use flag-bearer

remove dynamic limiter internal Box<dyn>
proxy: simplify dynamic limiter impl
2026-05-16 20:50:37 +00:00 · 2025-04-02 12:51:27 +01:00 · 2025-04-02 12:45:04 +01:00 · 2025-04-02 12:40:45 +01:00
214 changed files with 3998 additions and 7981 deletions
--- a/.dockerignore
+++ b/.dockerignore
@@ -19,7 +19,6 @@
 !pageserver/
 !pgxn/
 !proxy/
-!object_storage/
 !storage_scrubber/
 !safekeeper/
 !storage_broker/
--- a/.github/scripts/generate_image_maps.py
+++ b/.github/scripts/generate_image_maps.py
@@ -39,18 +39,12 @@ registries = {
    ],
 }

-release_branches = ["release", "release-proxy", "release-compute"]
-
 outputs: dict[str, dict[str, list[str]]] = {}

-target_tags = (
-    [target_tag, "latest"]
-    if branch == "main"
-    else [target_tag, "released"]
-    if branch in release_branches
-    else [target_tag]
+target_tags = [target_tag, "latest"] if branch == "main" else [target_tag]
+target_stages = (
+    ["dev", "prod"] if branch in ["release", "release-proxy", "release-compute"] else ["dev"]
 )
-target_stages = ["dev", "prod"] if branch in release_branches else ["dev"]

 for component_name, component_images in components.items():
    for stage in target_stages:
--- a/.github/scripts/push_with_image_map.py
+++ b/.github/scripts/push_with_image_map.py
@@ -2,9 +2,6 @@ import json
 import os
 import subprocess

-RED = "\033[91m"
-RESET = "\033[0m"
-
 image_map = os.getenv("IMAGE_MAP")
 if not image_map:
    raise ValueError("IMAGE_MAP environment variable is not set")
@@ -14,32 +11,12 @@ try:
 except json.JSONDecodeError as e:
    raise ValueError("Failed to parse IMAGE_MAP as JSON") from e

-failures = []
+for source, targets in parsed_image_map.items():
+    for target in targets:
+        cmd = ["docker", "buildx", "imagetools", "create", "-t", target, source]
+        print(f"Running: {' '.join(cmd)}")
+        result = subprocess.run(cmd, text=True, stdout=subprocess.PIPE, stderr=subprocess.STDOUT)

-pending = [(source, target) for source, targets in parsed_image_map.items() for target in targets]
-
-while len(pending) > 0:
-    if len(failures) > 10:
-        print("Error: more than 10 failures!")
-        for failure in failures:
-            print(f'"{failure[0]}" failed with the following output:')
-            print(failure[1])
-        raise RuntimeError("Retry limit reached.")
-
-    source, target = pending.pop(0)
-    cmd = ["docker", "buildx", "imagetools", "create", "-t", target, source]
-    print(f"Running: {' '.join(cmd)}")
-    result = subprocess.run(cmd, text=True, stdout=subprocess.PIPE, stderr=subprocess.STDOUT)
-
-    if result.returncode != 0:
-        failures.append((" ".join(cmd), result.stdout, target))
-        pending.append((source, target))
-        print(
-            f"{RED}[RETRY]{RESET} Push failed for {target}. Retrying... (failure count: {len(failures)})"
-        )
-        print(result.stdout)
-
-if len(failures) > 0 and (github_output := os.getenv("GITHUB_OUTPUT")):
-    failed_targets = [target for _, _, target in failures]
-    with open(github_output, "a") as f:
-        f.write(f"push_failures={json.dumps(failed_targets)}\n")
+        if result.returncode != 0:
+            print(f"Error: {result.stdout}")
+            raise RuntimeError(f"Command failed: {' '.join(cmd)}")
--- a/.github/workflows/_push-to-container-registry.yml
+++ b/.github/workflows/_push-to-container-registry.yml
@@ -104,25 +104,6 @@ jobs:
          password: ${{ secrets.NEON_DOCKERHUB_PASSWORD }}

      - name: Copy docker images to target registries
-        id: push
        run: python3 .github/scripts/push_with_image_map.py
        env:
          IMAGE_MAP: ${{ inputs.image-map }}
-
-      - name: Notify Slack if container image pushing fails
-        if: steps.push.outputs.push_failures || failure()
-        uses: slackapi/slack-github-action@485a9d42d3a73031f12ec201c457e2162c45d02d # v2.0.0
-        with:
-          method: chat.postMessage
-          token: ${{ secrets.SLACK_BOT_TOKEN }}
-          payload: |
-            channel: ${{ vars.SLACK_ON_CALL_DEVPROD_STREAM }}
-            text: >
-              *Container image pushing ${{
-                steps.push.outcome == 'failure' && 'failed completely' || 'succeeded with some retries'
-              }}* in
-              <${{ github.server_url }}/${{ github.repository }}/actions/runs/${{ github.run_id }}|GitHub Run>
-
-              ${{ steps.push.outputs.push_failures && format(
-                '*Failed targets:*\n• {0}', join(fromJson(steps.push.outputs.push_failures), '\n• ')
-              ) || '' }}
--- a/.github/workflows/build_and_test.yml
+++ b/.github/workflows/build_and_test.yml
@@ -89,8 +89,8 @@ jobs:

  check-codestyle-python:
    needs: [ meta, check-permissions, build-build-tools-image ]
-    # No need to run on `main` because we this in the merge queue. We do need to run this in `.*-rc-pr` because of hotfixes.
-    if: ${{ contains(fromJSON('["pr", "storage-rc-pr", "proxy-rc-pr", "compute-rc-pr"]'), needs.meta.outputs.run-kind) }}
+    # No need to run on `main` because we this in the merge queue
+    if: ${{ needs.meta.outputs.run-kind == 'pr' }}
    uses: ./.github/workflows/_check-codestyle-python.yml
    with:
      build-tools-image: ${{ needs.build-build-tools-image.outputs.image }}-bookworm
@@ -98,8 +98,7 @@ jobs:

  check-codestyle-jsonnet:
    needs: [ meta, check-permissions, build-build-tools-image ]
-    # We do need to run this in `.*-rc-pr` because of hotfixes.
-    if: ${{ contains(fromJSON('["pr", "push-main", "storage-rc-pr", "proxy-rc-pr", "compute-rc-pr"]'), needs.meta.outputs.run-kind) }}
+    if: ${{ contains(fromJSON('["pr", "push-main"]'), needs.meta.outputs.run-kind) }}
    runs-on: [ self-hosted, small ]
    container:
      image: ${{ needs.build-build-tools-image.outputs.image }}
@@ -182,8 +181,8 @@ jobs:

  check-codestyle-rust:
    needs: [ meta, check-permissions, build-build-tools-image ]
-    # No need to run on `main` because we this in the merge queue. We do need to run this in `.*-rc-pr` because of hotfixes.
-    if: ${{ contains(fromJSON('["pr", "storage-rc-pr", "proxy-rc-pr", "compute-rc-pr"]'), needs.meta.outputs.run-kind) }}
+    # No need to run on `main` because we this in the merge queue
+    if: ${{ needs.meta.outputs.run-kind == 'pr' }}
    uses: ./.github/workflows/_check-codestyle-rust.yml
    with:
      build-tools-image: ${{ needs.build-build-tools-image.outputs.image }}-bookworm
@@ -192,8 +191,7 @@ jobs:

  check-dependencies-rust:
    needs: [ meta, files-changed, build-build-tools-image ]
-    # No need to run on `main` because we this in the merge queue. We do need to run this in `.*-rc-pr` because of hotfixes.
-    if: ${{ needs.files-changed.outputs.check-rust-dependencies == 'true' && contains(fromJSON('["pr", "storage-rc-pr", "proxy-rc-pr", "compute-rc-pr"]'), needs.meta.outputs.run-kind) }}
+    if: ${{ needs.files-changed.outputs.check-rust-dependencies == 'true' && needs.meta.outputs.run-kind == 'pr' }}
    uses: ./.github/workflows/cargo-deny.yml
    with:
      build-tools-image: ${{ needs.build-build-tools-image.outputs.image }}-bookworm
@@ -201,8 +199,7 @@ jobs:

  build-and-test-locally:
    needs: [ meta, build-build-tools-image ]
-    # We do need to run this in `.*-rc-pr` because of hotfixes.
-    if: ${{ contains(fromJSON('["pr", "push-main", "storage-rc-pr", "proxy-rc-pr", "compute-rc-pr"]'), needs.meta.outputs.run-kind) }}
+    if: ${{ contains(fromJSON('["pr", "push-main"]'), needs.meta.outputs.run-kind) }}
    strategy:
      fail-fast: false
      matrix:
@@ -980,7 +977,7 @@ jobs:
          TEST_EXTENSIONS_TAG: >-
            ${{
              contains(fromJSON('["storage-rc-pr", "proxy-rc-pr"]'), needs.meta.outputs.run-kind)
-              && needs.meta.outputs.previous-compute-release
+              && 'latest'
              || needs.meta.outputs.build-tag
            }}
          TEST_VERSION_ONLY: ${{ matrix.pg_version }}
@@ -1568,10 +1565,10 @@ jobs:
        if: |
          contains(needs.*.result, 'failure')
          || contains(needs.*.result, 'cancelled')
-          || (needs.check-dependencies-rust.result == 'skipped' && needs.files-changed.outputs.check-rust-dependencies == 'true' && contains(fromJSON('["pr", "storage-rc-pr", "proxy-rc-pr", "compute-rc-pr"]'), needs.meta.outputs.run-kind))
-          || (needs.build-and-test-locally.result == 'skipped' && contains(fromJSON('["pr", "push-main", "storage-rc-pr", "proxy-rc-pr", "compute-rc-pr"]'), needs.meta.outputs.run-kind))
-          || (needs.check-codestyle-python.result == 'skipped' && contains(fromJSON('["pr", "storage-rc-pr", "proxy-rc-pr", "compute-rc-pr"]'), needs.meta.outputs.run-kind))
-          || (needs.check-codestyle-rust.result == 'skipped' && contains(fromJSON('["pr", "storage-rc-pr", "proxy-rc-pr", "compute-rc-pr"]'), needs.meta.outputs.run-kind))
+          || (needs.check-dependencies-rust.result == 'skipped' && needs.files-changed.outputs.check-rust-dependencies == 'true' && needs.meta.outputs.run-kind == 'pr')
+          || (needs.build-and-test-locally.result == 'skipped' && needs.meta.outputs.run-kind == 'pr')
+          || (needs.check-codestyle-python.result == 'skipped' && needs.meta.outputs.run-kind == 'pr')
+          || (needs.check-codestyle-rust.result == 'skipped' && needs.meta.outputs.run-kind == 'pr')
          || needs.files-changed.result == 'skipped'
          || (needs.push-compute-image-dev.result == 'skipped' && contains(fromJSON('["push-main", "pr", "compute-release", "compute-rc-pr"]'), needs.meta.outputs.run-kind))
          || (needs.push-neon-image-dev.result == 'skipped' && contains(fromJSON('["push-main", "pr", "storage-release", "storage-rc-pr", "proxy-release", "proxy-rc-pr"]'), needs.meta.outputs.run-kind))
--- a/.gitignore
+++ b/.gitignore
@@ -1,4 +1,3 @@
-/artifact_cache
 /pg_install
 /target
 /tmp_check
--- a/Cargo.lock
+++ b/Cargo.lock
@@ -2248,6 +2248,17 @@ version = "0.4.2"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "0ce7134b9999ecaf8bcd65542e436736ef32ddca1b3e06094cb6ec5755203b80"

+[[package]]
+name = "flag-bearer"
+version = "0.1.0-rc.4"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "e8feaa1b7a5ad6e6dd7791d42d36c1a25004f1c25eae9ab7b904c864109d8260"
+dependencies = [
+ "parking_lot 0.12.1",
+ "pin-list",
+ "pin-project-lite",
+]
+
 [[package]]
 name = "flagset"
 version = "0.4.6"
@@ -3991,33 +4002,6 @@ dependencies = [
 "memchr",
 ]

-[[package]]
-name = "object_storage"
-version = "0.0.1"
-dependencies = [
- "anyhow",
- "axum",
- "axum-extra",
- "camino",
- "camino-tempfile",
- "futures",
- "http-body-util",
- "itertools 0.10.5",
- "jsonwebtoken",
- "prometheus",
- "rand 0.8.5",
- "remote_storage",
- "serde",
- "serde_json",
- "test-log",
- "tokio",
- "tokio-util",
- "tower 0.5.2",
- "tracing",
- "utils",
- "workspace_hack",
-]
-
 [[package]]
 name = "once_cell"
 version = "1.20.2"
@@ -4356,7 +4340,6 @@ dependencies = [
 "strum",
 "strum_macros",
 "thiserror 1.0.69",
- "tracing-utils",
 "utils",
 ]

@@ -4582,6 +4565,16 @@ dependencies = [
 "siphasher",
 ]

+[[package]]
+name = "pin-list"
+version = "0.1.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "5a3c0987a7464afc0d593f13429732ef87e9a6c8e7909a1a22faeff7e1d2159d"
+dependencies = [
+ "pin-project-lite",
+ "pinned-aliasable",
+]
+
 [[package]]
 name = "pin-project"
 version = "1.1.9"
@@ -4614,6 +4607,12 @@ version = "0.1.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "8b870d8c151b6f2fb93e84a13146138f05d02ed11c7e7c54f8826aaaf7c9f184"

+[[package]]
+name = "pinned-aliasable"
+version = "0.1.3"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "5d0f9ae89bf0ed03b69ac1f3f7ea2e6e09b4fa5448011df2e67d581c2b850b7b"
+
 [[package]]
 name = "pkcs1"
 version = "0.7.5"
@@ -4720,7 +4719,7 @@ dependencies = [
 [[package]]
 name = "postgres-protocol"
 version = "0.6.6"
-source = "git+https://github.com/neondatabase/rust-postgres.git?branch=neon#f3cf448febde5fd298071d54d568a9c875a7a62b"
+source = "git+https://github.com/neondatabase/rust-postgres.git?branch=neon#1f21e7959a96a34dcfbfce1b14b73286cdadffe9"
 dependencies = [
 "base64 0.22.1",
 "byteorder",
@@ -4754,7 +4753,7 @@ dependencies = [
 [[package]]
 name = "postgres-types"
 version = "0.2.6"
-source = "git+https://github.com/neondatabase/rust-postgres.git?branch=neon#f3cf448febde5fd298071d54d568a9c875a7a62b"
+source = "git+https://github.com/neondatabase/rust-postgres.git?branch=neon#1f21e7959a96a34dcfbfce1b14b73286cdadffe9"
 dependencies = [
 "bytes",
 "chrono",
@@ -5107,6 +5106,7 @@ dependencies = [
 "ed25519-dalek",
 "env_logger",
 "fallible-iterator",
+ "flag-bearer",
 "flate2",
 "framed-websockets",
 "futures",
@@ -6952,28 +6952,6 @@ dependencies = [
 "syn 2.0.100",
 ]

-[[package]]
-name = "test-log"
-version = "0.2.17"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "e7f46083d221181166e5b6f6b1e5f1d499f3a76888826e6cb1d057554157cd0f"
-dependencies = [
- "env_logger",
- "test-log-macros",
- "tracing-subscriber",
-]
-
-[[package]]
-name = "test-log-macros"
-version = "0.2.17"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "888d0c3c6db53c0fdab160d2ed5e12ba745383d3e85813f2ea0f2b1475ab553f"
-dependencies = [
- "proc-macro2",
- "quote",
- "syn 2.0.100",
-]
-
 [[package]]
 name = "thiserror"
 version = "1.0.69"
@@ -7165,9 +7143,9 @@ dependencies = [

 [[package]]
 name = "tokio"
-version = "1.43.1"
+version = "1.43.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "492a604e2fd7f814268a378409e6c92b5525d747d10db9a229723f55a417958c"
+checksum = "3d61fa4ffa3de412bfea335c6ecff681de2b609ba3c77ef3e00e521813a9ed9e"
 dependencies = [
 "backtrace",
 "bytes",
@@ -7221,7 +7199,7 @@ dependencies = [
 [[package]]
 name = "tokio-postgres"
 version = "0.7.10"
-source = "git+https://github.com/neondatabase/rust-postgres.git?branch=neon#f3cf448febde5fd298071d54d568a9c875a7a62b"
+source = "git+https://github.com/neondatabase/rust-postgres.git?branch=neon#1f21e7959a96a34dcfbfce1b14b73286cdadffe9"
 dependencies = [
 "async-trait",
 "byteorder",
@@ -7653,7 +7631,6 @@ dependencies = [
 "opentelemetry-otlp",
 "opentelemetry-semantic-conventions",
 "opentelemetry_sdk",
- "pin-project-lite",
 "tokio",
 "tracing",
 "tracing-opentelemetry",
--- a/Cargo.toml
+++ b/Cargo.toml
@@ -40,7 +40,6 @@ members = [
    "libs/proxy/postgres-protocol2",
    "libs/proxy/postgres-types2",
    "libs/proxy/tokio-postgres2",
-    "object_storage",
 ]

 [workspace.package]
@@ -184,7 +183,7 @@ test-context = "0.3"
 thiserror = "1.0"
 tikv-jemallocator = { version = "0.6", features = ["profiling", "stats", "unprefixed_malloc_on_supported_platforms"] }
 tikv-jemalloc-ctl = { version = "0.6", features = ["stats"] }
-tokio = { version = "1.43.1", features = ["macros"] }
+tokio = { version = "1.41", features = ["macros"] }
 tokio-epoll-uring = { git = "https://github.com/neondatabase/tokio-epoll-uring.git" , branch = "main" }
 tokio-io-timeout = "1.2.0"
 tokio-postgres-rustls = "0.12.0"
@@ -209,7 +208,6 @@ tracing-opentelemetry = "0.28"
 tracing-serde = "0.2.0"
 tracing-subscriber = { version = "0.3", default-features = false, features = ["smallvec", "fmt", "tracing-log", "std", "env-filter", "json"] }
 try-lock = "0.2.5"
-test-log = { version = "0.2.17", default-features = false, features = ["log"] }
 twox-hash = { version = "1.6.3", default-features = false }
 typed-json = "0.1"
 url = "2.2"
--- a/2
+++ b/2
@@ -89,7 +89,6 @@ RUN set -e \
      --bin storage_broker  \
      --bin storage_controller  \
      --bin proxy  \
-      --bin object_storage \
      --bin neon_local \
      --bin storage_scrubber \
      --locked --release
@@ -122,7 +121,6 @@ COPY --from=build --chown=neon:neon /home/nonroot/target/release/safekeeper
 COPY --from=build --chown=neon:neon /home/nonroot/target/release/storage_broker      /usr/local/bin
 COPY --from=build --chown=neon:neon /home/nonroot/target/release/storage_controller  /usr/local/bin
 COPY --from=build --chown=neon:neon /home/nonroot/target/release/proxy               /usr/local/bin
-COPY --from=build --chown=neon:neon /home/nonroot/target/release/object_storage      /usr/local/bin
 COPY --from=build --chown=neon:neon /home/nonroot/target/release/neon_local          /usr/local/bin
 COPY --from=build --chown=neon:neon /home/nonroot/target/release/storage_scrubber    /usr/local/bin

--- a/build-tools.Dockerfile
+++ b/build-tools.Dockerfile
@@ -292,7 +292,7 @@ WORKDIR /home/nonroot

 # Rust
 # Please keep the version of llvm (installed above) in sync with rust llvm (`rustc --version --verbose | grep LLVM`)
-ENV RUSTC_VERSION=1.86.0
+ENV RUSTC_VERSION=1.85.0
 ENV RUSTUP_HOME="/home/nonroot/.rustup"
 ENV PATH="/home/nonroot/.cargo/bin:${PATH}"
 ARG RUSTFILT_VERSION=0.2.1
--- a/compute/compute-node.Dockerfile
+++ b/compute/compute-node.Dockerfile
@@ -369,7 +369,7 @@ FROM build-deps AS plv8-src
 ARG PG_VERSION
 WORKDIR /ext-src

-COPY compute/patches/plv8* .
+COPY compute/patches/plv8-3.1.10.patch .

 # plv8 3.2.3 supports v17
 # last release v3.2.3 - Sep 7, 2024
@@ -393,7 +393,7 @@ RUN case "${PG_VERSION:?}" in \
    git clone --recurse-submodules --depth 1 --branch ${PLV8_TAG} https://github.com/plv8/plv8.git plv8-src && \
    tar -czf plv8.tar.gz --exclude .git plv8-src && \
    cd plv8-src && \
-    if [[ "${PG_VERSION:?}" < "v17" ]]; then patch -p1 < /ext-src/plv8_v3.1.10.patch; else patch -p1 < /ext-src/plv8_v3.2.3.patch; fi
+    if [[ "${PG_VERSION:?}" < "v17" ]]; then patch -p1 < /ext-src/plv8-3.1.10.patch; fi

 # Step 1: Build the vendored V8 engine. It doesn't depend on PostgreSQL, so use
 # 'build-deps' as the base. This enables caching and avoids unnecessary rebuilds.
@@ -1022,6 +1022,67 @@ RUN make -j $(getconf _NPROCESSORS_ONLN) && \
    make -j $(getconf _NPROCESSORS_ONLN) install && \
    echo 'trusted = true' >> /usr/local/pgsql/share/extension/semver.control

+#########################################################################################
+#
+# Layer "pg_embedding-build"
+# compile pg_embedding extension
+#
+#########################################################################################
+FROM build-deps AS pg_embedding-src
+ARG PG_VERSION
+
+# This is our extension, support stopped in favor of pgvector
+# TODO: deprecate it
+WORKDIR /ext-src
+RUN case "${PG_VERSION:?}" in \
+      "v14" | "v15") \
+        export PG_EMBEDDING_VERSION=0.3.5 \
+        export PG_EMBEDDING_CHECKSUM=0e95b27b8b6196e2cf0a0c9ec143fe2219b82e54c5bb4ee064e76398cbe69ae9 \
+        ;; \
+      *) \
+        echo "pg_embedding not supported on this PostgreSQL version. Use pgvector instead." && exit 0;; \
+    esac && \
+    wget https://github.com/neondatabase/pg_embedding/archive/refs/tags/${PG_EMBEDDING_VERSION}.tar.gz -O pg_embedding.tar.gz && \
+    echo "${PG_EMBEDDING_CHECKSUM} pg_embedding.tar.gz" | sha256sum --check && \
+    mkdir pg_embedding-src && cd pg_embedding-src && tar xzf ../pg_embedding.tar.gz --strip-components=1 -C .
+
+FROM pg-build AS pg_embedding-build
+COPY --from=pg_embedding-src /ext-src/ /ext-src/
+WORKDIR /ext-src/
+RUN  if [ -d pg_embedding-src ]; then \
+        cd pg_embedding-src && \
+        make -j $(getconf _NPROCESSORS_ONLN) && \
+        make -j $(getconf _NPROCESSORS_ONLN) install; \
+    fi
+
+#########################################################################################
+#
+# Layer "pg_anon-build"
+# compile anon extension
+#
+#########################################################################################
+FROM build-deps AS pg_anon-src
+ARG PG_VERSION
+
+# This is an experimental extension, never got to real production.
+# !Do not remove! It can be present in shared_preload_libraries and compute will fail to start if library is not found.
+WORKDIR /ext-src
+RUN case "${PG_VERSION:?}" in "v17") \
+    echo "postgresql_anonymizer does not yet support PG17" && exit 0;; \
+    esac && \
+    wget  https://github.com/neondatabase/postgresql_anonymizer/archive/refs/tags/neon_1.1.1.tar.gz -O pg_anon.tar.gz && \
+    echo "321ea8d5c1648880aafde850a2c576e4a9e7b9933a34ce272efc839328999fa9  pg_anon.tar.gz" | sha256sum --check && \
+    mkdir pg_anon-src && cd pg_anon-src && tar xzf ../pg_anon.tar.gz --strip-components=1 -C .
+
+FROM pg-build AS pg_anon-build
+COPY --from=pg_anon-src /ext-src/ /ext-src/
+WORKDIR /ext-src
+RUN if [ -d pg_anon-src ]; then \
+        cd pg_anon-src && \
+        make -j $(getconf _NPROCESSORS_ONLN) install && \
+        echo 'trusted = true' >> /usr/local/pgsql/share/extension/anon.control; \
+    fi
+
 #########################################################################################
 #
 # Layer "pg build with nonroot user and cargo installed"
@@ -1305,8 +1366,8 @@ ARG PG_VERSION
 # Do not update without approve from proxy team
 # Make sure the version is reflected in proxy/src/serverless/local_conn_pool.rs
 WORKDIR /ext-src
-RUN wget https://github.com/neondatabase/pg_session_jwt/archive/refs/tags/v0.3.0.tar.gz -O pg_session_jwt.tar.gz && \
-    echo "19be2dc0b3834d643706ed430af998bb4c2cdf24b3c45e7b102bb3a550e8660c pg_session_jwt.tar.gz" | sha256sum --check && \
+RUN wget https://github.com/neondatabase/pg_session_jwt/archive/refs/tags/v0.2.0.tar.gz -O pg_session_jwt.tar.gz && \
+    echo "5ace028e591f2e000ca10afa5b1ca62203ebff014c2907c0ec3b29c36f28a1bb pg_session_jwt.tar.gz" | sha256sum --check && \
    mkdir pg_session_jwt-src && cd pg_session_jwt-src && tar xzf ../pg_session_jwt.tar.gz --strip-components=1 -C . && \
    sed -i 's/pgrx = "0.12.6"/pgrx = { version = "0.12.9", features = [ "unsafe-postgres" ] }/g' Cargo.toml && \
    sed -i 's/version = "0.12.6"/version = "0.12.9"/g' pgrx-tests/Cargo.toml && \
@@ -1614,7 +1675,9 @@ COPY --from=rdkit-build /usr/local/pgsql/ /usr/local/pgsql/
 COPY --from=pg_uuidv7-build /usr/local/pgsql/ /usr/local/pgsql/
 COPY --from=pg_roaringbitmap-build /usr/local/pgsql/ /usr/local/pgsql/
 COPY --from=pg_semver-build /usr/local/pgsql/ /usr/local/pgsql/
+COPY --from=pg_embedding-build /usr/local/pgsql/ /usr/local/pgsql/
 COPY --from=wal2json-build /usr/local/pgsql /usr/local/pgsql
+COPY --from=pg_anon-build /usr/local/pgsql/ /usr/local/pgsql/
 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/
@@ -1790,6 +1853,7 @@ COPY --from=pg_cron-src /ext-src/ /ext-src/
 COPY --from=pg_uuidv7-src /ext-src/ /ext-src/
 COPY --from=pg_roaringbitmap-src /ext-src/ /ext-src/
 COPY --from=pg_semver-src /ext-src/ /ext-src/
+#COPY --from=pg_embedding-src /ext-src/ /ext-src/
 #COPY --from=wal2json-src /ext-src/ /ext-src/
 COPY --from=pg_ivm-src /ext-src/ /ext-src/
 COPY --from=pg_partman-src /ext-src/ /ext-src/
--- a/compute/etc/neon_collector.jsonnet
+++ b/compute/etc/neon_collector.jsonnet
@@ -33,7 +33,6 @@
    import 'sql_exporter/lfc_hits.libsonnet',
    import 'sql_exporter/lfc_misses.libsonnet',
    import 'sql_exporter/lfc_used.libsonnet',
-    import 'sql_exporter/lfc_used_pages.libsonnet',
    import 'sql_exporter/lfc_writes.libsonnet',
    import 'sql_exporter/logical_slot_restart_lsn.libsonnet',
    import 'sql_exporter/max_cluster_size.libsonnet',
--- a/compute/etc/sql_exporter/lfc_used_pages.libsonnet
+++ b/compute/etc/sql_exporter/lfc_used_pages.libsonnet
@@ -1,10 +0,0 @@
-{
-  metric_name: 'lfc_used_pages',
-  type: 'gauge',
-  help: 'LFC pages used',
-  key_labels: null,
-  values: [
-    'lfc_used_pages',
-  ],
-  query: importstr 'sql_exporter/lfc_used_pages.sql',
-}
--- a/compute/etc/sql_exporter/lfc_used_pages.sql
+++ b/compute/etc/sql_exporter/lfc_used_pages.sql
@@ -1 +0,0 @@
-SELECT lfc_value AS lfc_used_pages FROM neon.neon_lfc_stats WHERE lfc_key = 'file_cache_used_pages';
--- a/compute/patches/cloud_regress_pg16.patch
+++ b/compute/patches/cloud_regress_pg16.patch
@@ -202,10 +202,10 @@ index cf0b80d616..e8e2a14a4a 100644
 COMMENT ON CONSTRAINT the_constraint ON constraint_comments_tbl IS 'no, the comment';
 ERROR:  must be owner of relation constraint_comments_tbl
 diff --git a/src/test/regress/expected/conversion.out b/src/test/regress/expected/conversion.out
-index d785f92561..16377e5ac9 100644
+index 442e7aff2b..525f732b03 100644
 --- a/src/test/regress/expected/conversion.out
 +++ b/src/test/regress/expected/conversion.out
-@@ -15,7 +15,7 @@ SELECT FROM test_enc_setup();
+@@ -8,7 +8,7 @@
 CREATE FUNCTION test_enc_conversion(bytea, name, name, bool, validlen OUT int, result OUT bytea)
     AS :'regresslib', 'test_enc_conversion'
     LANGUAGE C STRICT;
@@ -587,15 +587,16 @@ index f551624afb..57f1e432d4 100644
 SELECT *
    INTO TABLE ramp
 diff --git a/src/test/regress/expected/database.out b/src/test/regress/expected/database.out
-index 4cbdbdf84d..573362850e 100644
+index 454db91ec0..01378d7081 100644
 --- a/src/test/regress/expected/database.out
 +++ b/src/test/regress/expected/database.out
-@@ -1,8 +1,6 @@
+@@ -1,8 +1,7 @@
 CREATE DATABASE regression_tbd
 	ENCODING utf8 LC_COLLATE "C" LC_CTYPE "C" TEMPLATE template0;
 ALTER DATABASE regression_tbd RENAME TO regression_utf8;
 -ALTER DATABASE regression_utf8 SET TABLESPACE regress_tblspace;
 -ALTER DATABASE regression_utf8 RESET TABLESPACE;
+WARNING:  you need to manually restart any running background workers after this command
 ALTER DATABASE regression_utf8 CONNECTION_LIMIT 123;
 -- Test PgDatabaseToastTable.  Doing this with GRANT would be slow.
 BEGIN;
@@ -699,7 +700,7 @@ index 6ed50fdcfa..caa00a345d 100644
 COMMENT ON FOREIGN DATA WRAPPER dummy IS 'useless';
 CREATE FOREIGN DATA WRAPPER postgresql VALIDATOR postgresql_fdw_validator;
 diff --git a/src/test/regress/expected/foreign_key.out b/src/test/regress/expected/foreign_key.out
-index 84745b9f60..4883c12351 100644
+index 6b8c2f2414..8e13b7fa46 100644
 --- a/src/test/regress/expected/foreign_key.out
 +++ b/src/test/regress/expected/foreign_key.out
@@ -1985,7 +1985,7 @@ ALTER TABLE fk_partitioned_fk_6 ATTACH PARTITION fk_partitioned_pk_6 FOR VALUES
@@ -1111,7 +1112,7 @@ index 8475231735..0653946337 100644
 DROP ROLE regress_passwd_sha_len1;
 DROP ROLE regress_passwd_sha_len2;
 diff --git a/src/test/regress/expected/privileges.out b/src/test/regress/expected/privileges.out
-index 620fbe8c52..0570102357 100644
+index 5b9dba7b32..cc408dad42 100644
 --- a/src/test/regress/expected/privileges.out
 +++ b/src/test/regress/expected/privileges.out
@@ -20,19 +20,19 @@ SELECT lo_unlink(oid) FROM pg_largeobject_metadata WHERE oid >= 1000 AND oid < 3
@@ -1173,8 +1174,8 @@ index 620fbe8c52..0570102357 100644
 +CREATE GROUP regress_priv_group2 WITH ADMIN regress_priv_user1 PASSWORD NEON_PASSWORD_PLACEHOLDER USER regress_priv_user2;
 ALTER GROUP regress_priv_group1 ADD USER regress_priv_user4;
 GRANT regress_priv_group2 TO regress_priv_user2 GRANTED BY regress_priv_user1;
- SET SESSION AUTHORIZATION regress_priv_user3;
-@@ -246,12 +246,16 @@ GRANT regress_priv_role TO regress_priv_user1 WITH ADMIN OPTION GRANTED BY regre
+ SET SESSION AUTHORIZATION regress_priv_user1;
+@@ -239,12 +239,16 @@ GRANT regress_priv_role TO regress_priv_user1 WITH ADMIN OPTION GRANTED BY regre
 ERROR:  permission denied to grant privileges as role "regress_priv_role"
 DETAIL:  The grantor must have the ADMIN option on role "regress_priv_role".
 GRANT regress_priv_role TO regress_priv_user1 WITH ADMIN OPTION GRANTED BY CURRENT_ROLE;
@@ -1191,7 +1192,7 @@ index 620fbe8c52..0570102357 100644
 DROP ROLE regress_priv_role;
 SET SESSION AUTHORIZATION regress_priv_user1;
 SELECT session_user, current_user;
-@@ -1783,7 +1787,7 @@ SELECT has_table_privilege('regress_priv_user1', 'atest4', 'SELECT WITH GRANT OP
+@@ -1776,7 +1780,7 @@ SELECT has_table_privilege('regress_priv_user1', 'atest4', 'SELECT WITH GRANT OP
 
