cleanup logging

- move extension download to a separate thread;
- add timer around shared preload libraries downloading

.github/workflows/benchmarking.yml

@@ -180,7 +180,8 @@ jobs:
       image: 369495373322.dkr.ecr.eu-central-1.amazonaws.com/rust:pinned
       options: --init
 
-    timeout-minutes: 360 # 6h
+    # Increase timeout to 8h, default timeout is 6h
+    timeout-minutes: 480
 
     steps:
     - uses: actions/checkout@v3
@@ -321,8 +322,6 @@ jobs:
       image: 369495373322.dkr.ecr.eu-central-1.amazonaws.com/rust:pinned
       options: --init
 
-    timeout-minutes: 360 # 6h
-
     steps:
     - uses: actions/checkout@v3
 
@@ -414,8 +413,6 @@ jobs:
       image: 369495373322.dkr.ecr.eu-central-1.amazonaws.com/rust:pinned
       options: --init
 
-    timeout-minutes: 360 # 6h
-
     steps:
     - uses: actions/checkout@v3
 
@@ -501,8 +498,6 @@ jobs:
       image: 369495373322.dkr.ecr.eu-central-1.amazonaws.com/rust:pinned
       options: --init
 
-    timeout-minutes: 360 # 6h
-
     steps:
     - uses: actions/checkout@v3
 

.github/workflows/build_and_test.yml

@@ -722,6 +722,35 @@ jobs:
                            --dockerfile Dockerfile.compute-node
                            --destination 369495373322.dkr.ecr.eu-central-1.amazonaws.com/compute-node-${{ matrix.version }}:${{needs.tag.outputs.build-tag}}
                            --destination neondatabase/compute-node-${{ matrix.version }}:${{needs.tag.outputs.build-tag}}
+                           --cleanup
+
+      # Due to a kaniko bug, we can't use cache for extensions image, thus it takes about the same amount of time as compute-node image to build (~10 min)
+      # During the transition period we need to have extensions in both places (in S3 and in compute-node image),
+      # so we won't build extension twice, but extract them from compute-node.
+      #
+      # For now we use extensions image only for new custom extensitons
+      - name: Kaniko build extensions only
+        run: |
+          # Kaniko is suposed to clean up after itself if --cleanup flag is set, but it doesn't.
+          # Despite some fixes were made in https://github.com/GoogleContainerTools/kaniko/pull/2504 (in kaniko v1.11.0),
+          # it still fails with error:
+          #   error building image: could not save file: copying file: symlink postgres /kaniko/1/usr/local/pgsql/bin/postmaster: file exists
+          #
+          # Ref https://github.com/GoogleContainerTools/kaniko/issues/1406
+          find /kaniko -maxdepth 1 -mindepth 1 -type d -regex "/kaniko/[0-9]*" -exec rm -rv {} \;
+
+          /kaniko/executor --reproducible --snapshot-mode=redo --skip-unused-stages --cache=true \
+                           --cache-repo 369495373322.dkr.ecr.eu-central-1.amazonaws.com/cache \
+                           --context . \
+                           --build-arg GIT_VERSION=${{ github.sha }} \
+                           --build-arg PG_VERSION=${{ matrix.version }} \
+                           --build-arg BUILD_TAG=${{needs.tag.outputs.build-tag}} \
+                           --build-arg REPOSITORY=369495373322.dkr.ecr.eu-central-1.amazonaws.com \
+                           --dockerfile Dockerfile.compute-node \
+                           --destination 369495373322.dkr.ecr.eu-central-1.amazonaws.com/extensions-${{ matrix.version }}:${{needs.tag.outputs.build-tag}} \
+                           --destination neondatabase/extensions-${{ matrix.version }}:${{needs.tag.outputs.build-tag}} \
+                           --cleanup \
+                           --target postgres-extensions
 
       # Cleanup script fails otherwise - rm: cannot remove '/nvme/actions-runner/_work/_temp/_github_home/.ecr': Permission denied
       - name: Cleanup ECR folder
@@ -738,7 +767,7 @@ jobs:
       run:
         shell: sh -eu {0}
     env:
-      VM_BUILDER_VERSION: v0.8.0
+      VM_BUILDER_VERSION: v0.11.1
 
     steps:
       - name: Checkout
@@ -840,8 +869,10 @@ jobs:
           crane tag 369495373322.dkr.ecr.eu-central-1.amazonaws.com/compute-tools:${{needs.tag.outputs.build-tag}} latest
           crane tag 369495373322.dkr.ecr.eu-central-1.amazonaws.com/compute-node-v14:${{needs.tag.outputs.build-tag}} latest
           crane tag 369495373322.dkr.ecr.eu-central-1.amazonaws.com/vm-compute-node-v14:${{needs.tag.outputs.build-tag}} latest
+          crane tag 369495373322.dkr.ecr.eu-central-1.amazonaws.com/extensions-v14:${{needs.tag.outputs.build-tag}} latest
           crane tag 369495373322.dkr.ecr.eu-central-1.amazonaws.com/compute-node-v15:${{needs.tag.outputs.build-tag}} latest
           crane tag 369495373322.dkr.ecr.eu-central-1.amazonaws.com/vm-compute-node-v15:${{needs.tag.outputs.build-tag}} latest
+          crane tag 369495373322.dkr.ecr.eu-central-1.amazonaws.com/extensions-v15:${{needs.tag.outputs.build-tag}} latest
 
       - name: Push images to production ECR
         if: |
@@ -852,8 +883,10 @@ jobs:
           crane copy 369495373322.dkr.ecr.eu-central-1.amazonaws.com/compute-tools:${{needs.tag.outputs.build-tag}} 093970136003.dkr.ecr.eu-central-1.amazonaws.com/compute-tools:latest
           crane copy 369495373322.dkr.ecr.eu-central-1.amazonaws.com/compute-node-v14:${{needs.tag.outputs.build-tag}} 093970136003.dkr.ecr.eu-central-1.amazonaws.com/compute-node-v14:latest
           crane copy 369495373322.dkr.ecr.eu-central-1.amazonaws.com/vm-compute-node-v14:${{needs.tag.outputs.build-tag}} 093970136003.dkr.ecr.eu-central-1.amazonaws.com/vm-compute-node-v14:latest
+          crane copy 369495373322.dkr.ecr.eu-central-1.amazonaws.com/extensions-v14:${{needs.tag.outputs.build-tag}} 093970136003.dkr.ecr.eu-central-1.amazonaws.com/extensions-v14:latest
           crane copy 369495373322.dkr.ecr.eu-central-1.amazonaws.com/compute-node-v15:${{needs.tag.outputs.build-tag}} 093970136003.dkr.ecr.eu-central-1.amazonaws.com/compute-node-v15:latest
           crane copy 369495373322.dkr.ecr.eu-central-1.amazonaws.com/vm-compute-node-v15:${{needs.tag.outputs.build-tag}} 093970136003.dkr.ecr.eu-central-1.amazonaws.com/vm-compute-node-v15:latest
+          crane copy 369495373322.dkr.ecr.eu-central-1.amazonaws.com/extensions-v15:${{needs.tag.outputs.build-tag}} 093970136003.dkr.ecr.eu-central-1.amazonaws.com/extensions-v15:latest
 
       - name: Configure Docker Hub login
         run: |
@@ -875,16 +908,93 @@ jobs:
           crane tag neondatabase/compute-tools:${{needs.tag.outputs.build-tag}} latest
           crane tag neondatabase/compute-node-v14:${{needs.tag.outputs.build-tag}} latest
           crane tag neondatabase/vm-compute-node-v14:${{needs.tag.outputs.build-tag}} latest
+          crane tag neondatabase/extensions-v14:${{needs.tag.outputs.build-tag}} latest
           crane tag neondatabase/compute-node-v15:${{needs.tag.outputs.build-tag}} latest
           crane tag neondatabase/vm-compute-node-v15:${{needs.tag.outputs.build-tag}} latest
+          crane tag neondatabase/extensions-v15:${{needs.tag.outputs.build-tag}} latest
 
       - name: Cleanup ECR folder
         run: rm -rf ~/.ecr
 
+  upload-postgres-extensions-to-s3:
+    if: |
+      (github.ref_name == 'main' || github.ref_name == 'release') &&
+       github.event_name != 'workflow_dispatch'
+    runs-on: ${{ github.ref_name == 'release' && fromJSON('["self-hosted", "prod", "x64"]') || fromJSON('["self-hosted", "gen3", "small"]') }}
+    needs: [ tag, promote-images ]
+    strategy:
+      fail-fast: false
+      matrix:
+        version: [ v14, v15 ]
+
+    env:
+      # While on transition period we extract public extensions from compute-node image and custom extensions from extensions image.
+      # Later all the extensions will be moved to extensions image.
+      EXTENSIONS_IMAGE: ${{ github.ref_name == 'release' && '093970136003' || '369495373322'}}.dkr.ecr.eu-central-1.amazonaws.com/extensions-${{ matrix.version }}:latest
+      COMPUTE_NODE_IMAGE: ${{ github.ref_name == 'release' && '093970136003' || '369495373322'}}.dkr.ecr.eu-central-1.amazonaws.com/compute-node-${{ matrix.version }}:latest
+      AWS_ACCESS_KEY_ID: ${{ github.ref_name == 'release' && secrets.AWS_ACCESS_KEY_PROD || secrets.AWS_ACCESS_KEY_DEV }}
+      AWS_SECRET_ACCESS_KEY: ${{ github.ref_name == 'release' && secrets.AWS_SECRET_KEY_PROD || secrets.AWS_SECRET_KEY_DEV }}
+      S3_BUCKETS: |
+        ${{ github.ref_name == 'release' &&
+          'neon-prod-extensions-ap-southeast-1 neon-prod-extensions-eu-central-1 neon-prod-extensions-us-east-1 neon-prod-extensions-us-east-2 neon-prod-extensions-us-west-2' ||
+          'neon-dev-extensions-eu-central-1 neon-dev-extensions-eu-west-1 neon-dev-extensions-us-east-2' }}
+
+    steps:
+      - name: Pull postgres-extensions image
+        run: |
+          docker pull ${EXTENSIONS_IMAGE}
+          docker pull ${COMPUTE_NODE_IMAGE}
+
+      - name: Create postgres-extensions container
+        id: create-container
+        run: |
+          EID=$(docker create ${EXTENSIONS_IMAGE} true)
+          echo "EID=${EID}" >> $GITHUB_OUTPUT
+
+          CID=$(docker create ${COMPUTE_NODE_IMAGE} true)
+          echo "CID=${CID}" >> $GITHUB_OUTPUT
+
+      - name: Extract postgres-extensions from container
+        run: |
+          rm -rf ./extensions-to-upload ./custom-extensions # Just in case
+
+          # In compute image we have a bit different directory layout
+          mkdir -p extensions-to-upload/share
+          docker cp ${{ steps.create-container.outputs.CID }}:/usr/local/share/extension ./extensions-to-upload/share/extension
+          docker cp ${{ steps.create-container.outputs.CID }}:/usr/local/lib             ./extensions-to-upload/lib
+
+          # Delete Neon extensitons (they always present on compute-node image)
+          rm -rf ./extensions-to-upload/share/extension/neon*
+          rm -rf ./extensions-to-upload/lib/neon*
+
+          # Delete leftovers from the extension build step
+          rm -rf ./extensions-to-upload/lib/pgxs
+          rm -rf ./extensions-to-upload/lib/pkgconfig
+
+          docker cp ${{ steps.create-container.outputs.EID }}:/extensions ./custom-extensions
+          for EXT_NAME in $(ls ./custom-extensions); do
+            mkdir -p ./extensions-to-upload/${EXT_NAME}/share
+
+            mv ./custom-extensions/${EXT_NAME}/share/extension ./extensions-to-upload/${EXT_NAME}/share/extension
+            mv ./custom-extensions/${EXT_NAME}/lib             ./extensions-to-upload/${EXT_NAME}/lib
+          done
+
+      - name: Upload postgres-extensions to S3
+        run: |
+          for BUCKET in $(echo ${S3_BUCKETS}); do
+            aws s3 cp --recursive --only-show-errors ./extensions-to-upload s3://${BUCKET}/${{ needs.tag.outputs.build-tag }}/${{ matrix.version }}
+          done
+
+      - name: Cleanup
+        if: ${{ always() && (steps.create-container.outputs.CID || steps.create-container.outputs.EID) }}
+        run: |
+          docker rm ${{ steps.create-container.outputs.CID }} || true
+          docker rm ${{ steps.create-container.outputs.EID }} || true
+
   deploy:
     runs-on: [ self-hosted, gen3, small ]
     container: 369495373322.dkr.ecr.eu-central-1.amazonaws.com/ansible:latest
-    needs: [ promote-images, tag, regress-tests ]
+    needs: [ upload-postgres-extensions-to-s3, promote-images, tag, regress-tests ]
     if: ( github.ref_name == 'main' || github.ref_name == 'release' ) && github.event_name != 'workflow_dispatch'
     steps:
       - name: Fix git ownership
@@ -916,7 +1026,7 @@ jobs:
             exit 1
           fi
 
-      - name: Create tag "release-${{ needs.tag.outputs.build-tag }}"
+      - name: Create git tag
         if: github.ref_name == 'release'
         uses: actions/github-script@v6
         with:
@@ -926,7 +1036,7 @@ jobs:
             github.rest.git.createRef({
               owner: context.repo.owner,
               repo: context.repo.repo,
-              ref: "refs/tags/release-${{ needs.tag.outputs.build-tag }}",
+              ref: "refs/tags/${{ needs.tag.outputs.build-tag }}",
               sha: context.sha,
             })
 

Cargo.lock

@@ -200,17 +200,6 @@ dependencies = [
  "critical-section",
 ]
 
-[[package]]
-name = "atty"
-version = "0.2.14"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "d9b39be18770d11421cdb1b9947a45dd3f37e93092cbf377614828a319d5fee8"
-dependencies = [
- "hermit-abi 0.1.19",
- "libc",
- "winapi",
-]
-
 [[package]]
 name = "autocfg"
 version = "1.1.0"
@@ -604,7 +593,7 @@ dependencies = [
  "cc",
  "cfg-if",
  "libc",
- "miniz_oxide",
+ "miniz_oxide 0.6.2",
  "object",
  "rustc-demangle",
 ]
@@ -805,18 +794,6 @@ dependencies = [
  "libloading",
 ]
 
-[[package]]
-name = "clap"
-version = "3.2.25"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "4ea181bf566f71cb9a5d17a59e1871af638180a18fb0035c92ae62b705207123"
-dependencies = [
- "bitflags",
- "clap_lex 0.2.4",
- "indexmap",
- "textwrap",
-]
-
 [[package]]
 name = "clap"
 version = "4.3.0"
@@ -837,7 +814,7 @@ dependencies = [
  "anstream",
  "anstyle",
  "bitflags",
- "clap_lex 0.5.0",
+ "clap_lex",
  "strsim",
 ]
 
@@ -853,15 +830,6 @@ dependencies = [
  "syn 2.0.16",
 ]
 
-[[package]]
-name = "clap_lex"
-version = "0.2.4"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "2850f2f5a82cbf437dd5af4d49848fbdfc27c157c3d010345776f952765261c5"
-dependencies = [
- "os_str_bytes",
-]
-
 [[package]]
 name = "clap_lex"
 version = "0.5.0"
@@ -915,8 +883,9 @@ version = "0.1.0"
 dependencies = [
  "anyhow",
  "chrono",
- "clap 4.3.0",
+ "clap",
  "compute_api",
+ "flate2",
  "futures",
  "hyper",
  "notify",
@@ -924,12 +893,14 @@ dependencies = [
  "opentelemetry",
  "postgres",
  "regex",
+ "remote_storage",
  "reqwest",
  "serde",
  "serde_json",
  "tar",
  "tokio",
  "tokio-postgres",
+ "toml_edit",
  "tracing",
  "tracing-opentelemetry",
  "tracing-subscriber",
@@ -977,7 +948,7 @@ name = "control_plane"
 version = "0.1.0"
 dependencies = [
  "anyhow",
- "clap 4.3.0",
+ "clap",
  "comfy-table",
  "compute_api",
  "git-version",
@@ -997,6 +968,7 @@ dependencies = [
  "tar",
  "thiserror",
  "toml",
+ "tracing",
  "url",
  "utils",
  "workspace_hack",
@@ -1047,19 +1019,19 @@ dependencies = [
 
 [[package]]
 name = "criterion"
-version = "0.4.0"
+version = "0.5.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "e7c76e09c1aae2bc52b3d2f29e13c6572553b30c4aa1b8a49fd70de6412654cb"
+checksum = "f2b12d017a929603d80db1831cd3a24082f8137ce19c69e6447f54f5fc8d692f"
 dependencies = [
  "anes",
- "atty",
  "cast",
  "ciborium",
- "clap 3.2.25",
+ "clap",
  "criterion-plot",
+ "is-terminal",
  "itertools",
- "lazy_static",
  "num-traits",
+ "once_cell",
  "oorandom",
  "plotters",
  "rayon",
@@ -1140,7 +1112,7 @@ dependencies = [
  "crossterm_winapi",
  "libc",
  "mio",
- "parking_lot",
+ "parking_lot 0.12.1",
  "signal-hook",
  "signal-hook-mio",
  "winapi",
@@ -1210,7 +1182,7 @@ dependencies = [
  "hashbrown 0.12.3",
  "lock_api",
  "once_cell",
- "parking_lot_core",
+ "parking_lot_core 0.9.7",
 ]
 
 [[package]]
@@ -1399,6 +1371,16 @@ version = "0.4.2"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "0ce7134b9999ecaf8bcd65542e436736ef32ddca1b3e06094cb6ec5755203b80"
 
+[[package]]
+name = "flate2"
+version = "1.0.26"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "3b9429470923de8e8cbd4d2dc513535400b4b3fef0319fb5c4e1f520a7bef743"
+dependencies = [
+ "crc32fast",
+ "miniz_oxide 0.7.1",
+]
+
 [[package]]
 name = "fnv"
 version = "1.0.7"
@@ -1676,15 +1658,6 @@ version = "0.4.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "95505c38b4572b2d910cecb0281560f54b440a19336cbbcb27bf6ce6adc6f5a8"
 
-[[package]]
-name = "hermit-abi"
-version = "0.1.19"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "62b467343b94ba476dcb2500d242dadbb39557df889310ac77c5d99100aaac33"
-dependencies = [
- "libc",
-]
-
 [[package]]
 name = "hermit-abi"
 version = "0.2.6"
@@ -1939,6 +1912,9 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "7a5bbe824c507c5da5956355e86a746d82e0e1464f65d862cc5e71da70e94b2c"
 dependencies = [
  "cfg-if",
+ "js-sys",
+ "wasm-bindgen",
+ "web-sys",
 ]
 
 [[package]]
@@ -2189,6 +2165,15 @@ dependencies = [
  "adler",
 ]
 
+[[package]]
+name = "miniz_oxide"
+version = "0.7.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "e7810e0be55b428ada41041c41f32c9f1a42817901b4ccf45fa3d4b6561e74c7"
+dependencies = [
+ "adler",
+]
+
 [[package]]
 name = "mio"
 version = "0.8.6"
@@ -2267,16 +2252,6 @@ dependencies = [
  "windows-sys 0.45.0",
 ]
 
-[[package]]
-name = "nu-ansi-term"
-version = "0.46.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "77a8165726e8236064dbb45459242600304b42a5ea24ee2948e18e023bf7ba84"
-dependencies = [
- "overload",
- "winapi",
-]
-
 [[package]]
 name = "num-bigint"
 version = "0.4.3"
@@ -2504,31 +2479,19 @@ dependencies = [
  "winapi",
 ]
 
-[[package]]
-name = "os_str_bytes"
-version = "6.5.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "ceedf44fb00f2d1984b0bc98102627ce622e083e49a5bacdb3e514fa4238e267"
-
 [[package]]
 name = "outref"
 version = "0.5.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "4030760ffd992bef45b0ae3f10ce1aba99e33464c90d14dd7c039884963ddc7a"
 
-[[package]]
-name = "overload"
-version = "0.1.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "b15813163c1d831bf4a13c3610c05c0d03b39feb07f7e09fa234dac9b15aaf39"
-
 [[package]]
 name = "pagectl"
 version = "0.1.0"
 dependencies = [
  "anyhow",
  "bytes",
- "clap 4.3.0",
+ "clap",
  "git-version",
  "pageserver",
  "postgres_ffi",
@@ -2547,7 +2510,7 @@ dependencies = [
  "byteorder",
  "bytes",
  "chrono",
- "clap 4.3.0",
+ "clap",
  "close_fds",
  "const_format",
  "consumption_metrics",
@@ -2629,6 +2592,17 @@ dependencies = [
  "workspace_hack",
 ]
 
+[[package]]
+name = "parking_lot"
+version = "0.11.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "7d17b78036a60663b797adeaee46f5c9dfebb86948d1255007a1d6be0271ff99"
+dependencies = [
+ "instant",
+ "lock_api",
+ "parking_lot_core 0.8.6",
+]
+
 [[package]]
 name = "parking_lot"
 version = "0.12.1"
@@ -2636,7 +2610,21 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "3742b2c103b9f06bc9fff0a37ff4912935851bee6d36f3c02bcc755bcfec228f"
 dependencies = [
  "lock_api",
- "parking_lot_core",
+ "parking_lot_core 0.9.7",
+]
+
+[[package]]
+name = "parking_lot_core"
+version = "0.8.6"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "60a2cfe6f0ad2bfc16aefa463b497d5c7a5ecd44a23efa72aa342d90177356dc"
+dependencies = [
+ "cfg-if",
+ "instant",
+ "libc",
+ "redox_syscall 0.2.16",
+ "smallvec",
+ "winapi",
 ]
 
 [[package]]
@@ -2957,7 +2945,7 @@ dependencies = [
  "lazy_static",
  "libc",
  "memchr",
- "parking_lot",
+ "parking_lot 0.12.1",
  "procfs",
  "thiserror",
 ]
@@ -3022,12 +3010,11 @@ version = "0.1.0"
 dependencies = [
  "anyhow",
  "async-trait",
- "atty",
  "base64 0.13.1",
  "bstr",
  "bytes",
  "chrono",
- "clap 4.3.0",
+ "clap",
  "consumption_metrics",
  "futures",
  "git-version",
@@ -3045,7 +3032,7 @@ dependencies = [
  "native-tls",
  "once_cell",
  "opentelemetry",
- "parking_lot",
+ "parking_lot 0.12.1",
  "pin-project-lite",
  "postgres-native-tls",
  "postgres_backend",
@@ -3056,6 +3043,7 @@ dependencies = [
  "regex",
  "reqwest",
  "reqwest-middleware",
+ "reqwest-retry",
  "reqwest-tracing",
  "routerify",
  "rstest",
@@ -3291,6 +3279,29 @@ dependencies = [
  "thiserror",
 ]
 
+[[package]]
+name = "reqwest-retry"
+version = "0.2.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "48d0fd6ef4c6d23790399fe15efc8d12cd9f3d4133958f9bd7801ee5cbaec6c4"
+dependencies = [
+ "anyhow",
+ "async-trait",
+ "chrono",
+ "futures",
+ "getrandom",
+ "http",
+ "hyper",
+ "parking_lot 0.11.2",
+ "reqwest",
+ "reqwest-middleware",
+ "retry-policies",
+ "task-local-extensions",
+ "tokio",
+ "tracing",
+ "wasm-timer",
+]
+
 [[package]]
 name = "reqwest-tracing"
 version = "0.4.4"
@@ -3309,6 +3320,17 @@ dependencies = [
  "tracing-opentelemetry",
 ]
 
+[[package]]
+name = "retry-policies"
+version = "0.1.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "e09bbcb5003282bcb688f0bae741b278e9c7e8f378f561522c9806c58e075d9b"
+dependencies = [
+ "anyhow",
+ "chrono",
+ "rand",
+]
+
 [[package]]
 name = "ring"
 version = "0.16.20"
@@ -3507,7 +3529,7 @@ dependencies = [
  "byteorder",
  "bytes",
  "chrono",
- "clap 4.3.0",
+ "clap",
  "const_format",
  "crc32c",
  "fs2",
@@ -3518,7 +3540,7 @@ dependencies = [
  "hyper",
  "metrics",
  "once_cell",
- "parking_lot",
+ "parking_lot 0.12.1",
  "postgres",
  "postgres-protocol",
  "postgres_backend",
@@ -3937,7 +3959,7 @@ dependencies = [
  "anyhow",
  "async-stream",
  "bytes",
- "clap 4.3.0",
+ "clap",
  "const_format",
  "futures",
  "futures-core",
@@ -3947,7 +3969,7 @@ dependencies = [
  "hyper",
  "metrics",
  "once_cell",
- "parking_lot",
+ "parking_lot 0.12.1",
  "prost",
  "tokio",
  "tokio-stream",
@@ -4118,12 +4140,6 @@ dependencies = [
  "syn 1.0.109",
 ]
 
-[[package]]
-name = "textwrap"
-version = "0.16.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "222a222a5bfe1bba4a77b45ec488a741b3cb8872e5e499451fd7d0129c9c7c3d"
-
 [[package]]
 name = "thiserror"
 version = "1.0.40"
@@ -4281,7 +4297,7 @@ dependencies = [
  "futures-channel",
  "futures-util",
  "log",
- "parking_lot",
+ "parking_lot 0.12.1",
  "percent-encoding",
  "phf",
  "pin-project-lite",
@@ -4539,7 +4555,7 @@ name = "trace"
 version = "0.1.0"
 dependencies = [
  "anyhow",
- "clap 4.3.0",
+ "clap",
  "pageserver_api",
  "utils",
  "workspace_hack",
@@ -4641,7 +4657,6 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "30a651bc37f915e81f087d86e62a18eec5f79550c7faff886f7090b4ea757c77"
 dependencies = [
  "matchers",
- "nu-ansi-term",
  "once_cell",
  "regex",
  "serde",
@@ -4810,7 +4825,6 @@ version = "0.1.0"
 dependencies = [
  "anyhow",
  "async-trait",
- "atty",
  "bincode",
  "byteorder",
  "bytes",
@@ -4887,7 +4901,7 @@ name = "wal_craft"
 version = "0.1.0"
 dependencies = [
  "anyhow",
- "clap 4.3.0",
+ "clap",
  "env_logger",
  "log",
  "once_cell",
@@ -4991,6 +5005,21 @@ version = "0.2.86"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "ed9d5b4305409d1fc9482fee2d7f9bcbf24b3972bf59817ef757e23982242a93"
 
+[[package]]
+name = "wasm-timer"
+version = "0.2.5"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "be0ecb0db480561e9a7642b5d3e4187c128914e58aa84330b9493e3eb68c5e7f"
+dependencies = [
+ "futures",
+ "js-sys",
+ "parking_lot 0.11.2",
+ "pin-utils",
+ "wasm-bindgen",
+ "wasm-bindgen-futures",
+ "web-sys",
+]
+
 [[package]]
 name = "web-sys"
 version = "0.3.63"
@@ -5252,7 +5281,7 @@ dependencies = [
  "anyhow",
  "bytes",
  "chrono",
- "clap 4.3.0",
+ "clap",
  "clap_builder",
  "crossbeam-utils",
  "either",

Cargo.toml

@@ -34,7 +34,6 @@ license = "Apache-2.0"
 anyhow = { version = "1.0", features = ["backtrace"] }
 async-stream = "0.3"
 async-trait = "0.1"
-atty = "0.2.14"
 aws-config = { version = "0.55", default-features = false, features=["rustls"] }
 aws-sdk-s3 = "0.27"
 aws-smithy-http = "0.55"
@@ -95,6 +94,7 @@ regex = "1.4"
 reqwest = { version = "0.11", default-features = false, features = ["rustls-tls"] }
 reqwest-tracing = { version = "0.4.0", features = ["opentelemetry_0_18"] }
 reqwest-middleware = "0.2.0"
+reqwest-retry = "0.2.2"
 routerify = "3"
 rpds = "0.13"
 rustls = "0.20"
@@ -128,7 +128,7 @@ tonic = {version = "0.9", features = ["tls", "tls-roots"]}
 tracing = "0.1"
 tracing-error = "0.2.0"
 tracing-opentelemetry = "0.18.0"
-tracing-subscriber = { version = "0.3", features = ["env-filter"] }
+tracing-subscriber = { version = "0.3", default_features = false, features = ["smallvec", "fmt", "tracing-log", "std", "env-filter"] }
 url = "2.2"
 uuid = { version = "1.2", features = ["v4", "serde"] }
 walkdir = "2.3.2"
@@ -170,7 +170,7 @@ utils = { version = "0.1", path = "./libs/utils/" }
 workspace_hack = { version = "0.1", path = "./workspace_hack/" }
 
 ## Build dependencies
-criterion = "0.4"
+criterion = "0.5.1"
 rcgen = "0.10"
 rstest = "0.17"
 tempfile = "3.4"

Dockerfile.compute-node

@@ -481,6 +481,60 @@ RUN wget https://github.com/rdkit/rdkit/archive/refs/tags/Release_2023_03_1.tar.
     make -j $(getconf _NPROCESSORS_ONLN) install && \
     echo 'trusted = true' >> /usr/local/pgsql/share/extension/rdkit.control
 
+#########################################################################################
+#
+# Layer "pg-uuidv7-pg-build"
+# compile pg_uuidv7 extension
+#
+#########################################################################################
+FROM build-deps AS pg-uuidv7-pg-build
+COPY --from=pg-build /usr/local/pgsql/ /usr/local/pgsql/
+
+ENV PATH "/usr/local/pgsql/bin/:$PATH"
+RUN wget https://github.com/fboulnois/pg_uuidv7/archive/refs/tags/v1.0.1.tar.gz -O pg_uuidv7.tar.gz && \
+    echo "0d0759ab01b7fb23851ecffb0bce27822e1868a4a5819bfd276101c716637a7a pg_uuidv7.tar.gz" | sha256sum --check && \
+    mkdir pg_uuidv7-src && cd pg_uuidv7-src && tar xvzf ../pg_uuidv7.tar.gz --strip-components=1 -C . && \
+    make -j $(getconf _NPROCESSORS_ONLN) && \
+    make -j $(getconf _NPROCESSORS_ONLN) install && \
+    echo 'trusted = true' >> /usr/local/pgsql/share/extension/pg_uuidv7.control
+
+#########################################################################################
+#
+# Layer "pg-roaringbitmap-pg-build"
+# compile pg_roaringbitmap extension
+#
+#########################################################################################
+FROM build-deps AS pg-roaringbitmap-pg-build
+COPY --from=pg-build /usr/local/pgsql/ /usr/local/pgsql/
+
+ENV PATH "/usr/local/pgsql/bin/:$PATH"
+RUN wget https://github.com/ChenHuajun/pg_roaringbitmap/archive/refs/tags/v0.5.4.tar.gz -O pg_roaringbitmap.tar.gz && \
+    echo "b75201efcb1c2d1b014ec4ae6a22769cc7a224e6e406a587f5784a37b6b5a2aa pg_roaringbitmap.tar.gz" | sha256sum --check && \
+    mkdir pg_roaringbitmap-src && cd pg_roaringbitmap-src && tar xvzf ../pg_roaringbitmap.tar.gz --strip-components=1 -C . && \
+    make -j $(getconf _NPROCESSORS_ONLN) && \
+    make -j $(getconf _NPROCESSORS_ONLN) install && \
+    echo 'trusted = true' >> /usr/local/pgsql/share/extension/roaringbitmap.control
+
+#########################################################################################
+#
+# Layer "pg-anon-pg-build"
+# compile anon extension
+#
+#########################################################################################
+FROM build-deps AS pg-anon-pg-build
+COPY --from=pg-build /usr/local/pgsql/ /usr/local/pgsql/
+
+# Kaniko doesn't allow to do `${from#/usr/local/pgsql/}`, so we use `${from:17}` instead
+ENV PATH "/usr/local/pgsql/bin/:$PATH"
+RUN wget https://gitlab.com/dalibo/postgresql_anonymizer/-/archive/1.1.0/postgresql_anonymizer-1.1.0.tar.gz -O pg_anon.tar.gz && \
+    echo "08b09d2ff9b962f96c60db7e6f8e79cf7253eb8772516998fc35ece08633d3ad pg_anon.tar.gz" | sha256sum --check && \
+    mkdir pg_anon-src && cd pg_anon-src && tar xvzf ../pg_anon.tar.gz --strip-components=1 -C . && \
+    find /usr/local/pgsql -type f | sort  > /before.txt && \
+    make -j $(getconf _NPROCESSORS_ONLN) install PG_CONFIG=/usr/local/pgsql/bin/pg_config && \
+    echo 'trusted = true' >> /usr/local/pgsql/share/extension/anon.control && \
+    find /usr/local/pgsql -type f | sort  > /after.txt && \
+    /bin/bash -c 'for from in $(comm -13 /before.txt /after.txt); do to=/extensions/anon/${from:17} && mkdir -p $(dirname ${to}) && cp -a ${from} ${to}; done'
+
 #########################################################################################
 #
 # Layer "rust extensions"
@@ -589,6 +643,7 @@ RUN wget https://github.com/pksunkara/pgx_ulid/archive/refs/tags/v0.1.0.tar.gz -
 #
 #########################################################################################
 FROM build-deps AS neon-pg-ext-build
+# Public extensions
 COPY --from=postgis-build /usr/local/pgsql/ /usr/local/pgsql/
 COPY --from=postgis-build /sfcgal/* /
 COPY --from=plv8-build /usr/local/pgsql/ /usr/local/pgsql/
@@ -614,6 +669,8 @@ COPY --from=kq-imcx-pg-build /usr/local/pgsql/ /usr/local/pgsql/
 COPY --from=pg-cron-pg-build /usr/local/pgsql/ /usr/local/pgsql/
 COPY --from=pg-pgx-ulid-build /usr/local/pgsql/ /usr/local/pgsql/
 COPY --from=rdkit-pg-build /usr/local/pgsql/ /usr/local/pgsql/
+COPY --from=pg-uuidv7-pg-build /usr/local/pgsql/ /usr/local/pgsql/
+COPY --from=pg-roaringbitmap-pg-build /usr/local/pgsql/ /usr/local/pgsql/
 COPY pgxn/ pgxn/
 
