need additional grants according to Jelte F

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

@@ -1,91 +0,0 @@
-name: Check Codestyle Rust
-
-on:
-  workflow_call:
-    inputs:
-      build-tools-image:
-        description: "build-tools image"
-        required: true
-        type: string
-      archs:
-        description: "Json array of architectures to run on"
-        type: string
-
-
-defaults:
-  run:
-    shell: bash -euxo pipefail {0}
-
-jobs:
-  check-codestyle-rust:
-    strategy:
-      matrix:
-        arch: ${{ fromJson(inputs.archs) }}
-    runs-on: ${{ fromJson(format('["self-hosted", "{0}"]', matrix.arch == 'arm64' && 'small-arm64' || 'small')) }}
-
-    container:
-      image: ${{ inputs.build-tools-image }}
-      credentials:
-        username: ${{ secrets.NEON_DOCKERHUB_USERNAME }}
-        password: ${{ secrets.NEON_DOCKERHUB_PASSWORD }}
-      options: --init
-
-    steps:
-      - name: Checkout
-        uses: actions/checkout@v4
-        with:
-          submodules: true
-
-      - name: Cache cargo deps
-        uses: actions/cache@v4
-        with:
-          path: |
-            ~/.cargo/registry
-            !~/.cargo/registry/src
-            ~/.cargo/git
-            target
-          key: v1-${{ runner.os }}-${{ runner.arch }}-cargo-${{ hashFiles('./Cargo.lock') }}-${{ hashFiles('./rust-toolchain.toml') }}-rust
-
-      # Some of our rust modules use FFI and need those to be checked
-      - name: Get postgres headers
-        run: make postgres-headers -j$(nproc)
-
-      # cargo hack runs the given cargo subcommand (clippy in this case) for all feature combinations.
-      # This will catch compiler & clippy warnings in all feature combinations.
-      # TODO: use cargo hack for build and test as well, but, that's quite expensive.
-      # NB: keep clippy args in sync with ./run_clippy.sh
-      #
-      # The only difference between "clippy --debug" and "clippy --release" is that in --release mode,
-      # #[cfg(debug_assertions)] blocks are not built. It's not worth building everything for second
-      # time just for that, so skip "clippy --release".
-      - run: |
-          CLIPPY_COMMON_ARGS="$( source .neon_clippy_args; echo "$CLIPPY_COMMON_ARGS")"
-          if [ "$CLIPPY_COMMON_ARGS" = "" ]; then
-            echo "No clippy args found in .neon_clippy_args"
-            exit 1
-          fi
-          echo "CLIPPY_COMMON_ARGS=${CLIPPY_COMMON_ARGS}" >> $GITHUB_ENV
-      - name: Run cargo clippy (debug)
-        run: cargo hack --features default --ignore-unknown-features --feature-powerset clippy $CLIPPY_COMMON_ARGS
-
-      - name: Check documentation generation
-        run: cargo doc --workspace --no-deps --document-private-items
-        env:
-          RUSTDOCFLAGS: "-Dwarnings -Arustdoc::private_intra_doc_links"
-
-      # Use `${{ !cancelled() }}` to run quck tests after the longer clippy run
-      - name: Check formatting
-        if: ${{ !cancelled() }}
-        run: cargo fmt --all -- --check
-
-      # https://github.com/facebookincubator/cargo-guppy/tree/bec4e0eb29dcd1faac70b1b5360267fc02bf830e/tools/cargo-hakari#2-keep-the-workspace-hack-up-to-date-in-ci
-      - name: Check rust dependencies
-        if: ${{ !cancelled() }}
-        run: |
-          cargo hakari generate --diff  # workspace-hack Cargo.toml is up-to-date
-          cargo hakari manage-deps --dry-run  # all workspace crates depend on workspace-hack
-
-      # https://github.com/EmbarkStudios/cargo-deny
-      - name: Check rust licenses/bans/advisories/sources
-        if: ${{ !cancelled() }}
-        run: cargo deny check --hide-inclusion-graph

.github/workflows/build_and_test.yml

@@ -164,11 +164,77 @@ jobs:
 
   check-codestyle-rust:
     needs: [ check-permissions, build-build-tools-image ]
-    uses: ./.github/workflows/_check-codestyle-rust.yml
-    with:
-      build-tools-image: ${{ needs.build-build-tools-image.outputs.image }}-bookworm
-      archs: '["x64", "arm64"]'
-    secrets: inherit
+    strategy:
+      matrix:
+        arch: [ x64, arm64 ]
+    runs-on: ${{ fromJson(format('["self-hosted", "{0}"]', matrix.arch == 'arm64' && 'small-arm64' || 'small')) }}
+
+    container:
+      image: ${{ needs.build-build-tools-image.outputs.image }}-bookworm
+      credentials:
+        username: ${{ secrets.NEON_DOCKERHUB_USERNAME }}
+        password: ${{ secrets.NEON_DOCKERHUB_PASSWORD }}
+      options: --init
+
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v4
+        with:
+          submodules: true
+
+      - name: Cache cargo deps
+        uses: actions/cache@v4
+        with:
+          path: |
+            ~/.cargo/registry
+            !~/.cargo/registry/src
+            ~/.cargo/git
+            target
+          key: v1-${{ runner.os }}-${{ runner.arch }}-cargo-${{ hashFiles('./Cargo.lock') }}-${{ hashFiles('./rust-toolchain.toml') }}-rust
+
+      # Some of our rust modules use FFI and need those to be checked
+      - name: Get postgres headers
+        run: make postgres-headers -j$(nproc)
+
+      # cargo hack runs the given cargo subcommand (clippy in this case) for all feature combinations.
+      # This will catch compiler & clippy warnings in all feature combinations.
+      # TODO: use cargo hack for build and test as well, but, that's quite expensive.
+      # NB: keep clippy args in sync with ./run_clippy.sh
+      #
+      # The only difference between "clippy --debug" and "clippy --release" is that in --release mode,
+      # #[cfg(debug_assertions)] blocks are not built. It's not worth building everything for second
+      # time just for that, so skip "clippy --release".
+      - run: |
+          CLIPPY_COMMON_ARGS="$( source .neon_clippy_args; echo "$CLIPPY_COMMON_ARGS")"
+          if [ "$CLIPPY_COMMON_ARGS" = "" ]; then
+            echo "No clippy args found in .neon_clippy_args"
+            exit 1
+          fi
+          echo "CLIPPY_COMMON_ARGS=${CLIPPY_COMMON_ARGS}" >> $GITHUB_ENV
+      - name: Run cargo clippy (debug)
+        run: cargo hack --features default --ignore-unknown-features --feature-powerset clippy $CLIPPY_COMMON_ARGS
+
+      - name: Check documentation generation
+        run: cargo doc --workspace --no-deps --document-private-items
+        env:
+            RUSTDOCFLAGS: "-Dwarnings -Arustdoc::private_intra_doc_links"
+
+      # Use `${{ !cancelled() }}` to run quck tests after the longer clippy run
+      - name: Check formatting
+        if: ${{ !cancelled() }}
+        run: cargo fmt --all -- --check
+
+      # https://github.com/facebookincubator/cargo-guppy/tree/bec4e0eb29dcd1faac70b1b5360267fc02bf830e/tools/cargo-hakari#2-keep-the-workspace-hack-up-to-date-in-ci
+      - name: Check rust dependencies
+        if: ${{ !cancelled() }}
+        run: |
+          cargo hakari generate --diff  # workspace-hack Cargo.toml is up-to-date
+          cargo hakari manage-deps --dry-run  # all workspace crates depend on workspace-hack
+
+      # https://github.com/EmbarkStudios/cargo-deny
+      - name: Check rust licenses/bans/advisories/sources
+        if: ${{ !cancelled() }}
+        run: cargo deny check --hide-inclusion-graph
 
   build-and-test-locally:
     needs: [ tag, build-build-tools-image ]
@@ -824,7 +890,7 @@ jobs:
           docker compose -f ./docker-compose/docker-compose.yml down
 
   promote-images-dev:
-    needs: [ check-permissions, tag, vm-compute-node-image, neon-image ]
+    needs: [ check-permissions, tag, vm-compute-node-image ]
     runs-on: ubuntu-22.04
 
     permissions:

.github/workflows/pre-merge-checks.yml

@@ -1,12 +1,6 @@
 name: Pre-merge checks
 
 on:
-  pull_request:
-    paths:
-      - .github/workflows/_check-codestyle-python.yml
-      - .github/workflows/_check-codestyle-rust.yml
-      - .github/workflows/build-build-tools-image.yml
-      - .github/workflows/pre-merge-checks.yml
   merge_group:
     branches:
       - main
@@ -23,10 +17,8 @@ jobs:
     runs-on: ubuntu-22.04
     outputs:
       python-changed: ${{ steps.python-src.outputs.any_changed }}
-      rust-changed: ${{ steps.rust-src.outputs.any_changed }}
     steps:
       - uses: actions/checkout@v4
-
       - uses: tj-actions/changed-files@4edd678ac3f81e2dc578756871e4d00c19191daf # v45.0.4
         id: python-src
         with:
@@ -38,25 +30,11 @@ jobs:
             poetry.lock
             pyproject.toml
 
-      - uses: tj-actions/changed-files@4edd678ac3f81e2dc578756871e4d00c19191daf # v45.0.4
-        id: rust-src
-        with:
-          files: |
-            .github/workflows/_check-codestyle-rust.yml
-            .github/workflows/build-build-tools-image.yml
-            .github/workflows/pre-merge-checks.yml
-            **/**.rs
-            **/Cargo.toml
-            Cargo.toml
-            Cargo.lock
-
       - name: PRINT ALL CHANGED FILES FOR DEBUG PURPOSES
         env:
           PYTHON_CHANGED_FILES: ${{ steps.python-src.outputs.all_changed_files }}
-          RUST_CHANGED_FILES: ${{ steps.rust-src.outputs.all_changed_files }}
         run: |
           echo "${PYTHON_CHANGED_FILES}"
-          echo "${RUST_CHANGED_FILES}"
 
   build-build-tools-image:
     if: needs.get-changed-files.outputs.python-changed == 'true'
@@ -77,16 +55,6 @@ jobs:
       build-tools-image: ${{ needs.build-build-tools-image.outputs.image }}-bookworm-x64
     secrets: inherit
 
-  check-codestyle-rust:
-    if: needs.get-changed-files.outputs.rust-changed == 'true'
-    needs: [ get-changed-files, build-build-tools-image ]
-    uses: ./.github/workflows/_check-codestyle-rust.yml
-    with:
-      # `-bookworm-x64` suffix should match the combination in `build-build-tools-image`
-      build-tools-image: ${{ needs.build-build-tools-image.outputs.image }}-bookworm-x64
-      archs: '["x64"]'
-    secrets: inherit
-
   # To get items from the merge queue merged into main we need to satisfy "Status checks that are required".
   # Currently we require 2 jobs (checks with exact name):
   # - conclusion
@@ -99,7 +67,6 @@ jobs:
     needs:
       - get-changed-files
       - check-codestyle-python
-      - check-codestyle-rust
     runs-on: ubuntu-22.04
     steps:
       - name: Create fake `neon-cloud-e2e` check

Cargo.lock

@@ -1312,7 +1312,6 @@ dependencies = [
  "tracing-utils",
  "url",
  "utils",
- "uuid",
  "vm_monitor",
  "workspace_hack",
  "zstd",

compute/compute-node.Dockerfile

@@ -34,11 +34,11 @@ RUN case $DEBIAN_VERSION in \
         echo "Unknown Debian version ${DEBIAN_VERSION}" && exit 1 \
       ;; \
     esac && \
-    apt update &&  \
+    apt update && \
     apt install --no-install-recommends --no-install-suggests -y \
     ninja-build git autoconf automake libtool build-essential bison flex libreadline-dev \
     zlib1g-dev libxml2-dev libcurl4-openssl-dev libossp-uuid-dev wget ca-certificates pkg-config libssl-dev \
-    libicu-dev libxslt1-dev liblz4-dev libzstd-dev zstd \
+    libicu-dev libxslt1-dev liblz4-dev libzstd-dev zstd g++ \
     $VERSION_INSTALLS \
     && apt clean && rm -rf /var/lib/apt/lists/*
 
@@ -995,50 +995,24 @@ RUN wget https://github.com/kelvich/pg_tiktoken/archive/9118dd4549b7d8c0bbc98e04
 #########################################################################################
 #
 # Layer "pg-pgx-ulid-build"
-# Compile "pgx_ulid" extension for v16 and below
+# Compile "pgx_ulid" extension
 #
 #########################################################################################
 
 FROM rust-extensions-build AS pg-pgx-ulid-build
 ARG PG_VERSION
 
-RUN case "${PG_VERSION}" in \
-    "v14" | "v15" | "v16") \
-        ;; \
-    *) \
-        echo "skipping the version of pgx_ulid for $PG_VERSION" && exit 0 \
-        ;; \
+# doesn't support v17 yet
+# https://github.com/pksunkara/pgx_ulid/pull/52
+RUN case "${PG_VERSION}" in "v17") \
+    echo "pgx_ulid does not support pg17 as of the latest version (0.1.5)" && exit 0;; \
     esac && \
     wget https://github.com/pksunkara/pgx_ulid/archive/refs/tags/v0.1.5.tar.gz -O pgx_ulid.tar.gz && \
-    echo "9d1659a2da65af0133d5451c454de31b37364e3502087dadf579f790bc8bef17  pgx_ulid.tar.gz" | sha256sum --check && \
+    echo "9d1659a2da65af0133d5451c454de31b37364e3502087dadf579f790bc8bef17 pgx_ulid.tar.gz" | sha256sum --check && \
     mkdir pgx_ulid-src && cd pgx_ulid-src && tar xzf ../pgx_ulid.tar.gz --strip-components=1 -C . && \
-    sed -i 's/pgrx       = "^0.11.2"/pgrx       = { version = "0.11.3", features = [ "unsafe-postgres" ] }/g' Cargo.toml && \
+    sed -i 's/pgrx       = "^0.11.2"/pgrx = { version = "=0.11.3", features = [ "unsafe-postgres" ] }/g' Cargo.toml && \
     cargo pgrx install --release && \
-    echo 'trusted = true' >> /usr/local/pgsql/share/extension/ulid.control
-
-#########################################################################################
-#
-# Layer "pg-pgx-ulid-pgrx12-build"
-# Compile "pgx_ulid" extension for v17 and up
-#
-#########################################################################################
-
-FROM rust-extensions-build-pgrx12 AS pg-pgx-ulid-pgrx12-build
-ARG PG_VERSION
-
-RUN case "${PG_VERSION}" in \
-    "v17") \
-        ;; \
-    *) \
-        echo "skipping the version of pgx_ulid for $PG_VERSION" && exit 0 \
-        ;; \
-    esac && \
-    wget https://github.com/pksunkara/pgx_ulid/archive/refs/tags/v0.2.0.tar.gz -O pgx_ulid.tar.gz && \
-    echo "cef6a9a2e5e7bd1a10a18989286586ee9e6c1c06005a4055cff190de41bf3e9f pgx_ulid.tar.gz" | sha256sum --check && \
-    mkdir pgx_ulid-src && cd pgx_ulid-src && tar xzf ../pgx_ulid.tar.gz --strip-components=1 -C . && \
-    sed -i 's/pgrx       = "^0.12.7"/pgrx       = { version = "0.12.9", features = [ "unsafe-postgres" ] }/g' Cargo.toml && \
-    cargo pgrx install --release && \
-    echo 'trusted = true' >> /usr/local/pgsql/share/extension/pgx_ulid.control
+    echo "trusted = true" >> /usr/local/pgsql/share/extension/ulid.control
 
 #########################################################################################
 #
@@ -1133,6 +1107,46 @@ RUN wget https://github.com/Mooncake-Labs/pg_mooncake/releases/download/v0.1.0/p
     make install -j $(getconf _NPROCESSORS_ONLN) && \
     echo 'trusted = true' >> /usr/local/pgsql/share/extension/pg_mooncake.control
 
+#########################################################################################
+#
+# Layer "pg-duckdb-pg-build"
+# compile pg_duckdb extension
+#
+#########################################################################################
+
+FROM build-deps AS pg-duckdb-pg-build
+ARG PG_VERSION
+COPY --from=pg-build /usr/local/pgsql/ /usr/local/pgsql/
+COPY compute/patches/duckdb-v1-1-3.patch /duckdb-v1-1-3.patch
+COPY compute/patches/pg_duckdb-0-2-0.patch /pg_duckdb-0-2-0.patch
+
+ENV PATH="/usr/local/pgsql/bin/:$PATH"
+
+
+# pg_duckdb build requires source dir to be a git repo to get submodules
+# allow neon_superuser to execute some functions that in pg_duckdb are available to superuser only 
+# cache management functions duckdb.cache(), duckdb.cache_info(), duckdb.cache_delete()
+# extension management function duckdb.install_extension()
+# for debugging purposes raw query and reset ddb duckdb.raw_query(), duckdb.recycle_ddb()
+RUN git clone --depth 1 --branch v0.2.0 https://github.com/duckdb/pg_duckdb.git pg_duckdb-src && \
+    cd pg_duckdb-src && \
+    git submodule update --init --recursive && \
+    cd third_party/duckdb && \
+    patch -p1 < /duckdb-v1-1-3.patch && \
+    cd ../.. && \
+    patch -p1 < /pg_duckdb-0-2-0.patch && \
+    make install -j $(getconf _NPROCESSORS_ONLN) && \
+    echo 'trusted = true' >> /usr/local/pgsql/share/extension/pg_duckdb.control && \
+    file=/usr/local/pgsql/share/extension/pg_duckdb--0.1.0--0.2.0.sql && \
+    echo 'GRANT ALL ON FUNCTION duckdb.cache(TEXT, TEXT) TO neon_superuser;' >> $file && \
+    echo 'GRANT ALL ON FUNCTION duckdb.cache_info() TO neon_superuser;' >> $file && \
+    echo 'GRANT ALL ON FUNCTION duckdb.cache_delete(cache_key TEXT) TO neon_superuser;' >> $file && \
+    echo 'GRANT ALL ON FUNCTION duckdb.install_extension(TEXT) TO neon_superuser;' >> $file && \
+    echo 'GRANT ALL ON FUNCTION duckdb.raw_query(TEXT) TO neon_superuser;' >> $file && \
+    echo 'GRANT ALL ON PROCEDURE duckdb.recycle_ddb() TO neon_superuser;'  >> $file && \
+    echo 'GRANT ALL ON TABLE duckdb.extensions TO neon_superuser;' >> $file && \
+    echo 'GRANT ALL ON SEQUENCE duckdb.extensions_table_seq TO neon_superuser;' >> $file
+        
 #########################################################################################
 #
 # Layer "pg_repack"
@@ -1183,7 +1197,6 @@ COPY --from=timescaledb-pg-build /usr/local/pgsql/ /usr/local/pgsql/
 COPY --from=pg-hint-plan-pg-build /usr/local/pgsql/ /usr/local/pgsql/
 COPY --from=pg-cron-pg-build /usr/local/pgsql/ /usr/local/pgsql/
 COPY --from=pg-pgx-ulid-build /usr/local/pgsql/ /usr/local/pgsql/
-COPY --from=pg-pgx-ulid-pgrx12-build /usr/local/pgsql/ /usr/local/pgsql/
 COPY --from=pg-session-jwt-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/
@@ -1196,6 +1209,7 @@ COPY --from=pg-ivm-build /usr/local/pgsql/ /usr/local/pgsql/
 COPY --from=pg-partman-build /usr/local/pgsql/ /usr/local/pgsql/
 COPY --from=pg-mooncake-build /usr/local/pgsql/ /usr/local/pgsql/
 COPY --from=pg-repack-build /usr/local/pgsql/ /usr/local/pgsql/
+COPY --from=pg-duckdb-pg-build /usr/local/pgsql/ /usr/local/pgsql/
 COPY pgxn/ pgxn/
 
 RUN make -j $(getconf _NPROCESSORS_ONLN) \

compute/patches/duckdb-v1-1-3.patch

@@ -0,0 +1,122 @@
+diff --git a/src/common/virtual_file_system.cpp b/src/common/virtual_file_system.cpp
+index 74892a4e05..60e9e7af88 100644
+--- a/src/common/virtual_file_system.cpp
++++ b/src/common/virtual_file_system.cpp
+@@ -173,7 +173,9 @@ void VirtualFileSystem::SetDisabledFileSystems(const vector<string> &names) {
+ 
+ FileSystem &VirtualFileSystem::FindFileSystem(const string &path) {
+ 	auto &fs = FindFileSystemInternal(path);
+-	if (!disabled_file_systems.empty() && disabled_file_systems.find(fs.GetName()) != disabled_file_systems.end()) {
++	// we allow LocalFileSystem access to duckdb itself (by specifiying trustedContext=true) 
++	// but not to duckdb users if disabled_file_systems='LocalFileSystem'
++	if (!isTrustedContext && !disabled_file_systems.empty() && disabled_file_systems.find(fs.GetName()) != disabled_file_systems.end()) {
+ 		throw PermissionException("File system %s has been disabled by configuration", fs.GetName());
+ 	}
+ 	return fs;
+diff --git a/src/include/duckdb/common/file_system.hpp b/src/include/duckdb/common/file_system.hpp
+index 0b83c4f393..58d70889cf 100644
+--- a/src/include/duckdb/common/file_system.hpp
++++ b/src/include/duckdb/common/file_system.hpp
+@@ -256,6 +256,30 @@ public:
+ 		DynamicCastCheck<TARGET>(this);
+ 		return reinterpret_cast<const TARGET &>(*this);
+ 	}
++public:
++	DUCKDB_API virtual void setTrusted(){
++		this->isTrustedContext = true;
++	}
++	DUCKDB_API virtual void setUntrusted(){
++		this->isTrustedContext = false;
++	}
++
++public:
++  class TrustedContext {
++    public:
++        TrustedContext(FileSystem &fs) : fileSystem(fs) {
++            fileSystem.setTrusted();
++        }
++        ~TrustedContext() {
++            fileSystem.setUntrusted();
++        }
++
++    private:
++        FileSystem &fileSystem;
++    };
++
++protected:
++	bool isTrustedContext = false;
+ };
+ 
+ } // namespace duckdb
+diff --git a/src/include/duckdb/common/opener_file_system.hpp b/src/include/duckdb/common/opener_file_system.hpp
+index 2d35512b21..d1597173ee 100644
+--- a/src/include/duckdb/common/opener_file_system.hpp
++++ b/src/include/duckdb/common/opener_file_system.hpp
+@@ -143,6 +143,18 @@ public:
+ 	vector<string> ListSubSystems() override {
+ 		return GetFileSystem().ListSubSystems();
+ 	}
++
++private:
++
++	virtual void setTrusted() override {
++		this->isTrustedContext = true;
++		GetFileSystem().setTrusted();
++	}
++	
++	virtual void setUntrusted() override {
++		this->isTrustedContext = false;
++		GetFileSystem().setUntrusted();
++	}
+ };
+ 
+ } // namespace duckdb
+diff --git a/src/main/extension/extension_install.cpp b/src/main/extension/extension_install.cpp
+index 1258d95ead..def1ba71cb 100644
+--- a/src/main/extension/extension_install.cpp
++++ b/src/main/extension/extension_install.cpp
+@@ -57,6 +57,7 @@ const vector<string> ExtensionHelper::PathComponents() {
+ }
+ 
+ duckdb::string ExtensionHelper::DefaultExtensionFolder(FileSystem &fs) {
++	FileSystem::TrustedContext trusted(fs);
+ 	string home_directory = fs.GetHomeDirectory();
+ 	// exception if the home directory does not exist, don't create whatever we think is home
+ 	if (!fs.DirectoryExists(home_directory)) {
+@@ -84,6 +85,7 @@ string ExtensionHelper::ExtensionDirectory(DatabaseInstance &db, FileSystem &fs)
+ 		extension_directory = DefaultExtensionFolder(fs);
+ 	}
+ 	{
++		FileSystem::TrustedContext trusted(fs);
+ 		extension_directory = fs.ConvertSeparators(extension_directory);
+ 		// expand ~ in extension directory
+ 		extension_directory = fs.ExpandPath(extension_directory);
+@@ -107,6 +109,7 @@ string ExtensionHelper::ExtensionDirectory(DatabaseInstance &db, FileSystem &fs)
+ 
+ 	auto path_components = PathComponents();
+ 	for (auto &path_ele : path_components) {
++		FileSystem::TrustedContext trusted(fs);
+ 		extension_directory = fs.JoinPath(extension_directory, path_ele);
+ 		if (!fs.DirectoryExists(extension_directory)) {
+ 			fs.CreateDirectory(extension_directory);
+@@ -538,7 +541,7 @@ ExtensionHelper::InstallExtensionInternal(DatabaseInstance &db, FileSystem &fs,
+ 	if (!db.config.options.enable_external_access) {
+ 		throw PermissionException("Installing extensions is disabled through configuration");
+ 	}
+-
++	FileSystem::TrustedContext trusted(fs);
+ 	auto extension_name = ApplyExtensionAlias(fs.ExtractBaseName(extension));
+ 	string local_extension_path = fs.JoinPath(local_path, extension_name + ".duckdb_extension");
+ 	string temp_path = local_extension_path + ".tmp-" + UUID::ToString(UUID::GenerateRandomUUID());
+diff --git a/src/main/extension/extension_load.cpp b/src/main/extension/extension_load.cpp
+index b0282a7103..c2765f97c6 100644
+--- a/src/main/extension/extension_load.cpp
++++ b/src/main/extension/extension_load.cpp
+@@ -293,6 +293,7 @@ bool ExtensionHelper::TryInitialLoad(DatabaseInstance &db, FileSystem &fs, const
+ 	if (!db.config.options.enable_external_access) {
+ 		throw PermissionException("Loading external extensions is disabled through configuration");
+ 	}
++	FileSystem::TrustedContext trusted(fs);
+ 	auto filename = fs.ConvertSeparators(extension);
+ 
+ 	bool direct_load;