 -- security-restricted operations
 \c -
@@ -1200,7 +1201,7 @@ index 620fbe8c52..0570102357 100644
 -- Check that index expressions and predicates are run as the table's owner
 -- A dummy index function checking current_user
 CREATE FUNCTION sro_ifun(int) RETURNS int AS $$
-@@ -2675,8 +2679,8 @@ drop cascades to function testns.priv_testagg(integer)
+@@ -2668,8 +2672,8 @@ drop cascades to function testns.priv_testagg(integer)
 drop cascades to function testns.priv_testproc(integer)
 -- Change owner of the schema & and rename of new schema owner
 \c -
@@ -1211,7 +1212,7 @@ index 620fbe8c52..0570102357 100644
 SET SESSION ROLE regress_schemauser1;
 CREATE SCHEMA testns;
 SELECT nspname, rolname FROM pg_namespace, pg_roles WHERE pg_namespace.nspname = 'testns' AND pg_namespace.nspowner = pg_roles.oid;
-@@ -2799,7 +2803,7 @@ DROP USER regress_priv_user7;
+@@ -2792,7 +2796,7 @@ DROP USER regress_priv_user7;
 DROP USER regress_priv_user8; -- does not exist
 ERROR:  role "regress_priv_user8" does not exist
 -- permissions with LOCK TABLE
@@ -1220,7 +1221,7 @@ index 620fbe8c52..0570102357 100644
 CREATE TABLE lock_table (a int);
 -- LOCK TABLE and SELECT permission
 GRANT SELECT ON lock_table TO regress_locktable_user;
-@@ -2881,7 +2885,7 @@ DROP USER regress_locktable_user;
+@@ -2874,7 +2878,7 @@ DROP USER regress_locktable_user;
 -- pg_backend_memory_contexts.
 -- switch to superuser
 \c -
@@ -1229,7 +1230,7 @@ index 620fbe8c52..0570102357 100644
 SELECT has_table_privilege('regress_readallstats','pg_backend_memory_contexts','SELECT'); -- no
  has_table_privilege 
 ---------------------
-@@ -2925,10 +2929,10 @@ RESET ROLE;
+@@ -2918,10 +2922,10 @@ RESET ROLE;
 -- clean up
 DROP ROLE regress_readallstats;
 -- test role grantor machinery
@@ -1244,7 +1245,7 @@ index 620fbe8c52..0570102357 100644
 GRANT regress_group TO regress_group_direct_manager WITH INHERIT FALSE, ADMIN TRUE;
 GRANT regress_group_direct_manager TO regress_group_indirect_manager;
 SET SESSION AUTHORIZATION regress_group_direct_manager;
-@@ -2957,9 +2961,9 @@ DROP ROLE regress_group_direct_manager;
+@@ -2950,9 +2954,9 @@ DROP ROLE regress_group_direct_manager;
 DROP ROLE regress_group_indirect_manager;
 DROP ROLE regress_group_member;
 -- test SET and INHERIT options with object ownership changes
@@ -1840,7 +1841,7 @@ index 09a255649b..15895f0c53 100644
 CREATE TABLE ruletest_t2 (x int);
 CREATE VIEW ruletest_v1 WITH (security_invoker=true) AS
 diff --git a/src/test/regress/expected/security_label.out b/src/test/regress/expected/security_label.out
-index a8e01a6220..83543b250a 100644
+index a8e01a6220..5a9cef4ede 100644
 --- a/src/test/regress/expected/security_label.out
 +++ b/src/test/regress/expected/security_label.out
@@ -6,8 +6,8 @@ SET client_min_messages TO 'warning';
@@ -1854,6 +1855,34 @@ index a8e01a6220..83543b250a 100644
 CREATE TABLE seclabel_tbl1 (a int, b text);
 CREATE TABLE seclabel_tbl2 (x int, y text);
 CREATE VIEW seclabel_view1 AS SELECT * FROM seclabel_tbl2;
+@@ -19,21 +19,21 @@ ALTER TABLE seclabel_tbl2 OWNER TO regress_seclabel_user2;
+ -- Test of SECURITY LABEL statement without a plugin
+ --
+ SECURITY LABEL ON TABLE seclabel_tbl1 IS 'classified';			-- fail
+-ERROR:  no security label providers have been loaded
+ERROR:  must specify provider when multiple security label providers have been loaded
+ SECURITY LABEL FOR 'dummy' ON TABLE seclabel_tbl1 IS 'classified';		-- fail
+ ERROR:  security label provider "dummy" is not loaded
+ SECURITY LABEL ON TABLE seclabel_tbl1 IS '...invalid label...';		-- fail
+-ERROR:  no security label providers have been loaded
+ERROR:  must specify provider when multiple security label providers have been loaded
+ SECURITY LABEL ON TABLE seclabel_tbl3 IS 'unclassified';			-- fail
+-ERROR:  no security label providers have been loaded
+ERROR:  must specify provider when multiple security label providers have been loaded
+ SECURITY LABEL ON ROLE regress_seclabel_user1 IS 'classified';			-- fail
+-ERROR:  no security label providers have been loaded
+ERROR:  must specify provider when multiple security label providers have been loaded
+ SECURITY LABEL FOR 'dummy' ON ROLE regress_seclabel_user1 IS 'classified';		-- fail
+ ERROR:  security label provider "dummy" is not loaded
+ SECURITY LABEL ON ROLE regress_seclabel_user1 IS '...invalid label...';		-- fail
+-ERROR:  no security label providers have been loaded
+ERROR:  must specify provider when multiple security label providers have been loaded
+ SECURITY LABEL ON ROLE regress_seclabel_user3 IS 'unclassified';			-- fail
+-ERROR:  no security label providers have been loaded
+ERROR:  must specify provider when multiple security label providers have been loaded
+ -- clean up objects
+ DROP FUNCTION seclabel_four();
+ DROP DOMAIN seclabel_domain;
 diff --git a/src/test/regress/expected/select_into.out b/src/test/regress/expected/select_into.out
 index b79fe9a1c0..e29fab88ab 100644
 --- a/src/test/regress/expected/select_into.out
@@ -2384,10 +2413,10 @@ index e3e3bea709..fa86ddc326 100644
 COMMENT ON CONSTRAINT the_constraint ON constraint_comments_tbl IS 'no, the comment';
 COMMENT ON CONSTRAINT the_constraint ON DOMAIN constraint_comments_dom IS 'no, another comment';
 diff --git a/src/test/regress/sql/conversion.sql b/src/test/regress/sql/conversion.sql
-index b567a1a572..4d1ac2e631 100644
+index 9a65fca91f..58431a3056 100644
 --- a/src/test/regress/sql/conversion.sql
 +++ b/src/test/regress/sql/conversion.sql
-@@ -17,7 +17,7 @@ CREATE FUNCTION test_enc_conversion(bytea, name, name, bool, validlen OUT int, r
+@@ -12,7 +12,7 @@ CREATE FUNCTION test_enc_conversion(bytea, name, name, bool, validlen OUT int, r
     AS :'regresslib', 'test_enc_conversion'
     LANGUAGE C STRICT;
 
@@ -2751,7 +2780,7 @@ index ae6841308b..47bc792e30 100644
 
 SELECT *
 diff --git a/src/test/regress/sql/database.sql b/src/test/regress/sql/database.sql
-index 46ad263478..eb05584ed5 100644
+index 0367c0e37a..a23b98c4bd 100644
 --- a/src/test/regress/sql/database.sql
 +++ b/src/test/regress/sql/database.sql
@@ -1,8 +1,6 @@
@@ -2864,7 +2893,7 @@ index aa147b14a9..370e0dd570 100644
 CREATE FOREIGN DATA WRAPPER dummy;
 COMMENT ON FOREIGN DATA WRAPPER dummy IS 'useless';
 diff --git a/src/test/regress/sql/foreign_key.sql b/src/test/regress/sql/foreign_key.sql
-index 9f4210b26e..620d3fc87e 100644
+index 45c7a534cb..32dd26b8cd 100644
 --- a/src/test/regress/sql/foreign_key.sql
 +++ b/src/test/regress/sql/foreign_key.sql
@@ -1435,7 +1435,7 @@ ALTER TABLE fk_partitioned_fk_6 ATTACH PARTITION fk_partitioned_pk_6 FOR VALUES
@@ -3217,7 +3246,7 @@ index 53e86b0b6c..0303fdfe96 100644
 -- Check that the invalid secrets were re-hashed. A re-hashed secret
 -- should not contain the original salt.
 diff --git a/src/test/regress/sql/privileges.sql b/src/test/regress/sql/privileges.sql
-index 259f1aedd1..6e1a3d17b7 100644
+index 249df17a58..b258e7f26a 100644
 --- a/src/test/regress/sql/privileges.sql
 +++ b/src/test/regress/sql/privileges.sql
@@ -24,18 +24,18 @@ RESET client_min_messages;
@@ -3279,7 +3308,7 @@ index 259f1aedd1..6e1a3d17b7 100644
 
 ALTER GROUP regress_priv_group1 ADD USER regress_priv_user4;
 
-@@ -1160,7 +1160,7 @@ SELECT has_table_privilege('regress_priv_user1', 'atest4', 'SELECT WITH GRANT OP
+@@ -1157,7 +1157,7 @@ SELECT has_table_privilege('regress_priv_user1', 'atest4', 'SELECT WITH GRANT OP
 
 -- security-restricted operations
 \c -
@@ -3288,7 +3317,7 @@ index 259f1aedd1..6e1a3d17b7 100644
 
 -- Check that index expressions and predicates are run as the table's owner
 
-@@ -1656,8 +1656,8 @@ DROP SCHEMA testns CASCADE;
+@@ -1653,8 +1653,8 @@ DROP SCHEMA testns CASCADE;
 -- Change owner of the schema & and rename of new schema owner
 \c -
 
@@ -3299,7 +3328,7 @@ index 259f1aedd1..6e1a3d17b7 100644
 
 SET SESSION ROLE regress_schemauser1;
 CREATE SCHEMA testns;
-@@ -1751,7 +1751,7 @@ DROP USER regress_priv_user8; -- does not exist
+@@ -1748,7 +1748,7 @@ DROP USER regress_priv_user8; -- does not exist
 
 
 -- permissions with LOCK TABLE
@@ -3308,7 +3337,7 @@ index 259f1aedd1..6e1a3d17b7 100644
 CREATE TABLE lock_table (a int);
 
 -- LOCK TABLE and SELECT permission
-@@ -1839,7 +1839,7 @@ DROP USER regress_locktable_user;
+@@ -1836,7 +1836,7 @@ DROP USER regress_locktable_user;
 -- switch to superuser
 \c -
 
@@ -3317,7 +3346,7 @@ index 259f1aedd1..6e1a3d17b7 100644
 
 SELECT has_table_privilege('regress_readallstats','pg_backend_memory_contexts','SELECT'); -- no
 SELECT has_table_privilege('regress_readallstats','pg_shmem_allocations','SELECT'); -- no
-@@ -1859,10 +1859,10 @@ RESET ROLE;
+@@ -1856,10 +1856,10 @@ RESET ROLE;
 DROP ROLE regress_readallstats;
 
 -- test role grantor machinery
@@ -3332,7 +3361,7 @@ index 259f1aedd1..6e1a3d17b7 100644
 
 GRANT regress_group TO regress_group_direct_manager WITH INHERIT FALSE, ADMIN TRUE;
 GRANT regress_group_direct_manager TO regress_group_indirect_manager;
-@@ -1884,9 +1884,9 @@ DROP ROLE regress_group_indirect_manager;
+@@ -1881,9 +1881,9 @@ DROP ROLE regress_group_indirect_manager;
 DROP ROLE regress_group_member;
 
 -- test SET and INHERIT options with object ownership changes
--- a/compute/patches/cloud_regress_pg17.patch
+++ b/compute/patches/cloud_regress_pg17.patch
@@ -202,10 +202,10 @@ index cf0b80d616..e8e2a14a4a 100644
 COMMENT ON CONSTRAINT the_constraint ON constraint_comments_tbl IS 'no, the comment';
 ERROR:  must be owner of relation constraint_comments_tbl
 diff --git a/src/test/regress/expected/conversion.out b/src/test/regress/expected/conversion.out
-index d785f92561..16377e5ac9 100644
+index 442e7aff2b..525f732b03 100644
 --- a/src/test/regress/expected/conversion.out
 +++ b/src/test/regress/expected/conversion.out
-@@ -15,7 +15,7 @@ SELECT FROM test_enc_setup();
+@@ -8,7 +8,7 @@
 CREATE FUNCTION test_enc_conversion(bytea, name, name, bool, validlen OUT int, result OUT bytea)
     AS :'regresslib', 'test_enc_conversion'
     LANGUAGE C STRICT;
@@ -587,15 +587,16 @@ index f551624afb..57f1e432d4 100644
 SELECT *
    INTO TABLE ramp
 diff --git a/src/test/regress/expected/database.out b/src/test/regress/expected/database.out
-index 4cbdbdf84d..573362850e 100644
+index 454db91ec0..01378d7081 100644
 --- a/src/test/regress/expected/database.out
 +++ b/src/test/regress/expected/database.out
-@@ -1,8 +1,6 @@
+@@ -1,8 +1,7 @@
 CREATE DATABASE regression_tbd
 	ENCODING utf8 LC_COLLATE "C" LC_CTYPE "C" TEMPLATE template0;
 ALTER DATABASE regression_tbd RENAME TO regression_utf8;
 -ALTER DATABASE regression_utf8 SET TABLESPACE regress_tblspace;
 -ALTER DATABASE regression_utf8 RESET TABLESPACE;
+WARNING:  you need to manually restart any running background workers after this command
 ALTER DATABASE regression_utf8 CONNECTION_LIMIT 123;
 -- Test PgDatabaseToastTable.  Doing this with GRANT would be slow.
 BEGIN;
@@ -699,7 +700,7 @@ index 6ed50fdcfa..caa00a345d 100644
 COMMENT ON FOREIGN DATA WRAPPER dummy IS 'useless';
 CREATE FOREIGN DATA WRAPPER postgresql VALIDATOR postgresql_fdw_validator;
 diff --git a/src/test/regress/expected/foreign_key.out b/src/test/regress/expected/foreign_key.out
-index fe6a1015f2..614b387b7d 100644
+index 69994c98e3..129abcfbe8 100644
 --- a/src/test/regress/expected/foreign_key.out
 +++ b/src/test/regress/expected/foreign_key.out
@@ -1985,7 +1985,7 @@ ALTER TABLE fk_partitioned_fk_6 ATTACH PARTITION fk_partitioned_pk_6 FOR VALUES
@@ -1146,7 +1147,7 @@ index 924d6e001d..7fdda73439 100644
 DROP ROLE regress_passwd_sha_len1;
 DROP ROLE regress_passwd_sha_len2;
 diff --git a/src/test/regress/expected/privileges.out b/src/test/regress/expected/privileges.out
-index e8c668e0a1..03be5c2120 100644
+index 1296da0d57..f43fffa44c 100644
 --- a/src/test/regress/expected/privileges.out
 +++ b/src/test/regress/expected/privileges.out
@@ -20,19 +20,19 @@ SELECT lo_unlink(oid) FROM pg_largeobject_metadata WHERE oid >= 1000 AND oid < 3
@@ -1208,8 +1209,8 @@ index e8c668e0a1..03be5c2120 100644
 +CREATE GROUP regress_priv_group2 WITH ADMIN regress_priv_user1 PASSWORD NEON_PASSWORD_PLACEHOLDER USER regress_priv_user2;
 ALTER GROUP regress_priv_group1 ADD USER regress_priv_user4;
 GRANT regress_priv_group2 TO regress_priv_user2 GRANTED BY regress_priv_user1;
- SET SESSION AUTHORIZATION regress_priv_user3;
-@@ -246,12 +246,16 @@ GRANT regress_priv_role TO regress_priv_user1 WITH ADMIN OPTION GRANTED BY regre
+ SET SESSION AUTHORIZATION regress_priv_user1;
+@@ -239,12 +239,16 @@ GRANT regress_priv_role TO regress_priv_user1 WITH ADMIN OPTION GRANTED BY regre
 ERROR:  permission denied to grant privileges as role "regress_priv_role"
 DETAIL:  The grantor must have the ADMIN option on role "regress_priv_role".
 GRANT regress_priv_role TO regress_priv_user1 WITH ADMIN OPTION GRANTED BY CURRENT_ROLE;
@@ -1226,7 +1227,7 @@ index e8c668e0a1..03be5c2120 100644
 DROP ROLE regress_priv_role;
 SET SESSION AUTHORIZATION regress_priv_user1;
 SELECT session_user, current_user;
-@@ -1783,7 +1787,7 @@ SELECT has_table_privilege('regress_priv_user1', 'atest4', 'SELECT WITH GRANT OP
+@@ -1776,7 +1780,7 @@ SELECT has_table_privilege('regress_priv_user1', 'atest4', 'SELECT WITH GRANT OP
 
 -- security-restricted operations
 \c -
@@ -1235,7 +1236,7 @@ index e8c668e0a1..03be5c2120 100644
 -- Check that index expressions and predicates are run as the table's owner
 -- A dummy index function checking current_user
 CREATE FUNCTION sro_ifun(int) RETURNS int AS $$
-@@ -2675,8 +2679,8 @@ drop cascades to function testns.priv_testagg(integer)
+@@ -2668,8 +2672,8 @@ drop cascades to function testns.priv_testagg(integer)
 drop cascades to function testns.priv_testproc(integer)
 -- Change owner of the schema & and rename of new schema owner
 \c -
@@ -1246,7 +1247,7 @@ index e8c668e0a1..03be5c2120 100644
 SET SESSION ROLE regress_schemauser1;
 CREATE SCHEMA testns;
 SELECT nspname, rolname FROM pg_namespace, pg_roles WHERE pg_namespace.nspname = 'testns' AND pg_namespace.nspowner = pg_roles.oid;
-@@ -2799,7 +2803,7 @@ DROP USER regress_priv_user7;
+@@ -2792,7 +2796,7 @@ DROP USER regress_priv_user7;
 DROP USER regress_priv_user8; -- does not exist
 ERROR:  role "regress_priv_user8" does not exist
 -- permissions with LOCK TABLE
@@ -1255,7 +1256,7 @@ index e8c668e0a1..03be5c2120 100644
 CREATE TABLE lock_table (a int);
 -- LOCK TABLE and SELECT permission
 GRANT SELECT ON lock_table TO regress_locktable_user;
-@@ -2895,7 +2899,7 @@ DROP USER regress_locktable_user;
+@@ -2888,7 +2892,7 @@ DROP USER regress_locktable_user;
 -- pg_backend_memory_contexts.
 -- switch to superuser
 \c -
@@ -1264,7 +1265,7 @@ index e8c668e0a1..03be5c2120 100644
 SELECT has_table_privilege('regress_readallstats','pg_backend_memory_contexts','SELECT'); -- no
  has_table_privilege 
 ---------------------
-@@ -2939,10 +2943,10 @@ RESET ROLE;
+@@ -2932,10 +2936,10 @@ RESET ROLE;
 -- clean up
 DROP ROLE regress_readallstats;
 -- test role grantor machinery
@@ -1279,7 +1280,7 @@ index e8c668e0a1..03be5c2120 100644
 GRANT regress_group TO regress_group_direct_manager WITH INHERIT FALSE, ADMIN TRUE;
 GRANT regress_group_direct_manager TO regress_group_indirect_manager;
 SET SESSION AUTHORIZATION regress_group_direct_manager;
-@@ -2971,9 +2975,9 @@ DROP ROLE regress_group_direct_manager;
+@@ -2964,9 +2968,9 @@ DROP ROLE regress_group_direct_manager;
 DROP ROLE regress_group_indirect_manager;
 DROP ROLE regress_group_member;
 -- test SET and INHERIT options with object ownership changes
@@ -1292,7 +1293,7 @@ index e8c668e0a1..03be5c2120 100644
 CREATE SCHEMA regress_roleoption;
 GRANT CREATE, USAGE ON SCHEMA regress_roleoption TO PUBLIC;
 GRANT regress_roleoption_donor TO regress_roleoption_protagonist WITH INHERIT TRUE, SET FALSE;
-@@ -3002,9 +3006,9 @@ DROP ROLE regress_roleoption_protagonist;
+@@ -2995,9 +2999,9 @@ DROP ROLE regress_roleoption_protagonist;
 DROP ROLE regress_roleoption_donor;
 DROP ROLE regress_roleoption_recipient;
 -- MAINTAIN
@@ -2432,10 +2433,10 @@ index e3e3bea709..fa86ddc326 100644
 COMMENT ON CONSTRAINT the_constraint ON constraint_comments_tbl IS 'no, the comment';
 COMMENT ON CONSTRAINT the_constraint ON DOMAIN constraint_comments_dom IS 'no, another comment';
 diff --git a/src/test/regress/sql/conversion.sql b/src/test/regress/sql/conversion.sql
-index b567a1a572..4d1ac2e631 100644
+index 9a65fca91f..58431a3056 100644
 --- a/src/test/regress/sql/conversion.sql
 +++ b/src/test/regress/sql/conversion.sql
-@@ -17,7 +17,7 @@ CREATE FUNCTION test_enc_conversion(bytea, name, name, bool, validlen OUT int, r
+@@ -12,7 +12,7 @@ CREATE FUNCTION test_enc_conversion(bytea, name, name, bool, validlen OUT int, r
     AS :'regresslib', 'test_enc_conversion'
     LANGUAGE C STRICT;
 
@@ -2799,7 +2800,7 @@ index ae6841308b..47bc792e30 100644
 
 SELECT *
 diff --git a/src/test/regress/sql/database.sql b/src/test/regress/sql/database.sql
-index 46ad263478..eb05584ed5 100644
+index 0367c0e37a..a23b98c4bd 100644
 --- a/src/test/regress/sql/database.sql
 +++ b/src/test/regress/sql/database.sql
@@ -1,8 +1,6 @@
@@ -2912,7 +2913,7 @@ index aa147b14a9..370e0dd570 100644
 CREATE FOREIGN DATA WRAPPER dummy;
 COMMENT ON FOREIGN DATA WRAPPER dummy IS 'useless';
 diff --git a/src/test/regress/sql/foreign_key.sql b/src/test/regress/sql/foreign_key.sql
-index 8c4e4c7c83..e946cd2119 100644
+index 2e710e419c..89cd481a54 100644
 --- a/src/test/regress/sql/foreign_key.sql
 +++ b/src/test/regress/sql/foreign_key.sql
@@ -1435,7 +1435,7 @@ ALTER TABLE fk_partitioned_fk_6 ATTACH PARTITION fk_partitioned_pk_6 FOR VALUES
@@ -3300,7 +3301,7 @@ index bb82aa4aa2..dd8a05e24d 100644
 -- Check that the invalid secrets were re-hashed. A re-hashed secret
 -- should not contain the original salt.
 diff --git a/src/test/regress/sql/privileges.sql b/src/test/regress/sql/privileges.sql
-index b7e1cb6cdd..6e5a2217f1 100644
+index 5880bc018d..27aa952b18 100644
 --- a/src/test/regress/sql/privileges.sql
 +++ b/src/test/regress/sql/privileges.sql
@@ -24,18 +24,18 @@ RESET client_min_messages;
@@ -3362,7 +3363,7 @@ index b7e1cb6cdd..6e5a2217f1 100644
 
 ALTER GROUP regress_priv_group1 ADD USER regress_priv_user4;
 
-@@ -1160,7 +1160,7 @@ SELECT has_table_privilege('regress_priv_user1', 'atest4', 'SELECT WITH GRANT OP
+@@ -1157,7 +1157,7 @@ SELECT has_table_privilege('regress_priv_user1', 'atest4', 'SELECT WITH GRANT OP
 
 -- security-restricted operations
 \c -
@@ -3371,7 +3372,7 @@ index b7e1cb6cdd..6e5a2217f1 100644
 
 -- Check that index expressions and predicates are run as the table's owner
 
-@@ -1656,8 +1656,8 @@ DROP SCHEMA testns CASCADE;
+@@ -1653,8 +1653,8 @@ DROP SCHEMA testns CASCADE;
 -- Change owner of the schema & and rename of new schema owner
 \c -
 
@@ -3382,7 +3383,7 @@ index b7e1cb6cdd..6e5a2217f1 100644
 
 SET SESSION ROLE regress_schemauser1;
 CREATE SCHEMA testns;
-@@ -1751,7 +1751,7 @@ DROP USER regress_priv_user8; -- does not exist
+@@ -1748,7 +1748,7 @@ DROP USER regress_priv_user8; -- does not exist
 