 RUN make -j $(getconf _NPROCESSORS_ONLN) \
@@ -662,6 +719,22 @@ RUN rm -r /usr/local/pgsql/include
 # if they were to be used by other libraries.
 RUN rm /usr/local/pgsql/lib/lib*.a
 
+#########################################################################################
+#
+# Extenstion only
+#
+#########################################################################################
+FROM scratch AS postgres-extensions
+# After the transition this layer will include all extensitons.
+# As for now, it's only for new custom ones
+#
+# # Default extensions
+# COPY --from=postgres-cleanup-layer /usr/local/pgsql/share/extension /usr/local/pgsql/share/extension
+# COPY --from=postgres-cleanup-layer /usr/local/pgsql/lib             /usr/local/pgsql/lib
+# Custom extensions
+COPY --from=pg-anon-pg-build /extensions/anon/lib/ /extensions/anon/lib
+COPY --from=pg-anon-pg-build /extensions/anon/share/extension /extensions/anon/share/extension
+
 #########################################################################################
 #
 # Final layer

README.md

@@ -132,13 +132,13 @@ Python (3.9 or higher), and install python3 packages using `./scripts/pysync` (r
 # Create repository in .neon with proper paths to binaries and data
 # Later that would be responsibility of a package install script
 > cargo neon init
-Starting pageserver at '127.0.0.1:64000' in '.neon'.
+Initializing pageserver node 1 at '127.0.0.1:64000' in ".neon"
 
 # start pageserver, safekeeper, and broker for their intercommunication
 > cargo neon start
-Starting neon broker at 127.0.0.1:50051
+Starting neon broker at 127.0.0.1:50051.
 storage_broker started, pid: 2918372
-Starting pageserver at '127.0.0.1:64000' in '.neon'.
+Starting pageserver node 1 at '127.0.0.1:64000' in ".neon".
 pageserver started, pid: 2918386
 Starting safekeeper at '127.0.0.1:5454' in '.neon/safekeepers/sk1'.
 safekeeper 1 started, pid: 2918437
@@ -152,8 +152,7 @@ Setting tenant 9ef87a5bf0d92544f6fafeeb3239695c as a default one
 # start postgres compute node
 > cargo neon endpoint start main
 Starting new endpoint main (PostgreSQL v14) on timeline de200bd42b49cc1814412c7e592dd6e9 ...
-Extracting base backup to create postgres instance: path=.neon/pgdatadirs/tenants/9ef87a5bf0d92544f6fafeeb3239695c/main port=55432
+Starting postgres at 'postgresql://cloud_admin@127.0.0.1:55432/postgres'
-Starting postgres at 'host=127.0.0.1 port=55432 user=cloud_admin dbname=postgres'
 
 # check list of running postgres instances
 > cargo neon endpoint list
@@ -189,18 +188,17 @@ Created timeline 'b3b863fa45fa9e57e615f9f2d944e601' at Lsn 0/16F9A00 for tenant:
 # start postgres on that branch
 > cargo neon endpoint start migration_check --branch-name migration_check
 Starting new endpoint migration_check (PostgreSQL v14) on timeline b3b863fa45fa9e57e615f9f2d944e601 ...
-Extracting base backup to create postgres instance: path=.neon/pgdatadirs/tenants/9ef87a5bf0d92544f6fafeeb3239695c/migration_check port=55433
+Starting postgres at 'postgresql://cloud_admin@127.0.0.1:55434/postgres'
-Starting postgres at 'host=127.0.0.1 port=55433 user=cloud_admin dbname=postgres'
 
 # check the new list of running postgres instances
 > cargo neon endpoint list
  ENDPOINT         ADDRESS          TIMELINE                          BRANCH NAME      LSN        STATUS
  main             127.0.0.1:55432  de200bd42b49cc1814412c7e592dd6e9  main             0/16F9A38  running
- migration_check  127.0.0.1:55433  b3b863fa45fa9e57e615f9f2d944e601  migration_check  0/16F9A70  running
+ migration_check  127.0.0.1:55434  b3b863fa45fa9e57e615f9f2d944e601  migration_check  0/16F9A70  running
 
 # this new postgres instance will have all the data from 'main' postgres,
 # but all modifications would not affect data in original postgres
-> psql -p55433 -h 127.0.0.1 -U cloud_admin postgres
+> psql -p55434 -h 127.0.0.1 -U cloud_admin postgres
 postgres=# select * from t;
  key | value
 -----+-------

compute_tools/Cargo.toml

@@ -30,3 +30,6 @@ url.workspace = true
 compute_api.workspace = true
 utils.workspace = true
 workspace_hack.workspace = true
+toml_edit.workspace = true
+remote_storage = { version = "0.1", path = "../libs/remote_storage/" }
+flate2 = "1.0.26"

compute_tools/src/bin/compute_ctl.rs

@@ -5,6 +5,8 @@
 //! - `compute_ctl` accepts cluster (compute node) specification as a JSON file.
 //! - Every start is a fresh start, so the data directory is removed and
 //!   initialized again on each run.
+//! - If remote_extension_config is provided, it will be used to fetch extensions list
+//!  and download `shared_preload_libraries` from the remote storage.
 //! - Next it will put configuration files into the `PGDATA` directory.
 //! - Sync safekeepers and get commit LSN.
 //! - Get `basebackup` from pageserver using the returned on the previous step LSN.
@@ -27,7 +29,8 @@
 //! compute_ctl -D /var/db/postgres/compute \
 //!             -C 'postgresql://cloud_admin@localhost/postgres' \
 //!             -S /var/db/postgres/specs/current.json \
-//!             -b /usr/local/bin/postgres
+//!             -b /usr/local/bin/postgres \
+//!             -r {"bucket": "my-bucket", "region": "eu-central-1", "endpoint": "http:://localhost:9000"} \
 //! ```
 //!
 use std::collections::HashMap;
@@ -35,7 +38,7 @@ use std::fs::File;
 use std::panic;
 use std::path::Path;
 use std::process::exit;
-use std::sync::{mpsc, Arc, Condvar, Mutex};
+use std::sync::{mpsc, Arc, Condvar, Mutex, OnceLock};
 use std::{thread, time::Duration};
 
 use anyhow::{Context, Result};
@@ -48,6 +51,8 @@ use compute_api::responses::ComputeStatus;
 
 use compute_tools::compute::{ComputeNode, ComputeState, ParsedSpec};
 use compute_tools::configurator::launch_configurator;
+use compute_tools::extension_server::launch_download_extensions;
+use compute_tools::extension_server::{get_pg_version, init_remote_storage};
 use compute_tools::http::api::launch_http_server;
 use compute_tools::logger::*;
 use compute_tools::monitor::launch_monitor;
@@ -60,10 +65,21 @@ fn main() -> Result<()> {
     init_tracing_and_logging(DEFAULT_LOG_LEVEL)?;
 
     let build_tag = option_env!("BUILD_TAG").unwrap_or(BUILD_TAG_DEFAULT);
-
     info!("build_tag: {build_tag}");
 
     let matches = cli().get_matches();
+    let pgbin_default = String::from("postgres");
+    let pgbin = matches.get_one::<String>("pgbin").unwrap_or(&pgbin_default);
+
+    let remote_ext_config = matches.get_one::<String>("remote-ext-config");
+    // NOTE TODO: until control-plane changes, we can use the following line to forcibly enable remote extensions
+    // let remote_ext_config = Some(
+    //     r#"{"bucket": "neon-dev-extensions", "region": "eu-central-1", "endpoint": null, "prefix": "5555"}"#.to_string(),
+    // );
+    let ext_remote_storage = remote_ext_config.map(|x| {
+        init_remote_storage(x, build_tag)
+            .expect("cannot initialize remote extension storage from config")
+    });
 
     let http_port = *matches
         .get_one::<u16>("http-port")
@@ -128,9 +144,6 @@ fn main() -> Result<()> {
     let compute_id = matches.get_one::<String>("compute-id");
     let control_plane_uri = matches.get_one::<String>("control-plane-uri");
 
-    // Try to use just 'postgres' if no path is provided
-    let pgbin = matches.get_one::<String>("pgbin").unwrap();
-
     let spec;
     let mut live_config_allowed = false;
     match spec_json {
@@ -168,6 +181,7 @@ fn main() -> Result<()> {
 
     let mut new_state = ComputeState::new();
     let spec_set;
+
     if let Some(spec) = spec {
         let pspec = ParsedSpec::try_from(spec).map_err(|msg| anyhow::anyhow!(msg))?;
         new_state.pspec = Some(pspec);
@@ -179,9 +193,12 @@ fn main() -> Result<()> {
         connstr: Url::parse(connstr).context("cannot parse connstr as a URL")?,
         pgdata: pgdata.to_string(),
         pgbin: pgbin.to_string(),
+        pgversion: get_pg_version(pgbin),
         live_config_allowed,
         state: Mutex::new(new_state),
         state_changed: Condvar::new(),
+        ext_remote_storage,
+        available_extensions: OnceLock::new(),
     };
     let compute = Arc::new(compute_node);
 
@@ -190,6 +207,8 @@ fn main() -> Result<()> {
     let _http_handle =
         launch_http_server(http_port, &compute).expect("cannot launch http endpoint thread");
 
+    let extension_server_port: u16 = http_port;
+
     if !spec_set {
         // No spec provided, hang waiting for it.
         info!("no compute spec provided, waiting");
@@ -227,10 +246,13 @@ fn main() -> Result<()> {
     let _configurator_handle =
         launch_configurator(&compute).expect("cannot launch configurator thread");
 
+    let _download_extensions_handle =
+        launch_download_extensions(&compute).expect("cannot launch download extensions thread");
+
     // Start Postgres
     let mut delay_exit = false;
     let mut exit_code = None;
-    let pg = match compute.start_compute() {
+    let pg = match compute.start_compute(extension_server_port) {
         Ok(pg) => Some(pg),
         Err(err) => {
             error!("could not start the compute node: {:?}", err);
@@ -256,6 +278,16 @@ fn main() -> Result<()> {
         exit_code = ecode.code()
     }
 
+    // Maybe sync safekeepers again, to speed up next startup
+    let compute_state = compute.state.lock().unwrap().clone();
+    let pspec = compute_state.pspec.as_ref().expect("spec must be set");
+    if matches!(pspec.spec.mode, compute_api::spec::ComputeMode::Primary) {
+        info!("syncing safekeepers on shutdown");
+        let storage_auth_token = pspec.storage_auth_token.clone();
+        let lsn = compute.sync_safekeepers(storage_auth_token)?;
+        info!("synced safekeepers at lsn {lsn}");
+    }
+
     if let Err(err) = compute.check_for_core_dumps() {
         error!("error while checking for core dumps: {err:?}");
     }
@@ -349,6 +381,12 @@ fn cli() -> clap::Command {
                 .long("control-plane-uri")
                 .value_name("CONTROL_PLANE_API_BASE_URI"),
         )
+        .arg(
+            Arg::new("remote-ext-config")
+                .short('r')
+                .long("remote-ext-config")
+                .value_name("REMOTE_EXT_CONFIG"),
+        )
 }
 
 #[test]

compute_tools/src/compute.rs

@@ -1,13 +1,16 @@
+use std::collections::HashSet;
 use std::fs;
 use std::os::unix::fs::PermissionsExt;
 use std::path::Path;
 use std::process::{Command, Stdio};
 use std::str::FromStr;
-use std::sync::{Condvar, Mutex};
+use std::sync::{Condvar, Mutex, OnceLock};
 
 use anyhow::{Context, Result};
 use chrono::{DateTime, Utc};
+use futures::future::join_all;
 use postgres::{Client, NoTls};
+use tokio;
 use tokio_postgres;
 use tracing::{info, instrument, warn};
 use utils::id::{TenantId, TimelineId};
@@ -16,9 +19,11 @@ use utils::lsn::Lsn;
 use compute_api::responses::{ComputeMetrics, ComputeStatus};
 use compute_api::spec::{ComputeMode, ComputeSpec};
 
-use crate::config;
+use remote_storage::GenericRemoteStorage;
+
 use crate::pg_helpers::*;
 use crate::spec::*;
+use crate::{config, extension_server};
 
 /// Compute node info shared across several `compute_ctl` threads.
 pub struct ComputeNode {
@@ -26,6 +31,7 @@ pub struct ComputeNode {
     pub connstr: url::Url,
     pub pgdata: String,
     pub pgbin: String,
+    pub pgversion: String,
     /// We should only allow live re- / configuration of the compute node if
     /// it uses 'pull model', i.e. it can go to control-plane and fetch
     /// the latest configuration. Otherwise, there could be a case:
@@ -45,6 +51,10 @@ pub struct ComputeNode {
     pub state: Mutex<ComputeState>,
     /// `Condvar` to allow notifying waiters about state changes.
     pub state_changed: Condvar,
+    ///  the S3 bucket that we search for extensions in
+    pub ext_remote_storage: Option<GenericRemoteStorage>,
+    // cached lists of available extensions and libraries
+    pub available_extensions: OnceLock<HashSet<String>>,
 }
 
 #[derive(Clone, Debug)]
@@ -133,6 +143,84 @@ impl TryFrom<ComputeSpec> for ParsedSpec {
     }
 }
 
+/// Create special neon_superuser role, that's a slightly nerfed version of a real superuser
+/// that we give to customers
+fn create_neon_superuser(spec: &ComputeSpec, client: &mut Client) -> Result<()> {
+    let roles = spec
+        .cluster
+        .roles
+        .iter()
+        .map(|r| format!("'{}'", escape_literal(&r.name)))
+        .collect::<Vec<_>>();
+
+    let dbs = spec
+        .cluster
+        .databases
+        .iter()
+        .map(|db| format!("'{}'", escape_literal(&db.name)))
+        .collect::<Vec<_>>();
+
+    let roles_decl = if roles.is_empty() {
+        String::from("roles text[] := NULL;")
+    } else {
+        format!(
+            r#"
+               roles text[] := ARRAY(SELECT rolname
+                                     FROM pg_catalog.pg_roles
+                                     WHERE rolname IN ({}));"#,
+            roles.join(", ")
+        )
+    };
+
+    let database_decl = if dbs.is_empty() {
+        String::from("dbs text[] := NULL;")
+    } else {
+        format!(
+            r#"
+               dbs text[] := ARRAY(SELECT datname
+                                   FROM pg_catalog.pg_database
+                                   WHERE datname IN ({}));"#,
+            dbs.join(", ")
+        )
+    };
+
+    // ALL PRIVILEGES grants CREATE, CONNECT, and TEMPORARY on all databases
+    // (see https://www.postgresql.org/docs/current/ddl-priv.html)
+    let query = format!(
+        r#"
+            DO $$
+                DECLARE
+                    r text;
+                    {}
+                    {}
+                BEGIN
+                    IF NOT EXISTS (
+                        SELECT FROM pg_catalog.pg_roles WHERE rolname = 'neon_superuser')
+                    THEN
+                        CREATE ROLE neon_superuser CREATEDB CREATEROLE NOLOGIN IN ROLE pg_read_all_data, pg_write_all_data;
+                        IF array_length(roles, 1) IS NOT NULL THEN
+                            EXECUTE format('GRANT neon_superuser TO %s',
+                                           array_to_string(ARRAY(SELECT quote_ident(x) FROM unnest(roles) as x), ', '));
+                            FOREACH r IN ARRAY roles LOOP
+                                EXECUTE format('ALTER ROLE %s CREATEROLE CREATEDB', quote_ident(r));
+                            END LOOP;
+                        END IF;
+                        IF array_length(dbs, 1) IS NOT NULL THEN
+                            EXECUTE format('GRANT ALL PRIVILEGES ON DATABASE %s TO neon_superuser',
+                                           array_to_string(ARRAY(SELECT quote_ident(x) FROM unnest(dbs) as x), ', '));
+                        END IF;
+                    END IF;
+                END
+            $$;"#,
+        roles_decl, database_decl,
+    );
+    info!("Neon superuser created:\n{}", &query);
+    client
+        .simple_query(&query)
+        .map_err(|e| anyhow::anyhow!(e).context(query))?;
+    Ok(())
+}
+
 impl ComputeNode {
     pub fn set_status(&self, status: ComputeStatus) {
         let mut state = self.state.lock().unwrap();
@@ -157,7 +245,7 @@ impl ComputeNode {
 
     // Get basebackup from the libpq connection to pageserver using `connstr` and
     // unarchive it to `pgdata` directory overriding all its previous content.
-    #[instrument(skip(self, compute_state))]
+    #[instrument(skip_all, fields(%lsn))]
     fn get_basebackup(&self, compute_state: &ComputeState, lsn: Lsn) -> Result<()> {
         let spec = compute_state.pspec.as_ref().expect("spec must be set");
         let start_time = Utc::now();
@@ -199,8 +287,8 @@ impl ComputeNode {
 
     // Run `postgres` in a special mode with `--sync-safekeepers` argument
     // and return the reported LSN back to the caller.
-    #[instrument(skip(self, storage_auth_token))]
+    #[instrument(skip_all)]
-    fn sync_safekeepers(&self, storage_auth_token: Option<String>) -> Result<Lsn> {
+    pub fn sync_safekeepers(&self, storage_auth_token: Option<String>) -> Result<Lsn> {
         let start_time = Utc::now();
 
         let sync_handle = Command::new(&self.pgbin)
@@ -244,15 +332,23 @@ impl ComputeNode {
 
     /// Do all the preparations like PGDATA directory creation, configuration,
     /// safekeepers sync, basebackup, etc.
-    #[instrument(skip(self, compute_state))]
+    #[instrument(skip_all)]
-    pub fn prepare_pgdata(&self, compute_state: &ComputeState) -> Result<()> {
+    pub fn prepare_pgdata(
+        &self,
+        compute_state: &ComputeState,
+        extension_server_port: u16,
+    ) -> Result<()> {
         let pspec = compute_state.pspec.as_ref().expect("spec must be set");
         let spec = &pspec.spec;
         let pgdata_path = Path::new(&self.pgdata);
 
         // Remove/create an empty pgdata directory and put configuration there.
         self.create_pgdata()?;
-        config::write_postgres_conf(&pgdata_path.join("postgresql.conf"), &pspec.spec)?;
+        config::write_postgres_conf(
+            &pgdata_path.join("postgresql.conf"),
+            &pspec.spec,
+            Some(extension_server_port),
+        )?;
 
         // Syncing safekeepers is only safe with primary nodes: if a primary
         // is already connected it will be kicked out, so a secondary (standby)
@@ -302,7 +398,7 @@ impl ComputeNode {
 
     /// Start Postgres as a child process and manage DBs/roles.
     /// After that this will hang waiting on the postmaster process to exit.
-    #[instrument(skip(self))]
+    #[instrument(skip_all)]
     pub fn start_postgres(
         &self,
         storage_auth_token: Option<String>,
@@ -326,7 +422,7 @@ impl ComputeNode {
     }
 
     /// Do initial configuration of the already started Postgres.
-    #[instrument(skip(self, compute_state))]
+    #[instrument(skip_all)]
     pub fn apply_config(&self, compute_state: &ComputeState) -> Result<()> {
         // If connection fails,
         // it may be the old node with `zenith_admin` superuser.
@@ -347,6 +443,8 @@ impl ComputeNode {
                     .map_err(|_| anyhow::anyhow!("invalid connstr"))?;
 
                 let mut client = Client::connect(zenith_admin_connstr.as_str(), NoTls)?;
+                // Disable forwarding so that users don't get a cloud_admin role
+                client.simple_query("SET neon.forward_ddl = false")?;
                 client.simple_query("CREATE USER cloud_admin WITH SUPERUSER")?;
                 client.simple_query("GRANT zenith_admin TO cloud_admin")?;
                 drop(client);
@@ -357,14 +455,16 @@ impl ComputeNode {
             Ok(client) => client,
         };
 
-        // Proceed with post-startup configuration. Note, that order of operations is important.
         // Disable DDL forwarding because control plane already knows about these roles/databases.
         client.simple_query("SET neon.forward_ddl = false")?;
+
+        // Proceed with post-startup configuration. Note, that order of operations is important.
         let spec = &compute_state.pspec.as_ref().expect("spec must be set").spec;
+        create_neon_superuser(spec, &mut client)?;
         handle_roles(spec, &mut client)?;
         handle_databases(spec, &mut client)?;
         handle_role_deletions(spec, self.connstr.as_str(), &mut client)?;
-        handle_grants(spec, self.connstr.as_str(), &mut client)?;
+        handle_grants(spec, self.connstr.as_str())?;
         handle_extensions(spec, &mut client)?;
 
         // 'Close' connection
@@ -376,7 +476,7 @@ impl ComputeNode {
     // We could've wrapped this around `pg_ctl reload`, but right now we don't use
     // `pg_ctl` for start / stop, so this just seems much easier to do as we already
     // have opened connection to Postgres and superuser access.
-    #[instrument(skip(self, client))]
+    #[instrument(skip_all)]
     fn pg_reload_conf(&self, client: &mut Client) -> Result<()> {
         client.simple_query("SELECT pg_reload_conf()")?;
         Ok(())
@@ -384,13 +484,13 @@ impl ComputeNode {
 
     /// Similar to `apply_config()`, but does a bit different sequence of operations,
     /// as it's used to reconfigure a previously started and configured Postgres node.
-    #[instrument(skip(self))]
+    #[instrument(skip_all)]
     pub fn reconfigure(&self) -> Result<()> {
         let spec = self.state.lock().unwrap().pspec.clone().unwrap().spec;
 
         // Write new config
         let pgdata_path = Path::new(&self.pgdata);
-        config::write_postgres_conf(&pgdata_path.join("postgresql.conf"), &spec)?;
+        config::write_postgres_conf(&pgdata_path.join("postgresql.conf"), &spec, None)?;
 
         let mut client = Client::connect(self.connstr.as_str(), NoTls)?;
         self.pg_reload_conf(&mut client)?;
@@ -402,7 +502,7 @@ impl ComputeNode {
             handle_roles(&spec, &mut client)?;
             handle_databases(&spec, &mut client)?;
             handle_role_deletions(&spec, self.connstr.as_str(), &mut client)?;
-            handle_grants(&spec, self.connstr.as_str(), &mut client)?;
+            handle_grants(&spec, self.connstr.as_str())?;
             handle_extensions(&spec, &mut client)?;
         }
 
@@ -419,8 +519,8 @@ impl ComputeNode {
         Ok(())
     }
 
-    #[instrument(skip(self))]
+    #[instrument(skip_all)]
-    pub fn start_compute(&self) -> Result<std::process::Child> {
+    pub fn start_compute(&self, extension_server_port: u16) -> Result<std::process::Child> {
         let compute_state = self.state.lock().unwrap().clone();
         let pspec = compute_state.pspec.as_ref().expect("spec must be set");
         info!(
@@ -431,7 +531,26 @@ impl ComputeNode {
             pspec.timeline_id,
         );
 
-        self.prepare_pgdata(&compute_state)?;
+        // This part is sync, because we need to download
+        // remote shared_preload_libraries before postgres start (if any)
+        {
+            let library_load_start_time = Utc::now();
+            self.prepare_preload_libraries(&compute_state)?;
+
+            let library_load_time = Utc::now()
+                .signed_duration_since(library_load_start_time)
+                .to_std()
+                .unwrap()
+                .as_millis() as u64;
+            let mut state = self.state.lock().unwrap();
+            state.metrics.load_libraries_ms = library_load_time;
+            info!(
+                "Loading shared_preload_libraries took {:?}ms",
+                library_load_time
+            );
+        }
+
+        self.prepare_pgdata(&compute_state, extension_server_port)?;
 
         let start_time = Utc::now();
 
@@ -567,4 +686,92 @@ LIMIT 100",
             "{{\"pg_stat_statements\": []}}".to_string()
         }
     }
+
+    // If remote extension storage is configured,
+    // download extension control files
+    #[tokio::main]
+    pub async fn prepare_external_extensions(&self, compute_state: &ComputeState) -> Result<()> {
+        if let Some(ref ext_remote_storage) = self.ext_remote_storage {
+            let pspec = compute_state.pspec.as_ref().expect("spec must be set");
+            let spec = &pspec.spec;
+            let custom_ext_prefixes = spec.custom_extensions.clone().unwrap_or(Vec::new());
+            info!("custom_ext_prefixes: {:?}", &custom_ext_prefixes);
+            let available_extensions = extension_server::get_available_extensions(
+                ext_remote_storage,
+                &self.pgbin,
+                &self.pgversion,
+                &custom_ext_prefixes,
+            )
+            .await?;
+            self.available_extensions
+                .set(available_extensions)
+                .expect("available_extensions.set error");
+        }
+        Ok(())
+    }
+
+    pub async fn download_extension(&self, ext_name: &str) -> Result<()> {
+        match &self.ext_remote_storage {
+            None => anyhow::bail!("No remote extension storage"),
+            Some(remote_storage) => {
+                extension_server::download_extension(
+                    ext_name,
+                    remote_storage,
+                    &self.pgbin,
+                    &self.pgversion,
+                )
+                .await
+            }
+        }
+    }
+
+    #[tokio::main]
+    pub async fn prepare_preload_libraries(&self, compute_state: &ComputeState) -> Result<()> {
+        if self.ext_remote_storage.is_none() {
+            return Ok(());
+        }
+        let pspec = compute_state.pspec.as_ref().expect("spec must be set");
+        let spec = &pspec.spec;
+
+        info!("parse shared_preload_libraries from spec.cluster.settings");
+        let mut libs_vec = Vec::new();
+        if let Some(libs) = spec.cluster.settings.find("shared_preload_libraries") {
+            libs_vec = libs
+                .split(&[',', '\'', ' '])
+                .filter(|s| *s != "neon" && !s.is_empty())
+                .map(str::to_string)
+                .collect();
+        }
+        info!("parse shared_preload_libraries from provided postgresql.conf");
+        // that is used in neon_local and python tests
+        if let Some(conf) = &spec.cluster.postgresql_conf {
+            let conf_lines = conf.split('\n').collect::<Vec<&str>>();
+            let mut shared_preload_libraries_line = "";
+            for line in conf_lines {
+                if line.starts_with("shared_preload_libraries") {
+                    shared_preload_libraries_line = line;
+                }
+            }
+            let mut preload_libs_vec = Vec::new();
+            if let Some(libs) = shared_preload_libraries_line.split("='").nth(1) {
+                preload_libs_vec = libs
+                    .split(&[',', '\'', ' '])
+                    .filter(|s| *s != "neon" && !s.is_empty())
+                    .map(str::to_string)
+                    .collect();
+            }
+            libs_vec.extend(preload_libs_vec);
+        }
+
+        info!("Downloading to shared preload libraries: {:?}", &libs_vec);
+        let mut download_tasks = Vec::new();
+        for library in &libs_vec {
+            download_tasks.push(self.download_extension(library));
+        }
+        let results = join_all(download_tasks).await;
+        for result in results {
+            result?; // propogate any errors
+        }
+        Ok(())
+    }
 }

compute_tools/src/config.rs

@@ -33,7 +33,11 @@ pub fn line_in_file(path: &Path, line: &str) -> Result<bool> {
 }
 
 /// Create or completely rewrite configuration file specified by `path`
-pub fn write_postgres_conf(path: &Path, spec: &ComputeSpec) -> Result<()> {
+pub fn write_postgres_conf(
+    path: &Path,
+    spec: &ComputeSpec,
+    extension_server_port: Option<u16>,
+) -> Result<()> {
     // File::create() destroys the file content if it exists.
     let mut file = File::create(path)?;
 
@@ -95,5 +99,9 @@ pub fn write_postgres_conf(path: &Path, spec: &ComputeSpec) -> Result<()> {
         writeln!(file, "# Managed by compute_ctl: end")?;
     }
 
+    if let Some(port) = extension_server_port {
+        writeln!(file, "neon.extension_server_port={}", port)?;
+    }
+
     Ok(())
 }

compute_tools/src/configurator.rs

@@ -8,7 +8,7 @@ use compute_api::responses::ComputeStatus;
 
 use crate::compute::ComputeNode;
 
-#[instrument(skip(compute))]
+#[instrument(skip_all)]
 fn configurator_main_loop(compute: &Arc<ComputeNode>) {
     info!("waiting for reconfiguration requests");
     loop {
@@ -42,13 +42,15 @@ fn configurator_main_loop(compute: &Arc<ComputeNode>) {
     }
 }
 
-pub fn launch_configurator(compute: &Arc<ComputeNode>) -> Result<thread::JoinHandle<()>> {
+pub fn launch_configurator(
+    compute: &Arc<ComputeNode>,
+) -> Result<thread::JoinHandle<()>, std::io::Error> {
     let compute = Arc::clone(compute);
 