compute/patches/pg_duckdb-0-2-0.patch

@@ -0,0 +1,61 @@
+diff --git a/third_party/cached_httpfs/http_file_cache.cpp b/third_party/cached_httpfs/http_file_cache.cpp
+index 813db57..c6e4c3a 100644
+--- a/third_party/cached_httpfs/http_file_cache.cpp
++++ b/third_party/cached_httpfs/http_file_cache.cpp
+@@ -4,20 +4,22 @@ namespace duckdb {
+ 
+ CachedFile::CachedFile(const string &cache_dir, FileSystem &fs, const string &key, bool cache_file) : cache_directory(cache_dir), fs(fs) {
+ 	file_name = cache_dir + "/" + key;
+-
++	FileSystem::TrustedContext trusted1(fs);
+ 	GetDirectoryCacheLock(cache_dir);
+ 
+ 	FileOpenFlags flags =
+ 	    FileFlags::FILE_FLAGS_READ | FileFlags::FILE_FLAGS_NULL_IF_NOT_EXISTS | FileLockType::READ_LOCK;
++	FileSystem::TrustedContext trusted2(fs);
+ 	handle = fs.OpenFile(file_name, flags);
+ 	if (handle) {
+ 		initialized = true;
+ 		size = handle->GetFileSize();
+ 	} else if (cache_file) {
+ 		flags = FileFlags::FILE_FLAGS_WRITE | FileFlags::FILE_FLAGS_FILE_CREATE | FileLockType::WRITE_LOCK;
++		FileSystem::TrustedContext trusted3(fs);
+ 		handle = fs.OpenFile(file_name, flags);
+ 	}
+-
++	FileSystem::TrustedContext trusted4(fs);
+ 	ReleaseDirectoryCacheLock();
+ }
+ 
+@@ -30,6 +32,7 @@ void CachedFile::GetDirectoryCacheLock(const string &cache_dir) {
+ 	FileOpenFlags flags = FileFlags::FILE_FLAGS_WRITE | FileFlags::FILE_FLAGS_FILE_CREATE |
+ 	                      FileFlags::FILE_FLAGS_EXCLUSIVE_CREATE | FileFlags::FILE_FLAGS_NULL_IF_EXISTS |
+ 	                      FileLockType::WRITE_LOCK;
++	FileSystem::TrustedContext trusted(fs);
+ 	directory_lock_handle = fs.OpenFile(lock_file, flags);
+ 	if (directory_lock_handle == nullptr) {
+ 		flags = FileFlags::FILE_FLAGS_WRITE | FileLockType::WRITE_LOCK;
+@@ -52,6 +55,7 @@ void CachedFileHandle::WriteMetadata(const string &cache_key, const string &remo
+ 	D_ASSERT(!file->initialized);
+ 	string metadata_file_name = file->cache_directory + "/" + cache_key + ".meta";
+ 	FileOpenFlags flags = FileFlags::FILE_FLAGS_WRITE | FileFlags::FILE_FLAGS_FILE_CREATE | FileLockType::WRITE_LOCK;
++	FileSystem::TrustedContext trusted(file->fs);
+ 	auto handle = file->fs.OpenFile(metadata_file_name, flags);
+ 	auto cached_file_timestamp = std::chrono::system_clock::to_time_t(std::chrono::system_clock::now());
+ 	string metadata_info = cache_key + "," + remote_path + "," + std::to_string(total_size) + "," + std::to_string(cached_file_timestamp);
+@@ -66,6 +70,7 @@ void CachedFileHandle::SetInitialized(idx_t total_size) {
+ 	file->initialized = true;
+ 	file->handle->Close();
+ 	FileOpenFlags flags = FileFlags::FILE_FLAGS_READ | FileLockType::READ_LOCK;
++	FileSystem::TrustedContext trusted(file->fs);
+ 	file->handle = file->fs.OpenFile(file->file_name, flags);
+ }
+ 
+@@ -93,6 +98,7 @@ void CachedFileHandle::Read(void *buffer, idx_t length, idx_t offset) {
+ //! Get cache entry, create if not exists only if caching is enabled
+ shared_ptr<CachedFile> HTTPFileCache::GetCachedFile(const string &cache_dir, const string &key, bool cache_file) {
+ 	lock_guard<mutex> lock(cached_files_mutex);
++	FileSystem::TrustedContext trusted(db->GetFileSystem());
+ 	auto it = cached_files.find(key);
+ 	if (it != cached_files.end()) {
+ 		return it->second;

compute_tools/Cargo.toml

@@ -51,7 +51,6 @@ tracing-subscriber.workspace = true
 tracing-utils.workspace = true
 thiserror.workspace = true
 url.workspace = true
-uuid.workspace = true
 prometheus.workspace = true
 
 postgres_initdb.workspace = true

compute_tools/src/http/server.rs

@@ -1,14 +1,15 @@
 use std::{
     net::{IpAddr, Ipv6Addr, SocketAddr},
-    sync::Arc,
+    sync::{
+        atomic::{AtomicU64, Ordering},
+        Arc,
+    },
     thread,
     time::Duration,
 };
 
 use anyhow::Result;
 use axum::{
-    extract::Request,
-    middleware::{self, Next},
     response::{IntoResponse, Response},
     routing::{get, post},
     Router,
@@ -16,9 +17,11 @@ use axum::{
 use http::StatusCode;
 use tokio::net::TcpListener;
 use tower::ServiceBuilder;
-use tower_http::{request_id::PropagateRequestIdLayer, trace::TraceLayer};
+use tower_http::{
+    request_id::{MakeRequestId, PropagateRequestIdLayer, RequestId, SetRequestIdLayer},
+    trace::TraceLayer,
+};
 use tracing::{debug, error, info, Span};
-use uuid::Uuid;
 
 use super::routes::{
     check_writability, configure, database_schema, dbs_and_roles, extension_server, extensions,
@@ -31,24 +34,30 @@ async fn handle_404() -> Response {
     StatusCode::NOT_FOUND.into_response()
 }
 
-const X_REQUEST_ID: &str = "x-request-id";
+#[derive(Clone, Default)]
+struct ComputeMakeRequestId(Arc<AtomicU64>);
 
-/// This middleware function allows compute_ctl to generate its own request ID
-/// if one isn't supplied. The control plane will always send one as a UUID. The
-/// neon Postgres extension on the other hand does not send one.
-async fn maybe_add_request_id_header(mut request: Request, next: Next) -> Response {
-    let headers = request.headers_mut();
+impl MakeRequestId for ComputeMakeRequestId {
+    fn make_request_id<B>(
+        &mut self,
+        _request: &http::Request<B>,
+    ) -> Option<tower_http::request_id::RequestId> {
+        let request_id = self
+            .0
+            .fetch_add(1, Ordering::SeqCst)
+            .to_string()
+            .parse()
+            .unwrap();
 
-    if headers.get(X_REQUEST_ID).is_none() {
-        headers.append(X_REQUEST_ID, Uuid::new_v4().to_string().parse().unwrap());
+        Some(RequestId::new(request_id))
     }
-
-    next.run(request).await
 }
 
 /// Run the HTTP server and wait on it forever.
 #[tokio::main]
 async fn serve(port: u16, compute: Arc<ComputeNode>) {
+    const X_REQUEST_ID: &str = "x-request-id";
+
     let mut app = Router::new()
         .route("/check_writability", post(check_writability::is_writable))
         .route("/configure", post(configure::configure))
@@ -73,8 +82,9 @@ async fn serve(port: u16, compute: Arc<ComputeNode>) {
         .fallback(handle_404)
         .layer(
             ServiceBuilder::new()
-                // Add this middleware since we assume the request ID exists
-                .layer(middleware::from_fn(maybe_add_request_id_header))
+                .layer(SetRequestIdLayer::x_request_id(
+                    ComputeMakeRequestId::default(),
+                ))
                 .layer(
                     TraceLayer::new_for_http()
                         .on_request(|request: &http::Request<_>, _span: &Span| {

control_plane/src/pageserver.rs

@@ -418,11 +418,6 @@ impl PageServerNode {
                 .map(serde_json::from_str)
                 .transpose()
                 .context("parse `wal_receiver_protocol_override` from json")?,
-            rel_size_v2_enabled: settings
-                .remove("rel_size_v2_enabled")
-                .map(|x| x.parse::<bool>())
-                .transpose()
-                .context("Failed to parse 'rel_size_v2_enabled' as bool")?,
         };
         if !settings.is_empty() {
             bail!("Unrecognized tenant settings: {settings:?}")

control_plane/storcon_cli/src/main.rs

@@ -298,7 +298,14 @@ impl FromStr for SkSchedulingPolicyArg {
     type Err = anyhow::Error;
 
     fn from_str(s: &str) -> Result<Self, Self::Err> {
-        SkSchedulingPolicy::from_str(s).map(Self)
+        match s {
+            "active" => Ok(Self(SkSchedulingPolicy::Active)),
+            "disabled" => Ok(Self(SkSchedulingPolicy::Disabled)),
+            "decomissioned" => Ok(Self(SkSchedulingPolicy::Decomissioned)),
+            _ => Err(anyhow::anyhow!(
+                "Unknown scheduling policy '{s}', try active,disabled,decomissioned"
+            )),
+        }
     }
 }
 

docs/rfcs/041-sharded-ingest.md

@@ -1,255 +0,0 @@
-# 
-Created on Aug 2024
-Implemented on Jan 2025
-
-## Summary
-
-Data in large tenants is split up between multiple pageservers according to key hashes, as
-introduced in the [sharding RFC](031-sharding-static.md) and [shard splitting RFC](032-shard-splitting.md).
-
-Whereas currently we send all WAL to all pageserver shards, and each shard filters out the data that it needs,
-in this RFC we add a mechanism to filter the WAL on the safekeeper, so that each shard receives
-only the data it needs.
-
-This will place some extra CPU load on the safekeepers, in exchange for reducing the network bandwidth
-for ingesting WAL back to scaling as O(1) with shard count, rather than O(N_shards).
-
-## Motivation
-
-1. Large databases require higher shard counts.  Whereas currently we run with up to 8 shards for tenants
-with a few TB of storage, the next order of magnitude capacity increase will require tens of shards, such
-that sending all WAL to all shards is impractical in terms of bandwidth.
-2. For contemporary database sizes (~2TB), the pageserver is the bottleneck for ingest: since each
-   shard has to decode and process the whole WAL, sharding doesn't fully relieve this bottleneck.  To achieve significantly higher ingest speeds, we need to filter the WAL earlier so that each pageserver
-   only has to process relevant parts.
-
-## Non Goals (if relevant)
-
-We do not seek to introduce multiple WALs per timeline, or to share the work of handling a timeline's
-WAL across safekeepers (beyond simple 3x replication).  This RFC may be thought of as an incremental
-move of the ingestion bottleneck up the stack: instead of high write rates bottlenecking on the
-pageserver, they will bottleneck on the safekeeper.
-
-## Impacted components (e.g. pageserver, safekeeper, console, etc)
-
-Safekeeper, pageserver.
-
-There will be no control plane or storage controller coordination needed, as pageservers will directly
-indicate their sharding parameters to the safekeeper when subscribing for WAL.
-
-## Proposed implementation
-
-Terminology:
-- "Data pages" refers to postgres relation blocks, and SLRU blocks.
-- "Metadata pages" refers to everything else the pageserver stores, such as relation sizes and
-  directories of relations.
-
-### Phase 1: Refactor ingest
-
-Currently, pageserver ingest code is structured approximately as follows:
-1. `handle_walreceiver_connection` reads a stream of binary WAL records off a network
-   socket
-2. `WalIngest::ingest_record` to translate the record into a series of page-level modifications
-3. `DatadirModification` accumulates page updates from several `ingest_record` calls, and when
-   its `commit()` method is called, flushes these into a Timeline's open `InMemoryLayer`.
-
-This process currently assumes access to a pageserver `Timeline` throughout `ingest_record` and
-from `DatadirModification`, which is used to do read-modify-write cycles on metadata pages
-such as relation sizes and the master DBDIR page.  It also assumes that records are ingested
-strictly one after the other: they cannot be ingested in parallel because each record assumes
-that earlier records' changes have already been applied to `Timeline`.
-
-This code will be refactored to disentangle the simple, fast decode of relation page writes
-from the more complex logic for updating internal metadata.  An intermediate representation
-called `InterpretedWalRecords` will be introduced.  This is similar to the internal state of
-a `DatadirModification`, but does not require access to a Timeline.  Instead of storing
-metadata updates as materialized writes to pages, it will accumulate these as abstract operations,
-for example rather than including a write to a relation size key, this structure will include
-an operation that indicates "Update relation _foo_'s size to the max of its current value and
-_bar_", such that these may be applied later to a real Timeline.
-
-The `DatadirModification` will be aware of the `EphemeralFile` format, so that as it accumulates
-simple page writes of relation blocks, it can write them directly into a buffer in the serialized
-format.  This will avoid the need to later deserialize/reserialize this data when passing the
-structure between safekeeper and pageserver.
-
-The new pipeline will be:
-1. `handle_walreceiver_connection` reads a stream of binary WAL records off a network
-2. A `InterpretedWalRecords` is generated from the incoming WAL records.  This does not
-   require a reference to a Timeline.
-3. The logic that is current spread between `WalIngest` and `DatadirModification` for updating
-   metadata will be refactored to consume the metadata operations from the `InterpretedWalRecords`
-   and turn them into literal writes to metadata pages.  This part must be done sequentially.
-4. The resulting buffer of metadata page writes is combined with the buffer of relation block
-   writes, and written into the `InMemoryLayer`.
-
-Implemented in:
-1. https://github.com/neondatabase/neon/pull/9472
-2. https://github.com/neondatabase/neon/pull/9504
-3. https://github.com/neondatabase/neon/pull/9524
-
-### Phase 2: Decode & filter on safekeeper
-
-In the previous phase, the ingest code was modified to be able to do most of its work without access to
-a Timeline: this first stage of ingest simply converts a series of binary wal records into
-a buffer of relation/SLRU page writes, and a buffer of abstract metadata writes.
-
-The modified ingest code may be transplanted from pageserver to safekeeper (probably via a
-shared crate).  The safekeeper->pageserver network protocol is modified to:
- - in subscription requests, send the `ShardIdentity` from the pageserver to the safekeeper
- - in responses, transmit a `InterpretedWalRecords` instead of a raw `WalRecord`.
- - use the `ShardIdentity` to filter the `ProcessedWalIngest` to relevant content for
-   the subscribing shard before transmitting it.
-
-The overall behavior of the pageserver->safekeeper interaction remains the same, in terms of
-consistent LSN feedback, and connection management.  Only the payload of the subscriptions
-changes, to express an LSN range of WAL as a filtered `ProcessedWalIngest` instead of the
-raw data.
-
-The ingest code on the pageserver can now skip the part where it does the first phase of
-processing, as it will receive pre-processed, compressed data off the wire.
-
-Note that `InterpretedWalRecord` batches multiple `InterpretedWalRecord(s)` in the same network
-message. Safekeeper reads WAL in chunks of 16 blocks and then decodes as many Postgres WAL records
-as possible. Each Postgres WAL record maps to one `InterpretedWalRecord` for potentially multiple shards.
-Hence, the size of the batch is given by the number of Postgres WAL records that fit in 16 blocks.
-
-The protocol needs to support evolution. Protobuf was chosen here with the view that, in the future,
-we may migrate it to GRPC altogether
-
-Implemented in:
-1. https://github.com/neondatabase/neon/pull/9746
-2. https://github.com/neondatabase/neon/pull/9821
-
-### Phase 3: Fan out interpreted WAL
-
-In the previous phase, the initial processing of WAL was moved to the safekeeper, but it is still
-done once for each shard: this will generate O(N_shards) CPU work on the safekeeper (especially
-when considering converting to Protobuf format and compression).
-
-To avoid this, we fan-out WAL from one (tenant, timeline, shard) to all other shards subscribed on
-the same safekeeper. Under normal operation, the WAL will be read from disk, decoded and interpreted
-_only_ once per (safekeeper, timeline).
-
-When the first shard of a sharded timeline subscribes to a given safekeeper a task is spawned
-for the WAL reader (`InterpretedWalReader`). This task reads WAL, decodes, interprets it and sends
-it to the sender (`InterpretedWalSender`). The sender is a future that is polled from the connection
-task. When further shards subscribe on the safekeeper they will attach themselves to the existing WAL reader.
-There's two cases to consider:
-1. The shard's requested `start_lsn` is ahead of the current position of the WAL reader. In this case, the shard
-will start receiving data when the reader reaches that LSN. The intuition here is that there's little to gain
-by letting shards "front-run" since compute backpressure is based on the laggard LSN.
-2. The shard's requested `start_lsn` is below the current position of the WAL reader. In this case, the WAL reader
-gets reset to this requested position (same intuition). Special care is taken such that advanced shards do not receive
-interpreted WAL records below their current position.
-
-The approach above implies that there is at most one WAL reader per (tenant, timeline) on a given safekeeper at any point in time.
-If this turns out to be operationally problematic, there's a trick we can deploy: `--max-delta-for-fanout` is an optional safekeeper
-argument that controls the max absolute delta between a new shard and the current WAL position of the WAL reader. If the absolute
-delta is above that value, a new reader is spawned. Note that there's currently no concurrency control on the number of WAL readers,
-so it's recommended to use large values to avoid pushing CPU utilisation too high.
-
-Unsharded tenants do not spawn a separate task for the interpreted WAL reader since there's no benefit to it. Instead they poll
-the reader and sender concurrently from the connection task.
-
-Shard splits are interesting here because it is the only case when the same shard might have two subscriptions at the same time.
-This is handled by giving readers a unique identifier. Both shards will receive the same data while respecting their requested start
-position.
-
-Implemented in:
-1. https://github.com/neondatabase/neon/pull/10190
-
-## Deployment
-
-Each phase shall be deployed independently. Special care should be taken around protocol changes.
-
-## Observability Tips
-
-* The safekeeper logs the protocol requested by the pageserver
-along with the pageserver ID, tenant, timeline and shard: `starting streaming from`.
-* There's metrics for the number of wal readers:
-  * `safekeeper_wal_readers{kind="task", target=~"pageserver.*"}` gives the number of wal reader tasks for each SK
-  * `safekeeper_wal_readers{kind="future", target=~"pageserver.*"}` gives the numer of wal readers polled inline by each SK
-  * `safekeeper_interpreted_wal_reader_tasks` gives the number of wal reader tasks per tenant, timeline
-* Interesting log lines for the fan-out reader:
-  * `Spawning interpreted`: first shard creates the interpreted wal reader
-  * `Fanning out`: a subsequent shard attaches itself to an interpreted wal reader
-  * `Aborting interpreted`: all senders have finished and the reader task is being aborted
-
-## Future Optimizations
-
-This sections describes some improvement areas which may be revisited in the future.
-
-### Buffering of Interpreted WAL
-
-The interpreted WAL reader may buffer interpreted WAL records in user space to help with serving
-subscriptions that are lagging behind the current position of the reader.
-
-Counterpoints:
-* Safekeepers serve many thousands of timelines and allocating a buffer for each might be wasteful,
-especially given that it would go unused on the happy path.
-* WAL is buffered in the kernel page cache. Usually we'd only pay the CPU cost of decoding and interpreting.
-
-### Tweaking the Pagserver Safekeeper Selection Algorithm
-
-We could make the pageserver aware of which safekeeper's already host shards for the timeline along
-with their current WAL positions. The pageserver should then prefer safkeepers that are in the same
-AZ _and_ already have a shard with a position close to the desired start position.
-
-We currently run one safekeeper per AZ, so the point is mute until that changes.
-
-### Pipelining first ingest phase
-
-The first ingest phase is a stateless transformation of a binary WAL record into a pre-processed
-output per shard.  To put multiple CPUs to work, we may pipeline this processing up to some defined buffer
-depth.
-
-## Alternatives considered
-
-### Give safekeepers enough state to fully decode WAL
-
-In this RFC, we only do the first phase of ingest on the safekeeper, because this is
-the phase that is stateless.  Subsequent changes then happen on the pageserver, with
-access to the `Timeline` state.
-
-We could do more work on the safekeeper if we transmitted metadata state to the safekeeper
-when subscribing to the WAL: for example, by telling the safekeeper all the relation sizes,
-so that it could then generate all the metadata writes for relation sizes.
-
-We avoid doing this for several reasons:
-1. Complexity: it's a more invasive protocol change
-2. Decoupling: having the safekeeper understand the `ProcessedWalIngest` already somewhat
-   infects it with knowledge of the pageserver, but this is mainly an abstract structure
-   that describes postgres writes.  However, if we taught the safekeeper about the exact
-   way that pageserver deals with metadata keys, this would be a much tighter coupling.
-3. Load: once the WAL has been processed to the point that it can be split between shards,
-   it is preferable to share out work on the remaining shards rather than adding extra CPU
-   load to the safekeeper.
-
-### Do pre-processing on the compute instead of the safekeeper
-
-Since our first stage of ingest is stateless, it could be done at any stage in the pipeline,
-all the way up to the compute.
-
-We choose not to do this, because it is useful for the safekeeper to store the raw WAL rather
-than just the preprocessed WAL:
-- The safekeeper still needs to be able to serve raw WAL back to postgres for e.g. physical replication
-- It simplifies our paxos implementation to have the offset in the write log be literally
-  the same as the LSN
-- Raw WAL must have a stable protocol since we might have to re-ingest it at arbitrary points in the future.
-  Storing raw WAL give us more flexibility to evolve the pageserver, safekeeper protocol.
-
-### Do wal pre-processing on shard 0 or a separate service, send it to other shards from there
-
-If we wanted to keep the safekeepers as entirely pure stores of raw WAL bytes, then
-we could do the initial decode and shard-splitting in some other location:
-- Shard 0 could subscribe to the full WAL and then send writes to other shards
-- A new intermediate service between the safekeeper and pageserver could do the splitting.
-
-So why not?
-- Extra network hop from shard 0 to the final destination shard
-- Clearly there is more infrastructure involved here compared with doing it inline on the safekeeper.
-- Safekeepers already have very light CPU load: typical cloud instances shapes with appropriate
-  disks for the safekeepers effectively have "free" CPU resources.
-- Doing extra work on shard 0 would complicate scheduling of shards on pageservers, because
-  shard 0 would have significantly higher CPU load under write workloads than other shards.

libs/pageserver_api/src/config.rs

@@ -301,10 +301,6 @@ pub struct TenantConfigToml {
     pub timeline_offloading: bool,
 
     pub wal_receiver_protocol_override: Option<PostgresClientProtocol>,
-
-    /// Enable rel_size_v2 for this tenant. Once enabled, the tenant will persist this information into
-    /// `index_part.json`, and it cannot be reversed.
-    pub rel_size_v2_enabled: Option<bool>,
 }
 
 pub mod defaults {
@@ -542,7 +538,6 @@ impl Default for TenantConfigToml {
             lsn_lease_length_for_ts: LsnLease::DEFAULT_LENGTH_FOR_TS,
             timeline_offloading: false,
             wal_receiver_protocol_override: None,
-            rel_size_v2_enabled: None,
         }
     }
 }

libs/pageserver_api/src/controller_api.rs

@@ -324,7 +324,7 @@ impl From<NodeSchedulingPolicy> for String {
 #[derive(Serialize, Deserialize, Clone, Copy, Eq, PartialEq, Debug)]
 pub enum SkSchedulingPolicy {
     Active,
-    Pause,
+    Disabled,
     Decomissioned,
 }
 
@@ -334,13 +334,9 @@ impl FromStr for SkSchedulingPolicy {
     fn from_str(s: &str) -> Result<Self, Self::Err> {
         Ok(match s {
             "active" => Self::Active,
-            "pause" => Self::Pause,
+            "disabled" => Self::Disabled,
             "decomissioned" => Self::Decomissioned,
-            _ => {
-                return Err(anyhow::anyhow!(
-                    "Unknown scheduling policy '{s}', try active,pause,decomissioned"
-                ))
-            }
+            _ => return Err(anyhow::anyhow!("Unknown scheduling state '{s}'")),
         })
     }
 }
@@ -350,7 +346,7 @@ impl From<SkSchedulingPolicy> for String {
         use SkSchedulingPolicy::*;
         match value {
             Active => "active",
-            Pause => "pause",
+            Disabled => "disabled",
             Decomissioned => "decomissioned",
         }
         .to_string()

libs/pageserver_api/src/models.rs

@@ -33,6 +33,7 @@ use crate::{
     reltag::RelTag,
     shard::{ShardCount, ShardStripeSize, TenantShardId},
 };
+use anyhow::bail;
 use bytes::{Buf, BufMut, Bytes, BytesMut};
 
 /// The state of a tenant in this pageserver.
@@ -497,8 +498,6 @@ pub struct TenantConfigPatch {
     pub timeline_offloading: FieldPatch<bool>,
     #[serde(skip_serializing_if = "FieldPatch::is_noop")]
     pub wal_receiver_protocol_override: FieldPatch<PostgresClientProtocol>,
-    #[serde(skip_serializing_if = "FieldPatch::is_noop")]
-    pub rel_size_v2_enabled: FieldPatch<bool>,
 }
 
 /// An alternative representation of `pageserver::tenant::TenantConf` with
@@ -530,7 +529,6 @@ pub struct TenantConfig {
     pub lsn_lease_length_for_ts: Option<String>,
     pub timeline_offloading: Option<bool>,
     pub wal_receiver_protocol_override: Option<PostgresClientProtocol>,
-    pub rel_size_v2_enabled: Option<bool>,
 }
 
 impl TenantConfig {
@@ -560,7 +558,6 @@ impl TenantConfig {
             mut lsn_lease_length_for_ts,
             mut timeline_offloading,
             mut wal_receiver_protocol_override,
-            mut rel_size_v2_enabled,
         } = self;
 
         patch.checkpoint_distance.apply(&mut checkpoint_distance);
@@ -605,7 +602,6 @@ impl TenantConfig {
         patch
             .wal_receiver_protocol_override
             .apply(&mut wal_receiver_protocol_override);
-        patch.rel_size_v2_enabled.apply(&mut rel_size_v2_enabled);
 