 
 -- permissions with LOCK TABLE
@@ -3391,7 +3392,7 @@ index b7e1cb6cdd..6e5a2217f1 100644
 CREATE TABLE lock_table (a int);
 
 -- LOCK TABLE and SELECT permission
-@@ -1854,7 +1854,7 @@ DROP USER regress_locktable_user;
+@@ -1851,7 +1851,7 @@ DROP USER regress_locktable_user;
 -- switch to superuser
 \c -
 
@@ -3400,7 +3401,7 @@ index b7e1cb6cdd..6e5a2217f1 100644
 
 SELECT has_table_privilege('regress_readallstats','pg_backend_memory_contexts','SELECT'); -- no
 SELECT has_table_privilege('regress_readallstats','pg_shmem_allocations','SELECT'); -- no
-@@ -1874,10 +1874,10 @@ RESET ROLE;
+@@ -1871,10 +1871,10 @@ RESET ROLE;
 DROP ROLE regress_readallstats;
 
 -- test role grantor machinery
@@ -3415,7 +3416,7 @@ index b7e1cb6cdd..6e5a2217f1 100644
 
 GRANT regress_group TO regress_group_direct_manager WITH INHERIT FALSE, ADMIN TRUE;
 GRANT regress_group_direct_manager TO regress_group_indirect_manager;
-@@ -1899,9 +1899,9 @@ DROP ROLE regress_group_indirect_manager;
+@@ -1896,9 +1896,9 @@ DROP ROLE regress_group_indirect_manager;
 DROP ROLE regress_group_member;
 
 -- test SET and INHERIT options with object ownership changes
@@ -3428,7 +3429,7 @@ index b7e1cb6cdd..6e5a2217f1 100644
 CREATE SCHEMA regress_roleoption;
 GRANT CREATE, USAGE ON SCHEMA regress_roleoption TO PUBLIC;
 GRANT regress_roleoption_donor TO regress_roleoption_protagonist WITH INHERIT TRUE, SET FALSE;
-@@ -1929,9 +1929,9 @@ DROP ROLE regress_roleoption_donor;
+@@ -1926,9 +1926,9 @@ DROP ROLE regress_roleoption_donor;
 DROP ROLE regress_roleoption_recipient;
 
 -- MAINTAIN
--- a/compute/patches/pg_hint_plan_v16.patch
+++ b/compute/patches/pg_hint_plan_v16.patch
@@ -2,6 +2,23 @@ diff --git a/expected/ut-A.out b/expected/ut-A.out
 index da723b8..5328114 100644
 --- a/expected/ut-A.out
 +++ b/expected/ut-A.out
+@@ -9,13 +9,16 @@ SET search_path TO public;
+ ----
+ -- No.A-1-1-3
+ CREATE EXTENSION pg_hint_plan;
+LOG:  Sending request to compute_ctl: http://localhost:3081/extension_server/pg_hint_plan
+ -- No.A-1-2-3
+ DROP EXTENSION pg_hint_plan;
+ -- No.A-1-1-4
+ CREATE SCHEMA other_schema;
+ CREATE EXTENSION pg_hint_plan SCHEMA other_schema;
+LOG:  Sending request to compute_ctl: http://localhost:3081/extension_server/pg_hint_plan
+ ERROR:  extension "pg_hint_plan" must be installed in schema "hint_plan"
+ CREATE EXTENSION pg_hint_plan;
+LOG:  Sending request to compute_ctl: http://localhost:3081/extension_server/pg_hint_plan
+ DROP SCHEMA other_schema;
+ ----
+ ---- No. A-5-1 comment pattern
@@ -3175,6 +3178,7 @@ SELECT s.query, s.calls
   FROM public.pg_stat_statements s
   JOIN pg_catalog.pg_database d
@@ -10,6 +27,18 @@ index da723b8..5328114 100644
  ORDER BY 1;
                 query                 | calls 
 --------------------------------------+-------
+diff --git a/expected/ut-fdw.out b/expected/ut-fdw.out
+index d372459..6282afe 100644
+--- a/expected/ut-fdw.out
+++ b/expected/ut-fdw.out
+@@ -7,6 +7,7 @@ SET pg_hint_plan.debug_print TO on;
+ SET client_min_messages TO LOG;
+ SET pg_hint_plan.enable_hint TO on;
+ CREATE EXTENSION file_fdw;
+LOG:  Sending request to compute_ctl: http://localhost:3081/extension_server/file_fdw
+ CREATE SERVER file_server FOREIGN DATA WRAPPER file_fdw;
+ CREATE USER MAPPING FOR PUBLIC SERVER file_server;
+ CREATE FOREIGN TABLE ft1 (id int, val int) SERVER file_server OPTIONS (format 'csv', filename :'filename');
 diff --git a/sql/ut-A.sql b/sql/ut-A.sql
 index 7c7d58a..4fd1a07 100644
 --- a/sql/ut-A.sql
--- a/compute/patches/pg_hint_plan_v17.patch
+++ b/compute/patches/pg_hint_plan_v17.patch
@@ -1,3 +1,24 @@
+diff --git a/expected/ut-A.out b/expected/ut-A.out
+index e7d68a1..65a056c 100644
+--- a/expected/ut-A.out
+++ b/expected/ut-A.out
+@@ -9,13 +9,16 @@ SET search_path TO public;
+ ----
+ -- No.A-1-1-3
+ CREATE EXTENSION pg_hint_plan;
+LOG:  Sending request to compute_ctl: http://localhost:3081/extension_server/pg_hint_plan
+ -- No.A-1-2-3
+ DROP EXTENSION pg_hint_plan;
+ -- No.A-1-1-4
+ CREATE SCHEMA other_schema;
+ CREATE EXTENSION pg_hint_plan SCHEMA other_schema;
+LOG:  Sending request to compute_ctl: http://localhost:3081/extension_server/pg_hint_plan
+ ERROR:  extension "pg_hint_plan" must be installed in schema "hint_plan"
+ CREATE EXTENSION pg_hint_plan;
+LOG:  Sending request to compute_ctl: http://localhost:3081/extension_server/pg_hint_plan
+ DROP SCHEMA other_schema;
+ ----
+ ---- No. A-5-1 comment pattern
 diff --git a/expected/ut-J.out b/expected/ut-J.out
 index 2fa3c70..314e929 100644
 --- a/expected/ut-J.out
@@ -139,3 +160,15 @@ index a09bd34..0ad227c 100644
 error hint:
 
                     explain_filter                    
+diff --git a/expected/ut-fdw.out b/expected/ut-fdw.out
+index 017fa4b..98d989b 100644
+--- a/expected/ut-fdw.out
+++ b/expected/ut-fdw.out
+@@ -7,6 +7,7 @@ SET pg_hint_plan.debug_print TO on;
+ SET client_min_messages TO LOG;
+ SET pg_hint_plan.enable_hint TO on;
+ CREATE EXTENSION file_fdw;
+LOG:  Sending request to compute_ctl: http://localhost:3081/extension_server/file_fdw
+ CREATE SERVER file_server FOREIGN DATA WRAPPER file_fdw;
+ CREATE USER MAPPING FOR PUBLIC SERVER file_server;
+ CREATE FOREIGN TABLE ft1 (id int, val int) SERVER file_server OPTIONS (format 'csv', filename :'filename');
--- a/compute/patches/plv8_v3.1.10.patch
+++ b/compute/patches/plv8_v3.1.10.patch
@@ -1,6 +1,12 @@
+commit 46b38d3e46f9cd6c70d9b189dd6ff4abaa17cf5e
+Author: Alexander Bayandin <alexander@neon.tech>
+Date:   Sat Nov 30 18:29:32 2024 +0000
+
+    Fix v8 9.7.37 compilation on Debian 12
+
 diff --git a/patches/code/84cf3230a9680aac3b73c410c2b758760b6d3066.patch b/patches/code/84cf3230a9680aac3b73c410c2b758760b6d3066.patch
 new file mode 100644
-index 0000000..fae1cb3
+index 0000000..f0a5dc7
 --- /dev/null
 +++ b/patches/code/84cf3230a9680aac3b73c410c2b758760b6d3066.patch
@@ -0,0 +1,30 @@
@@ -29,21 +35,8 @@ index 0000000..fae1cb3
 +@@ -5,6 +5,7 @@
 + #ifndef V8_HEAP_CPPGC_PREFINALIZER_HANDLER_H_
 + #define V8_HEAP_CPPGC_PREFINALIZER_HANDLER_H_
-+
+ 
 ++#include <utility>
 + #include <vector>
-+
+ 
 + #include "include/cppgc/prefinalizer.h"
-diff --git a/plv8.cc b/plv8.cc
-index c1ce883..6e47e94 100644
--- a/plv8.cc
-+++ b/plv8.cc
-@@ -379,7 +379,7 @@ _PG_init(void)
- 							   NULL,
- 							   &plv8_v8_flags,
- 							   NULL,
-							   PGC_USERSET, 0,
-+							   PGC_SUSET, 0,
- #if PG_VERSION_NUM >= 90100
- 							   NULL,
- #endif
--- a/compute/patches/plv8_v3.2.3.patch
+++ b/compute/patches/plv8_v3.2.3.patch
@@ -1,13 +0,0 @@
-diff --git a/plv8.cc b/plv8.cc
-index edfa2aa..623e7f2 100644
--- a/plv8.cc
-+++ b/plv8.cc
-@@ -385,7 +385,7 @@ _PG_init(void)
-                                    NULL,
-                                    &plv8_v8_flags,
-                                    NULL,
-                                   PGC_USERSET, 0,
-+                                   PGC_SUSET, 0,
- #if PG_VERSION_NUM >= 90100
-                                    NULL,
- #endif
--- a/compute_tools/src/catalog.rs
+++ b/compute_tools/src/catalog.rs
@@ -98,15 +98,13 @@ pub async fn get_database_schema(
        .kill_on_drop(true)
        .spawn()?;

-    let stdout = cmd
-        .stdout
-        .take()
-        .ok_or_else(|| std::io::Error::other("Failed to capture stdout."))?;
+    let stdout = cmd.stdout.take().ok_or_else(|| {
+        std::io::Error::new(std::io::ErrorKind::Other, "Failed to capture stdout.")
+    })?;

-    let stderr = cmd
-        .stderr
-        .take()
-        .ok_or_else(|| std::io::Error::other("Failed to capture stderr."))?;
+    let stderr = cmd.stderr.take().ok_or_else(|| {
+        std::io::Error::new(std::io::ErrorKind::Other, "Failed to capture stderr.")
+    })?;

    let mut stdout_reader = FramedRead::new(stdout, BytesCodec::new());
    let stderr_reader = BufReader::new(stderr);
@@ -130,7 +128,8 @@ pub async fn get_database_schema(
                }
            });

-            return Err(SchemaDumpError::IO(std::io::Error::other(
+            return Err(SchemaDumpError::IO(std::io::Error::new(
+                std::io::ErrorKind::Other,
                "failed to start pg_dump",
            )));
        }
--- a/compute_tools/src/compute.rs
+++ b/compute_tools/src/compute.rs
@@ -188,7 +188,7 @@ impl ComputeState {

        COMPUTE_CTL_UP.reset();
        COMPUTE_CTL_UP
-            .with_label_values(&[&BUILD_TAG, status.to_string().as_str()])
+            .with_label_values(&[&BUILD_TAG, format!("{}", status).as_str()])
            .set(1);
    }

@@ -360,14 +360,6 @@ impl ComputeNode {
            this.prewarm_postgres()?;
        }

-        // Set the up metric with Empty status before starting the HTTP server.
-        // That way on the first metric scrape, an external observer will see us
-        // as 'up' and 'empty' (unless the compute was started with a spec or
-        // already configured by control plane).
-        COMPUTE_CTL_UP
-            .with_label_values(&[&BUILD_TAG, ComputeStatus::Empty.to_string().as_str()])
-            .set(1);
-
        // Launch the external HTTP server first, so that we can serve control plane
        // requests while configuration is still in progress.
        crate::http::server::Server::External {
@@ -377,13 +369,19 @@ impl ComputeNode {
        }
        .launch(&this);

-        // The internal HTTP server could be launched later, but there isn't much
-        // sense in waiting.
+        // The internal HTTP server is needed for a further activation by control plane
+        // if compute was started for a pool, so we have to start server before hanging
+        // waiting for a spec.
        crate::http::server::Server::Internal {
            port: this.params.internal_http_port,
        }
        .launch(&this);

+        // HTTP server is running, so we can officially declare compute_ctl as 'up'
+        COMPUTE_CTL_UP
+            .with_label_values(&[&BUILD_TAG, ComputeStatus::Empty.to_string().as_str()])
+            .set(1);
+
        // If we got a spec from the CLI already, use that. Otherwise wait for the
        // control plane to pass it to us with a /configure HTTP request
        let pspec = if let Some(cli_spec) = cli_spec {
@@ -661,8 +659,15 @@ impl ComputeNode {
        }

        // Configure and start rsyslog for Postgres logs export
-        let conf = PostgresLogsRsyslogConfig::new(pspec.spec.logs_export_host.as_deref());
-        configure_postgres_logs_export(conf)?;
+        if self.has_feature(ComputeFeature::PostgresLogsExport) {
+            if let Some(ref project_id) = pspec.spec.cluster.cluster_id {
+                let host = PostgresLogsRsyslogConfig::default_host(project_id);
+                let conf = PostgresLogsRsyslogConfig::new(Some(&host));
+                configure_postgres_logs_export(conf)?;
+            } else {
+                warn!("not configuring rsyslog for Postgres logs export: project ID is missing")
+            }
+        }

        // Launch remaining service threads
        let _monitor_handle = launch_monitor(self);
@@ -1566,10 +1571,6 @@ impl ComputeNode {
            });
        }

-        // Reconfigure rsyslog for Postgres logs export
-        let conf = PostgresLogsRsyslogConfig::new(spec.logs_export_host.as_deref());
-        configure_postgres_logs_export(conf)?;
-
        // Write new config
        let pgdata_path = Path::new(&self.params.pgdata);
        config::write_postgres_conf(
--- a/compute_tools/src/config.rs
+++ b/compute_tools/src/config.rs
@@ -7,7 +7,7 @@ use std::io::prelude::*;
 use std::path::Path;

 use compute_api::responses::TlsConfig;
-use compute_api::spec::{ComputeAudit, ComputeMode, ComputeSpec, GenericOption};
+use compute_api::spec::{ComputeAudit, ComputeFeature, ComputeMode, ComputeSpec, GenericOption};

 use crate::pg_helpers::{
    GenericOptionExt, GenericOptionsSearch, PgOptionsSerialize, escape_conf_value,
@@ -255,7 +255,7 @@ pub fn write_postgres_conf(

    // We need Postgres to send logs to rsyslog so that we can forward them
    // further to customers' log aggregation systems.
-    if spec.logs_export_host.is_some() {
+    if spec.features.contains(&ComputeFeature::PostgresLogsExport) {
        writeln!(file, "log_destination='stderr,syslog'")?;
    }

--- a/compute_tools/src/http/middleware/authorize.rs
+++ b/compute_tools/src/http/middleware/authorize.rs
@@ -59,12 +59,9 @@ impl AsyncAuthorizeRequest<Body> for Authorize {
        Box::pin(async move {
            let request_id = request.extract_parts::<RequestId>().await.unwrap();

-            // TODO: Remove this stanza after teaching neon_local and the
-            // regression tests to use a JWT + JWKS.
-            //
-            // https://github.com/neondatabase/neon/issues/11316
-            if cfg!(feature = "testing") {
-                warn!(%request_id, "Skipping compute_ctl authorization check");
+            // TODO: Remove this check after a successful rollout
+            if jwks.keys.is_empty() {
+                warn!(%request_id, "Authorization has not been configured");

                return Ok(request);
            }
@@ -113,6 +110,8 @@ impl AsyncAuthorizeRequest<Body> for Authorize {
 impl Authorize {
    /// Verify the token using the JSON Web Key set and return the token data.
    fn verify(jwks: &JwkSet, token: &str, validation: &Validation) -> Result<TokenData<Claims>> {
+        debug_assert!(!jwks.keys.is_empty());
+
        for jwk in jwks.keys.iter() {
            let decoding_key = match DecodingKey::from_jwk(jwk) {
                Ok(key) => key,
--- a/compute_tools/src/http/openapi_spec.yaml
+++ b/compute_tools/src/http/openapi_spec.yaml
@@ -306,6 +306,36 @@ paths:
              schema:
                $ref: "#/components/schemas/GenericError"

+  /configure_telemetry:
+    post:
+      tags:
+        - Configure
+      summary: Configure rsyslog
+      description: |
+        This API endpoint configures rsyslog to forward Postgres logs
+        to a specified otel collector.
+      operationId: configureTelemetry
+      requestBody:
+        required: true
+        content:
+          application/json:
+            schema:
+              type: object
+              properties:
+                logs_export_host:
+                  type: string
+                  description: |
+                    Hostname and the port of the otel collector. Leave empty to disable logs forwarding.
+                    Example: config-shy-breeze-123-collector-monitoring.neon-telemetry.svc.cluster.local:54526
+      responses:
+        204:
+          description: "Telemetry configured successfully"
+        500:
+          content:
+            application/json:
+              schema:
+                $ref: "#/components/schemas/GenericError"
+
 components:
  securitySchemes:
    JWT:
--- a/compute_tools/src/http/routes/configure.rs
+++ b/compute_tools/src/http/routes/configure.rs
@@ -1,9 +1,11 @@
 use std::sync::Arc;

+use axum::body::Body;
 use axum::extract::State;
 use axum::response::Response;
-use compute_api::requests::ConfigurationRequest;
+use compute_api::requests::{ConfigurationRequest, ConfigureTelemetryRequest};
 use compute_api::responses::{ComputeStatus, ComputeStatusResponse};
+use compute_api::spec::ComputeFeature;
 use http::StatusCode;
 use tokio::task;
 use tracing::info;
@@ -11,6 +13,7 @@ use tracing::info;
 use crate::compute::{ComputeNode, ParsedSpec};
 use crate::http::JsonResponse;
 use crate::http::extract::Json;
+use crate::rsyslog::{PostgresLogsRsyslogConfig, configure_postgres_logs_export};

 // Accept spec in JSON format and request compute configuration. If anything
 // goes wrong after we set the compute status to `ConfigurationPending` and
@@ -92,3 +95,25 @@ pub(in crate::http) async fn configure(

    JsonResponse::success(StatusCode::OK, body)
 }
+
+pub(in crate::http) async fn configure_telemetry(
+    State(compute): State<Arc<ComputeNode>>,
+    request: Json<ConfigureTelemetryRequest>,
+) -> Response {
+    if !compute.has_feature(ComputeFeature::PostgresLogsExport) {
+        return JsonResponse::error(
+            StatusCode::PRECONDITION_FAILED,
+            "Postgres logs export feature is not enabled".to_string(),
+        );
+    }
+
+    let conf = PostgresLogsRsyslogConfig::new(request.logs_export_host.as_deref());
+    if let Err(err) = configure_postgres_logs_export(conf) {
+        return JsonResponse::error(StatusCode::INTERNAL_SERVER_ERROR, err.to_string());
+    }
+
+    Response::builder()
+        .status(StatusCode::NO_CONTENT)
+        .body(Body::from(""))
+        .unwrap()
+}
--- a/compute_tools/src/http/server.rs
+++ b/compute_tools/src/http/server.rs
@@ -87,6 +87,7 @@ impl From<&Server> for Router<Arc<ComputeNode>> {
                let authenticated_router = Router::<Arc<ComputeNode>>::new()
                    .route("/check_writability", post(check_writability::is_writable))
                    .route("/configure", post(configure::configure))
+                    .route("/configure_telemetry", post(configure::configure_telemetry))
                    .route("/database_schema", get(database_schema::get_schema_dump))
                    .route("/dbs_and_roles", get(dbs_and_roles::get_catalog_objects))
                    .route("/insights", get(insights::get_insights))
--- a/compute_tools/src/rsyslog.rs
+++ b/compute_tools/src/rsyslog.rs
@@ -119,9 +119,16 @@ impl<'a> PostgresLogsRsyslogConfig<'a> {
        };
        Ok(config_content)
    }
+
+    /// Returns the default host for otel collector that receives Postgres logs
+    pub fn default_host(project_id: &str) -> String {
+        format!(
+            "config-{}-collector.neon-telemetry.svc.cluster.local:10514",
+            project_id
+        )
+    }
 }

-/// Writes rsyslogd configuration for Postgres logs export and restarts rsyslog.
 pub fn configure_postgres_logs_export(conf: PostgresLogsRsyslogConfig) -> Result<()> {
    let new_config = conf.build()?;
    let current_config = PostgresLogsRsyslogConfig::current_config()?;
@@ -254,5 +261,16 @@ mod tests {
            let res = conf.build();
            assert!(res.is_err());
        }
+
+        {
+            // Verify config with default host
+            let host = PostgresLogsRsyslogConfig::default_host("shy-breeze-123");
+            let conf = PostgresLogsRsyslogConfig::new(Some(&host));
+            let res = conf.build();
+            assert!(res.is_ok());
+            let conf_str = res.unwrap();
+            assert!(conf_str.contains(r#"shy-breeze-123"#));
+            assert!(conf_str.contains(r#"port="10514""#));
+        }
    }
 }
--- a/compute_tools/src/spec_apply.rs
+++ b/compute_tools/src/spec_apply.rs
@@ -419,7 +419,7 @@ impl ComputeNode {
                .iter()
                .filter_map(|val| val.parse::<usize>().ok())
                .map(|val| if val > 1 { val - 1 } else { 1 })
-                .next_back()
+                .last()
                .unwrap_or(3)
        }
    }
--- a/control_plane/src/bin/neon_local.rs
+++ b/control_plane/src/bin/neon_local.rs
@@ -20,10 +20,8 @@ use compute_api::spec::ComputeMode;
 use control_plane::endpoint::ComputeControlPlane;
 use control_plane::local_env::{
    InitForceMode, LocalEnv, NeonBroker, NeonLocalInitConf, NeonLocalInitPageserverConf,
-    ObjectStorageConf, SafekeeperConf,
+    SafekeeperConf,
 };
-use control_plane::object_storage::OBJECT_STORAGE_DEFAULT_PORT;
-use control_plane::object_storage::ObjectStorage;
 use control_plane::pageserver::PageServerNode;
 use control_plane::safekeeper::SafekeeperNode;
 use control_plane::storage_controller::{
@@ -41,7 +39,7 @@ use pageserver_api::controller_api::{
 use pageserver_api::models::{
    ShardParameters, TenantConfigRequest, TimelineCreateRequest, TimelineInfo,
 };
-use pageserver_api::shard::{DEFAULT_STRIPE_SIZE, ShardCount, ShardStripeSize, TenantShardId};
+use pageserver_api::shard::{ShardCount, ShardStripeSize, TenantShardId};
 use postgres_backend::AuthType;
 use postgres_connection::parse_host_port;
 use safekeeper_api::membership::SafekeeperGeneration;
@@ -93,8 +91,6 @@ enum NeonLocalCmd {
    #[command(subcommand)]
    Safekeeper(SafekeeperCmd),
    #[command(subcommand)]
-    ObjectStorage(ObjectStorageCmd),
-    #[command(subcommand)]
    Endpoint(EndpointCmd),
    #[command(subcommand)]
    Mappings(MappingsCmd),
@@ -458,32 +454,6 @@ enum SafekeeperCmd {
    Restart(SafekeeperRestartCmdArgs),
 }

-#[derive(clap::Subcommand)]
-#[clap(about = "Manage object storage")]
-enum ObjectStorageCmd {
-    Start(ObjectStorageStartCmd),
-    Stop(ObjectStorageStopCmd),
-}
-
-#[derive(clap::Args)]
-#[clap(about = "Start object storage")]
-struct ObjectStorageStartCmd {
-    #[clap(short = 't', long, help = "timeout until we fail the command")]
-    #[arg(default_value = "10s")]
-    start_timeout: humantime::Duration,
-}
-
-#[derive(clap::Args)]
-#[clap(about = "Stop object storage")]
-struct ObjectStorageStopCmd {
-    #[arg(value_enum, default_value = "fast")]
-    #[clap(
-        short = 'm',
-        help = "If 'immediate', don't flush repository data at shutdown"
-    )]
-    stop_mode: StopMode,
-}
-
 #[derive(clap::Args)]
 #[clap(about = "Start local safekeeper")]
 struct SafekeeperStartCmdArgs {
@@ -789,7 +759,6 @@ fn main() -> Result<()> {
            }
            NeonLocalCmd::StorageBroker(subcmd) => rt.block_on(handle_storage_broker(&subcmd, env)),
            NeonLocalCmd::Safekeeper(subcmd) => rt.block_on(handle_safekeeper(&subcmd, env)),
-            NeonLocalCmd::ObjectStorage(subcmd) => rt.block_on(handle_object_storage(&subcmd, env)),
            NeonLocalCmd::Endpoint(subcmd) => rt.block_on(handle_endpoint(&subcmd, env)),
            NeonLocalCmd::Mappings(subcmd) => handle_mappings(&subcmd, env),
        };
@@ -1006,9 +975,6 @@ fn handle_init(args: &InitCmdArgs) -> anyhow::Result<LocalEnv> {
                    }
                })
                .collect(),
-            object_storage: ObjectStorageConf {
-                port: OBJECT_STORAGE_DEFAULT_PORT,
-            },
            pg_distrib_dir: None,
            neon_distrib_dir: None,
            default_tenant_id: TenantId::from_array(std::array::from_fn(|_| 0)),
@@ -1117,7 +1083,7 @@ async fn handle_tenant(subcmd: &TenantCmd, env: &mut local_env::LocalEnv) -> any
                        stripe_size: args
                            .shard_stripe_size
                            .map(ShardStripeSize)
-                            .unwrap_or(DEFAULT_STRIPE_SIZE),
+                            .unwrap_or(ShardParameters::DEFAULT_STRIPE_SIZE),
                    },
                    placement_policy: args.placement_policy.clone(),
                    config: tenant_conf,
@@ -1430,7 +1396,7 @@ async fn handle_endpoint(subcmd: &EndpointCmd, env: &local_env::LocalEnv) -> Res
                    vec![(parsed.0, parsed.1.unwrap_or(5432))],
                    // If caller is telling us what pageserver to use, this is not a tenant which is
                    // full managed by storage controller, therefore not sharded.
-                    DEFAULT_STRIPE_SIZE,
+                    ShardParameters::DEFAULT_STRIPE_SIZE,
                )
            } else {
                // Look up the currently attached location of the tenant, and its striping metadata,
@@ -1717,41 +1683,6 @@ async fn handle_safekeeper(subcmd: &SafekeeperCmd, env: &local_env::LocalEnv) ->
    Ok(())
 }

-async fn handle_object_storage(subcmd: &ObjectStorageCmd, env: &local_env::LocalEnv) -> Result<()> {
-    use ObjectStorageCmd::*;
-    let storage = ObjectStorage::from_env(env);
-
-    // In tests like test_forward_compatibility or test_graceful_cluster_restart
-    // old neon binaries (without object_storage) are present
-    if !storage.bin.exists() {
-        eprintln!(
-            "{} binary not found. Ignore if this is a compatibility test",
-            storage.bin
-        );
-        return Ok(());
-    }
-
-    match subcmd {
-        Start(ObjectStorageStartCmd { start_timeout }) => {
-            if let Err(e) = storage.start(start_timeout).await {
-                eprintln!("object_storage start failed: {e}");
-                exit(1);
-            }
-        }
-        Stop(ObjectStorageStopCmd { stop_mode }) => {
-            let immediate = match stop_mode {
-                StopMode::Fast => false,
-                StopMode::Immediate => true,
-            };
-            if let Err(e) = storage.stop(immediate) {
-                eprintln!("proxy stop failed: {e}");
-                exit(1);
-            }
-        }
-    };
-    Ok(())
-}
-
 async fn handle_storage_broker(subcmd: &StorageBrokerCmd, env: &local_env::LocalEnv) -> Result<()> {
    match subcmd {
        StorageBrokerCmd::Start(args) => {
@@ -1846,13 +1777,6 @@ async fn handle_start_all_impl(
                    .map_err(|e| e.context(format!("start safekeeper {}", safekeeper.id)))
            });
        }
-
-        js.spawn(async move {
-            ObjectStorage::from_env(env)
-                .start(&retry_timeout)
-                .await
-                .map_err(|e| e.context("start object_storage"))
-        });
    })();

    let mut errors = Vec::new();
@@ -1950,11 +1874,6 @@ async fn try_stop_all(env: &local_env::LocalEnv, immediate: bool) {
        }
    }

-    let storage = ObjectStorage::from_env(env);
-    if let Err(e) = storage.stop(immediate) {
-        eprintln!("object_storage stop failed: {:#}", e);
-    }
-
    for ps_conf in &env.pageservers {
        let pageserver = PageServerNode::from_env(env, ps_conf);
        if let Err(e) = pageserver.stop(immediate) {
--- a/control_plane/src/endpoint.rs
+++ b/control_plane/src/endpoint.rs
@@ -670,7 +670,6 @@ impl Endpoint {
            reconfigure_concurrency: self.reconfigure_concurrency,
            drop_subscriptions_before_start: self.drop_subscriptions_before_start,
            audit_log_level: ComputeAudit::Disabled,
-            logs_export_host: None::<String>,
        };