-    Ok(thread::Builder::new()
+    thread::Builder::new()
         .name("compute-configurator".into())
         .spawn(move || {
             configurator_main_loop(&compute);
             info!("configurator thread is exited");
-        })?)
+        })
 }

compute_tools/src/extension_server.rs

@@ -0,0 +1,237 @@
+// Download extension files from the extension store
+// and put them in the right place in the postgres directory
+/*
+The layout of the S3 bucket is as follows:
+
+v14/ext_index.json
+    -- this contains information necessary to create control files
+v14/extensions/test_ext1.tar.gz
+    -- this contains the library files and sql files necessary to create this extension
+v14/extensions/custom_ext1.tar.gz
+
+The difference between a private and public extensions is determined by who can
+load the extension this is specified in ext_index.json
+
+Speicially, ext_index.json has a list of public extensions, and a list of
+extensions enabled for specific tenant-ids.
+*/
+use crate::compute::ComputeNode;
+use anyhow::Context;
+use anyhow::{self, Result};
+use flate2::read::GzDecoder;
+use remote_storage::*;
+use serde_json::{self, Value};
+use std::collections::HashSet;
+use std::num::{NonZeroU32, NonZeroUsize};
+use std::path::Path;
+use std::str;
+use std::sync::Arc;
+use std::thread;
+use tar::Archive;
+use tokio::io::AsyncReadExt;
+use tracing::info;
+
+fn get_pg_config(argument: &str, pgbin: &str) -> String {
+    // gives the result of `pg_config [argument]`
+    // where argument is a flag like `--version` or `--sharedir`
+    let pgconfig = pgbin
+        .strip_suffix("postgres")
+        .expect("bad pgbin")
+        .to_owned()
+        + "/pg_config";
+    let config_output = std::process::Command::new(pgconfig)
+        .arg(argument)
+        .output()
+        .expect("pg_config error");
+    std::str::from_utf8(&config_output.stdout)
+        .expect("pg_config error")
+        .trim()
+        .to_string()
+}
+
+pub fn get_pg_version(pgbin: &str) -> String {
+    // pg_config --version returns a (platform specific) human readable string
+    // such as "PostgreSQL 15.4". We parse this to v14/v15
+    let human_version = get_pg_config("--version", pgbin);
+    if human_version.contains("15") {
+        return "v15".to_string();
+    } else if human_version.contains("14") {
+        return "v14".to_string();
+    }
+    panic!("Unsuported postgres version {human_version}");
+}
+
+// download extension control files
+// if custom_ext_prefixes is provided - search also in custom extension paths
+pub async fn get_available_extensions(
+    remote_storage: &GenericRemoteStorage,
+    pgbin: &str,
+    pg_version: &str,
+    custom_ext_prefixes: &[String],
+) -> Result<HashSet<String>> {
+    let local_sharedir = Path::new(&get_pg_config("--sharedir", pgbin)).join("extension");
+    let index_path = pg_version.to_owned() + "/ext_index.json";
+    let index_path = RemotePath::new(Path::new(&index_path)).context("error forming path")?;
+    info!("download ext_index.json: {:?}", &index_path);
+
+    // TODO: potential optimization: cache ext_index.json
+    let mut download = remote_storage.download(&index_path).await?;
+    let mut write_data_buffer = Vec::new();
+    download
+        .download_stream
+        .read_to_end(&mut write_data_buffer)
+        .await?;
+    let ext_index_str = match str::from_utf8(&write_data_buffer) {
+        Ok(v) => v,
+        Err(e) => panic!("Invalid UTF-8 sequence: {}", e),
+    };
+
+    let ext_index_full: Value = serde_json::from_str(ext_index_str)?;
+    let ext_index_full = ext_index_full.as_object().context("error parsing json")?;
+    let control_data = ext_index_full["control_data"]
+        .as_object()
+        .context("json parse error")?;
+    let enabled_extensions = ext_index_full["enabled_extensions"]
+        .as_object()
+        .context("json parse error")?;
+    info!("{:?}", control_data.clone());
+    info!("{:?}", enabled_extensions.clone());
+
+    let mut prefixes = vec!["public".to_string()];
+    prefixes.extend(custom_ext_prefixes.to_owned());
+    info!("{:?}", &prefixes);
+    let mut all_extensions = HashSet::new();
+    for prefix in prefixes {
+        let prefix_extensions = match enabled_extensions.get(&prefix) {
+            Some(Value::Array(ext_name)) => ext_name,
+            _ => {
+                info!("prefix {} has no extensions", prefix);
+                continue;
+            }
+        };
+        info!("{:?}", prefix_extensions);
+        for ext_name in prefix_extensions {
+            all_extensions.insert(ext_name.as_str().context("json parse error")?.to_string());
+        }
+    }
+
+    for prefix in &all_extensions {
+        let control_contents = control_data[prefix].as_str().context("json parse error")?;
+        let control_path = local_sharedir.join(prefix.to_owned() + ".control");
+
+        info!("WRITING FILE {:?}{:?}", control_path, control_contents);
+        std::fs::write(control_path, control_contents)?;
+    }
+
+    Ok(all_extensions.into_iter().collect())
+}
+
+// download all sqlfiles (and possibly data files) for a given extension name
+pub async fn download_extension(
+    ext_name: &str,
+    remote_storage: &GenericRemoteStorage,
+    pgbin: &str,
+    pg_version: &str,
+) -> Result<()> {
+    // TODO: potential optimization: only download the extension if it doesn't exist
+    // problem: how would we tell if it exists?
+    let ext_name = ext_name.replace(".so", "");
+    let ext_name_targz = ext_name.to_owned() + ".tar.gz";
+    if Path::new(&ext_name_targz).exists() {
+        info!("extension {:?} already exists", ext_name_targz);
+        return Ok(());
+    }
+    let ext_path = RemotePath::new(
+        &Path::new(pg_version)
+            .join("extensions")
+            .join(ext_name_targz.clone()),
+    )?;
+    info!(
+        "Start downloading extension {:?} from {:?}",
+        ext_name, ext_path
+    );
+    let mut download = remote_storage.download(&ext_path).await?;
+    let mut write_data_buffer = Vec::new();
+    download
+        .download_stream
+        .read_to_end(&mut write_data_buffer)
+        .await?;
+    let unzip_dest = pgbin.strip_suffix("/bin/postgres").expect("bad pgbin");
+    let tar = GzDecoder::new(std::io::Cursor::new(write_data_buffer));
+    let mut archive = Archive::new(tar);
+    archive.unpack(unzip_dest)?;
+    info!("Download + unzip {:?} completed successfully", &ext_path);
+
+    let local_sharedir = Path::new(&get_pg_config("--sharedir", pgbin)).join("extension");
+    let zip_sharedir = format!("{unzip_dest}/extensions/{ext_name}/share/extension");
+    info!("mv {zip_sharedir:?}/* {local_sharedir:?}");
+    for file in std::fs::read_dir(zip_sharedir)? {
+        let old_file = file?.path();
+        let new_file =
+            Path::new(&local_sharedir).join(old_file.file_name().context("error parsing file")?);
+        std::fs::rename(old_file, new_file)?;
+    }
+    let local_libdir = Path::new(&get_pg_config("--libdir", pgbin)).join("postgresql");
+    let zip_libdir = format!("{unzip_dest}/extensions/{ext_name}/lib");
+    info!("mv {zip_libdir:?}/* {local_libdir:?}");
+    for file in std::fs::read_dir(zip_libdir)? {
+        let old_file = file?.path();
+        let new_file =
+            Path::new(&local_libdir).join(old_file.file_name().context("error parsing file")?);
+        std::fs::rename(old_file, new_file)?;
+    }
+    Ok(())
+}
+
+// This function initializes the necessary structs to use remmote storage (should be fairly cheap)
+pub fn init_remote_storage(
+    remote_ext_config: &str,
+    default_prefix: &str,
+) -> anyhow::Result<GenericRemoteStorage> {
+    let remote_ext_config: serde_json::Value = serde_json::from_str(remote_ext_config)?;
+
+    let remote_ext_bucket = remote_ext_config["bucket"]
+        .as_str()
+        .context("config parse error")?;
+    let remote_ext_region = remote_ext_config["region"]
+        .as_str()
+        .context("config parse error")?;
+    let remote_ext_endpoint = remote_ext_config["endpoint"].as_str();
+    let remote_ext_prefix = remote_ext_config["prefix"]
+        .as_str()
+        .unwrap_or(default_prefix)
+        .to_string();
+
+    // TODO: potentially allow modification of other parameters
+    // however, default values should be fine for now
+    let config = S3Config {
+        bucket_name: remote_ext_bucket.to_string(),
+        bucket_region: remote_ext_region.to_string(),
+        prefix_in_bucket: Some(remote_ext_prefix),
+        endpoint: remote_ext_endpoint.map(|x| x.to_string()),
+        concurrency_limit: NonZeroUsize::new(100).expect("100 != 0"),
+        max_keys_per_list_response: None,
+    };
+    let config = RemoteStorageConfig {
+        max_concurrent_syncs: NonZeroUsize::new(100).expect("100 != 0"),
+        max_sync_errors: NonZeroU32::new(100).expect("100 != 0"),
+        storage: RemoteStorageKind::AwsS3(config),
+    };
+    GenericRemoteStorage::from_config(&config)
+}
+
+pub fn launch_download_extensions(
+    compute: &Arc<ComputeNode>,
+) -> Result<thread::JoinHandle<()>, std::io::Error> {
+    let compute = Arc::clone(compute);
+    thread::Builder::new()
+        .name("download-extensions".into())
+        .spawn(move || {
+            info!("start download_extension_files");
+            let compute_state = compute.state.lock().expect("error unlocking compute.state");
+            compute
+                .prepare_external_extensions(&compute_state)
+                .expect("error preparing extensions");
+            info!("download_extension_files done, exiting thread");
+        })
+}

compute_tools/src/http/api.rs

@@ -121,6 +121,27 @@ async fn routes(req: Request<Body>, compute: &Arc<ComputeNode>) -> Response<Body
             }
         }
 
+        // download extension files from S3 on demand
+        (&Method::POST, route) if route.starts_with("/extension_server/") => {
+            info!("serving {:?} POST request", route);
+            info!("req.uri {:?}", req.uri());
+            let filename = route.split('/').last().unwrap().to_string();
+            info!(
+                "serving /extension_server POST request, filename: {:?}",
+                &filename
+            );
+
+            match compute.download_extension(&filename).await {
+                Ok(_) => Response::new(Body::from("OK")),
+                Err(e) => {
+                    error!("extension download failed: {}", e);
+                    let mut resp = Response::new(Body::from(e.to_string()));
+                    *resp.status_mut() = StatusCode::INTERNAL_SERVER_ERROR;
+                    resp
+                }
+            }
+        }
+
         // Return the `404 Not Found` for any other routes.
         _ => {
             let mut not_found = Response::new(Body::from("404 Not Found"));

compute_tools/src/http/openapi_spec.yaml

@@ -139,6 +139,34 @@ paths:
             application/json:
               schema:
                 $ref: "#/components/schemas/GenericError"
+  /extension_server:
+    post:
+      tags:
+      - Extension
+      summary: Download extension from S3 to local folder.
+      description: ""
+      operationId: downloadExtension
+      responses:
+        200:
+          description: Extension downloaded
+          content:
+            text/plain:
+              schema:
+                type: string
+                description: Error text or 'OK' if download succeeded.
+                example: "OK"
+        400:
+        description: Request is invalid.
+        content:
+          application/json:
+            schema:
+              $ref: "#/components/schemas/GenericError"
+        500:
+        description: Extension download request failed.
+        content:
+          application/json:
+            schema:
+              $ref: "#/components/schemas/GenericError"
 
 components:
   securitySchemes:

compute_tools/src/lib.rs

@@ -9,6 +9,7 @@ pub mod http;
 #[macro_use]
 pub mod logger;
 pub mod compute;
+pub mod extension_server;
 pub mod monitor;
 pub mod params;
 pub mod pg_helpers;

compute_tools/src/logger.rs

@@ -18,6 +18,7 @@ pub fn init_tracing_and_logging(default_log_level: &str) -> anyhow::Result<()> {
         .unwrap_or_else(|_| tracing_subscriber::EnvFilter::new(default_log_level));
 
     let fmt_layer = tracing_subscriber::fmt::layer()
+        .with_ansi(false)
         .with_target(false)
         .with_writer(std::io::stderr);
 

compute_tools/src/monitor.rs

@@ -105,10 +105,10 @@ fn watch_compute_activity(compute: &ComputeNode) {
 }
 
 /// Launch a separate compute monitor thread and return its `JoinHandle`.
-pub fn launch_monitor(state: &Arc<ComputeNode>) -> Result<thread::JoinHandle<()>> {
+pub fn launch_monitor(state: &Arc<ComputeNode>) -> Result<thread::JoinHandle<()>, std::io::Error> {
     let state = Arc::clone(state);
 
-    Ok(thread::Builder::new()
+    thread::Builder::new()
         .name("compute-monitor".into())
-        .spawn(move || watch_compute_activity(&state))?)
+        .spawn(move || watch_compute_activity(&state))
 }

compute_tools/src/pg_helpers.rs

@@ -17,7 +17,7 @@ use compute_api::spec::{Database, GenericOption, GenericOptions, PgIdent, Role};
 const POSTGRES_WAIT_TIMEOUT: Duration = Duration::from_millis(60 * 1000); // milliseconds
 
 /// Escape a string for including it in a SQL literal
-fn escape_literal(s: &str) -> String {
+pub fn escape_literal(s: &str) -> String {
     s.replace('\'', "''").replace('\\', "\\\\")
 }
 
@@ -215,7 +215,7 @@ pub fn get_existing_dbs(client: &mut Client) -> Result<Vec<Database>> {
 /// Wait for Postgres to become ready to accept connections. It's ready to
 /// accept connections when the state-field in `pgdata/postmaster.pid` says
 /// 'ready'.
-#[instrument(skip(pg))]
+#[instrument(skip_all, fields(pgdata = %pgdata.display()))]
 pub fn wait_for_postgres(pg: &mut Child, pgdata: &Path) -> Result<()> {
     let pid_path = pgdata.join("postmaster.pid");
 

compute_tools/src/spec.rs

@@ -124,7 +124,7 @@ pub fn get_spec_from_control_plane(
 pub fn handle_configuration(spec: &ComputeSpec, pgdata_path: &Path) -> Result<()> {
     // File `postgresql.conf` is no longer included into `basebackup`, so just
     // always write all config into it creating new file.
-    config::write_postgres_conf(&pgdata_path.join("postgresql.conf"), spec)?;
+    config::write_postgres_conf(&pgdata_path.join("postgresql.conf"), spec, None)?;
 
     update_pg_hba(pgdata_path)?;
 
@@ -269,17 +269,13 @@ pub fn handle_roles(spec: &ComputeSpec, client: &mut Client) -> Result<()> {
                 xact.execute(query.as_str(), &[])?;
             }
             RoleAction::Create => {
-                let mut query: String = format!("CREATE ROLE {} ", name.pg_quote());
+                let mut query: String = format!(
+                    "CREATE ROLE {} CREATEROLE CREATEDB IN ROLE neon_superuser",
+                    name.pg_quote()
+                );
                 info!("role create query: '{}'", &query);
                 query.push_str(&role.to_pg_options());
                 xact.execute(query.as_str(), &[])?;
-
-                let grant_query = format!(
-                    "GRANT pg_read_all_data, pg_write_all_data TO {}",
-                    name.pg_quote()
-                );
-                xact.execute(grant_query.as_str(), &[])?;
-                info!("role grant query: '{}'", &grant_query);
             }
         }
 
@@ -476,6 +472,11 @@ pub fn handle_databases(spec: &ComputeSpec, client: &mut Client) -> Result<()> {
                 query.push_str(&db.to_pg_options());
                 let _guard = info_span!("executing", query).entered();
                 client.execute(query.as_str(), &[])?;
+                let grant_query: String = format!(
+                    "GRANT ALL PRIVILEGES ON DATABASE {} TO neon_superuser",
+                    name.pg_quote()
+                );
+                client.execute(grant_query.as_str(), &[])?;
             }
         };
 
@@ -495,35 +496,9 @@ pub fn handle_databases(spec: &ComputeSpec, client: &mut Client) -> Result<()> {
 /// Grant CREATE ON DATABASE to the database owner and do some other alters and grants
 /// to allow users creating trusted extensions and re-creating `public` schema, for example.
 #[instrument(skip_all)]
-pub fn handle_grants(spec: &ComputeSpec, connstr: &str, client: &mut Client) -> Result<()> {
+pub fn handle_grants(spec: &ComputeSpec, connstr: &str) -> Result<()> {
     info!("cluster spec grants:");
 
-    // We now have a separate `web_access` role to connect to the database
-    // via the web interface and proxy link auth. And also we grant a
-    // read / write all data privilege to every role. So also grant
-    // create to everyone.
-    // XXX: later we should stop messing with Postgres ACL in such horrible
-    // ways.
-    let roles = spec
-        .cluster
-        .roles
-        .iter()
-        .map(|r| r.name.pg_quote())
-        .collect::<Vec<_>>();
-
-    for db in &spec.cluster.databases {
-        let dbname = &db.name;
-
-        let query: String = format!(
-            "GRANT CREATE ON DATABASE {} TO {}",
-            dbname.pg_quote(),
-            roles.join(", ")
-        );
-        info!("grant query {}", &query);
-
-        client.execute(query.as_str(), &[])?;
-    }
-
     // Do some per-database access adjustments. We'd better do this at db creation time,
     // but CREATE DATABASE isn't transactional. So we cannot create db + do some grants
     // atomically.

control_plane/Cargo.toml

@@ -32,3 +32,4 @@ utils.workspace = true
 
 compute_api.workspace = true
 workspace_hack.workspace = true
+tracing.workspace = true

control_plane/src/background_process.rs

@@ -180,6 +180,11 @@ pub fn stop_process(immediate: bool, process_name: &str, pid_file: &Path) -> any
     }
 
     // Wait until process is gone
+    wait_until_stopped(process_name, pid)?;
+    Ok(())
+}
+
+pub fn wait_until_stopped(process_name: &str, pid: Pid) -> anyhow::Result<()> {
     for retries in 0..RETRIES {
         match process_has_stopped(pid) {
             Ok(true) => {

control_plane/src/bin/neon_local.rs

@@ -308,7 +308,8 @@ fn handle_init(init_match: &ArgMatches) -> anyhow::Result<LocalEnv> {
 
     let mut env =
         LocalEnv::parse_config(&toml_file).context("Failed to create neon configuration")?;
-    env.init(pg_version)
+    let force = init_match.get_flag("force");
+    env.init(pg_version, force)
         .context("Failed to initialize neon repository")?;
 
     // Initialize pageserver, create initial tenant and timeline.
@@ -657,6 +658,8 @@ fn handle_endpoint(ep_match: &ArgMatches, env: &local_env::LocalEnv) -> Result<(
                 .get_one::<String>("endpoint_id")
                 .ok_or_else(|| anyhow!("No endpoint ID was provided to start"))?;
 
+            let remote_ext_config = sub_args.get_one::<String>("remote-ext-config");
+
             // If --safekeepers argument is given, use only the listed safekeeper nodes.
             let safekeepers =
                 if let Some(safekeepers_str) = sub_args.get_one::<String>("safekeepers") {
@@ -698,7 +701,7 @@ fn handle_endpoint(ep_match: &ArgMatches, env: &local_env::LocalEnv) -> Result<(
                     _ => {}
                 }
                 println!("Starting existing endpoint {endpoint_id}...");
-                endpoint.start(&auth_token, safekeepers)?;
+                endpoint.start(&auth_token, safekeepers, remote_ext_config)?;
             } else {
                 let branch_name = sub_args
                     .get_one::<String>("branch-name")
@@ -742,7 +745,7 @@ fn handle_endpoint(ep_match: &ArgMatches, env: &local_env::LocalEnv) -> Result<(
                     pg_version,
                     mode,
                 )?;
-                ep.start(&auth_token, safekeepers)?;
+                ep.start(&auth_token, safekeepers, remote_ext_config)?;
             }
         }
         "stop" => {
@@ -1002,6 +1005,12 @@ fn cli() -> Command {
         .help("Additional pageserver's configuration options or overrides, refer to pageserver's 'config-override' CLI parameter docs for more")
         .required(false);
 
+    let remote_ext_config_args = Arg::new("remote-ext-config")
+        .long("remote-ext-config")
+        .num_args(1)
+        .help("Configure the S3 bucket that we search for extensions in.")
+        .required(false);
+
     let lsn_arg = Arg::new("lsn")
         .long("lsn")
         .help("Specify Lsn on the timeline to start from. By default, end of the timeline would be used.")
@@ -1013,6 +1022,13 @@ fn cli() -> Command {
         .help("If set, the node will be a hot replica on the specified timeline")
         .required(false);
 
+    let force_arg = Arg::new("force")
+        .value_parser(value_parser!(bool))
+        .long("force")
+        .action(ArgAction::SetTrue)
+        .help("Force initialization even if the repository is not empty")
+        .required(false);
+
     Command::new("Neon CLI")
         .arg_required_else_help(true)
         .version(GIT_VERSION)
@@ -1028,6 +1044,7 @@ fn cli() -> Command {
                         .value_name("config"),
                 )
                 .arg(pg_version_arg.clone())
+                .arg(force_arg)
         )
         .subcommand(
             Command::new("timeline")
@@ -1152,6 +1169,7 @@ fn cli() -> Command {
                     .arg(pg_version_arg)
                     .arg(hot_standby_arg)
                     .arg(safekeepers_arg)
+                    .arg(remote_ext_config_args)
                 )
                 .subcommand(
                     Command::new("stop")

control_plane/src/endpoint.rs

@@ -311,7 +311,7 @@ impl Endpoint {
 
                 // TODO: use future host field from safekeeper spec
                 // Pass the list of safekeepers to the replica so that it can connect to any of them,
-                // whichever is availiable.
+                // whichever is available.
                 let sk_ports = self
                     .env
                     .safekeepers
@@ -405,10 +405,25 @@ impl Endpoint {
                 String::from_utf8_lossy(&pg_ctl.stderr),
             );
         }
+
+        // Also wait for the compute_ctl process to die. It might have some cleanup
+        // work to do after postgres stops, like syncing safekeepers, etc.
+        //
+        // TODO use background_process::stop_process instead
+        let pidfile_path = self.endpoint_path().join("compute_ctl.pid");
+        let pid: u32 = std::fs::read_to_string(pidfile_path)?.parse()?;
+        let pid = nix::unistd::Pid::from_raw(pid as i32);
+        crate::background_process::wait_until_stopped("compute_ctl", pid)?;
+
         Ok(())
     }
 
-    pub fn start(&self, auth_token: &Option<String>, safekeepers: Vec<NodeId>) -> Result<()> {
+    pub fn start(
+        &self,
+        auth_token: &Option<String>,
+        safekeepers: Vec<NodeId>,
+        remote_ext_config: Option<&String>,
+    ) -> Result<()> {
         if self.status() == "running" {
             anyhow::bail!("The endpoint is already running");
         }
@@ -476,6 +491,15 @@ impl Endpoint {
             pageserver_connstring: Some(pageserver_connstring),
             safekeeper_connstrings,
             storage_auth_token: auth_token.clone(),
+            // TODO FIXME: This is a hack to test custom extensions locally.
+            // In test_download_extensions, we assume that the custom extension
+            // prefix is the tenant ID. So we set it here.
+            //
+            // The proper way to implement this is to pass the custom extension
+            // in spec, but we don't have a way to do that yet in the python tests.
+            // NEW HACK: we enable the anon custom extension for everyone! this is of course just for testing
+            // how will we do it for real?
+            custom_extensions: Some(vec!["123454321".to_string(), self.tenant_id.to_string()]),
         };
         let spec_path = self.endpoint_path().join("spec.json");
         std::fs::write(spec_path, serde_json::to_string_pretty(&spec)?)?;
@@ -507,7 +531,18 @@ impl Endpoint {
             .stdin(std::process::Stdio::null())
             .stderr(logfile.try_clone()?)
             .stdout(logfile);
-        let _child = cmd.spawn()?;
+
+        if let Some(remote_ext_config) = remote_ext_config {
+            cmd.args(["--remote-ext-config", remote_ext_config]);
+        }
+
+        let child = cmd.spawn()?;
+
+        // Write down the pid so we can wait for it when we want to stop
+        // TODO use background_process::start_process instead
+        let pid = child.id();
+        let pidfile_path = self.endpoint_path().join("compute_ctl.pid");
+        std::fs::write(pidfile_path, pid.to_string())?;
 
         // Wait for it to start
         let mut attempt = 0;

control_plane/src/local_env.rs

@@ -364,7 +364,7 @@ impl LocalEnv {
     //
     // Initialize a new Neon repository
     //
-    pub fn init(&mut self, pg_version: u32) -> anyhow::Result<()> {
+    pub fn init(&mut self, pg_version: u32, force: bool) -> anyhow::Result<()> {
         // check if config already exists
         let base_path = &self.base_data_dir;
         ensure!(
@@ -372,11 +372,29 @@ impl LocalEnv {
             "repository base path is missing"
         );
 
-        ensure!(
+        if base_path.exists() {
-            !base_path.exists(),
+            if force {
-            "directory '{}' already exists. Perhaps already initialized?",
+                println!("removing all contents of '{}'", base_path.display());
-            base_path.display()
+                // instead of directly calling `remove_dir_all`, we keep the original dir but removing
-        );
+                // all contents inside. This helps if the developer symbol links another directory (i.e.,
+                // S3 local SSD) to the `.neon` base directory.
+                for entry in std::fs::read_dir(base_path)? {
+                    let entry = entry?;
+                    let path = entry.path();
+                    if path.is_dir() {
+                        fs::remove_dir_all(&path)?;
+                    } else {
+                        fs::remove_file(&path)?;
+                    }
+                }
+            } else {
+                bail!(
+                    "directory '{}' already exists. Perhaps already initialized? (Hint: use --force to remove all contents)",
+                    base_path.display()
+                );
+            }
+        }
+
         if !self.pg_bin_dir(pg_version)?.join("postgres").exists() {
             bail!(
                 "Can't find postgres binary at {}",
@@ -392,7 +410,9 @@ impl LocalEnv {
             }
         }
 
-        fs::create_dir(base_path)?;
+        if !base_path.exists() {
+            fs::create_dir(base_path)?;
+        }
 
         // Generate keypair for JWT.
         //

docs/rfcs/024-extension-loading.md

@@ -0,0 +1,183 @@
+# Supporting custom user Extensions (Dynamic Extension Loading)
+Created 2023-05-03
+
+## Motivation
+
+There are many extensions in the PostgreSQL ecosystem, and not all extensions
+are of a quality that we can confidently support them. Additionally, our
+current extension inclusion mechanism has several problems because we build all
+extensions into the primary Compute image: We build the extensions every time
+we build the compute image regardless of whether we actually need to rebuild
+the image, and the inclusion of these extensions in the image adds a hard
+dependency on all supported extensions - thus increasing the image size, and
+with it the time it takes to download that image - increasing first start
+latency.
+
+This RFC proposes a dynamic loading mechanism that solves most of these
+problems.
+
+## Summary
+
+`compute_ctl` is made responsible for loading extensions on-demand into
+the container's file system for dynamically loaded extensions, and will also
+make sure that the extensions in `shared_preload_libraries` are downloaded
+before the compute node starts.
+
+## Components
+
+compute_ctl, PostgreSQL, neon (extension), Compute Host Node, Extension Store
+
+## Requirements
+
+Compute nodes with no extra extensions should not be negatively impacted by
+the existence of support for many extensions.
+
+Installing an extension into PostgreSQL should be easy.
+
+Non-preloaded extensions shouldn't impact startup latency.
+
+Uninstalled extensions shouldn't impact query latency.
+
+A small latency penalty for dynamically loaded extensions is acceptable in
+the first seconds of compute startup, but not in steady-state operations.
+
+## Proposed implementation
+
+### On-demand, JIT-loading of extensions
+
+Before postgres starts we download 
+- control files for all extensions available to that compute node;
+- all `shared_preload_libraries`;
+
+After postgres is running, `compute_ctl` listens for requests to load files.
+When PostgreSQL requests a file, `compute_ctl` downloads it.
+
+PostgreSQL requests files in the following cases:
+- When loading a preload library set in `local_preload_libraries`
+- When explicitly loading a library with `LOAD`
+- Wnen creating extension with `CREATE EXTENSION` (download sql scripts, (optional) extension data files and (optional) library files)))
+
+
+#### Summary
+
+Pros:
+ - Startup is only as slow as it takes to load all (shared_)preload_libraries
+ - Supports BYO Extension
+
+Cons:
+ - O(sizeof(extensions)) IO requirement for loading all extensions.
+
+### Alternative solutions
+
+1. Allow users to add their extensions to the base image
+   
+   Pros:
+    - Easy to deploy
+
+   Cons:
+    - Doesn't scale - first start size is dependent on image size;
+    - All extensions are shared across all users: It doesn't allow users to
+      bring their own restrictive-licensed extensions
+
+2. Bring Your Own compute image
+   
+   Pros:
+    - Still easy to deploy
+    - User can bring own patched version of PostgreSQL
+
+   Cons:
+    - First start latency is O(sizeof(extensions image))
+    - Warm instance pool for skipping pod schedule latency is not feasible with
+      O(n) custom images
+    - Support channels are difficult to manage
+
+3. Download all user extensions in bulk on compute start
+   
+   Pros:
+    - Easy to deploy
+    - No startup latency issues for "clean" users.
+    - Warm instance pool for skipping pod schedule latency is possible
+
+   Cons:
+    - Downloading all extensions in advance takes a lot of time, thus startup
+      latency issues
+
+4. Store user's extensions in persistent storage
+   
+   Pros:
+    - Easy to deploy
+    - No startup latency issues
+    - Warm instance pool for skipping pod schedule latency is possible
+
+   Cons:
+    - EC2 instances have only limited number of attachments shared between EBS
+      volumes, direct-attached NVMe drives, and ENIs.
+    - Compute instance migration isn't trivially solved for EBS mounts (e.g.
+      the device is unavailable whilst moving the mount between instances).
+    - EBS can only mount on one instance at a time (except the expensive IO2
+      device type).
+
+5. Store user's extensions in network drive
+   
+   Pros:
+    - Easy to deploy
+    - Few startup latency issues
+    - Warm instance pool for skipping pod schedule latency is possible
+
+   Cons:
+    - We'd need networked drives, and a lot of them, which would store many
+      duplicate extensions.
+    - **UNCHECKED:** Compute instance migration may not work nicely with
+      networked IOs
+
+
+### Idea extensions
+
+The extension store does not have to be S3 directly, but could be a Node-local
+caching service on top of S3. This would reduce the load on the network for
+popular extensions.
+
+## Extension Storage implementation
+
+Extension Storage in our case is an S3 bucket with a "directory" per build and postgres version,
+where extension files are stored as plain files in the bucket following the same directory structure as in the postgres.
+
+i.e.
+
+`s3://<the-bucket>/<build-version>/<postgres-version>/lib/postgis-3.1.so`
+`s3://<the-bucket>/<build-version>/<postgres-version>/share/extension/postgis.control`
+`s3://<the-bucket>/<build-version>/<postgres-version>/share/extension/postgis--3.1.sql`
+
+To handle custom extensions, that available only to specific users, we use per-extension subdirectories:
+
+i.e.
+`s3://<the-bucket>/<build-version>/<postgres-version>/<custom-ext-prefix>/lib/ext-name.so`, etc.
+`s3://<the-bucket>/<build-version>/<postgres-version>/<custom-ext-prefix>/share/extension/ext-name.control`, etc.
+
+On compute start, `compute_ctl` accepts a list of custom_ext_prefixes.
+
+To get the list of available extensions,`compute_ctl` downloads control files from all prefixes:
+
+`s3://<the-bucket>/<build-version>/<postgres-version>/share/extension/`
+`s3://<the-bucket>/<build-version>/<postgres-version>/<custom-ext-prefix1>/share/extension/`
+`s3://<the-bucket>/<build-version>/<postgres-version>/<custom-ext-prefix2>/share/extension/`
+
+
+
+### How to add new extension to the Extension Storage?
+
+Simply upload build artifacts to the S3 bucket.
+Implement a CI step for that. Splitting it from ompute-node-image build.
+
+### How do we deal with extension versions and updates?
+
+Currently, we rebuild extensions on every compute-node-image build and store them in the <build-version> prefix.
+This is needed to ensure that `/share` and `/lib` files are in sync.
+
+For extension updates, we rely on the PostgreSQL extension versioning mechanism (sql update scripts) and extension authors to not break backwards compatibility within one major version of PostgreSQL.
+
+### Alternatives
+
+For extensions written on trusted languages we can also adopt
+`dbdev` PostgreSQL Package Manager based on `pg_tle` by Supabase.
+This will increase the amount supported extensions and decrease the amount of work required to support them.

libs/compute_api/src/responses.rs

@@ -73,6 +73,7 @@ pub struct ComputeMetrics {
     pub basebackup_ms: u64,
     pub config_ms: u64,
     pub total_startup_ms: u64,
+    pub load_libraries_ms: u64,
 }
 
 /// Response of the `/computes/{compute_id}/spec` control-plane API.

libs/compute_api/src/spec.rs

@@ -60,6 +60,9 @@ pub struct ComputeSpec {
     /// If set, 'storage_auth_token' is used as the password to authenticate to
     /// the pageserver and safekeepers.
     pub storage_auth_token: Option<String>,
+
+    // list of prefixes to search for custom extensions in remote extension storage
+    pub custom_extensions: Option<Vec<String>>,
 }
 
 #[serde_as]

libs/metrics/src/lib.rs

@@ -23,6 +23,7 @@ use prometheus::{Registry, Result};
 pub mod launch_timestamp;
 mod wrappers;
 pub use wrappers::{CountedReader, CountedWriter};
+pub mod metric_vec_duration;
 
 pub type UIntGauge = GenericGauge<AtomicU64>;
 pub type UIntGaugeVec = GenericGaugeVec<AtomicU64>;

libs/metrics/src/metric_vec_duration.rs

@@ -0,0 +1,23 @@
+//! Helpers for observing duration on HistogramVec / CounterVec / GaugeVec / MetricVec<T>.
+
+use std::{future::Future, time::Instant};
+
+pub trait DurationResultObserver {
+    fn observe_result<T, E>(&self, res: &Result<T, E>, duration: std::time::Duration);
+}
+
+pub async fn observe_async_block_duration_by_result<
+    T,
+    E,
+    F: Future<Output = Result<T, E>>,
+    O: DurationResultObserver,
+>(
+    observer: &O,
+    block: F,
+) -> Result<T, E> {
+    let start = Instant::now();
+    let result = block.await;
+    let duration = start.elapsed();
+    observer.observe_result(&result, duration);
+    result
+}

libs/remote_storage/src/lib.rs

@@ -184,6 +184,20 @@ pub enum GenericRemoteStorage {
 }
 
 impl GenericRemoteStorage {
+    // A function for listing all the files in a "directory"
+    // Example:
+    // list_files("foo/bar") = ["foo/bar/a.txt", "foo/bar/b.txt"]
+    pub async fn list_files(&self, folder: Option<&RemotePath>) -> anyhow::Result<Vec<RemotePath>> {
+        match self {
+            Self::LocalFs(s) => s.list_files(folder).await,
+            Self::AwsS3(s) => s.list_files(folder).await,
+            Self::Unreliable(s) => s.list_files(folder).await,
+        }
+    }
+
+    // lists common *prefixes*, if any of files
+    // Example:
+    // list_prefixes("foo123","foo567","bar123","bar432") = ["foo", "bar"]
     pub async fn list_prefixes(
         &self,
         prefix: Option<&RemotePath>,
@@ -195,14 +209,6 @@ impl GenericRemoteStorage {
         }
     }
 
-    pub async fn list_files(&self, folder: Option<&RemotePath>) -> anyhow::Result<Vec<RemotePath>> {
-        match self {
-            Self::LocalFs(s) => s.list_files(folder).await,
-            Self::AwsS3(s) => s.list_files(folder).await,
-            Self::Unreliable(s) => s.list_files(folder).await,
-        }
-    }
-
     pub async fn upload(
         &self,
         from: impl io::AsyncRead + Unpin + Send + Sync + 'static,

libs/remote_storage/src/s3_bucket.rs

@@ -349,10 +349,17 @@ impl RemoteStorage for S3Bucket {
 
     /// See the doc for `RemoteStorage::list_files`
     async fn list_files(&self, folder: Option<&RemotePath>) -> anyhow::Result<Vec<RemotePath>> {
-        let folder_name = folder
+        let mut folder_name = folder
             .map(|p| self.relative_path_to_s3_object(p))
             .or_else(|| self.prefix_in_bucket.clone());
 
+        // remove leading "/" if one exists
+        if let Some(folder_name_slash) = folder_name.clone() {
+            if folder_name_slash.starts_with(REMOTE_STORAGE_PREFIX_SEPARATOR) {
+                folder_name = Some(folder_name_slash[1..].to_string());
+            }
+        }
+
         // AWS may need to break the response into several parts
         let mut continuation_token = None;
         let mut all_files = vec![];

libs/remote_storage/tests/test_real_s3.rs

@@ -173,10 +173,15 @@ async fn s3_delete_objects_works(ctx: &mut MaybeEnabledS3) -> anyhow::Result<()>
     let path2 = RemotePath::new(&PathBuf::from(format!("{}/path2", ctx.base_prefix,)))
         .with_context(|| "RemotePath conversion")?;
 
+    let path3 = RemotePath::new(&PathBuf::from(format!("{}/path3", ctx.base_prefix,)))
+        .with_context(|| "RemotePath conversion")?;
+
     let data1 = "remote blob data1".as_bytes();
     let data1_len = data1.len();
     let data2 = "remote blob data2".as_bytes();
     let data2_len = data2.len();
+    let data3 = "remote blob data3".as_bytes();
+    let data3_len = data3.len();
     ctx.client
         .upload(std::io::Cursor::new(data1), data1_len, &path1, None)
         .await?;
@@ -185,8 +190,18 @@ async fn s3_delete_objects_works(ctx: &mut MaybeEnabledS3) -> anyhow::Result<()>
         .upload(std::io::Cursor::new(data2), data2_len, &path2, None)
         .await?;
 
+    ctx.client
+        .upload(std::io::Cursor::new(data3), data3_len, &path3, None)
+        .await?;
+
     ctx.client.delete_objects(&[path1, path2]).await?;
 