         Self {
             checkpoint_distance,
@@ -632,7 +628,6 @@ impl TenantConfig {
             lsn_lease_length_for_ts,
             timeline_offloading,
             wal_receiver_protocol_override,
-            rel_size_v2_enabled,
         }
     }
 }
@@ -1405,8 +1400,6 @@ pub enum PagestreamFeMessage {
     GetPage(PagestreamGetPageRequest),
     DbSize(PagestreamDbSizeRequest),
     GetSlruSegment(PagestreamGetSlruSegmentRequest),
-    #[cfg(feature = "testing")]
-    Test(PagestreamTestRequest),
 }
 
 // Wrapped in libpq CopyData
@@ -1418,22 +1411,6 @@ pub enum PagestreamBeMessage {
     Error(PagestreamErrorResponse),
     DbSize(PagestreamDbSizeResponse),
     GetSlruSegment(PagestreamGetSlruSegmentResponse),
-    #[cfg(feature = "testing")]
-    Test(PagestreamTestResponse),
-}
-
-// Keep in sync with `pagestore_client.h`
-#[repr(u8)]
-enum PagestreamFeMessageTag {
-    Exists = 0,
-    Nblocks = 1,
-    GetPage = 2,
-    DbSize = 3,
-    GetSlruSegment = 4,
-    /* future tags above this line */
-    /// For testing purposes, not available in production.
-    #[cfg(feature = "testing")]
-    Test = 99,
 }
 
 // Keep in sync with `pagestore_client.h`
@@ -1445,28 +1422,7 @@ enum PagestreamBeMessageTag {
     Error = 103,
     DbSize = 104,
     GetSlruSegment = 105,
-    /* future tags above this line */
-    /// For testing purposes, not available in production.
-    #[cfg(feature = "testing")]
-    Test = 199,
 }
-
-impl TryFrom<u8> for PagestreamFeMessageTag {
-    type Error = u8;
-    fn try_from(value: u8) -> Result<Self, u8> {
-        match value {
-            0 => Ok(PagestreamFeMessageTag::Exists),
-            1 => Ok(PagestreamFeMessageTag::Nblocks),
-            2 => Ok(PagestreamFeMessageTag::GetPage),
-            3 => Ok(PagestreamFeMessageTag::DbSize),
-            4 => Ok(PagestreamFeMessageTag::GetSlruSegment),
-            #[cfg(feature = "testing")]
-            99 => Ok(PagestreamFeMessageTag::Test),
-            _ => Err(value),
-        }
-    }
-}
-
 impl TryFrom<u8> for PagestreamBeMessageTag {
     type Error = u8;
     fn try_from(value: u8) -> Result<Self, u8> {
@@ -1477,8 +1433,6 @@ impl TryFrom<u8> for PagestreamBeMessageTag {
             103 => Ok(PagestreamBeMessageTag::Error),
             104 => Ok(PagestreamBeMessageTag::DbSize),
             105 => Ok(PagestreamBeMessageTag::GetSlruSegment),
-            #[cfg(feature = "testing")]
-            199 => Ok(PagestreamBeMessageTag::Test),
             _ => Err(value),
         }
     }
@@ -1596,20 +1550,6 @@ pub struct PagestreamDbSizeResponse {
     pub db_size: i64,
 }
 
-#[cfg(feature = "testing")]
-#[derive(Debug, PartialEq, Eq, Clone)]
-pub struct PagestreamTestRequest {
-    pub hdr: PagestreamRequest,
-    pub batch_key: u64,
-    pub message: String,
-}
-
-#[cfg(feature = "testing")]
-#[derive(Debug)]
-pub struct PagestreamTestResponse {
-    pub req: PagestreamTestRequest,
-}
-
 // This is a cut-down version of TenantHistorySize from the pageserver crate, omitting fields
 // that require pageserver-internal types.  It is sufficient to get the total size.
 #[derive(Serialize, Deserialize, Debug)]
@@ -1629,7 +1569,7 @@ impl PagestreamFeMessage {
 
         match self {
             Self::Exists(req) => {
-                bytes.put_u8(PagestreamFeMessageTag::Exists as u8);
+                bytes.put_u8(0);
                 bytes.put_u64(req.hdr.reqid);
                 bytes.put_u64(req.hdr.request_lsn.0);
                 bytes.put_u64(req.hdr.not_modified_since.0);
@@ -1640,7 +1580,7 @@ impl PagestreamFeMessage {
             }
 
             Self::Nblocks(req) => {
-                bytes.put_u8(PagestreamFeMessageTag::Nblocks as u8);
+                bytes.put_u8(1);
                 bytes.put_u64(req.hdr.reqid);
                 bytes.put_u64(req.hdr.request_lsn.0);
                 bytes.put_u64(req.hdr.not_modified_since.0);
@@ -1651,7 +1591,7 @@ impl PagestreamFeMessage {
             }
 
             Self::GetPage(req) => {
-                bytes.put_u8(PagestreamFeMessageTag::GetPage as u8);
+                bytes.put_u8(2);
                 bytes.put_u64(req.hdr.reqid);
                 bytes.put_u64(req.hdr.request_lsn.0);
                 bytes.put_u64(req.hdr.not_modified_since.0);
@@ -1663,7 +1603,7 @@ impl PagestreamFeMessage {
             }
 
             Self::DbSize(req) => {
-                bytes.put_u8(PagestreamFeMessageTag::DbSize as u8);
+                bytes.put_u8(3);
                 bytes.put_u64(req.hdr.reqid);
                 bytes.put_u64(req.hdr.request_lsn.0);
                 bytes.put_u64(req.hdr.not_modified_since.0);
@@ -1671,24 +1611,13 @@ impl PagestreamFeMessage {
             }
 
             Self::GetSlruSegment(req) => {
-                bytes.put_u8(PagestreamFeMessageTag::GetSlruSegment as u8);
+                bytes.put_u8(4);
                 bytes.put_u64(req.hdr.reqid);
                 bytes.put_u64(req.hdr.request_lsn.0);
                 bytes.put_u64(req.hdr.not_modified_since.0);
                 bytes.put_u8(req.kind);
                 bytes.put_u32(req.segno);
             }
-            #[cfg(feature = "testing")]
-            Self::Test(req) => {
-                bytes.put_u8(PagestreamFeMessageTag::Test as u8);
-                bytes.put_u64(req.hdr.reqid);
-                bytes.put_u64(req.hdr.request_lsn.0);
-                bytes.put_u64(req.hdr.not_modified_since.0);
-                bytes.put_u64(req.batch_key);
-                let message = req.message.as_bytes();
-                bytes.put_u64(message.len() as u64);
-                bytes.put_slice(message);
-            }
         }
 
         bytes.into()
@@ -1716,66 +1645,56 @@ impl PagestreamFeMessage {
             ),
         };
 
-        match PagestreamFeMessageTag::try_from(msg_tag)
-            .map_err(|tag: u8| anyhow::anyhow!("invalid tag {tag}"))?
-        {
-            PagestreamFeMessageTag::Exists => {
-                Ok(PagestreamFeMessage::Exists(PagestreamExistsRequest {
-                    hdr: PagestreamRequest {
-                        reqid,
-                        request_lsn,
-                        not_modified_since,
-                    },
-                    rel: RelTag {
-                        spcnode: body.read_u32::<BigEndian>()?,
-                        dbnode: body.read_u32::<BigEndian>()?,
-                        relnode: body.read_u32::<BigEndian>()?,
-                        forknum: body.read_u8()?,
-                    },
-                }))
-            }
-            PagestreamFeMessageTag::Nblocks => {
-                Ok(PagestreamFeMessage::Nblocks(PagestreamNblocksRequest {
-                    hdr: PagestreamRequest {
-                        reqid,
-                        request_lsn,
-                        not_modified_since,
-                    },
-                    rel: RelTag {
-                        spcnode: body.read_u32::<BigEndian>()?,
-                        dbnode: body.read_u32::<BigEndian>()?,
-                        relnode: body.read_u32::<BigEndian>()?,
-                        forknum: body.read_u8()?,
-                    },
-                }))
-            }
-            PagestreamFeMessageTag::GetPage => {
-                Ok(PagestreamFeMessage::GetPage(PagestreamGetPageRequest {
-                    hdr: PagestreamRequest {
-                        reqid,
-                        request_lsn,
-                        not_modified_since,
-                    },
-                    rel: RelTag {
-                        spcnode: body.read_u32::<BigEndian>()?,
-                        dbnode: body.read_u32::<BigEndian>()?,
-                        relnode: body.read_u32::<BigEndian>()?,
-                        forknum: body.read_u8()?,
-                    },
-                    blkno: body.read_u32::<BigEndian>()?,
-                }))
-            }
-            PagestreamFeMessageTag::DbSize => {
-                Ok(PagestreamFeMessage::DbSize(PagestreamDbSizeRequest {
-                    hdr: PagestreamRequest {
-                        reqid,
-                        request_lsn,
-                        not_modified_since,
-                    },
+        match msg_tag {
+            0 => Ok(PagestreamFeMessage::Exists(PagestreamExistsRequest {
+                hdr: PagestreamRequest {
+                    reqid,
+                    request_lsn,
+                    not_modified_since,
+                },
+                rel: RelTag {
+                    spcnode: body.read_u32::<BigEndian>()?,
                     dbnode: body.read_u32::<BigEndian>()?,
-                }))
-            }
-            PagestreamFeMessageTag::GetSlruSegment => Ok(PagestreamFeMessage::GetSlruSegment(
+                    relnode: body.read_u32::<BigEndian>()?,
+                    forknum: body.read_u8()?,
+                },
+            })),
+            1 => Ok(PagestreamFeMessage::Nblocks(PagestreamNblocksRequest {
+                hdr: PagestreamRequest {
+                    reqid,
+                    request_lsn,
+                    not_modified_since,
+                },
+                rel: RelTag {
+                    spcnode: body.read_u32::<BigEndian>()?,
+                    dbnode: body.read_u32::<BigEndian>()?,
+                    relnode: body.read_u32::<BigEndian>()?,
+                    forknum: body.read_u8()?,
+                },
+            })),
+            2 => Ok(PagestreamFeMessage::GetPage(PagestreamGetPageRequest {
+                hdr: PagestreamRequest {
+                    reqid,
+                    request_lsn,
+                    not_modified_since,
+                },
+                rel: RelTag {
+                    spcnode: body.read_u32::<BigEndian>()?,
+                    dbnode: body.read_u32::<BigEndian>()?,
+                    relnode: body.read_u32::<BigEndian>()?,
+                    forknum: body.read_u8()?,
+                },
+                blkno: body.read_u32::<BigEndian>()?,
+            })),
+            3 => Ok(PagestreamFeMessage::DbSize(PagestreamDbSizeRequest {
+                hdr: PagestreamRequest {
+                    reqid,
+                    request_lsn,
+                    not_modified_since,
+                },
+                dbnode: body.read_u32::<BigEndian>()?,
+            })),
+            4 => Ok(PagestreamFeMessage::GetSlruSegment(
                 PagestreamGetSlruSegmentRequest {
                     hdr: PagestreamRequest {
                         reqid,
@@ -1786,21 +1705,7 @@ impl PagestreamFeMessage {
                     segno: body.read_u32::<BigEndian>()?,
                 },
             )),
-            #[cfg(feature = "testing")]
-            PagestreamFeMessageTag::Test => Ok(PagestreamFeMessage::Test(PagestreamTestRequest {
-                hdr: PagestreamRequest {
-                    reqid,
-                    request_lsn,
-                    not_modified_since,
-                },
-                batch_key: body.read_u64::<BigEndian>()?,
-                message: {
-                    let len = body.read_u64::<BigEndian>()?;
-                    let mut buf = vec![0; len as usize];
-                    body.read_exact(&mut buf)?;
-                    String::from_utf8(buf)?
-                },
-            })),
+            _ => bail!("unknown smgr message tag: {:?}", msg_tag),
         }
     }
 }
@@ -1843,15 +1748,6 @@ impl PagestreamBeMessage {
                         bytes.put_u32((resp.segment.len() / BLCKSZ as usize) as u32);
                         bytes.put(&resp.segment[..]);
                     }
-
-                    #[cfg(feature = "testing")]
-                    Self::Test(resp) => {
-                        bytes.put_u8(Tag::Test as u8);
-                        bytes.put_u64(resp.req.batch_key);
-                        let message = resp.req.message.as_bytes();
-                        bytes.put_u64(message.len() as u64);
-                        bytes.put_slice(message);
-                    }
                 }
             }
             PagestreamProtocolVersion::V3 => {
@@ -1920,18 +1816,6 @@ impl PagestreamBeMessage {
                         bytes.put_u32((resp.segment.len() / BLCKSZ as usize) as u32);
                         bytes.put(&resp.segment[..]);
                     }
-
-                    #[cfg(feature = "testing")]
-                    Self::Test(resp) => {
-                        bytes.put_u8(Tag::Test as u8);
-                        bytes.put_u64(resp.req.hdr.reqid);
-                        bytes.put_u64(resp.req.hdr.request_lsn.0);
-                        bytes.put_u64(resp.req.hdr.not_modified_since.0);
-                        bytes.put_u64(resp.req.batch_key);
-                        let message = resp.req.message.as_bytes();
-                        bytes.put_u64(message.len() as u64);
-                        bytes.put_slice(message);
-                    }
                 }
             }
         }
@@ -2074,28 +1958,6 @@ impl PagestreamBeMessage {
                         segment: segment.into(),
                     })
                 }
-                #[cfg(feature = "testing")]
-                Tag::Test => {
-                    let reqid = buf.read_u64::<BigEndian>()?;
-                    let request_lsn = Lsn(buf.read_u64::<BigEndian>()?);
-                    let not_modified_since = Lsn(buf.read_u64::<BigEndian>()?);
-                    let batch_key = buf.read_u64::<BigEndian>()?;
-                    let len = buf.read_u64::<BigEndian>()?;
-                    let mut msg = vec![0; len as usize];
-                    buf.read_exact(&mut msg)?;
-                    let message = String::from_utf8(msg)?;
-                    Self::Test(PagestreamTestResponse {
-                        req: PagestreamTestRequest {
-                            hdr: PagestreamRequest {
-                                reqid,
-                                request_lsn,
-                                not_modified_since,
-                            },
-                            batch_key,
-                            message,
-                        },
-                    })
-                }
             };
         let remaining = buf.into_inner();
         if !remaining.is_empty() {
@@ -2115,8 +1977,6 @@ impl PagestreamBeMessage {
             Self::Error(_) => "Error",
             Self::DbSize(_) => "DbSize",
             Self::GetSlruSegment(_) => "GetSlruSegment",
-            #[cfg(feature = "testing")]
-            Self::Test(_) => "Test",
         }
     }
 }

libs/safekeeper_api/src/models.rs

@@ -277,8 +277,3 @@ pub struct TimelineTermBumpResponse {
     pub previous_term: u64,
     pub current_term: u64,
 }
-
-#[derive(Debug, Clone, Deserialize, Serialize)]
-pub struct SafekeeperUtilization {
-    pub timeline_count: u64,
-}

libs/utils/src/guard_arc_swap.rs

@@ -1,54 +0,0 @@
-//! A wrapper around `ArcSwap` that ensures there is only one writer at a time and writes
-//! don't block reads.
-
-use arc_swap::ArcSwap;
-use std::sync::Arc;
-use tokio::sync::TryLockError;
-
-pub struct GuardArcSwap<T> {
-    inner: ArcSwap<T>,
-    guard: tokio::sync::Mutex<()>,
-}
-
-pub struct Guard<'a, T> {
-    _guard: tokio::sync::MutexGuard<'a, ()>,
-    inner: &'a ArcSwap<T>,
-}
-
-impl<T> GuardArcSwap<T> {
-    pub fn new(inner: T) -> Self {
-        Self {
-            inner: ArcSwap::new(Arc::new(inner)),
-            guard: tokio::sync::Mutex::new(()),
-        }
-    }
-
-    pub fn read(&self) -> Arc<T> {
-        self.inner.load_full()
-    }
-
-    pub async fn write_guard(&self) -> Guard<'_, T> {
-        Guard {
-            _guard: self.guard.lock().await,
-            inner: &self.inner,
-        }
-    }
-
-    pub fn try_write_guard(&self) -> Result<Guard<'_, T>, TryLockError> {
-        let guard = self.guard.try_lock()?;
-        Ok(Guard {
-            _guard: guard,
-            inner: &self.inner,
-        })
-    }
-}
-
-impl<T> Guard<'_, T> {
-    pub fn read(&self) -> Arc<T> {
-        self.inner.load_full()
-    }
-
-    pub fn write(&mut self, value: T) {
-        self.inner.store(Arc::new(value));
-    }
-}

libs/utils/src/lib.rs

@@ -98,8 +98,6 @@ pub mod try_rcu;
 
 pub mod pprof;
 
-pub mod guard_arc_swap;
-
 // Re-export used in macro. Avoids adding git-version as dep in target crates.
 #[doc(hidden)]
 pub use git_version;

pageserver/Cargo.toml

@@ -8,7 +8,7 @@ license.workspace = true
 default = []
 # Enables test-only APIs, incuding failpoints. In particular, enables the `fail_point!` macro,
 # which adds some runtime cost to run tests on outage conditions
-testing = ["fail/failpoints", "pageserver_api/testing", "wal_decoder/testing", "pageserver_client/testing"]
+testing = ["fail/failpoints", "pageserver_api/testing", "wal_decoder/testing"]
 
 [dependencies]
 anyhow.workspace = true
@@ -114,7 +114,3 @@ harness = false
 [[bench]]
 name = "upload_queue"
 harness = false
-
-[[bin]]
-name = "test_helper_slow_client_reads"
-required-features = [ "testing" ]

pageserver/client/Cargo.toml

@@ -4,9 +4,6 @@ version = "0.1.0"
 edition.workspace = true
 license.workspace = true
 
-[features]
-testing = [ "pageserver_api/testing" ]
-
 [dependencies]
 pageserver_api.workspace = true
 thiserror.workspace = true

pageserver/client/src/page_service.rs

@@ -1,9 +1,6 @@
-use std::sync::{Arc, Mutex};
+use std::pin::Pin;
 
-use futures::{
-    stream::{SplitSink, SplitStream},
-    SinkExt, StreamExt,
-};
+use futures::SinkExt;
 use pageserver_api::{
     models::{
         PagestreamBeMessage, PagestreamFeMessage, PagestreamGetPageRequest,
@@ -13,6 +10,7 @@ use pageserver_api::{
 };
 use tokio::task::JoinHandle;
 use tokio_postgres::CopyOutStream;
+use tokio_stream::StreamExt;
 use tokio_util::sync::CancellationToken;
 use utils::{
     id::{TenantId, TimelineId},
@@ -64,28 +62,15 @@ impl Client {
             .client
             .copy_both_simple(&format!("pagestream_v3 {tenant_id} {timeline_id}"))
             .await?;
-        let (sink, stream) = copy_both.split(); // TODO: actually support splitting of the CopyBothDuplex so the lock inside this split adaptor goes away.
         let Client {
             cancel_on_client_drop,
             conn_task,
             client: _,
         } = self;
-        let shared = Arc::new(Mutex::new(PagestreamShared::ConnTaskRunning(
-            ConnTaskRunning {
-                cancel_on_client_drop,
-                conn_task,
-            },
-        )));
         Ok(PagestreamClient {
-            sink: PagestreamSender {
-                shared: shared.clone(),
-                sink,
-            },
-            stream: PagestreamReceiver {
-                shared: shared.clone(),
-                stream,
-            },
-            shared,
+            copy_both: Box::pin(copy_both),
+            conn_task,
+            cancel_on_client_drop,
         })
     }
 
@@ -112,28 +97,7 @@ impl Client {
 
 /// Create using [`Client::pagestream`].
 pub struct PagestreamClient {
-    shared: Arc<Mutex<PagestreamShared>>,
-    sink: PagestreamSender,
-    stream: PagestreamReceiver,
-}
-
-pub struct PagestreamSender {
-    #[allow(dead_code)]
-    shared: Arc<Mutex<PagestreamShared>>,
-    sink: SplitSink<tokio_postgres::CopyBothDuplex<bytes::Bytes>, bytes::Bytes>,
-}
-
-pub struct PagestreamReceiver {
-    #[allow(dead_code)]
-    shared: Arc<Mutex<PagestreamShared>>,
-    stream: SplitStream<tokio_postgres::CopyBothDuplex<bytes::Bytes>>,
-}
-
-enum PagestreamShared {
-    ConnTaskRunning(ConnTaskRunning),
-    ConnTaskCancelledJoinHandleReturnedOrDropped,
-}
-struct ConnTaskRunning {
+    copy_both: Pin<Box<tokio_postgres::CopyBothDuplex<bytes::Bytes>>>,
     cancel_on_client_drop: Option<tokio_util::sync::DropGuard>,
     conn_task: JoinHandle<()>,
 }
@@ -146,11 +110,11 @@ pub struct RelTagBlockNo {
 impl PagestreamClient {
     pub async fn shutdown(self) {
         let Self {
-            shared,
-            sink,
-            stream,
-        } = { self };
-        // The `copy_both` split into `sink` and `stream` contains internal channel sender, the receiver of which is polled by `conn_task`.
+            copy_both,
+            cancel_on_client_drop: cancel_conn_task,
+            conn_task,
+        } = self;
+        // The `copy_both` contains internal channel sender, the receiver of which is polled by `conn_task`.
         // When `conn_task` observes the sender has been dropped, it sends a `FeMessage::CopyFail` into the connection.
         // (see https://github.com/neondatabase/rust-postgres/blob/2005bf79573b8add5cf205b52a2b208e356cc8b0/tokio-postgres/src/copy_both.rs#L56).
         //
@@ -167,77 +131,27 @@ impl PagestreamClient {
         //
         // NB: page_service doesn't have a use case to exit the `pagestream` mode currently.
         // => https://github.com/neondatabase/neon/issues/6390
-        let ConnTaskRunning {
-            cancel_on_client_drop,
-            conn_task,
-        } = {
-            let mut guard = shared.lock().unwrap();
-            match std::mem::replace(
-                &mut *guard,
-                PagestreamShared::ConnTaskCancelledJoinHandleReturnedOrDropped,
-            ) {
-                PagestreamShared::ConnTaskRunning(conn_task_running) => conn_task_running,
-                PagestreamShared::ConnTaskCancelledJoinHandleReturnedOrDropped => unreachable!(),
-            }
-        };
-        let _ = cancel_on_client_drop.unwrap();
+        let _ = cancel_conn_task.unwrap();
         conn_task.await.unwrap();
-
-        // Now drop the split copy_both.
-        drop(sink);
-        drop(stream);
-    }
-
-    pub fn split(self) -> (PagestreamSender, PagestreamReceiver) {
-        let Self {
-            shared: _,
-            sink,
-            stream,
-        } = self;
-        (sink, stream)
+        drop(copy_both);
     }
 
     pub async fn getpage(
         &mut self,
         req: PagestreamGetPageRequest,
     ) -> anyhow::Result<PagestreamGetPageResponse> {
-        self.getpage_send(req).await?;
-        self.getpage_recv().await
-    }
+        let req = PagestreamFeMessage::GetPage(req);
+        let req: bytes::Bytes = req.serialize();
+        // let mut req = tokio_util::io::ReaderStream::new(&req);
+        let mut req = tokio_stream::once(Ok(req));
 
-    pub async fn getpage_send(&mut self, req: PagestreamGetPageRequest) -> anyhow::Result<()> {
-        self.sink.getpage_send(req).await
-    }
+        self.copy_both.send_all(&mut req).await?;
 
-    pub async fn getpage_recv(&mut self) -> anyhow::Result<PagestreamGetPageResponse> {
-        self.stream.getpage_recv().await
-    }
-}
-
-impl PagestreamSender {
-    // TODO: maybe make this impl Sink instead for better composability?
-    pub async fn send(&mut self, msg: PagestreamFeMessage) -> anyhow::Result<()> {
-        let msg = msg.serialize();
-        self.sink.send_all(&mut tokio_stream::once(Ok(msg))).await?;
-        Ok(())
-    }
-
-    pub async fn getpage_send(&mut self, req: PagestreamGetPageRequest) -> anyhow::Result<()> {
-        self.send(PagestreamFeMessage::GetPage(req)).await
-    }
-}
-
-impl PagestreamReceiver {
-    // TODO: maybe make this impl Stream instead for better composability?
-    pub async fn recv(&mut self) -> anyhow::Result<PagestreamBeMessage> {
-        let next: Option<Result<bytes::Bytes, _>> = self.stream.next().await;
+        let next: Option<Result<bytes::Bytes, _>> = self.copy_both.next().await;
         let next: bytes::Bytes = next.unwrap()?;
-        PagestreamBeMessage::deserialize(next)
-    }
 