        // this strange code is needed to support respec() in tests
--- a/control_plane/src/lib.rs
+++ b/control_plane/src/lib.rs
@@ -10,7 +10,6 @@ mod background_process;
 pub mod broker;
 pub mod endpoint;
 pub mod local_env;
-pub mod object_storage;
 pub mod pageserver;
 pub mod postgresql_conf;
 pub mod safekeeper;
--- a/control_plane/src/local_env.rs
+++ b/control_plane/src/local_env.rs
@@ -15,10 +15,9 @@ use clap::ValueEnum;
 use postgres_backend::AuthType;
 use reqwest::Url;
 use serde::{Deserialize, Serialize};
-use utils::auth::encode_from_key_file;
+use utils::auth::{Claims, encode_from_key_file};
 use utils::id::{NodeId, TenantId, TenantTimelineId, TimelineId};

-use crate::object_storage::{OBJECT_STORAGE_REMOTE_STORAGE_DIR, ObjectStorage};
 use crate::pageserver::{PAGESERVER_REMOTE_STORAGE_DIR, PageServerNode};
 use crate::safekeeper::SafekeeperNode;

@@ -56,7 +55,6 @@ pub struct LocalEnv {

    // used to issue tokens during e.g pg start
    pub private_key_path: PathBuf,
-    pub public_key_path: PathBuf,

    pub broker: NeonBroker,

@@ -70,8 +68,6 @@ pub struct LocalEnv {

    pub safekeepers: Vec<SafekeeperConf>,

-    pub object_storage: ObjectStorageConf,
-
    // Control plane upcall API for pageserver: if None, we will not run storage_controller  If set, this will
    // be propagated into each pageserver's configuration.
    pub control_plane_api: Url,
@@ -99,7 +95,6 @@ pub struct OnDiskConfig {
    pub neon_distrib_dir: PathBuf,
    pub default_tenant_id: Option<TenantId>,
    pub private_key_path: PathBuf,
-    pub public_key_path: PathBuf,
    pub broker: NeonBroker,
    pub storage_controller: NeonStorageControllerConf,
    #[serde(
@@ -108,7 +103,6 @@ pub struct OnDiskConfig {
    )]
    pub pageservers: Vec<PageServerConf>,
    pub safekeepers: Vec<SafekeeperConf>,
-    pub object_storage: ObjectStorageConf,
    pub control_plane_api: Option<Url>,
    pub control_plane_hooks_api: Option<Url>,
    pub control_plane_compute_hook_api: Option<Url>,
@@ -142,18 +136,11 @@ pub struct NeonLocalInitConf {
    pub storage_controller: Option<NeonStorageControllerConf>,
    pub pageservers: Vec<NeonLocalInitPageserverConf>,
    pub safekeepers: Vec<SafekeeperConf>,
-    pub object_storage: ObjectStorageConf,
    pub control_plane_api: Option<Url>,
    pub control_plane_hooks_api: Option<Url>,
    pub generate_local_ssl_certs: bool,
 }

-#[derive(Serialize, Default, Deserialize, PartialEq, Eq, Clone, Debug)]
-#[serde(default)]
-pub struct ObjectStorageConf {
-    pub port: u16,
-}
-
 /// Broker config for cluster internal communication.
 #[derive(Serialize, Deserialize, PartialEq, Eq, Clone, Debug)]
 #[serde(default)]
@@ -411,10 +398,6 @@ impl LocalEnv {
        self.pg_dir(pg_version, "lib")
    }

-    pub fn object_storage_bin(&self) -> PathBuf {
-        self.neon_distrib_dir.join("object_storage")
-    }
-
    pub fn pageserver_bin(&self) -> PathBuf {
        self.neon_distrib_dir.join("pageserver")
    }
@@ -448,10 +431,6 @@ impl LocalEnv {
        self.base_data_dir.join("safekeepers").join(data_dir_name)
    }

-    pub fn object_storage_data_dir(&self) -> PathBuf {
-        self.base_data_dir.join("object_storage")
-    }
-
    pub fn get_pageserver_conf(&self, id: NodeId) -> anyhow::Result<&PageServerConf> {
        if let Some(conf) = self.pageservers.iter().find(|node| node.id == id) {
            Ok(conf)
@@ -603,7 +582,6 @@ impl LocalEnv {
                neon_distrib_dir,
                default_tenant_id,
                private_key_path,
-                public_key_path,
                broker,
                storage_controller,
                pageservers,
@@ -613,7 +591,6 @@ impl LocalEnv {
                control_plane_compute_hook_api: _,
                branch_name_mappings,
                generate_local_ssl_certs,
-                object_storage,
            } = on_disk_config;
            LocalEnv {
                base_data_dir: repopath.to_owned(),
@@ -621,7 +598,6 @@ impl LocalEnv {
                neon_distrib_dir,
                default_tenant_id,
                private_key_path,
-                public_key_path,
                broker,
                storage_controller,
                pageservers,
@@ -630,7 +606,6 @@ impl LocalEnv {
                control_plane_hooks_api,
                branch_name_mappings,
                generate_local_ssl_certs,
-                object_storage,
            }
        };

@@ -730,7 +705,6 @@ impl LocalEnv {
                neon_distrib_dir: self.neon_distrib_dir.clone(),
                default_tenant_id: self.default_tenant_id,
                private_key_path: self.private_key_path.clone(),
-                public_key_path: self.public_key_path.clone(),
                broker: self.broker.clone(),
                storage_controller: self.storage_controller.clone(),
                pageservers: vec![], // it's skip_serializing anyway
@@ -740,7 +714,6 @@ impl LocalEnv {
                control_plane_compute_hook_api: None,
                branch_name_mappings: self.branch_name_mappings.clone(),
                generate_local_ssl_certs: self.generate_local_ssl_certs,
-                object_storage: self.object_storage.clone(),
            },
        )
    }
@@ -757,7 +730,7 @@ impl LocalEnv {
    }

    // this function is used only for testing purposes in CLI e g generate tokens during init
-    pub fn generate_auth_token<S: Serialize>(&self, claims: &S) -> anyhow::Result<String> {
+    pub fn generate_auth_token(&self, claims: &Claims) -> anyhow::Result<String> {
        let private_key_path = self.get_private_key_path();
        let key_data = fs::read(private_key_path)?;
        encode_from_key_file(claims, &key_data)
@@ -824,7 +797,6 @@ impl LocalEnv {
            control_plane_api,
            generate_local_ssl_certs,
            control_plane_hooks_api,
-            object_storage,
        } = conf;

        // Find postgres binaries.
@@ -856,7 +828,6 @@ impl LocalEnv {
        )
        .context("generate auth keys")?;
        let private_key_path = PathBuf::from("auth_private_key.pem");
-        let public_key_path = PathBuf::from("auth_public_key.pem");

        // create the runtime type because the remaining initialization code below needs
        // a LocalEnv instance op operation
@@ -867,7 +838,6 @@ impl LocalEnv {
            neon_distrib_dir,
            default_tenant_id: Some(default_tenant_id),
            private_key_path,
-            public_key_path,
            broker,
            storage_controller: storage_controller.unwrap_or_default(),
            pageservers: pageservers.iter().map(Into::into).collect(),
@@ -876,7 +846,6 @@ impl LocalEnv {
            control_plane_hooks_api,
            branch_name_mappings: Default::default(),
            generate_local_ssl_certs,
-            object_storage,
        };

        if generate_local_ssl_certs {
@@ -904,13 +873,8 @@ impl LocalEnv {
                .context("pageserver init failed")?;
        }

-        ObjectStorage::from_env(&env)
-            .init()
-            .context("object storage init failed")?;
-
        // setup remote remote location for default LocalFs remote storage
        std::fs::create_dir_all(env.base_data_dir.join(PAGESERVER_REMOTE_STORAGE_DIR))?;
-        std::fs::create_dir_all(env.base_data_dir.join(OBJECT_STORAGE_REMOTE_STORAGE_DIR))?;

        env.persist_config()
    }
--- a/control_plane/src/object_storage.rs
+++ b/control_plane/src/object_storage.rs
@@ -1,107 +0,0 @@
-use crate::background_process::{self, start_process, stop_process};
-use crate::local_env::LocalEnv;
-use anyhow::anyhow;
-use anyhow::{Context, Result};
-use camino::Utf8PathBuf;
-use std::io::Write;
-use std::time::Duration;
-
-/// Directory within .neon which will be used by default for LocalFs remote storage.
-pub const OBJECT_STORAGE_REMOTE_STORAGE_DIR: &str = "local_fs_remote_storage/object_storage";
-pub const OBJECT_STORAGE_DEFAULT_PORT: u16 = 9993;
-
-pub struct ObjectStorage {
-    pub bin: Utf8PathBuf,
-    pub data_dir: Utf8PathBuf,
-    pub pemfile: Utf8PathBuf,
-    pub port: u16,
-}
-
-impl ObjectStorage {
-    pub fn from_env(env: &LocalEnv) -> ObjectStorage {
-        ObjectStorage {
-            bin: Utf8PathBuf::from_path_buf(env.object_storage_bin()).unwrap(),
-            data_dir: Utf8PathBuf::from_path_buf(env.object_storage_data_dir()).unwrap(),
-            pemfile: Utf8PathBuf::from_path_buf(env.public_key_path.clone()).unwrap(),
-            port: env.object_storage.port,
-        }
-    }
-
-    fn config_path(&self) -> Utf8PathBuf {
-        self.data_dir.join("object_storage.json")
-    }
-
-    fn listen_addr(&self) -> Utf8PathBuf {
-        format!("127.0.0.1:{}", self.port).into()
-    }
-
-    pub fn init(&self) -> Result<()> {
-        println!("Initializing object storage in {:?}", self.data_dir);
-        let parent = self.data_dir.parent().unwrap();
-
-        #[derive(serde::Serialize)]
-        struct Cfg {
-            listen: Utf8PathBuf,
-            pemfile: Utf8PathBuf,
-            local_path: Utf8PathBuf,
-            r#type: String,
-        }
-        let cfg = Cfg {
-            listen: self.listen_addr(),
-            pemfile: parent.join(self.pemfile.clone()),
-            local_path: parent.join(OBJECT_STORAGE_REMOTE_STORAGE_DIR),
-            r#type: "LocalFs".to_string(),
-        };
-        std::fs::create_dir_all(self.config_path().parent().unwrap())?;
-        std::fs::write(self.config_path(), serde_json::to_string(&cfg)?)
-            .context("write object storage config")?;
-        Ok(())
-    }
-
-    pub async fn start(&self, retry_timeout: &Duration) -> Result<()> {
-        println!("Starting s3 proxy at {}", self.listen_addr());
-        std::io::stdout().flush().context("flush stdout")?;
-
-        let process_status_check = || async {
-            tokio::time::sleep(Duration::from_millis(500)).await;
-            let res = reqwest::Client::new()
-                .get(format!("http://{}/metrics", self.listen_addr()))
-                .send()
-                .await;
-            match res {
-                Ok(response) if response.status().is_success() => Ok(true),
-                Ok(_) => Err(anyhow!("Failed to query /metrics")),
-                Err(e) => Err(anyhow!("Failed to check node status: {e}")),
-            }
-        };
-
-        let res = start_process(
-            "object_storage",
-            &self.data_dir.clone().into_std_path_buf(),
-            &self.bin.clone().into_std_path_buf(),
-            vec![self.config_path().to_string()],
-            vec![("RUST_LOG".into(), "debug".into())],
-            background_process::InitialPidFile::Create(self.pid_file()),
-            retry_timeout,
-            process_status_check,
-        )
-        .await;
-        if res.is_err() {
-            eprintln!("Logs:\n{}", std::fs::read_to_string(self.log_file())?);
-        }
-
-        res
-    }
-
-    pub fn stop(&self, immediate: bool) -> anyhow::Result<()> {
-        stop_process(immediate, "object_storage", &self.pid_file())
-    }
-
-    fn log_file(&self) -> Utf8PathBuf {
-        self.data_dir.join("object_storage.log")
-    }
-
-    fn pid_file(&self) -> Utf8PathBuf {
-        self.data_dir.join("object_storage.pid")
-    }
-}
--- a/control_plane/src/pageserver.rs
+++ b/control_plane/src/pageserver.rs
@@ -545,11 +545,6 @@ impl PageServerNode {
                .map(|x| x.parse::<u64>())
                .transpose()
                .context("Failed to parse 'gc_compaction_ratio_percent' as integer")?,
-            sampling_ratio: settings
-                .remove("sampling_ratio")
-                .map(serde_json::from_str)
-                .transpose()
-                .context("Falied to parse 'sampling_ratio'")?,
        };
        if !settings.is_empty() {
            bail!("Unrecognized tenant settings: {settings:?}")
--- a/control_plane/storcon_cli/src/main.rs
+++ b/control_plane/storcon_cli/src/main.rs
@@ -385,6 +385,8 @@ where
 async fn main() -> anyhow::Result<()> {
    let cli = Cli::parse();

+    let storcon_client = Client::new(cli.api.clone(), cli.jwt.clone());
+
    let ssl_ca_certs = match &cli.ssl_ca_file {
        Some(ssl_ca_file) => {
            let buf = tokio::fs::read(ssl_ca_file).await?;
@@ -399,11 +401,9 @@ async fn main() -> anyhow::Result<()> {
    }
    let http_client = http_client.build()?;

-    let storcon_client = Client::new(http_client.clone(), cli.api.clone(), cli.jwt.clone());
-
    let mut trimmed = cli.api.to_string();
    trimmed.pop();
-    let vps_client = mgmt_api::Client::new(http_client.clone(), trimmed, cli.jwt.as_deref());
+    let vps_client = mgmt_api::Client::new(http_client, trimmed, cli.jwt.as_deref());

    match cli.command {
        Command::NodeRegister {
@@ -941,7 +941,7 @@ async fn main() -> anyhow::Result<()> {
            let mut node_to_fill_descs = Vec::new();

            for desc in node_descs {
-                let to_drain = nodes.contains(&desc.id);
+                let to_drain = nodes.iter().any(|id| *id == desc.id);
                if to_drain {
                    node_to_drain_descs.push(desc);
                } else {
@@ -1056,7 +1056,7 @@ async fn main() -> anyhow::Result<()> {
            const DEFAULT_MIGRATE_CONCURRENCY: usize = 8;
            let mut stream = futures::stream::iter(moves)
                .map(|mv| {
-                    let client = Client::new(http_client.clone(), cli.api.clone(), cli.jwt.clone());
+                    let client = Client::new(cli.api.clone(), cli.jwt.clone());
                    async move {
                        client
                            .dispatch::<TenantShardMigrateRequest, TenantShardMigrateResponse>(
--- a/docs/SUMMARY.md
+++ b/docs/SUMMARY.md
@@ -21,7 +21,6 @@ in this repository.
    - [WAL Redo](./pageserver-walredo.md)
    - [Page cache](./pageserver-pagecache.md)
    - [Storage](./pageserver-storage.md)
-    - [Compaction](./pageserver-compaction.md)
    - [Processing a GetPage request](./pageserver-processing-getpage.md)
    - [Processing WAL](./pageserver-processing-wal.md)