+    let prefixes = ctx.client.list_prefixes(None).await?;
+
+    assert_eq!(prefixes.len(), 1);
+
+    ctx.client.delete_objects(&[path3]).await?;
+
     Ok(())
 }
 

libs/utils/Cargo.toml

@@ -5,7 +5,6 @@ edition.workspace = true
 license.workspace = true
 
 [dependencies]
-atty.workspace = true
 sentry.workspace = true
 async-trait.workspace = true
 anyhow.workspace = true

libs/utils/src/logging.rs

@@ -84,7 +84,7 @@ pub fn init(
     let r = r.with({
         let log_layer = tracing_subscriber::fmt::layer()
             .with_target(false)
-            .with_ansi(atty::is(atty::Stream::Stdout))
+            .with_ansi(false)
             .with_writer(std::io::stdout);
         let log_layer = match log_format {
             LogFormat::Json => log_layer.json().boxed(),

pageserver/benches/bench_layer_map.rs

@@ -1,22 +1,23 @@
 use pageserver::keyspace::{KeyPartitioning, KeySpace};
 use pageserver::repository::Key;
 use pageserver::tenant::layer_map::LayerMap;
-use pageserver::tenant::storage_layer::{Layer, LayerDescriptor, LayerFileName};
+use pageserver::tenant::storage_layer::{tests::LayerDescriptor, Layer, LayerFileName};
+use pageserver::tenant::storage_layer::{PersistentLayer, PersistentLayerDesc};
 use rand::prelude::{SeedableRng, SliceRandom, StdRng};
 use std::cmp::{max, min};
 use std::fs::File;
 use std::io::{BufRead, BufReader};
 use std::path::PathBuf;
 use std::str::FromStr;
-use std::sync::Arc;
 use std::time::Instant;
+use utils::id::{TenantId, TimelineId};
 
 use utils::lsn::Lsn;
 
 use criterion::{black_box, criterion_group, criterion_main, Criterion};
 
-fn build_layer_map(filename_dump: PathBuf) -> LayerMap<LayerDescriptor> {
+fn build_layer_map(filename_dump: PathBuf) -> LayerMap {
-    let mut layer_map = LayerMap::<LayerDescriptor>::default();
+    let mut layer_map = LayerMap::default();
 
     let mut min_lsn = Lsn(u64::MAX);
     let mut max_lsn = Lsn(0);
@@ -33,7 +34,7 @@ fn build_layer_map(filename_dump: PathBuf) -> LayerMap<LayerDescriptor> {
         min_lsn = min(min_lsn, lsn_range.start);
         max_lsn = max(max_lsn, Lsn(lsn_range.end.0 - 1));
 
-        updates.insert_historic(layer.get_persistent_layer_desc(), Arc::new(layer));
+        updates.insert_historic(layer.layer_desc().clone());
     }
 
     println!("min: {min_lsn}, max: {max_lsn}");
@@ -43,7 +44,7 @@ fn build_layer_map(filename_dump: PathBuf) -> LayerMap<LayerDescriptor> {
 }
 
 /// Construct a layer map query pattern for benchmarks
-fn uniform_query_pattern(layer_map: &LayerMap<LayerDescriptor>) -> Vec<(Key, Lsn)> {
+fn uniform_query_pattern(layer_map: &LayerMap) -> Vec<(Key, Lsn)> {
     // For each image layer we query one of the pages contained, at LSN right
     // before the image layer was created. This gives us a somewhat uniform
     // coverage of both the lsn and key space because image layers have
@@ -69,7 +70,7 @@ fn uniform_query_pattern(layer_map: &LayerMap<LayerDescriptor>) -> Vec<(Key, Lsn
 
 // Construct a partitioning for testing get_difficulty map when we
 // don't have an exact result of `collect_keyspace` to work with.
-fn uniform_key_partitioning(layer_map: &LayerMap<LayerDescriptor>, _lsn: Lsn) -> KeyPartitioning {
+fn uniform_key_partitioning(layer_map: &LayerMap, _lsn: Lsn) -> KeyPartitioning {
     let mut parts = Vec::new();
 
     // We add a partition boundary at the start of each image layer,
@@ -209,13 +210,15 @@ fn bench_sequential(c: &mut Criterion) {
     for i in 0..100_000 {
         let i32 = (i as u32) % 100;
         let zero = Key::from_hex("000000000000000000000000000000000000").unwrap();
-        let layer = LayerDescriptor {
+        let layer = LayerDescriptor::from(PersistentLayerDesc::new_img(
-            key: zero.add(10 * i32)..zero.add(10 * i32 + 1),
+            TenantId::generate(),
-            lsn: Lsn(i)..Lsn(i + 1),
+            TimelineId::generate(),
-            is_incremental: false,
+            zero.add(10 * i32)..zero.add(10 * i32 + 1),
-            short_id: format!("Layer {}", i),
+            Lsn(i),
-        };
+            false,
-        updates.insert_historic(layer.get_persistent_layer_desc(), Arc::new(layer));
+            0,
+        ));
+        updates.insert_historic(layer.layer_desc().clone());
     }
     updates.flush();
     println!("Finished layer map init in {:?}", now.elapsed());

pageserver/src/bin/pageserver.rs

@@ -495,50 +495,50 @@ fn start_pageserver(
                 Ok(())
             },
         );
+    }
 
-        if let Some(metric_collection_endpoint) = &conf.metric_collection_endpoint {
+    if let Some(metric_collection_endpoint) = &conf.metric_collection_endpoint {
-            let background_jobs_barrier = background_jobs_barrier;
+        let background_jobs_barrier = background_jobs_barrier;
-            let metrics_ctx = RequestContext::todo_child(
+        let metrics_ctx = RequestContext::todo_child(
-                TaskKind::MetricsCollection,
+            TaskKind::MetricsCollection,
-                // This task itself shouldn't download anything.
+            // This task itself shouldn't download anything.
-                // The actual size calculation does need downloads, and
+            // The actual size calculation does need downloads, and
-                // creates a child context with the right DownloadBehavior.
+            // creates a child context with the right DownloadBehavior.
-                DownloadBehavior::Error,
+            DownloadBehavior::Error,
-            );
+        );
-            task_mgr::spawn(
+        task_mgr::spawn(
-                MGMT_REQUEST_RUNTIME.handle(),
+            crate::BACKGROUND_RUNTIME.handle(),
-                TaskKind::MetricsCollection,
+            TaskKind::MetricsCollection,
-                None,
+            None,
-                None,
+            None,
-                "consumption metrics collection",
+            "consumption metrics collection",
-                true,
+            true,
-                async move {
+            async move {
-                    // first wait until background jobs are cleared to launch.
+                // first wait until background jobs are cleared to launch.
-                    //
+                //
-                    // this is because we only process active tenants and timelines, and the
+                // this is because we only process active tenants and timelines, and the
-                    // Timeline::get_current_logical_size will spawn the logical size calculation,
+                // Timeline::get_current_logical_size will spawn the logical size calculation,
-                    // which will not be rate-limited.
+                // which will not be rate-limited.
-                    let cancel = task_mgr::shutdown_token();
+                let cancel = task_mgr::shutdown_token();
 
-                    tokio::select! {
+                tokio::select! {
-                        _ = cancel.cancelled() => { return Ok(()); },
+                    _ = cancel.cancelled() => { return Ok(()); },
-                        _ = background_jobs_barrier.wait() => {}
+                    _ = background_jobs_barrier.wait() => {}
-                    };
+                };
 
-                    pageserver::consumption_metrics::collect_metrics(
+                pageserver::consumption_metrics::collect_metrics(
-                        metric_collection_endpoint,
+                    metric_collection_endpoint,
-                        conf.metric_collection_interval,
+                    conf.metric_collection_interval,
-                        conf.cached_metric_collection_interval,
+                    conf.cached_metric_collection_interval,
-                        conf.synthetic_size_calculation_interval,
+                    conf.synthetic_size_calculation_interval,
-                        conf.id,
+                    conf.id,
-                        metrics_ctx,
+                    metrics_ctx,
-                    )
+                )
-                    .instrument(info_span!("metrics_collection"))
+                .instrument(info_span!("metrics_collection"))
-                    .await?;
+                .await?;
-                    Ok(())
+                Ok(())
-                },
+            },
-            );
+        );
-        }
     }
 
     // Spawn a task to listen for libpq connections. It will spawn further tasks

pageserver/src/config.rs

@@ -96,12 +96,12 @@ pub mod defaults {
 
 #background_task_maximum_delay = '{DEFAULT_BACKGROUND_TASK_MAXIMUM_DELAY}'
 
-# [tenant_config]
+[tenant_config]
 #checkpoint_distance = {DEFAULT_CHECKPOINT_DISTANCE} # in bytes
 #checkpoint_timeout = {DEFAULT_CHECKPOINT_TIMEOUT}
 #compaction_target_size = {DEFAULT_COMPACTION_TARGET_SIZE} # in bytes
 #compaction_period = '{DEFAULT_COMPACTION_PERIOD}'
-#compaction_threshold = '{DEFAULT_COMPACTION_THRESHOLD}'
+#compaction_threshold = {DEFAULT_COMPACTION_THRESHOLD}
 
 #gc_period = '{DEFAULT_GC_PERIOD}'
 #gc_horizon = {DEFAULT_GC_HORIZON}
@@ -111,7 +111,8 @@ pub mod defaults {
 #min_resident_size_override = .. # in bytes
 #evictions_low_residence_duration_metric_threshold = '{DEFAULT_EVICTIONS_LOW_RESIDENCE_DURATION_METRIC_THRESHOLD}'
 #gc_feedback = false
-# [remote_storage]
+
+[remote_storage]
 
 "###
     );

pageserver/src/consumption_metrics.rs

@@ -24,6 +24,8 @@ const RESIDENT_SIZE: &str = "resident_size";
 const REMOTE_STORAGE_SIZE: &str = "remote_storage_size";
 const TIMELINE_LOGICAL_SIZE: &str = "timeline_logical_size";
 
+const DEFAULT_HTTP_REPORTING_TIMEOUT: Duration = Duration::from_secs(60);
+
 #[serde_as]
 #[derive(Serialize, Debug)]
 struct Ids {
@@ -73,7 +75,10 @@ pub async fn collect_metrics(
     );
 
     // define client here to reuse it for all requests
-    let client = reqwest::Client::new();
+    let client = reqwest::ClientBuilder::new()
+        .timeout(DEFAULT_HTTP_REPORTING_TIMEOUT)
+        .build()
+        .expect("Failed to create http client with timeout");
     let mut cached_metrics: HashMap<PageserverConsumptionMetricsKey, u64> = HashMap::new();
     let mut prev_iteration_time: std::time::Instant = std::time::Instant::now();
 
@@ -83,7 +88,7 @@ pub async fn collect_metrics(
                 info!("collect_metrics received cancellation request");
                 return Ok(());
             },
-            _ = ticker.tick() => {
+            tick_at = ticker.tick() => {
 
                 // send cached metrics every cached_metric_collection_interval
                 let send_cached = prev_iteration_time.elapsed() >= cached_metric_collection_interval;
@@ -93,6 +98,12 @@ pub async fn collect_metrics(
                 }
 
                 collect_metrics_iteration(&client, &mut cached_metrics, metric_collection_endpoint, node_id, &ctx, send_cached).await;
+
+                crate::tenant::tasks::warn_when_period_overrun(
+                    tick_at.elapsed(),
+                    metric_collection_interval,
+                    "consumption_metrics_collect_metrics",
+                );
             }
         }
     }
@@ -273,31 +284,42 @@ pub async fn collect_metrics_iteration(
         })
         .expect("PageserverConsumptionMetric should not fail serialization");
 
-        let res = client
+        const MAX_RETRIES: u32 = 3;
-            .post(metric_collection_endpoint.clone())
-            .json(&chunk_json)
-            .send()
-            .await;
 
-        match res {
+        for attempt in 0..MAX_RETRIES {
-            Ok(res) => {
+            let res = client
-                if res.status().is_success() {
+                .post(metric_collection_endpoint.clone())
-                    // update cached metrics after they were sent successfully
+                .json(&chunk_json)
-                    for (curr_key, curr_val) in chunk.iter() {
+                .send()
-                        cached_metrics.insert(curr_key.clone(), *curr_val);
+                .await;
-                    }
+
-                } else {
+            match res {
-                    error!("metrics endpoint refused the sent metrics: {:?}", res);
+                Ok(res) => {
-                    for metric in chunk_to_send.iter() {
+                    if res.status().is_success() {
-                        // Report if the metric value is suspiciously large
+                        // update cached metrics after they were sent successfully
-                        if metric.value > (1u64 << 40) {
+                        for (curr_key, curr_val) in chunk.iter() {
+                            cached_metrics.insert(curr_key.clone(), *curr_val);
+                        }
+                    } else {
+                        error!("metrics endpoint refused the sent metrics: {:?}", res);
+                        for metric in chunk_to_send
+                            .iter()
+                            .filter(|metric| metric.value > (1u64 << 40))
+                        {
+                            // Report if the metric value is suspiciously large
                             error!("potentially abnormal metric value: {:?}", metric);
                         }
                     }
+                    break;
+                }
+                Err(err) if err.is_timeout() => {
+                    error!(attempt, "timeout sending metrics, retrying immediately");
+                    continue;
+                }
+                Err(err) => {
+                    error!(attempt, ?err, "failed to send metrics");
+                    break;
                 }
-            }
-            Err(err) => {
-                error!("failed to send metrics: {:?}", err);
             }
         }
     }
@@ -317,7 +339,7 @@ pub async fn calculate_synthetic_size_worker(
             _ = task_mgr::shutdown_watcher() => {
                 return Ok(());
             },
-        _ = ticker.tick() => {
+        tick_at = ticker.tick() => {
 
                 let tenants = match mgr::list_tenants().await {
                     Ok(tenants) => tenants,
@@ -343,6 +365,12 @@ pub async fn calculate_synthetic_size_worker(
                     }
 
                 }
+
+                crate::tenant::tasks::warn_when_period_overrun(
+                    tick_at.elapsed(),
+                    synthetic_size_calculation_interval,
+                    "consumption_metrics_synthetic_size_worker",
+                );
             }
         }
     }

pageserver/src/http/openapi_spec.yml

@@ -186,10 +186,8 @@ paths:
               schema:
                 $ref: "#/components/schemas/Error"
     delete:
-      description: "Attempts to delete specified timeline. On 500 errors should be retried"
+      description: "Attempts to delete specified timeline. 500 and 409 errors should be retried"
       responses:
-        "200":
-          description: Ok
         "400":
           description: Error when no tenant id found in path or no timeline id
           content:
@@ -214,6 +212,12 @@ paths:
             application/json:
               schema:
                 $ref: "#/components/schemas/NotFoundError"
+        "409":
+          description: Deletion is already in progress, continue polling
+          content:
+            application/json:
+              schema:
+                $ref: "#/components/schemas/ConflictError"
         "412":
           description: Tenant is missing, or timeline has children
           content:
@@ -718,6 +722,12 @@ paths:
             application/json:
               schema:
                 $ref: "#/components/schemas/ForbiddenError"
+        "406":
+          description: Permanently unsatisfiable request, don't retry.
+          content:
+            application/json:
+              schema:
+                $ref: "#/components/schemas/Error"
         "409":
           description: Timeline already exists, creation skipped
           content:

pageserver/src/http/routes.rs

@@ -23,7 +23,6 @@ use super::models::{
     TimelineCreateRequest, TimelineGcRequest, TimelineInfo,
 };
 use crate::context::{DownloadBehavior, RequestContext};
-use crate::disk_usage_eviction_task;
 use crate::metrics::{StorageTimeOperation, STORAGE_TIME_GLOBAL};
 use crate::pgdatadir_mapping::LsnForTimestamp;
 use crate::task_mgr::TaskKind;
@@ -35,6 +34,7 @@ use crate::tenant::size::ModelInputs;
 use crate::tenant::storage_layer::LayerAccessStatsReset;
 use crate::tenant::{LogicalSizeCalculationCause, PageReconstructError, Timeline};
 use crate::{config::PageServerConf, tenant::mgr};
+use crate::{disk_usage_eviction_task, tenant};
 use utils::{
     auth::JwtAuth,
     http::{
@@ -187,6 +187,7 @@ impl From<crate::tenant::DeleteTimelineError> for ApiError {
                 format!("Cannot delete timeline which has child timelines: {children:?}")
                     .into_boxed_str(),
             ),
+            a @ AlreadyInProgress => ApiError::Conflict(a.to_string()),
             Other(e) => ApiError::InternalServerError(e),
         }
     }
@@ -327,15 +328,22 @@ async fn timeline_create_handler(
             &ctx,
         )
         .await {
-            Ok(Some(new_timeline)) => {
+            Ok(new_timeline) => {
                 // Created. Construct a TimelineInfo for it.
                 let timeline_info = build_timeline_info_common(&new_timeline, &ctx)
                     .await
                     .map_err(ApiError::InternalServerError)?;
                 json_response(StatusCode::CREATED, timeline_info)
             }
-            Ok(None) => json_response(StatusCode::CONFLICT, ()), // timeline already exists
+            Err(tenant::CreateTimelineError::AlreadyExists) => {
-            Err(err) => Err(ApiError::InternalServerError(err)),
+                json_response(StatusCode::CONFLICT, ())
+            }
+            Err(tenant::CreateTimelineError::AncestorLsn(err)) => {
+                json_response(StatusCode::NOT_ACCEPTABLE, HttpErrorBody::from_msg(
+                    format!("{err:#}")
+                ))
+            }
+            Err(tenant::CreateTimelineError::Other(err)) => Err(ApiError::InternalServerError(err)),
         }
     }
     .instrument(info_span!("timeline_create", tenant = %tenant_id, timeline_id = %new_timeline_id, lsn=?request_data.ancestor_start_lsn, pg_version=?request_data.pg_version))
@@ -1128,8 +1136,6 @@ async fn disk_usage_eviction_run(
         freed_bytes: 0,
     };
 
-    use crate::task_mgr::MGMT_REQUEST_RUNTIME;
-
     let (tx, rx) = tokio::sync::oneshot::channel();
 
     let state = get_state(&r);
@@ -1147,7 +1153,7 @@ async fn disk_usage_eviction_run(
     let _g = cancel.drop_guard();
 
     crate::task_mgr::spawn(
-        MGMT_REQUEST_RUNTIME.handle(),
+        crate::task_mgr::BACKGROUND_RUNTIME.handle(),
         TaskKind::DiskUsageEviction,
         None,
         None,

pageserver/src/metrics.rs

@@ -1,3 +1,4 @@
+use metrics::metric_vec_duration::DurationResultObserver;
 use metrics::{
     register_counter_vec, register_histogram, register_histogram_vec, register_int_counter,
     register_int_counter_vec, register_int_gauge, register_int_gauge_vec, register_uint_gauge_vec,
@@ -203,11 +204,11 @@ pub static TENANT_STATE_METRIC: Lazy<UIntGaugeVec> = Lazy::new(|| {
 
 pub static TENANT_SYNTHETIC_SIZE_METRIC: Lazy<UIntGaugeVec> = Lazy::new(|| {
     register_uint_gauge_vec!(
-        "pageserver_tenant_synthetic_size",
+        "pageserver_tenant_synthetic_cached_size_bytes",
-        "Synthetic size of each tenant",
+        "Synthetic size of each tenant in bytes",
         &["tenant_id"]
     )
-    .expect("Failed to register pageserver_tenant_synthetic_size metric")
+    .expect("Failed to register pageserver_tenant_synthetic_cached_size_bytes metric")
 });
 
 // Metrics for cloud upload. These metrics reflect data uploaded to cloud storage,
@@ -424,6 +425,27 @@ pub static SMGR_QUERY_TIME: Lazy<HistogramVec> = Lazy::new(|| {
     .expect("failed to define a metric")
 });
 
+pub struct BasebackupQueryTime(HistogramVec);
+pub static BASEBACKUP_QUERY_TIME: Lazy<BasebackupQueryTime> = Lazy::new(|| {
+    BasebackupQueryTime({
+        register_histogram_vec!(
+            "pageserver_basebackup_query_seconds",
+            "Histogram of basebackup queries durations, by result type",
+            &["result"],
+            CRITICAL_OP_BUCKETS.into(),
+        )
+        .expect("failed to define a metric")
+    })
+});
+
+impl DurationResultObserver for BasebackupQueryTime {
+    fn observe_result<T, E>(&self, res: &Result<T, E>, duration: std::time::Duration) {
+        let label_value = if res.is_ok() { "ok" } else { "error" };
+        let metric = self.0.get_metric_with_label_values(&[label_value]).unwrap();
+        metric.observe(duration.as_secs_f64());
+    }
+}
+
 pub static LIVE_CONNECTIONS_COUNT: Lazy<IntGaugeVec> = Lazy::new(|| {
     register_int_gauge_vec!(
         "pageserver_live_connections",
@@ -823,11 +845,6 @@ impl TimelineMetrics {
         let evictions_with_low_residence_duration =
             evictions_with_low_residence_duration_builder.build(&tenant_id, &timeline_id);
 
-        // TODO(chi): remove this once we remove Lazy for all metrics. Otherwise this will not appear in the exporter
-        // and integration test will error.
-        MATERIALIZED_PAGE_CACHE_HIT_DIRECT.get();
-        MATERIALIZED_PAGE_CACHE_HIT.get();
-
         TimelineMetrics {
             tenant_id,
             timeline_id,
@@ -1302,4 +1319,8 @@ pub fn preinitialize_metrics() {
 
     // Same as above for this metric, but, it's a Vec-type metric for which we don't know all the labels.
     BACKGROUND_LOOP_PERIOD_OVERRUN_COUNT.reset();
+
+    // Python tests need these.
+    MATERIALIZED_PAGE_CACHE_HIT_DIRECT.get();
+    MATERIALIZED_PAGE_CACHE_HIT.get();
 }

pageserver/src/page_service.rs

@@ -904,7 +904,7 @@ where
 
             self.check_permission(Some(tenant_id))?;
 
-            let lsn = if params.len() == 3 {
+            let lsn = if params.len() >= 3 {
                 Some(
                     Lsn::from_str(params[2])
                         .with_context(|| format!("Failed to parse Lsn from {}", params[2]))?,
@@ -913,10 +913,24 @@ where
                 None
             };
 
-            // Check that the timeline exists
+            metrics::metric_vec_duration::observe_async_block_duration_by_result(
-            self.handle_basebackup_request(pgb, tenant_id, timeline_id, lsn, None, false, ctx)
+                &*crate::metrics::BASEBACKUP_QUERY_TIME,
-                .await?;
+                async move {
-            pgb.write_message_noflush(&BeMessage::CommandComplete(b"SELECT 1"))?;
+                    self.handle_basebackup_request(
+                        pgb,
+                        tenant_id,
+                        timeline_id,
+                        lsn,
+                        None,
+                        false,
+                        ctx,
+                    )
+                    .await?;
+                    pgb.write_message_noflush(&BeMessage::CommandComplete(b"SELECT 1"))?;
+                    anyhow::Ok(())
+                },
+            )
+            .await?;
         }
         // return pair of prev_lsn and last_lsn
         else if query_string.starts_with("get_last_record_rlsn ") {

pageserver/src/pgdatadir_mapping.rs

@@ -887,7 +887,7 @@ impl<'a> DatadirModification<'a> {
         ctx: &RequestContext,
     ) -> Result<(), RelationError> {
         if rel.relnode == 0 {
-            return Err(RelationError::AlreadyExists);
+            return Err(RelationError::InvalidRelnode);
         }
         // It's possible that this is the first rel for this db in this
         // tablespace.  Create the reldir entry for it if so.

pageserver/src/task_mgr.rs

@@ -506,17 +506,17 @@ pub async fn shutdown_tasks(
                     warn!(name = task.name, tenant_id = ?tenant_id, timeline_id = ?timeline_id, kind = ?task_kind, "stopping left-over");
                 }
             }
-            let completed = tokio::select! {
+            let join_handle = tokio::select! {
                 biased;
-                _ = &mut join_handle => { true },
+                _ = &mut join_handle => { None },
                 _ = tokio::time::sleep(std::time::Duration::from_secs(1)) => {
                     // allow some time to elapse before logging to cut down the number of log
                     // lines.
                     info!("waiting for {} to shut down", task.name);
-                    false
+                    Some(join_handle)
                 }
             };
-            if !completed {
+            if let Some(join_handle) = join_handle {
                 // we never handled this return value, but:
                 // - we don't deschedule which would lead to is_cancelled
                 // - panics are already logged (is_panicked)

pageserver/src/tenant.rs

@@ -440,8 +440,13 @@ pub enum GetTimelineError {
 pub enum DeleteTimelineError {
     #[error("NotFound")]
     NotFound,
+
     #[error("HasChildren")]
     HasChildren(Vec<TimelineId>),
+
+    #[error("Timeline deletion is already in progress")]
+    AlreadyInProgress,
+
     #[error(transparent)]
     Other(#[from] anyhow::Error),
 }
@@ -496,6 +501,16 @@ impl DeletionGuard {
     }
 }
 
+#[derive(thiserror::Error, Debug)]
+pub enum CreateTimelineError {
+    #[error("a timeline with the given ID already exists")]
+    AlreadyExists,
+    #[error(transparent)]
+    AncestorLsn(anyhow::Error),
+    #[error(transparent)]
+    Other(#[from] anyhow::Error),
+}
+
 impl Tenant {
     /// Yet another helper for timeline initialization.
     /// Contains the common part of `load_local_timeline` and `load_remote_timeline`.
@@ -585,6 +600,7 @@ impl Tenant {
                     .layers
                     .read()
                     .await
+                    .0
                     .iter_historic_layers()
                     .next()
                     .is_some(),
@@ -1369,8 +1385,7 @@ impl Tenant {
     /// Returns the new timeline ID and reference to its Timeline object.
     ///
     /// If the caller specified the timeline ID to use (`new_timeline_id`), and timeline with
-    /// the same timeline ID already exists, returns None. If `new_timeline_id` is not given,
+    /// the same timeline ID already exists, returns CreateTimelineError::AlreadyExists.
-    /// a new unique ID is generated.
     pub async fn create_timeline(
         &self,
         new_timeline_id: TimelineId,
@@ -1379,11 +1394,12 @@ impl Tenant {
         pg_version: u32,
         broker_client: storage_broker::BrokerClientChannel,
         ctx: &RequestContext,
-    ) -> anyhow::Result<Option<Arc<Timeline>>> {
+    ) -> Result<Arc<Timeline>, CreateTimelineError> {
-        anyhow::ensure!(
+        if !self.is_active() {
-            self.is_active(),
+            return Err(CreateTimelineError::Other(anyhow::anyhow!(
-            "Cannot create timelines on inactive tenant"
+                "Cannot create timelines on inactive tenant"
-        );
+            )));
+        }
 
         if let Ok(existing) = self.get_timeline(new_timeline_id, false) {
             debug!("timeline {new_timeline_id} already exists");
@@ -1403,7 +1419,7 @@ impl Tenant {
                     .context("wait for timeline uploads to complete")?;
             }
 
-            return Ok(None);
+            return Err(CreateTimelineError::AlreadyExists);
         }
 
         let loaded_timeline = match ancestor_timeline_id {
@@ -1418,12 +1434,12 @@ impl Tenant {
                     let ancestor_ancestor_lsn = ancestor_timeline.get_ancestor_lsn();
                     if ancestor_ancestor_lsn > *lsn {
                         // can we safely just branch from the ancestor instead?
-                        bail!(
+                        return Err(CreateTimelineError::AncestorLsn(anyhow::anyhow!(
                             "invalid start lsn {} for ancestor timeline {}: less than timeline ancestor lsn {}",
                             lsn,
                             ancestor_timeline_id,
                             ancestor_ancestor_lsn,
-                        );
+                        )));
                     }
 
                     // Wait for the WAL to arrive and be processed on the parent branch up
@@ -1457,7 +1473,7 @@ impl Tenant {
             })?;
         }
 
-        Ok(Some(loaded_timeline))
+        Ok(loaded_timeline)
     }
 
     /// perform one garbage collection iteration, removing old data files from disk.
@@ -1755,14 +1771,11 @@ impl Tenant {
             timeline = Arc::clone(timeline_entry.get());
 
             // Prevent two tasks from trying to delete the timeline at the same time.
-            delete_lock_guard =
+            delete_lock_guard = DeletionGuard(
-                DeletionGuard(Arc::clone(&timeline.delete_lock).try_lock_owned().map_err(
+                Arc::clone(&timeline.delete_lock)
-                    |_| {
+                    .try_lock_owned()
-                        DeleteTimelineError::Other(anyhow::anyhow!(
+                    .map_err(|_| DeleteTimelineError::AlreadyInProgress)?,
-                            "timeline deletion is already in progress"
+            );
-                        ))
-                    },
-                )?);
 
             // If another task finished the deletion just before we acquired the lock,
             // return success.
@@ -2704,7 +2717,7 @@ impl Tenant {
         dst_id: TimelineId,
         start_lsn: Option<Lsn>,
         ctx: &RequestContext,
-    ) -> anyhow::Result<Arc<Timeline>> {
+    ) -> Result<Arc<Timeline>, CreateTimelineError> {
         let tl = self
             .branch_timeline_impl(src_timeline, dst_id, start_lsn, ctx)
             .await?;
@@ -2721,7 +2734,7 @@ impl Tenant {
         dst_id: TimelineId,
         start_lsn: Option<Lsn>,
         ctx: &RequestContext,
-    ) -> anyhow::Result<Arc<Timeline>> {
+    ) -> Result<Arc<Timeline>, CreateTimelineError> {
         self.branch_timeline_impl(src_timeline, dst_id, start_lsn, ctx)
             .await
     }
@@ -2732,7 +2745,7 @@ impl Tenant {
         dst_id: TimelineId,
         start_lsn: Option<Lsn>,
         _ctx: &RequestContext,
-    ) -> anyhow::Result<Arc<Timeline>> {
+    ) -> Result<Arc<Timeline>, CreateTimelineError> {
         let src_id = src_timeline.timeline_id;
 
         // If no start LSN is specified, we branch the new timeline from the source timeline's last record LSN
@@ -2772,16 +2785,17 @@ impl Tenant {
             .context(format!(
                 "invalid branch start lsn: less than latest GC cutoff {}",
                 *latest_gc_cutoff_lsn,
-            ))?;
+            ))
+            .map_err(CreateTimelineError::AncestorLsn)?;
 
         // and then the planned GC cutoff
         {
             let gc_info = src_timeline.gc_info.read().unwrap();
             let cutoff = min(gc_info.pitr_cutoff, gc_info.horizon_cutoff);
             if start_lsn < cutoff {
-                bail!(format!(
+                return Err(CreateTimelineError::AncestorLsn(anyhow::anyhow!(
                     "invalid branch start lsn: less than planned GC cutoff {cutoff}"
-                ));
+                )));
             }
         }
 
@@ -3814,6 +3828,9 @@ mod tests {
         {
             Ok(_) => panic!("branching should have failed"),
             Err(err) => {
+                let CreateTimelineError::AncestorLsn(err) = err else {
+                    panic!("wrong error type")
+                };
                 assert!(err.to_string().contains("invalid branch start lsn"));
                 assert!(err
                     .source()
@@ -3843,6 +3860,9 @@ mod tests {
         {
             Ok(_) => panic!("branching should have failed"),
             Err(err) => {
+                let CreateTimelineError::AncestorLsn(err) = err else {
+                    panic!("wrong error type");
+                };
                 assert!(&err.to_string().contains("invalid branch start lsn"));
                 assert!(&err
                     .source()

pageserver/src/tenant/layer_map.rs

@@ -51,25 +51,23 @@ use crate::keyspace::KeyPartitioning;
 use crate::repository::Key;
 use crate::tenant::storage_layer::InMemoryLayer;
 use crate::tenant::storage_layer::Layer;
-use anyhow::Context;
 use anyhow::Result;
-use std::collections::HashMap;
 use std::collections::VecDeque;
 use std::ops::Range;
 use std::sync::Arc;
 use utils::lsn::Lsn;
 
 use historic_layer_coverage::BufferedHistoricLayerCoverage;
-pub use historic_layer_coverage::Replacement;
+pub use historic_layer_coverage::LayerKey;
 
 use super::storage_layer::range_eq;
 use super::storage_layer::PersistentLayerDesc;
-use super::storage_layer::PersistentLayerKey;
 
 ///
 /// LayerMap tracks what layers exist on a timeline.
 ///
-pub struct LayerMap<L: ?Sized> {
+#[derive(Default)]
+pub struct LayerMap {
     //
     // 'open_layer' holds the current InMemoryLayer that is accepting new
     // records. If it is None, 'next_open_layer_at' will be set instead, indicating
@@ -95,24 +93,6 @@ pub struct LayerMap<L: ?Sized> {
     /// L0 layers have key range Key::MIN..Key::MAX, and locating them using R-Tree search is very inefficient.
     /// So L0 layers are held in l0_delta_layers vector, in addition to the R-tree.
     l0_delta_layers: Vec<Arc<PersistentLayerDesc>>,
-
-    /// Mapping from persistent layer key to the actual layer object. Currently, it stores delta, image, and
-    /// remote layers. In future refactors, this will be eventually moved out of LayerMap into Timeline, and
-    /// RemoteLayer will be removed.
-    mapping: HashMap<PersistentLayerKey, Arc<L>>,
-}
-
-impl<L: ?Sized> Default for LayerMap<L> {
-    fn default() -> Self {
-        Self {
-            open_layer: None,
-            next_open_layer_at: None,
-            frozen_layers: VecDeque::default(),
-            l0_delta_layers: Vec::default(),
-            historic: BufferedHistoricLayerCoverage::default(),
-            mapping: HashMap::default(),
-        }
-    }
 }
 