-    pub async fn getpage_recv(&mut self) -> anyhow::Result<PagestreamGetPageResponse> {
-        let next: PagestreamBeMessage = self.recv().await?;
-        match next {
+        let msg = PagestreamBeMessage::deserialize(next)?;
+        match msg {
             PagestreamBeMessage::GetPage(p) => Ok(p),
             PagestreamBeMessage::Error(e) => anyhow::bail!("Error: {:?}", e),
             PagestreamBeMessage::Exists(_)
@@ -246,14 +160,7 @@ impl PagestreamReceiver {
             | PagestreamBeMessage::GetSlruSegment(_) => {
                 anyhow::bail!(
                     "unexpected be message kind in response to getpage request: {}",
-                    next.kind()
-                )
-            }
-            #[cfg(feature = "testing")]
-            PagestreamBeMessage::Test(_) => {
-                anyhow::bail!(
-                    "unexpected be message kind in response to getpage request: {}",
-                    next.kind()
+                    msg.kind()
                 )
             }
         }

pageserver/src/bin/test_helper_slow_client_reads.rs

@@ -1,65 +0,0 @@
-use std::{
-    io::{stdin, stdout, Read, Write},
-    time::Duration,
-};
-
-use clap::Parser;
-use pageserver_api::models::{PagestreamRequest, PagestreamTestRequest};
-use utils::{
-    id::{TenantId, TimelineId},
-    lsn::Lsn,
-};
-
-#[derive(clap::Parser)]
-struct Args {
-    connstr: String,
-    tenant_id: TenantId,
-    timeline_id: TimelineId,
-}
-
-#[tokio::main]
-async fn main() -> anyhow::Result<()> {
-    let Args {
-        connstr,
-        tenant_id,
-        timeline_id,
-    } = Args::parse();
-    let client = pageserver_client::page_service::Client::new(connstr).await?;
-    let client = client.pagestream(tenant_id, timeline_id).await?;
-    let (mut sender, _receiver) = client.split();
-
-    eprintln!("filling the pipe");
-    let mut msg = 0;
-    loop {
-        msg += 1;
-        let fut = sender.send(pageserver_api::models::PagestreamFeMessage::Test(
-            PagestreamTestRequest {
-                hdr: PagestreamRequest {
-                    reqid: 0,
-                    request_lsn: Lsn(23),
-                    not_modified_since: Lsn(23),
-                },
-                batch_key: 42,
-                message: format!("message {}", msg),
-            },
-        ));
-        let Ok(res) = tokio::time::timeout(Duration::from_secs(10), fut).await else {
-            eprintln!("pipe seems full");
-            break;
-        };
-        let _: () = res?;
-    }
-
-    let n = stdout().write(b"R")?;
-    assert_eq!(n, 1);
-    stdout().flush()?;
-
-    eprintln!("waiting for signal to tell us to exit");
-
-    let mut buf = [0u8; 1];
-    stdin().read_exact(&mut buf)?;
-
-    eprintln!("termination signal received, exiting");
-
-    anyhow::Ok(())
-}

pageserver/src/metrics.rs

@@ -100,32 +100,6 @@ pub(crate) static VEC_READ_NUM_LAYERS_VISITED: Lazy<Histogram> = Lazy::new(|| {
     .expect("failed to define a metric")
 });
 
-pub(crate) static CONCURRENT_INITDBS: Lazy<UIntGauge> = Lazy::new(|| {
-    register_uint_gauge!(
-        "pageserver_concurrent_initdb",
-        "Number of initdb processes running"
-    )
-    .expect("failed to define a metric")
-});
-
-pub(crate) static INITDB_SEMAPHORE_ACQUISITION_TIME: Lazy<Histogram> = Lazy::new(|| {
-    register_histogram!(
-        "pageserver_initdb_semaphore_seconds_global",
-        "Time spent getting a permit from the global initdb semaphore",
-        STORAGE_OP_BUCKETS.into()
-    )
-    .expect("failed to define metric")
-});
-
-pub(crate) static INITDB_RUN_TIME: Lazy<Histogram> = Lazy::new(|| {
-    register_histogram!(
-        "pageserver_initdb_seconds_global",
-        "Time spent performing initdb",
-        STORAGE_OP_BUCKETS.into()
-    )
-    .expect("failed to define metric")
-});
-
 // Metrics collected on operations on the storage repository.
 #[derive(
     Clone, Copy, enum_map::Enum, strum_macros::EnumString, strum_macros::Display, IntoStaticStr,
@@ -1489,8 +1463,6 @@ pub enum SmgrQueryType {
     GetPageAtLsn,
     GetDbSize,
     GetSlruSegment,
-    #[cfg(feature = "testing")]
-    Test,
 }
 
 pub(crate) struct SmgrQueryTimePerTimeline {

pageserver/src/page_service.rs

@@ -555,52 +555,37 @@ struct BatchedGetPageRequest {
     timer: SmgrOpTimer,
 }
 
-#[cfg(feature = "testing")]
-struct BatchedTestRequest {
-    req: models::PagestreamTestRequest,
-    timer: SmgrOpTimer,
-}
-
-/// NB: we only hold [`timeline::handle::WeakHandle`] inside this enum,
-/// so that we don't keep the [`Timeline::gate`] open while the batch
-/// is being built up inside the [`spsc_fold`] (pagestream pipelining).
 enum BatchedFeMessage {
     Exists {
         span: Span,
         timer: SmgrOpTimer,
-        shard: timeline::handle::WeakHandle<TenantManagerTypes>,
+        shard: timeline::handle::Handle<TenantManagerTypes>,
         req: models::PagestreamExistsRequest,
     },
     Nblocks {
         span: Span,
         timer: SmgrOpTimer,
-        shard: timeline::handle::WeakHandle<TenantManagerTypes>,
+        shard: timeline::handle::Handle<TenantManagerTypes>,
         req: models::PagestreamNblocksRequest,
     },
     GetPage {
         span: Span,
-        shard: timeline::handle::WeakHandle<TenantManagerTypes>,
+        shard: timeline::handle::Handle<TenantManagerTypes>,
         effective_request_lsn: Lsn,
         pages: smallvec::SmallVec<[BatchedGetPageRequest; 1]>,
     },
     DbSize {
         span: Span,
         timer: SmgrOpTimer,
-        shard: timeline::handle::WeakHandle<TenantManagerTypes>,
+        shard: timeline::handle::Handle<TenantManagerTypes>,
         req: models::PagestreamDbSizeRequest,
     },
     GetSlruSegment {
         span: Span,
         timer: SmgrOpTimer,
-        shard: timeline::handle::WeakHandle<TenantManagerTypes>,
+        shard: timeline::handle::Handle<TenantManagerTypes>,
         req: models::PagestreamGetSlruSegmentRequest,
     },
-    #[cfg(feature = "testing")]
-    Test {
-        span: Span,
-        shard: timeline::handle::WeakHandle<TenantManagerTypes>,
-        requests: Vec<BatchedTestRequest>,
-    },
     RespondError {
         span: Span,
         error: BatchedPageStreamError,
@@ -621,12 +606,6 @@ impl BatchedFeMessage {
                     page.timer.observe_execution_start(at);
                 }
             }
-            #[cfg(feature = "testing")]
-            BatchedFeMessage::Test { requests, .. } => {
-                for req in requests {
-                    req.timer.observe_execution_start(at);
-                }
-            }
             BatchedFeMessage::RespondError { .. } => {}
         }
     }
@@ -756,7 +735,7 @@ impl PageServerHandler {
                 BatchedFeMessage::Exists {
                     span,
                     timer,
-                    shard: shard.downgrade(),
+                    shard,
                     req,
                 }
             }
@@ -775,7 +754,7 @@ impl PageServerHandler {
                 BatchedFeMessage::Nblocks {
                     span,
                     timer,
-                    shard: shard.downgrade(),
+                    shard,
                     req,
                 }
             }
@@ -794,7 +773,7 @@ impl PageServerHandler {
                 BatchedFeMessage::DbSize {
                     span,
                     timer,
-                    shard: shard.downgrade(),
+                    shard,
                     req,
                 }
             }
@@ -813,7 +792,7 @@ impl PageServerHandler {
                 BatchedFeMessage::GetSlruSegment {
                     span,
                     timer,
-                    shard: shard.downgrade(),
+                    shard,
                     req,
                 }
             }
@@ -865,7 +844,6 @@ impl PageServerHandler {
                 )
                 .await?;
 
-                // We're holding the Handle
                 let effective_request_lsn = match Self::wait_or_get_last_lsn(
                     &shard,
                     req.hdr.request_lsn,
@@ -883,27 +861,11 @@ impl PageServerHandler {
                 };
                 BatchedFeMessage::GetPage {
                     span,
-                    shard: shard.downgrade(),
+                    shard,
                     effective_request_lsn,
                     pages: smallvec::smallvec![BatchedGetPageRequest { req, timer }],
                 }
             }
-            #[cfg(feature = "testing")]
-            PagestreamFeMessage::Test(req) => {
-                let span = tracing::info_span!(parent: parent_span, "handle_test_request");
-                let shard = timeline_handles
-                    .get(tenant_id, timeline_id, ShardSelector::Zero)
-                    .instrument(span.clone()) // sets `shard_id` field
-                    .await?;
-                let timer =
-                    record_op_start_and_throttle(&shard, metrics::SmgrQueryType::Test, received_at)
-                        .await?;
-                BatchedFeMessage::Test {
-                    span,
-                    shard: shard.downgrade(),
-                    requests: vec![BatchedTestRequest { req, timer }],
-                }
-            }
         };
         Ok(Some(batched_msg))
     }
@@ -945,7 +907,9 @@ impl PageServerHandler {
                     assert_eq!(accum_pages.len(), max_batch_size.get());
                     return false;
                 }
-                if !accum_shard.is_same_handle_as(&this_shard) {
+                if (accum_shard.tenant_shard_id, accum_shard.timeline_id)
+                    != (this_shard.tenant_shard_id, this_shard.timeline_id)
+                {
                     trace!(%accum_lsn, %this_lsn, "stopping batching because timeline object mismatch");
                     // TODO: we _could_ batch & execute each shard seperately (and in parallel).
                     // But the current logic for keeping responses in order does not support that.
@@ -964,44 +928,6 @@ impl PageServerHandler {
                 accum_pages.extend(this_pages);
                 Ok(())
             }
-            #[cfg(feature = "testing")]
-            (
-                Ok(BatchedFeMessage::Test {
-                    shard: accum_shard,
-                    requests: accum_requests,
-                    ..
-                }),
-                BatchedFeMessage::Test {
-                    shard: this_shard,
-                    requests: this_requests,
-                    ..
-                },
-            ) if (|| {
-                assert!(this_requests.len() == 1);
-                if accum_requests.len() >= max_batch_size.get() {
-                    trace!(%max_batch_size, "stopping batching because of batch size");
-                    assert_eq!(accum_requests.len(), max_batch_size.get());
-                    return false;
-                }
-                if !accum_shard.is_same_handle_as(&this_shard) {
-                    trace!("stopping batching because timeline object mismatch");
-                    // TODO: we _could_ batch & execute each shard seperately (and in parallel).
-                    // But the current logic for keeping responses in order does not support that.
-                    return false;
-                }
-                let this_batch_key = this_requests[0].req.batch_key;
-                let accum_batch_key = accum_requests[0].req.batch_key;
-                if this_requests[0].req.batch_key != accum_requests[0].req.batch_key {
-                    trace!(%accum_batch_key, %this_batch_key, "stopping batching because batch key changed");
-                    return false;
-                }
-                true
-            })() =>
-            {
-                // ok to batch
-                accum_requests.extend(this_requests);
-                Ok(())
-            }
             // something batched already but this message is unbatchable
             (_, this_msg) => {
                 // by default, don't continue batching
@@ -1043,7 +969,7 @@ impl PageServerHandler {
                 fail::fail_point!("ps::handle-pagerequest-message::exists");
                 (
                     vec![self
-                        .handle_get_rel_exists_request(&*shard.upgrade()?, &req, ctx)
+                        .handle_get_rel_exists_request(&shard, &req, ctx)
                         .instrument(span.clone())
                         .await
                         .map(|msg| (msg, timer))
@@ -1060,7 +986,7 @@ impl PageServerHandler {
                 fail::fail_point!("ps::handle-pagerequest-message::nblocks");
                 (
                     vec![self
-                        .handle_get_nblocks_request(&*shard.upgrade()?, &req, ctx)
+                        .handle_get_nblocks_request(&shard, &req, ctx)
                         .instrument(span.clone())
                         .await
                         .map(|msg| (msg, timer))
@@ -1081,7 +1007,7 @@ impl PageServerHandler {
                         trace!(npages, "handling getpage request");
                         let res = self
                             .handle_get_page_at_lsn_request_batched(
-                                &*shard.upgrade()?,
+                                &shard,
                                 effective_request_lsn,
                                 pages,
                                 ctx,
@@ -1103,7 +1029,7 @@ impl PageServerHandler {
                 fail::fail_point!("ps::handle-pagerequest-message::dbsize");
                 (
                     vec![self
-                        .handle_db_size_request(&*shard.upgrade()?, &req, ctx)
+                        .handle_db_size_request(&shard, &req, ctx)
                         .instrument(span.clone())
                         .await
                         .map(|msg| (msg, timer))
@@ -1120,7 +1046,7 @@ impl PageServerHandler {
                 fail::fail_point!("ps::handle-pagerequest-message::slrusegment");
                 (
                     vec![self
-                        .handle_get_slru_segment_request(&*shard.upgrade()?, &req, ctx)
+                        .handle_get_slru_segment_request(&shard, &req, ctx)
                         .instrument(span.clone())
                         .await
                         .map(|msg| (msg, timer))
@@ -1128,27 +1054,6 @@ impl PageServerHandler {
                     span,
                 )
             }
-            #[cfg(feature = "testing")]
-            BatchedFeMessage::Test {
-                span,
-                shard,
-                requests,
-            } => {
-                fail::fail_point!("ps::handle-pagerequest-message::test");
-                (
-                    {
-                        let npages = requests.len();
-                        trace!(npages, "handling getpage request");
-                        let res = self
-                            .handle_test_request_batch(&*shard.upgrade()?, requests, ctx)
-                            .instrument(span.clone())
-                            .await;
-                        assert_eq!(res.len(), npages);
-                        res
-                    },
-                    span,
-                )
-            }
             BatchedFeMessage::RespondError { span, error } => {
                 // We've already decided to respond with an error, so we don't need to
                 // call the handler.
@@ -1886,51 +1791,6 @@ impl PageServerHandler {
         ))
     }
 
-    // NB: this impl mimics what we do for batched getpage requests.
-    #[cfg(feature = "testing")]
-    #[instrument(skip_all, fields(shard_id))]
-    async fn handle_test_request_batch(
-        &mut self,
-        timeline: &Timeline,
-        requests: Vec<BatchedTestRequest>,
-        _ctx: &RequestContext,
-    ) -> Vec<Result<(PagestreamBeMessage, SmgrOpTimer), BatchedPageStreamError>> {
-        // real requests would do something with the timeline
-        let mut results = Vec::with_capacity(requests.len());
-        for _req in requests.iter() {
-            tokio::task::yield_now().await;
-
-            results.push({
-                if timeline.cancel.is_cancelled() {
-                    Err(PageReconstructError::Cancelled)
-                } else {
-                    Ok(())
-                }
-            });
-        }
-
-        // TODO: avoid creating the new Vec here
-        Vec::from_iter(
-            requests
-                .into_iter()
-                .zip(results.into_iter())
-                .map(|(req, res)| {
-                    res.map(|()| {
-                        (
-                            PagestreamBeMessage::Test(models::PagestreamTestResponse {
-                                req: req.req.clone(),
-                            }),
-                            req.timer,
-                        )
-                    })
-                    .map_err(|e| BatchedPageStreamError {
-                        err: PageStreamError::from(e),
-                        req: req.req.hdr,
-                    })
-                }),
-        )
-    }
-
     /// Note on "fullbackup":
     /// Full basebackups should only be used for debugging purposes.
     /// Originally, it was introduced to enable breaking storage format changes,
@@ -2546,14 +2406,6 @@ impl From<GetActiveTimelineError> for QueryError {
     }
 }
 
-impl From<crate::tenant::timeline::handle::HandleUpgradeError> for QueryError {
-    fn from(e: crate::tenant::timeline::handle::HandleUpgradeError) -> Self {
-        match e {
-            crate::tenant::timeline::handle::HandleUpgradeError::ShutDown => QueryError::Shutdown,
-        }
-    }
-}
-
 fn set_tracing_field_shard_id(timeline: &Timeline) {
     debug_assert_current_span_has_tenant_and_timeline_id_no_shard_id();
     tracing::Span::current().record(

pageserver/src/tenant.rs

@@ -95,9 +95,6 @@ use crate::deletion_queue::DeletionQueueError;
 use crate::import_datadir;
 use crate::is_uninit_mark;
 use crate::l0_flush::L0FlushGlobalState;
-use crate::metrics::CONCURRENT_INITDBS;
-use crate::metrics::INITDB_RUN_TIME;
-use crate::metrics::INITDB_SEMAPHORE_ACQUISITION_TIME;
 use crate::metrics::TENANT;
 use crate::metrics::{
     remove_tenant_metrics, BROKEN_TENANTS_SET, CIRCUIT_BREAKERS_BROKEN, CIRCUIT_BREAKERS_UNBROKEN,
@@ -5350,17 +5347,8 @@ async fn run_initdb(
         initdb_bin_path, initdb_target_dir, initdb_lib_dir,
     );
 
-    let _permit = {
-        let _timer = INITDB_SEMAPHORE_ACQUISITION_TIME.start_timer();
-        INIT_DB_SEMAPHORE.acquire().await
-    };
+    let _permit = INIT_DB_SEMAPHORE.acquire().await;
 
-    CONCURRENT_INITDBS.inc();
-    scopeguard::defer! {
-        CONCURRENT_INITDBS.dec();
-    }
-
-    let _timer = INITDB_RUN_TIME.start_timer();
     let res = postgres_initdb::do_run_initdb(postgres_initdb::RunInitdbArgs {
         superuser: &conf.superuser,
         locale: &conf.locale,
@@ -5475,7 +5463,6 @@ pub(crate) mod harness {
                 lsn_lease_length_for_ts: Some(tenant_conf.lsn_lease_length_for_ts),
                 timeline_offloading: Some(tenant_conf.timeline_offloading),
                 wal_receiver_protocol_override: tenant_conf.wal_receiver_protocol_override,
-                rel_size_v2_enabled: tenant_conf.rel_size_v2_enabled,
             }
         }
     }

pageserver/src/tenant/config.rs

@@ -357,9 +357,6 @@ pub struct TenantConfOpt {
 
     #[serde(skip_serializing_if = "Option::is_none")]
     pub wal_receiver_protocol_override: Option<PostgresClientProtocol>,
-
-    #[serde(skip_serializing_if = "Option::is_none")]
-    pub rel_size_v2_enabled: Option<bool>,
 }
 
 impl TenantConfOpt {
@@ -428,7 +425,6 @@ impl TenantConfOpt {
             wal_receiver_protocol_override: self
                 .wal_receiver_protocol_override
                 .or(global_conf.wal_receiver_protocol_override),
-            rel_size_v2_enabled: self.rel_size_v2_enabled.or(global_conf.rel_size_v2_enabled),
         }
     }
 
@@ -458,7 +454,6 @@ impl TenantConfOpt {
             mut lsn_lease_length_for_ts,
             mut timeline_offloading,
             mut wal_receiver_protocol_override,
-            mut rel_size_v2_enabled,
         } = self;
 
         patch.checkpoint_distance.apply(&mut checkpoint_distance);
@@ -527,7 +522,6 @@ impl TenantConfOpt {
         patch
             .wal_receiver_protocol_override
             .apply(&mut wal_receiver_protocol_override);
-        patch.rel_size_v2_enabled.apply(&mut rel_size_v2_enabled);
 
         Ok(Self {
             checkpoint_distance,
@@ -554,7 +548,6 @@ impl TenantConfOpt {
             lsn_lease_length_for_ts,
             timeline_offloading,
             wal_receiver_protocol_override,
-            rel_size_v2_enabled,
         })
     }
 }
@@ -610,7 +603,6 @@ impl From<TenantConfOpt> for models::TenantConfig {
             lsn_lease_length_for_ts: value.lsn_lease_length_for_ts.map(humantime),
             timeline_offloading: value.timeline_offloading,
             wal_receiver_protocol_override: value.wal_receiver_protocol_override,
-            rel_size_v2_enabled: value.rel_size_v2_enabled,
         }
     }
 }

pageserver/src/tenant/remote_timeline_client.rs

@@ -382,12 +382,6 @@ pub(crate) struct RemoteTimelineClient {
     cancel: CancellationToken,
 }
 
-impl Drop for RemoteTimelineClient {
-    fn drop(&mut self) {
-        debug!("dropping RemoteTimelineClient");
-    }
-}
-
 impl RemoteTimelineClient {
     ///
     /// Create a remote storage client for given timeline
@@ -803,12 +797,6 @@ impl RemoteTimelineClient {
 
         upload_queue.dirty.metadata.apply(update);
 
-        // Defense in depth: if we somehow generated invalid metadata, do not persist it.
-        upload_queue
-            .dirty
-            .validate()
-            .map_err(|e| anyhow::anyhow!(e))?;
-
         self.schedule_index_upload(upload_queue);
 
         Ok(())

pageserver/src/tenant/remote_timeline_client/index.rs

@@ -79,24 +79,6 @@ pub struct IndexPart {
     /// when this flag is introduced.
     #[serde(skip_serializing_if = "Option::is_none", default)]
     pub(crate) last_aux_file_policy: Option<AuxFilePolicy>,
-
-    #[serde(skip_serializing_if = "Option::is_none", default)]
-    pub(crate) rel_size_migration: Option<RelSizeMigration>,
-}
-
-#[derive(Debug, Clone, PartialEq, Eq, Serialize, Deserialize)]
-#[serde(rename_all = "camelCase")]
-pub enum RelSizeMigration {
-    /// The tenant is using the old rel_size format.
-    /// Note that this enum is persisted as `Option<RelSizeMigration>` in the index part, so
-    /// `None` is the same as `Some(RelSizeMigration::Legacy)`.
-    Legacy,
-    /// The tenant is migrating to the new rel_size format. Both old and new rel_size format are
-    /// persisted in the index part. The read path will read both formats and merge them.
-    Migrating,
-    /// The tenant has migrated to the new rel_size format. Only the new rel_size format is persisted
-    /// in the index part, and the read path will not read the old format.
-    Migrated,
 }
 
 impl IndexPart {
@@ -115,11 +97,10 @@ impl IndexPart {
     /// - 8: added `archived_at`
     /// - 9: +gc_blocking
     /// - 10: +import_pgdata
-    /// - 11: +rel_size_migration
-    const LATEST_VERSION: usize = 11;
+    const LATEST_VERSION: usize = 10;
 
     // Versions we may see when reading from a bucket.
-    pub const KNOWN_VERSIONS: &'static [usize] = &[1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11];
+    pub const KNOWN_VERSIONS: &'static [usize] = &[1, 2, 3, 4, 5, 6, 7, 8, 9, 10];
 
     pub const FILE_NAME: &'static str = "index_part.json";
 
@@ -135,7 +116,6 @@ impl IndexPart {
             gc_blocking: None,
             last_aux_file_policy: None,
             import_pgdata: None,
-            rel_size_migration: None,
         }
     }
 
@@ -172,21 +152,6 @@ impl IndexPart {
         };
         is_same_remote_layer_path(name, metadata, name, index_metadata)
     }
-
-    /// Check for invariants in the index: this is useful when uploading an index to ensure that if
-    /// we encounter a bug, we do not persist buggy metadata.
-    pub(crate) fn validate(&self) -> Result<(), String> {
-        if self.import_pgdata.is_none()
-            && self.metadata.ancestor_timeline().is_none()
-            && self.layer_metadata.is_empty()
-        {
-            // Unless we're in the middle of a raw pgdata import, or this is a child timeline,the index must
-            // always have at least one layer.
-            return Err("Index has no ancestor and no layers".to_string());
-        }
-
-        Ok(())
-    }
 }
 