--- a/docs/pageserver-compaction.md
+++ b/docs/pageserver-compaction.md
@@ -1,110 +0,0 @@
-# Pageserver Compaction
-
-Lifted from <https://www.notion.so/neondatabase/Rough-Notes-on-Compaction-1baf189e004780859e65ef63b85cfa81?pvs=4>.
-
-Updated 2025-03-26.
-
-## Pages and WAL
-
-Postgres stores data in 8 KB pages, identified by a page number.
-
-The WAL contains a sequence of page writes: either images (complete page contents) or deltas (patches applied to images). Each write is identified by its byte position in the WAL, aka LSN. 
-
-Each page version is thus identified by page@LSN. Postgres may read pages at past LSNs.
-
-Pageservers ingest WAL by writing WAL records into a key/value store keyed by page@LSN.
-
-Pageservers materialize pages for Postgres reads by finding the most recent page image and applying all subsequent page deltas, up to the read LSN.
-
-## Compaction: Why?
-
-Pageservers store page@LSN keys in a key/value store using a custom variant of an LSM tree. Each timeline on each tenant shard has its own LSM tree.
-
-When Pageservers write new page@LSN entries, they are appended unordered to an ephemeral layer file. When the ephemeral layer file exceeds `checkpoint_distance` (default 256 MB), the key/value pairs are sorted by key and written out to a layer file (for efficient lookups).
-
-As WAL writes continue, more layer files accumulate.
-
-Reads must search through the layer files to find the page’s image and deltas. The more layer files accumulate, the more la yer files reads must search through before they find a page image, aka read amplification.
-
-Compaction’s job is to:
-
- Reduce read amplification by reorganizing and combining layer files.
- Remove old garbage from layer files.
-
-As part of this, it may combine several page deltas into a single page image where possible.
-
-## Compaction: How?
-
-Neon uses a non-standard variant of an LSM tree made up of two levels of layer files: L0 and L1.
-
-Compaction runs in two phases: L0→L1 compaction, and L1 image compaction.
-
-L0 contains a stack of L0 layers at decreasing LSN ranges. These have been flushed sequentially from ephemeral layers. Each L0 layer covers the entire page space (page 0 to ~infinity) and the LSN range that was ingested into it. L0 layers are therefore particularly bad for read amp, since every read must search all L0 layers below the read LSN. For example:
-
-```
-| Page 0-99 @ LSN 0400-04ff |
-| Page 0-99 @ LSN 0300-03ff |
-| Page 0-99 @ LSN 0200-02ff |
-| Page 0-99 @ LSN 0100-01ff |
-| Page 0-99 @ LSN 0000-00ff |
-```
-
-L0→L1 compaction takes the bottom-most chunk of L0 layer files of between `compaction_threshold` (default 10) and `compaction_upper_limit` (default 20) layers. It uses merge-sort to write out sorted L1 delta layers of size `compaction_target_size` (default 128 MB).
-
-L1 typically consists of a “bed” of image layers with materialized page images at a specific LSN, and then delta layers of various page/LSN ranges above them with page deltas. For example:
-
-```
-Delta layers:               |     30-84@0310-04ff      |
-Delta layers:    | 10-42@0200-02ff |           | 65-92@0174-02aa |
-Image layers: |    0-39@0100    |    40-79@0100    |    80-99@0100    |
-```
-
-L1 image compaction scans across the L1 keyspace at some LSN, materializes page images by reading the image and delta layers below the LSN (via vectored reads), and writes out new sorted image layers of roughly size `compaction_target_size` (default 128 MB) at that LSN.
-
-Layer files below the new image files’ LSN can be garbage collected when they are no longer needed for PITR.
-
-Even though the old layer files are not immediately garbage collected, the new image layers help with read amp because reads can stop traversing the layer stack as soon as they encounter a page image.
-
-## Compaction: When?
-
-Pageservers run a `compaction_loop` background task for each tenant shard. Every `compaction_period` (default 20 seconds) it will wake up and check if any of the shard’s timelines need compaction. Additionally, L0 layer flushes will eagerly wake the compaction loop if the L0 count exceeds `compaction_threshold` (default 10).
-
-L0 compaction runs if the number of L0 layers exceeds `compaction_threshold` (default 10).
-
-L1 image compaction runs across sections of the L1 keyspace that have at least `image_creation_threshold` (default 3) delta layers overlapping image layers.
-
-At most `CONCURRENT_BACKGROUND_TASKS` (default 3 / 4 * CPUs = 6) background tasks can run concurrently on a Pageserver, including compaction. Further compaction tasks must wait.
-
-Because L0 layers cause the most read amp (they overlap the entire keyspace and only contain page deltas), they are aggressively compacted down:
-
- L0 is compacted down across all tenant timelines before L1 compaction is attempted (`compaction_l0_first`).
- L0 compaction uses a separate concurrency limit of `CONCURRENT_L0_COMPACTION_TASKS` (default 3 / 4 * CPUs = 6) to avoid waiting for other tasks (`compaction_l0_semaphore`).
- If L0 compaction is needed on any tenant timeline, L1 image compaction will yield to start an immediate L0 compaction run (except for compaction run via admin APIs).
-
-## Backpressure
-
-With sustained heavy write loads, new L0 layers may be flushed faster than they can be compacted down. This can cause an unbounded buildup of read amplification and compaction debt, which can take hours to resolve even after the writes stop.
-
-To avoid this and allow compaction to keep up, layer flushes will slow writes down to apply backpressure on the workload:
-
- At `l0_flush_delay_threshold` (default 30) L0 layers, layer flushes are delayed by the flush duration, such that they take 2x as long.
- At `l0_flush_stall_threshold` (default disabled) L0 layers, layer flushes stall entirely until the L0 count falls back below the threshold. This is currently disabled because we don’t trust L0 compaction to be responsive enough.
-
-This backpressure is propagated to the compute by waiting for layer flushes when WAL ingestion rolls the ephemeral layer. The compute will significantly slow down WAL writes at:
-
- `max_replication_write_lag` (default 500 MB), when Pageserver WAL ingestion lags
- `max_replication_flush_lag` (default 10 GB), when Pageserver L0 flushes lag
-
-Combined, this means that the compute will backpressure when there are 30 L0 layers (30 * 256 MB = 7.7 GB) and the Pageserver WAL ingestion lags the compute by 500 MB, for a total of ~8 GB L0+ephemeral compaction debt on a single shard.
-
-Since we only delay L0 flushes by 2x when backpressuring, and haven’t enabled stalls, it is still possible for read amp to increase unbounded if compaction is too slow (although we haven’t seen this in practice). But this is considered better than stalling flushes and causing unavailability for as long as it takes L0 compaction to react, since we don’t trust it to be fast enough — at the expense of continually increasing read latency and CPU usage for this tenant. We should either enable stalls when we have enough confidence in L0 compaction, or scale the flush delay by the number of L0 layers to apply increasing backpressure.
-
-## Circuit Breaker
-
-Compaction can fail, often repeatedly. This can happen e.g. due to data corruption, faulty hardware, S3 outages, etc.
-
-If compaction fails, the compaction loop will naïvely try and fail again almost immediately. It may only fail after doing a significant amount of wasted work, while holding onto the background task semaphore.
-
-To avoid repeatedly doing wasted work and starving out other compaction jobs, each tenant has a compaction circuit breaker. After 5 repeated compaction failures, the circuit breaker trips and disables compaction for the next 24 hours, before resetting the breaker and trying again. This disables compaction across all tenant timelines (faulty or not).
-
-Disabling compaction for a long time is dangerous, since it can lead to unbounded read amp and compaction debt, and continuous workload backpressure. However, continually failing would not help either. Tripped circuit breakers trigger an alert and must be investigated promptly.
--- a/docs/storage_controller.md
+++ b/docs/storage_controller.md
@@ -151,7 +151,7 @@ Example body:
 ```
 {
  "tenant_id": "1f359dd625e519a1a4e8d7509690f6fc",
-  "stripe_size": 2048,
+  "stripe_size": 32768,
  "shards": [
      {"node_id": 344, "shard_number": 0},
      {"node_id": 722, "shard_number": 1},
--- a/libs/compute_api/src/requests.rs
+++ b/libs/compute_api/src/requests.rs
@@ -30,3 +30,9 @@ pub struct SetRoleGrantsRequest {
    pub privileges: Vec<Privilege>,
    pub role: PgIdent,
 }
+
+/// Request of the /configure_telemetry API
+#[derive(Debug, Deserialize, Serialize)]
+pub struct ConfigureTelemetryRequest {
+    pub logs_export_host: Option<String>,
+}
--- a/libs/compute_api/src/spec.rs
+++ b/libs/compute_api/src/spec.rs
@@ -168,10 +168,6 @@ pub struct ComputeSpec {
    /// Extensions should be present in shared_preload_libraries
    #[serde(default)]
    pub audit_log_level: ComputeAudit,
-
-    /// Hostname and the port of the otel collector. Leave empty to disable Postgres logs forwarding.
-    /// Example: config-shy-breeze-123-collector-monitoring.neon-telemetry.svc.cluster.local:10514
-    pub logs_export_host: Option<String>,
 }

 /// Feature flag to signal `compute_ctl` to enable certain experimental functionality.
@@ -183,6 +179,9 @@ pub enum ComputeFeature {
    /// track short-lived connections as user activity.
    ActivityMonitorExperimental,

+    /// Allow to configure rsyslog for Postgres logs export
+    PostgresLogsExport,
+
    /// This is a special feature flag that is used to represent unknown feature flags.
    /// Basically all unknown to enum flags are represented as this one. See unit test
    /// `parse_unknown_features()` for more details.
--- a/libs/http-utils/src/server.rs
+++ b/libs/http-utils/src/server.rs
@@ -91,14 +91,14 @@ impl Server {
                                        Ok(tls_stream) => tls_stream,
                                        Err(err) => {
                                            if !suppress_io_error(&err) {
-                                                info!(%remote_addr, "Failed to accept TLS connection: {err:#}");
+                                                info!("Failed to accept TLS connection: {err:#}");
                                            }
                                            return;
                                        }
                                    };
                                    if let Err(err) = Self::serve_connection(tls_stream, service, cancel).await {
                                        if !suppress_hyper_error(&err) {
-                                            info!(%remote_addr, "Failed to serve HTTPS connection: {err:#}");
+                                            info!("Failed to serve HTTPS connection: {err:#}");
                                        }
                                    }
                                }
@@ -106,7 +106,7 @@ impl Server {
                                    // Handle HTTP connection.
                                    if let Err(err) = Self::serve_connection(tcp_stream, service, cancel).await {
                                        if !suppress_hyper_error(&err) {
-                                            info!(%remote_addr, "Failed to serve HTTP connection: {err:#}");
+                                            info!("Failed to serve HTTP connection: {err:#}");
                                        }
                                    }
                                }
--- a/libs/pageserver_api/Cargo.toml
+++ b/libs/pageserver_api/Cargo.toml
@@ -34,7 +34,6 @@ postgres_backend.workspace = true
 nix = {workspace = true, optional = true}
 reqwest.workspace = true
 rand.workspace = true
-tracing-utils.workspace = true

 [dev-dependencies]
 bincode.workspace = true
--- a/libs/pageserver_api/src/config.rs
+++ b/libs/pageserver_api/src/config.rs
@@ -51,54 +51,9 @@ pub struct NodeMetadata {
 /// If there cannot be a static default value because we need to make runtime
 /// checks to determine the default, make it an `Option` (which defaults to None).
 /// The runtime check should be done in the consuming crate, i.e., `pageserver`.
-///
-/// Unknown fields are silently ignored during deserialization.
-/// The alternative, which we used in the past, was to set `deny_unknown_fields`,
-/// which fails deserialization, and hence pageserver startup, if there is an unknown field.
-/// The reason we don't do that anymore is that it complicates
-/// usage of config fields for feature flagging, which we commonly do for
-/// region-by-region rollouts.
-/// The complications mainly arise because the `pageserver.toml` contents on a
-/// prod server have a separate lifecycle from the pageserver binary.
-/// For instance, `pageserver.toml` contents today are defined in the internal
-/// infra repo, and thus introducing a new config field to pageserver and
-/// rolling it out to prod servers are separate commits in separate repos
-/// that can't be made or rolled back atomically.
-/// Rollbacks in particular pose a risk with deny_unknown_fields because
-/// the old pageserver binary may reject a new config field, resulting in
-/// an outage unless the person doing the pageserver rollback remembers
-/// to also revert the commit that added the config field in to the
-/// `pageserver.toml` templates in the internal infra repo.
-/// (A pre-deploy config check would eliminate this risk during rollbacks,
-///  cf [here](https://github.com/neondatabase/cloud/issues/24349).)
-/// In addition to this compatibility problem during emergency rollbacks,
-/// deny_unknown_fields adds further complications when decomissioning a feature
-/// flag: with deny_unknown_fields, we can't remove a flag from the [`ConfigToml`]
-/// until all prod servers' `pageserver.toml` files have been updated to a version
-/// that doesn't specify the flag. Otherwise new software would fail to start up.
-/// This adds the requirement for an intermediate step where the new config field
-/// is accepted but ignored, prolonging the decomissioning process by an entire
-/// release cycle.
-/// By contrast  with unknown fields silently ignored, decomissioning a feature
-/// flag is a one-step process: we can skip the intermediate step and straight
-/// remove the field from the [`ConfigToml`]. We leave the field in the
-/// `pageserver.toml` files on prod servers until we reach certainty that we
-/// will not roll back to old software whose behavior was dependent on config.
-/// Then we can remove the field from the templates in the internal infra repo.
-/// This process is [documented internally](
-/// https://docs.neon.build/storage/pageserver_configuration.html).
-///
-/// Note that above relaxed compatbility for the config format does NOT APPLY
-/// TO THE STORAGE FORMAT. As general guidance, when introducing storage format
-/// changes, ensure that the potential rollback target version will be compatible
-/// with the new format. This must hold regardless of what flags are set in in the `pageserver.toml`:
-/// any format version that exists in an environment must be compatible with the software that runs there.
-/// Use a pageserver.toml flag only to gate whether software _writes_ the new format.
-/// For more compatibility considerations, refer to [internal docs](
-/// https://docs.neon.build/storage/compat.html?highlight=compat#format-versions--compatibility)
 #[serde_as]
 #[derive(Clone, Debug, serde::Deserialize, serde::Serialize)]
-#[serde(default)]
+#[serde(default, deny_unknown_fields)]
 pub struct ConfigToml {
    // types mapped 1:1 into the runtime PageServerConfig type
    pub listen_pg_addr: String,
@@ -179,10 +134,10 @@ pub struct ConfigToml {
    pub load_previous_heatmap: Option<bool>,
    #[serde(skip_serializing_if = "Option::is_none")]
    pub generate_unarchival_heatmap: Option<bool>,
-    pub tracing: Option<Tracing>,
 }

 #[derive(Debug, Clone, PartialEq, Eq, serde::Serialize, serde::Deserialize)]
+#[serde(deny_unknown_fields)]
 pub struct DiskUsageEvictionTaskConfig {
    pub max_usage_pct: utils::serde_percent::Percent,
    pub min_avail_bytes: u64,
@@ -197,11 +152,13 @@ pub struct DiskUsageEvictionTaskConfig {

 #[derive(Debug, Clone, PartialEq, Eq, serde::Serialize, serde::Deserialize)]
 #[serde(tag = "mode", rename_all = "kebab-case")]
+#[serde(deny_unknown_fields)]
 pub enum PageServicePipeliningConfig {
    Serial,
    Pipelined(PageServicePipeliningConfigPipelined),
 }
 #[derive(Debug, Clone, PartialEq, Eq, serde::Serialize, serde::Deserialize)]
+#[serde(deny_unknown_fields)]
 pub struct PageServicePipeliningConfigPipelined {
    /// Causes runtime errors if larger than max get_vectored batch size.
    pub max_batch_size: NonZeroUsize,
@@ -217,6 +174,7 @@ pub enum PageServiceProtocolPipelinedExecutionStrategy {

 #[derive(Debug, Clone, PartialEq, Eq, serde::Serialize, serde::Deserialize)]
 #[serde(tag = "mode", rename_all = "kebab-case")]
+#[serde(deny_unknown_fields)]
 pub enum GetVectoredConcurrentIo {
    /// The read path is fully sequential: layers are visited
    /// one after the other and IOs are issued and waited upon
@@ -233,54 +191,6 @@ pub enum GetVectoredConcurrentIo {
    SidecarTask,
 }

-#[derive(Debug, Copy, Clone, PartialEq, Eq, serde::Serialize, serde::Deserialize)]
-pub struct Ratio {
-    pub numerator: usize,
-    pub denominator: usize,
-}
-
-#[derive(Debug, Clone, PartialEq, Eq, serde::Serialize, serde::Deserialize)]
-pub struct OtelExporterConfig {
-    pub endpoint: String,
-    pub protocol: OtelExporterProtocol,
-    #[serde(with = "humantime_serde")]
-    pub timeout: Duration,
-}
-
-#[derive(Debug, Copy, Clone, PartialEq, Eq, serde::Serialize, serde::Deserialize)]
-#[serde(rename_all = "kebab-case")]
-pub enum OtelExporterProtocol {
-    Grpc,
-    HttpBinary,
-    HttpJson,
-}
-
-#[derive(Debug, Clone, PartialEq, Eq, serde::Serialize, serde::Deserialize)]
-pub struct Tracing {
-    pub sampling_ratio: Ratio,
-    pub export_config: OtelExporterConfig,
-}
-
-impl From<&OtelExporterConfig> for tracing_utils::ExportConfig {
-    fn from(val: &OtelExporterConfig) -> Self {
-        tracing_utils::ExportConfig {
-            endpoint: Some(val.endpoint.clone()),
-            protocol: val.protocol.into(),
-            timeout: val.timeout,
-        }
-    }
-}
-
-impl From<OtelExporterProtocol> for tracing_utils::Protocol {
-    fn from(val: OtelExporterProtocol) -> Self {
-        match val {
-            OtelExporterProtocol::Grpc => tracing_utils::Protocol::Grpc,
-            OtelExporterProtocol::HttpJson => tracing_utils::Protocol::HttpJson,
-            OtelExporterProtocol::HttpBinary => tracing_utils::Protocol::HttpBinary,
-        }
-    }
-}
-
 pub mod statvfs {
    pub mod mock {
        #[derive(Debug, Clone, PartialEq, Eq, serde::Serialize, serde::Deserialize)]
@@ -335,7 +245,7 @@ pub struct MaxVectoredReadBytes(pub NonZeroUsize);

 /// Tenant-level configuration values, used for various purposes.
 #[derive(Debug, Clone, PartialEq, Eq, serde::Serialize, serde::Deserialize)]
-#[serde(default)]
+#[serde(deny_unknown_fields, default)]
 pub struct TenantConfigToml {
    // Flush out an inmemory layer, if it's holding WAL older than this
    // This puts a backstop on how much WAL needs to be re-digested if the
@@ -457,9 +367,6 @@ pub struct TenantConfigToml {
    /// The ratio that triggers the auto gc-compaction. If (the total size of layers between L2 LSN and gc-horizon) / (size below the L2 LSN)
    /// is above this ratio, gc-compaction will be triggered.
    pub gc_compaction_ratio_percent: u64,
-    /// Tenant level performance sampling ratio override. Controls the ratio of get page requests
-    /// that will get perf sampling for the tenant.
-    pub sampling_ratio: Option<Ratio>,
 }

 pub mod defaults {
@@ -630,7 +537,6 @@ impl Default for ConfigToml {
            validate_wal_contiguity: None,
            load_previous_heatmap: None,
            generate_unarchival_heatmap: None,
-            tracing: None,
        }
    }
 }
@@ -746,7 +652,6 @@ impl Default for TenantConfigToml {
            gc_compaction_enabled: DEFAULT_GC_COMPACTION_ENABLED,
            gc_compaction_initial_threshold_kb: DEFAULT_GC_COMPACTION_INITIAL_THRESHOLD_KB,
            gc_compaction_ratio_percent: DEFAULT_GC_COMPACTION_RATIO_PERCENT,
-            sampling_ratio: None,
        }
    }
 }
--- a/libs/pageserver_api/src/keyspace.rs
+++ b/libs/pageserver_api/src/keyspace.rs
@@ -613,7 +613,8 @@ mod tests {
    use rand::{RngCore, SeedableRng};

    use super::*;
-    use crate::shard::{DEFAULT_STRIPE_SIZE, ShardCount, ShardNumber, ShardStripeSize};
+    use crate::models::ShardParameters;
+    use crate::shard::{ShardCount, ShardNumber};

    // Helper function to create a key range.
    //
@@ -963,8 +964,12 @@ mod tests {
    }
    #[test]
    fn sharded_range_relation_gap() {
-        let shard_identity =
-            ShardIdentity::new(ShardNumber(0), ShardCount::new(4), DEFAULT_STRIPE_SIZE).unwrap();
+        let shard_identity = ShardIdentity::new(
+            ShardNumber(0),
+            ShardCount::new(4),
+            ShardParameters::DEFAULT_STRIPE_SIZE,
+        )
+        .unwrap();

        let range = ShardedRange::new(
            Range {
@@ -980,8 +985,12 @@ mod tests {

    #[test]
    fn shard_identity_keyspaces_single_key() {
-        let shard_identity =
-            ShardIdentity::new(ShardNumber(1), ShardCount::new(4), DEFAULT_STRIPE_SIZE).unwrap();
+        let shard_identity = ShardIdentity::new(
+            ShardNumber(1),
+            ShardCount::new(4),
+            ShardParameters::DEFAULT_STRIPE_SIZE,
+        )
+        .unwrap();

        let range = ShardedRange::new(
            Range {
@@ -1025,8 +1034,12 @@ mod tests {

    #[test]
    fn shard_identity_keyspaces_forkno_gap() {
-        let shard_identity =
-            ShardIdentity::new(ShardNumber(1), ShardCount::new(4), DEFAULT_STRIPE_SIZE).unwrap();
+        let shard_identity = ShardIdentity::new(
+            ShardNumber(1),
+            ShardCount::new(4),
+            ShardParameters::DEFAULT_STRIPE_SIZE,
+        )
+        .unwrap();

        let range = ShardedRange::new(
            Range {
@@ -1048,7 +1061,7 @@ mod tests {
            let shard_identity = ShardIdentity::new(
                ShardNumber(shard_number),
                ShardCount::new(4),
-                DEFAULT_STRIPE_SIZE,
+                ShardParameters::DEFAULT_STRIPE_SIZE,
            )
            .unwrap();

@@ -1131,44 +1144,37 @@ mod tests {
    /// for a single tenant.
    #[test]
    fn sharded_range_fragment_simple() {
-        const SHARD_COUNT: u8 = 4;
-        const STRIPE_SIZE: u32 = DEFAULT_STRIPE_SIZE.0;
-
        let shard_identity = ShardIdentity::new(
            ShardNumber(0),
-            ShardCount::new(SHARD_COUNT),
-            ShardStripeSize(STRIPE_SIZE),
+            ShardCount::new(4),
+            ShardParameters::DEFAULT_STRIPE_SIZE,
        )
        .unwrap();

        // A range which we happen to know covers exactly one stripe which belongs to this shard
        let input_start = Key::from_hex("000000067f00000001000000ae0000000000").unwrap();
-        let mut input_end = input_start;
-        input_end.field6 += STRIPE_SIZE; // field6 is block number
+        let input_end = Key::from_hex("000000067f00000001000000ae0000008000").unwrap();

        // Ask for stripe_size blocks, we get the whole stripe
        assert_eq!(
-            do_fragment(input_start, input_end, &shard_identity, STRIPE_SIZE),
-            (STRIPE_SIZE, vec![(STRIPE_SIZE, input_start..input_end)])
+            do_fragment(input_start, input_end, &shard_identity, 32768),
+            (32768, vec![(32768, input_start..input_end)])
        );

        // Ask for more, we still get the whole stripe
        assert_eq!(
-            do_fragment(input_start, input_end, &shard_identity, 10 * STRIPE_SIZE),
-            (STRIPE_SIZE, vec![(STRIPE_SIZE, input_start..input_end)])
+            do_fragment(input_start, input_end, &shard_identity, 10000000),
+            (32768, vec![(32768, input_start..input_end)])
        );

        // Ask for target_nblocks of half the stripe size, we get two halves
        assert_eq!(
-            do_fragment(input_start, input_end, &shard_identity, STRIPE_SIZE / 2),
+            do_fragment(input_start, input_end, &shard_identity, 16384),
            (
-                STRIPE_SIZE,
+                32768,
                vec![
-                    (
-                        STRIPE_SIZE / 2,
-                        input_start..input_start.add(STRIPE_SIZE / 2)
-                    ),
-                    (STRIPE_SIZE / 2, input_start.add(STRIPE_SIZE / 2)..input_end)
+                    (16384, input_start..input_start.add(16384)),
+                    (16384, input_start.add(16384)..input_end)
                ]
            )
        );
@@ -1176,53 +1182,40 @@ mod tests {

    #[test]
    fn sharded_range_fragment_multi_stripe() {
-        const SHARD_COUNT: u8 = 4;
-        const STRIPE_SIZE: u32 = DEFAULT_STRIPE_SIZE.0;
-        const RANGE_SIZE: u32 = SHARD_COUNT as u32 * STRIPE_SIZE;
-
        let shard_identity = ShardIdentity::new(
            ShardNumber(0),
-            ShardCount::new(SHARD_COUNT),
-            ShardStripeSize(STRIPE_SIZE),
+            ShardCount::new(4),
+            ShardParameters::DEFAULT_STRIPE_SIZE,
        )
        .unwrap();

        // A range which covers multiple stripes, exactly one of which belongs to the current shard.
        let input_start = Key::from_hex("000000067f00000001000000ae0000000000").unwrap();
-        let mut input_end = input_start;
-        input_end.field6 += RANGE_SIZE; // field6 is block number
-
+        let input_end = Key::from_hex("000000067f00000001000000ae0000020000").unwrap();
        // Ask for all the blocks, get a fragment that covers the whole range but reports
        // its size to be just the blocks belonging to our shard.
        assert_eq!(
-            do_fragment(input_start, input_end, &shard_identity, RANGE_SIZE),
-            (STRIPE_SIZE, vec![(STRIPE_SIZE, input_start..input_end)])
+            do_fragment(input_start, input_end, &shard_identity, 131072),
+            (32768, vec![(32768, input_start..input_end)])
        );

-        // Ask for a sub-stripe quantity that results in 3 fragments.
-        let limit = STRIPE_SIZE / 3 + 1;
+        // Ask for a sub-stripe quantity
        assert_eq!(
-            do_fragment(input_start, input_end, &shard_identity, limit),
+            do_fragment(input_start, input_end, &shard_identity, 16000),
            (
-                STRIPE_SIZE,
+                32768,
                vec![
-                    (limit, input_start..input_start.add(limit)),
-                    (limit, input_start.add(limit)..input_start.add(2 * limit)),
-                    (
-                        STRIPE_SIZE - 2 * limit,
-                        input_start.add(2 * limit)..input_end
-                    ),
+                    (16000, input_start..input_start.add(16000)),
+                    (16000, input_start.add(16000)..input_start.add(32000)),
+                    (768, input_start.add(32000)..input_end),
                ]
            )
        );

        // Try on a range that starts slightly after our owned stripe
        assert_eq!(
-            do_fragment(input_start.add(1), input_end, &shard_identity, RANGE_SIZE),
-            (
-                STRIPE_SIZE - 1,
-                vec![(STRIPE_SIZE - 1, input_start.add(1)..input_end)]
-            )
+            do_fragment(input_start.add(1), input_end, &shard_identity, 131072),
+            (32767, vec![(32767, input_start.add(1)..input_end)])
        );
    }

@@ -1230,40 +1223,32 @@ mod tests {
    /// a previous relation.
    #[test]
    fn sharded_range_fragment_starting_from_logical_size() {
-        const SHARD_COUNT: u8 = 4;
-        const STRIPE_SIZE: u32 = DEFAULT_STRIPE_SIZE.0;
-        const RANGE_SIZE: u32 = SHARD_COUNT as u32 * STRIPE_SIZE;
-
        let input_start = Key::from_hex("000000067f00000001000000ae00ffffffff").unwrap();
-        let mut input_end = Key::from_hex("000000067f00000001000000ae0100000000").unwrap();
-        input_end.field6 += RANGE_SIZE; // field6 is block number
+        let input_end = Key::from_hex("000000067f00000001000000ae0100008000").unwrap();

        // Shard 0 owns the first stripe in the relation, and the preceding logical size is shard local too
        let shard_identity = ShardIdentity::new(
            ShardNumber(0),
-            ShardCount::new(SHARD_COUNT),
-            ShardStripeSize(STRIPE_SIZE),
+            ShardCount::new(4),
+            ShardParameters::DEFAULT_STRIPE_SIZE,
        )
        .unwrap();
        assert_eq!(
-            do_fragment(input_start, input_end, &shard_identity, 2 * STRIPE_SIZE),
-            (
-                STRIPE_SIZE + 1,
-                vec![(STRIPE_SIZE + 1, input_start..input_end)]
-            )
+            do_fragment(input_start, input_end, &shard_identity, 0x10000),
+            (0x8001, vec![(0x8001, input_start..input_end)])
        );

        // Shard 1 does not own the first stripe in the relation, but it does own the logical size (all shards
        // store all logical sizes)
        let shard_identity = ShardIdentity::new(
            ShardNumber(1),
-            ShardCount::new(SHARD_COUNT),
-            ShardStripeSize(STRIPE_SIZE),
+            ShardCount::new(4),
+            ShardParameters::DEFAULT_STRIPE_SIZE,
        )
        .unwrap();
        assert_eq!(
-            do_fragment(input_start, input_end, &shard_identity, 2 * STRIPE_SIZE),
-            (1, vec![(1, input_start..input_end)])
+            do_fragment(input_start, input_end, &shard_identity, 0x10000),
+            (0x1, vec![(0x1, input_start..input_end)])
        );
    }

@@ -1299,8 +1284,12 @@ mod tests {
        );

        // Same, but using a sharded identity
-        let shard_identity =
-            ShardIdentity::new(ShardNumber(0), ShardCount::new(4), DEFAULT_STRIPE_SIZE).unwrap();
+        let shard_identity = ShardIdentity::new(
+            ShardNumber(0),
+            ShardCount::new(4),
+            ShardParameters::DEFAULT_STRIPE_SIZE,
+        )
+        .unwrap();
        assert_eq!(
            do_fragment(input_start, input_end, &shard_identity, 0x8000),
            (u32::MAX, vec![(u32::MAX, input_start..input_end),])
@@ -1342,7 +1331,7 @@ mod tests {
                ShardIdentity::new(
                    ShardNumber((prng.next_u32() % shard_count) as u8),
                    ShardCount::new(shard_count as u8),
-                    DEFAULT_STRIPE_SIZE,
+                    ShardParameters::DEFAULT_STRIPE_SIZE,
                )
                .unwrap()
            };
--- a/libs/pageserver_api/src/models.rs
+++ b/libs/pageserver_api/src/models.rs