 /// The primary update API for the layer map.
@@ -120,24 +100,21 @@ impl<L: ?Sized> Default for LayerMap<L> {
 /// Batching historic layer insertions and removals is good for
 /// performance and this struct helps us do that correctly.
 #[must_use]
-pub struct BatchedUpdates<'a, L: ?Sized + Layer> {
+pub struct BatchedUpdates<'a> {
     // While we hold this exclusive reference to the layer map the type checker
     // will prevent us from accidentally reading any unflushed updates.
-    layer_map: &'a mut LayerMap<L>,
+    layer_map: &'a mut LayerMap,
 }
 
 /// Provide ability to batch more updates while hiding the read
 /// API so we don't accidentally read without flushing.
-impl<L> BatchedUpdates<'_, L>
+impl BatchedUpdates<'_> {
-where
-    L: ?Sized + Layer,
-{
     ///
     /// Insert an on-disk layer.
     ///
     // TODO remove the `layer` argument when `mapping` is refactored out of `LayerMap`
-    pub fn insert_historic(&mut self, layer_desc: PersistentLayerDesc, layer: Arc<L>) {
+    pub fn insert_historic(&mut self, layer_desc: PersistentLayerDesc) {
-        self.layer_map.insert_historic_noflush(layer_desc, layer)
+        self.layer_map.insert_historic_noflush(layer_desc)
     }
 
     ///
@@ -145,31 +122,8 @@ where
     ///
     /// This should be called when the corresponding file on disk has been deleted.
     ///
-    pub fn remove_historic(&mut self, layer_desc: PersistentLayerDesc, layer: Arc<L>) {
+    pub fn remove_historic(&mut self, layer_desc: PersistentLayerDesc) {
-        self.layer_map.remove_historic_noflush(layer_desc, layer)
+        self.layer_map.remove_historic_noflush(layer_desc)
-    }
-
-    /// Replaces existing layer iff it is the `expected`.
-    ///
-    /// If the expected layer has been removed it will not be inserted by this function.
-    ///
-    /// Returned `Replacement` describes succeeding in replacement or the reason why it could not
-    /// be done.
-    ///
-    /// TODO replacement can be done without buffering and rebuilding layer map updates.
-    ///      One way to do that is to add a layer of indirection for returned values, so
-    ///      that we can replace values only by updating a hashmap.
-    pub fn replace_historic(
-        &mut self,
-        expected_desc: PersistentLayerDesc,
-        expected: &Arc<L>,
-        new_desc: PersistentLayerDesc,
-        new: Arc<L>,
-    ) -> anyhow::Result<Replacement<Arc<L>>> {
-        fail::fail_point!("layermap-replace-notfound", |_| Ok(Replacement::NotFound));
-
-        self.layer_map
-            .replace_historic_noflush(expected_desc, expected, new_desc, new)
     }
 
     // We will flush on drop anyway, but this method makes it
@@ -185,25 +139,19 @@ where
 // than panic later or read without flushing.
 //
 // TODO maybe warn if flush hasn't explicitly been called
-impl<L> Drop for BatchedUpdates<'_, L>
+impl Drop for BatchedUpdates<'_> {
-where
-    L: ?Sized + Layer,
-{
     fn drop(&mut self) {
         self.layer_map.flush_updates();
     }
 }
 
 /// Return value of LayerMap::search
-pub struct SearchResult<L: ?Sized> {
+pub struct SearchResult {
-    pub layer: Arc<L>,
+    pub layer: Arc<PersistentLayerDesc>,
     pub lsn_floor: Lsn,
 }
 
-impl<L> LayerMap<L>
+impl LayerMap {
-where
-    L: ?Sized + Layer,
-{
     ///
     /// Find the latest layer (by lsn.end) that covers the given
     /// 'key', with lsn.start < 'end_lsn'.
@@ -235,7 +183,7 @@ where
     /// NOTE: This only searches the 'historic' layers, *not* the
     /// 'open' and 'frozen' layers!
     ///
-    pub fn search(&self, key: Key, end_lsn: Lsn) -> Option<SearchResult<L>> {
+    pub fn search(&self, key: Key, end_lsn: Lsn) -> Option<SearchResult> {
         let version = self.historic.get().unwrap().get_version(end_lsn.0 - 1)?;
         let latest_delta = version.delta_coverage.query(key.to_i128());
         let latest_image = version.image_coverage.query(key.to_i128());
@@ -244,7 +192,6 @@ where
             (None, None) => None,
             (None, Some(image)) => {
                 let lsn_floor = image.get_lsn_range().start;
-                let image = self.get_layer_from_mapping(&image.key()).clone();
                 Some(SearchResult {
                     layer: image,
                     lsn_floor,
@@ -252,7 +199,6 @@ where
             }
             (Some(delta), None) => {
                 let lsn_floor = delta.get_lsn_range().start;
-                let delta = self.get_layer_from_mapping(&delta.key()).clone();
                 Some(SearchResult {
                     layer: delta,
                     lsn_floor,
@@ -263,7 +209,6 @@ where
                 let image_is_newer = image.get_lsn_range().end >= delta.get_lsn_range().end;
                 let image_exact_match = img_lsn + 1 == end_lsn;
                 if image_is_newer || image_exact_match {
-                    let image = self.get_layer_from_mapping(&image.key()).clone();
                     Some(SearchResult {
                         layer: image,
                         lsn_floor: img_lsn,
@@ -271,7 +216,6 @@ where
                 } else {
                     let lsn_floor =
                         std::cmp::max(delta.get_lsn_range().start, image.get_lsn_range().start + 1);
-                    let delta = self.get_layer_from_mapping(&delta.key()).clone();
                     Some(SearchResult {
                         layer: delta,
                         lsn_floor,
@@ -282,7 +226,7 @@ where
     }
 
     /// Start a batch of updates, applied on drop
-    pub fn batch_update(&mut self) -> BatchedUpdates<'_, L> {
+    pub fn batch_update(&mut self) -> BatchedUpdates<'_> {
         BatchedUpdates { layer_map: self }
     }
 
@@ -292,48 +236,32 @@ where
     /// Helper function for BatchedUpdates::insert_historic
     ///
     /// TODO(chi): remove L generic so that we do not need to pass layer object.
-    pub(self) fn insert_historic_noflush(
+    pub(self) fn insert_historic_noflush(&mut self, layer_desc: PersistentLayerDesc) {
-        &mut self,
-        layer_desc: PersistentLayerDesc,
-        layer: Arc<L>,
-    ) {
-        self.mapping.insert(layer_desc.key(), layer.clone());
-
         // TODO: See #3869, resulting #4088, attempted fix and repro #4094
 
-        if Self::is_l0(&layer) {
+        if Self::is_l0(&layer_desc) {
             self.l0_delta_layers.push(layer_desc.clone().into());
         }
 
         self.historic.insert(
-            historic_layer_coverage::LayerKey::from(&*layer),
+            historic_layer_coverage::LayerKey::from(&layer_desc),
             layer_desc.into(),
         );
     }
 
-    fn get_layer_from_mapping(&self, key: &PersistentLayerKey) -> &Arc<L> {
-        let layer = self
-            .mapping
-            .get(key)
-            .with_context(|| format!("{key:?}"))
-            .expect("inconsistent layer mapping");
-        layer
-    }
-
     ///
     /// Remove an on-disk layer from the map.
     ///
     /// Helper function for BatchedUpdates::remove_historic
     ///
-    pub fn remove_historic_noflush(&mut self, layer_desc: PersistentLayerDesc, layer: Arc<L>) {
+    pub fn remove_historic_noflush(&mut self, layer_desc: PersistentLayerDesc) {
         self.historic
-            .remove(historic_layer_coverage::LayerKey::from(&*layer));
+            .remove(historic_layer_coverage::LayerKey::from(&layer_desc));
-        if Self::is_l0(&layer) {
+        let layer_key = layer_desc.key();
+        if Self::is_l0(&layer_desc) {
             let len_before = self.l0_delta_layers.len();
             let mut l0_delta_layers = std::mem::take(&mut self.l0_delta_layers);
-            l0_delta_layers.retain(|other| {
+            l0_delta_layers.retain(|other| other.key() != layer_key);
-                !Self::compare_arced_layers(self.get_layer_from_mapping(&other.key()), &layer)
-            });
             self.l0_delta_layers = l0_delta_layers;
             // this assertion is related to use of Arc::ptr_eq in Self::compare_arced_layers,
             // there's a chance that the comparison fails at runtime due to it comparing (pointer,
@@ -344,69 +272,6 @@ where
                 "failed to locate removed historic layer from l0_delta_layers"
             );
         }
-        self.mapping.remove(&layer_desc.key());
-    }
-
-    pub(self) fn replace_historic_noflush(
-        &mut self,
-        expected_desc: PersistentLayerDesc,
-        expected: &Arc<L>,
-        new_desc: PersistentLayerDesc,
-        new: Arc<L>,
-    ) -> anyhow::Result<Replacement<Arc<L>>> {
-        let key = historic_layer_coverage::LayerKey::from(&**expected);
-        let other = historic_layer_coverage::LayerKey::from(&*new);
-
-        let expected_l0 = Self::is_l0(expected);
-        let new_l0 = Self::is_l0(&new);
-
-        anyhow::ensure!(
-            key == other,
-            "expected and new must have equal LayerKeys: {key:?} != {other:?}"
-        );
-
-        anyhow::ensure!(
-            expected_l0 == new_l0,
-            "expected and new must both be l0 deltas or neither should be: {expected_l0} != {new_l0}"
-        );
-
-        let l0_index = if expected_l0 {
-            // find the index in case replace worked, we need to replace that as well
-            let pos = self.l0_delta_layers.iter().position(|slot| {
-                Self::compare_arced_layers(self.get_layer_from_mapping(&slot.key()), expected)
-            });
-
-            if pos.is_none() {
-                return Ok(Replacement::NotFound);
-            }
-            pos
-        } else {
-            None
-        };
-
-        let new_desc = Arc::new(new_desc);
-        let replaced = self.historic.replace(&key, new_desc.clone(), |existing| {
-            **existing == expected_desc
-        });
-
-        if let Replacement::Replaced { .. } = &replaced {
-            self.mapping.remove(&expected_desc.key());
-            self.mapping.insert(new_desc.key(), new);
-            if let Some(index) = l0_index {
-                self.l0_delta_layers[index] = new_desc;
-            }
-        }
-
-        let replaced = match replaced {
-            Replacement::Replaced { in_buffered } => Replacement::Replaced { in_buffered },
-            Replacement::NotFound => Replacement::NotFound,
-            Replacement::RemovalBuffered => Replacement::RemovalBuffered,
-            Replacement::Unexpected(x) => {
-                Replacement::Unexpected(self.get_layer_from_mapping(&x.key()).clone())
-            }
-        };
-
-        Ok(replaced)
     }
 
     /// Helper function for BatchedUpdates::drop.
@@ -454,10 +319,8 @@ where
         Ok(true)
     }
 
-    pub fn iter_historic_layers(&self) -> impl '_ + Iterator<Item = Arc<L>> {
+    pub fn iter_historic_layers(&self) -> impl '_ + Iterator<Item = Arc<PersistentLayerDesc>> {
-        self.historic
+        self.historic.iter()
-            .iter()
-            .map(|x| self.get_layer_from_mapping(&x.key()).clone())
     }
 
     ///
@@ -472,7 +335,7 @@ where
         &self,
         key_range: &Range<Key>,
         lsn: Lsn,
-    ) -> Result<Vec<(Range<Key>, Option<Arc<L>>)>> {
+    ) -> Result<Vec<(Range<Key>, Option<Arc<PersistentLayerDesc>>)>> {
         let version = match self.historic.get().unwrap().get_version(lsn.0) {
             Some(v) => v,
             None => return Ok(vec![]),
@@ -482,36 +345,26 @@ where
         let end = key_range.end.to_i128();
 
         // Initialize loop variables
-        let mut coverage: Vec<(Range<Key>, Option<Arc<L>>)> = vec![];
+        let mut coverage: Vec<(Range<Key>, Option<Arc<PersistentLayerDesc>>)> = vec![];
         let mut current_key = start;
         let mut current_val = version.image_coverage.query(start);
 
         // Loop through the change events and push intervals
         for (change_key, change_val) in version.image_coverage.range(start..end) {
             let kr = Key::from_i128(current_key)..Key::from_i128(change_key);
-            coverage.push((
+            coverage.push((kr, current_val.take()));
-                kr,
-                current_val
-                    .take()
-                    .map(|l| self.get_layer_from_mapping(&l.key()).clone()),
-            ));
             current_key = change_key;
             current_val = change_val.clone();
         }
 
         // Add the final interval
         let kr = Key::from_i128(current_key)..Key::from_i128(end);
-        coverage.push((
+        coverage.push((kr, current_val.take()));
-            kr,
-            current_val
-                .take()
-                .map(|l| self.get_layer_from_mapping(&l.key()).clone()),
-        ));
 
         Ok(coverage)
     }
 
-    pub fn is_l0(layer: &L) -> bool {
+    pub fn is_l0(layer: &PersistentLayerDesc) -> bool {
         range_eq(&layer.get_key_range(), &(Key::MIN..Key::MAX))
     }
 
@@ -537,7 +390,7 @@ where
     /// TODO The optimal number should probably be slightly higher than 1, but to
     ///      implement that we need to plumb a lot more context into this function
     ///      than just the current partition_range.
-    pub fn is_reimage_worthy(layer: &L, partition_range: &Range<Key>) -> bool {
+    pub fn is_reimage_worthy(layer: &PersistentLayerDesc, partition_range: &Range<Key>) -> bool {
         // Case 1
         if !Self::is_l0(layer) {
             return true;
@@ -595,9 +448,7 @@ where
                     let kr = Key::from_i128(current_key)..Key::from_i128(change_key);
                     let lr = lsn.start..val.get_lsn_range().start;
                     if !kr.is_empty() {
-                        let base_count =
+                        let base_count = Self::is_reimage_worthy(&val, key) as usize;
-                            Self::is_reimage_worthy(self.get_layer_from_mapping(&val.key()), key)
-                                as usize;
                         let new_limit = limit.map(|l| l - base_count);
                         let max_stacked_deltas_underneath =
                             self.count_deltas(&kr, &lr, new_limit)?;
@@ -620,9 +471,7 @@ where
                 let lr = lsn.start..val.get_lsn_range().start;
 
                 if !kr.is_empty() {
-                    let base_count =
+                    let base_count = Self::is_reimage_worthy(&val, key) as usize;
-                        Self::is_reimage_worthy(self.get_layer_from_mapping(&val.key()), key)
-                            as usize;
                     let new_limit = limit.map(|l| l - base_count);
                     let max_stacked_deltas_underneath = self.count_deltas(&kr, &lr, new_limit)?;
                     max_stacked_deltas = std::cmp::max(
@@ -772,12 +621,8 @@ where
     }
 
     /// Return all L0 delta layers
-    pub fn get_level0_deltas(&self) -> Result<Vec<Arc<L>>> {
+    pub fn get_level0_deltas(&self) -> Result<Vec<Arc<PersistentLayerDesc>>> {
-        Ok(self
+        Ok(self.l0_delta_layers.to_vec())
-            .l0_delta_layers
-            .iter()
-            .map(|x| self.get_layer_from_mapping(&x.key()).clone())
-            .collect())
     }
 
     /// debugging function to print out the contents of the layer map
@@ -802,72 +647,51 @@ where
         println!("End dump LayerMap");
         Ok(())
     }
-
-    /// Similar to `Arc::ptr_eq`, but only compares the object pointers, not vtables.
-    ///
-    /// Returns `true` if the two `Arc` point to the same layer, false otherwise.
-    #[inline(always)]
-    pub fn compare_arced_layers(left: &Arc<L>, right: &Arc<L>) -> bool {
-        // "dyn Trait" objects are "fat pointers" in that they have two components:
-        // - pointer to the object
-        // - pointer to the vtable
-        //
-        // rust does not provide a guarantee that these vtables are unique, but however
-        // `Arc::ptr_eq` as of writing (at least up to 1.67) uses a comparison where both the
-        // pointer and the vtable need to be equal.
-        //
-        // See: https://github.com/rust-lang/rust/issues/103763
-        //
-        // A future version of rust will most likely use this form below, where we cast each
-        // pointer into a pointer to unit, which drops the inaccessible vtable pointer, making it
-        // not affect the comparison.
-        //
-        // See: https://github.com/rust-lang/rust/pull/106450
-        let left = Arc::as_ptr(left) as *const ();
-        let right = Arc::as_ptr(right) as *const ();
-
-        left == right
-    }
 }
 
 #[cfg(test)]
 mod tests {
-    use super::{LayerMap, Replacement};
+    use super::LayerMap;
-    use crate::tenant::storage_layer::{Layer, LayerDescriptor, LayerFileName};
+    use crate::tenant::storage_layer::{tests::LayerDescriptor, LayerFileName};
     use std::str::FromStr;
     use std::sync::Arc;
 
     mod l0_delta_layers_updated {
 
+        use crate::tenant::{
+            storage_layer::{PersistentLayer, PersistentLayerDesc},
+            timeline::LayerFileManager,
+        };
+
         use super::*;
 
         #[test]
         fn for_full_range_delta() {
             // l0_delta_layers are used by compaction, and should observe all buffered updates
             l0_delta_layers_updated_scenario(
-                "000000000000000000000000000000000000-FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF__0000000053423C21-0000000053424D69",
+                 "000000000000000000000000000000000000-FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF__0000000053423C21-0000000053424D69",
-                true
+                 true
-            )
+             )
         }
 
         #[test]
         fn for_non_full_range_delta() {
             // has minimal uncovered areas compared to l0_delta_layers_updated_on_insert_replace_remove_for_full_range_delta
             l0_delta_layers_updated_scenario(
-                "000000000000000000000000000000000001-FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFE__0000000053423C21-0000000053424D69",
+                 "000000000000000000000000000000000001-FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFE__0000000053423C21-0000000053424D69",
-                // because not full range
+                 // because not full range
-                false
+                 false
-            )
+             )
         }
 
         #[test]
         fn for_image() {
             l0_delta_layers_updated_scenario(
-                "000000000000000000000000000000000000-000000000000000000000000000000010000__0000000053424D69",
+                 "000000000000000000000000000000000000-000000000000000000000000000000010000__0000000053424D69",
-                // code only checks if it is a full range layer, doesn't care about images, which must
+                 // code only checks if it is a full range layer, doesn't care about images, which must
-                // mean we should in practice never have full range images
+                 // mean we should in practice never have full range images
-                false
+                 false
-            )
+             )
         }
 
         #[test]
@@ -883,16 +707,16 @@ mod tests {
             let not_found = Arc::new(layer.clone());
             let new_version = Arc::new(layer);
 
-            let mut map = LayerMap::default();
+            // after the immutable storage state refactor, the replace operation
+            // will not use layer map any more. We keep it here for consistency in test cases
+            // and can remove it in the future.
+            let _map = LayerMap::default();
 
-            let res = map.batch_update().replace_historic(
+            let mut mapping = LayerFileManager::new();
-                not_found.get_persistent_layer_desc(),
-                &not_found,
-                new_version.get_persistent_layer_desc(),
-                new_version,
-            );
 
-            assert!(matches!(res, Ok(Replacement::NotFound)), "{res:?}");
+            mapping
+                .replace_and_verify(not_found, new_version)
+                .unwrap_err();
         }
 
         fn l0_delta_layers_updated_scenario(layer_name: &str, expected_l0: bool) {
@@ -903,49 +727,44 @@ mod tests {
             let downloaded = Arc::new(skeleton);
 
             let mut map = LayerMap::default();
+            let mut mapping = LayerFileManager::new();
 
             // two disjoint Arcs in different lifecycle phases. even if it seems they must be the
             // same layer, we use LayerMap::compare_arced_layers as the identity of layers.
-            assert!(!LayerMap::compare_arced_layers(&remote, &downloaded));
+            assert_eq!(remote.layer_desc(), downloaded.layer_desc());
 
             let expected_in_counts = (1, usize::from(expected_l0));
 
             map.batch_update()
-                .insert_historic(remote.get_persistent_layer_desc(), remote.clone());
+                .insert_historic(remote.layer_desc().clone());
-            assert_eq!(count_layer_in(&map, &remote), expected_in_counts);
+            mapping.insert(remote.clone());
-
+            assert_eq!(
-            let replaced = map
+                count_layer_in(&map, remote.layer_desc()),
-                .batch_update()
+                expected_in_counts
-                .replace_historic(
+            );
-                    remote.get_persistent_layer_desc(),
+
-                    &remote,
+            mapping
-                    downloaded.get_persistent_layer_desc(),
+                .replace_and_verify(remote, downloaded.clone())
-                    downloaded.clone(),
+                .expect("name derived attributes are the same");
-                )
+            assert_eq!(
-                .expect("name derived attributes are the same");
+                count_layer_in(&map, downloaded.layer_desc()),
-            assert!(
+                expected_in_counts
-                matches!(replaced, Replacement::Replaced { .. }),
-                "{replaced:?}"
             );
-            assert_eq!(count_layer_in(&map, &downloaded), expected_in_counts);
 
             map.batch_update()
-                .remove_historic(downloaded.get_persistent_layer_desc(), downloaded.clone());
+                .remove_historic(downloaded.layer_desc().clone());
-            assert_eq!(count_layer_in(&map, &downloaded), (0, 0));
+            assert_eq!(count_layer_in(&map, downloaded.layer_desc()), (0, 0));
         }
 
-        fn count_layer_in<L: Layer + ?Sized>(map: &LayerMap<L>, layer: &Arc<L>) -> (usize, usize) {
+        fn count_layer_in(map: &LayerMap, layer: &PersistentLayerDesc) -> (usize, usize) {
             let historic = map
                 .iter_historic_layers()
-                .filter(|x| LayerMap::compare_arced_layers(x, layer))
+                .filter(|x| x.key() == layer.key())
                 .count();
             let l0s = map
                 .get_level0_deltas()
                 .expect("why does this return a result");
-            let l0 = l0s
+            let l0 = l0s.iter().filter(|x| x.key() == layer.key()).count();
-                .iter()
-                .filter(|x| LayerMap::compare_arced_layers(x, layer))
-                .count();
 
             (historic, l0)
         }

pageserver/src/tenant/layer_map/historic_layer_coverage.rs

@@ -3,6 +3,8 @@ use std::ops::Range;
 
 use tracing::info;
 
+use crate::tenant::storage_layer::PersistentLayerDesc;
+
 use super::layer_coverage::LayerCoverageTuple;
 
 /// Layers in this module are identified and indexed by this data.
@@ -41,8 +43,8 @@ impl Ord for LayerKey {
     }
 }
 
-impl<'a, L: crate::tenant::storage_layer::Layer + ?Sized> From<&'a L> for LayerKey {
+impl From<&PersistentLayerDesc> for LayerKey {
-    fn from(layer: &'a L) -> Self {
+    fn from(layer: &PersistentLayerDesc) -> Self {
         let kr = layer.get_key_range();
         let lr = layer.get_lsn_range();
         LayerKey {
@@ -454,59 +456,6 @@ impl<Value: Clone> BufferedHistoricLayerCoverage<Value> {
         self.buffer.insert(layer_key, None);
     }
 
-    /// Replaces a previous layer with a new layer value.
-    ///
-    /// The replacement is conditional on:
-    /// - there is an existing `LayerKey` record
-    /// - there is no buffered removal for the given `LayerKey`
-    /// - the given closure returns true for the current `Value`
-    ///
-    /// The closure is used to compare the latest value (buffered insert, or existing layer)
-    /// against some expectation. This allows to use `Arc::ptr_eq` or similar which would be
-    /// inaccessible via `PartialEq` trait.
-    ///
-    /// Returns a `Replacement` value describing the outcome; only the case of
-    /// `Replacement::Replaced` modifies the map and requires a rebuild.
-    pub fn replace<F>(
-        &mut self,
-        layer_key: &LayerKey,
-        new: Value,
-        check_expected: F,
-    ) -> Replacement<Value>
-    where
-        F: FnOnce(&Value) -> bool,
-    {
-        let (slot, in_buffered) = match self.buffer.get(layer_key) {
-            Some(inner @ Some(_)) => {
-                // we compare against the buffered version, because there will be a later
-                // rebuild before querying
-                (inner.as_ref(), true)
-            }
-            Some(None) => {
-                // buffer has removal for this key; it will not be equivalent by any check_expected.
-                return Replacement::RemovalBuffered;
-            }
-            None => {
-                // no pending modification for the key, check layers
-                (self.layers.get(layer_key), false)
-            }
-        };
-
-        match slot {
-            Some(existing) if !check_expected(existing) => {
-                // unfortunate clone here, but otherwise the nll borrowck grows the region of
-                // 'a to cover the whole function, and we could not mutate in the other
-                // Some(existing) branch
-                Replacement::Unexpected(existing.clone())
-            }
-            None => Replacement::NotFound,
-            Some(_existing) => {
-                self.insert(layer_key.to_owned(), new);
-                Replacement::Replaced { in_buffered }
-            }
-        }
-    }
-
     pub fn rebuild(&mut self) {
         // Find the first LSN that needs to be rebuilt
         let rebuild_since: u64 = match self.buffer.iter().next() {
@@ -575,22 +524,6 @@ impl<Value: Clone> BufferedHistoricLayerCoverage<Value> {
     }
 }
 
-/// Outcome of the replace operation.
-#[derive(Debug)]
-pub enum Replacement<Value> {
-    /// Previous value was replaced with the new value.
-    Replaced {
-        /// Replacement happened for a scheduled insert.
-        in_buffered: bool,
-    },
-    /// Key was not found buffered updates or existing layers.
-    NotFound,
-    /// Key has been scheduled for removal, it was not replaced.
-    RemovalBuffered,
-    /// Previous value was rejected by the closure.
-    Unexpected(Value),
-}
-
 #[test]
 fn test_retroactive_regression_1() {
     let mut map = BufferedHistoricLayerCoverage::new();
@@ -699,139 +632,3 @@ fn test_retroactive_simple() {
         assert_eq!(version.image_coverage.query(8), Some("Image 4".to_string()));
     }
 }
-
-#[test]
-fn test_retroactive_replacement() {
-    let mut map = BufferedHistoricLayerCoverage::new();
-
-    let keys = [
-        LayerKey {
-            key: 0..5,
-            lsn: 100..101,
-            is_image: true,
-        },
-        LayerKey {
-            key: 3..9,
-            lsn: 110..111,
-            is_image: true,
-        },
-        LayerKey {
-            key: 4..6,
-            lsn: 120..121,
-            is_image: true,
-        },
-    ];
-
-    let layers = [
-        "Image 1".to_string(),
-        "Image 2".to_string(),
-        "Image 3".to_string(),
-    ];
-
-    for (key, layer) in keys.iter().zip(layers.iter()) {
-        map.insert(key.to_owned(), layer.to_owned());
-    }
-
-    // rebuild is not necessary here, because replace works for both buffered updates and existing
-    // layers.
-
-    for (key, orig_layer) in keys.iter().zip(layers.iter()) {
-        let replacement = format!("Remote {orig_layer}");
-
-        // evict
-        let ret = map.replace(key, replacement.clone(), |l| l == orig_layer);
-        assert!(
-            matches!(ret, Replacement::Replaced { .. }),
-            "replace {orig_layer}: {ret:?}"
-        );
-        map.rebuild();
-
-        let at = key.lsn.end + 1;
-
-        let version = map.get().expect("rebuilt").get_version(at).unwrap();
-        assert_eq!(
-            version.image_coverage.query(4).as_deref(),
-            Some(replacement.as_str()),
-            "query for 4 at version {at} after eviction",
-        );
-
-        // download
-        let ret = map.replace(key, orig_layer.clone(), |l| l == &replacement);
-        assert!(
-            matches!(ret, Replacement::Replaced { .. }),
-            "replace {orig_layer} back: {ret:?}"
-        );
-        map.rebuild();
-        let version = map.get().expect("rebuilt").get_version(at).unwrap();
-        assert_eq!(
-            version.image_coverage.query(4).as_deref(),
-            Some(orig_layer.as_str()),
-            "query for 4 at version {at} after download",
-        );
-    }
-}
-
-#[test]
-fn missing_key_is_not_inserted_with_replace() {
-    let mut map = BufferedHistoricLayerCoverage::new();
-    let key = LayerKey {
-        key: 0..5,
-        lsn: 100..101,
-        is_image: true,
-    };
-
-    let ret = map.replace(&key, "should not replace", |_| true);
-    assert!(matches!(ret, Replacement::NotFound), "{ret:?}");
-    map.rebuild();
-    assert!(map
-        .get()
-        .expect("no changes to rebuild")
-        .get_version(102)
-        .is_none());
-}
-
-#[test]
-fn replacing_buffered_insert_and_remove() {
-    let mut map = BufferedHistoricLayerCoverage::new();
-    let key = LayerKey {
-        key: 0..5,
-        lsn: 100..101,
-        is_image: true,
-    };
-
-    map.insert(key.clone(), "Image 1");
-    let ret = map.replace(&key, "Remote Image 1", |&l| l == "Image 1");
-    assert!(
-        matches!(ret, Replacement::Replaced { in_buffered: true }),
-        "{ret:?}"
-    );
-    map.rebuild();
-
-    assert_eq!(
-        map.get()
-            .expect("rebuilt")
-            .get_version(102)
-            .unwrap()
-            .image_coverage
-            .query(4),
-        Some("Remote Image 1")
-    );
-
-    map.remove(key.clone());
-    let ret = map.replace(&key, "should not replace", |_| true);
-    assert!(
-        matches!(ret, Replacement::RemovalBuffered),
-        "cannot replace after scheduled remove: {ret:?}"
-    );
-
-    map.rebuild();
-
-    let ret = map.replace(&key, "should not replace", |_| true);
-    assert!(
-        matches!(ret, Replacement::NotFound),
-        "cannot replace after remove + rebuild: {ret:?}"
-    );
-
-    let at_version = map.get().expect("rebuilt").get_version(102);
-    assert!(at_version.is_none());
-}

pageserver/src/tenant/remote_timeline_client.rs

@@ -862,10 +862,8 @@ impl RemoteTimelineClient {
                 "Found {} files not bound to index_file.json, proceeding with their deletion",
                 remaining.len()
             );
-            for file in remaining {
+            warn!("About to remove {} files", remaining.len());
-                warn!("Removing {}", file.object_name().unwrap_or_default());
+            self.storage_impl.delete_objects(&remaining).await?;
-                self.storage_impl.delete(&file).await?;
-            }
         }
 
         let index_file_path = timeline_storage_path.join(Path::new(IndexPart::FILE_NAME));

pageserver/src/tenant/storage_layer.rs

@@ -176,13 +176,10 @@ impl LayerAccessStats {
     /// Create an empty stats object and record a [`LayerLoad`] event with the given residence status.
     ///
     /// See [`record_residence_event`] for why you need to do this while holding the layer map lock.
-    pub(crate) fn for_loading_layer<L>(
+    pub(crate) fn for_loading_layer(
-        layer_map_lock_held_witness: &BatchedUpdates<'_, L>,
+        layer_map_lock_held_witness: &BatchedUpdates<'_>,
         status: LayerResidenceStatus,
-    ) -> Self
+    ) -> Self {
-    where
-        L: ?Sized + Layer,
-    {
         let new = LayerAccessStats(Mutex::new(LayerAccessStatsLocked::default()));
         new.record_residence_event(
             layer_map_lock_held_witness,
@@ -197,14 +194,11 @@ impl LayerAccessStats {
     /// The `new_status` is not recorded in `self`.
     ///
     /// See [`record_residence_event`] for why you need to do this while holding the layer map lock.
-    pub(crate) fn clone_for_residence_change<L>(
+    pub(crate) fn clone_for_residence_change(
         &self,
-        layer_map_lock_held_witness: &BatchedUpdates<'_, L>,
+        layer_map_lock_held_witness: &BatchedUpdates<'_>,
         new_status: LayerResidenceStatus,
-    ) -> LayerAccessStats
+    ) -> LayerAccessStats {
-    where
-        L: ?Sized + Layer,
-    {
         let clone = {
             let inner = self.0.lock().unwrap();
             inner.clone()
@@ -232,14 +226,12 @@ impl LayerAccessStats {
     /// - Compact: Grab layer map lock, add the new L1 to layer map and remove the L0s, release layer map lock.
     /// - Eviction: observes the new L1 layer whose only activity timestamp is the LayerCreate event.
     ///
-    pub(crate) fn record_residence_event<L>(
+    pub(crate) fn record_residence_event(
         &self,
-        _layer_map_lock_held_witness: &BatchedUpdates<'_, L>,
+        _layer_map_lock_held_witness: &BatchedUpdates<'_>,
         status: LayerResidenceStatus,
         reason: LayerResidenceEventReason,
-    ) where
+    ) {
-        L: ?Sized + Layer,
-    {
         let mut locked = self.0.lock().unwrap();
         locked.iter_mut().for_each(|inner| {
             inner
@@ -473,94 +465,125 @@ pub fn downcast_remote_layer(
     }
 }
 
-/// Holds metadata about a layer without any content. Used mostly for testing.
+pub mod tests {
-///
+    use super::*;
-/// To use filenames as fixtures, parse them as [`LayerFileName`] then convert from that to a
-/// LayerDescriptor.
-#[derive(Clone, Debug)]
-pub struct LayerDescriptor {
-    pub key: Range<Key>,
-    pub lsn: Range<Lsn>,
-    pub is_incremental: bool,
-    pub short_id: String,
-}
 
-impl LayerDescriptor {
+    /// Holds metadata about a layer without any content. Used mostly for testing.
-    /// `LayerDescriptor` is only used for testing purpose so it does not matter whether it is image / delta,
+    ///
-    /// and the tenant / timeline id does not matter.
+    /// To use filenames as fixtures, parse them as [`LayerFileName`] then convert from that to a
-    pub fn get_persistent_layer_desc(&self) -> PersistentLayerDesc {
+    /// LayerDescriptor.
-        PersistentLayerDesc::new_delta(
+    #[derive(Clone, Debug)]
-            TenantId::from_array([0; 16]),
+    pub struct LayerDescriptor {
-            TimelineId::from_array([0; 16]),
+        base: PersistentLayerDesc,
-            self.key.clone(),
-            self.lsn.clone(),
-            233,
-        )
-    }
-}
-
-impl Layer for LayerDescriptor {
-    fn get_key_range(&self) -> Range<Key> {
-        self.key.clone()
     }
 