 /// Metadata gathered for each of the layer files.
@@ -436,7 +401,6 @@ mod tests {
             gc_blocking: None,
             last_aux_file_policy: None,
             import_pgdata: None,
-            rel_size_migration: None,
         };
 
         let part = IndexPart::from_json_bytes(example.as_bytes()).unwrap();
@@ -482,7 +446,6 @@ mod tests {
             gc_blocking: None,
             last_aux_file_policy: None,
             import_pgdata: None,
-            rel_size_migration: None,
         };
 
         let part = IndexPart::from_json_bytes(example.as_bytes()).unwrap();
@@ -529,7 +492,6 @@ mod tests {
             gc_blocking: None,
             last_aux_file_policy: None,
             import_pgdata: None,
-            rel_size_migration: None,
         };
 
         let part = IndexPart::from_json_bytes(example.as_bytes()).unwrap();
@@ -579,7 +541,6 @@ mod tests {
             gc_blocking: None,
             last_aux_file_policy: None,
             import_pgdata: None,
-            rel_size_migration: None,
         };
 
         let empty_layers_parsed = IndexPart::from_json_bytes(empty_layers_json.as_bytes()).unwrap();
@@ -624,7 +585,6 @@ mod tests {
             gc_blocking: None,
             last_aux_file_policy: None,
             import_pgdata: None,
-            rel_size_migration: None,
         };
 
         let part = IndexPart::from_json_bytes(example.as_bytes()).unwrap();
@@ -672,7 +632,6 @@ mod tests {
             gc_blocking: None,
             last_aux_file_policy: None,
             import_pgdata: None,
-            rel_size_migration: None,
         };
 
         let part = IndexPart::from_json_bytes(example.as_bytes()).unwrap();
@@ -725,7 +684,6 @@ mod tests {
             gc_blocking: None,
             last_aux_file_policy: Some(AuxFilePolicy::V2),
             import_pgdata: None,
-            rel_size_migration: None,
         };
 
         let part = IndexPart::from_json_bytes(example.as_bytes()).unwrap();
@@ -783,7 +741,6 @@ mod tests {
             gc_blocking: None,
             last_aux_file_policy: Default::default(),
             import_pgdata: None,
-            rel_size_migration: None,
         };
 
         let part = IndexPart::from_json_bytes(example.as_bytes()).unwrap();
@@ -842,7 +799,6 @@ mod tests {
             gc_blocking: None,
             last_aux_file_policy: Default::default(),
             import_pgdata: None,
-            rel_size_migration: None,
         };
 
         let part = IndexPart::from_json_bytes(example.as_bytes()).unwrap();
@@ -906,7 +862,6 @@ mod tests {
             last_aux_file_policy: Default::default(),
             archived_at: None,
             import_pgdata: None,
-            rel_size_migration: None,
         };
 
         let part = IndexPart::from_json_bytes(example.as_bytes()).unwrap();
@@ -982,86 +937,7 @@ mod tests {
                 started_at: parse_naive_datetime("2024-11-13T09:23:42.123000000"),
                 finished_at: parse_naive_datetime("2024-11-13T09:42:23.123000000"),
                 idempotency_key: import_pgdata::index_part_format::IdempotencyKey::new("specified-by-client-218a5213-5044-4562-a28d-d024c5f057f5".to_string()),
-            }))),
-            rel_size_migration: None,
-        };
-
-        let part = IndexPart::from_json_bytes(example.as_bytes()).unwrap();
-        assert_eq!(part, expected);
-    }
-
-    #[test]
-    fn v11_rel_size_migration_is_parsed() {
-        let example = r#"{
-            "version": 11,
-            "layer_metadata":{
-                "000000000000000000000000000000000000-FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF__0000000001696070-00000000016960E9": { "file_size": 25600000 },
-                "000000000000000000000000000000000000-FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF__00000000016B59D8-00000000016B5A51": { "file_size": 9007199254741001 }
-            },
-            "disk_consistent_lsn":"0/16960E8",
-            "metadata": {
-                "disk_consistent_lsn": "0/16960E8",
-                "prev_record_lsn": "0/1696070",
-                "ancestor_timeline": "e45a7f37d3ee2ff17dc14bf4f4e3f52e",
-                "ancestor_lsn": "0/0",
-                "latest_gc_cutoff_lsn": "0/1696070",
-                "initdb_lsn": "0/1696070",
-                "pg_version": 14
-            },
-            "gc_blocking": {
-                "started_at": "2024-07-19T09:00:00.123",
-                "reasons": ["DetachAncestor"]
-            },
-            "import_pgdata": {
-                "V1": {
-                    "Done": {
-                        "idempotency_key": "specified-by-client-218a5213-5044-4562-a28d-d024c5f057f5",
-                        "started_at": "2024-11-13T09:23:42.123",
-                        "finished_at": "2024-11-13T09:42:23.123"
-                    }
-                }
-            },
-            "rel_size_migration": "legacy"
-        }"#;
-
-        let expected = IndexPart {
-            version: 11,
-            layer_metadata: HashMap::from([
-                ("000000000000000000000000000000000000-FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF__0000000001696070-00000000016960E9".parse().unwrap(), LayerFileMetadata {
-                    file_size: 25600000,
-                    generation: Generation::none(),
-                    shard: ShardIndex::unsharded()
-                }),
-                ("000000000000000000000000000000000000-FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF__00000000016B59D8-00000000016B5A51".parse().unwrap(), LayerFileMetadata {
-                    file_size: 9007199254741001,
-                    generation: Generation::none(),
-                    shard: ShardIndex::unsharded()
-                })
-            ]),
-            disk_consistent_lsn: "0/16960E8".parse::<Lsn>().unwrap(),
-            metadata: TimelineMetadata::new(
-                Lsn::from_str("0/16960E8").unwrap(),
-                Some(Lsn::from_str("0/1696070").unwrap()),
-                Some(TimelineId::from_str("e45a7f37d3ee2ff17dc14bf4f4e3f52e").unwrap()),
-                Lsn::INVALID,
-                Lsn::from_str("0/1696070").unwrap(),
-                Lsn::from_str("0/1696070").unwrap(),
-                14,
-            ).with_recalculated_checksum().unwrap(),
-            deleted_at: None,
-            lineage: Default::default(),
-            gc_blocking: Some(GcBlocking {
-                started_at: parse_naive_datetime("2024-07-19T09:00:00.123000000"),
-                reasons: enumset::EnumSet::from_iter([GcBlockingReason::DetachAncestor]),
-            }),
-            last_aux_file_policy: Default::default(),
-            archived_at: None,
-            import_pgdata: Some(import_pgdata::index_part_format::Root::V1(import_pgdata::index_part_format::V1::Done(import_pgdata::index_part_format::Done{
-                started_at: parse_naive_datetime("2024-11-13T09:23:42.123000000"),
-                finished_at: parse_naive_datetime("2024-11-13T09:42:23.123000000"),
-                idempotency_key: import_pgdata::index_part_format::IdempotencyKey::new("specified-by-client-218a5213-5044-4562-a28d-d024c5f057f5".to_string()),
-            }))),
-            rel_size_migration: Some(RelSizeMigration::Legacy),
+            })))
         };
 
         let part = IndexPart::from_json_bytes(example.as_bytes()).unwrap();

pageserver/src/tenant/remote_timeline_client/upload.rs

@@ -40,10 +40,6 @@ pub(crate) async fn upload_index_part(
     });
     pausable_failpoint!("before-upload-index-pausable");
 
-    // Safety: refuse to persist invalid index metadata, to mitigate the impact of any bug that produces this
-    // (this should never happen)
-    index_part.validate().map_err(|e| anyhow::anyhow!(e))?;
-
     // FIXME: this error comes too late
     let serialized = index_part.to_json_bytes()?;
     let serialized = Bytes::from(serialized);

pageserver/src/tenant/storage_layer/layer/tests.rs

@@ -1,6 +1,6 @@
 use std::time::UNIX_EPOCH;
 
-use pageserver_api::key::{Key, CONTROLFILE_KEY};
+use pageserver_api::key::CONTROLFILE_KEY;
 use tokio::task::JoinSet;
 use utils::{
     completion::{self, Completion},
@@ -9,10 +9,7 @@ use utils::{
 
 use super::failpoints::{Failpoint, FailpointKind};
 use super::*;
-use crate::{
-    context::DownloadBehavior,
-    tenant::{harness::test_img, storage_layer::LayerVisibilityHint},
-};
+use crate::{context::DownloadBehavior, tenant::storage_layer::LayerVisibilityHint};
 use crate::{task_mgr::TaskKind, tenant::harness::TenantHarness};
 
 /// Used in tests to advance a future to wanted await point, and not futher.
@@ -34,51 +31,20 @@ async fn smoke_test() {
 
     let ctx = RequestContext::new(TaskKind::UnitTest, DownloadBehavior::Download);
 
-    let image_layers = vec![(
-        Lsn(0x40),
-        vec![(
-            Key::from_hex("620000000033333333444444445500000000").unwrap(),
-            test_img("foo"),
-        )],
-    )];
-
-    // Create a test timeline with one real layer, and one synthetic test layer.  The synthetic
-    // one is only there so that we can GC the real one without leaving the timeline's metadata
-    // empty, which is an illegal state (see [`IndexPart::validate`]).
     let timeline = tenant
-        .create_test_timeline_with_layers(
-            TimelineId::generate(),
-            Lsn(0x10),
-            14,
-            &ctx,
-            Default::default(),
-            image_layers,
-            Lsn(0x100),
-        )
+        .create_test_timeline(TimelineId::generate(), Lsn(0x10), 14, &ctx)
         .await
         .unwrap();
 
-    // Grab one of the timeline's layers to exercise in the test, and the other layer that is just
-    // there to avoid the timeline being illegally empty
-    let (layer, dummy_layer) = {
+    let layer = {
         let mut layers = {
             let layers = timeline.layers.read().await;
             layers.likely_resident_layers().cloned().collect::<Vec<_>>()
         };
 
-        assert_eq!(layers.len(), 2);
+        assert_eq!(layers.len(), 1);
 
-        layers.sort_by_key(|l| l.layer_desc().get_key_range().start);
-        let synthetic_layer = layers.pop().unwrap();
-        let real_layer = layers.pop().unwrap();
-        tracing::info!(
-            "real_layer={:?} ({}), synthetic_layer={:?} ({})",
-            real_layer,
-            real_layer.layer_desc().file_size,
-            synthetic_layer,
-            synthetic_layer.layer_desc().file_size
-        );
-        (real_layer, synthetic_layer)
+        layers.swap_remove(0)
     };
 
     // all layers created at pageserver are like `layer`, initialized with strong
@@ -207,13 +173,10 @@ async fn smoke_test() {
 
     let rtc = &timeline.remote_client;
 
-    // Simulate GC removing our test layer.
     {
-        let mut g = timeline.layers.write().await;
-
         let layers = &[layer];
+        let mut g = timeline.layers.write().await;
         g.open_mut().unwrap().finish_gc_timeline(layers);
-
         // this just updates the remote_physical_size for demonstration purposes
         rtc.schedule_gc_update(layers).unwrap();
     }
@@ -228,10 +191,7 @@ async fn smoke_test() {
 
     rtc.wait_completion().await.unwrap();
 
-    assert_eq!(
-        rtc.get_remote_physical_size(),
-        dummy_layer.metadata().file_size
-    );
+    assert_eq!(rtc.get_remote_physical_size(), 0);
     assert_eq!(0, LAYER_IMPL_METRICS.inits_cancelled.get())
 }
 

pageserver/src/tenant/timeline.rs

@@ -51,9 +51,7 @@ use tokio::{
 use tokio_util::sync::CancellationToken;
 use tracing::*;
 use utils::{
-    fs_ext,
-    guard_arc_swap::GuardArcSwap,
-    pausable_failpoint,
+    fs_ext, pausable_failpoint,
     postgres_client::PostgresClientProtocol,
     sync::gate::{Gate, GateGuard},
 };
@@ -76,7 +74,6 @@ use std::{pin::pin, sync::OnceLock};
 
 use crate::{
     aux_file::AuxFileSizeEstimator,
-    page_service::TenantManagerTypes,
     tenant::{
         config::AttachmentMode,
         layer_map::{LayerMap, SearchResult},
@@ -356,8 +353,8 @@ pub struct Timeline {
     // though let's keep them both for better error visibility.
     pub initdb_lsn: Lsn,
 
-    /// The repartitioning result. Allows a single writer and multiple readers.
-    pub(crate) partitioning: GuardArcSwap<((KeyPartitioning, SparseKeyPartitioning), Lsn)>,
+    /// When did we last calculate the partitioning? Make it pub to test cases.
+    pub(super) partitioning: tokio::sync::Mutex<((KeyPartitioning, SparseKeyPartitioning), Lsn)>,
 
     /// Configuration: how often should the partitioning be recalculated.
     repartition_threshold: u64,
@@ -432,7 +429,7 @@ pub struct Timeline {
 
     pub(crate) l0_flush_global_state: L0FlushGlobalState,
 
-    pub(crate) handles: handle::PerTimelineState<TenantManagerTypes>,
+    pub(crate) handles: handle::PerTimelineState<crate::page_service::TenantManagerTypes>,
 
     pub(crate) attach_wal_lag_cooldown: Arc<OnceLock<WalLagCooldown>>,
 
@@ -2343,8 +2340,7 @@ impl Timeline {
                     // initial logical size is 0.
                     LogicalSize::empty_initial()
                 },
-
-                partitioning: GuardArcSwap::new((
+                partitioning: tokio::sync::Mutex::new((
                     (KeyPartitioning::new(), KeyPartitioning::new().into_sparse()),
                     Lsn(0),
                 )),
@@ -4032,15 +4028,18 @@ impl Timeline {
         flags: EnumSet<CompactFlags>,
         ctx: &RequestContext,
     ) -> Result<((KeyPartitioning, SparseKeyPartitioning), Lsn), CompactionError> {
-        let Ok(mut guard) = self.partitioning.try_write_guard() else {
+        let Ok(mut partitioning_guard) = self.partitioning.try_lock() else {
             // NB: there are two callers, one is the compaction task, of which there is only one per struct Tenant and hence Timeline.
             // The other is the initdb optimization in flush_frozen_layer, used by `boostrap_timeline`, which runs before `.activate()`
             // and hence before the compaction task starts.
+            // Note that there are a third "caller" that will take the `partitioning` lock. It is `gc_compaction_split_jobs` for
+            // gc-compaction where it uses the repartition data to determine the split jobs. In the future, it might use its own
+            // heuristics, but for now, we should allow concurrent access to it and let the caller retry compaction.
             return Err(CompactionError::Other(anyhow!(
-                "repartition() called concurrently"
+                "repartition() called concurrently, this is rare and a retry should be fine"
             )));
         };
-        let ((dense_partition, sparse_partition), partition_lsn) = &*guard.read();
+        let ((dense_partition, sparse_partition), partition_lsn) = &*partitioning_guard;
         if lsn < *partition_lsn {
             return Err(CompactionError::Other(anyhow!(
                 "repartition() called with LSN going backwards, this should not happen"
@@ -4068,9 +4067,9 @@ impl Timeline {
         let sparse_partitioning = SparseKeyPartitioning {
             parts: vec![sparse_ks],
         }; // no partitioning for metadata keys for now
-        let result = ((dense_partitioning, sparse_partitioning), lsn);
-        guard.write(result.clone());
-        Ok(result)
+        *partitioning_guard = ((dense_partitioning, sparse_partitioning), lsn);
+
+        Ok((partitioning_guard.0.clone(), partitioning_guard.1))
     }
 
     // Is it time to create a new image layer for the given partition?
@@ -4626,10 +4625,6 @@ impl Drop for Timeline {
                 }
             }
         }
-        info!(
-            "Timeline {} for tenant {} is being dropped",
-            self.timeline_id, self.tenant_shard_id.tenant_id
-        );
     }
 }
 
@@ -5678,17 +5673,9 @@ impl Timeline {
         info!("force created image layer {}", image_layer.local_path());
         {
             let mut guard = self.layers.write().await;
-            guard
-                .open_mut()
-                .unwrap()
-                .force_insert_layer(image_layer.clone());
+            guard.open_mut().unwrap().force_insert_layer(image_layer);
         }
 
-        // Update remote_timeline_client state to reflect existence of this layer
-        self.remote_client
-            .schedule_layer_file_upload(image_layer)
-            .unwrap();
-
         Ok(())
     }
 
@@ -5739,17 +5726,9 @@ impl Timeline {
         info!("force created delta layer {}", delta_layer.local_path());
         {
             let mut guard = self.layers.write().await;
-            guard
-                .open_mut()
-                .unwrap()
-                .force_insert_layer(delta_layer.clone());
+            guard.open_mut().unwrap().force_insert_layer(delta_layer);
         }
 
-        // Update remote_timeline_client state to reflect existence of this layer
-        self.remote_client
-            .schedule_layer_file_upload(delta_layer)
-            .unwrap();
-
         Ok(())
     }
 

pageserver/src/tenant/timeline/compaction.rs

@@ -2146,7 +2146,12 @@ impl Timeline {
         let mut compact_jobs = Vec::new();
         // For now, we simply use the key partitioning information; we should do a more fine-grained partitioning
         // by estimating the amount of files read for a compaction job. We should also partition on LSN.
-        let ((dense_ks, sparse_ks), _) = self.partitioning.read().as_ref().clone();
+        let ((dense_ks, sparse_ks), _) = {
+            let Ok(partition) = self.partitioning.try_lock() else {
+                bail!("failed to acquire partition lock during gc-compaction");
+            };
+            partition.clone()
+        };
         // Truncate the key range to be within user specified compaction range.
         fn truncate_to(
             source_start: &Key,

pageserver/src/tenant/timeline/delete.rs

@@ -112,7 +112,7 @@ pub(super) async fn delete_local_timeline_directory(
 }
 
 /// It is important that this gets called when DeletionGuard is being held.
-/// For more context see comments in [`make_timeline_delete_guard`]
+/// For more context see comments in [`DeleteTimelineFlow::prepare`]
 async fn remove_maybe_offloaded_timeline_from_tenant(
     tenant: &Tenant,
     timeline: &TimelineOrOffloaded,
@@ -193,8 +193,10 @@ impl DeleteTimelineFlow {
     ) -> Result<(), DeleteTimelineError> {
         super::debug_assert_current_span_has_tenant_and_timeline_id();
 
+        let allow_offloaded_children = false;
+        let set_stopping = true;
         let (timeline, mut guard) =
-            make_timeline_delete_guard(tenant, timeline_id, TimelineDeleteGuardKind::Delete)?;
+            Self::prepare(tenant, timeline_id, allow_offloaded_children, set_stopping)?;
 
         guard.mark_in_progress()?;
 
@@ -331,6 +333,75 @@ impl DeleteTimelineFlow {
         Ok(())
     }
 
+    pub(super) fn prepare(
+        tenant: &Tenant,
+        timeline_id: TimelineId,
+        allow_offloaded_children: bool,
+        set_stopping: bool,
+    ) -> Result<(TimelineOrOffloaded, DeletionGuard), DeleteTimelineError> {
+        // Note the interaction between this guard and deletion guard.
+        // Here we attempt to lock deletion guard when we're holding a lock on timelines.
+        // This is important because when you take into account `remove_timeline_from_tenant`
+        // we remove timeline from memory when we still hold the deletion guard.
+        // So here when timeline deletion is finished timeline wont be present in timelines map at all
+        // which makes the following sequence impossible:
+        // T1: get preempted right before the try_lock on `Timeline::delete_progress`
+        // T2: do a full deletion, acquire and drop `Timeline::delete_progress`
+        // T1: acquire deletion lock, do another `DeleteTimelineFlow::run`
+        // For more context see this discussion: `https://github.com/neondatabase/neon/pull/4552#discussion_r1253437346`
+        let timelines = tenant.timelines.lock().unwrap();
+        let timelines_offloaded = tenant.timelines_offloaded.lock().unwrap();
+
+        let timeline = match timelines.get(&timeline_id) {
+            Some(t) => TimelineOrOffloaded::Timeline(Arc::clone(t)),
+            None => match timelines_offloaded.get(&timeline_id) {
+                Some(t) => TimelineOrOffloaded::Offloaded(Arc::clone(t)),
+                None => return Err(DeleteTimelineError::NotFound),
+            },
+        };
+
+        // Ensure that there are no child timelines, because we are about to remove files,
+        // which will break child branches
+        let mut children = Vec::new();
+        if !allow_offloaded_children {
+            children.extend(timelines_offloaded.iter().filter_map(|(id, entry)| {
+                (entry.ancestor_timeline_id == Some(timeline_id)).then_some(*id)
+            }));
+        }
+        children.extend(timelines.iter().filter_map(|(id, entry)| {
+            (entry.get_ancestor_timeline_id() == Some(timeline_id)).then_some(*id)
+        }));
+
+        if !children.is_empty() {
+            return Err(DeleteTimelineError::HasChildren(children));
+        }
+
+        // Note that using try_lock here is important to avoid a deadlock.
+        // Here we take lock on timelines and then the deletion guard.
+        // At the end of the operation we're holding the guard and need to lock timelines map
+        // to remove the timeline from it.
+        // Always if you have two locks that are taken in different order this can result in a deadlock.
+
+        let delete_progress = Arc::clone(timeline.delete_progress());
+        let delete_lock_guard = match delete_progress.try_lock_owned() {
+            Ok(guard) => DeletionGuard(guard),
+            Err(_) => {
+                // Unfortunately if lock fails arc is consumed.
+                return Err(DeleteTimelineError::AlreadyInProgress(Arc::clone(
+                    timeline.delete_progress(),
+                )));
+            }
+        };
+
+        if set_stopping {
+            if let TimelineOrOffloaded::Timeline(timeline) = &timeline {
+                timeline.set_state(TimelineState::Stopping);
+            }
+        }
+
+        Ok((timeline, delete_lock_guard))
+    }
+
     fn schedule_background(
         guard: DeletionGuard,
         conf: &'static PageServerConf,
@@ -412,80 +483,6 @@ impl DeleteTimelineFlow {
     }
 }
 
-#[derive(Copy, Clone, PartialEq, Eq)]
-pub(super) enum TimelineDeleteGuardKind {
-    Offload,
-    Delete,
-}
-
-pub(super) fn make_timeline_delete_guard(
-    tenant: &Tenant,
-    timeline_id: TimelineId,
-    guard_kind: TimelineDeleteGuardKind,
-) -> Result<(TimelineOrOffloaded, DeletionGuard), DeleteTimelineError> {
-    // Note the interaction between this guard and deletion guard.
-    // Here we attempt to lock deletion guard when we're holding a lock on timelines.
-    // This is important because when you take into account `remove_timeline_from_tenant`
-    // we remove timeline from memory when we still hold the deletion guard.
-    // So here when timeline deletion is finished timeline wont be present in timelines map at all
-    // which makes the following sequence impossible:
-    // T1: get preempted right before the try_lock on `Timeline::delete_progress`
-    // T2: do a full deletion, acquire and drop `Timeline::delete_progress`
-    // T1: acquire deletion lock, do another `DeleteTimelineFlow::run`
-    // For more context see this discussion: `https://github.com/neondatabase/neon/pull/4552#discussion_r1253437346`
-    let timelines = tenant.timelines.lock().unwrap();
-    let timelines_offloaded = tenant.timelines_offloaded.lock().unwrap();
-
-    let timeline = match timelines.get(&timeline_id) {
-        Some(t) => TimelineOrOffloaded::Timeline(Arc::clone(t)),
-        None => match timelines_offloaded.get(&timeline_id) {
-            Some(t) => TimelineOrOffloaded::Offloaded(Arc::clone(t)),
-            None => return Err(DeleteTimelineError::NotFound),
-        },
-    };
-
-    // Ensure that there are no child timelines, because we are about to remove files,
-    // which will break child branches
-    let mut children = Vec::new();
-    if guard_kind == TimelineDeleteGuardKind::Delete {
-        children.extend(timelines_offloaded.iter().filter_map(|(id, entry)| {
-            (entry.ancestor_timeline_id == Some(timeline_id)).then_some(*id)
-        }));
-    }
-    children.extend(timelines.iter().filter_map(|(id, entry)| {
-        (entry.get_ancestor_timeline_id() == Some(timeline_id)).then_some(*id)
-    }));
-
-    if !children.is_empty() {
-        return Err(DeleteTimelineError::HasChildren(children));
-    }
-
-    // Note that using try_lock here is important to avoid a deadlock.
-    // Here we take lock on timelines and then the deletion guard.
-    // At the end of the operation we're holding the guard and need to lock timelines map
-    // to remove the timeline from it.
-    // Always if you have two locks that are taken in different order this can result in a deadlock.
-
-    let delete_progress = Arc::clone(timeline.delete_progress());
-    let delete_lock_guard = match delete_progress.try_lock_owned() {
-        Ok(guard) => DeletionGuard(guard),
-        Err(_) => {
-            // Unfortunately if lock fails arc is consumed.
-            return Err(DeleteTimelineError::AlreadyInProgress(Arc::clone(
-                timeline.delete_progress(),
-            )));
-        }
-    };
-
-    if guard_kind == TimelineDeleteGuardKind::Delete {
-        if let TimelineOrOffloaded::Timeline(timeline) = &timeline {
-            timeline.set_state(TimelineState::Stopping);
-        }
-    }
-
-    Ok((timeline, delete_lock_guard))
-}
-
 pub(super) struct DeletionGuard(OwnedMutexGuard<DeleteTimelineFlow>);
 
