeleteTimelineFlow {
     /// Shortcut to create Timeline in stopping state and spawn deletion task.
     #[instrument(skip_all, fields(%timeline_id))]
     pub(crate) async fn resume_deletion(
-        tenant: Arc<TenantShard>,
+        tenant: Arc<Tenant>,
         timeline_id: TimelineId,
         local_metadata: &TimelineMetadata,
         remote_client: RemoteTimelineClient,
@@ -338,7 +338,7 @@ impl DeleteTimelineFlow {
     fn schedule_background(
         guard: DeletionGuard,
         conf: &'static PageServerConf,
-        tenant: Arc<TenantShard>,
+        tenant: Arc<Tenant>,
         timeline: TimelineOrOffloaded,
         remote_client: Arc<RemoteTimelineClient>,
     ) {
@@ -381,7 +381,7 @@ impl DeleteTimelineFlow {
     async fn background(
         mut guard: DeletionGuard,
         conf: &PageServerConf,
-        tenant: &TenantShard,
+        tenant: &Tenant,
         timeline: &TimelineOrOffloaded,
         remote_client: Arc<RemoteTimelineClient>,
     ) -> Result<(), DeleteTimelineError> {
@@ -435,7 +435,7 @@ pub(super) enum TimelineDeleteGuardKind {
 }
 
 pub(super) fn make_timeline_delete_guard(
-    tenant: &TenantShard,
+    tenant: &Tenant,
     timeline_id: TimelineId,
     guard_kind: TimelineDeleteGuardKind,
 ) -> Result<(TimelineOrOffloaded, DeletionGuard), DeleteTimelineError> {