-    fn get_lsn_range(&self) -> Range<Lsn> {
+    impl From<PersistentLayerDesc> for LayerDescriptor {
-        self.lsn.clone()
+        fn from(base: PersistentLayerDesc) -> Self {
-    }
+            Self { base }
-
-    fn is_incremental(&self) -> bool {
-        self.is_incremental
-    }
-
-    fn get_value_reconstruct_data(
-        &self,
-        _key: Key,
-        _lsn_range: Range<Lsn>,
-        _reconstruct_data: &mut ValueReconstructState,
-        _ctx: &RequestContext,
-    ) -> Result<ValueReconstructResult> {
-        todo!("This method shouldn't be part of the Layer trait")
-    }
-
-    fn short_id(&self) -> String {
-        self.short_id.clone()
-    }
-
-    fn dump(&self, _verbose: bool, _ctx: &RequestContext) -> Result<()> {
-        todo!()
-    }
-}
-
-impl From<DeltaFileName> for LayerDescriptor {
-    fn from(value: DeltaFileName) -> Self {
-        let short_id = value.to_string();
-        LayerDescriptor {
-            key: value.key_range,
-            lsn: value.lsn_range,
-            is_incremental: true,
-            short_id,
         }
     }
-}
 
-impl From<ImageFileName> for LayerDescriptor {
+    impl Layer for LayerDescriptor {
-    fn from(value: ImageFileName) -> Self {
+        fn get_value_reconstruct_data(
-        let short_id = value.to_string();
+            &self,
-        let lsn = value.lsn_as_range();
+            _key: Key,
-        LayerDescriptor {
+            _lsn_range: Range<Lsn>,
-            key: value.key_range,
+            _reconstruct_data: &mut ValueReconstructState,
-            lsn,
+            _ctx: &RequestContext,
-            is_incremental: false,
+        ) -> Result<ValueReconstructResult> {
-            short_id,
+            todo!("This method shouldn't be part of the Layer trait")
+        }
+
+        fn dump(&self, _verbose: bool, _ctx: &RequestContext) -> Result<()> {
+            todo!()
+        }
+
+        /// Boilerplate to implement the Layer trait, always use layer_desc for persistent layers.
+        fn get_key_range(&self) -> Range<Key> {
+            self.layer_desc().key_range.clone()
+        }
+
+        /// Boilerplate to implement the Layer trait, always use layer_desc for persistent layers.
+        fn get_lsn_range(&self) -> Range<Lsn> {
+            self.layer_desc().lsn_range.clone()
+        }
+
+        /// Boilerplate to implement the Layer trait, always use layer_desc for persistent layers.
+        fn is_incremental(&self) -> bool {
+            self.layer_desc().is_incremental
+        }
+
+        /// Boilerplate to implement the Layer trait, always use layer_desc for persistent layers.
+        fn short_id(&self) -> String {
+            self.layer_desc().short_id()
         }
     }
-}
 
-impl From<LayerFileName> for LayerDescriptor {
+    impl PersistentLayer for LayerDescriptor {
-    fn from(value: LayerFileName) -> Self {
+        fn layer_desc(&self) -> &PersistentLayerDesc {
-        match value {
+            &self.base
-            LayerFileName::Delta(d) => Self::from(d),
+        }
-            LayerFileName::Image(i) => Self::from(i),
+
+        fn local_path(&self) -> Option<PathBuf> {
+            unimplemented!()
+        }
+
+        fn iter(&self, _: &RequestContext) -> Result<LayerIter<'_>> {
+            unimplemented!()
+        }
+
+        fn key_iter(&self, _: &RequestContext) -> Result<LayerKeyIter<'_>> {
+            unimplemented!()
+        }
+
+        fn delete_resident_layer_file(&self) -> Result<()> {
+            unimplemented!()
+        }
+
+        fn info(&self, _: LayerAccessStatsReset) -> HistoricLayerInfo {
+            unimplemented!()
+        }
+
+        fn access_stats(&self) -> &LayerAccessStats {
+            unimplemented!()
+        }
+    }
+
+    impl From<DeltaFileName> for LayerDescriptor {
+        fn from(value: DeltaFileName) -> Self {
+            LayerDescriptor {
+                base: PersistentLayerDesc::new_delta(
+                    TenantId::from_array([0; 16]),
+                    TimelineId::from_array([0; 16]),
+                    value.key_range,
+                    value.lsn_range,
+                    233,
+                ),
+            }
+        }
+    }
+
+    impl From<ImageFileName> for LayerDescriptor {
+        fn from(value: ImageFileName) -> Self {
+            LayerDescriptor {
+                base: PersistentLayerDesc::new_img(
+                    TenantId::from_array([0; 16]),
+                    TimelineId::from_array([0; 16]),
+                    value.key_range,
+                    value.lsn,
+                    false,
+                    233,
+                ),
+            }
+        }
+    }
+
+    impl From<LayerFileName> for LayerDescriptor {
+        fn from(value: LayerFileName) -> Self {
+            match value {
+                LayerFileName::Delta(d) => Self::from(d),
+                LayerFileName::Image(i) => Self::from(i),
+            }
         }
     }
 }

pageserver/src/tenant/storage_layer/remote_layer.rs

@@ -218,15 +218,12 @@ impl RemoteLayer {
     }
 
     /// Create a Layer struct representing this layer, after it has been downloaded.
-    pub fn create_downloaded_layer<L>(
+    pub fn create_downloaded_layer(
         &self,
-        layer_map_lock_held_witness: &BatchedUpdates<'_, L>,
+        layer_map_lock_held_witness: &BatchedUpdates<'_>,
         conf: &'static PageServerConf,
         file_size: u64,
-    ) -> Arc<dyn PersistentLayer>
+    ) -> Arc<dyn PersistentLayer> {
-    where
-        L: ?Sized + Layer,
-    {
         if self.desc.is_delta {
             let fname = self.desc.delta_file_name();
             Arc::new(DeltaLayer::new(

pageserver/src/tenant/timeline/eviction_task.rs

@@ -197,9 +197,11 @@ impl Timeline {
         // We don't want to hold the layer map lock during eviction.
         // So, we just need to deal with this.
         let candidates: Vec<Arc<dyn PersistentLayer>> = {
-            let layers = self.layers.read().await;
+            let guard = self.layers.read().await;
+            let (layers, mapping) = &*guard;
             let mut candidates = Vec::new();
             for hist_layer in layers.iter_historic_layers() {
+                let hist_layer = mapping.get_from_desc(&hist_layer);
                 if hist_layer.is_remote_layer() {
                     continue;
                 }

pgxn/neon/Makefile

@@ -4,6 +4,7 @@
 MODULE_big = neon
 OBJS = \
 	$(WIN32RES) \
+	extension_server.o \
 	file_cache.o \
 	libpagestore.o \
 	libpqwalproposer.o \

pgxn/neon/control_plane_connector.c

@@ -32,6 +32,7 @@
 #include "port.h"
 #include <curl/curl.h>
 #include "utils/jsonb.h"
+#include "libpq/crypt.h"
 
 static ProcessUtility_hook_type PreviousProcessUtilityHook = NULL;
 
@@ -161,7 +162,22 @@ ConstructDeltaMessage()
 			PushKeyValue(&state, "name", entry->name);
 			if (entry->password)
 			{
+#if PG_MAJORVERSION_NUM == 14
+				char	   *logdetail;
+#else
+				const char *logdetail;
+#endif
 				PushKeyValue(&state, "password", (char *) entry->password);
+				char	   *encrypted_password = get_role_password(entry->name, &logdetail);
+
+				if (encrypted_password)
+				{
+					PushKeyValue(&state, "encrypted_password", encrypted_password);
+				}
+				else
+				{
+					elog(ERROR, "Failed to get encrypted password: %s", logdetail);
+				}
 			}
 			if (entry->old_name[0] != '\0')
 			{

pgxn/neon/extension_server.c

@@ -0,0 +1,104 @@
+
+/*-------------------------------------------------------------------------
+ *
+ * extension_server.c
+ *	  Request compute_ctl to download extension files.
+ *
+ * IDENTIFICATION
+ *	 contrib/neon/extension_server.c
+ *
+ *-------------------------------------------------------------------------
+ */
+#include "postgres.h"
+#include "tcop/pquery.h"
+#include "tcop/utility.h"
+#include "access/xact.h"
+#include "utils/hsearch.h"
+#include "utils/memutils.h"
+#include "commands/defrem.h"
+#include "miscadmin.h"
+#include "utils/acl.h"
+#include "fmgr.h"
+#include "utils/guc.h"
+#include "port.h"
+#include "fmgr.h"
+
+#include <curl/curl.h>
+
+static int extension_server_port = 0;
+
+static download_extension_file_hook_type prev_download_extension_file_hook = NULL;
+
+// to download all SQL (and data) files for an extension:
+// curl -X POST http://localhost:8080/extension_server/postgis
+// it covers two possible extension files layouts:
+// 1. extension_name--version--platform.sql
+// 2. extension_name/extension_name--version.sql
+//    extension_name/extra_files.csv
+//
+// to download specific library file:
+// curl -X POST http://localhost:8080/extension_server/postgis-3.so?is_library=true
+static bool
+neon_download_extension_file_http(const char *filename, bool is_library)
+{
+    CURL *curl;
+    CURLcode res;
+    char *compute_ctl_url;
+    char *postdata;
+    bool ret = false;
+
+    if ((curl = curl_easy_init()) == NULL)
+    {
+        elog(ERROR, "Failed to initialize curl handle");
+    }
+
+    compute_ctl_url = psprintf("http://localhost:%d/extension_server/%s%s",
+                               extension_server_port, filename, is_library ? "?is_library=true" : "");
+
+    elog(LOG, "Sending request to compute_ctl: %s", compute_ctl_url);
+
+    curl_easy_setopt(curl, CURLOPT_CUSTOMREQUEST, "POST");
+    curl_easy_setopt(curl, CURLOPT_URL, compute_ctl_url);
+    // NOTE: 15L may be insufficient time for large extensions like postgis
+    curl_easy_setopt(curl, CURLOPT_TIMEOUT, 15L /* seconds */);
+
+    if (curl)
+    {
+        /* Perform the request, res will get the return code */
+        res = curl_easy_perform(curl);
+        /* Check for errors */
+        if (res == CURLE_OK)
+        {
+            ret = true;
+        }
+        else
+        {
+            // Don't error here because postgres will try to find the file
+            // and will fail with some proper error message if it's not found.
+            elog(WARNING, "neon_download_extension_file_http failed: %s\n", curl_easy_strerror(res));
+        }
+
+        /* always cleanup */
+        curl_easy_cleanup(curl);
+    }
+
+    return ret;
+}
+
+void pg_init_extension_server()
+{
+    // Port to connect to compute_ctl on localhost
+    // to request extension files.
+    DefineCustomIntVariable("neon.extension_server_port",
+                            "connection string to the compute_ctl",
+                            NULL,
+                            &extension_server_port,
+                            0, 0, INT_MAX,
+                            PGC_POSTMASTER,
+                            0, /* no flags required */
+                            NULL, NULL, NULL);
+
+    // set download_extension_file_hook
+    prev_download_extension_file_hook = download_extension_file_hook;
+    download_extension_file_hook = neon_download_extension_file_http;
+}

pgxn/neon/extension_server.h

@@ -0,0 +1 @@
+

pgxn/neon/file_cache.c

@@ -190,7 +190,7 @@ lfc_change_limit_hook(int newval, void *extra)
 		hash_search(lfc_hash, &victim->key, HASH_REMOVE, NULL);
 		lfc_ctl->used -= 1;
 	}
-	elog(LOG, "set local file cache limit to %d", new_size);
+	elog(DEBUG1, "set local file cache limit to %d", new_size);
 	LWLockRelease(lfc_lock);
 }
 

pgxn/neon/neon.c

@@ -35,8 +35,11 @@ _PG_init(void)
 {
 	pg_init_libpagestore();
 	pg_init_walproposer();
+
 	InitControlPlaneConnector();
 
+	pg_init_extension_server();
+
         // Important: This must happen after other parts of the extension
         // are loaded, otherwise any settings to GUCs that were set before
         // the extension was loaded will be removed.

pgxn/neon/neon.h

@@ -21,6 +21,8 @@ extern char *neon_tenant;
 extern void pg_init_libpagestore(void);
 extern void pg_init_walproposer(void);
 
+extern void pg_init_extension_server(void);
+
 /*
  * Returns true if we shouldn't do REDO on that block in record indicated by
  * block_id; false otherwise.

pgxn/neon/pagestore_smgr.c

@@ -2675,7 +2675,6 @@ bool
 neon_redo_read_buffer_filter(XLogReaderState *record, uint8 block_id)
 {
 	XLogRecPtr	end_recptr = record->EndRecPtr;
-	XLogRecPtr	prev_end_recptr = record->ReadRecPtr - 1;
 	RelFileNode	rnode;
 	ForkNumber	forknum;
 	BlockNumber	blkno;
@@ -2719,16 +2718,15 @@ neon_redo_read_buffer_filter(XLogReaderState *record, uint8 block_id)
 
 	no_redo_needed = buffer < 0;
 
-	/* we don't have the buffer in memory, update lwLsn past this record */
+	/* In both cases st lwlsn past this WAL record */
+	SetLastWrittenLSNForBlock(end_recptr, rnode, forknum, blkno);
+
+	/* we don't have the buffer in memory, update lwLsn past this record,
+	 * also evict page fro file cache
+	 */
 	if (no_redo_needed)
-	{
-		SetLastWrittenLSNForBlock(end_recptr, rnode, forknum, blkno);
 		lfc_evict(rnode, forknum, blkno);
-	}
+
-	else
-	{
-		SetLastWrittenLSNForBlock(prev_end_recptr, rnode, forknum, blkno);
-	}
 
 	LWLockRelease(partitionLock);
 
@@ -2736,7 +2734,10 @@ neon_redo_read_buffer_filter(XLogReaderState *record, uint8 block_id)
 	if (get_cached_relsize(rnode, forknum, &relsize))
 	{
 		if (relsize < blkno + 1)
+		{
 			update_cached_relsize(rnode, forknum, blkno + 1);
+			SetLastWrittenLSNForRelation(end_recptr, rnode, forknum);
+		}
 	}
 	else
 	{
@@ -2768,6 +2769,7 @@ neon_redo_read_buffer_filter(XLogReaderState *record, uint8 block_id)
 		Assert(nbresponse->n_blocks > blkno);
 
 		set_cached_relsize(rnode, forknum, nbresponse->n_blocks);
+		SetLastWrittenLSNForRelation(end_recptr, rnode, forknum);
 
 		elog(SmgrTrace, "Set length to %d", nbresponse->n_blocks);
 	}

poetry.lock

@@ -1654,71 +1654,74 @@ test = ["enum34", "ipaddress", "mock", "pywin32", "wmi"]
 
 [[package]]
 name = "psycopg2-binary"
-version = "2.9.3"
+version = "2.9.6"
 description = "psycopg2 - Python-PostgreSQL Database Adapter"
 category = "main"
 optional = false
 python-versions = ">=3.6"
 files = [
-    {file = "psycopg2-binary-2.9.3.tar.gz", hash = "sha256:761df5313dc15da1502b21453642d7599d26be88bff659382f8f9747c7ebea4e"},
+    {file = "psycopg2-binary-2.9.6.tar.gz", hash = "sha256:1f64dcfb8f6e0c014c7f55e51c9759f024f70ea572fbdef123f85318c297947c"},
-    {file = "psycopg2_binary-2.9.3-cp310-cp310-macosx_10_14_x86_64.macosx_10_9_intel.macosx_10_9_x86_64.macosx_10_10_intel.macosx_10_10_x86_64.whl", hash = "sha256:539b28661b71da7c0e428692438efbcd048ca21ea81af618d845e06ebfd29478"},
+    {file = "psycopg2_binary-2.9.6-cp310-cp310-macosx_10_9_x86_64.whl", hash = "sha256:d26e0342183c762de3276cca7a530d574d4e25121ca7d6e4a98e4f05cb8e4df7"},
-    {file = "psycopg2_binary-2.9.3-cp310-cp310-macosx_11_0_arm64.whl", hash = "sha256:2f2534ab7dc7e776a263b463a16e189eb30e85ec9bbe1bff9e78dae802608932"},
+    {file = "psycopg2_binary-2.9.6-cp310-cp310-macosx_11_0_arm64.whl", hash = "sha256:c48d8f2db17f27d41fb0e2ecd703ea41984ee19362cbce52c097963b3a1b4365"},
-    {file = "psycopg2_binary-2.9.3-cp310-cp310-manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:6e82d38390a03da28c7985b394ec3f56873174e2c88130e6966cb1c946508e65"},
+    {file = "psycopg2_binary-2.9.6-cp310-cp310-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:ffe9dc0a884a8848075e576c1de0290d85a533a9f6e9c4e564f19adf8f6e54a7"},
-    {file = "psycopg2_binary-2.9.3-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:57804fc02ca3ce0dbfbef35c4b3a4a774da66d66ea20f4bda601294ad2ea6092"},
+    {file = "psycopg2_binary-2.9.6-cp310-cp310-manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:8a76e027f87753f9bd1ab5f7c9cb8c7628d1077ef927f5e2446477153a602f2c"},
-    {file = "psycopg2_binary-2.9.3-cp310-cp310-manylinux_2_24_aarch64.whl", hash = "sha256:083a55275f09a62b8ca4902dd11f4b33075b743cf0d360419e2051a8a5d5ff76"},
+    {file = "psycopg2_binary-2.9.6-cp310-cp310-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:6460c7a99fc939b849431f1e73e013d54aa54293f30f1109019c56a0b2b2ec2f"},
-    {file = "psycopg2_binary-2.9.3-cp310-cp310-manylinux_2_24_ppc64le.whl", hash = "sha256:0a29729145aaaf1ad8bafe663131890e2111f13416b60e460dae0a96af5905c9"},
+    {file = "psycopg2_binary-2.9.6-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:ae102a98c547ee2288637af07393dd33f440c25e5cd79556b04e3fca13325e5f"},
-    {file = "psycopg2_binary-2.9.3-cp310-cp310-musllinux_1_1_aarch64.whl", hash = "sha256:3a79d622f5206d695d7824cbf609a4f5b88ea6d6dab5f7c147fc6d333a8787e4"},
+    {file = "psycopg2_binary-2.9.6-cp310-cp310-musllinux_1_1_aarch64.whl", hash = "sha256:9972aad21f965599ed0106f65334230ce826e5ae69fda7cbd688d24fa922415e"},
-    {file = "psycopg2_binary-2.9.3-cp310-cp310-musllinux_1_1_i686.whl", hash = "sha256:090f3348c0ab2cceb6dfbe6bf721ef61262ddf518cd6cc6ecc7d334996d64efa"},
+    {file = "psycopg2_binary-2.9.6-cp310-cp310-musllinux_1_1_i686.whl", hash = "sha256:7a40c00dbe17c0af5bdd55aafd6ff6679f94a9be9513a4c7e071baf3d7d22a70"},
-    {file = "psycopg2_binary-2.9.3-cp310-cp310-musllinux_1_1_ppc64le.whl", hash = "sha256:a9e1f75f96ea388fbcef36c70640c4efbe4650658f3d6a2967b4cc70e907352e"},
+    {file = "psycopg2_binary-2.9.6-cp310-cp310-musllinux_1_1_ppc64le.whl", hash = "sha256:cacbdc5839bdff804dfebc058fe25684cae322987f7a38b0168bc1b2df703fb1"},
-    {file = "psycopg2_binary-2.9.3-cp310-cp310-musllinux_1_1_x86_64.whl", hash = "sha256:c3ae8e75eb7160851e59adc77b3a19a976e50622e44fd4fd47b8b18208189d42"},
+    {file = "psycopg2_binary-2.9.6-cp310-cp310-musllinux_1_1_x86_64.whl", hash = "sha256:7f0438fa20fb6c7e202863e0d5ab02c246d35efb1d164e052f2f3bfe2b152bd0"},
-    {file = "psycopg2_binary-2.9.3-cp310-cp310-win32.whl", hash = "sha256:7b1e9b80afca7b7a386ef087db614faebbf8839b7f4db5eb107d0f1a53225029"},
+    {file = "psycopg2_binary-2.9.6-cp310-cp310-win32.whl", hash = "sha256:b6c8288bb8a84b47e07013bb4850f50538aa913d487579e1921724631d02ea1b"},
-    {file = "psycopg2_binary-2.9.3-cp310-cp310-win_amd64.whl", hash = "sha256:8b344adbb9a862de0c635f4f0425b7958bf5a4b927c8594e6e8d261775796d53"},
+    {file = "psycopg2_binary-2.9.6-cp310-cp310-win_amd64.whl", hash = "sha256:61b047a0537bbc3afae10f134dc6393823882eb263088c271331602b672e52e9"},
-    {file = "psycopg2_binary-2.9.3-cp36-cp36m-macosx_10_14_x86_64.macosx_10_9_intel.macosx_10_9_x86_64.macosx_10_10_intel.macosx_10_10_x86_64.whl", hash = "sha256:e847774f8ffd5b398a75bc1c18fbb56564cda3d629fe68fd81971fece2d3c67e"},
+    {file = "psycopg2_binary-2.9.6-cp311-cp311-macosx_10_9_x86_64.whl", hash = "sha256:964b4dfb7c1c1965ac4c1978b0f755cc4bd698e8aa2b7667c575fb5f04ebe06b"},
-    {file = "psycopg2_binary-2.9.3-cp36-cp36m-manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:68641a34023d306be959101b345732360fc2ea4938982309b786f7be1b43a4a1"},
+    {file = "psycopg2_binary-2.9.6-cp311-cp311-macosx_11_0_arm64.whl", hash = "sha256:afe64e9b8ea66866a771996f6ff14447e8082ea26e675a295ad3bdbffdd72afb"},
-    {file = "psycopg2_binary-2.9.3-cp36-cp36m-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:3303f8807f342641851578ee7ed1f3efc9802d00a6f83c101d21c608cb864460"},
+    {file = "psycopg2_binary-2.9.6-cp311-cp311-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:15e2ee79e7cf29582ef770de7dab3d286431b01c3bb598f8e05e09601b890081"},
-    {file = "psycopg2_binary-2.9.3-cp36-cp36m-manylinux_2_24_aarch64.whl", hash = "sha256:e3699852e22aa68c10de06524a3721ade969abf382da95884e6a10ff798f9281"},
+    {file = "psycopg2_binary-2.9.6-cp311-cp311-manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:dfa74c903a3c1f0d9b1c7e7b53ed2d929a4910e272add6700c38f365a6002820"},
-    {file = "psycopg2_binary-2.9.3-cp36-cp36m-manylinux_2_24_ppc64le.whl", hash = "sha256:526ea0378246d9b080148f2d6681229f4b5964543c170dd10bf4faaab6e0d27f"},
+    {file = "psycopg2_binary-2.9.6-cp311-cp311-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:b83456c2d4979e08ff56180a76429263ea254c3f6552cd14ada95cff1dec9bb8"},
-    {file = "psycopg2_binary-2.9.3-cp36-cp36m-musllinux_1_1_aarch64.whl", hash = "sha256:b1c8068513f5b158cf7e29c43a77eb34b407db29aca749d3eb9293ee0d3103ca"},
+    {file = "psycopg2_binary-2.9.6-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:0645376d399bfd64da57148694d78e1f431b1e1ee1054872a5713125681cf1be"},
-    {file = "psycopg2_binary-2.9.3-cp36-cp36m-musllinux_1_1_i686.whl", hash = "sha256:15803fa813ea05bef089fa78835118b5434204f3a17cb9f1e5dbfd0b9deea5af"},
+    {file = "psycopg2_binary-2.9.6-cp311-cp311-musllinux_1_1_aarch64.whl", hash = "sha256:e99e34c82309dd78959ba3c1590975b5d3c862d6f279f843d47d26ff89d7d7e1"},
-    {file = "psycopg2_binary-2.9.3-cp36-cp36m-musllinux_1_1_ppc64le.whl", hash = "sha256:152f09f57417b831418304c7f30d727dc83a12761627bb826951692cc6491e57"},
+    {file = "psycopg2_binary-2.9.6-cp311-cp311-musllinux_1_1_i686.whl", hash = "sha256:4ea29fc3ad9d91162c52b578f211ff1c931d8a38e1f58e684c45aa470adf19e2"},
-    {file = "psycopg2_binary-2.9.3-cp36-cp36m-musllinux_1_1_x86_64.whl", hash = "sha256:404224e5fef3b193f892abdbf8961ce20e0b6642886cfe1fe1923f41aaa75c9d"},
+    {file = "psycopg2_binary-2.9.6-cp311-cp311-musllinux_1_1_ppc64le.whl", hash = "sha256:4ac30da8b4f57187dbf449294d23b808f8f53cad6b1fc3623fa8a6c11d176dd0"},
-    {file = "psycopg2_binary-2.9.3-cp36-cp36m-win32.whl", hash = "sha256:1f6b813106a3abdf7b03640d36e24669234120c72e91d5cbaeb87c5f7c36c65b"},
+    {file = "psycopg2_binary-2.9.6-cp311-cp311-musllinux_1_1_x86_64.whl", hash = "sha256:e78e6e2a00c223e164c417628572a90093c031ed724492c763721c2e0bc2a8df"},
-    {file = "psycopg2_binary-2.9.3-cp36-cp36m-win_amd64.whl", hash = "sha256:2d872e3c9d5d075a2e104540965a1cf898b52274a5923936e5bfddb58c59c7c2"},
+    {file = "psycopg2_binary-2.9.6-cp311-cp311-win32.whl", hash = "sha256:1876843d8e31c89c399e31b97d4b9725a3575bb9c2af92038464231ec40f9edb"},
-    {file = "psycopg2_binary-2.9.3-cp37-cp37m-macosx_10_14_x86_64.macosx_10_9_intel.macosx_10_9_x86_64.macosx_10_10_intel.macosx_10_10_x86_64.whl", hash = "sha256:10bb90fb4d523a2aa67773d4ff2b833ec00857f5912bafcfd5f5414e45280fb1"},
+    {file = "psycopg2_binary-2.9.6-cp311-cp311-win_amd64.whl", hash = "sha256:b4b24f75d16a89cc6b4cdff0eb6a910a966ecd476d1e73f7ce5985ff1328e9a6"},
-    {file = "psycopg2_binary-2.9.3-cp37-cp37m-manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:874a52ecab70af13e899f7847b3e074eeb16ebac5615665db33bce8a1009cf33"},
+    {file = "psycopg2_binary-2.9.6-cp36-cp36m-win32.whl", hash = "sha256:498807b927ca2510baea1b05cc91d7da4718a0f53cb766c154c417a39f1820a0"},
-    {file = "psycopg2_binary-2.9.3-cp37-cp37m-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:a29b3ca4ec9defec6d42bf5feb36bb5817ba3c0230dd83b4edf4bf02684cd0ae"},
+    {file = "psycopg2_binary-2.9.6-cp36-cp36m-win_amd64.whl", hash = "sha256:0d236c2825fa656a2d98bbb0e52370a2e852e5a0ec45fc4f402977313329174d"},
-    {file = "psycopg2_binary-2.9.3-cp37-cp37m-manylinux_2_24_aarch64.whl", hash = "sha256:12b11322ea00ad8db8c46f18b7dfc47ae215e4df55b46c67a94b4effbaec7094"},
+    {file = "psycopg2_binary-2.9.6-cp37-cp37m-macosx_10_9_x86_64.whl", hash = "sha256:34b9ccdf210cbbb1303c7c4db2905fa0319391bd5904d32689e6dd5c963d2ea8"},
-    {file = "psycopg2_binary-2.9.3-cp37-cp37m-manylinux_2_24_ppc64le.whl", hash = "sha256:53293533fcbb94c202b7c800a12c873cfe24599656b341f56e71dd2b557be063"},
+    {file = "psycopg2_binary-2.9.6-cp37-cp37m-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:84d2222e61f313c4848ff05353653bf5f5cf6ce34df540e4274516880d9c3763"},
-    {file = "psycopg2_binary-2.9.3-cp37-cp37m-musllinux_1_1_aarch64.whl", hash = "sha256:c381bda330ddf2fccbafab789d83ebc6c53db126e4383e73794c74eedce855ef"},
+    {file = "psycopg2_binary-2.9.6-cp37-cp37m-manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:30637a20623e2a2eacc420059be11527f4458ef54352d870b8181a4c3020ae6b"},
-    {file = "psycopg2_binary-2.9.3-cp37-cp37m-musllinux_1_1_i686.whl", hash = "sha256:9d29409b625a143649d03d0fd7b57e4b92e0ecad9726ba682244b73be91d2fdb"},
+    {file = "psycopg2_binary-2.9.6-cp37-cp37m-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:8122cfc7cae0da9a3077216528b8bb3629c43b25053284cc868744bfe71eb141"},
-    {file = "psycopg2_binary-2.9.3-cp37-cp37m-musllinux_1_1_ppc64le.whl", hash = "sha256:183a517a3a63503f70f808b58bfbf962f23d73b6dccddae5aa56152ef2bcb232"},
+    {file = "psycopg2_binary-2.9.6-cp37-cp37m-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:38601cbbfe600362c43714482f43b7c110b20cb0f8172422c616b09b85a750c5"},
-    {file = "psycopg2_binary-2.9.3-cp37-cp37m-musllinux_1_1_x86_64.whl", hash = "sha256:15c4e4cfa45f5a60599d9cec5f46cd7b1b29d86a6390ec23e8eebaae84e64554"},
+    {file = "psycopg2_binary-2.9.6-cp37-cp37m-musllinux_1_1_aarch64.whl", hash = "sha256:c7e62ab8b332147a7593a385d4f368874d5fe4ad4e341770d4983442d89603e3"},
-    {file = "psycopg2_binary-2.9.3-cp37-cp37m-win32.whl", hash = "sha256:adf20d9a67e0b6393eac162eb81fb10bc9130a80540f4df7e7355c2dd4af9fba"},
+    {file = "psycopg2_binary-2.9.6-cp37-cp37m-musllinux_1_1_i686.whl", hash = "sha256:2ab652e729ff4ad76d400df2624d223d6e265ef81bb8aa17fbd63607878ecbee"},
-    {file = "psycopg2_binary-2.9.3-cp37-cp37m-win_amd64.whl", hash = "sha256:2f9ffd643bc7349eeb664eba8864d9e01f057880f510e4681ba40a6532f93c71"},
+    {file = "psycopg2_binary-2.9.6-cp37-cp37m-musllinux_1_1_ppc64le.whl", hash = "sha256:c83a74b68270028dc8ee74d38ecfaf9c90eed23c8959fca95bd703d25b82c88e"},
-    {file = "psycopg2_binary-2.9.3-cp38-cp38-macosx_10_14_x86_64.macosx_10_9_intel.macosx_10_9_x86_64.macosx_10_10_intel.macosx_10_10_x86_64.whl", hash = "sha256:def68d7c21984b0f8218e8a15d514f714d96904265164f75f8d3a70f9c295667"},
+    {file = "psycopg2_binary-2.9.6-cp37-cp37m-musllinux_1_1_x86_64.whl", hash = "sha256:d4e6036decf4b72d6425d5b29bbd3e8f0ff1059cda7ac7b96d6ac5ed34ffbacd"},
-    {file = "psycopg2_binary-2.9.3-cp38-cp38-macosx_11_0_arm64.whl", hash = "sha256:e6aa71ae45f952a2205377773e76f4e3f27951df38e69a4c95440c779e013560"},
+    {file = "psycopg2_binary-2.9.6-cp37-cp37m-win32.whl", hash = "sha256:a8c28fd40a4226b4a84bdf2d2b5b37d2c7bd49486b5adcc200e8c7ec991dfa7e"},
-    {file = "psycopg2_binary-2.9.3-cp38-cp38-manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:dffc08ca91c9ac09008870c9eb77b00a46b3378719584059c034b8945e26b272"},
+    {file = "psycopg2_binary-2.9.6-cp37-cp37m-win_amd64.whl", hash = "sha256:51537e3d299be0db9137b321dfb6a5022caaab275775680e0c3d281feefaca6b"},
-    {file = "psycopg2_binary-2.9.3-cp38-cp38-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:280b0bb5cbfe8039205c7981cceb006156a675362a00fe29b16fbc264e242834"},
+    {file = "psycopg2_binary-2.9.6-cp38-cp38-macosx_10_9_x86_64.whl", hash = "sha256:cf4499e0a83b7b7edcb8dabecbd8501d0d3a5ef66457200f77bde3d210d5debb"},
-    {file = "psycopg2_binary-2.9.3-cp38-cp38-manylinux_2_24_aarch64.whl", hash = "sha256:af9813db73395fb1fc211bac696faea4ca9ef53f32dc0cfa27e4e7cf766dcf24"},
+    {file = "psycopg2_binary-2.9.6-cp38-cp38-macosx_11_0_arm64.whl", hash = "sha256:7e13a5a2c01151f1208d5207e42f33ba86d561b7a89fca67c700b9486a06d0e2"},
-    {file = "psycopg2_binary-2.9.3-cp38-cp38-manylinux_2_24_ppc64le.whl", hash = "sha256:63638d875be8c2784cfc952c9ac34e2b50e43f9f0a0660b65e2a87d656b3116c"},
+    {file = "psycopg2_binary-2.9.6-cp38-cp38-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:0e0f754d27fddcfd74006455b6e04e6705d6c31a612ec69ddc040a5468e44b4e"},
-    {file = "psycopg2_binary-2.9.3-cp38-cp38-musllinux_1_1_aarch64.whl", hash = "sha256:ffb7a888a047696e7f8240d649b43fb3644f14f0ee229077e7f6b9f9081635bd"},
+    {file = "psycopg2_binary-2.9.6-cp38-cp38-manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:d57c3fd55d9058645d26ae37d76e61156a27722097229d32a9e73ed54819982a"},
-    {file = "psycopg2_binary-2.9.3-cp38-cp38-musllinux_1_1_i686.whl", hash = "sha256:0c9d5450c566c80c396b7402895c4369a410cab5a82707b11aee1e624da7d004"},
+    {file = "psycopg2_binary-2.9.6-cp38-cp38-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:71f14375d6f73b62800530b581aed3ada394039877818b2d5f7fc77e3bb6894d"},
-    {file = "psycopg2_binary-2.9.3-cp38-cp38-musllinux_1_1_ppc64le.whl", hash = "sha256:d1c1b569ecafe3a69380a94e6ae09a4789bbb23666f3d3a08d06bbd2451f5ef1"},
+    {file = "psycopg2_binary-2.9.6-cp38-cp38-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:441cc2f8869a4f0f4bb408475e5ae0ee1f3b55b33f350406150277f7f35384fc"},
-    {file = "psycopg2_binary-2.9.3-cp38-cp38-musllinux_1_1_x86_64.whl", hash = "sha256:8fc53f9af09426a61db9ba357865c77f26076d48669f2e1bb24d85a22fb52307"},
+    {file = "psycopg2_binary-2.9.6-cp38-cp38-musllinux_1_1_aarch64.whl", hash = "sha256:65bee1e49fa6f9cf327ce0e01c4c10f39165ee76d35c846ade7cb0ec6683e303"},
-    {file = "psycopg2_binary-2.9.3-cp38-cp38-win32.whl", hash = "sha256:6472a178e291b59e7f16ab49ec8b4f3bdada0a879c68d3817ff0963e722a82ce"},
+    {file = "psycopg2_binary-2.9.6-cp38-cp38-musllinux_1_1_i686.whl", hash = "sha256:af335bac6b666cc6aea16f11d486c3b794029d9df029967f9938a4bed59b6a19"},
-    {file = "psycopg2_binary-2.9.3-cp38-cp38-win_amd64.whl", hash = "sha256:35168209c9d51b145e459e05c31a9eaeffa9a6b0fd61689b48e07464ffd1a83e"},
+    {file = "psycopg2_binary-2.9.6-cp38-cp38-musllinux_1_1_ppc64le.whl", hash = "sha256:cfec476887aa231b8548ece2e06d28edc87c1397ebd83922299af2e051cf2827"},
-    {file = "psycopg2_binary-2.9.3-cp39-cp39-macosx_10_14_x86_64.macosx_10_9_intel.macosx_10_9_x86_64.macosx_10_10_intel.macosx_10_10_x86_64.whl", hash = "sha256:47133f3f872faf28c1e87d4357220e809dfd3fa7c64295a4a148bcd1e6e34ec9"},
+    {file = "psycopg2_binary-2.9.6-cp38-cp38-musllinux_1_1_x86_64.whl", hash = "sha256:65c07febd1936d63bfde78948b76cd4c2a411572a44ac50719ead41947d0f26b"},
-    {file = "psycopg2_binary-2.9.3-cp39-cp39-macosx_11_0_arm64.whl", hash = "sha256:b3a24a1982ae56461cc24f6680604fffa2c1b818e9dc55680da038792e004d18"},
+    {file = "psycopg2_binary-2.9.6-cp38-cp38-win32.whl", hash = "sha256:4dfb4be774c4436a4526d0c554af0cc2e02082c38303852a36f6456ece7b3503"},
-    {file = "psycopg2_binary-2.9.3-cp39-cp39-manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:91920527dea30175cc02a1099f331aa8c1ba39bf8b7762b7b56cbf54bc5cce42"},
+    {file = "psycopg2_binary-2.9.6-cp38-cp38-win_amd64.whl", hash = "sha256:02c6e3cf3439e213e4ee930308dc122d6fb4d4bea9aef4a12535fbd605d1a2fe"},
-    {file = "psycopg2_binary-2.9.3-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:887dd9aac71765ac0d0bac1d0d4b4f2c99d5f5c1382d8b770404f0f3d0ce8a39"},
+    {file = "psycopg2_binary-2.9.6-cp39-cp39-macosx_10_9_x86_64.whl", hash = "sha256:e9182eb20f41417ea1dd8e8f7888c4d7c6e805f8a7c98c1081778a3da2bee3e4"},
-    {file = "psycopg2_binary-2.9.3-cp39-cp39-manylinux_2_24_aarch64.whl", hash = "sha256:1f14c8b0942714eb3c74e1e71700cbbcb415acbc311c730370e70c578a44a25c"},
+    {file = "psycopg2_binary-2.9.6-cp39-cp39-macosx_11_0_arm64.whl", hash = "sha256:8a6979cf527e2603d349a91060f428bcb135aea2be3201dff794813256c274f1"},
-    {file = "psycopg2_binary-2.9.3-cp39-cp39-manylinux_2_24_ppc64le.whl", hash = "sha256:7af0dd86ddb2f8af5da57a976d27cd2cd15510518d582b478fbb2292428710b4"},
+    {file = "psycopg2_binary-2.9.6-cp39-cp39-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:8338a271cb71d8da40b023a35d9c1e919eba6cbd8fa20a54b748a332c355d896"},
-    {file = "psycopg2_binary-2.9.3-cp39-cp39-musllinux_1_1_aarch64.whl", hash = "sha256:93cd1967a18aa0edd4b95b1dfd554cf15af657cb606280996d393dadc88c3c35"},
+    {file = "psycopg2_binary-2.9.6-cp39-cp39-manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:e3ed340d2b858d6e6fb5083f87c09996506af483227735de6964a6100b4e6a54"},
-    {file = "psycopg2_binary-2.9.3-cp39-cp39-musllinux_1_1_i686.whl", hash = "sha256:bda845b664bb6c91446ca9609fc69f7db6c334ec5e4adc87571c34e4f47b7ddb"},
+    {file = "psycopg2_binary-2.9.6-cp39-cp39-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:f81e65376e52f03422e1fb475c9514185669943798ed019ac50410fb4c4df232"},
-    {file = "psycopg2_binary-2.9.3-cp39-cp39-musllinux_1_1_ppc64le.whl", hash = "sha256:01310cf4cf26db9aea5158c217caa92d291f0500051a6469ac52166e1a16f5b7"},
+    {file = "psycopg2_binary-2.9.6-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:bfb13af3c5dd3a9588000910178de17010ebcccd37b4f9794b00595e3a8ddad3"},
-    {file = "psycopg2_binary-2.9.3-cp39-cp39-musllinux_1_1_x86_64.whl", hash = "sha256:99485cab9ba0fa9b84f1f9e1fef106f44a46ef6afdeec8885e0b88d0772b49e8"},
+    {file = "psycopg2_binary-2.9.6-cp39-cp39-musllinux_1_1_aarch64.whl", hash = "sha256:4c727b597c6444a16e9119386b59388f8a424223302d0c06c676ec8b4bc1f963"},
-    {file = "psycopg2_binary-2.9.3-cp39-cp39-win32.whl", hash = "sha256:46f0e0a6b5fa5851bbd9ab1bc805eef362d3a230fbdfbc209f4a236d0a7a990d"},
+    {file = "psycopg2_binary-2.9.6-cp39-cp39-musllinux_1_1_i686.whl", hash = "sha256:4d67fbdaf177da06374473ef6f7ed8cc0a9dc640b01abfe9e8a2ccb1b1402c1f"},
-    {file = "psycopg2_binary-2.9.3-cp39-cp39-win_amd64.whl", hash = "sha256:accfe7e982411da3178ec690baaceaad3c278652998b2c45828aaac66cd8285f"},
+    {file = "psycopg2_binary-2.9.6-cp39-cp39-musllinux_1_1_ppc64le.whl", hash = "sha256:0892ef645c2fabb0c75ec32d79f4252542d0caec1d5d949630e7d242ca4681a3"},
+    {file = "psycopg2_binary-2.9.6-cp39-cp39-musllinux_1_1_x86_64.whl", hash = "sha256:02c0f3757a4300cf379eb49f543fb7ac527fb00144d39246ee40e1df684ab514"},
+    {file = "psycopg2_binary-2.9.6-cp39-cp39-win32.whl", hash = "sha256:c3dba7dab16709a33a847e5cd756767271697041fbe3fe97c215b1fc1f5c9848"},
+    {file = "psycopg2_binary-2.9.6-cp39-cp39-win_amd64.whl", hash = "sha256:f6a88f384335bb27812293fdb11ac6aee2ca3f51d3c7820fe03de0a304ab6249"},
 ]
 