 impl Deref for DeletionGuard {

pageserver/src/tenant/timeline/handle.rs

@@ -32,151 +32,54 @@
 //!
 //! # Design
 //!
-//! ## Data Structures
-//!
 //! There are three user-facing data structures:
 //! - `PerTimelineState`: a struct embedded into each Timeline struct. Lifetime == Timeline lifetime.
 //! - `Cache`: a struct private to each connection handler; Lifetime == connection lifetime.
 //! - `Handle`: a smart pointer that holds the Timeline gate open and derefs to `&Timeline`.
-//! - `WeakHandle`: downgrade of a `Handle` that does not keep the gate open, but allows
-//!   trying to ugprade back to a `Handle`, guaranteeing it's the same `Timeline` *object*.
+//!   Lifetime: for a single request dispatch on the Timeline (i.e., one getpage request)
 //!
-//! Internally, there is 0 or 1 `HandleInner` per `(Cache,Timeline)`.
-//! Since Cache:Connection is 1:1, there is 0 or 1 `HandleInner` per `(Connection,Timeline)`.
+//! The `Handle` is just a wrapper around an `Arc<HandleInner>`.
 //!
-//! The `HandleInner`  is allocated as a `Arc<Mutex<HandleInner>>` and
-//! referenced weakly and strongly from various places which we are now illustrating.
-//! For brevity, we will omit the `Arc<Mutex<>>` part in the following and instead
-//! use `strong ref` and `weak ref` when referring to the `Arc<Mutex<HandleInner>>`
-//! or `Weak<Mutex<HandleInner>>`, respectively.
-//!
-//! - The `Handle` is a strong ref.
-//! - The `WeakHandle` is a weak ref.
-//! - The `PerTimelineState` contains a `HashMap<CacheId, strong ref>`.
-//! - The `Cache` is a `HashMap<unique identifier for the shard, weak ref>`.
-//!
-//! Lifetimes:
-//! - `WeakHandle` and `Handle`: single pagestream request.
-//! - `Cache`: single page service connection.
-//! - `PerTimelineState`:  lifetime of the Timeline object (i.e., i.e., till `Timeline::shutdown`).
-//!
-//! ## Request Handling Flow (= filling and using the `Cache``)
+//! There is one long-lived `Arc<HandleInner>`, which is stored in the `PerTimelineState`.
+//! The `Cache` stores a `Weak<HandleInner>` for each cached Timeline.
 //!
 //! To dispatch a request, the page service connection calls `Cache::get`.
 //!
 //! A cache miss means we consult the tenant manager for shard routing,
-//! resulting in an `Arc<Timeline>`. We enter its gate _once_ and store it in the the
-//! `Arc<Mutex<HandleInner>>>`. A weak ref is stored in the `Cache`
-//! and a strong ref in the `PerTimelineState`.
-//! A strong ref is returned wrapped in a `Handle`.
+//! resulting in an `Arc<Timeline>`. We enter its gate _once_ and construct an
+//! `Arc<HandleInner>`. We store a `Weak<HandleInner>` in the cache
+//! and the `Arc<HandleInner>` in the `PerTimelineState`.
 //!
 //! For subsequent requests, `Cache::get` will perform a "fast path" shard routing
-//! and find the weak ref in the cache.
-//! We upgrade the weak ref to a strong ref and return it wrapped in a `Handle`.
+//! and find the `Weak<HandleInner>` in the cache.
+//! We upgrade the `Weak<HandleInner>` to an `Arc<HandleInner>` and wrap it in the user-facing `Handle` type.
 //!
-//! The pagestream processing is pipelined and involves a batching step.
-//! While a request is batching, the `Handle` is downgraded to a `WeakHandle`.
-//! When the batch is ready to be executed, the `WeakHandle` is upgraded back to a `Handle`
-//! and the request handler dispatches the request to the right `<Handle as Deref<Target = Timeline>>::$request_method`.
+//! The request handler dispatches the request to the right `<Handle as Deref<Target = Timeline>>::$request_method`.
 //! It then drops the `Handle`, which drops the `Arc<HandleInner>`.
 //!
-//! # Performance
+//! # Memory Management / How The Reference Cycle Is Broken
 //!
-//! Remember from the introductory section:
+//! The attentive reader may have noticed the strong reference cycle
+//! from `Arc<HandleInner>` to `PerTimelineState` to `Arc<Timeline>`.
 //!
-//! > However, we want to avoid the overhead of entering the gate for every
-//! > method invocation.
-//!
-//! Why do we want to avoid that?
-//! Because the gate is a shared location in memory and entering it involves
-//! bumping refcounts, which leads to cache contention if done frequently
-//! from multiple cores in parallel.
-//!
-//! So, we only acquire the `GateGuard` once on `Cache` miss, and wrap it in an `Arc`.
-//! That `Arc` is private to the `HandleInner` and hence to the connection.
-//! (Review the "Data Structures" section if that is unclear to you.)
-//!
-//! A `WeakHandle` is a weak ref to the `HandleInner`.
-//! When upgrading a `WeakHandle`, we upgrade to a strong ref to the `HandleInner` and
-//! further acquire an additional strong ref to the `Arc<GateGuard>` inside it.
-//! Again, this manipulation of ref counts is is cheap because `Arc` is private to the connection.
-//!
-//! When downgrading a `Handle` to a `WeakHandle`, we drop the `Arc<GateGuard>`.
-//! Again, this is cheap because the `Arc` is private to the connection.
-//!
-//! In addition to the GateGuard, we need to provide `Deref<Target=Timeline>` impl.
-//! For this, both `Handle` need infallible access to an `Arc<Timeline>`.
-//! We could clone the `Arc<Timeline>` when upgrading a `WeakHandle`, but that would cause contention
-//! on the shared memory location that trakcs the refcount of the `Arc<Timeline>`.
-//! Instead, we wrap the `Arc<Timeline>` into another `Arc`.
-//! so that we can clone it cheaply when upgrading a `WeakHandle`.
-//!
-//! # Shutdown
-//!
-//! The attentive reader may have noticed the following reference cycle around the `Arc<Timeline>`:
-//!
-//! ```text
-//! Timeline --owns--> PerTimelineState --strong--> HandleInner --strong--> Timeline
-//! ```
-//!
-//! Further, there is this cycle:
-//!
-//! ```text
-//! Timeline --owns--> PerTimelineState --strong--> HandleInner --strong--> GateGuard --keepalive--> Timeline
-//! ```
-//!
-//! The former cycle is a memory leak if not broken.
-//! The latter cycle further prevents the Timeline from shutting down
-//! because we certainly won't drop the Timeline while the GateGuard is alive.
-//! Preventing shutdown is the whole point of this handle/cache system,
-//! but when the Timeline needs to shut down, we need to break the cycle.
+//! This cycle is intentional: while it exists, the `Cache` can upgrade its
+//! `Weak<HandleInner>` to an `Arc<HandleInner>` in a single atomic operation.
 //!
 //! The cycle is broken by either
-//! - Timeline shutdown (=> `PerTimelineState::shutdown`)
-//! - Connection shutdown (=> dropping the `Cache`).
+//! - `PerTimelineState::shutdown` or
+//! - dropping the `Cache`.
 //!
-//! Both transition the `HandleInner` from [`HandleInner::KeepingTimelineGateOpen`] to
-//! [`HandleInner::ShutDown`], which drops the only long-lived strong ref to the
-//! `Arc<GateGuard>`.
-//!
-//! `PerTimelineState::shutdown` drops all the `HandleInners` it contains,
-//! thereby breaking the cycle.
-//! It also initiates draining of already existing `Handle`s by
-//! poisoning things so that no new `HandleInner`'s can be added
-//! to the `PerTimelineState`, which will make subsequent `Cache::get` fail.
-//!
-//! Concurrently existing / already upgraded `Handle`s will extend the
-//! lifetime of the `Arc<Mutex<HandleInner>>` and hence cycles.
+//! Concurrently existing `Handle`s will extend the existence of the cycle.
 //! However, since `Handle`s are short-lived and new `Handle`s are not
-//! handed out from `Cache::get` or `WeakHandle::upgrade` after
-//! `PerTimelineState::shutdown`, that extension of the cycle is bounded.
-//!
-//! Concurrently existing `WeakHandle`s will fail to `upgrade()`:
-//! while they will succeed in upgrading `Weak<Mutex<HandleInner>>`,
-//! they will find the inner in state `HandleInner::ShutDown` state where the
-//! `Arc<GateGuard>` and Timeline has already been dropped.
-//!
-//! Dropping the `Cache` undoes the registration of this `Cache`'s
-//! `HandleInner`s from all the `PerTimelineState`s, i.e., it
-//! removes the strong ref to each of its `HandleInner`s
-//! from all the `PerTimelineState`.
-//!
-//! # Locking Rules
-//!
-//! To prevent deadlocks we:
-//!
-//! 1. Only ever hold one of the locks at a time.
-//! 2. Don't add more than one Drop impl that locks on the
-//!    cycles above.
-//!
-//! As per (2), that impl is in `Drop for Cache`.
+//! handed out after either `PerTimelineState::shutdown` or `Cache` drop,
+//! that extension of the cycle is bounded.
 //!
 //! # Fast Path for Shard Routing
 //!
 //! The `Cache` has a fast path for shard routing to avoid calling into
 //! the tenant manager for every request.
 //!
-//! The `Cache` maintains a hash map of `ShardTimelineId` to `WeakHandle`s.
+//! The `Cache` maintains a hash map of `ShardTimelineId` to `Weak<HandleInner>`.
 //!
 //! The current implementation uses the first entry in the hash map
 //! to determine the `ShardParameters` and derive the correct
@@ -184,18 +87,18 @@
 //!
 //! It then looks up the hash map for that `ShardTimelineId := {ShardIndex,TimelineId}`.
 //!
-//! If the lookup is successful and the `WeakHandle` can be upgraded,
+//! If the lookup is successful and the `Weak<HandleInner>` can be upgraded,
 //! it's a hit.
 //!
 //! ## Cache invalidation
 //!
-//! The insight is that cache invalidation is sufficient and most efficiently if done lazily.
+//! The insight is that cache invalidation is sufficient and most efficiently done lazily.
 //! The only reasons why an entry in the cache can become stale are:
 //! 1. The `PerTimelineState` / Timeline is shutting down e.g. because the shard is
 //!    being detached, timeline or shard deleted, or pageserver is shutting down.
 //! 2. We're doing a shard split and new traffic should be routed to the child shards.
 //!
-//! Regarding (1), we will eventually fail to upgrade the `WeakHandle` once the
+//! Regarding (1), we will eventually fail to upgrade the `Weak<HandleInner>` once the
 //! timeline has shut down, and when that happens, we remove the entry from the cache.
 //!
 //! Regarding (2), the insight is that it is toally fine to keep dispatching requests
@@ -204,6 +107,8 @@
 
 use std::collections::hash_map;
 use std::collections::HashMap;
+use std::sync::atomic::AtomicBool;
+use std::sync::atomic::Ordering;
 use std::sync::Arc;
 use std::sync::Mutex;
 use std::sync::Weak;
@@ -247,7 +152,7 @@ pub(crate) struct Cache<T: Types> {
     map: Map<T>,
 }
 
-type Map<T> = HashMap<ShardTimelineId, WeakHandle<T>>;
+type Map<T> = HashMap<ShardTimelineId, Weak<HandleInner<T>>>;
 
 impl<T: Types> Default for Cache<T> {
     fn default() -> Self {
@@ -265,22 +170,12 @@ pub(crate) struct ShardTimelineId {
 }
 
 /// See module-level comment.
-pub(crate) struct Handle<T: Types> {
-    timeline: Arc<T::Timeline>,
-    #[allow(dead_code)] // the field exists to keep the gate open
-    gate_guard: Arc<utils::sync::gate::GateGuard>,
-    inner: Arc<Mutex<HandleInner<T>>>,
-}
-pub(crate) struct WeakHandle<T: Types> {
-    inner: Weak<Mutex<HandleInner<T>>>,
-}
-enum HandleInner<T: Types> {
-    KeepingTimelineGateOpen {
-        #[allow(dead_code)]
-        gate_guard: Arc<utils::sync::gate::GateGuard>,
-        timeline: Arc<T::Timeline>,
-    },
-    ShutDown,
+pub(crate) struct Handle<T: Types>(Arc<HandleInner<T>>);
+struct HandleInner<T: Types> {
+    shut_down: AtomicBool,
+    timeline: T::Timeline,
+    // The timeline's gate held open.
+    _gate_guard: utils::sync::gate::GateGuard,
 }
 
 /// Embedded in each [`Types::Timeline`] as the anchor for the only long-lived strong ref to `HandleInner`.
@@ -288,8 +183,7 @@ enum HandleInner<T: Types> {
 /// See module-level comment for details.
 pub struct PerTimelineState<T: Types> {
     // None = shutting down
-    #[allow(clippy::type_complexity)]
-    handles: Mutex<Option<HashMap<CacheId, Arc<Mutex<HandleInner<T>>>>>>,
+    handles: Mutex<Option<HashMap<CacheId, Arc<HandleInner<T>>>>>,
 }
 
 impl<T: Types> Default for PerTimelineState<T> {
@@ -349,24 +243,49 @@ impl<T: Types> Cache<T> {
         shard_selector: ShardSelector,
         tenant_manager: &T::TenantManager,
     ) -> Result<Handle<T>, GetError<T>> {
-        // terminates because when every iteration we remove an element from the map
-        let miss: ShardSelector = loop {
+        // terminates because each iteration removes an element from the map
+        loop {
+            let handle = self
+                .get_impl(timeline_id, shard_selector, tenant_manager)
+                .await?;
+            if handle.0.shut_down.load(Ordering::Relaxed) {
+                let removed = self
+                    .map
+                    .remove(&handle.0.timeline.shard_timeline_id())
+                    .expect("invariant of get_impl is that the returned handle is in the map");
+                assert!(
+                    Weak::ptr_eq(&removed, &Arc::downgrade(&handle.0)),
+                    "shard_timeline_id() incorrect?"
+                );
+            } else {
+                return Ok(handle);
+            }
+        }
+    }
+
+    #[instrument(level = "trace", skip_all)]
+    async fn get_impl(
+        &mut self,
+        timeline_id: TimelineId,
+        shard_selector: ShardSelector,
+        tenant_manager: &T::TenantManager,
+    ) -> Result<Handle<T>, GetError<T>> {
+        let miss: ShardSelector = {
             let routing_state = self.shard_routing(timeline_id, shard_selector);
             match routing_state {
                 RoutingResult::FastPath(handle) => return Ok(handle),
                 RoutingResult::SlowPath(key) => match self.map.get(&key) {
                     Some(cached) => match cached.upgrade() {
-                        Ok(upgraded) => return Ok(upgraded),
-                        Err(HandleUpgradeError::ShutDown) => {
-                            // TODO: dedup with shard_routing()
+                        Some(upgraded) => return Ok(Handle(upgraded)),
+                        None => {
                             trace!("handle cache stale");
                             self.map.remove(&key).unwrap();
-                            continue;
+                            ShardSelector::Known(key.shard_index)
                         }
                     },
-                    None => break ShardSelector::Known(key.shard_index),
+                    None => ShardSelector::Known(key.shard_index),
                 },
-                RoutingResult::NeedConsultTenantManager => break shard_selector,
+                RoutingResult::NeedConsultTenantManager => shard_selector,
             }
         };
         self.get_miss(timeline_id, miss, tenant_manager).await
@@ -383,7 +302,7 @@ impl<T: Types> Cache<T> {
             let Some((first_key, first_handle)) = self.map.iter().next() else {
                 return RoutingResult::NeedConsultTenantManager;
             };
-            let Ok(first_handle) = first_handle.upgrade() else {
+            let Some(first_handle) = first_handle.upgrade() else {
                 // TODO: dedup with get()
                 trace!("handle cache stale");
                 let first_key_owned = *first_key;
@@ -391,7 +310,7 @@ impl<T: Types> Cache<T> {
                 continue;
             };
 
-            let first_handle_shard_identity = first_handle.get_shard_identity();
+            let first_handle_shard_identity = first_handle.timeline.get_shard_identity();
             let make_shard_index = |shard_num: ShardNumber| ShardIndex {
                 shard_number: shard_num,
                 shard_count: first_handle_shard_identity.count,
@@ -410,11 +329,11 @@ impl<T: Types> Cache<T> {
             };
             let first_handle_shard_timeline_id = ShardTimelineId {
                 shard_index: first_handle_shard_identity.shard_index(),
-                timeline_id: first_handle.shard_timeline_id().timeline_id,
+                timeline_id: first_handle.timeline.shard_timeline_id().timeline_id,
             };
 
             if need_shard_timeline_id == first_handle_shard_timeline_id {
-                return RoutingResult::FastPath(first_handle);
+                return RoutingResult::FastPath(Handle(first_handle));
             } else {
                 return RoutingResult::SlowPath(need_shard_timeline_id);
             }
@@ -438,30 +357,23 @@ impl<T: Types> Cache<T> {
                     ShardSelector::Known(idx) => assert_eq!(idx, &key.shard_index),
                 }
 
-                trace!("creating new HandleInner");
-                let handle_inner_arc = Arc::new(Mutex::new(HandleInner::KeepingTimelineGateOpen {
-                    gate_guard: Arc::new(
-                        // this enter() is expensive in production code because
-                        // it hits the global Arc<Timeline>::gate refcounts
-                        match timeline.gate().enter() {
-                            Ok(guard) => guard,
-                            Err(_) => {
-                                return Err(GetError::TimelineGateClosed);
-                            }
-                        },
-                    ),
-                    // this clone is expensive in production code because
-                    // it hits the global Arc<Timeline>::clone refcounts
-                    timeline: Arc::new(timeline.clone()),
-                }));
-                let handle_weak = WeakHandle {
-                    inner: Arc::downgrade(&handle_inner_arc),
+                let gate_guard = match timeline.gate().enter() {
+                    Ok(guard) => guard,
+                    Err(_) => {
+                        return Err(GetError::TimelineGateClosed);
+                    }
                 };
-                let handle = handle_weak
-                    .upgrade()
-                    .ok()
-                    .expect("we just created it and it's not linked anywhere yet");
-                {
+                trace!("creating new HandleInner");
+                let handle = Arc::new(
+                    // TODO: global metric that keeps track of the number of live HandlerTimeline instances
+                    // so we can identify reference cycle bugs.
+                    HandleInner {
+                        shut_down: AtomicBool::new(false),
+                        _gate_guard: gate_guard,
+                        timeline: timeline.clone(),
+                    },
+                );
+                let handle = {
                     let mut lock_guard = timeline
                         .per_timeline_state()
                         .handles
@@ -469,8 +381,7 @@ impl<T: Types> Cache<T> {
                         .expect("mutex poisoned");
                     match &mut *lock_guard {
                         Some(per_timeline_state) => {
-                            let replaced =
-                                per_timeline_state.insert(self.id, Arc::clone(&handle_inner_arc));
+                            let replaced = per_timeline_state.insert(self.id, Arc::clone(&handle));
                             assert!(replaced.is_none(), "some earlier code left a stale handle");
                             match self.map.entry(key) {
                                 hash_map::Entry::Occupied(_o) => {
@@ -481,7 +392,8 @@ impl<T: Types> Cache<T> {
                                     unreachable!()
                                 }
                                 hash_map::Entry::Vacant(v) => {
-                                    v.insert(handle_weak);
+                                    v.insert(Arc::downgrade(&handle));
+                                    handle
                                 }
                             }
                         }
@@ -489,62 +401,14 @@ impl<T: Types> Cache<T> {
                             return Err(GetError::PerTimelineStateShutDown);
                         }
                     }
-                }
-                Ok(handle)
+                };
+                Ok(Handle(handle))
             }
             Err(e) => Err(GetError::TenantManager(e)),
         }
     }
 }
 
-pub(crate) enum HandleUpgradeError {
-    ShutDown,
-}
-
-impl<T: Types> WeakHandle<T> {
-    pub(crate) fn upgrade(&self) -> Result<Handle<T>, HandleUpgradeError> {
-        let Some(inner) = Weak::upgrade(&self.inner) else {
-            return Err(HandleUpgradeError::ShutDown);
-        };
-        let lock_guard = inner.lock().expect("poisoned");
-        match &*lock_guard {
-            HandleInner::KeepingTimelineGateOpen {
-                timeline,
-                gate_guard,
-            } => {
-                let gate_guard = Arc::clone(gate_guard);
-                let timeline = Arc::clone(timeline);
-                drop(lock_guard);
-                Ok(Handle {
-                    timeline,
-                    gate_guard,
-                    inner,
-                })
-            }
-            HandleInner::ShutDown => Err(HandleUpgradeError::ShutDown),
-        }
-    }
-
-    pub(crate) fn is_same_handle_as(&self, other: &WeakHandle<T>) -> bool {
-        Weak::ptr_eq(&self.inner, &other.inner)
-    }
-}
-
-impl<T: Types> std::ops::Deref for Handle<T> {
-    type Target = T::Timeline;
-    fn deref(&self) -> &Self::Target {
-        &self.timeline
-    }
-}
-
-impl<T: Types> Handle<T> {
-    pub(crate) fn downgrade(&self) -> WeakHandle<T> {
-        WeakHandle {
-            inner: Arc::downgrade(&self.inner),
-        }
-    }
-}
-
 impl<T: Types> PerTimelineState<T> {
     /// After this method returns, [`Cache::get`] will never again return a [`Handle`]
     /// to the [`Types::Timeline`] that embeds this per-timeline state.
@@ -566,62 +430,43 @@ impl<T: Types> PerTimelineState<T> {
             trace!("already shut down");
             return;
         };
-        for handle_inner_arc in handles.values() {
+        for handle in handles.values() {
             // Make hits fail.
-            let mut lock_guard = handle_inner_arc.lock().expect("poisoned");
-            lock_guard.shutdown();
+            handle.shut_down.store(true, Ordering::Relaxed);
         }
         drop(handles);
     }
 }
 
+impl<T: Types> std::ops::Deref for Handle<T> {
+    type Target = T::Timeline;
+    fn deref(&self) -> &Self::Target {
+        &self.0.timeline
+    }
+}
+
+#[cfg(test)]
+impl<T: Types> Drop for HandleInner<T> {
+    fn drop(&mut self) {
+        trace!("HandleInner dropped");
+    }
+}
+
 // When dropping a [`Cache`], prune its handles in the [`PerTimelineState`] to break the reference cycle.
 impl<T: Types> Drop for Cache<T> {
     fn drop(&mut self) {
-        for (
-            _,
-            WeakHandle {
-                inner: handle_inner_weak,
-            },
-        ) in self.map.drain()
-        {
-            let Some(handle_inner_arc) = handle_inner_weak.upgrade() else {
-                continue;
-            };
-            let Some(handle_timeline) = handle_inner_arc
-                // locking rules: drop lock before acquiring other lock below
-                .lock()
-                .expect("poisoned")
-                .shutdown()
-            else {
-                // Concurrent PerTimelineState::shutdown.
-                continue;
-            };
-            // Clean up per_timeline_state so the HandleInner allocation can be dropped.
-            let per_timeline_state = handle_timeline.per_timeline_state();
-            let mut handles_lock_guard = per_timeline_state.handles.lock().expect("mutex poisoned");
-            let Some(handles) = &mut *handles_lock_guard else {
-                continue;
-            };
-            let Some(removed_handle_inner_arc) = handles.remove(&self.id) else {
-                // Concurrent PerTimelineState::shutdown.
-                continue;
-            };
-            drop(handles_lock_guard); // locking rules!
-            assert!(Arc::ptr_eq(&removed_handle_inner_arc, &handle_inner_arc));
-        }
-    }
-}
-
-impl<T: Types> HandleInner<T> {
-    fn shutdown(&mut self) -> Option<Arc<T::Timeline>> {
-        match std::mem::replace(self, HandleInner::ShutDown) {
-            HandleInner::KeepingTimelineGateOpen { timeline, .. } => Some(timeline),
-            HandleInner::ShutDown => {
-                // Duplicate shutdowns are possible because both Cache::drop and PerTimelineState::shutdown
-                // may do it concurrently, but locking rules disallow holding per-timeline-state lock and
-                // the handle lock at the same time.
-                None
+        for (_, weak) in self.map.drain() {
+            if let Some(strong) = weak.upgrade() {
+                // handle is still being kept alive in PerTimelineState
+                let timeline = strong.timeline.per_timeline_state();
+                let mut handles = timeline.handles.lock().expect("mutex poisoned");
+                if let Some(handles) = &mut *handles {
+                    let Some(removed) = handles.remove(&self.id) else {
+                        // There could have been a shutdown inbetween us upgrading the weak and locking the mutex.
+                        continue;
+                    };
+                    assert!(Arc::ptr_eq(&removed, &strong));
+                }
             }
         }
     }
@@ -629,8 +474,6 @@ impl<T: Types> HandleInner<T> {
 
 #[cfg(test)]
 mod tests {
-    use std::sync::Weak;
-
     use pageserver_api::{
         key::{rel_block_to_key, Key, DBDIR_KEY},
         models::ShardParameters,
@@ -740,13 +583,39 @@ mod tests {
         //
         // fill the cache
         //
+        assert_eq!(
+            (Arc::strong_count(&shard0), Arc::weak_count(&shard0)),
+            (2, 1),
+            "strong: shard0, mgr; weak: myself"
+        );
+
         let handle: Handle<_> = cache
             .get(timeline_id, ShardSelector::Page(key), &mgr)
             .await
             .expect("we have the timeline");
+        let handle_inner_weak = Arc::downgrade(&handle.0);
         assert!(Weak::ptr_eq(&handle.myself, &shard0.myself));
+        assert_eq!(
+            (
+                Weak::strong_count(&handle_inner_weak),
+                Weak::weak_count(&handle_inner_weak)
+            ),
+            (2, 2),
+            "strong: handle, per_timeline_state, weak: handle_inner_weak, cache"
+        );
         assert_eq!(cache.map.len(), 1);
+
+        assert_eq!(
+            (Arc::strong_count(&shard0), Arc::weak_count(&shard0)),
+            (3, 1),
+            "strong: handleinner(per_timeline_state), shard0, mgr; weak: myself"
+        );
         drop(handle);
+        assert_eq!(
+            (Arc::strong_count(&shard0), Arc::weak_count(&shard0)),
+            (3, 1),
+            "strong: handleinner(per_timeline_state), shard0, mgr; weak: myself"
+        );
 
         //
         // demonstrate that Handle holds up gate closure
@@ -771,11 +640,21 @@ mod tests {
         // SHUTDOWN
         shard0.per_timeline_state.shutdown(); // keeping handle alive across shutdown
 
+        assert_eq!(
+            1,
+            Weak::strong_count(&handle_inner_weak),
+            "through local var handle"
+        );
         assert_eq!(
             cache.map.len(),
             1,
             "this is an implementation detail but worth pointing out: we can't clear the cache from shutdown(), it's cleared on first access after"
         );
+        assert_eq!(
+            (Arc::strong_count(&shard0), Arc::weak_count(&shard0)),
+            (3, 1),
+            "strong: handleinner(via handle), shard0, mgr; weak: myself"
+        );
 
         // this handle is perfectly usable
         handle.getpage();
@@ -799,6 +678,16 @@ mod tests {
         }
 
         drop(handle);
+        assert_eq!(
+            0,
+            Weak::strong_count(&handle_inner_weak),
+            "the HandleInner destructor already ran"
+        );
+        assert_eq!(
+            (Arc::strong_count(&shard0), Arc::weak_count(&shard0)),
+            (2, 1),
+            "strong: shard0, mgr; weak: myself"
+        );
 
         // closing gate succeeds after dropping handle
         tokio::select! {
@@ -817,8 +706,10 @@ mod tests {
         assert_eq!(cache.map.len(), 0);
 
         // ensure all refs to shard0 are gone and we're not leaking anything
+        let myself = Weak::clone(&shard0.myself);
         drop(shard0);
         drop(mgr);
+        assert_eq!(Weak::strong_count(&myself), 0);
     }
 
     #[tokio::test]
@@ -1057,11 +948,15 @@ mod tests {
                 handle
             };
             handle.getpage();
-            used_handles.push(Arc::downgrade(&handle.timeline));
+            used_handles.push(Arc::downgrade(&handle.0));
         }
 