@@ -23,10 +23,9 @@ use utils::lsn::Lsn;
 use utils::postgres_client::PostgresClientProtocol;
 use utils::{completion, serde_system_time};

-use crate::config::Ratio;
 use crate::key::{CompactKey, Key};
 use crate::reltag::RelTag;
-use crate::shard::{DEFAULT_STRIPE_SIZE, ShardCount, ShardStripeSize, TenantShardId};
+use crate::shard::{ShardCount, ShardStripeSize, TenantShardId};

 /// The state of a tenant in this pageserver.
 ///
@@ -80,22 +79,10 @@ pub enum TenantState {
    ///
    /// Transitions out of this state are possible through `set_broken()`.
    Stopping {
-        /// The barrier can be used to wait for shutdown to complete. The first caller to set
-        /// Some(Barrier) is responsible for driving shutdown to completion. Subsequent callers
-        /// will wait for the first caller's existing barrier.
-        ///
-        /// None is set when an attach is cancelled, to signal to shutdown that the attach has in
-        /// fact cancelled:
-        ///
-        /// 1. `shutdown` sees `TenantState::Attaching`, and cancels the tenant.
-        /// 2. `attach` sets `TenantState::Stopping(None)` and exits.
-        /// 3. `set_stopping` waits for `TenantState::Stopping(None)` and sets
-        ///    `TenantState::Stopping(Some)` to claim the barrier as the shutdown owner.
-        //
        // Because of https://github.com/serde-rs/serde/issues/2105 this has to be a named field,
        // otherwise it will not be skipped during deserialization
        #[serde(skip)]
-        progress: Option<completion::Barrier>,
+        progress: completion::Barrier,
    },
    /// The tenant is recognized by the pageserver, but can no longer be used for
    /// any operations.
@@ -438,6 +425,8 @@ pub struct ShardParameters {
 }

 impl ShardParameters {
+    pub const DEFAULT_STRIPE_SIZE: ShardStripeSize = ShardStripeSize(256 * 1024 / 8);
+
    pub fn is_unsharded(&self) -> bool {
        self.count.is_unsharded()
    }
@@ -447,7 +436,7 @@ impl Default for ShardParameters {
    fn default() -> Self {
        Self {
            count: ShardCount::new(0),
-            stripe_size: DEFAULT_STRIPE_SIZE,
+            stripe_size: Self::DEFAULT_STRIPE_SIZE,
        }
    }
 }
@@ -579,8 +568,6 @@ pub struct TenantConfigPatch {
    pub gc_compaction_initial_threshold_kb: FieldPatch<u64>,
    #[serde(skip_serializing_if = "FieldPatch::is_noop")]
    pub gc_compaction_ratio_percent: FieldPatch<u64>,
-    #[serde(skip_serializing_if = "FieldPatch::is_noop")]
-    pub sampling_ratio: FieldPatch<Option<Ratio>>,
 }

 /// Like [`crate::config::TenantConfigToml`], but preserves the information
@@ -701,9 +688,6 @@ pub struct TenantConfig {

    #[serde(skip_serializing_if = "Option::is_none")]
    pub gc_compaction_ratio_percent: Option<u64>,
-
-    #[serde(skip_serializing_if = "Option::is_none")]
-    pub sampling_ratio: Option<Option<Ratio>>,
 }

 impl TenantConfig {
@@ -746,7 +730,6 @@ impl TenantConfig {
            mut gc_compaction_enabled,
            mut gc_compaction_initial_threshold_kb,
            mut gc_compaction_ratio_percent,
-            mut sampling_ratio,
        } = self;

        patch.checkpoint_distance.apply(&mut checkpoint_distance);
@@ -841,7 +824,6 @@ impl TenantConfig {
        patch
            .gc_compaction_ratio_percent
            .apply(&mut gc_compaction_ratio_percent);
-        patch.sampling_ratio.apply(&mut sampling_ratio);

        Ok(Self {
            checkpoint_distance,
@@ -878,7 +860,6 @@ impl TenantConfig {
            gc_compaction_enabled,
            gc_compaction_initial_threshold_kb,
            gc_compaction_ratio_percent,
-            sampling_ratio,
        })
    }

@@ -980,7 +961,6 @@ impl TenantConfig {
            gc_compaction_ratio_percent: self
                .gc_compaction_ratio_percent
                .unwrap_or(global_conf.gc_compaction_ratio_percent),
-            sampling_ratio: self.sampling_ratio.unwrap_or(global_conf.sampling_ratio),
        }
    }
 }
@@ -1114,7 +1094,7 @@ pub struct CompactionAlgorithmSettings {
 }

 #[derive(Debug, PartialEq, Eq, Clone, Deserialize, Serialize)]
-#[serde(tag = "mode", rename_all = "kebab-case")]
+#[serde(tag = "mode", rename_all = "kebab-case", deny_unknown_fields)]
 pub enum L0FlushConfig {
    #[serde(rename_all = "snake_case")]
    Direct { max_concurrency: NonZeroUsize },
@@ -1438,6 +1418,11 @@ pub struct TimelineInfo {
    pub last_record_lsn: Lsn,
    pub prev_record_lsn: Option<Lsn>,

+    /// Legacy field, retained for one version to enable old storage controller to
+    /// decode (it was a mandatory field).
+    #[serde(default, rename = "latest_gc_cutoff_lsn")]
+    pub _unused: Lsn,
+
    /// The LSN up to which GC has advanced: older data may still exist but it is not available for clients.
    /// This LSN is not suitable for deciding where to create branches etc: use [`TimelineInfo::min_readable_lsn`] instead,
    /// as it is easier to reason about.
@@ -1678,7 +1663,6 @@ pub struct SecondaryProgress {
 pub struct TenantScanRemoteStorageShard {
    pub tenant_shard_id: TenantShardId,
    pub generation: Option<u32>,
-    pub stripe_size: Option<ShardStripeSize>,
 }

 #[derive(Serialize, Deserialize, Debug, Default)]
@@ -2730,15 +2714,10 @@ mod tests {
                "Activating",
            ),
            (line!(), TenantState::Active, "Active"),
-            (
-                line!(),
-                TenantState::Stopping { progress: None },
-                "Stopping",
-            ),
            (
                line!(),
                TenantState::Stopping {
-                    progress: Some(completion::Barrier::default()),
+                    progress: utils::completion::Barrier::default(),
                },
                "Stopping",
            ),
--- a/libs/pageserver_api/src/record.rs
+++ b/libs/pageserver_api/src/record.rs
@@ -58,8 +58,6 @@ pub enum NeonWalRecord {
        /// to true. This record does not need the history WALs to reconstruct. See [`NeonWalRecord::will_init`] and
        /// its references in `timeline.rs`.
        will_init: bool,
-        /// Only append the record if the current image is the same as the one specified in this field.
-        only_if: Option<String>,
    },
 }

@@ -83,17 +81,6 @@ impl NeonWalRecord {
            append: s.as_ref().to_string(),
            clear: false,
            will_init: false,
-            only_if: None,
-        }
-    }
-
-    #[cfg(feature = "testing")]
-    pub fn wal_append_conditional(s: impl AsRef<str>, only_if: impl AsRef<str>) -> Self {
-        Self::Test {
-            append: s.as_ref().to_string(),
-            clear: false,
-            will_init: false,
-            only_if: Some(only_if.as_ref().to_string()),
        }
    }

@@ -103,7 +90,6 @@ impl NeonWalRecord {
            append: s.as_ref().to_string(),
            clear: true,
            will_init: false,
-            only_if: None,
        }
    }

@@ -113,7 +99,6 @@ impl NeonWalRecord {
            append: s.as_ref().to_string(),
            clear: true,
            will_init: true,
-            only_if: None,
        }
    }
 }
--- a/libs/pageserver_api/src/shard.rs
+++ b/libs/pageserver_api/src/shard.rs
@@ -78,12 +78,6 @@ impl Default for ShardStripeSize {
    }
 }

-impl std::fmt::Display for ShardStripeSize {
-    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
-        self.0.fmt(f)
-    }
-}
-
 /// Layout version: for future upgrades where we might change how the key->shard mapping works
 #[derive(Clone, Copy, Serialize, Deserialize, Eq, PartialEq, Hash, Debug)]
 pub struct ShardLayout(u8);
@@ -92,11 +86,8 @@ const LAYOUT_V1: ShardLayout = ShardLayout(1);
 /// ShardIdentity uses a magic layout value to indicate if it is unusable
 const LAYOUT_BROKEN: ShardLayout = ShardLayout(255);

-/// The default stripe size in pages. 16 MiB divided by 8 kiB page size.
-///
-/// A lower stripe size distributes ingest load better across shards, but reduces IO amortization.
-/// 16 MiB appears to be a reasonable balance: <https://github.com/neondatabase/neon/pull/10510>.
-pub const DEFAULT_STRIPE_SIZE: ShardStripeSize = ShardStripeSize(16 * 1024 / 8);
+/// Default stripe size in pages: 256MiB divided by 8kiB page size.
+const DEFAULT_STRIPE_SIZE: ShardStripeSize = ShardStripeSize(256 * 1024 / 8);

 #[derive(thiserror::Error, Debug, PartialEq, Eq)]
 pub enum ShardConfigError {
@@ -546,7 +537,7 @@ mod tests {
            field6: 0x7d06,
        };

-        let shard = key_to_shard_number(ShardCount(10), ShardStripeSize(32768), &key);
+        let shard = key_to_shard_number(ShardCount(10), DEFAULT_STRIPE_SIZE, &key);
        assert_eq!(shard, ShardNumber(8));
    }

--- a/libs/postgres_backend/src/lib.rs
+++ b/libs/postgres_backend/src/lib.rs
@@ -5,6 +5,7 @@
 #![deny(unsafe_code)]
 #![deny(clippy::undocumented_unsafe_blocks)]
 use std::future::Future;
+use std::io::ErrorKind;
 use std::net::SocketAddr;
 use std::os::fd::{AsRawFd, RawFd};
 use std::pin::Pin;
@@ -226,7 +227,7 @@ impl<IO: AsyncRead + AsyncWrite + Unpin> MaybeWriteOnly<IO> {
        match self {
            MaybeWriteOnly::Full(framed) => framed.read_startup_message().await,
            MaybeWriteOnly::WriteOnly(_) => {
-                Err(io::Error::other("reading from write only half").into())
+                Err(io::Error::new(ErrorKind::Other, "reading from write only half").into())
            }
            MaybeWriteOnly::Broken => panic!("IO on invalid MaybeWriteOnly"),
        }
@@ -236,7 +237,7 @@ impl<IO: AsyncRead + AsyncWrite + Unpin> MaybeWriteOnly<IO> {
        match self {
            MaybeWriteOnly::Full(framed) => framed.read_message().await,
            MaybeWriteOnly::WriteOnly(_) => {
-                Err(io::Error::other("reading from write only half").into())
+                Err(io::Error::new(ErrorKind::Other, "reading from write only half").into())
            }
            MaybeWriteOnly::Broken => panic!("IO on invalid MaybeWriteOnly"),
        }
@@ -974,7 +975,7 @@ impl<IO: AsyncRead + AsyncWrite + Unpin> AsyncWrite for CopyDataWriter<'_, IO> {
            .write_message_noflush(&BeMessage::CopyData(buf))
            // write_message only writes to the buffer, so it can fail iff the
            // message is invaid, but CopyData can't be invalid.
-            .map_err(|_| io::Error::other("failed to serialize CopyData"))?;
+            .map_err(|_| io::Error::new(ErrorKind::Other, "failed to serialize CopyData"))?;

        Poll::Ready(Ok(buf.len()))
    }
--- a/libs/postgres_backend/tests/simple_select.rs
+++ b/libs/postgres_backend/tests/simple_select.rs
@@ -85,8 +85,8 @@ static KEY: Lazy<rustls::pki_types::PrivateKeyDer<'static>> = Lazy::new(|| {

 static CERT: Lazy<rustls::pki_types::CertificateDer<'static>> = Lazy::new(|| {
    let mut cursor = Cursor::new(include_bytes!("cert.pem"));
-
-    rustls_pemfile::certs(&mut cursor).next().unwrap().unwrap()
+    let cert = rustls_pemfile::certs(&mut cursor).next().unwrap().unwrap();
+    cert
 });

 // test that basic select with ssl works
--- a/libs/pq_proto/src/framed.rs
+++ b/libs/pq_proto/src/framed.rs
@@ -35,7 +35,7 @@ impl ConnectionError {
    pub fn into_io_error(self) -> io::Error {
        match self {
            ConnectionError::Io(io) => io,
-            ConnectionError::Protocol(pe) => io::Error::other(pe.to_string()),
+            ConnectionError::Protocol(pe) => io::Error::new(io::ErrorKind::Other, pe.to_string()),
        }
    }
 }
--- a/libs/pq_proto/src/lib.rs
+++ b/libs/pq_proto/src/lib.rs
@@ -257,7 +257,7 @@ pub enum ProtocolError {
 impl ProtocolError {
    /// Proxy stream.rs uses only io::Error; provide it.
    pub fn into_io_error(self) -> io::Error {
-        io::Error::other(self.to_string())
+        io::Error::new(io::ErrorKind::Other, self.to_string())
    }
 }

--- a/libs/proxy/postgres-protocol2/src/authentication/sasl.rs
+++ b/libs/proxy/postgres-protocol2/src/authentication/sasl.rs
@@ -212,7 +212,7 @@ impl ScramSha256 {
                    password,
                    channel_binding,
                } => (nonce, password, channel_binding),
-                _ => return Err(io::Error::other("invalid SCRAM state")),
+                _ => return Err(io::Error::new(io::ErrorKind::Other, "invalid SCRAM state")),
            };

        let message =
@@ -291,7 +291,7 @@ impl ScramSha256 {
                server_key,
                auth_message,
            } => (server_key, auth_message),
-            _ => return Err(io::Error::other("invalid SCRAM state")),
+            _ => return Err(io::Error::new(io::ErrorKind::Other, "invalid SCRAM state")),
        };

        let message =
@@ -301,7 +301,10 @@ impl ScramSha256 {

        let verifier = match parsed {
            ServerFinalMessage::Error(e) => {
-                return Err(io::Error::other(format!("SCRAM error: {}", e)));
+                return Err(io::Error::new(
+                    io::ErrorKind::Other,
+                    format!("SCRAM error: {}", e),
+                ));
            }
            ServerFinalMessage::Verifier(verifier) => verifier,
        };
--- a/libs/remote_storage/Cargo.toml
+++ b/libs/remote_storage/Cargo.toml
@@ -28,7 +28,7 @@ toml_edit.workspace = true
 tracing.workspace = true
 scopeguard.workspace = true
 metrics.workspace = true
-utils = { path = "../utils", default-features = false }
+utils.workspace = true
 pin-project-lite.workspace = true

 azure_core.workspace = true
--- a/libs/remote_storage/src/azure_blob.rs
+++ b/libs/remote_storage/src/azure_blob.rs
@@ -801,7 +801,8 @@ where
            // that support needs to be hacked in.
            //
            // including {self:?} into the message would be useful, but unsure how to unproject.
-            _ => std::task::Poll::Ready(Err(std::io::Error::other(
+            _ => std::task::Poll::Ready(Err(std::io::Error::new(
+                std::io::ErrorKind::Other,
                "cloned or initial values cannot be read",
            ))),
        }
@@ -854,7 +855,7 @@ where
        };
        Err(azure_core::error::Error::new(
            azure_core::error::ErrorKind::Io,
-            std::io::Error::other(msg),
+            std::io::Error::new(std::io::ErrorKind::Other, msg),
        ))
    }

--- a/libs/remote_storage/tests/test_real_s3.rs
+++ b/libs/remote_storage/tests/test_real_s3.rs
@@ -558,7 +558,7 @@ async fn upload_large_enough_file(
 ) -> usize {
    let header = bytes::Bytes::from_static("remote blob data content".as_bytes());
    let body = bytes::Bytes::from(vec![0u8; 1024]);
-    let contents = std::iter::once(header).chain(std::iter::repeat_n(body, 128));
+    let contents = std::iter::once(header).chain(std::iter::repeat(body).take(128));

    let len = contents.clone().fold(0, |acc, next| acc + next.len());

--- a/libs/safekeeper_api/src/models.rs
+++ b/libs/safekeeper_api/src/models.rs
@@ -71,7 +71,6 @@ pub struct PeerInfo {
    pub ts: Instant,
    pub pg_connstr: String,
    pub http_connstr: String,
-    pub https_connstr: Option<String>,
 }

 pub type FullTransactionId = u64;
@@ -228,8 +227,6 @@ pub struct TimelineDeleteResult {
    pub dir_existed: bool,
 }

-pub type TenantDeleteResult = std::collections::HashMap<String, TimelineDeleteResult>;
-
 fn lsn_invalid() -> Lsn {
    Lsn::INVALID
 }
@@ -262,8 +259,6 @@ pub struct SkTimelineInfo {
    pub safekeeper_connstr: Option<String>,
    #[serde(default)]
    pub http_connstr: Option<String>,
-    #[serde(default)]
-    pub https_connstr: Option<String>,
    // Minimum of all active RO replicas flush LSN
    #[serde(default = "lsn_invalid")]
    pub standby_horizon: Lsn,
--- a/libs/tracing-utils/Cargo.toml
+++ b/libs/tracing-utils/Cargo.toml
@@ -14,7 +14,6 @@ tokio = { workspace = true, features = ["rt", "rt-multi-thread"] }
 tracing.workspace = true
 tracing-opentelemetry.workspace = true
 tracing-subscriber.workspace = true
-pin-project-lite.workspace = true

 [dev-dependencies]
 tracing-subscriber.workspace = true    # For examples in docs
--- a/libs/tracing-utils/src/lib.rs
+++ b/libs/tracing-utils/src/lib.rs
@@ -31,10 +31,10 @@
 //!         .init();
 //! }
 //! ```
+#![deny(unsafe_code)]
 #![deny(clippy::undocumented_unsafe_blocks)]

 pub mod http;
-pub mod perf_span;

 use opentelemetry::KeyValue;
 use opentelemetry::trace::TracerProvider;
--- a/libs/tracing-utils/src/perf_span.rs
+++ b/libs/tracing-utils/src/perf_span.rs
@@ -1,144 +0,0 @@
-//! Crutch module to work around tracing infrastructure deficiencies
-//!
-//! We wish to collect granular request spans without impacting performance
-//! by much. Ideally, we should have zero overhead for a sampling rate of 0.
-//!
-//! The approach taken by the pageserver crate is to use a completely different
-//! span hierarchy for the performance spans. Spans are explicitly stored in
-//! the request context and use a different [`tracing::Subscriber`] in order
-//! to avoid expensive filtering.
-//!
-//! [`tracing::Span`] instances record their [`tracing::Dispatch`] and, implcitly,
-//! their [`tracing::Subscriber`] at creation time. However, upon exiting the span,
-//! the global default [`tracing::Dispatch`] is used. This is problematic if one
-//! wishes to juggle different subscribers.
-//!
-//! In order to work around this, this module provides a [`PerfSpan`] type which
-//! wraps a [`Span`] and sets the default subscriber when exiting the span. This
-//! achieves the correct routing.
-//!
-//! There's also a modified version of [`tracing::Instrument`] which works with
-//! [`PerfSpan`].
-
-use core::{
-    future::Future,
-    marker::Sized,
-    mem::ManuallyDrop,
-    pin::Pin,
-    task::{Context, Poll},
-};
-use pin_project_lite::pin_project;
-use tracing::{Dispatch, span::Span};
-
-#[derive(Debug, Clone)]
-pub struct PerfSpan {
-    inner: ManuallyDrop<Span>,
-    dispatch: Dispatch,
-}
-
-#[must_use = "once a span has been entered, it should be exited"]
-pub struct PerfSpanEntered<'a> {
-    span: &'a PerfSpan,
-}
-
-impl PerfSpan {
-    pub fn new(span: Span, dispatch: Dispatch) -> Self {
-        Self {
-            inner: ManuallyDrop::new(span),
-            dispatch,
-        }
-    }
-
-    pub fn enter(&self) -> PerfSpanEntered {
-        if let Some(ref id) = self.inner.id() {
-            self.dispatch.enter(id);
-        }
-
-        PerfSpanEntered { span: self }
-    }
-
-    pub fn inner(&self) -> &Span {
-        &self.inner
-    }
-}
-
-impl Drop for PerfSpan {
-    fn drop(&mut self) {
-        // Bring the desired dispatch into scope before explicitly calling
-        // the span destructor. This routes the span exit to the correct
-        // [`tracing::Subscriber`].
-        let _dispatch_guard = tracing::dispatcher::set_default(&self.dispatch);
-        // SAFETY: ManuallyDrop in Drop implementation
-        unsafe { ManuallyDrop::drop(&mut self.inner) }
-    }
-}
-
-impl Drop for PerfSpanEntered<'_> {
-    fn drop(&mut self) {
-        assert!(self.span.inner.id().is_some());
-
-        let _dispatch_guard = tracing::dispatcher::set_default(&self.span.dispatch);
-        self.span.dispatch.exit(&self.span.inner.id().unwrap());
-    }
-}
-
-pub trait PerfInstrument: Sized {
-    fn instrument(self, span: PerfSpan) -> PerfInstrumented<Self> {
-        PerfInstrumented {
-            inner: ManuallyDrop::new(self),
-            span,
-        }
-    }
-}
-
-pin_project! {
-    #[project = PerfInstrumentedProj]
-    #[derive(Debug, Clone)]
-    #[must_use = "futures do nothing unless you `.await` or poll them"]
-    pub struct PerfInstrumented<T> {
-        // `ManuallyDrop` is used here to to enter instrument `Drop` by entering
-        // `Span` and executing `ManuallyDrop::drop`.
-        #[pin]
-        inner: ManuallyDrop<T>,
-        span: PerfSpan,
-    }
-
-    impl<T> PinnedDrop for PerfInstrumented<T> {
-        fn drop(this: Pin<&mut Self>) {
-            let this = this.project();
-            let _enter = this.span.enter();
-            // SAFETY: 1. `Pin::get_unchecked_mut()` is safe, because this isn't
-            //             different from wrapping `T` in `Option` and calling
-            //             `Pin::set(&mut this.inner, None)`, except avoiding
-            //             additional memory overhead.
-            //         2. `ManuallyDrop::drop()` is safe, because
-            //            `PinnedDrop::drop()` is guaranteed to be called only
-            //            once.
-            unsafe { ManuallyDrop::drop(this.inner.get_unchecked_mut()) }
-        }
-    }
-}
-
-impl<'a, T> PerfInstrumentedProj<'a, T> {
-    /// Get a mutable reference to the [`Span`] a pinned mutable reference to
-    /// the wrapped type.
-    fn span_and_inner_pin_mut(self) -> (&'a mut PerfSpan, Pin<&'a mut T>) {
-        // SAFETY: As long as `ManuallyDrop<T>` does not move, `T` won't move
-        //         and `inner` is valid, because `ManuallyDrop::drop` is called
-        //         only inside `Drop` of the `Instrumented`.
-        let inner = unsafe { self.inner.map_unchecked_mut(|v| &mut **v) };
-        (self.span, inner)
-    }
-}
-
-impl<T: Future> Future for PerfInstrumented<T> {
-    type Output = T::Output;
-
-    fn poll(self: Pin<&mut Self>, cx: &mut Context<'_>) -> Poll<Self::Output> {
-        let (span, inner) = self.project().span_and_inner_pin_mut();
-        let _enter = span.enter();
-        inner.poll(cx)
-    }
-}
-
-impl<T: Sized> PerfInstrument for T {}
--- a/libs/utils/Cargo.toml
+++ b/libs/utils/Cargo.toml
@@ -5,8 +5,7 @@ edition.workspace = true
 license.workspace = true

 [features]
-default = ["rename_noreplace"]
-rename_noreplace = []
+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"]
@@ -36,7 +35,7 @@ serde_with.workspace = true
 serde_json.workspace = true
 signal-hook.workspace = true
 thiserror.workspace = true
-tokio = { workspace = true, features = ["signal"] }
+tokio.workspace = true
 tokio-tar.workspace = true
 tokio-util.workspace = true
 toml_edit = { workspace = true, features = ["serde"] }
--- a/libs/utils/src/auth.rs
+++ b/libs/utils/src/auth.rs
@@ -173,7 +173,7 @@ impl std::fmt::Debug for JwtAuth {
 }

 // this function is used only for testing purposes in CLI e g generate tokens during init
-pub fn encode_from_key_file<S: Serialize>(claims: &S, key_data: &[u8]) -> Result<String> {
+pub fn encode_from_key_file(claims: &Claims, key_data: &[u8]) -> Result<String> {
    let key = EncodingKey::from_ed_pem(key_data)?;
    Ok(encode(&Header::new(STORAGE_TOKEN_ALGORITHM), claims, &key)?)
 }
--- a/libs/utils/src/crashsafe.rs
+++ b/libs/utils/src/crashsafe.rs
@@ -81,9 +81,12 @@ pub fn path_with_suffix_extension(
 }

 pub fn fsync_file_and_parent(file_path: &Utf8Path) -> io::Result<()> {
-    let parent = file_path
-        .parent()
-        .ok_or_else(|| io::Error::other(format!("File {file_path:?} has no parent")))?;
+    let parent = file_path.parent().ok_or_else(|| {
+        io::Error::new(
+            io::ErrorKind::Other,
+            format!("File {file_path:?} has no parent"),
+        )
+    })?;

    fsync(file_path)?;
    fsync(parent)?;
--- a/libs/utils/src/elapsed_accum.rs
+++ b/libs/utils/src/elapsed_accum.rs
@@ -1,26 +0,0 @@
-use std::time::{Duration, Instant};
-
-#[derive(Default)]
-pub struct ElapsedAccum {
-    accum: Duration,
-}
-
-impl ElapsedAccum {
-    pub fn get(&self) -> Duration {
-        self.accum
-    }
-    pub fn guard(&mut self) -> impl Drop + '_ {
-        let start = Instant::now();
-        scopeguard::guard(start, |last_wait_at| {
-            self.accum += Instant::now() - last_wait_at;
-        })
-    }
-
-    pub async fn measure<Fut, O>(&mut self, fut: Fut) -> O
-    where
-        Fut: Future<Output = O>,
-    {
-        let _guard = self.guard();
-        fut.await
-    }
-}
--- a/libs/utils/src/fs_ext.rs
+++ b/libs/utils/src/fs_ext.rs
@@ -3,9 +3,7 @@ use std::{fs, io, path::Path};

 use anyhow::Context;

-#[cfg(feature = "rename_noreplace")]
 mod rename_noreplace;
-#[cfg(feature = "rename_noreplace")]
 pub use rename_noreplace::rename_noreplace;

 pub trait PathExt {
--- a/libs/utils/src/fs_ext/rename_noreplace.rs
+++ b/libs/utils/src/fs_ext/rename_noreplace.rs
@@ -8,7 +8,7 @@ pub fn rename_noreplace<P1: ?Sized + NixPath, P2: ?Sized + NixPath>(
    dst: &P2,
 ) -> nix::Result<()> {
    {
-        #[cfg(all(target_os = "linux", target_env = "gnu"))]
+        #[cfg(target_os = "linux")]
        {
            nix::fcntl::renameat2(
                None,
@@ -29,7 +29,7 @@ pub fn rename_noreplace<P1: ?Sized + NixPath, P2: ?Sized + NixPath>(
            })??;
            nix::errno::Errno::result(res).map(drop)
        }
-        #[cfg(not(any(all(target_os = "linux", target_env = "gnu"), target_os = "macos")))]
+        #[cfg(not(any(target_os = "linux", target_os = "macos")))]
        {
            std::compile_error!("OS does not support no-replace renames");
        }
--- a/libs/utils/src/lib.rs
+++ b/libs/utils/src/lib.rs
@@ -93,8 +93,6 @@ pub mod try_rcu;

 pub mod guard_arc_swap;

-pub mod elapsed_accum;
-
 #[cfg(target_os = "linux")]
 pub mod linux_socket_ioctl;

--- a/libs/utils/src/signals.rs
+++ b/libs/utils/src/signals.rs
@@ -1,8 +1,6 @@
 pub use signal_hook::consts::TERM_SIGNALS;
 pub use signal_hook::consts::signal::*;
 use signal_hook::iterator::Signals;
-use tokio::signal::unix::{SignalKind, signal};
-use tracing::info;

 pub enum Signal {
    Quit,
@@ -38,30 +36,3 @@ impl ShutdownSignals {
        Ok(())
    }
 }
-
-/// Runs in a loop since we want to be responsive to multiple signals
-/// even after triggering shutdown (e.g. a SIGQUIT after a slow SIGTERM shutdown)
-/// <https://github.com/neondatabase/neon/issues/9740>
-pub async fn signal_handler(token: tokio_util::sync::CancellationToken) {
-    let mut sigint = signal(SignalKind::interrupt()).unwrap();
-    let mut sigterm = signal(SignalKind::terminate()).unwrap();
-    let mut sigquit = signal(SignalKind::quit()).unwrap();
-
-    loop {
-        let signal = tokio::select! {
-            _ = sigquit.recv() => {
-                info!("Got signal SIGQUIT. Terminating in immediate shutdown mode.");
-                std::process::exit(111);
-            }
-            _ = sigint.recv() => "SIGINT",
-            _ = sigterm.recv() => "SIGTERM",
-        };
-
-        if !token.is_cancelled() {
-            info!("Got signal {signal}. Terminating gracefully in fast shutdown mode.");
-            token.cancel();
-        } else {
-            info!("Got signal {signal}. Already shutting down.");
-        }
-    }
-}
--- a/libs/utils/src/sync/heavier_once_cell.rs
+++ b/libs/utils/src/sync/heavier_once_cell.rs
@@ -111,17 +111,9 @@ impl<T> OnceCell<T> {
        }
    }

-    /// Like [`Self::get_or_init_detached_measured`], but without out parameter for time spent waiting.
-    pub async fn get_or_init_detached(&self) -> Result<Guard<'_, T>, InitPermit> {
-        self.get_or_init_detached_measured(None).await
-    }
-
    /// Returns a guard to an existing initialized value, or returns an unique initialization
    /// permit which can be used to initialize this `OnceCell` using `OnceCell::set`.
-    pub async fn get_or_init_detached_measured(
-        &self,
-        mut wait_time: Option<&mut crate::elapsed_accum::ElapsedAccum>,
-    ) -> Result<Guard<'_, T>, InitPermit> {
+    pub async fn get_or_init_detached(&self) -> Result<Guard<'_, T>, InitPermit> {
        // It looks like OnceCell::get_or_init could be implemented using this method instead of
        // duplication. However, that makes the future be !Send due to possibly holding on to the
        // MutexGuard over an await point.
@@ -133,16 +125,12 @@ impl<T> OnceCell<T> {
                }
                guard.init_semaphore.clone()
            };
+
            {
                let permit = {
                    // increment the count for the duration of queued
                    let _guard = CountWaitingInitializers::start(self);
-                    let fut = sem.acquire();
-                    if let Some(wait_time) = wait_time.as_mut() {
-                        wait_time.measure(fut).await
-                    } else {
-                        fut.await
-                    }
+                    sem.acquire().await
                };

                let Ok(permit) = permit else {
--- a/object_storage/Cargo.toml
+++ b/object_storage/Cargo.toml
@@ -1,28 +0,0 @@
-[package]
-name = "object_storage"
-version = "0.0.1"
-edition.workspace = true
-license.workspace = true
-[dependencies]
-anyhow.workspace = true
-axum-extra.workspace = true
-axum.workspace = true
-camino.workspace = true
-futures.workspace = true
-jsonwebtoken.workspace = true
-prometheus.workspace = true
-remote_storage.workspace = true
-serde.workspace = true
-serde_json.workspace = true
-tokio-util.workspace = true
-tokio.workspace = true
-tracing.workspace = true
-utils = { path = "../libs/utils", default-features = false }
-workspace_hack.workspace = true
-[dev-dependencies]
-camino-tempfile.workspace = true
-http-body-util.workspace = true
-itertools.workspace = true
-rand.workspace = true
-test-log.workspace = true
-tower.workspace = true
--- a/object_storage/src/app.rs
+++ b/object_storage/src/app.rs
@@ -1,561 +0,0 @@
-use anyhow::anyhow;
-use axum::body::{Body, Bytes};
-use axum::response::{IntoResponse, Response};
-use axum::{Router, http::StatusCode};
-use object_storage::{PrefixS3Path, S3Path, Storage, bad_request, internal_error, not_found, ok};
-use remote_storage::TimeoutOrCancel;
-use remote_storage::{DownloadError, DownloadOpts, GenericRemoteStorage, RemotePath};
-use std::{sync::Arc, time::SystemTime, time::UNIX_EPOCH};
-use tokio_util::sync::CancellationToken;
-use tracing::{error, info};
-use utils::backoff::retry;
-
-pub fn app(state: Arc<Storage>) -> Router<()> {
-    use axum::routing::{delete as _delete, get as _get};
-    let delete_prefix = _delete(delete_prefix);
-    Router::new()
-        .route(
-            "/{tenant_id}/{timeline_id}/{endpoint_id}/{*path}",
-            _get(get).put(set).delete(delete),
-        )
-        .route(
-            "/{tenant_id}/{timeline_id}/{endpoint_id}",
-            delete_prefix.clone(),
-        )
-        .route("/{tenant_id}/{timeline_id}", delete_prefix.clone())
-        .route("/{tenant_id}", delete_prefix)
-        .route("/metrics", _get(metrics))
-        .route("/status", _get(async || StatusCode::OK.into_response()))
-        .with_state(state)
-}
-
-type Result = anyhow::Result<Response, Response>;
-type State = axum::extract::State<Arc<Storage>>;
-
-const CONTENT_TYPE: &str = "content-type";
-const APPLICATION_OCTET_STREAM: &str = "application/octet-stream";
-const WARN_THRESHOLD: u32 = 3;
-const MAX_RETRIES: u32 = 10;
-
-async fn metrics() -> Result {
-    prometheus::TextEncoder::new()
-        .encode_to_string(&prometheus::gather())
-        .map(|s| s.into_response())
-        .map_err(|e| internal_error(e, "/metrics", "collecting metrics"))
-}
-
-async fn get(S3Path { path }: S3Path, state: State) -> Result {
-    info!(%path, "downloading");
-    let download_err = |e| {
-        if let DownloadError::NotFound = e {