 [[package]]

proxy/Cargo.toml

@@ -7,7 +7,6 @@ license.workspace = true
 [dependencies]
 anyhow.workspace = true
 async-trait.workspace = true
-atty.workspace = true
 base64.workspace = true
 bstr.workspace = true
 bytes = { workspace = true, features = ["serde"] }
@@ -38,6 +37,7 @@ rand.workspace = true
 regex.workspace = true
 reqwest = { workspace = true, features = ["json"] }
 reqwest-middleware.workspace = true
+reqwest-retry.workspace = true
 reqwest-tracing.workspace = true
 routerify.workspace = true
 rustls-pemfile.workspace = true

proxy/src/console/mgmt.rs

@@ -8,7 +8,7 @@ use postgres_backend::{self, AuthType, PostgresBackend, PostgresBackendTCP, Quer
 use pq_proto::{BeMessage, SINGLE_COL_ROWDESC};
 use std::future;
 use tokio::net::{TcpListener, TcpStream};
-use tracing::{error, info, info_span};
+use tracing::{error, info, info_span, Instrument};
 
 static CPLANE_WAITERS: Lazy<Waiters<ComputeReady>> = Lazy::new(Default::default);
 
@@ -44,19 +44,30 @@ pub async fn task_main(listener: TcpListener) -> anyhow::Result<()> {
             .set_nodelay(true)
             .context("failed to set client socket option")?;
 
-        tokio::task::spawn(async move {
+        let span = info_span!("mgmt", peer = %peer_addr);
-            let span = info_span!("mgmt", peer = %peer_addr);
-            let _enter = span.enter();
 
-            info!("started a new console management API thread");
+        tokio::task::spawn(
-            scopeguard::defer! {
+            async move {
-                info!("console management API thread is about to finish");
+                info!("serving a new console management API connection");
-            }
 
-            if let Err(e) = handle_connection(socket).await {
+                // these might be long running connections, have a separate logging for cancelling
-                error!("thread failed with an error: {e}");
+                // on shutdown and other ways of stopping.
+                let cancelled = scopeguard::guard(tracing::Span::current(), |span| {
+                    let _e = span.entered();
+                    info!("console management API task cancelled");
+                });
+
+                if let Err(e) = handle_connection(socket).await {
+                    error!("serving failed with an error: {e}");
+                } else {
+                    info!("serving completed");
+                }
+
+                // we can no longer get dropped
+                scopeguard::ScopeGuard::into_inner(cancelled);
             }
-        });
+            .instrument(span),
+        );
     }
 }
 
@@ -77,14 +88,14 @@ impl postgres_backend::Handler<tokio::net::TcpStream> for MgmtHandler {
         pgb: &mut PostgresBackendTCP,
         query: &str,
     ) -> Result<(), QueryError> {
-        try_process_query(pgb, query).await.map_err(|e| {
+        try_process_query(pgb, query).map_err(|e| {
             error!("failed to process response: {e:?}");
             e
         })
     }
 }
 
-async fn try_process_query(pgb: &mut PostgresBackendTCP, query: &str) -> Result<(), QueryError> {
+fn try_process_query(pgb: &mut PostgresBackendTCP, query: &str) -> Result<(), QueryError> {
     let resp: KickSession = serde_json::from_str(query).context("Failed to parse query as json")?;
 
     let span = info_span!("event", session_id = resp.session_id);

proxy/src/http.rs

@@ -6,8 +6,11 @@ pub mod server;
 pub mod sql_over_http;
 pub mod websocket;
 
+use std::time::Duration;
+
 pub use reqwest::{Request, Response, StatusCode};
 pub use reqwest_middleware::{ClientWithMiddleware, Error};
+pub use reqwest_retry::{policies::ExponentialBackoff, RetryTransientMiddleware};
 
 use crate::url::ApiUrl;
 use reqwest_middleware::RequestBuilder;
@@ -21,6 +24,24 @@ pub fn new_client() -> ClientWithMiddleware {
         .build()
 }
 
+pub fn new_client_with_timeout(default_timout: Duration) -> ClientWithMiddleware {
+    let timeout_client = reqwest::ClientBuilder::new()
+        .timeout(default_timout)
+        .build()
+        .expect("Failed to create http client with timeout");
+
+    let retry_policy =
+        ExponentialBackoff::builder().build_with_total_retry_duration(default_timout);
+
+    reqwest_middleware::ClientBuilder::new(timeout_client)
+        .with(reqwest_tracing::TracingMiddleware::default())
+        // As per docs, "This middleware always errors when given requests with streaming bodies".
+        // That's all right because we only use this client to send `serde_json::RawValue`, which
+        // is not a stream.
+        .with(RetryTransientMiddleware::new_with_policy(retry_policy))
+        .build()
+}
+
 /// Thin convenience wrapper for an API provided by an http endpoint.
 #[derive(Debug, Clone)]
 pub struct Endpoint {

proxy/src/logging.rs

@@ -18,7 +18,7 @@ pub async fn init() -> anyhow::Result<LoggingGuard> {
         .from_env_lossy();
 
     let fmt_layer = tracing_subscriber::fmt::layer()
-        .with_ansi(atty::is(atty::Stream::Stderr))
+        .with_ansi(false)
         .with_writer(std::io::stderr)
         .with_target(false);
 

proxy/src/metrics.rs

@@ -4,11 +4,13 @@ use crate::{config::MetricCollectionConfig, http};
 use chrono::{DateTime, Utc};
 use consumption_metrics::{idempotency_key, Event, EventChunk, EventType, CHUNK_SIZE};
 use serde::Serialize;
-use std::collections::HashMap;
+use std::{collections::HashMap, time::Duration};
 use tracing::{error, info, instrument, trace, warn};
 
 const PROXY_IO_BYTES_PER_CLIENT: &str = "proxy_io_bytes_per_client";
 
+const DEFAULT_HTTP_REPORTING_TIMEOUT: Duration = Duration::from_secs(60);
+
 ///
 /// Key that uniquely identifies the object, this metric describes.
 /// Currently, endpoint_id is enough, but this may change later,
@@ -30,7 +32,7 @@ pub async fn task_main(config: &MetricCollectionConfig) -> anyhow::Result<()> {
         info!("metrics collector has shut down");
     }
 
-    let http_client = http::new_client();
+    let http_client = http::new_client_with_timeout(DEFAULT_HTTP_REPORTING_TIMEOUT);
     let mut cached_metrics: HashMap<Ids, (u64, DateTime<Utc>)> = HashMap::new();
     let hostname = hostname::get()?.as_os_str().to_string_lossy().into_owned();
 
@@ -182,36 +184,36 @@ async fn collect_metrics_iteration(
             }
         };
 
-        if res.status().is_success() {
+        if !res.status().is_success() {
-            // update cached metrics after they were sent successfully
-            for send_metric in chunk {
-                let stop_time = match send_metric.kind {
-                    EventType::Incremental { stop_time, .. } => stop_time,
-                    _ => unreachable!(),
-                };
-
-                cached_metrics
-                    .entry(Ids {
-                        endpoint_id: send_metric.extra.endpoint_id.clone(),
-                        branch_id: send_metric.extra.branch_id.clone(),
-                    })
-                    // update cached value (add delta) and time
-                    .and_modify(|e| {
-                        e.0 = e.0.saturating_add(send_metric.value);
-                        e.1 = stop_time
-                    })
-                    // cache new metric
-                    .or_insert((send_metric.value, stop_time));
-            }
-        } else {
             error!("metrics endpoint refused the sent metrics: {:?}", res);
-            for metric in chunk.iter() {
+            for metric in chunk.iter().filter(|metric| metric.value > (1u64 << 40)) {
                 // Report if the metric value is suspiciously large
-                if metric.value > (1u64 << 40) {
+                error!("potentially abnormal metric value: {:?}", metric);
-                    error!("potentially abnormal metric value: {:?}", metric);
-                }
             }
         }
+        // update cached metrics after they were sent
+        // (to avoid sending the same metrics twice)
+        // see the relevant discussion on why to do so even if the status is not success:
+        // https://github.com/neondatabase/neon/pull/4563#discussion_r1246710956
+        for send_metric in chunk {
+            let stop_time = match send_metric.kind {
+                EventType::Incremental { stop_time, .. } => stop_time,
+                _ => unreachable!(),
+            };
+
+            cached_metrics
+                .entry(Ids {
+                    endpoint_id: send_metric.extra.endpoint_id.clone(),
+                    branch_id: send_metric.extra.branch_id.clone(),
+                })
+                // update cached value (add delta) and time
+                .and_modify(|e| {
+                    e.0 = e.0.saturating_add(send_metric.value);
+                    e.1 = stop_time
+                })
+                // cache new metric
+                .or_insert((send_metric.value, stop_time));
+        }
     }
     Ok(())
 }

pyproject.toml

@@ -7,7 +7,7 @@ authors = []
 [tool.poetry.dependencies]
 python = "^3.9"
 pytest = "^7.3.1"
-psycopg2-binary = "^2.9.1"
+psycopg2-binary = "^2.9.6"
 typing-extensions = "^4.6.1"
 PyJWT = {version = "^2.1.0", extras = ["crypto"]}
 requests = "^2.31.0"

rust-toolchain.toml

@@ -1,5 +1,5 @@
 [toolchain]
-channel = "1.68.2"
+channel = "1.70.0"
 profile = "default"
 # The default profile includes rustc, rust-std, cargo, rust-docs, rustfmt and clippy.
 # https://rust-lang.github.io/rustup/concepts/profiles.html

scripts/comment-test-report.js

@@ -144,17 +144,24 @@ const reportSummary = async (params) => {
                 }
                 summary += `- \`${testName}\`: ${links.join(", ")}\n`
             }
-
-            const testsToRerun = Object.values(failedTests[pgVersion]).map(x => x[0].name)
-            const command = `DEFAULT_PG_VERSION=${pgVersion} scripts/pytest -k "${testsToRerun.join(" or ")}"`
-
-            summary += "```\n"
-            summary += `# Run failed on Postgres ${pgVersion} tests locally:\n`
-            summary += `${command}\n`
-            summary += "```\n"
         }
     }
 
+    if (failedTestsCount > 0) {
+        const testsToRerun = []
+        for (const pgVersion of Object.keys(failedTests)) {
+            for (const testName of Object.keys(failedTests[pgVersion])) {
+                testsToRerun.push(...failedTests[pgVersion][testName].map(test => test.name))
+            }
+        }
+        const command = `scripts/pytest -vv -n $(nproc) -k "${testsToRerun.join(" or ")}"`
+
+        summary += "```\n"
+        summary += `# Run all failed tests locally:\n`
+        summary += `${command}\n`
+        summary += "```\n"
+    }
+
     if (flakyTestsCount > 0) {
         summary += `<details>\n<summary>Flaky tests (${flakyTestsCount})</summary>\n\n`
         for (const pgVersion of Array.from(pgVersions).sort().reverse()) {
@@ -164,8 +171,7 @@ const reportSummary = async (params) => {
                     const links = []
                     for (const test of tests) {
                         const allureLink = `${reportUrl}#suites/${test.parentUid}/${test.uid}/retries`
-                        const status = test.status === "passed" ? ":white_check_mark:" : ":x:"
+                        links.push(`[${test.buildType}](${allureLink})`)
-                        links.push(`[${status} ${test.buildType}](${allureLink})`)
                     }
                     summary += `- \`${testName}\`: ${links.join(", ")}\n`
                 }

test_runner/conftest.py

@@ -1,6 +1,6 @@
 pytest_plugins = (
     "fixtures.pg_version",
-    "fixtures.allure",
+    "fixtures.parametrize",
     "fixtures.neon_fixtures",
     "fixtures.benchmark_fixture",
     "fixtures.pg_stats",

test_runner/fixtures/allure.py

@@ -1,25 +0,0 @@
-import os
-
-import pytest
-
-from fixtures.pg_version import DEFAULT_VERSION, PgVersion
-
-"""
-Set of utilities to make Allure report more informative.
-
-- It adds BUILD_TYPE and DEFAULT_PG_VERSION to the test names (only in test_runner/regress)
-to make tests distinguishable in Allure report.
-"""
-
-
-@pytest.fixture(scope="function", autouse=True)
-def allure_noop():
-    pass
-
-
-def pytest_generate_tests(metafunc):
-    if "test_runner/regress" in metafunc.definition._nodeid:
-        build_type = os.environ.get("BUILD_TYPE", "DEBUG").lower()
-        pg_version = PgVersion(os.environ.get("DEFAULT_PG_VERSION", DEFAULT_VERSION))
-
-        metafunc.parametrize("allure_noop", [f"{build_type}-pg{pg_version}"])

test_runner/fixtures/metrics.py

@@ -62,6 +62,7 @@ PAGESERVER_GLOBAL_METRICS: Tuple[str, ...] = (
     "pageserver_getpage_reconstruct_seconds_bucket",
     "pageserver_getpage_reconstruct_seconds_count",
     "pageserver_getpage_reconstruct_seconds_sum",
+    *[f"pageserver_basebackup_query_seconds_{x}" for x in ["bucket", "count", "sum"]],
 )
 
 PAGESERVER_PER_TENANT_METRICS: Tuple[str, ...] = (

test_runner/fixtures/neon_fixtures.py

@@ -102,8 +102,8 @@ def base_dir() -> Iterator[Path]:
     yield base_dir
 
 
-@pytest.fixture(scope="session")
+@pytest.fixture(scope="function")
-def neon_binpath(base_dir: Path) -> Iterator[Path]:
+def neon_binpath(base_dir: Path, build_type: str) -> Iterator[Path]:
     if os.getenv("REMOTE_ENV"):
         # we are in remote env and do not have neon binaries locally
         # this is the case for benchmarks run on self-hosted runner
@@ -113,7 +113,6 @@ def neon_binpath(base_dir: Path) -> Iterator[Path]:
     if env_neon_bin := os.environ.get("NEON_BIN"):
         binpath = Path(env_neon_bin)
     else:
-        build_type = os.environ.get("BUILD_TYPE", "debug")
         binpath = base_dir / "target" / build_type
     log.info(f"neon_binpath is {binpath}")
 
@@ -123,7 +122,7 @@ def neon_binpath(base_dir: Path) -> Iterator[Path]:
     yield binpath
 
 
-@pytest.fixture(scope="session")
+@pytest.fixture(scope="function")
 def pg_distrib_dir(base_dir: Path) -> Iterator[Path]:
     if env_postgres_bin := os.environ.get("POSTGRES_DISTRIB_DIR"):
         distrib_dir = Path(env_postgres_bin).resolve()
@@ -147,7 +146,7 @@ def top_output_dir(base_dir: Path) -> Iterator[Path]:
     yield output_dir
 
 
-@pytest.fixture(scope="session")
+@pytest.fixture(scope="function")
 def versioned_pg_distrib_dir(pg_distrib_dir: Path, pg_version: PgVersion) -> Iterator[Path]:
     versioned_dir = pg_distrib_dir / pg_version.v_prefixed
 
@@ -174,7 +173,23 @@ def shareable_scope(fixture_name: str, config: Config) -> Literal["session", "fu
     def myfixture(...)
        ...
     """
-    return "function" if os.environ.get("TEST_SHARED_FIXTURES") is None else "session"
+    scope: Literal["session", "function"]
+
+    if os.environ.get("TEST_SHARED_FIXTURES") is None:
+        # Create the environment in the per-test output directory
+        scope = "function"
+    elif (
+        os.environ.get("BUILD_TYPE") is not None
+        and os.environ.get("DEFAULT_PG_VERSION") is not None
+    ):
+        scope = "session"
+    else:
+        pytest.fail(
+            "Shared environment(TEST_SHARED_FIXTURES) requires BUILD_TYPE and DEFAULT_PG_VERSION to be set",
+            pytrace=False,
+        )
+
+    return scope
 
 
 @pytest.fixture(scope="session")
@@ -514,6 +529,16 @@ def available_remote_storages() -> List[RemoteStorageKind]:
     return remote_storages
 
 
+def available_s3_storages() -> List[RemoteStorageKind]:
+    remote_storages = [RemoteStorageKind.MOCK_S3]
+    if os.getenv("ENABLE_REAL_S3_REMOTE_STORAGE") is not None:
+        remote_storages.append(RemoteStorageKind.REAL_S3)
+        log.info("Enabling real s3 storage for tests")
+    else:
+        log.info("Using mock implementations to test remote storage")
+    return remote_storages
+
+
 @dataclass
 class LocalFsStorage:
     root: Path
@@ -534,6 +559,16 @@ class S3Storage:
             "AWS_SECRET_ACCESS_KEY": self.secret_key,
         }
 
+    def to_string(self) -> str:
+        return json.dumps(
+            {
+                "bucket": self.bucket_name,
+                "region": self.bucket_region,
+                "endpoint": self.endpoint,
+                "prefix": self.prefix_in_bucket,
+            }
+        )
+
 
 RemoteStorage = Union[LocalFsStorage, S3Storage]
 
@@ -600,10 +635,12 @@ class NeonEnvBuilder:
         self.rust_log_override = rust_log_override
         self.port_distributor = port_distributor
         self.remote_storage = remote_storage
+        self.ext_remote_storage: Optional[S3Storage] = None
+        self.remote_storage_client: Optional[Any] = None
         self.remote_storage_users = remote_storage_users
         self.broker = broker
         self.run_id = run_id
-        self.mock_s3_server = mock_s3_server
+        self.mock_s3_server: MockS3Server = mock_s3_server
         self.pageserver_config_override = pageserver_config_override
         self.num_safekeepers = num_safekeepers
         self.safekeepers_id_start = safekeepers_id_start
@@ -651,15 +688,24 @@ class NeonEnvBuilder:
         remote_storage_kind: RemoteStorageKind,
         test_name: str,
         force_enable: bool = True,
+        enable_remote_extensions: bool = False,
     ):
         if remote_storage_kind == RemoteStorageKind.NOOP:
             return
         elif remote_storage_kind == RemoteStorageKind.LOCAL_FS:
             self.enable_local_fs_remote_storage(force_enable=force_enable)
         elif remote_storage_kind == RemoteStorageKind.MOCK_S3:
-            self.enable_mock_s3_remote_storage(bucket_name=test_name, force_enable=force_enable)
+            self.enable_mock_s3_remote_storage(
+                bucket_name=test_name,
+                force_enable=force_enable,
+                enable_remote_extensions=enable_remote_extensions,
+            )
         elif remote_storage_kind == RemoteStorageKind.REAL_S3:
-            self.enable_real_s3_remote_storage(test_name=test_name, force_enable=force_enable)
+            self.enable_real_s3_remote_storage(
+                test_name=test_name,
+                force_enable=force_enable,
+                enable_remote_extensions=enable_remote_extensions,
+            )
         else:
             raise RuntimeError(f"Unknown storage type: {remote_storage_kind}")
 
@@ -673,11 +719,15 @@ class NeonEnvBuilder:
         assert force_enable or self.remote_storage is None, "remote storage is enabled already"
         self.remote_storage = LocalFsStorage(Path(self.repo_dir / "local_fs_remote_storage"))
 
-    def enable_mock_s3_remote_storage(self, bucket_name: str, force_enable: bool = True):
+    def enable_mock_s3_remote_storage(
+        self, bucket_name: str, force_enable: bool = True, enable_remote_extensions: bool = False
+    ):
         """
         Sets up the pageserver to use the S3 mock server, creates the bucket, if it's not present already.
         Starts up the mock server, if that does not run yet.
         Errors, if the pageserver has some remote storage configuration already, unless `force_enable` is not set to `True`.
+
+        Also creates the bucket for extensions, self.ext_remote_storage bucket
         """
         assert force_enable or self.remote_storage is None, "remote storage is enabled already"
         mock_endpoint = self.mock_s3_server.endpoint()
@@ -698,9 +748,22 @@ class NeonEnvBuilder:
             bucket_region=mock_region,
             access_key=self.mock_s3_server.access_key(),
             secret_key=self.mock_s3_server.secret_key(),
+            prefix_in_bucket="pageserver",
         )
 
-    def enable_real_s3_remote_storage(self, test_name: str, force_enable: bool = True):
+        if enable_remote_extensions:
+            self.ext_remote_storage = S3Storage(
+                bucket_name=bucket_name,
+                endpoint=mock_endpoint,
+                bucket_region=mock_region,
+                access_key=self.mock_s3_server.access_key(),
+                secret_key=self.mock_s3_server.secret_key(),
+                prefix_in_bucket="ext",
+            )
+
+    def enable_real_s3_remote_storage(
+        self, test_name: str, force_enable: bool = True, enable_remote_extensions: bool = False
+    ):
         """
         Sets up configuration to use real s3 endpoint without mock server
         """
@@ -740,6 +803,15 @@ class NeonEnvBuilder:
             prefix_in_bucket=self.remote_storage_prefix,
         )
 
+        if enable_remote_extensions:
+            self.ext_remote_storage = S3Storage(
+                bucket_name="neon-dev-extensions",
+                bucket_region="eu-central-1",
+                access_key=access_key,
+                secret_key=secret_key,
+                prefix_in_bucket="5555",
+            )
+
     def cleanup_local_storage(self):
         if self.preserve_database_files:
             return
@@ -773,6 +845,7 @@ class NeonEnvBuilder:
         # `self.remote_storage_prefix` is coupled with `S3Storage` storage type,
         # so this line effectively a no-op
         assert isinstance(self.remote_storage, S3Storage)
+        assert self.remote_storage_client is not None
 
         if self.keep_remote_storage_contents:
             log.info("keep_remote_storage_contents skipping remote storage cleanup")
@@ -902,6 +975,8 @@ class NeonEnv:
         self.neon_binpath = config.neon_binpath
         self.pg_distrib_dir = config.pg_distrib_dir
         self.endpoint_counter = 0
+        self.remote_storage_client = config.remote_storage_client
+        self.ext_remote_storage = config.ext_remote_storage
 
         # generate initial tenant ID here instead of letting 'neon init' generate it,
         # so that we don't need to dig it out of the config file afterwards.
@@ -1488,6 +1563,7 @@ class NeonCli(AbstractNeonCli):
         safekeepers: Optional[List[int]] = None,
         tenant_id: Optional[TenantId] = None,
         lsn: Optional[Lsn] = None,
+        remote_ext_config: Optional[str] = None,
     ) -> "subprocess.CompletedProcess[str]":
         args = [
             "endpoint",
@@ -1497,6 +1573,8 @@ class NeonCli(AbstractNeonCli):
             "--pg-version",
             self.env.pg_version,
         ]
+        if remote_ext_config is not None:
+            args.extend(["--remote-ext-config", remote_ext_config])
         if lsn is not None:
             args.append(f"--lsn={lsn}")
         args.extend(["--pg-port", str(pg_port)])
@@ -2358,7 +2436,7 @@ class Endpoint(PgProtocol):
 
         return self
 
-    def start(self) -> "Endpoint":
+    def start(self, remote_ext_config: Optional[str] = None) -> "Endpoint":
         """
         Start the Postgres instance.
         Returns self.
@@ -2374,6 +2452,7 @@ class Endpoint(PgProtocol):
             http_port=self.http_port,
             tenant_id=self.tenant_id,
             safekeepers=self.active_safekeepers,
+            remote_ext_config=remote_ext_config,
         )
         self.running = True
 
@@ -2463,6 +2542,7 @@ class Endpoint(PgProtocol):
         hot_standby: bool = False,
         lsn: Optional[Lsn] = None,
         config_lines: Optional[List[str]] = None,
+        remote_ext_config: Optional[str] = None,
     ) -> "Endpoint":
         """
         Create an endpoint, apply config, and start Postgres.
@@ -2477,7 +2557,7 @@ class Endpoint(PgProtocol):
             config_lines=config_lines,
             hot_standby=hot_standby,
             lsn=lsn,
-        ).start()
+        ).start(remote_ext_config=remote_ext_config)
 
         log.info(f"Postgres startup took {time.time() - started_at} seconds")
 
@@ -2511,6 +2591,7 @@ class EndpointFactory:
         lsn: Optional[Lsn] = None,
         hot_standby: bool = False,
         config_lines: Optional[List[str]] = None,
+        remote_ext_config: Optional[str] = None,
     ) -> Endpoint:
         ep = Endpoint(
             self.env,
@@ -2527,6 +2608,7 @@ class EndpointFactory:
             hot_standby=hot_standby,
             config_lines=config_lines,
             lsn=lsn,
+            remote_ext_config=remote_ext_config,
         )
 
     def create(

test_runner/fixtures/pageserver/http.py

@@ -21,6 +21,18 @@ class PageserverApiException(Exception):
         self.status_code = status_code
 