-        // No handles exist, thus gates are closed and don't require shutdown.
-        // Thus the gate should close immediately, even without shutdown.
+        // No handles exist, thus gates are closed and don't require shutdown
+        assert!(used_handles
+            .iter()
+            .all(|weak| Weak::strong_count(weak) == 0));
+
+        // ... thus the gate should close immediately, even without shutdown
         tokio::select! {
             _ = shard0.gate.close() => { }
             _ = tokio::time::sleep(FOREVER) => {
@@ -1069,172 +964,4 @@ mod tests {
             }
         }
     }
-
-    #[tokio::test(start_paused = true)]
-    async fn test_weak_handles() {
-        crate::tenant::harness::setup_logging();
-        let timeline_id = TimelineId::generate();
-        let shard0 = Arc::new_cyclic(|myself| StubTimeline {
-            gate: Default::default(),
-            id: timeline_id,
-            shard: ShardIdentity::unsharded(),
-            per_timeline_state: PerTimelineState::default(),
-            myself: myself.clone(),
-        });
-        let mgr = StubManager {
-            shards: vec![shard0.clone()],
-        };
-
-        let refcount_start = Arc::strong_count(&shard0);
-
-        let key = DBDIR_KEY;
-
-        let mut cache = Cache::<TestTypes>::default();
-
-        let handle = cache
-            .get(timeline_id, ShardSelector::Page(key), &mgr)
-            .await
-            .expect("we have the timeline");
-        assert!(Weak::ptr_eq(&handle.myself, &shard0.myself));
-
-        let weak_handle = handle.downgrade();
-
-        drop(handle);
-
-        let upgraded_handle = weak_handle.upgrade().ok().expect("we can upgrade it");
-
-        // Start shutdown
-        shard0.per_timeline_state.shutdown();
-
-        // Upgrades during shutdown don't work, even if upgraded_handle exists.
-        weak_handle
-            .upgrade()
-            .err()
-            .expect("can't upgrade weak handle as soon as shutdown started");
-
-        // But upgraded_handle is still alive, so the gate won't close.
-        tokio::select! {
-            _ = shard0.gate.close() => {
-                panic!("handle is keeping gate open");
-            }
-            _ = tokio::time::sleep(FOREVER) => { }
-        }
-
-        // Drop the last handle.
-        drop(upgraded_handle);
-
-        // The gate should close now, despite there still being a weak_handle.
-        tokio::select! {
-            _ = shard0.gate.close() => { }
-            _ = tokio::time::sleep(FOREVER) => {
-                panic!("only strong handle is dropped and we shut down per-timeline-state")
-            }
-        }
-
-        // The weak handle still can't be upgraded.
-        weak_handle
-            .upgrade()
-            .err()
-            .expect("still shouldn't be able to upgrade the weak handle");
-
-        // There should be no strong references to the timeline object except the one on "stack".
-        assert_eq!(Arc::strong_count(&shard0), refcount_start);
-    }
-
-    #[tokio::test(start_paused = true)]
-    async fn test_reference_cycle_broken_when_cache_is_dropped() {
-        crate::tenant::harness::setup_logging();
-        let timeline_id = TimelineId::generate();
-        let shard0 = Arc::new_cyclic(|myself| StubTimeline {
-            gate: Default::default(),
-            id: timeline_id,
-            shard: ShardIdentity::unsharded(),
-            per_timeline_state: PerTimelineState::default(),
-            myself: myself.clone(),
-        });
-        let mgr = StubManager {
-            shards: vec![shard0.clone()],
-        };
-        let key = DBDIR_KEY;
-
-        let mut cache = Cache::<TestTypes>::default();
-
-        // helper to check if a handle is referenced by per_timeline_state
-        let per_timeline_state_refs_handle = |handle_weak: &Weak<Mutex<HandleInner<_>>>| {
-            let per_timeline_state = shard0.per_timeline_state.handles.lock().unwrap();
-            let per_timeline_state = per_timeline_state.as_ref().unwrap();
-            per_timeline_state
-                .values()
-                .any(|v| Weak::ptr_eq(&Arc::downgrade(v), handle_weak))
-        };
-
-        // Fill the cache.
-        let handle = cache
-            .get(timeline_id, ShardSelector::Page(key), &mgr)
-            .await
-            .expect("we have the timeline");
-        assert!(Weak::ptr_eq(&handle.myself, &shard0.myself));
-        let handle_inner_weak = Arc::downgrade(&handle.inner);
-        assert!(
-            per_timeline_state_refs_handle(&handle_inner_weak),
-            "we still hold `handle` _and_ haven't dropped `cache` yet"
-        );
-
-        // Drop the cache.
-        drop(cache);
-
-        assert!(
-            !(per_timeline_state_refs_handle(&handle_inner_weak)),
-            "nothing should reference the handle allocation anymore"
-        );
-        assert!(
-            Weak::upgrade(&handle_inner_weak).is_some(),
-            "the local `handle` still keeps the allocation alive"
-        );
-        // but obviously the cache is gone so no new allocations can be handed out.
-
-        // Drop handle.
-        drop(handle);
-        assert!(
-            Weak::upgrade(&handle_inner_weak).is_none(),
-            "the local `handle` is dropped, so the allocation should be dropped by now"
-        );
-    }
-
-    #[tokio::test(start_paused = true)]
-    async fn test_reference_cycle_broken_when_per_timeline_state_shutdown() {
-        crate::tenant::harness::setup_logging();
-        let timeline_id = TimelineId::generate();
-        let shard0 = Arc::new_cyclic(|myself| StubTimeline {
-            gate: Default::default(),
-            id: timeline_id,
-            shard: ShardIdentity::unsharded(),
-            per_timeline_state: PerTimelineState::default(),
-            myself: myself.clone(),
-        });
-        let mgr = StubManager {
-            shards: vec![shard0.clone()],
-        };
-        let key = DBDIR_KEY;
-
-        let mut cache = Cache::<TestTypes>::default();
-        let handle = cache
-            .get(timeline_id, ShardSelector::Page(key), &mgr)
-            .await
-            .expect("we have the timeline");
-        // grab a weak reference to the inner so can later try to Weak::upgrade it and assert that fails
-        let handle_inner_weak = Arc::downgrade(&handle.inner);
-
-        // drop the handle, obviously the lifetime of `inner` is at least as long as each strong reference to it
-        drop(handle);
-        assert!(Weak::upgrade(&handle_inner_weak).is_some(), "can still");
-
-        // Shutdown the per_timeline_state.
-        shard0.per_timeline_state.shutdown();
-        assert!(Weak::upgrade(&handle_inner_weak).is_none(), "can no longer");
-
-        // cache only contains Weak's, so, it can outlive the per_timeline_state without
-        // Drop explicitly solely to make this point.
-        drop(cache);
-    }
 }

pageserver/src/tenant/timeline/offload.rs

@@ -2,11 +2,10 @@ use std::sync::Arc;
 
 use pageserver_api::models::{TenantState, TimelineState};
 
-use super::delete::{delete_local_timeline_directory, DeletionGuard};
+use super::delete::{delete_local_timeline_directory, DeleteTimelineFlow, DeletionGuard};
 use super::Timeline;
 use crate::span::debug_assert_current_span_has_tenant_and_timeline_id;
 use crate::tenant::remote_timeline_client::ShutdownIfArchivedError;
-use crate::tenant::timeline::delete::{make_timeline_delete_guard, TimelineDeleteGuardKind};
 use crate::tenant::{OffloadedTimeline, Tenant, TenantManifestError, TimelineOrOffloaded};
 
 #[derive(thiserror::Error, Debug)]
@@ -37,10 +36,13 @@ pub(crate) async fn offload_timeline(
     debug_assert_current_span_has_tenant_and_timeline_id();
     tracing::info!("offloading archived timeline");
 
-    let (timeline, guard) = make_timeline_delete_guard(
+    let allow_offloaded_children = true;
+    let set_stopping = false;
+    let (timeline, guard) = DeleteTimelineFlow::prepare(
         tenant,
         timeline.timeline_id,
-        TimelineDeleteGuardKind::Offload,
+        allow_offloaded_children,
+        set_stopping,
     )
     .map_err(|e| OffloadError::Other(anyhow::anyhow!(e)))?;
 
@@ -104,7 +106,7 @@ pub(crate) async fn offload_timeline(
 }
 
 /// It is important that this gets called when DeletionGuard is being held.
-/// For more context see comments in [`make_timeline_delete_guard`]
+/// For more context see comments in [`DeleteTimelineFlow::prepare`]
 ///
 /// Returns the strong count of the timeline `Arc`
 fn remove_timeline_from_tenant(

pageserver/src/tenant/timeline/walreceiver/walreceiver_connection.rs

@@ -140,7 +140,7 @@ pub(super) async fn handle_walreceiver_connection(
 
     let (replication_client, connection) = {
         let mut config = wal_source_connconf.to_tokio_postgres_config();
-        config.application_name(format!("pageserver-{}", node.0).as_str());
+        config.application_name("pageserver");
         config.replication_mode(tokio_postgres::config::ReplicationMode::Physical);
         match time::timeout(connect_timeout, config.connect(postgres::NoTls)).await {
             Ok(client_and_conn) => client_and_conn?,

pgxn/neon/file_cache.c

@@ -911,85 +911,57 @@ lfc_writev(NRelFileInfo rinfo, ForkNumber forkNum, BlockNumber blkno,
 			if (entry->access_count++ == 0)
 				dlist_delete(&entry->list_node);
 		}
-		/*-----------
-		 * If the chunk wasn't already in the LFC then we have these
-		 * options, in order of preference:
-		 *
-		 * Unless there is no space available, we can:
-		 *  1. Use an entry from the `holes` list, and
-		 *  2. Create a new entry.
-		 * We can always, regardless of space in the LFC:
-		 *  3. evict an entry from LRU, and
-		 *  4. ignore the write operation (the least favorite option)
-		 */
-		else if (lfc_ctl->used < lfc_ctl->limit)
+		else
 		{
-			if (!dlist_is_empty(&lfc_ctl->holes))
+			/*
+			 * We have two choices if all cache pages are pinned (i.e. used in IO
+			 * operations):
+			 *
+			 * 1) Wait until some of this operation is completed and pages is
+			 * unpinned.
+			 *
+			 * 2) Allocate one more chunk, so that specified cache size is more
+			 * recommendation than hard limit.
+			 *
+			 * As far as probability of such event (that all pages are pinned) is
+			 * considered to be very very small: there are should be very large
+			 * number of concurrent IO operations and them are limited by
+			 * max_connections, we prefer not to complicate code and use second
+			 * approach.
+			 */
+			if (lfc_ctl->used >= lfc_ctl->limit && !dlist_is_empty(&lfc_ctl->lru))
+			{
+				/* Cache overflow: evict least recently used chunk */
+				FileCacheEntry *victim = dlist_container(FileCacheEntry, list_node, dlist_pop_head_node(&lfc_ctl->lru));
+	
+				for (int i = 0; i < BLOCKS_PER_CHUNK; i++)
+				{
+					lfc_ctl->used_pages -= (victim->bitmap[i >> 5] >> (i & 31)) & 1;
+				}
+				CriticalAssert(victim->access_count == 0);
+				entry->offset = victim->offset; /* grab victim's chunk */
+				hash_search_with_hash_value(lfc_hash, &victim->key, victim->hash, HASH_REMOVE, NULL);
+				neon_log(DEBUG2, "Swap file cache page");
+			}
+			else if (!dlist_is_empty(&lfc_ctl->holes))
 			{
 				/* We can reuse a hole that was left behind when the LFC was shrunk previously */
-				FileCacheEntry *hole = dlist_container(FileCacheEntry, list_node,
-													   dlist_pop_head_node(&lfc_ctl->holes));
-				uint32 offset = hole->offset;
-				bool hole_found;
-
-				hash_search_with_hash_value(lfc_hash, &hole->key,
-											hole->hash, HASH_REMOVE, &hole_found);
+				FileCacheEntry *hole = dlist_container(FileCacheEntry, list_node, dlist_pop_head_node(&lfc_ctl->holes));
+				uint32		offset = hole->offset;
+				bool		hole_found;
+	
+				hash_search_with_hash_value(lfc_hash, &hole->key, hole->hash, HASH_REMOVE, &hole_found);
 				CriticalAssert(hole_found);
-
+	
 				lfc_ctl->used += 1;
-				entry->offset = offset;			/* reuse the hole */
+				entry->offset = offset;	/* reuse the hole */
 			}
 			else
 			{
 				lfc_ctl->used += 1;
-				entry->offset = lfc_ctl->size++;/* allocate new chunk at end
-												 * of file */
+				entry->offset = lfc_ctl->size++;	/* allocate new chunk at end
+													 * of file */
 			}
-		}
-		/*
-		 * We've already used up all allocated LFC entries.
-		 *
-		 * If we can clear an entry from the LRU, do that.
-		 * If we can't (e.g. because all other slots are being accessed)
-		 * then we will remove this entry from the hash and continue
-		 * on to the next chunk, as we may not exceed the limit.
-		 */
-		else if (!dlist_is_empty(&lfc_ctl->lru))
-		{
-			/* Cache overflow: evict least recently used chunk */
-			FileCacheEntry *victim = dlist_container(FileCacheEntry, list_node,
-													 dlist_pop_head_node(&lfc_ctl->lru));
-
-			for (int i = 0; i < BLOCKS_PER_CHUNK; i++)
-			{
-				lfc_ctl->used_pages -= (victim->bitmap[i >> 5] >> (i & 31)) & 1;
-			}
-
-			CriticalAssert(victim->access_count == 0);
-			entry->offset = victim->offset; /* grab victim's chunk */
-			hash_search_with_hash_value(lfc_hash, &victim->key,
-										victim->hash, HASH_REMOVE, NULL);
-			neon_log(DEBUG2, "Swap file cache page");
-		}
-		else
-		{
-			/* Can't add this chunk - we don't have the space for it */
-			hash_search_with_hash_value(lfc_hash, &entry->key, hash,
-										HASH_REMOVE, NULL);
-
-			/*
-			 * We can't process this chunk due to lack of space in LFC,
-			 * so skip to the next one
-			 */
-			LWLockRelease(lfc_lock);
-			blkno += blocks_in_chunk;
-			buf_offset += blocks_in_chunk;
-			nblocks -= blocks_in_chunk;
-			continue;
-		}
-
-		if (!found)
-		{
 			entry->access_count = 1;
 			entry->hash = hash;
 			memset(entry->bitmap, 0, sizeof entry->bitmap);

pgxn/neon/libpagestore.c

@@ -373,8 +373,8 @@ pageserver_connect(shardno_t shard_no, int elevel)
 	{
 	case PS_Disconnected:
 	{
-		const char *keywords[4];
-		const char *values[4];
+		const char *keywords[3];
+		const char *values[3];
 		int			n_pgsql_params;
 		TimestampTz	now;
 		int64		us_since_last_attempt;
@@ -421,14 +421,12 @@ pageserver_connect(shardno_t shard_no, int elevel)
 		 */
 		keywords[0] = "dbname";
 		values[0] = connstr;
-		keywords[1] = "keepalives";
-		values[1] = "1";
-		n_pgsql_params = 2;
+		n_pgsql_params = 1;
 
 		if (neon_auth_token)
 		{
-			keywords[2] = "password";
-			values[2] = neon_auth_token;
+			keywords[1] = "password";
+			values[1] = neon_auth_token;
 			n_pgsql_params++;
 		}
 

pgxn/neon/pagestore_client.h

@@ -34,8 +34,6 @@ typedef enum
 	T_NeonGetPageRequest,
 	T_NeonDbSizeRequest,
 	T_NeonGetSlruSegmentRequest,
-	/* future tags above this line */
-	T_NeonTestRequest = 99, /* only in cfg(feature = "testing") */
 
 	/* pagestore -> pagestore_client */
 	T_NeonExistsResponse = 100,
@@ -44,8 +42,6 @@ typedef enum
 	T_NeonErrorResponse,
 	T_NeonDbSizeResponse,
 	T_NeonGetSlruSegmentResponse,
-	/* future tags above this line */
-	T_NeonTestResponse = 199, /* only in cfg(feature = "testing") */
 } NeonMessageTag;
 
 typedef uint64 NeonRequestId;

safekeeper/client/src/mgmt_api.rs

@@ -102,11 +102,6 @@ impl Client {
         self.get(&uri).await
     }
 
-    pub async fn utilization(&self) -> Result<reqwest::Response> {
-        let uri = format!("{}/v1/utilization/", self.mgmt_api_endpoint);
-        self.get(&uri).await
-    }
-
     async fn get<U: IntoUrl>(&self, uri: U) -> Result<reqwest::Response> {
         self.request(Method::GET, uri, ()).await
     }

safekeeper/src/http/routes.rs

@@ -127,13 +127,6 @@ async fn timeline_create_handler(mut request: Request<Body>) -> Result<Response<
     json_response(StatusCode::OK, ())
 }
 
-async fn utilization_handler(request: Request<Body>) -> Result<Response<Body>, ApiError> {
-    check_permission(&request, None)?;
-    let global_timelines = get_global_timelines(&request);
-    let utilization = global_timelines.get_timeline_counts();
-    json_response(StatusCode::OK, utilization)
-}
-
 /// List all (not deleted) timelines.
 /// Note: it is possible to do the same with debug_dump.
 async fn timeline_list_handler(request: Request<Body>) -> Result<Response<Body>, ApiError> {
@@ -627,7 +620,6 @@ pub fn make_router(
                 failpoints_handler(r, cancel).await
             })
         })
-        .get("/v1/uzilization", |r| request_span(r, utilization_handler))
         .delete("/v1/tenant/:tenant_id", |r| {
             request_span(r, tenant_delete_handler)
         })

safekeeper/src/timelines_global_map.rs

@@ -13,7 +13,6 @@ use anyhow::{bail, Context, Result};
 use camino::Utf8PathBuf;
 use camino_tempfile::Utf8TempDir;
 use safekeeper_api::membership::Configuration;
-use safekeeper_api::models::SafekeeperUtilization;
 use safekeeper_api::ServerInfo;
 use serde::Serialize;
 use std::collections::HashMap;
@@ -417,20 +416,6 @@ impl GlobalTimelines {
             .collect()
     }
 
-    /// Returns statistics about timeline counts
-    pub fn get_timeline_counts(&self) -> SafekeeperUtilization {
-        let global_lock = self.state.lock().unwrap();
-        let timeline_count = global_lock
-            .timelines
-            .values()
-            .filter(|t| match t {
-                GlobalMapTimeline::CreationInProgress => false,
-                GlobalMapTimeline::Timeline(t) => !t.is_cancelled(),
-            })
-            .count() as u64;
-        SafekeeperUtilization { timeline_count }
-    }
-
     /// Returns all timelines belonging to a given tenant. Used for deleting all timelines of a tenant,
     /// and that's why it can return cancelled timelines, to retry deleting them.
     fn get_all_for_tenant(&self, tenant_id: TenantId) -> Vec<Arc<Timeline>> {

storage_controller/migrations/2025-01-15-181207_safekeepers_disabled_to_pause/down.sql

@@ -1,2 +0,0 @@
-ALTER TABLE safekeepers ALTER COLUMN scheduling_policy SET DEFAULT 'disabled';
-UPDATE safekeepers SET scheduling_policy = 'disabled' WHERE scheduling_policy = 'pause';

storage_controller/migrations/2025-01-15-181207_safekeepers_disabled_to_pause/up.sql

@@ -1,2 +0,0 @@
-ALTER TABLE safekeepers ALTER COLUMN scheduling_policy SET DEFAULT 'pause';
-UPDATE safekeepers SET scheduling_policy = 'pause' WHERE scheduling_policy = 'disabled';

storage_controller/src/service.rs

@@ -5411,15 +5411,6 @@ impl Service {
 
         expect_shards.sort_by_key(|tsp| (tsp.tenant_id.clone(), tsp.shard_number, tsp.shard_count));
 
-        // Because JSON contents of persistent tenants might disagree with the fields in current `TenantConfig`
-        // definition, we will do an encode/decode cycle to ensure any legacy fields are dropped and any new
-        // fields are added, before doing a comparison.
-        for tsp in &mut persistent_shards {
-            let config: TenantConfig = serde_json::from_str(&tsp.config)
-                .map_err(|e| ApiError::InternalServerError(e.into()))?;
-            tsp.config = serde_json::to_string(&config).expect("Encoding config is infallible");
-        }
-
         if persistent_shards != expect_shards {
             tracing::error!("Consistency check failed on shards.");
 
@@ -7279,14 +7270,19 @@ impl Service {
         Ok(())
     }
 
-    /// Create a node fill plan (pick secondaries to promote), based on:
-    /// 1. Shards which have a secondary on this node, and this node is in their home AZ, and are currently attached to a node
-    ///    outside their home AZ, should be migrated back here.
-    /// 2. If after step 1 we have not migrated enough shards for this node to have its fair share of
-    ///    attached shards, we will promote more shards from the nodes with the most attached shards, unless
-    ///    those shards have a home AZ that doesn't match the node we're filling.
+    /// Create a node fill plan (pick secondaries to promote) that meets the following requirements:
+    /// 1. The node should be filled until it reaches the expected cluster average of
+    ///    attached shards. If there are not enough secondaries on the node, the plan stops early.
+    /// 2. Select tenant shards to promote such that the number of attached shards is balanced
+    ///    throughout the cluster. We achieve this by picking tenant shards from each node,
+    ///    starting from the ones with the largest number of attached shards, until the node
+    ///    reaches the expected cluster average.
+    /// 3. Avoid promoting more shards of the same tenant than required. The upper bound
+    ///    for the number of tenants from the same shard promoted to the node being filled is:
+    ///    shard count for the tenant divided by the number of nodes in the cluster.
     fn fill_node_plan(&self, node_id: NodeId) -> Vec<TenantShardId> {
         let mut locked = self.inner.write().unwrap();
+        let fill_requirement = locked.scheduler.compute_fill_requirement(node_id);
         let (nodes, tenants, _scheduler) = locked.parts_mut();
 
         let node_az = nodes
@@ -7295,79 +7291,53 @@ impl Service {
             .get_availability_zone_id()
             .clone();
 
-        // The tenant shard IDs that we plan to promote from secondary to attached on this node
-        let mut plan = Vec::new();
+        let mut tids_by_node = tenants
+            .iter_mut()
+            .filter_map(|(tid, tenant_shard)| {
+                if !matches!(
+                    tenant_shard.get_scheduling_policy(),
+                    ShardSchedulingPolicy::Active
+                ) {
+                    // Only include tenants in fills if they have a normal (Active) scheduling policy.  We
+                    // even exclude Essential, because moving to fill a node is not essential to keeping this
+                    // tenant available.
+                    return None;
+                }
 
-        // Collect shards which do not have a preferred AZ & are elegible for moving in stage 2
-        let mut free_tids_by_node: HashMap<NodeId, Vec<TenantShardId>> = HashMap::new();
-
-        // Don't respect AZ preferences if there is only one AZ.  This comes up in tests, but it could
-        // conceivably come up in real life if deploying a single-AZ region intentionally.
-        let respect_azs = nodes
-            .values()
-            .map(|n| n.get_availability_zone_id())
-            .unique()
-            .count()
-            > 1;
-
-        // Step 1: collect all shards that we are required to migrate back to this node because their AZ preference
-        // requires it.
-        for (tsid, tenant_shard) in tenants {
-            if !tenant_shard.intent.get_secondary().contains(&node_id) {
-                // Shard doesn't have a secondary on this node, ignore it.
-                continue;
-            }
-
-            // AZ check: when filling nodes after a restart, our intent is to move _back_ the
-            // shards which belong on this node, not to promote shards whose scheduling preference
-            // would be on their currently attached node.  So will avoid promoting shards whose
-            // home AZ doesn't match the AZ of the node we're filling.
-            match tenant_shard.preferred_az() {
-                _ if !respect_azs => {
-                    if let Some(primary) = tenant_shard.intent.get_attached() {
-                        free_tids_by_node.entry(*primary).or_default().push(*tsid);
+                // AZ check: when filling nodes after a restart, our intent is to move _back_ the
+                // shards which belong on this node, not to promote shards whose scheduling preference
+                // would be on their currently attached node.  So will avoid promoting shards whose
+                // home AZ doesn't match the AZ of the node we're filling.
+                match tenant_shard.preferred_az() {
+                    None => {
+                        // Shard doesn't have an AZ preference: it is elegible to be moved.
+                    }
+                    Some(az) if az == &node_az => {
+                        // This shard's home AZ is equal to the node we're filling: it is
+                        // elegible to be moved: fall through;
+                    }
+                    Some(_) => {
+                        // This shard's home AZ is somewhere other than the node we're filling:
+                        // do not include it in the fill plan.
+                        return None;
                     }
                 }
-                None => {
-                    // Shard doesn't have an AZ preference: it is elegible to be moved, but we
-                    // will only do so if our target shard count requires it.
-                    if let Some(primary) = tenant_shard.intent.get_attached() {
-                        free_tids_by_node.entry(*primary).or_default().push(*tsid);
-                    }
-                }
-                Some(az) if az == &node_az => {
-                    // This shard's home AZ is equal to the node we're filling: it should
-                    // be moved back to this node as part of filling, unless its currently
-                    // attached location is also in its home AZ.
-                    if let Some(primary) = tenant_shard.intent.get_attached() {
-                        if nodes
-                            .get(primary)
-                            .expect("referenced node must exist")
-                            .get_availability_zone_id()
-                            != tenant_shard
-                                .preferred_az()
-                                .expect("tenant must have an AZ preference")
-                        {
-                            plan.push(*tsid)
-                        }
-                    } else {
-                        plan.push(*tsid)
-                    }
-                }
-                Some(_) => {
-                    // This shard's home AZ is somewhere other than the node we're filling,
-                    // it may not be moved back to this node as part of filling.  Ignore it
-                }
-            }
-        }
 
-        // Step 2: also promote any AZ-agnostic shards as required to achieve the target number of attachments
-        let fill_requirement = locked.scheduler.compute_fill_requirement(node_id);
+                if tenant_shard.intent.get_secondary().contains(&node_id) {
+                    if let Some(primary) = tenant_shard.intent.get_attached() {
+                        return Some((*primary, *tid));
+                    }
+                }
+
+                None
+            })
+            .into_group_map();
 
         let expected_attached = locked.scheduler.expected_attached_shard_count();
         let nodes_by_load = locked.scheduler.nodes_by_attached_shard_count();
 
         let mut promoted_per_tenant: HashMap<TenantId, usize> = HashMap::new();
+        let mut plan = Vec::new();
 
         for (node_id, attached) in nodes_by_load {
             let available = locked.nodes.get(&node_id).is_some_and(|n| n.is_available());
@@ -7376,7 +7346,7 @@ impl Service {
             }
 
             if plan.len() >= fill_requirement
-                || free_tids_by_node.is_empty()
+                || tids_by_node.is_empty()
                 || attached <= expected_attached
             {
                 break;
@@ -7388,7 +7358,7 @@ impl Service {
 
             let mut remove_node = false;
             while take > 0 {
-                match free_tids_by_node.get_mut(&node_id) {
+                match tids_by_node.get_mut(&node_id) {
                     Some(tids) => match tids.pop() {
                         Some(tid) => {
                             let max_promote_for_tenant = std::cmp::max(
@@ -7414,7 +7384,7 @@ impl Service {
             }
 
             if remove_node {
-                free_tids_by_node.remove(&node_id);
+                tids_by_node.remove(&node_id);
             }
         }
 

storage_controller/src/service/chaos_injector.rs

@@ -1,17 +1,11 @@
-use std::{
-    collections::{BTreeMap, HashMap},
-    sync::Arc,
-    time::Duration,
-};
+use std::{sync::Arc, time::Duration};
 
 use pageserver_api::controller_api::ShardSchedulingPolicy;
 use rand::seq::SliceRandom;
 use rand::thread_rng;
 use tokio_util::sync::CancellationToken;
-use utils::id::NodeId;
-use utils::shard::TenantShardId;
 