-            info!(%path, %e, "downloading"); // 404 is not an issue of _this_ service
-            return not_found(&path);
-        }
-        internal_error(e, &path, "downloading")
-    };
-    let cancel = state.cancel.clone();
-    let opts = &DownloadOpts::default();
-
-    let stream = retry(
-        async || state.storage.download(&path, opts, &cancel).await,
-        DownloadError::is_permanent,
-        WARN_THRESHOLD,
-        MAX_RETRIES,
-        "downloading",
-        &cancel,
-    )
-    .await
-    .unwrap_or(Err(DownloadError::Cancelled))
-    .map_err(download_err)?
-    .download_stream;
-
-    Response::builder()
-        .status(StatusCode::OK)
-        .header(CONTENT_TYPE, APPLICATION_OCTET_STREAM)
-        .body(Body::from_stream(stream))
-        .map_err(|e| internal_error(e, path, "reading response"))
-}
-
-// Best solution for files is multipart upload, but remote_storage doesn't support it,
-// so we can either read Bytes in memory and push at once or forward BodyDataStream to
-// remote_storage. The latter may seem more peformant, but BodyDataStream doesn't have a
-// guaranteed size() which may produce issues while uploading to s3.
-// So, currently we're going with an in-memory copy plus a boundary to prevent uploading
-// very large files.
-async fn set(S3Path { path }: S3Path, state: State, bytes: Bytes) -> Result {
-    info!(%path, "uploading");
-    let request_len = bytes.len();
-    let max_len = state.max_upload_file_limit;
-    if request_len > max_len {
-        return Err(bad_request(
-            anyhow!("File size {request_len} exceeds max {max_len}"),
-            "uploading",
-        ));
-    }
-
-    let cancel = state.cancel.clone();
-    let fun = async || {
-        let stream = bytes_to_stream(bytes.clone());
-        state
-            .storage
-            .upload(stream, request_len, &path, None, &cancel)
-            .await
-    };
-    retry(
-        fun,
-        TimeoutOrCancel::caused_by_cancel,
-        WARN_THRESHOLD,
-        MAX_RETRIES,
-        "uploading",
-        &cancel,
-    )
-    .await
-    .unwrap_or(Err(anyhow!("uploading cancelled")))
-    .map_err(|e| internal_error(e, path, "reading response"))?;
-    Ok(ok())
-}
-
-async fn delete(S3Path { path }: S3Path, state: State) -> Result {
-    info!(%path, "deleting");
-    let cancel = state.cancel.clone();
-    retry(
-        async || state.storage.delete(&path, &cancel).await,
-        TimeoutOrCancel::caused_by_cancel,
-        WARN_THRESHOLD,
-        MAX_RETRIES,
-        "deleting",
-        &cancel,
-    )
-    .await
-    .unwrap_or(Err(anyhow!("deleting cancelled")))
-    .map_err(|e| internal_error(e, path, "deleting"))?;
-    Ok(ok())
-}
-
-async fn delete_prefix(PrefixS3Path { path }: PrefixS3Path, state: State) -> Result {
-    info!(%path, "deleting prefix");
-    let cancel = state.cancel.clone();
-    retry(
-        async || state.storage.delete_prefix(&path, &cancel).await,
-        TimeoutOrCancel::caused_by_cancel,
-        WARN_THRESHOLD,
-        MAX_RETRIES,
-        "deleting prefix",
-        &cancel,
-    )
-    .await
-    .unwrap_or(Err(anyhow!("deleting prefix cancelled")))
-    .map_err(|e| internal_error(e, path, "deleting prefix"))?;
-    Ok(ok())
-}
-
-pub async fn check_storage_permissions(
-    client: &GenericRemoteStorage,
-    cancel: CancellationToken,
-) -> anyhow::Result<()> {
-    info!("storage permissions check");
-
-    // as_nanos() as multiple instances proxying same bucket may be started at once
-    let now = SystemTime::now()
-        .duration_since(UNIX_EPOCH)?
-        .as_nanos()
-        .to_string();
-
-    let path = RemotePath::from_string(&format!("write_access_{now}"))?;
-    info!(%path, "uploading");
-
-    let body = now.to_string();
-    let stream = bytes_to_stream(Bytes::from(body.clone()));
-    client
-        .upload(stream, body.len(), &path, None, &cancel)
-        .await?;
-
-    use tokio::io::AsyncReadExt;
-    info!(%path, "downloading");
-    let download_opts = DownloadOpts {
-        kind: remote_storage::DownloadKind::Small,
-        ..Default::default()
-    };
-    let mut body_read_buf = Vec::new();
-    let stream = client
-        .download(&path, &download_opts, &cancel)
-        .await?
-        .download_stream;
-    tokio_util::io::StreamReader::new(stream)
-        .read_to_end(&mut body_read_buf)
-        .await?;
-    let body_read = String::from_utf8(body_read_buf)?;
-    if body != body_read {
-        error!(%body, %body_read, "File contents do not match");
-        anyhow::bail!("Read back file doesn't match original")
-    }
-
-    info!(%path, "removing");
-    client.delete(&path, &cancel).await
-}
-
-fn bytes_to_stream(bytes: Bytes) -> impl futures::Stream<Item = std::io::Result<Bytes>> {
-    futures::stream::once(futures::future::ready(Ok(bytes)))
-}
-
-#[cfg(test)]
-mod tests {
-    use super::*;
-    use axum::{body::Body, extract::Request, response::Response};
-    use http_body_util::BodyExt;
-    use itertools::iproduct;
-    use std::env::var;
-    use std::sync::Arc;
-    use std::time::Duration;
-    use test_log::test as testlog;
-    use tower::{Service, util::ServiceExt};
-    use utils::id::{TenantId, TimelineId};
-
-    // see libs/remote_storage/tests/test_real_s3.rs
-    const REAL_S3_ENV: &str = "ENABLE_REAL_S3_REMOTE_STORAGE";
-    const REAL_S3_BUCKET: &str = "REMOTE_STORAGE_S3_BUCKET";
-    const REAL_S3_REGION: &str = "REMOTE_STORAGE_S3_REGION";
-
-    async fn proxy() -> (Storage, Option<camino_tempfile::Utf8TempDir>) {
-        let cancel = CancellationToken::new();
-        let (dir, storage) = if var(REAL_S3_ENV).is_err() {
-            // tests execute in parallel and we need a new directory for each of them
-            let dir = camino_tempfile::tempdir().unwrap();
-            let fs =
-                remote_storage::LocalFs::new(dir.path().into(), Duration::from_secs(5)).unwrap();
-            (Some(dir), GenericRemoteStorage::LocalFs(fs))
-        } else {
-            // test_real_s3::create_s3_client is hard to reference, reimplementing here
-            let millis = SystemTime::now()
-                .duration_since(UNIX_EPOCH)
-                .unwrap()
-                .as_millis();
-            use rand::Rng;
-            let random = rand::thread_rng().r#gen::<u32>();
-
-            let s3_config = remote_storage::S3Config {
-                bucket_name: var(REAL_S3_BUCKET).unwrap(),
-                bucket_region: var(REAL_S3_REGION).unwrap(),
-                prefix_in_bucket: Some(format!("test_{millis}_{random:08x}/")),
-                endpoint: None,
-                concurrency_limit: std::num::NonZeroUsize::new(100).unwrap(),
-                max_keys_per_list_response: None,
-                upload_storage_class: None,
-            };
-            let bucket = remote_storage::S3Bucket::new(&s3_config, Duration::from_secs(1))
-                .await
-                .unwrap();
-            (None, GenericRemoteStorage::AwsS3(Arc::new(bucket)))
-        };
-
-        let proxy = Storage {
-            auth: object_storage::JwtAuth::new(TEST_PUB_KEY_ED25519).unwrap(),
-            storage,
-            cancel: cancel.clone(),
-            max_upload_file_limit: usize::MAX,
-        };
-        check_storage_permissions(&proxy.storage, cancel)
-            .await
-            .unwrap();
-        (proxy, dir)
-    }
-
-    // see libs/utils/src/auth.rs
-    const TEST_PUB_KEY_ED25519: &[u8] = b"
-----BEGIN PUBLIC KEY-----
-MCowBQYDK2VwAyEARYwaNBayR+eGI0iXB4s3QxE3Nl2g1iWbr6KtLWeVD/w=
-----END PUBLIC KEY-----
-";
-
-    const TEST_PRIV_KEY_ED25519: &[u8] = br#"
-----BEGIN PRIVATE KEY-----
-MC4CAQAwBQYDK2VwBCIEID/Drmc1AA6U/znNRWpF3zEGegOATQxfkdWxitcOMsIH
-----END PRIVATE KEY-----
-"#;
-
-    async fn request(req: Request<Body>) -> Response<Body> {
-        let (proxy, _) = proxy().await;
-        app(Arc::new(proxy))
-            .into_service()
-            .oneshot(req)
-            .await
-            .unwrap()
-    }
-
-    #[testlog(tokio::test)]
-    async fn status() {
-        let res = Request::builder()
-            .uri("/status")
-            .body(Body::empty())
-            .map(request)
-            .unwrap()
-            .await;
-        assert_eq!(res.status(), StatusCode::OK);
-    }
-
-    fn routes() -> impl Iterator<Item = (&'static str, &'static str)> {
-        iproduct!(
-            vec!["/1", "/1/2", "/1/2/3", "/1/2/3/4"],
-            vec!["GET", "PUT", "DELETE"]
-        )
-    }
-
-    #[testlog(tokio::test)]
-    async fn no_token() {
-        for (uri, method) in routes() {
-            info!(%uri, %method);
-            let res = Request::builder()
-                .uri(uri)
-                .method(method)
-                .body(Body::empty())
-                .map(request)
-                .unwrap()
-                .await;
-            assert!(matches!(
-                res.status(),
-                StatusCode::METHOD_NOT_ALLOWED | StatusCode::BAD_REQUEST
-            ));
-        }
-    }
-
-    #[testlog(tokio::test)]
-    async fn invalid_token() {
-        for (uri, method) in routes() {
-            info!(%uri, %method);
-            let status = Request::builder()
-                .uri(uri)
-                .header("Authorization", "Bearer 123")
-                .method(method)
-                .body(Body::empty())
-                .map(request)
-                .unwrap()
-                .await;
-            assert!(matches!(
-                status.status(),
-                StatusCode::METHOD_NOT_ALLOWED | StatusCode::BAD_REQUEST
-            ));
-        }
-    }
-
-    const TENANT_ID: TenantId =
-        TenantId::from_array([0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 1, 2, 3, 4, 5, 6]);
-    const TIMELINE_ID: TimelineId =
-        TimelineId::from_array([0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 1, 2, 3, 4, 5, 7]);
-    const ENDPOINT_ID: &str = "ep-winter-frost-a662z3vg";
-    fn token() -> String {
-        let claims = object_storage::Claims {
-            tenant_id: TENANT_ID,
-            timeline_id: TIMELINE_ID,
-            endpoint_id: ENDPOINT_ID.into(),
-            exp: u64::MAX,
-        };
-        let key = jsonwebtoken::EncodingKey::from_ed_pem(TEST_PRIV_KEY_ED25519).unwrap();
-        let header = jsonwebtoken::Header::new(object_storage::VALIDATION_ALGO);
-        jsonwebtoken::encode(&header, &claims, &key).unwrap()
-    }
-
-    #[testlog(tokio::test)]
-    async fn unauthorized() {
-        let (proxy, _) = proxy().await;
-        let mut app = app(Arc::new(proxy)).into_service();
-        let token = token();
-        let args = itertools::iproduct!(
-            vec![TENANT_ID.to_string(), TenantId::generate().to_string()],
-            vec![TIMELINE_ID.to_string(), TimelineId::generate().to_string()],
-            vec![ENDPOINT_ID, "ep-ololo"]
-        )
-        .skip(1);
-
-        for ((uri, method), (tenant, timeline, endpoint)) in iproduct!(routes(), args) {
-            info!(%uri, %method, %tenant, %timeline, %endpoint);
-            let request = Request::builder()
-                .uri(format!("/{tenant}/{timeline}/{endpoint}/sub/path/key"))
-                .method(method)
-                .header("Authorization", format!("Bearer {}", token))
-                .body(Body::empty())
-                .unwrap();
-            let status = ServiceExt::ready(&mut app)
-                .await
-                .unwrap()
-                .call(request)
-                .await
-                .unwrap()
-                .status();
-            assert_eq!(status, StatusCode::UNAUTHORIZED);
-        }
-    }
-
-    #[testlog(tokio::test)]
-    async fn method_not_allowed() {
-        let token = token();
-        let iter = iproduct!(vec!["", "/.."], vec!["GET", "PUT"]);
-        for (key, method) in iter {
-            let status = Request::builder()
-                .uri(format!("/{TENANT_ID}/{TIMELINE_ID}/{ENDPOINT_ID}{key}"))
-                .method(method)
-                .header("Authorization", format!("Bearer {token}"))
-                .body(Body::empty())
-                .map(request)
-                .unwrap()
-                .await
-                .status();
-            assert!(matches!(
-                status,
-                StatusCode::BAD_REQUEST | StatusCode::METHOD_NOT_ALLOWED
-            ));
-        }
-    }
-
-    async fn requests_chain(
-        chain: impl Iterator<Item = (String, &str, &'static str, StatusCode, bool)>,
-        token: impl Fn(&str) -> String,
-    ) {
-        let (proxy, _) = proxy().await;
-        let mut app = app(Arc::new(proxy)).into_service();
-        for (uri, method, body, expected_status, compare_body) in chain {
-            info!(%uri, %method, %body, %expected_status);
-            let bearer = format!("Bearer {}", token(&uri));
-            let request = Request::builder()
-                .uri(uri)
-                .method(method)
-                .header("Authorization", &bearer)
-                .body(Body::from(body))
-                .unwrap();
-            let response = ServiceExt::ready(&mut app)
-                .await
-                .unwrap()
-                .call(request)
-                .await
-                .unwrap();
-            assert_eq!(response.status(), expected_status);
-            if !compare_body {
-                continue;
-            }
-            let read_body = response.into_body().collect().await.unwrap().to_bytes();
-            assert_eq!(body, read_body);
-        }
-    }
-
-    #[testlog(tokio::test)]
-    async fn metrics() {
-        let uri = format!("/{TENANT_ID}/{TIMELINE_ID}/{ENDPOINT_ID}/key");
-        let req = vec![
-            (uri.clone(), "PUT", "body", StatusCode::OK, false),
-            (uri.clone(), "DELETE", "", StatusCode::OK, false),
-        ];
-        requests_chain(req.into_iter(), |_| token()).await;
-
-        let res = Request::builder()
-            .uri("/metrics")
-            .body(Body::empty())
-            .map(request)
-            .unwrap()
-            .await;
-        assert_eq!(res.status(), StatusCode::OK);
-        let body = res.into_body().collect().await.unwrap().to_bytes();
-        let body = String::from_utf8_lossy(&body);
-        tracing::debug!(%body);
-        // Storage metrics are not gathered for LocalFs
-        if var(REAL_S3_ENV).is_ok() {
-            assert!(body.contains("remote_storage_s3_deleted_objects_total"));
-        }
-        assert!(body.contains("process_threads"));
-    }
-
-    #[testlog(tokio::test)]
-    async fn insert_retrieve_remove() {
-        let uri = format!("/{TENANT_ID}/{TIMELINE_ID}/{ENDPOINT_ID}/key");
-        let chain = vec![
-            (uri.clone(), "GET", "", StatusCode::NOT_FOUND, false),
-            (uri.clone(), "PUT", "пыщьпыщь", StatusCode::OK, false),
-            (uri.clone(), "GET", "пыщьпыщь", StatusCode::OK, true),
-            (uri.clone(), "DELETE", "", StatusCode::OK, false),
-            (uri, "GET", "", StatusCode::NOT_FOUND, false),
-        ];
-        requests_chain(chain.into_iter(), |_| token()).await;
-    }
-
-    fn delete_prefix_token(uri: &str) -> String {
-        use serde::Serialize;
-        let parts = uri.split("/").collect::<Vec<&str>>();
-        #[derive(Serialize)]
-        struct PrefixClaims {
-            tenant_id: TenantId,
-            timeline_id: Option<TimelineId>,
-            endpoint_id: Option<object_storage::EndpointId>,
-            exp: u64,
-        }
-        let claims = PrefixClaims {
-            tenant_id: parts.get(1).map(|c| c.parse().unwrap()).unwrap(),
-            timeline_id: parts.get(2).map(|c| c.parse().unwrap()),
-            endpoint_id: parts.get(3).map(ToString::to_string),
-            exp: u64::MAX,
-        };
-        let key = jsonwebtoken::EncodingKey::from_ed_pem(TEST_PRIV_KEY_ED25519).unwrap();
-        let header = jsonwebtoken::Header::new(object_storage::VALIDATION_ALGO);
-        jsonwebtoken::encode(&header, &claims, &key).unwrap()
-    }
-
-    // Can't use single digit numbers as they won't be validated as TimelineId and EndpointId
-    #[testlog(tokio::test)]
-    async fn delete_prefix() {
-        let tenant_id =
-            TenantId::from_array([1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]).to_string();
-        let t2 = TimelineId::from_array([2, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]);
-        let t3 = TimelineId::from_array([3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]);
-        let t4 = TimelineId::from_array([4, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]);
-        let f = |timeline, path| format!("/{tenant_id}/{timeline}{path}");
-        // Why extra slash in string literals? Axum is weird with URIs:
-        // /1/2 and 1/2/ match different routes, thus first yields OK and second NOT_FOUND
-        //  as it matches /tenant/timeline/endpoint, see https://stackoverflow.com/a/75355932
-        // The cost of removing trailing slash is suprisingly hard:
-        // * Add tower dependency with NormalizePath layer
-        // * wrap Router<()> in this layer https://github.com/tokio-rs/axum/discussions/2377
-        // * Rewrite make_service() -> into_make_service()
-        // * Rewrite oneshot() (not available for NormalizePath)
-        // I didn't manage to get it working correctly
-        let chain = vec![
-            // create 1/2/3/4, 1/2/3/5, delete prefix 1/2/3 -> empty
-            (f(t2, "/3/4"), "PUT", "", StatusCode::OK, false),
-            (f(t2, "/3/4"), "PUT", "", StatusCode::OK, false), // we can override file contents
-            (f(t2, "/3/5"), "PUT", "", StatusCode::OK, false),
-            (f(t2, "/3"), "DELETE", "", StatusCode::OK, false),
-            (f(t2, "/3/4"), "GET", "", StatusCode::NOT_FOUND, false),
-            (f(t2, "/3/5"), "GET", "", StatusCode::NOT_FOUND, false),
-            // create 1/2/3/4, 1/2/5/6, delete prefix 1/2/3 -> 1/2/5/6
-            (f(t2, "/3/4"), "PUT", "", StatusCode::OK, false),
-            (f(t2, "/5/6"), "PUT", "", StatusCode::OK, false),
-            (f(t2, "/3"), "DELETE", "", StatusCode::OK, false),
-            (f(t2, "/3/4"), "GET", "", StatusCode::NOT_FOUND, false),
-            (f(t2, "/5/6"), "GET", "", StatusCode::OK, false),
-            // create 1/2/3/4, 1/2/7/8, delete prefix 1/2 -> empty
-            (f(t2, "/3/4"), "PUT", "", StatusCode::OK, false),
-            (f(t2, "/7/8"), "PUT", "", StatusCode::OK, false),
-            (f(t2, ""), "DELETE", "", StatusCode::OK, false),
-            (f(t2, "/3/4"), "GET", "", StatusCode::NOT_FOUND, false),
-            (f(t2, "/7/8"), "GET", "", StatusCode::NOT_FOUND, false),
-            // create 1/2/3/4, 1/2/5/6, 1/3/8/9, delete prefix 1/2/3 -> 1/2/5/6, 1/3/8/9
-            (f(t2, "/3/4"), "PUT", "", StatusCode::OK, false),
-            (f(t2, "/5/6"), "PUT", "", StatusCode::OK, false),
-            (f(t3, "/8/9"), "PUT", "", StatusCode::OK, false),
-            (f(t2, "/3"), "DELETE", "", StatusCode::OK, false),
-            (f(t2, "/3/4"), "GET", "", StatusCode::NOT_FOUND, false),
-            (f(t2, "/5/6"), "GET", "", StatusCode::OK, false),
-            (f(t3, "/8/9"), "GET", "", StatusCode::OK, false),
-            // create 1/4/5/6, delete prefix 1/2 -> 1/3/8/9, 1/4/5/6
-            (f(t4, "/5/6"), "PUT", "", StatusCode::OK, false),
-            (f(t2, ""), "DELETE", "", StatusCode::OK, false),
-            (f(t2, "/3/4"), "GET", "", StatusCode::NOT_FOUND, false),
-            (f(t2, "/5/6"), "GET", "", StatusCode::NOT_FOUND, false),
-            (f(t3, "/8/9"), "GET", "", StatusCode::OK, false),
-            (f(t4, "/5/6"), "GET", "", StatusCode::OK, false),
-            // delete prefix 1 -> empty
-            (format!("/{tenant_id}"), "DELETE", "", StatusCode::OK, false),
-            (f(t2, "/3/4"), "GET", "", StatusCode::NOT_FOUND, false),
-            (f(t2, "/5/6"), "GET", "", StatusCode::NOT_FOUND, false),
-            (f(t3, "/8/9"), "GET", "", StatusCode::NOT_FOUND, false),
-            (f(t4, "/5/6"), "GET", "", StatusCode::NOT_FOUND, false),
-        ];
-        requests_chain(chain.into_iter(), delete_prefix_token).await;
-    }
-}
--- a/object_storage/src/lib.rs
+++ b/object_storage/src/lib.rs
@@ -1,344 +0,0 @@
-use anyhow::Result;
-use axum::extract::{FromRequestParts, Path};
-use axum::response::{IntoResponse, Response};
-use axum::{RequestPartsExt, http::StatusCode, http::request::Parts};
-use axum_extra::TypedHeader;
-use axum_extra::headers::{Authorization, authorization::Bearer};
-use camino::Utf8PathBuf;
-use jsonwebtoken::{DecodingKey, Validation};
-use remote_storage::{GenericRemoteStorage, RemotePath};
-use serde::{Deserialize, Serialize};
-use std::fmt::Display;
-use std::result::Result as StdResult;
-use std::sync::Arc;
-use tokio_util::sync::CancellationToken;
-use tracing::{debug, error};
-use utils::id::{TenantId, TimelineId};
-
-// simplified version of utils::auth::JwtAuth
-pub struct JwtAuth {
-    decoding_key: DecodingKey,
-    validation: Validation,
-}
-
-pub const VALIDATION_ALGO: jsonwebtoken::Algorithm = jsonwebtoken::Algorithm::EdDSA;
-impl JwtAuth {
-    pub fn new(key: &[u8]) -> Result<Self> {
-        Ok(Self {
-            decoding_key: DecodingKey::from_ed_pem(key)?,
-            validation: Validation::new(VALIDATION_ALGO),
-        })
-    }
-
-    pub fn decode<T: serde::de::DeserializeOwned>(&self, token: &str) -> Result<T> {
-        Ok(jsonwebtoken::decode(token, &self.decoding_key, &self.validation).map(|t| t.claims)?)
-    }
-}
-
-fn normalize_key(key: &str) -> StdResult<Utf8PathBuf, String> {
-    let key = clean_utf8(&Utf8PathBuf::from(key));
-    if key.starts_with("..") || key == "." || key == "/" {
-        return Err(format!("invalid key {key}"));
-    }
-    match key.strip_prefix("/").map(Utf8PathBuf::from) {
-        Ok(p) => Ok(p),
-        _ => Ok(key),
-    }
-}
-
-// Copied from path_clean crate with PathBuf->Utf8PathBuf
-fn clean_utf8(path: &camino::Utf8Path) -> Utf8PathBuf {
-    use camino::Utf8Component as Comp;
-    let mut out = Vec::new();
-    for comp in path.components() {
-        match comp {
-            Comp::CurDir => (),
-            Comp::ParentDir => match out.last() {
-                Some(Comp::RootDir) => (),
-                Some(Comp::Normal(_)) => {
-                    out.pop();
-                }
-                None | Some(Comp::CurDir) | Some(Comp::ParentDir) | Some(Comp::Prefix(_)) => {
-                    out.push(comp)
-                }
-            },
-            comp => out.push(comp),
-        }
-    }
-    if !out.is_empty() {
-        out.iter().collect()
-    } else {
-        Utf8PathBuf::from(".")
-    }
-}
-
-pub struct Storage {
-    pub auth: JwtAuth,
-    pub storage: GenericRemoteStorage,
-    pub cancel: CancellationToken,
-    pub max_upload_file_limit: usize,
-}
-
-pub type EndpointId = String; // If needed, reuse small string from proxy/src/types.rc
-
-#[derive(Deserialize, Serialize, PartialEq)]
-pub struct Claims {
-    pub tenant_id: TenantId,
-    pub timeline_id: TimelineId,
-    pub endpoint_id: EndpointId,
-    pub exp: u64,
-}
-
-impl Display for Claims {
-    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
-        write!(
-            f,
-            "Claims(tenant_id {} timeline_id {} endpoint_id {} exp {})",
-            self.tenant_id, self.timeline_id, self.endpoint_id, self.exp
-        )
-    }
-}
-
-#[derive(Deserialize, Serialize)]
-struct KeyRequest {
-    tenant_id: TenantId,
-    timeline_id: TimelineId,
-    endpoint_id: EndpointId,
-    path: String,
-}
-
-#[derive(Debug, PartialEq)]
-pub struct S3Path {
-    pub path: RemotePath,
-}
-
-impl TryFrom<&KeyRequest> for S3Path {
-    type Error = String;
-    fn try_from(req: &KeyRequest) -> StdResult<Self, Self::Error> {
-        let KeyRequest {
-            tenant_id,
-            timeline_id,
-            endpoint_id,
-            path,
-        } = &req;
-        let prefix = format!("{tenant_id}/{timeline_id}/{endpoint_id}",);
-        let path = Utf8PathBuf::from(prefix).join(normalize_key(path)?);
-        let path = RemotePath::new(&path).unwrap(); // unwrap() because the path is already relative
-        Ok(S3Path { path })
-    }
-}
-
-fn unauthorized(route: impl Display, claims: impl Display) -> Response {
-    debug!(%route, %claims, "route doesn't match claims");
-    StatusCode::UNAUTHORIZED.into_response()
-}
-
-pub fn bad_request(err: impl Display, desc: &'static str) -> Response {
-    debug!(%err, desc);
-    (StatusCode::BAD_REQUEST, err.to_string()).into_response()
-}
-
-pub fn ok() -> Response {
-    StatusCode::OK.into_response()
-}
-
-pub fn internal_error(err: impl Display, path: impl Display, desc: &'static str) -> Response {
-    error!(%err, %path, desc);
-    StatusCode::INTERNAL_SERVER_ERROR.into_response()
-}
-
-pub fn not_found(key: impl ToString) -> Response {
-    (StatusCode::NOT_FOUND, key.to_string()).into_response()
-}
-
-impl FromRequestParts<Arc<Storage>> for S3Path {
-    type Rejection = Response;
-    async fn from_request_parts(
-        parts: &mut Parts,
-        state: &Arc<Storage>,
-    ) -> Result<Self, Self::Rejection> {
-        let Path(path): Path<KeyRequest> = parts
-            .extract()
-            .await
-            .map_err(|e| bad_request(e, "invalid route"))?;
-        let TypedHeader(Authorization(bearer)) = parts
-            .extract::<TypedHeader<Authorization<Bearer>>>()
-            .await
-            .map_err(|e| bad_request(e, "invalid token"))?;
-        let claims: Claims = state
-            .auth
-            .decode(bearer.token())
-            .map_err(|e| bad_request(e, "decoding token"))?;
-        let route = Claims {
-            tenant_id: path.tenant_id,
-            timeline_id: path.timeline_id,
-            endpoint_id: path.endpoint_id.clone(),
-            exp: claims.exp,
-        };
-        if route != claims {
-            return Err(unauthorized(route, claims));
-        }
-        (&path)
-            .try_into()
-            .map_err(|e| bad_request(e, "invalid route"))
-    }
-}
-
-#[derive(Deserialize, Serialize, PartialEq)]
-pub struct PrefixKeyPath {
-    pub tenant_id: TenantId,
-    pub timeline_id: Option<TimelineId>,
-    pub endpoint_id: Option<EndpointId>,
-}
-
-impl Display for PrefixKeyPath {
-    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
-        write!(
-            f,
-            "PrefixKeyPath(tenant_id {} timeline_id {} endpoint_id {})",
-            self.tenant_id,
-            self.timeline_id
-                .as_ref()
-                .map(ToString::to_string)
-                .unwrap_or("".to_string()),
-            self.endpoint_id
-                .as_ref()
-                .map(ToString::to_string)
-                .unwrap_or("".to_string())
-        )
-    }
-}
-
-#[derive(Debug, PartialEq)]
-pub struct PrefixS3Path {
-    pub path: RemotePath,
-}
-
-impl From<&PrefixKeyPath> for PrefixS3Path {
-    fn from(path: &PrefixKeyPath) -> Self {
-        let timeline_id = path
-            .timeline_id
-            .as_ref()
-            .map(ToString::to_string)
-            .unwrap_or("".to_string());
-        let endpoint_id = path
-            .endpoint_id
-            .as_ref()
-            .map(ToString::to_string)
-            .unwrap_or("".to_string());
-        let path = Utf8PathBuf::from(path.tenant_id.to_string())
-            .join(timeline_id)
-            .join(endpoint_id);
-        let path = RemotePath::new(&path).unwrap(); // unwrap() because the path is already relative
-        PrefixS3Path { path }
-    }
-}
-
-impl FromRequestParts<Arc<Storage>> for PrefixS3Path {
-    type Rejection = Response;
-    async fn from_request_parts(
-        parts: &mut Parts,
-        state: &Arc<Storage>,
-    ) -> Result<Self, Self::Rejection> {
-        let Path(path) = parts
-            .extract::<Path<PrefixKeyPath>>()
-            .await
-            .map_err(|e| bad_request(e, "invalid route"))?;
-        let TypedHeader(Authorization(bearer)) = parts
-            .extract::<TypedHeader<Authorization<Bearer>>>()
-            .await
-            .map_err(|e| bad_request(e, "invalid token"))?;
-        let claims: PrefixKeyPath = state
-            .auth
-            .decode(bearer.token())
-            .map_err(|e| bad_request(e, "invalid token"))?;
-        if path != claims {
-            return Err(unauthorized(path, claims));
-        }
-        Ok((&path).into())
-    }
-}
-
-#[cfg(test)]
-mod tests {
-    use super::*;
-
-    #[test]
-    fn normalize_key() {
-        let f = super::normalize_key;
-        assert_eq!(f("hello/world/..").unwrap(), Utf8PathBuf::from("hello"));
-        assert_eq!(
-            f("ololo/1/../../not_ololo").unwrap(),
-            Utf8PathBuf::from("not_ololo")
-        );
-        assert!(f("ololo/1/../../../").is_err());
-        assert!(f(".").is_err());
-        assert!(f("../").is_err());
-        assert!(f("").is_err());
-        assert_eq!(f("/1/2/3").unwrap(), Utf8PathBuf::from("1/2/3"));
-        assert!(f("/1/2/3/../../../").is_err());
-        assert!(f("/1/2/3/../../../../").is_err());
-    }
-
-    const TENANT_ID: TenantId =
-        TenantId::from_array([1, 1, 2, 3, 4, 5, 6, 7, 8, 9, 1, 2, 3, 4, 5, 6]);
-    const TIMELINE_ID: TimelineId =
-        TimelineId::from_array([1, 1, 2, 3, 4, 5, 6, 7, 8, 9, 1, 2, 3, 4, 5, 7]);
-    const ENDPOINT_ID: &str = "ep-winter-frost-a662z3vg";
-
-    #[test]
-    fn s3_path() {
-        let auth = Claims {
-            tenant_id: TENANT_ID,
-            timeline_id: TIMELINE_ID,
-            endpoint_id: ENDPOINT_ID.into(),
-            exp: u64::MAX,
-        };
-        let s3_path = |key| {
-            let path = &format!("{TENANT_ID}/{TIMELINE_ID}/{ENDPOINT_ID}/{key}");
-            let path = RemotePath::from_string(path).unwrap();
-            S3Path { path }
-        };
-
-        let path = "cache_key".to_string();
-        let mut key_path = KeyRequest {
-            path,
-            tenant_id: auth.tenant_id,
-            timeline_id: auth.timeline_id,
-            endpoint_id: auth.endpoint_id,
-        };
-        assert_eq!(S3Path::try_from(&key_path).unwrap(), s3_path(key_path.path));
-
-        key_path.path = "we/can/have/nested/paths".to_string();
-        assert_eq!(S3Path::try_from(&key_path).unwrap(), s3_path(key_path.path));
-
-        key_path.path = "../error/hello/../".to_string();
-        assert!(S3Path::try_from(&key_path).is_err());
-    }
-
-    #[test]
-    fn prefix_s3_path() {
-        let mut path = PrefixKeyPath {
-            tenant_id: TENANT_ID,
-            timeline_id: None,
-            endpoint_id: None,
-        };
-        let prefix_path = |s: String| RemotePath::from_string(&s).unwrap();
-        assert_eq!(
-            PrefixS3Path::from(&path).path,
-            prefix_path(format!("{TENANT_ID}"))
-        );
-
-        path.timeline_id = Some(TIMELINE_ID);
-        assert_eq!(
-            PrefixS3Path::from(&path).path,