 
+class TimelineCreate406(PageserverApiException):
+    def __init__(self, res: requests.Response):
+        assert res.status_code == 406
+        super().__init__(res.json()["msg"], res.status_code)
+
+
+class TimelineCreate409(PageserverApiException):
+    def __init__(self, res: requests.Response):
+        assert res.status_code == 409
+        super().__init__("", res.status_code)
+
+
 @dataclass
 class InMemoryLayerInfo:
     kind: str
@@ -309,9 +321,12 @@ class PageserverHttpClient(requests.Session):
         res = self.post(
             f"http://localhost:{self.port}/v1/tenant/{tenant_id}/timeline", json=body, **kwargs
         )
-        self.verbose_error(res)
         if res.status_code == 409:
-            raise Exception(f"could not create timeline: already exists for id {new_timeline_id}")
+            raise TimelineCreate409(res)
+        if res.status_code == 406:
+            raise TimelineCreate406(res)
+
+        self.verbose_error(res)
 
         res_json = res.json()
         assert isinstance(res_json, dict)

test_runner/fixtures/parametrize.py

@@ -0,0 +1,50 @@
+import os
+from typing import Optional
+
+import pytest
+from _pytest.fixtures import FixtureRequest
+from _pytest.python import Metafunc
+
+from fixtures.pg_version import PgVersion
+
+"""
+Dynamically parametrize tests by Postgres version and build type (debug/release/remote)
+"""
+
+
+@pytest.fixture(scope="function", autouse=True)
+def pg_version(request: FixtureRequest) -> Optional[PgVersion]:
+    # Do not parametrize performance tests yet, we need to prepare grafana charts first
+    if "test_runner/performance" in str(request.node.path):
+        v = os.environ.get("DEFAULT_PG_VERSION")
+        return PgVersion(v)
+
+    return None
+
+
+@pytest.fixture(scope="function", autouse=True)
+def build_type(request: FixtureRequest) -> Optional[str]:
+    # Do not parametrize performance tests yet, we need to prepare grafana charts first
+    if "test_runner/performance" in str(request.node.path):
+        return os.environ.get("BUILD_TYPE", "").lower()
+
+    return None
+
+
+def pytest_generate_tests(metafunc: Metafunc):
+    # Do not parametrize performance tests yet, we need to prepare grafana charts first
+    if "test_runner/performance" in metafunc.definition._nodeid:
+        return
+
+    if (v := os.environ.get("DEFAULT_PG_VERSION")) is None:
+        pg_versions = [version for version in PgVersion if version != PgVersion.NOT_SET]
+    else:
+        pg_versions = [PgVersion(v)]
+
+    if (bt := os.environ.get("BUILD_TYPE")) is None:
+        build_types = ["debug", "release"]
+    else:
+        build_types = [bt.lower()]
+
+    metafunc.parametrize("build_type", build_types)
+    metafunc.parametrize("pg_version", pg_versions, ids=map(lambda v: f"pg{v}", pg_versions))

test_runner/fixtures/pg_version.py

@@ -1,12 +1,10 @@
 import enum
 import os
-from typing import Iterator, Optional
+from typing import Optional
 
 import pytest
+from _pytest.config import Config
 from _pytest.config.argparsing import Parser
-from pytest import FixtureRequest
-
-from fixtures.log_helper import log
 
 """
 This fixture is used to determine which version of Postgres to use for tests.
@@ -75,18 +73,10 @@ def pytest_addoption(parser: Parser):
         "--pg-version",
         action="store",
         type=PgVersion,
-        help="Postgres version to use for tests",
+        help="DEPRECATED: Postgres version to use for tests",
     )
 
 
-@pytest.fixture(scope="session")
+def pytest_configure(config: Config):
-def pg_version(request: FixtureRequest) -> Iterator[PgVersion]:
+    if config.getoption("--pg-version"):
-    if v := request.config.getoption("--pg-version"):
+        raise Exception("--pg-version is deprecated, use DEFAULT_PG_VERSION env var instead")
-        version, source = v, "from --pg-version command-line argument"
-    elif v := os.environ.get("DEFAULT_PG_VERSION"):
-        version, source = PgVersion(v), "from DEFAULT_PG_VERSION environment variable"
-    else:
-        version, source = DEFAULT_VERSION, "default version"
-
-    log.info(f"pg_version is {version} ({source})")
-    yield version

test_runner/fixtures/types.py

@@ -89,6 +89,9 @@ class TenantId(Id):
     def __repr__(self) -> str:
         return f'`TenantId("{self.id.hex()}")'
 
+    def __str__(self) -> str:
+        return self.id.hex()
+
 
 class TimelineId(Id):
     def __repr__(self) -> str:

test_runner/regress/data/extension_test/v14/ext_index.json

@@ -0,0 +1,14 @@
+{
+  "enabled_extensions": {
+    "123454321": [
+      "anon"
+    ],
+    "public": [
+      "embedding"
+    ]
+  },
+  "control_data": {
+    "embedding": "comment = 'hnsw index' \ndefault_version = '0.1.0' \nmodule_pathname = '$libdir/embedding' \nrelocatable = true \ntrusted = true",
+    "anon": "# PostgreSQL Anonymizer (anon) extension \ncomment = 'Data anonymization tools' \ndefault_version = '1.1.0' \ndirectory='extension/anon' \nrelocatable = false \nrequires = 'pgcrypto' \nsuperuser = false \nmodule_pathname = '$libdir/anon' \ntrusted = true \n"
+  }
+}

test_runner/regress/data/extension_test/v15/ext_index.json

@@ -0,0 +1,14 @@
+{
+  "enabled_extensions": {
+    "123454321": [
+      "anon"
+    ],
+    "public": [
+      "embedding"
+    ]
+  },
+  "control_data": {
+    "embedding": "comment = 'hnsw index' \ndefault_version = '0.1.0' \nmodule_pathname = '$libdir/embedding' \nrelocatable = true \ntrusted = true",
+    "anon": "# PostgreSQL Anonymizer (anon) extension \ncomment = 'Data anonymization tools' \ndefault_version = '1.1.0' \ndirectory='extension/anon' \nrelocatable = false \nrequires = 'pgcrypto' \nsuperuser = false \nmodule_pathname = '$libdir/anon' \ntrusted = true \n"
+  }
+}

test_runner/regress/test_branch_and_gc.py

@@ -4,7 +4,8 @@ import time
 import pytest
 from fixtures.log_helper import log
 from fixtures.neon_fixtures import NeonEnv
-from fixtures.types import Lsn
+from fixtures.pageserver.http import TimelineCreate406
+from fixtures.types import Lsn, TimelineId
 from fixtures.utils import query_scalar
 
 
@@ -173,5 +174,12 @@ def test_branch_creation_before_gc(neon_simple_env: NeonEnv):
     # The starting LSN is invalid as the corresponding record is scheduled to be removed by in-queue GC.
     with pytest.raises(Exception, match="invalid branch start lsn: .*"):
         env.neon_cli.create_branch("b1", "b0", tenant_id=tenant, ancestor_start_lsn=lsn)
+    # retry the same with the HTTP API, so that we can inspect the status code
+    with pytest.raises(TimelineCreate406):
+        new_timeline_id = TimelineId.generate()
+        log.info(
+            f"Expecting failure for branch behind gc'ing LSN, new_timeline_id={new_timeline_id}"
+        )
+        pageserver_http_client.timeline_create(env.pg_version, tenant, new_timeline_id, b0, lsn)
 
     thread.join()

test_runner/regress/test_branch_behind.py

@@ -1,6 +1,7 @@
 import pytest
 from fixtures.log_helper import log
 from fixtures.neon_fixtures import NeonEnvBuilder
+from fixtures.pageserver.http import TimelineCreate406
 from fixtures.types import Lsn, TimelineId
 from fixtures.utils import print_gc_result, query_scalar
 
@@ -17,7 +18,7 @@ def test_branch_behind(neon_env_builder: NeonEnvBuilder):
     env.pageserver.allowed_errors.append(".*invalid start lsn .* for ancestor timeline.*")
 
     # Branch at the point where only 100 rows were inserted
-    env.neon_cli.create_branch("test_branch_behind")
+    branch_behind_timeline_id = env.neon_cli.create_branch("test_branch_behind")
     endpoint_main = env.endpoints.create_start("test_branch_behind")
     log.info("postgres is running on 'test_branch_behind' branch")
 
@@ -89,6 +90,7 @@ def test_branch_behind(neon_env_builder: NeonEnvBuilder):
     assert query_scalar(main_cur, "SELECT count(*) FROM foo") == 400100
 
     # Check bad lsn's for branching
+    pageserver_http = env.pageserver.http_client()
 
     # branch at segment boundary
     env.neon_cli.create_branch(
@@ -97,27 +99,52 @@ def test_branch_behind(neon_env_builder: NeonEnvBuilder):
     endpoint = env.endpoints.create_start("test_branch_segment_boundary")
     assert endpoint.safe_psql("SELECT 1")[0][0] == 1
 
-    # branch at pre-initdb lsn
+    # branch at pre-initdb lsn (from main branch)
     with pytest.raises(Exception, match="invalid branch start lsn: .*"):
         env.neon_cli.create_branch("test_branch_preinitdb", ancestor_start_lsn=Lsn("0/42"))
+    # retry the same with the HTTP API, so that we can inspect the status code
+    with pytest.raises(TimelineCreate406):
+        new_timeline_id = TimelineId.generate()
+        log.info(f"Expecting failure for branch pre-initdb LSN, new_timeline_id={new_timeline_id}")
+        pageserver_http.timeline_create(
+            env.pg_version, env.initial_tenant, new_timeline_id, env.initial_timeline, Lsn("0/42")
+        )
 
     # branch at pre-ancestor lsn
     with pytest.raises(Exception, match="less than timeline ancestor lsn"):
         env.neon_cli.create_branch(
             "test_branch_preinitdb", "test_branch_behind", ancestor_start_lsn=Lsn("0/42")
         )
+    # retry the same with the HTTP API, so that we can inspect the status code
+    with pytest.raises(TimelineCreate406):
+        new_timeline_id = TimelineId.generate()
+        log.info(
+            f"Expecting failure for branch pre-ancestor LSN, new_timeline_id={new_timeline_id}"
+        )
+        pageserver_http.timeline_create(
+            env.pg_version,
+            env.initial_tenant,
+            new_timeline_id,
+            branch_behind_timeline_id,
+            Lsn("0/42"),
+        )
 
     # check that we cannot create branch based on garbage collected data
-    with env.pageserver.http_client() as pageserver_http:
+    pageserver_http.timeline_checkpoint(env.initial_tenant, timeline)
-        pageserver_http.timeline_checkpoint(env.initial_tenant, timeline)
+    gc_result = pageserver_http.timeline_gc(env.initial_tenant, timeline, 0)
-        gc_result = pageserver_http.timeline_gc(env.initial_tenant, timeline, 0)
+    print_gc_result(gc_result)
-        print_gc_result(gc_result)
-
     with pytest.raises(Exception, match="invalid branch start lsn: .*"):
         # this gced_lsn is pretty random, so if gc is disabled this woudln't fail
         env.neon_cli.create_branch(
             "test_branch_create_fail", "test_branch_behind", ancestor_start_lsn=gced_lsn
         )
+    # retry the same with the HTTP API, so that we can inspect the status code
+    with pytest.raises(TimelineCreate406):
+        new_timeline_id = TimelineId.generate()
+        log.info(f"Expecting failure for branch behind gc'd LSN, new_timeline_id={new_timeline_id}")
+        pageserver_http.timeline_create(
+            env.pg_version, env.initial_tenant, new_timeline_id, branch_behind_timeline_id, gced_lsn
+        )
 
     # check that after gc everything is still there
     assert query_scalar(hundred_cur, "SELECT count(*) FROM foo") == 100

test_runner/regress/test_compatibility.py

@@ -2,6 +2,7 @@ import copy
 import os
 import shutil
 import subprocess
+import tempfile
 from pathlib import Path
 from typing import Any, Optional
 
@@ -448,7 +449,7 @@ def dump_differs(first: Path, second: Path, output: Path) -> bool:
     """
 
     with output.open("w") as stdout:
-        rv = subprocess.run(
+        res = subprocess.run(
             [
                 "diff",
                 "--unified",  # Make diff output more readable
@@ -460,4 +461,53 @@ def dump_differs(first: Path, second: Path, output: Path) -> bool:
             stdout=stdout,
         )
 
-    return rv.returncode != 0
+    differs = res.returncode != 0
+
+    # TODO: Remove after https://github.com/neondatabase/neon/pull/4425 is merged, and a couple of releases are made
+    if differs:
+        with tempfile.NamedTemporaryFile(mode="w") as tmp:
+            tmp.write(PR4425_ALLOWED_DIFF)
+            tmp.flush()
+
+            allowed = subprocess.run(
+                [
+                    "diff",
+                    "--unified",  # Make diff output more readable
+                    r"--ignore-matching-lines=^---",  # Ignore diff headers
+                    r"--ignore-matching-lines=^\+\+\+",  # Ignore diff headers
+                    "--ignore-matching-lines=^@@",  # Ignore diff blocks location
+                    "--ignore-matching-lines=^ *$",  # Ignore lines with only spaces
+                    "--ignore-matching-lines=^ --.*",  # Ignore the " --" lines for compatibility with PG14
+                    "--ignore-blank-lines",
+                    str(output),
+                    str(tmp.name),
+                ],
+            )
+
+            differs = allowed.returncode != 0
+
+    return differs
+
+
+PR4425_ALLOWED_DIFF = """
+--- /tmp/test_output/test_backward_compatibility[release-pg15]/compatibility_snapshot/dump.sql 2023-06-08 18:12:45.000000000 +0000
++++ /tmp/test_output/test_backward_compatibility[release-pg15]/dump.sql        2023-06-13 07:25:35.211733653 +0000
+@@ -13,12 +13,20 @@
+
+ CREATE ROLE cloud_admin;
+ ALTER ROLE cloud_admin WITH SUPERUSER INHERIT CREATEROLE CREATEDB LOGIN REPLICATION BYPASSRLS;
++CREATE ROLE neon_superuser;
++ALTER ROLE neon_superuser WITH NOSUPERUSER INHERIT CREATEROLE CREATEDB NOLOGIN NOREPLICATION NOBYPASSRLS;
+
+ --
+ -- User Configurations
+ --
+
+
++--
++-- Role memberships
++--
++
++GRANT pg_read_all_data TO neon_superuser GRANTED BY cloud_admin;
++GRANT pg_write_all_data TO neon_superuser GRANTED BY cloud_admin;
+"""

test_runner/regress/test_ddl_forwarding.py

@@ -33,6 +33,7 @@ def handle_role(dbs, roles, operation):
                     dbs[db] = operation["name"]
         if "password" in operation:
             roles[operation["name"]] = operation["password"]
+            assert "encrypted_password" in operation
     elif operation["op"] == "del":
         if "old_name" in operation:
             roles.pop(operation["old_name"])

test_runner/regress/test_disk_usage_eviction.py

@@ -1,7 +1,6 @@
 import shutil
 import time
 from dataclasses import dataclass
-from pathlib import Path
 from typing import Dict, Tuple
 
 import pytest
@@ -428,14 +427,14 @@ def poor_mans_du(
     largest_layer = 0
     smallest_layer = None
     for tenant_id, timeline_id in timelines:
-        dir = Path(env.repo_dir) / "tenants" / str(tenant_id) / "timelines" / str(timeline_id)
+        timeline_dir = env.timeline_dir(tenant_id, timeline_id)
-        assert dir.exists(), f"timeline dir does not exist: {dir}"
+        assert timeline_dir.exists(), f"timeline dir does not exist: {timeline_dir}"
-        sum = 0
+        total = 0
-        for file in dir.iterdir():
+        for file in timeline_dir.iterdir():
             if "__" not in file.name:
                 continue
             size = file.stat().st_size
-            sum += size
+            total += size
             largest_layer = max(largest_layer, size)
             if smallest_layer:
                 smallest_layer = min(smallest_layer, size)
@@ -443,8 +442,8 @@ def poor_mans_du(
                 smallest_layer = size
             log.info(f"{tenant_id}/{timeline_id} => {file.name} {size}")
 
-        log.info(f"{tenant_id}/{timeline_id}: sum {sum}")
+        log.info(f"{tenant_id}/{timeline_id}: sum {total}")
-        total_on_disk += sum
+        total_on_disk += total
 
     assert smallest_layer is not None or total_on_disk == 0 and largest_layer == 0
     return (total_on_disk, largest_layer, smallest_layer or 0)

test_runner/regress/test_download_extensions.py

@@ -0,0 +1,122 @@
+import os
+import shutil
+from contextlib import closing
+
+import pytest
+from fixtures.log_helper import log
+from fixtures.neon_fixtures import (
+    NeonEnvBuilder,
+    RemoteStorageKind,
+    available_s3_storages,
+)
+from fixtures.pg_version import PgVersion
+
+# Generate mock extension files and upload them to the mock bucket.
+#
+# NOTE: You must have appropriate AWS credentials to run REAL_S3 test.
+# It may also be necessary to set the following environment variables for MOCK_S3 test:
+#   export AWS_ACCESS_KEY_ID='test' #   export AWS_SECRET_ACCESS_KEY='test'
+#   export AWS_SECURITY_TOKEN='test' #   export AWS_SESSION_TOKEN='test'
+#   export AWS_DEFAULT_REGION='us-east-1'
+
+
+@pytest.mark.parametrize("remote_storage_kind", available_s3_storages())
+def test_remote_extensions(
+    neon_env_builder: NeonEnvBuilder,
+    remote_storage_kind: RemoteStorageKind,
+    pg_version: PgVersion,
+):
+    neon_env_builder.enable_remote_storage(
+        remote_storage_kind=remote_storage_kind,
+        test_name="test_remote_extensions",
+        enable_remote_extensions=True,
+    )
+    neon_env_builder.num_safekeepers = 3
+    env = neon_env_builder.init_start()
+    tenant_id, _ = env.neon_cli.create_tenant()
+    env.neon_cli.create_timeline("test_remote_extensions", tenant_id=tenant_id)
+
+    # For MOCK_S3 we upload test files.
+    # For REAL_S3 we use the files already in the bucket
+    if remote_storage_kind == RemoteStorageKind.MOCK_S3:
+        log.info("Uploading test files to mock bucket")
+
+        def upload_test_file(from_path, to_path):
+            assert env.ext_remote_storage is not None  # satisfy mypy
+            assert env.remote_storage_client is not None  # satisfy mypy
+            with open(
+                f"test_runner/regress/data/extension_test/v{pg_version}/{from_path}", "rb"
+            ) as f:
+                env.remote_storage_client.upload_fileobj(
+                    f,
+                    env.ext_remote_storage.bucket_name,
+                    f"ext/v{pg_version}/{to_path}",
+                )
+
+        upload_test_file("ext_index.json", "ext_index.json")
+        upload_test_file("anon.tar.gz", "extensions/anon.tar.gz")
+        upload_test_file("embedding.tar.gz", "extensions/embedding.tar.gz")
+
+    assert env.ext_remote_storage is not None  # satisfy mypy
+    assert env.remote_storage_client is not None  # satisfy mypy
+    try:
+        # Start a compute node and check that it can download the extensions
+        # and use them to CREATE EXTENSION and LOAD
+        endpoint = env.endpoints.create_start(
+            "test_remote_extensions",
+            tenant_id=tenant_id,
+            remote_ext_config=env.ext_remote_storage.to_string(),
+            # config_lines=["log_min_messages=debug3"],
+        )
+        with closing(endpoint.connect()) as conn:
+            with conn.cursor() as cur:
+                # Check that appropriate control files were downloaded
+                cur.execute("SELECT * FROM pg_available_extensions")
+                all_extensions = [x[0] for x in cur.fetchall()]
+                log.info(all_extensions)
+                assert "anon" in all_extensions
+                assert "embedding" in all_extensions
+                # TODO: check that we cant't download custom extensions for other tenant ids
+
+                # check that we can download public extension
+                cur.execute("CREATE EXTENSION embedding")
+                cur.execute("SELECT extname FROM pg_extension")
+                assert "embedding" in [x[0] for x in cur.fetchall()]
+
+                # check that we can download private extension
+                try:
+                    cur.execute("CREATE EXTENSION anon")
+                except Exception as err:
+                    log.info("error creating anon extension")
+                    assert "pgcrypto" in str(err), "unexpected error creating anon extension"
+
+                # TODO: try to load libraries as well
+
+    finally:
+        # Cleaning up downloaded files is important for local tests
+        # or else one test could reuse the files from another test or another test run
+        cleanup_files = [
+            "lib/postgresql/anon.so",
+            "lib/postgresql/embedding.so",
+            "share/postgresql/extension/anon.control",
+            "share/postgresql/extension/embedding--0.1.0.sql",
+            "share/postgresql/extension/embedding.control",
+        ]
+        cleanup_files = [f"pg_install/v{pg_version}/" + x for x in cleanup_files]
+        cleanup_folders = [
+            "extensions",
+            f"pg_install/v{pg_version}/share/postgresql/extension/anon",
+            f"pg_install/v{pg_version}/extensions",
+        ]
+        for file in cleanup_files:
+            try:
+                os.remove(file)
+                log.info(f"removed file {file}")
+            except Exception as err:
+                log.info(f"error removing file {file}: {err}")
+        for folder in cleanup_folders:
+            try:
+                shutil.rmtree(folder)
+                log.info(f"removed folder {folder}")
+            except Exception as err:
+                log.info(f"error removing file {file}: {err}")

test_runner/regress/test_hot_standby.py

@@ -1,3 +1,5 @@
+import time
+
 import pytest
 from fixtures.neon_fixtures import NeonEnv
 
@@ -10,9 +12,10 @@ def test_hot_standby(neon_simple_env: NeonEnv):
         branch_name="main",
         endpoint_id="primary",
     ) as primary:
+        time.sleep(1)
         with env.endpoints.new_replica_start(origin=primary, endpoint_id="secondary") as secondary:
             primary_lsn = None
-            cought_up = False
+            caught_up = False
             queries = [
                 "SHOW neon.timeline_id",
                 "SHOW neon.tenant_id",
@@ -56,7 +59,7 @@ def test_hot_standby(neon_simple_env: NeonEnv):
                     res = s_cur.fetchone()
                     assert res is not None
 
-                while not cought_up:
+                while not caught_up:
                     with s_con.cursor() as secondary_cursor:
                         secondary_cursor.execute("SELECT pg_last_wal_replay_lsn()")
                         res = secondary_cursor.fetchone()
@@ -66,7 +69,7 @@ def test_hot_standby(neon_simple_env: NeonEnv):
                         # due to e.g. autovacuum, but that shouldn't impact the content
                         # of the tables, so we check whether we've replayed up to at
                         # least after the commit of the `test` table.
-                        cought_up = secondary_lsn >= primary_lsn
+                        caught_up = secondary_lsn >= primary_lsn
 
                 # Explicit commit to flush any transient transaction-level state.
                 s_con.commit()

test_runner/regress/test_layer_eviction.py

@@ -58,7 +58,7 @@ def test_basic_eviction(
     for sk in env.safekeepers:
         sk.stop()
 
-    timeline_path = env.repo_dir / "tenants" / str(tenant_id) / "timelines" / str(timeline_id)
+    timeline_path = env.timeline_dir(tenant_id, timeline_id)
     initial_local_layers = sorted(
         list(filter(lambda path: path.name != "metadata", timeline_path.glob("*")))
     )

test_runner/regress/test_ondemand_download.py

@@ -713,9 +713,7 @@ def test_ondemand_download_failure_to_replace(
         # error message is not useful
         pageserver_http.timeline_detail(tenant_id, timeline_id, True, timeout=2)
 
-    actual_message = (
+    actual_message = ".* ERROR .*layermap-replace-notfound"
-        ".* ERROR .*replacing downloaded layer into layermap failed because layer was not found"
-    )
     assert env.pageserver.log_contains(actual_message) is not None
     env.pageserver.allowed_errors.append(actual_message)
 

test_runner/regress/test_remote_storage.py

@@ -535,7 +535,7 @@ def test_timeline_deletion_with_files_stuck_in_upload_queue(
             "pitr_interval": "0s",
         }
     )
-    timeline_path = env.repo_dir / "tenants" / str(tenant_id) / "timelines" / str(timeline_id)
+    timeline_path = env.timeline_dir(tenant_id, timeline_id)
 
     client = env.pageserver.http_client()
 

test_runner/regress/test_tenant_detach.py

@@ -632,14 +632,14 @@ def test_ignored_tenant_download_missing_layers(
 
     # ignore the tenant and remove its layers
     pageserver_http.tenant_ignore(tenant_id)
-    tenant_timeline_dir = env.repo_dir / "tenants" / str(tenant_id) / "timelines" / str(timeline_id)
+    timeline_dir = env.timeline_dir(tenant_id, timeline_id)
     layers_removed = False
-    for dir_entry in tenant_timeline_dir.iterdir():
+    for dir_entry in timeline_dir.iterdir():
         if dir_entry.name.startswith("00000"):
             # Looks like a layer file. Remove it
             dir_entry.unlink()
             layers_removed = True
-    assert layers_removed, f"Found no layers for tenant {tenant_timeline_dir}"
+    assert layers_removed, f"Found no layers for tenant {timeline_dir}"
 
     # now, load it from the local files and expect it to work due to remote storage restoration
     pageserver_http.tenant_load(tenant_id=tenant_id)
@@ -688,14 +688,14 @@ def test_ignored_tenant_stays_broken_without_metadata(
 
     # ignore the tenant and remove its metadata
     pageserver_http.tenant_ignore(tenant_id)
-    tenant_timeline_dir = env.repo_dir / "tenants" / str(tenant_id) / "timelines" / str(timeline_id)
+    timeline_dir = env.timeline_dir(tenant_id, timeline_id)
     metadata_removed = False
-    for dir_entry in tenant_timeline_dir.iterdir():
+    for dir_entry in timeline_dir.iterdir():
         if dir_entry.name == "metadata":
             # Looks like a layer file. Remove it
             dir_entry.unlink()
             metadata_removed = True
-    assert metadata_removed, f"Failed to find metadata file in {tenant_timeline_dir}"
+    assert metadata_removed, f"Failed to find metadata file in {timeline_dir}"
 
     env.pageserver.allowed_errors.append(
         f".*{tenant_id}.*: load failed.*: failed to load metadata.*"

test_runner/regress/test_tenant_relocation.py

@@ -214,9 +214,7 @@ def switch_pg_to_new_pageserver(
 
     endpoint.start()
 
-    timeline_to_detach_local_path = (
+    timeline_to_detach_local_path = env.timeline_dir(tenant_id, timeline_id)
-        env.repo_dir / "tenants" / str(tenant_id) / "timelines" / str(timeline_id)
-    )
     files_before_detach = os.listdir(timeline_to_detach_local_path)
     assert (
         "metadata" in files_before_detach
@@ -419,8 +417,6 @@ def test_tenant_relocation(
             new_pageserver_http.tenant_attach(tenant_id)
 
             # wait for tenant to finish attaching
-            tenant_status = new_pageserver_http.tenant_status(tenant_id=tenant_id)
-            assert tenant_status["state"]["slug"] in ["Attaching", "Active"]
             wait_until(
                 number_of_iterations=10,
                 interval=1,

test_runner/regress/test_tenant_size.py

@@ -16,6 +16,7 @@ from fixtures.pg_version import PgVersion, xfail_on_postgres
 from fixtures.types import Lsn, TenantId, TimelineId
 
 
+@pytest.mark.xfail
 def test_empty_tenant_size(neon_simple_env: NeonEnv, test_output_dir: Path):
     env = neon_simple_env
     (tenant_id, _) = env.neon_cli.create_tenant()
@@ -44,12 +45,16 @@ def test_empty_tenant_size(neon_simple_env: NeonEnv, test_output_dir: Path):
         # we've disabled the autovacuum and checkpoint
         # so background processes should not change the size.
         # If this test will flake we should probably loosen the check
-        assert size == initial_size, "starting idle compute should not change the tenant size"
+        assert (
+            size == initial_size
+        ), f"starting idle compute should not change the tenant size (Currently {size}, expected {initial_size})"
 
     # the size should be the same, until we increase the size over the
     # gc_horizon
     size, inputs = http_client.tenant_size_and_modelinputs(tenant_id)
-    assert size == initial_size, "tenant_size should not be affected by shutdown of compute"
+    assert (
+        size == initial_size
+    ), f"tenant_size should not be affected by shutdown of compute (Currently {size}, expected {initial_size})"
 
     expected_inputs = {
         "segments": [
@@ -318,6 +323,7 @@ def test_only_heads_within_horizon(neon_simple_env: NeonEnv, test_output_dir: Pa
     size_debug_file.write(size_debug)
 
 
+@pytest.mark.xfail
 def test_single_branch_get_tenant_size_grows(
     neon_env_builder: NeonEnvBuilder, test_output_dir: Path, pg_version: PgVersion
 ):
@@ -333,13 +339,13 @@ def test_single_branch_get_tenant_size_grows(
     # inserts is larger than gc_horizon. for example 0x20000 here hid the fact
     # that there next_gc_cutoff could be smaller than initdb_lsn, which will
     # obviously lead to issues when calculating the size.
-    gc_horizon = 0x38000
+    gc_horizon = 0x3BA00
 
     # it's a bit of a hack, but different versions of postgres have different
     # amount of WAL generated for the same amount of data. so we need to
     # adjust the gc_horizon accordingly.
     if pg_version == PgVersion.V14:
-        gc_horizon = 0x40000
+        gc_horizon = 0x4A000
 
     neon_env_builder.pageserver_config_override = f"tenant_config={{compaction_period='0s', gc_period='0s', pitr_interval='0sec', gc_horizon={gc_horizon}}}"
 
@@ -360,11 +366,11 @@ def test_single_branch_get_tenant_size_grows(
         if current_lsn - initdb_lsn >= gc_horizon:
             assert (
                 size >= prev_size
-            ), "tenant_size may grow or not grow, because we only add gc_horizon amount of WAL to initial snapshot size"
+            ), f"tenant_size may grow or not grow, because we only add gc_horizon amount of WAL to initial snapshot size (Currently at: {current_lsn}, Init at: {initdb_lsn})"
         else:
             assert (
                 size > prev_size
-            ), "tenant_size should grow, because we continue to add WAL to initial snapshot size"
+            ), f"tenant_size should grow, because we continue to add WAL to initial snapshot size (Currently at: {current_lsn}, Init at: {initdb_lsn})"
 
     def get_current_consistent_size(
         env: NeonEnv,

test_runner/regress/test_tenants_with_remote_storage.py

@@ -257,7 +257,7 @@ def test_tenant_redownloads_truncated_file_on_startup(
     env.endpoints.stop_all()
     env.pageserver.stop()
 
-    timeline_dir = Path(env.repo_dir) / "tenants" / str(tenant_id) / "timelines" / str(timeline_id)
+    timeline_dir = env.timeline_dir(tenant_id, timeline_id)
     local_layer_truncated = None
     for path in Path.iterdir(timeline_dir):
         if path.name.startswith("00000"):

test_runner/regress/test_timeline_delete.py

@@ -275,6 +275,7 @@ def assert_prefix_empty(neon_env_builder: NeonEnvBuilder, prefix: Optional[str]
     assert isinstance(neon_env_builder.remote_storage, S3Storage)
 
     # Note that this doesnt use pagination, so list is not guaranteed to be exhaustive.
+    assert neon_env_builder.remote_storage_client is not None
     response = neon_env_builder.remote_storage_client.list_objects_v2(
         Bucket=neon_env_builder.remote_storage.bucket_name,
         Prefix=prefix or neon_env_builder.remote_storage.prefix_in_bucket or "",
@@ -463,10 +464,10 @@ def test_concurrent_timeline_delete_stuck_on(
 
         # make the second call and assert behavior
         log.info("second call start")
-        error_msg_re = "timeline deletion is already in progress"
+        error_msg_re = "Timeline deletion is already in progress"
         with pytest.raises(PageserverApiException, match=error_msg_re) as second_call_err:
             ps_http.timeline_delete(env.initial_tenant, child_timeline_id)
-        assert second_call_err.value.status_code == 500
+        assert second_call_err.value.status_code == 409
         env.pageserver.allowed_errors.append(f".*{child_timeline_id}.*{error_msg_re}.*")
         # the second call will try to transition the timeline into Stopping state as well
         env.pageserver.allowed_errors.append(
@@ -518,9 +519,9 @@ def test_delete_timeline_client_hangup(neon_env_builder: NeonEnvBuilder):
         ps_http.timeline_delete(env.initial_tenant, child_timeline_id, timeout=2)
 
     env.pageserver.allowed_errors.append(
-        f".*{child_timeline_id}.*timeline deletion is already in progress.*"
+        f".*{child_timeline_id}.*Timeline deletion is already in progress.*"
     )
-    with pytest.raises(PageserverApiException, match="timeline deletion is already in progress"):
+    with pytest.raises(PageserverApiException, match="Timeline deletion is already in progress"):
         ps_http.timeline_delete(env.initial_tenant, child_timeline_id, timeout=2)
 
     # make sure the timeout was due to the failpoint
@@ -628,7 +629,7 @@ def test_timeline_delete_works_for_remote_smoke(
         )
 
     # for some reason the check above doesnt immediately take effect for the below.
-    # Assume it is mock server incosistency and check twice.
+    # Assume it is mock server inconsistency and check twice.
     wait_until(
         2,
         0.5,

test_runner/regress/test_timeline_size.py

@@ -416,6 +416,7 @@ def test_timeline_physical_size_post_compaction(
     wait_for_last_flush_lsn(env, endpoint, env.initial_tenant, new_timeline_id)
 
     # shutdown safekeepers to prevent new data from coming in
+    endpoint.stop()  # We can't gracefully stop after safekeepers die
     for sk in env.safekeepers:
         sk.stop()
 

test_runner/regress/test_wal_acceptor.py

@@ -1210,6 +1210,10 @@ def test_delete_force(neon_env_builder: NeonEnvBuilder, auth_enabled: bool):
             with conn.cursor() as cur:
                 cur.execute("INSERT INTO t (key) VALUES (1)")
 
+    # Stop all computes gracefully before safekeepers stop responding to them
+    endpoint_1.stop_and_destroy()
+    endpoint_3.stop_and_destroy()
+
     # Remove initial tenant's br1 (active)
     assert sk_http.timeline_delete_force(tenant_id, timeline_id_1)["dir_existed"]
     assert not (sk_data_dir / str(tenant_id) / str(timeline_id_1)).exists()