-use super::{Node, Scheduler, Service, TenantShard};
+use super::Service;
 
 pub struct ChaosInjector {
     service: Arc<Service>,
@@ -41,86 +35,50 @@ impl ChaosInjector {
         }
     }
 
-    /// If a shard has a secondary and attached location, then re-assign the secondary to be
-    /// attached and the attached to be secondary.
-    ///
-    /// Only modifies tenants if they're in Active scheduling policy.
-    fn maybe_migrate_to_secondary(
-        &self,
-        tenant_shard_id: TenantShardId,
-        nodes: &Arc<HashMap<NodeId, Node>>,
-        tenants: &mut BTreeMap<TenantShardId, TenantShard>,
-        scheduler: &mut Scheduler,
-    ) {
-        let shard = tenants
-            .get_mut(&tenant_shard_id)
-            .expect("Held lock between choosing ID and this get");
-
-        if !matches!(shard.get_scheduling_policy(), ShardSchedulingPolicy::Active) {
-            // Skip non-active scheduling policies, so that a shard with a policy like Pause can
-            // be pinned without being disrupted by us.
-            tracing::info!(
-                "Skipping shard {tenant_shard_id}: scheduling policy is {:?}",
-                shard.get_scheduling_policy()
-            );
-            return;
-        }
-
-        // Pick a secondary to promote
-        let Some(new_location) = shard
-            .intent
-            .get_secondary()
-            .choose(&mut thread_rng())
-            .cloned()
-        else {
-            tracing::info!(
-                "Skipping shard {tenant_shard_id}: no secondary location, can't migrate"
-            );
-            return;
-        };
-
-        let Some(old_location) = *shard.intent.get_attached() else {
-            tracing::info!("Skipping shard {tenant_shard_id}: currently has no attached location");
-            return;
-        };
-
-        tracing::info!("Injecting chaos: migrate {tenant_shard_id} {old_location}->{new_location}");
-
-        shard.intent.demote_attached(scheduler, old_location);
-        shard.intent.promote_attached(scheduler, new_location);
-        self.service.maybe_reconcile_shard(shard, nodes);
-    }
-
     async fn inject_chaos(&mut self) {
         // Pick some shards to interfere with
         let batch_size = 128;
         let mut inner = self.service.inner.write().unwrap();
         let (nodes, tenants, scheduler) = inner.parts_mut();
         let tenant_ids = tenants.keys().cloned().collect::<Vec<_>>();
-
-        // Prefer to migrate tenants that are currently outside their home AZ.  This avoids the chaos injector
-        // continuously pushing tenants outside their home AZ: instead, we'll tend to cycle between picking some
-        // random tenants to move, and then on next chaos iteration moving them back, then picking some new
-        // random tenants on the next iteration.
-        let mut victims = Vec::with_capacity(batch_size);
-        for shard in tenants.values() {
-            if shard.is_attached_outside_preferred_az(nodes) {
-                victims.push(shard.tenant_shard_id);
-            }
-
-            if victims.len() >= batch_size {
-                break;
-            }
-        }
-
-        let choose_random = batch_size.saturating_sub(victims.len());
-        tracing::info!("Injecting chaos: found {} shards to migrate back to home AZ, picking {choose_random} random shards to migrate", victims.len());
-
-        let random_victims = tenant_ids.choose_multiple(&mut thread_rng(), choose_random);
-        victims.extend(random_victims);
+        let victims = tenant_ids.choose_multiple(&mut thread_rng(), batch_size);
 
         for victim in victims {
-            self.maybe_migrate_to_secondary(victim, nodes, tenants, scheduler);
+            let shard = tenants
+                .get_mut(victim)
+                .expect("Held lock between choosing ID and this get");
+
+            if !matches!(shard.get_scheduling_policy(), ShardSchedulingPolicy::Active) {
+                // Skip non-active scheduling policies, so that a shard with a policy like Pause can
+                // be pinned without being disrupted by us.
+                tracing::info!(
+                    "Skipping shard {victim}: scheduling policy is {:?}",
+                    shard.get_scheduling_policy()
+                );
+                continue;
+            }
+
+            // Pick a secondary to promote
+            let Some(new_location) = shard
+                .intent
+                .get_secondary()
+                .choose(&mut thread_rng())
+                .cloned()
+            else {
+                tracing::info!("Skipping shard {victim}: no secondary location, can't migrate");
+                continue;
+            };
+
+            let Some(old_location) = *shard.intent.get_attached() else {
+                tracing::info!("Skipping shard {victim}: currently has no attached location");
+                continue;
+            };
+
+            tracing::info!("Injecting chaos: migrate {victim} {old_location}->{new_location}");
+
+            shard.intent.demote_attached(scheduler, old_location);
+            shard.intent.promote_attached(scheduler, new_location);
+            self.service.maybe_reconcile_shard(shard, nodes);
         }
     }
 }

storage_controller/src/tenant_shard.rs

@@ -1793,23 +1793,6 @@ impl TenantShard {
             }
         }
     }
-
-    /// Returns true if the tenant shard is attached to a node that is outside the preferred AZ.
-    ///
-    /// If the shard does not have a preferred AZ, returns false.
-    pub(crate) fn is_attached_outside_preferred_az(&self, nodes: &HashMap<NodeId, Node>) -> bool {
-        self.intent
-            .get_attached()
-            .map(|node_id| {
-                Some(
-                    nodes
-                        .get(&node_id)
-                        .expect("referenced node exists")
-                        .get_availability_zone_id(),
-                ) == self.intent.preferred_az_id.as_ref()
-            })
-            .unwrap_or(false)
-    }
 }
 
 impl Drop for TenantShard {

storage_scrubber/src/pageserver_physical_gc.rs

@@ -8,8 +8,6 @@ use crate::checks::{
 };
 use crate::metadata_stream::{stream_tenant_timelines, stream_tenants};
 use crate::{init_remote, BucketConfig, NodeKind, RootTarget, TenantShardTimelineId, MAX_RETRIES};
-use async_stream::try_stream;
-use futures::future::Either;
 use futures_util::{StreamExt, TryStreamExt};
 use pageserver::tenant::remote_timeline_client::index::LayerFileMetadata;
 use pageserver::tenant::remote_timeline_client::manifest::OffloadedTimelineManifest;
@@ -580,7 +578,7 @@ async fn gc_timeline(
     target: &RootTarget,
     mode: GcMode,
     ttid: TenantShardTimelineId,
-    accumulator: &std::sync::Mutex<TenantRefAccumulator>,
+    accumulator: &Arc<std::sync::Mutex<TenantRefAccumulator>>,
     tenant_manifest_info: Arc<Option<RemoteTenantManifestInfo>>,
 ) -> anyhow::Result<GcSummary> {
     let mut summary = GcSummary::default();
@@ -723,9 +721,9 @@ pub async fn pageserver_physical_gc(
 
     let remote_client = Arc::new(remote_client);
     let tenants = if tenant_shard_ids.is_empty() {
-        Either::Left(stream_tenants(&remote_client, &target))
+        futures::future::Either::Left(stream_tenants(&remote_client, &target))
     } else {
-        Either::Right(futures::stream::iter(tenant_shard_ids.into_iter().map(Ok)))
+        futures::future::Either::Right(futures::stream::iter(tenant_shard_ids.into_iter().map(Ok)))
     };
 
     // How many tenants to process in parallel.  We need to be mindful of pageservers
@@ -733,16 +731,16 @@ pub async fn pageserver_physical_gc(
     const CONCURRENCY: usize = 32;
 
     // Accumulate information about each tenant for cross-shard GC step we'll do at the end
-    let accumulator = std::sync::Mutex::new(TenantRefAccumulator::default());
-
-    // Accumulate information about how many manifests we have GCd
-    let manifest_gc_summary = std::sync::Mutex::new(GcSummary::default());
+    let accumulator = Arc::new(std::sync::Mutex::new(TenantRefAccumulator::default()));
 
     // Generate a stream of TenantTimelineId
+    enum GcSummaryOrContent<T> {
+        Content(T),
+        GcSummary(GcSummary),
+    }
     let timelines = tenants.map_ok(|tenant_shard_id| {
         let target_ref = &target;
         let remote_client_ref = &remote_client;
-        let manifest_gc_summary_ref = &manifest_gc_summary;
         async move {
             let gc_manifest_result = gc_tenant_manifests(
                 remote_client_ref,
@@ -759,48 +757,55 @@ pub async fn pageserver_physical_gc(
                     (GcSummary::default(), None)
                 }
             };
-            manifest_gc_summary_ref
-                .lock()
-                .unwrap()
-                .merge(summary_from_manifest);
             let tenant_manifest_arc = Arc::new(tenant_manifest_opt);
-            let mut timelines = Box::pin(
-                stream_tenant_timelines(remote_client_ref, target_ref, tenant_shard_id).await?,
-            );
-            Ok(try_stream! {
-                while let Some(ttid_res) = timelines.next().await {
-                    let ttid = ttid_res?;
-                    yield (ttid, tenant_manifest_arc.clone());
-                }
-            })
+            let summary_from_manifest = Ok(GcSummaryOrContent::<(_, _)>::GcSummary(
+                summary_from_manifest,
+            ));
+            stream_tenant_timelines(remote_client_ref, target_ref, tenant_shard_id)
+                .await
+                .map(|stream| {
+                    stream
+                        .zip(futures::stream::iter(std::iter::repeat(
+                            tenant_manifest_arc,
+                        )))
+                        .map(|(ttid_res, tenant_manifest_arc)| {
+                            ttid_res.map(move |ttid| {
+                                GcSummaryOrContent::Content((ttid, tenant_manifest_arc))
+                            })
+                        })
+                        .chain(futures::stream::iter([summary_from_manifest].into_iter()))
+                })
         }
     });
+    let timelines = std::pin::pin!(timelines.try_buffered(CONCURRENCY));
+    let timelines = timelines.try_flatten();
 
     let mut summary = GcSummary::default();
-    {
-        let timelines = std::pin::pin!(timelines.try_buffered(CONCURRENCY));
-        let timelines = timelines.try_flatten();
 
-        let timelines = timelines.map_ok(|(ttid, tenant_manifest_arc)| {
-            gc_timeline(
-                &remote_client,
-                &min_age,
-                &target,
-                mode,
-                ttid,
-                &accumulator,
-                tenant_manifest_arc,
-            )
+    // Drain futures for per-shard GC, populating accumulator as a side effect
+    {
+        let timelines = timelines.map_ok(|summary_or_ttid| match summary_or_ttid {
+            GcSummaryOrContent::Content((ttid, tenant_manifest_arc)) => {
+                futures::future::Either::Left(gc_timeline(
+                    &remote_client,
+                    &min_age,
+                    &target,
+                    mode,
+                    ttid,
+                    &accumulator,
+                    tenant_manifest_arc,
+                ))
+            }
+            GcSummaryOrContent::GcSummary(gc_summary) => {
+                futures::future::Either::Right(futures::future::ok(gc_summary))
+            }
         });
         let mut timelines = std::pin::pin!(timelines.try_buffered(CONCURRENCY));
 
-        // Drain futures for per-shard GC, populating accumulator as a side effect
         while let Some(i) = timelines.next().await {
             summary.merge(i?);
         }
     }
-    // Streams are lazily evaluated, so only now do we have access to the inner object
-    summary.merge(manifest_gc_summary.into_inner().unwrap());
 
     // Execute cross-shard GC, using the accumulator's full view of all the shards built in the per-shard GC
     let Some(client) = controller_client else {
@@ -808,7 +813,8 @@ pub async fn pageserver_physical_gc(
         return Ok(summary);
     };
 
-    let (ancestor_shards, ancestor_refs) = accumulator
+    let (ancestor_shards, ancestor_refs) = Arc::into_inner(accumulator)
+        .unwrap()
         .into_inner()
         .unwrap()
         .into_gc_ancestors(client, &mut summary)

test_runner/fixtures/neon_fixtures.py

@@ -370,7 +370,6 @@ class NeonEnvBuilder:
         pageserver_config_override: str | Callable[[dict[str, Any]], None] | None = None,
         num_safekeepers: int = 1,
         num_pageservers: int = 1,
-        num_azs: int = 1,
         # Use non-standard SK ids to check for various parsing bugs
         safekeepers_id_start: int = 0,
         # fsync is disabled by default to make the tests go faster
@@ -402,7 +401,6 @@ class NeonEnvBuilder:
         self.pageserver_config_override = pageserver_config_override
         self.num_safekeepers = num_safekeepers
         self.num_pageservers = num_pageservers
-        self.num_azs = num_azs
         self.safekeepers_id_start = safekeepers_id_start
         self.safekeepers_enable_fsync = safekeepers_enable_fsync
         self.auth_enabled = auth_enabled
@@ -992,7 +990,6 @@ class NeonEnv:
         self.endpoints = EndpointFactory(self)
         self.safekeepers: list[Safekeeper] = []
         self.pageservers: list[NeonPageserver] = []
-        self.num_azs = config.num_azs
         self.broker = NeonBroker(self)
         self.pageserver_remote_storage = config.pageserver_remote_storage
         self.safekeepers_remote_storage = config.safekeepers_remote_storage
@@ -1093,21 +1090,14 @@ class NeonEnv:
                 http=self.port_distributor.get_port(),
             )
 
-            # Availabilty zones may also be configured manually with `NeonEnvBuilder.pageserver_config_override`
-            if self.num_azs > 1:
-                # Round-robin assignment of AZ names like us-east-2a, us-east-2b, etc.
-                az_prefix = DEFAULT_AZ_ID[:-1]
-                availability_zone = f"{az_prefix}{chr(ord('a') + (ps_id - 1) % self.num_azs)}"
-            else:
-                availability_zone = DEFAULT_AZ_ID
-
             ps_cfg: dict[str, Any] = {
                 "id": ps_id,
                 "listen_pg_addr": f"localhost:{pageserver_port.pg}",
                 "listen_http_addr": f"localhost:{pageserver_port.http}",
                 "pg_auth_type": pg_auth_type,
                 "http_auth_type": http_auth_type,
-                "availability_zone": availability_zone,
+                # Default which can be overriden with `NeonEnvBuilder.pageserver_config_override`
+                "availability_zone": DEFAULT_AZ_ID,
                 # Disable pageserver disk syncs in tests: when running tests concurrently, this avoids
                 # the pageserver taking a long time to start up due to syncfs flushing other tests' data
                 "no_sync": True,

test_runner/regress/test_attach_tenant_config.py

@@ -176,7 +176,6 @@ def test_fully_custom_config(positive_env: NeonEnv):
             "type": "interpreted",
             "args": {"format": "bincode", "compression": {"zstd": {"level": 1}}},
         },
-        "rel_size_v2_enabled": True,
     }
 
     vps_http = env.storage_controller.pageserver_api()

test_runner/regress/test_local_file_cache.py

@@ -7,78 +7,9 @@ import threading
 import time
 
 import pytest
-from fixtures.neon_fixtures import NeonEnv, NeonEnvBuilder
+from fixtures.neon_fixtures import NeonEnvBuilder
 from fixtures.utils import USE_LFC, query_scalar
 
-"""
-Test whether LFC doesn't error out when the LRU is empty, but the LFC is
-already at its maximum size.
-
-If we don't handle this safely, we might allocate more hash entries than
-otherwise considered safe, thus causing ERRORs in hash_search(HASH_ENTER) once
-we hit lfc->used >= lfc->limit.
-"""
-
-
-@pytest.mark.skipif(not USE_LFC, reason="LFC is disabled, skipping")
-def test_local_file_cache_all_pinned(neon_simple_env: NeonEnv):
-    env = neon_simple_env
-    endpoint = env.endpoints.create_start(
-        "main",
-        config_lines=[
-            "neon.max_file_cache_size='1MB'",
-            "neon.file_cache_size_limit='1MB'",
-        ],
-    )
-    top_cur = endpoint.connect().cursor()
-
-    stop = threading.Event()
-    n_rows = 10000
-    n_threads = 5
-    n_updates_per_connection = 1000
-
-    top_cur.execute("CREATE TABLE lfctest (id int4 PRIMARY KEY, n int) WITH (fillfactor=10)")
-    top_cur.execute(f"INSERT INTO lfctest SELECT g, 1 FROM generate_series(1, {n_rows}) g")
-
-    # Start threads that will perform random UPDATEs. Each UPDATE
-    # increments the counter on the row, so that we can check at the
-    # end that the sum of all the counters match the number of updates
-    # performed (plus the initial 1 on each row).
-    #
-    # Furthermore, each thread will reconnect between every 1000 updates.
-    def run_updates(n_updates_performed_q: queue.Queue[int]):
-        n_updates_performed = 0
-        conn = endpoint.connect()
-        cur = conn.cursor()
-        while not stop.is_set():
-            id = random.randint(1, n_rows)
-            cur.execute(f"UPDATE lfctest SET n = n + 1 WHERE id = {id}")
-            n_updates_performed += 1
-            if n_updates_performed % n_updates_per_connection == 0:
-                cur.close()
-                conn.close()
-                conn = endpoint.connect()
-                cur = conn.cursor()
-        n_updates_performed_q.put(n_updates_performed)
-
-    n_updates_performed_q: queue.Queue[int] = queue.Queue()
-    threads: list[threading.Thread] = []
-    for _i in range(n_threads):
-        thread = threading.Thread(target=run_updates, args=(n_updates_performed_q,), daemon=True)
-        thread.start()
-        threads.append(thread)
-
-    time.sleep(15)
-
-    stop.set()
-
-    n_updates_performed = 0
-    for thread in threads:
-        thread.join()
-        n_updates_performed += n_updates_performed_q.get()
-
-    assert query_scalar(top_cur, "SELECT SUM(n) FROM lfctest") == n_rows + n_updates_performed
-
 
 @pytest.mark.skipif(not USE_LFC, reason="LFC is disabled, skipping")
 def test_local_file_cache_unlink(neon_env_builder: NeonEnvBuilder):

test_runner/regress/test_page_service_batching_regressions.py

@@ -1,60 +0,0 @@
-# NB: there are benchmarks that double-serve as tests inside the `performance` directory.
-
-import subprocess
-from pathlib import Path
-
-import pytest
-from fixtures.log_helper import log
-from fixtures.neon_fixtures import NeonEnvBuilder
-
-
-@pytest.mark.timeout(30)  # test takes <20s if pageserver impl is correct
-@pytest.mark.parametrize("kind", ["pageserver-stop", "tenant-detach"])
-def test_slow_flush(neon_env_builder: NeonEnvBuilder, neon_binpath: Path, kind: str):
-    def patch_pageserver_toml(config):
-        config["page_service_pipelining"] = {
-            "mode": "pipelined",
-            "max_batch_size": 32,
-            "execution": "concurrent-futures",
-        }
-
-    neon_env_builder.pageserver_config_override = patch_pageserver_toml
-    env = neon_env_builder.init_start()
-
-    log.info("make flush appear slow")
-
-    log.info("sending requests until pageserver accepts no more")
-    # TODO: extract this into a helper, like subprocess_capture,
-    # so that we capture the stderr from the helper somewhere.
-    child = subprocess.Popen(
-        [
-            neon_binpath / "test_helper_slow_client_reads",
-            env.pageserver.connstr(),
-            str(env.initial_tenant),
-            str(env.initial_timeline),
-        ],
-        bufsize=0,  # unbuffered
-        stdin=subprocess.PIPE,
-        stdout=subprocess.PIPE,
-    )
-    assert child.stdout is not None
-    buf = child.stdout.read(1)
-    if len(buf) != 1:
-        raise Exception("unexpected EOF")
-    if buf != b"R":
-        raise Exception(f"unexpected data: {buf!r}")
-    log.info("helper reports pageserver accepts no more requests")
-    log.info(
-        "assuming pageserver connection handle is in a state where TCP has backpressured pageserver=>client response flush() into userspace"
-    )
-
-    if kind == "pageserver-stop":
-        log.info("try to shut down the pageserver cleanly")
-        env.pageserver.stop()
-    elif kind == "tenant-detach":
-        log.info("try to shut down the tenant")
-        env.pageserver.tenant_detach(env.initial_tenant)
-    else:
-        raise ValueError(f"unexpected kind: {kind}")
-
-    log.info("shutdown did not time out, test passed")

test_runner/regress/test_storage_controller.py

@@ -2394,7 +2394,6 @@ def test_storage_controller_node_deletion(
     Test that deleting a node works & properly reschedules everything that was on the node.
     """
     neon_env_builder.num_pageservers = 3
-    neon_env_builder.num_azs = 3
     env = neon_env_builder.init_configs()
     env.start()
 
@@ -2408,9 +2407,6 @@ def test_storage_controller_node_deletion(
             tid, placement_policy='{"Attached":1}', shard_count=shard_count_per_tenant
         )
 
-    # Sanity check: initial creations should not leave the system in an unstable scheduling state
-    assert env.storage_controller.reconcile_all() == 0
-
     victim = env.pageservers[-1]
 
     # The procedure a human would follow is:
@@ -3215,7 +3211,7 @@ def test_safekeeper_deployment_time_update(neon_env_builder: NeonEnvBuilder):
     # some small tests for the scheduling policy querying and returning APIs
     newest_info = target.get_safekeeper(inserted["id"])
     assert newest_info
-    assert newest_info["scheduling_policy"] == "Pause"
+    assert newest_info["scheduling_policy"] == "Disabled"
     target.safekeeper_scheduling_policy(inserted["id"], "Decomissioned")
     newest_info = target.get_safekeeper(inserted["id"])
     assert newest_info

test_runner/regress/test_tenant_delete.py

@@ -1,7 +1,6 @@
 from __future__ import annotations
 
 import json
-from concurrent.futures import ThreadPoolExecutor
 from threading import Thread
 
 import pytest
@@ -254,8 +253,29 @@ def test_tenant_delete_races_timeline_creation(neon_env_builder: NeonEnvBuilder)
     ps_http.configure_failpoints((BEFORE_INITDB_UPLOAD_FAILPOINT, "pause"))
 
     def timeline_create():
-        ps_http.timeline_create(env.pg_version, tenant_id, TimelineId.generate(), timeout=1)
-        raise RuntimeError("creation succeeded even though it shouldn't")
+        try:
+            ps_http.timeline_create(env.pg_version, tenant_id, TimelineId.generate(), timeout=1)
+            raise RuntimeError("creation succeeded even though it shouldn't")
+        except ReadTimeout:
+            pass
+
+    Thread(target=timeline_create).start()
+
+    def hit_initdb_upload_failpoint():
+        env.pageserver.assert_log_contains(f"at failpoint {BEFORE_INITDB_UPLOAD_FAILPOINT}")
+
+    wait_until(hit_initdb_upload_failpoint)
+
+    def creation_connection_timed_out():
+        env.pageserver.assert_log_contains(
+            "POST.*/timeline.* request was dropped before completing"
+        )
+
+    # Wait so that we hit the timeout and the connection is dropped
+    # (But timeline creation still continues)
+    wait_until(creation_connection_timed_out)
+
+    ps_http.configure_failpoints((DELETE_BEFORE_CLEANUP_FAILPOINT, "pause"))
 
     def tenant_delete():
         def tenant_delete_inner():
@@ -263,46 +283,21 @@ def test_tenant_delete_races_timeline_creation(neon_env_builder: NeonEnvBuilder)
 
         wait_until(tenant_delete_inner)
 
-    # We will spawn background threads for timeline creation and tenant deletion.  They will both
-    # get blocked on our failpoint.
-    with ThreadPoolExecutor(max_workers=1) as executor:
-        create_fut = executor.submit(timeline_create)
+    Thread(target=tenant_delete).start()
 
-        def hit_initdb_upload_failpoint():
-            env.pageserver.assert_log_contains(f"at failpoint {BEFORE_INITDB_UPLOAD_FAILPOINT}")
+    def deletion_arrived():
+        env.pageserver.assert_log_contains(
+            f"cfg failpoint: {DELETE_BEFORE_CLEANUP_FAILPOINT} pause"
+        )
 
-        wait_until(hit_initdb_upload_failpoint)
+    wait_until(deletion_arrived)
 
-        def creation_connection_timed_out():
-            env.pageserver.assert_log_contains(
-                "POST.*/timeline.* request was dropped before completing"
-            )
+    ps_http.configure_failpoints((DELETE_BEFORE_CLEANUP_FAILPOINT, "off"))
 
-        # Wait so that we hit the timeout and the connection is dropped
-        # (But timeline creation still continues)
-        wait_until(creation_connection_timed_out)
+    # Disable the failpoint and wait for deletion to finish
+    ps_http.configure_failpoints((BEFORE_INITDB_UPLOAD_FAILPOINT, "off"))
 
-        with pytest.raises(ReadTimeout):
-            # Our creation failed from the client's point of view.
-            create_fut.result()
-
-        ps_http.configure_failpoints((DELETE_BEFORE_CLEANUP_FAILPOINT, "pause"))
-
-        delete_fut = executor.submit(tenant_delete)
-
-        def deletion_arrived():
-            env.pageserver.assert_log_contains(
-                f"cfg failpoint: {DELETE_BEFORE_CLEANUP_FAILPOINT} pause"
-            )
-
-        wait_until(deletion_arrived)
-
-        ps_http.configure_failpoints((DELETE_BEFORE_CLEANUP_FAILPOINT, "off"))
-
-        # Disable the failpoint and wait for deletion to finish
-        ps_http.configure_failpoints((BEFORE_INITDB_UPLOAD_FAILPOINT, "off"))
-
-        delete_fut.result()
+    ps_http.tenant_delete(tenant_id)
 
     # Physical deletion should have happened
     assert_prefix_empty(

test_runner/regress/test_tenants.py

@@ -194,7 +194,7 @@ def test_metrics_normal_work(neon_env_builder: NeonEnvBuilder):
         io_metrics = query_all_safekeepers(
             "safekeeper_pg_io_bytes_total",
             {
-                "app_name": f"pageserver-{env.pageserver.id}",
+                "app_name": "pageserver",
                 "client_az": "test_ps_az",
                 "dir": io_direction,
                 "same_az": "false",