-            prefix_path(format!("{TENANT_ID}/{TIMELINE_ID}"))
-        );
-
-        path.endpoint_id = Some(ENDPOINT_ID.into());
-        assert_eq!(
-            PrefixS3Path::from(&path).path,
-            prefix_path(format!("{TENANT_ID}/{TIMELINE_ID}/{ENDPOINT_ID}"))
-        );
-    }
-}
--- a/object_storage/src/main.rs
+++ b/object_storage/src/main.rs
@@ -1,65 +0,0 @@
-//! `object_storage` is a service which provides API for uploading and downloading
-//! files. It is used by compute and control plane for accessing LFC prewarm data.
-//! This service is deployed either as a separate component or as part of compute image
-//! for large computes.
-mod app;
-use anyhow::Context;
-use tracing::info;
-use utils::logging;
-
-//see set()
-const fn max_upload_file_limit() -> usize {
-    100 * 1024 * 1024
-}
-
-#[derive(serde::Deserialize)]
-#[serde(tag = "type")]
-struct Config {
-    listen: std::net::SocketAddr,
-    pemfile: camino::Utf8PathBuf,
-    #[serde(flatten)]
-    storage_config: remote_storage::RemoteStorageConfig,
-    #[serde(default = "max_upload_file_limit")]
-    max_upload_file_limit: usize,
-}
-
-#[tokio::main]
-async fn main() -> anyhow::Result<()> {
-    logging::init(
-        logging::LogFormat::Plain,
-        logging::TracingErrorLayerEnablement::EnableWithRustLogFilter,
-        logging::Output::Stdout,
-    )?;
-
-    let config: String = std::env::args().skip(1).take(1).collect();
-    if config.is_empty() {
-        anyhow::bail!("Usage: object_storage config.json")
-    }
-    info!("Reading config from {config}");
-    let config = std::fs::read_to_string(config.clone())?;
-    let config: Config = serde_json::from_str(&config).context("parsing config")?;
-    info!("Reading pemfile from {}", config.pemfile.clone());
-    let pemfile = std::fs::read(config.pemfile.clone())?;
-    info!("Loading public key from {}", config.pemfile.clone());
-    let auth = object_storage::JwtAuth::new(&pemfile)?;
-
-    let listener = tokio::net::TcpListener::bind(config.listen).await.unwrap();
-    info!("listening on {}", listener.local_addr().unwrap());
-
-    let storage = remote_storage::GenericRemoteStorage::from_config(&config.storage_config).await?;
-    let cancel = tokio_util::sync::CancellationToken::new();
-    app::check_storage_permissions(&storage, cancel.clone()).await?;
-
-    let proxy = std::sync::Arc::new(object_storage::Storage {
-        auth,
-        storage,
-        cancel: cancel.clone(),
-        max_upload_file_limit: config.max_upload_file_limit,
-    });
-
-    tokio::spawn(utils::signals::signal_handler(cancel.clone()));
-    axum::serve(listener, app::app(proxy))
-        .with_graceful_shutdown(async move { cancel.cancelled().await })
-        .await?;
-    Ok(())
-}
--- a/pageserver/client/src/mgmt_api.rs
+++ b/pageserver/client/src/mgmt_api.rs
@@ -86,17 +86,17 @@ impl Client {
        resp.json().await.map_err(Error::ReceiveBody)
    }

-    /// Send an HTTP request to an arbitrary path with a desired HTTP method and returning a streaming
-    /// Response.  This function is suitable for pass-through/proxy use cases where we don't care
-    /// what the response content looks like.
+    /// Get an arbitrary path and returning a streaming Response.  This function is suitable
+    /// for pass-through/proxy use cases where we don't care what the response content looks
+    /// like.
    ///
    /// Use/add one of the properly typed methods below if you know aren't proxying, and
    /// know what kind of response you expect.
-    pub async fn op_raw(&self, method: Method, path: String) -> Result<reqwest::Response> {
+    pub async fn get_raw(&self, path: String) -> Result<reqwest::Response> {
        debug_assert!(path.starts_with('/'));
        let uri = format!("{}{}", self.mgmt_api_endpoint, path);

-        let mut req = self.client.request(method, uri);
+        let mut req = self.client.request(Method::GET, uri);
        if let Some(value) = &self.authorization_header {
            req = req.header(reqwest::header::AUTHORIZATION, value);
        }
--- a/pageserver/src/bin/pageserver.rs
+++ b/pageserver/src/bin/pageserver.rs
@@ -16,7 +16,7 @@ use http_utils::tls_certs::ReloadingCertificateResolver;
 use metrics::launch_timestamp::{LaunchTimestamp, set_launch_timestamp_metric};
 use metrics::set_build_info_metric;
 use nix::sys::socket::{setsockopt, sockopt};
-use pageserver::config::{PageServerConf, PageserverIdentity, ignored_fields};
+use pageserver::config::{PageServerConf, PageserverIdentity};
 use pageserver::controller_upcall_client::StorageControllerUpcallClient;
 use pageserver::deletion_queue::DeletionQueue;
 use pageserver::disk_usage_eviction_task::{self, launch_disk_usage_global_eviction_task};
@@ -31,10 +31,10 @@ use pageserver::{
 };
 use postgres_backend::AuthType;
 use remote_storage::GenericRemoteStorage;
+use tokio::signal::unix::SignalKind;
 use tokio::time::Instant;
 use tokio_util::sync::CancellationToken;
 use tracing::*;
-use tracing_utils::OtelGuard;
 use utils::auth::{JwtAuth, SwappableJwtAuth};
 use utils::crashsafe::syncfs;
 use utils::logging::TracingErrorLayerEnablement;
@@ -97,7 +97,7 @@ fn main() -> anyhow::Result<()> {
    env::set_current_dir(&workdir)
        .with_context(|| format!("Failed to set application's current dir to '{workdir}'"))?;

-    let (conf, ignored) = initialize_config(&identity_file_path, &cfg_file_path, &workdir)?;
+    let conf = initialize_config(&identity_file_path, &cfg_file_path, &workdir)?;

    // Initialize logging.
    //
@@ -118,21 +118,6 @@ fn main() -> anyhow::Result<()> {
        logging::Output::Stdout,
    )?;

-    let otel_enablement = match &conf.tracing {
-        Some(cfg) => tracing_utils::OtelEnablement::Enabled {
-            service_name: "pageserver".to_string(),
-            export_config: (&cfg.export_config).into(),
-            runtime: *COMPUTE_REQUEST_RUNTIME,
-        },
-        None => tracing_utils::OtelEnablement::Disabled,
-    };
-
-    let otel_guard = tracing_utils::init_performance_tracing(otel_enablement);
-
-    if otel_guard.is_some() {
-        info!(?conf.tracing, "starting with OTEL tracing enabled");
-    }
-
    // mind the order required here: 1. logging, 2. panic_hook, 3. sentry.
    // disarming this hook on pageserver, because we never tear down tracing.
    logging::replace_panic_hook_with_tracing_panic_hook().forget();
@@ -143,17 +128,7 @@ fn main() -> anyhow::Result<()> {
        &[("node_id", &conf.id.to_string())],
    );

-    // Warn about ignored config items; see pageserver_api::config::ConfigToml
-    // doc comment for rationale why we prefer this over serde(deny_unknown_fields).
-    {
-        let ignored_fields::Paths { paths } = &ignored;
-        for path in paths {
-            warn!(?path, "ignoring unknown configuration item");
-        }
-    }
-
-    // Log configuration items for feature-flag-like config
-    // (maybe we should automate this with a visitor?).
+    // after setting up logging, log the effective IO engine choice and read path implementations
    info!(?conf.virtual_file_io_engine, "starting with virtual_file IO engine");
    info!(?conf.virtual_file_io_mode, "starting with virtual_file IO mode");
    info!(?conf.wal_receiver_protocol, "starting with WAL receiver protocol");
@@ -216,7 +191,7 @@ fn main() -> anyhow::Result<()> {
    tracing::info!("Initializing page_cache...");
    page_cache::init(conf.page_cache_size);

-    start_pageserver(launch_ts, conf, ignored, otel_guard).context("Failed to start pageserver")?;
+    start_pageserver(launch_ts, conf).context("Failed to start pageserver")?;

    scenario.teardown();
    Ok(())
@@ -226,7 +201,7 @@ fn initialize_config(
    identity_file_path: &Utf8Path,
    cfg_file_path: &Utf8Path,
    workdir: &Utf8Path,
-) -> anyhow::Result<(&'static PageServerConf, ignored_fields::Paths)> {
+) -> anyhow::Result<&'static PageServerConf> {
    // The deployment orchestrator writes out an indentity file containing the node id
    // for all pageservers. This file is the source of truth for the node id. In order
    // to allow for rolling back pageserver releases, the node id is also included in
@@ -255,36 +230,16 @@ fn initialize_config(

    let config_file_contents =
        std::fs::read_to_string(cfg_file_path).context("read config file from filesystem")?;
+    let config_toml = serde_path_to_error::deserialize(
+        toml_edit::de::Deserializer::from_str(&config_file_contents)
+            .context("build toml deserializer")?,
+    )
+    .context("deserialize config toml")?;

-    // Deserialize the config file contents into a ConfigToml.
-    let config_toml: pageserver_api::config::ConfigToml = {
-        let deserializer = toml_edit::de::Deserializer::from_str(&config_file_contents)
-            .context("build toml deserializer")?;
-        let mut path_to_error_track = serde_path_to_error::Track::new();
-        let deserializer =
-            serde_path_to_error::Deserializer::new(deserializer, &mut path_to_error_track);
-        serde::Deserialize::deserialize(deserializer).context("deserialize config toml")?
-    };
-
-    // Find unknown fields by re-serializing the parsed ConfigToml and comparing it to the on-disk file.
-    // Any fields that are only in the on-disk version are unknown.
-    // (The assumption here is that the ConfigToml doesn't to skip_serializing_if.)
-    // (Make sure to read the ConfigToml doc comment on why we only want to warn about, but not fail startup, on unknown fields).
-    let ignored = {
-        let ondisk_toml = config_file_contents
-            .parse::<toml_edit::DocumentMut>()
-            .context("parse original config as toml document")?;
-        let parsed_toml = toml_edit::ser::to_document(&config_toml)
-            .context("re-serialize config to toml document")?;
-        pageserver::config::ignored_fields::find(ondisk_toml, parsed_toml)
-    };
-
-    // Construct the runtime god object (it's called PageServerConf but actually is just global shared state).
    let conf = PageServerConf::parse_and_validate(identity.id, config_toml, workdir)
        .context("runtime-validation of config toml")?;
-    let conf = Box::leak(Box::new(conf));

-    Ok((conf, ignored))
+    Ok(Box::leak(Box::new(conf)))
 }

 struct WaitForPhaseResult<F: std::future::Future + Unpin> {
@@ -335,8 +290,6 @@ fn startup_checkpoint(started_at: Instant, phase: &str, human_phase: &str) {
 fn start_pageserver(
    launch_ts: &'static LaunchTimestamp,
    conf: &'static PageServerConf,
-    ignored: ignored_fields::Paths,
-    otel_guard: Option<OtelGuard>,
 ) -> anyhow::Result<()> {
    // Monotonic time for later calculating startup duration
    let started_startup_at = Instant::now();
@@ -359,7 +312,7 @@ fn start_pageserver(
        pageserver::metrics::tokio_epoll_uring::Collector::new(),
    ))
    .unwrap();
-    pageserver::preinitialize_metrics(conf, ignored);
+    pageserver::preinitialize_metrics(conf);

    // If any failpoints were set from FAILPOINTS environment variable,
    // print them to the log for debugging purposes
@@ -722,28 +675,45 @@ fn start_pageserver(

    // Spawn a task to listen for libpq connections. It will spawn further tasks
    // for each connection. We created the listener earlier already.
-    let perf_trace_dispatch = otel_guard.as_ref().map(|g| g.dispatch.clone());
-    let page_service = page_service::spawn(
-        conf,
-        tenant_manager.clone(),
-        pg_auth,
-        perf_trace_dispatch,
-        {
-            let _entered = COMPUTE_REQUEST_RUNTIME.enter(); // TcpListener::from_std requires it
-            pageserver_listener
-                .set_nonblocking(true)
-                .context("set listener to nonblocking")?;
-            tokio::net::TcpListener::from_std(pageserver_listener)
-                .context("create tokio listener")?
-        },
-    );
+    let page_service = page_service::spawn(conf, tenant_manager.clone(), pg_auth, {
+        let _entered = COMPUTE_REQUEST_RUNTIME.enter(); // TcpListener::from_std requires it
+        pageserver_listener
+            .set_nonblocking(true)
+            .context("set listener to nonblocking")?;
+        tokio::net::TcpListener::from_std(pageserver_listener).context("create tokio listener")?
+    });

    // All started up! Now just sit and wait for shutdown signal.
    BACKGROUND_RUNTIME.block_on(async move {
        let signal_token = CancellationToken::new();
        let signal_cancel = signal_token.child_token();

-        tokio::spawn(utils::signals::signal_handler(signal_token));
+        // Spawn signal handlers. Runs in a loop since we want to be responsive to multiple signals
+        // even after triggering shutdown (e.g. a SIGQUIT after a slow SIGTERM shutdown). See:
+        // https://github.com/neondatabase/neon/issues/9740.
+        tokio::spawn(async move {
+            let mut sigint = tokio::signal::unix::signal(SignalKind::interrupt()).unwrap();
+            let mut sigterm = tokio::signal::unix::signal(SignalKind::terminate()).unwrap();
+            let mut sigquit = tokio::signal::unix::signal(SignalKind::quit()).unwrap();
+
+            loop {
+                let signal = tokio::select! {
+                    _ = sigquit.recv() => {
+                        info!("Got signal SIGQUIT. Terminating in immediate shutdown mode.");
+                        std::process::exit(111);
+                    }
+                    _ = sigint.recv() => "SIGINT",
+                    _ = sigterm.recv() => "SIGTERM",
+                };
+
+                if !signal_token.is_cancelled() {
+                    info!("Got signal {signal}. Terminating gracefully in fast shutdown mode.");
+                    signal_token.cancel();
+                } else {
+                    info!("Got signal {signal}. Already shutting down.");
+                }
+            }
+        });

        // Wait for cancellation signal and shut down the pageserver.
        //
--- a/pageserver/src/config.rs
+++ b/pageserver/src/config.rs
@@ -4,8 +4,6 @@
 //! file, or on the command line.
 //! See also `settings.md` for better description on every parameter.

-pub mod ignored_fields;
-
 use std::env;
 use std::num::NonZeroUsize;
 use std::sync::Arc;
@@ -217,8 +215,6 @@ pub struct PageServerConf {

    /// When set, include visible layers in the next uploaded heatmaps of an unarchived timeline.
    pub generate_unarchival_heatmap: bool,
-
-    pub tracing: Option<pageserver_api::config::Tracing>,
 }

 /// Token for authentication to safekeepers
@@ -390,7 +386,6 @@ impl PageServerConf {
            validate_wal_contiguity,
            load_previous_heatmap,
            generate_unarchival_heatmap,
-            tracing,
        } = config_toml;

        let mut conf = PageServerConf {
@@ -440,7 +435,6 @@ impl PageServerConf {
            wal_receiver_protocol,
            page_service_pipelining,
            get_vectored_concurrent_io,
-            tracing,

            // ------------------------------------------------------------
            // fields that require additional validation or custom handling
@@ -512,17 +506,6 @@ impl PageServerConf {
            );
        }

-        if let Some(tracing_config) = conf.tracing.as_ref() {
-            let ratio = &tracing_config.sampling_ratio;
-            ensure!(
-                ratio.denominator != 0 && ratio.denominator >= ratio.numerator,
-                format!(
-                    "Invalid sampling ratio: {}/{}",
-                    ratio.numerator, ratio.denominator
-                )
-            );
-        }
-
        IndexEntry::validate_checkpoint_distance(conf.default_tenant_conf.checkpoint_distance)
            .map_err(anyhow::Error::msg)
            .with_context(|| {
@@ -562,6 +545,7 @@ impl PageServerConf {
 }

 #[derive(serde::Deserialize, serde::Serialize)]
+#[serde(deny_unknown_fields)]
 pub struct PageserverIdentity {
    pub id: NodeId,
 }
@@ -633,4 +617,82 @@ mod tests {
        PageServerConf::parse_and_validate(NodeId(0), config_toml, &workdir)
            .expect("parse_and_validate");
    }
+
+    /// If there's a typo in the pageserver config, we'd rather catch that typo
+    /// and fail pageserver startup than silently ignoring the typo, leaving whoever
+    /// made it in the believe that their config change is effective.
+    ///
+    /// The default in serde is to allow unknown fields, so, we rely
+    /// on developer+review discipline to add `deny_unknown_fields` when adding
+    /// new structs to the config, and these tests here as a regression test.
+    ///
+    /// The alternative to all of this would be to allow unknown fields in the config.
+    /// To catch them, we could have a config check tool or mgmt API endpoint that
+    /// compares the effective config with the TOML on disk and makes sure that
+    /// the on-disk TOML is a strict subset of the effective config.
+    mod unknown_fields_handling {
+        macro_rules! test {
+            ($short_name:ident, $input:expr) => {
+                #[test]
+                fn $short_name() {
+                    let input = $input;
+                    let err = toml_edit::de::from_str::<pageserver_api::config::ConfigToml>(&input)
+                        .expect_err("some_invalid_field is an invalid field");
+                    dbg!(&err);
+                    assert!(err.to_string().contains("some_invalid_field"));
+                }
+            };
+        }
+        use indoc::indoc;
+
+        test!(
+            toplevel,
+            indoc! {r#"
+                some_invalid_field = 23
+            "#}
+        );
+
+        test!(
+            toplevel_nested,
+            indoc! {r#"
+                [some_invalid_field]
+                foo = 23
+            "#}
+        );
+
+        test!(
+            disk_usage_based_eviction,
+            indoc! {r#"
+                [disk_usage_based_eviction]
+                some_invalid_field = 23
+            "#}
+        );
+
+        test!(
+            tenant_config,
+            indoc! {r#"
+                [tenant_config]
+                some_invalid_field = 23
+            "#}
+        );
+
+        test!(
+            l0_flush,
+            indoc! {r#"
+                [l0_flush]
+                mode = "direct"
+                some_invalid_field = 23
+            "#}
+        );
+
+        // TODO: fix this => https://github.com/neondatabase/neon/issues/8915
+        // test!(
+        //     remote_storage_config,
+        //     indoc! {r#"
+        //         [remote_storage_config]
+        //         local_path = "/nonexistent"
+        //         some_invalid_field = 23
+        //     "#}
+        // );
+    }
 }
--- a/pageserver/src/config/ignored_fields.rs
+++ b/pageserver/src/config/ignored_fields.rs
@@ -1,179 +0,0 @@
-//! Check for fields in the on-disk config file that were ignored when
-//! deserializing [`pageserver_api::config::ConfigToml`].
-//!
-//! This could have been part of the [`pageserver_api::config`] module,
-//! but the way we identify unused fields in this module
-//! is specific to the format (TOML) and the implementation of the
-//! deserialization for that format ([`toml_edit`]).
-
-use std::collections::HashSet;
-
-use itertools::Itertools;
-
-/// Pass in the user-specified config and the re-serialized [`pageserver_api::config::ConfigToml`].
-/// The returned [`Paths`] contains the paths to the fields that were ignored by deserialization
-/// of the [`pageserver_api::config::ConfigToml`].
-pub fn find(user_specified: toml_edit::DocumentMut, reserialized: toml_edit::DocumentMut) -> Paths {
-    let user_specified = paths(user_specified);
-    let reserialized = paths(reserialized);
-    fn paths(doc: toml_edit::DocumentMut) -> HashSet<String> {
-        let mut out = Vec::new();
-        let mut visitor = PathsVisitor::new(&mut out);
-        visitor.visit_table_like(doc.as_table());
-        HashSet::from_iter(out)
-    }
-
-    let mut ignored = HashSet::new();
-
-    // O(n) because of HashSet
-    for path in user_specified {
-        if !reserialized.contains(&path) {
-            ignored.insert(path);
-        }
-    }
-
-    Paths {
-        paths: ignored
-            .into_iter()
-            // sort lexicographically for deterministic output
-            .sorted()
-            .collect(),
-    }
-}
-
-pub struct Paths {
-    pub paths: Vec<String>,
-}
-
-struct PathsVisitor<'a> {
-    stack: Vec<String>,
-    out: &'a mut Vec<String>,
-}
-
-impl<'a> PathsVisitor<'a> {
-    fn new(out: &'a mut Vec<String>) -> Self {
-        Self {
-            stack: Vec::new(),
-            out,
-        }
-    }
-
-    fn visit_table_like(&mut self, table_like: &dyn toml_edit::TableLike) {
-        for (entry, item) in table_like.iter() {
-            self.stack.push(entry.to_string());
-            self.visit_item(item);
-            self.stack.pop();
-        }
-    }
-
-    fn visit_item(&mut self, item: &toml_edit::Item) {
-        match item {
-            toml_edit::Item::None => (),
-            toml_edit::Item::Value(value) => self.visit_value(value),
-            toml_edit::Item::Table(table) => {
-                self.visit_table_like(table);
-            }
-            toml_edit::Item::ArrayOfTables(array_of_tables) => {
-                for (i, table) in array_of_tables.iter().enumerate() {
-                    self.stack.push(format!("[{i}]"));
-                    self.visit_table_like(table);
-                    self.stack.pop();
-                }
-            }
-        }
-    }
-
-    fn visit_value(&mut self, value: &toml_edit::Value) {
-        match value {
-            toml_edit::Value::String(_)
-            | toml_edit::Value::Integer(_)
-            | toml_edit::Value::Float(_)
-            | toml_edit::Value::Boolean(_)
-            | toml_edit::Value::Datetime(_) => self.out.push(self.stack.join(".")),
-            toml_edit::Value::Array(array) => {
-                for (i, value) in array.iter().enumerate() {
-                    self.stack.push(format!("[{i}]"));
-                    self.visit_value(value);
-                    self.stack.pop();
-                }
-            }
-            toml_edit::Value::InlineTable(inline_table) => self.visit_table_like(inline_table),
-        }
-    }
-}
-
-#[cfg(test)]
-pub(crate) mod tests {
-
-    fn test_impl(original: &str, parsed: &str, expect: [&str; 1]) {
-        let original: toml_edit::DocumentMut = original.parse().expect("parse original config");
-        let parsed: toml_edit::DocumentMut = parsed.parse().expect("parse re-serialized config");
-
-        let super::Paths { paths: actual } = super::find(original, parsed);
-        assert_eq!(actual, &expect);
-    }
-
-    #[test]
-    fn top_level() {
-        test_impl(
-            r#"
-                [a]
-                b = 1
-                c = 2
-                d = 3
-            "#,
-            r#"
-                [a]
-                b = 1
-                c = 2
-            "#,
-            ["a.d"],
-        );
-    }
-
-    #[test]
-    fn nested() {
-        test_impl(
-            r#"
-                [a.b.c]
-                d = 23
-            "#,
-            r#"
-                [a]
-                e = 42
-            "#,
-            ["a.b.c.d"],
-        );
-    }
-
-    #[test]
-    fn array_of_tables() {
-        test_impl(
-            r#"
-                [[a]]
-                b = 1
-                c = 2
-                d = 3
-            "#,
-            r#"
-                [[a]]
-                b = 1
-                c = 2
-            "#,
-            ["a.[0].d"],
-        );
-    }
-
-    #[test]
-    fn array() {
-        test_impl(
-            r#"
-            foo = [ {bar = 23} ]
-            "#,
-            r#"
-            foo = [ { blup = 42 }]
-            "#,
-            ["foo.[0].bar"],
-        );
-    }
-}
--- a/pageserver/src/context.rs
+++ b/pageserver/src/context.rs
@@ -89,7 +89,7 @@
 //! [`RequestContext`] argument. Functions in the middle of the call chain
 //! only need to pass it on.

-use std::{sync::Arc, time::Duration};
+use std::sync::Arc;

 use once_cell::sync::Lazy;
 use tracing::warn;
@@ -100,12 +100,6 @@ use crate::{
    task_mgr::TaskKind,
    tenant::Timeline,
 };
-use futures::FutureExt;
-use futures::future::BoxFuture;
-use std::future::Future;
-use tracing_utils::perf_span::{PerfInstrument, PerfSpan};
-
-use tracing::{Dispatch, Span};

 // The main structure of this module, see module-level comment.
 pub struct RequestContext {
@@ -115,8 +109,6 @@ pub struct RequestContext {
    page_content_kind: PageContentKind,
    read_path_debug: bool,
    scope: Scope,
-    perf_span: Option<PerfSpan>,
-    perf_span_dispatch: Option<Dispatch>,
 }

 #[derive(Clone)]
@@ -271,15 +263,22 @@ impl RequestContextBuilder {
                page_content_kind: PageContentKind::Unknown,
                read_path_debug: false,
                scope: Scope::new_global(),
-                perf_span: None,
-                perf_span_dispatch: None,
            },
        }
    }

-    pub fn from(original: &RequestContext) -> Self {
+    pub fn extend(original: &RequestContext) -> Self {
        Self {
-            inner: original.clone(),
+            // This is like a Copy, but avoid implementing Copy because ordinary users of
+            // RequestContext should always move or ref it.
+            inner: RequestContext {
+                task_kind: original.task_kind,
+                download_behavior: original.download_behavior,
+                access_stats_behavior: original.access_stats_behavior,
+                page_content_kind: original.page_content_kind,
+                read_path_debug: original.read_path_debug,
+                scope: original.scope.clone(),
+            },
        }
    }

@@ -317,74 +316,12 @@ impl RequestContextBuilder {
        self
    }

-    pub(crate) fn perf_span_dispatch(mut self, dispatch: Option<Dispatch>) -> Self {
-        self.inner.perf_span_dispatch = dispatch;
-        self
-    }
-
-    pub fn root_perf_span<Fn>(mut self, make_span: Fn) -> Self
-    where
-        Fn: FnOnce() -> Span,
-    {
-        assert!(self.inner.perf_span.is_none());
-        assert!(self.inner.perf_span_dispatch.is_some());
-
-        let dispatcher = self.inner.perf_span_dispatch.as_ref().unwrap();
-        let new_span = tracing::dispatcher::with_default(dispatcher, make_span);
-
-        self.inner.perf_span = Some(PerfSpan::new(new_span, dispatcher.clone()));
-
-        self
-    }
-
-    pub fn perf_span<Fn>(mut self, make_span: Fn) -> Self
-    where
-        Fn: FnOnce(&Span) -> Span,
-    {
-        if let Some(ref perf_span) = self.inner.perf_span {
-            assert!(self.inner.perf_span_dispatch.is_some());
-            let dispatcher = self.inner.perf_span_dispatch.as_ref().unwrap();
-
-            let new_span =
-                tracing::dispatcher::with_default(dispatcher, || make_span(perf_span.inner()));
-
-            self.inner.perf_span = Some(PerfSpan::new(new_span, dispatcher.clone()));
-        }
-
-        self
-    }
-
-    pub fn root(self) -> RequestContext {
-        self.inner
-    }
-
-    pub fn attached_child(self) -> RequestContext {
-        self.inner
-    }
-
-    pub fn detached_child(self) -> RequestContext {
+    pub fn build(self) -> RequestContext {
        self.inner
    }
 }

 impl RequestContext {
-    /// Private clone implementation
-    ///
-    /// Callers should use the [`RequestContextBuilder`] or child spaning APIs of
-    /// [`RequestContext`].
-    fn clone(&self) -> Self {
-        Self {
-            task_kind: self.task_kind,
-            download_behavior: self.download_behavior,
-            access_stats_behavior: self.access_stats_behavior,
-            page_content_kind: self.page_content_kind,
-            read_path_debug: self.read_path_debug,
-            scope: self.scope.clone(),
-            perf_span: self.perf_span.clone(),
-            perf_span_dispatch: self.perf_span_dispatch.clone(),
-        }
-    }
-
    /// Create a new RequestContext that has no parent.
    ///
    /// The function is called `new` because, once we add children
@@ -400,7 +337,7 @@ impl RequestContext {
    pub fn new(task_kind: TaskKind, download_behavior: DownloadBehavior) -> Self {
        RequestContextBuilder::new(task_kind)
            .download_behavior(download_behavior)
-            .root()
+            .build()
    }

    /// Create a detached child context for a task that may outlive `self`.
@@ -421,10 +358,7 @@ impl RequestContext {
    ///
    /// We could make new calls to this function fail if `self` is already canceled.
    pub fn detached_child(&self, task_kind: TaskKind, download_behavior: DownloadBehavior) -> Self {
-        RequestContextBuilder::from(self)
-            .task_kind(task_kind)
-            .download_behavior(download_behavior)
-            .detached_child()
+        self.child_impl(task_kind, download_behavior)
    }

    /// Create a child of context `self` for a task that shall not outlive `self`.
@@ -448,7 +382,7 @@ impl RequestContext {
    /// The method to wait for child tasks would return an error, indicating
    /// that the child task was not started because the context was canceled.
    pub fn attached_child(&self) -> Self {
-        RequestContextBuilder::from(self).attached_child()
+        self.child_impl(self.task_kind(), self.download_behavior())
    }

    /// Use this function when you should be creating a child context using
@@ -463,10 +397,17 @@ impl RequestContext {
        Self::new(task_kind, download_behavior)
    }

+    fn child_impl(&self, task_kind: TaskKind, download_behavior: DownloadBehavior) -> Self {
+        RequestContextBuilder::extend(self)
+            .task_kind(task_kind)
+            .download_behavior(download_behavior)
+            .build()
+    }
+
    pub fn with_scope_timeline(&self, timeline: &Arc<Timeline>) -> Self {
-        RequestContextBuilder::from(self)
+        RequestContextBuilder::extend(self)
            .scope(Scope::new_timeline(timeline))
-            .attached_child()
+            .build()
    }

    pub(crate) fn with_scope_page_service_pagestream(
@@ -475,9 +416,9 @@ impl RequestContext {
            crate::page_service::TenantManagerTypes,
        >,
    ) -> Self {
-        RequestContextBuilder::from(self)
+        RequestContextBuilder::extend(self)
            .scope(Scope::new_page_service_pagestream(timeline_handle))
-            .attached_child()
+            .build()
    }

    pub fn with_scope_secondary_timeline(
@@ -485,30 +426,28 @@ impl RequestContext {
        tenant_shard_id: &TenantShardId,
        timeline_id: &TimelineId,
    ) -> Self {
-        RequestContextBuilder::from(self)
+        RequestContextBuilder::extend(self)
            .scope(Scope::new_secondary_timeline(tenant_shard_id, timeline_id))
-            .attached_child()
+            .build()
    }

    pub fn with_scope_secondary_tenant(&self, tenant_shard_id: &TenantShardId) -> Self {
-        RequestContextBuilder::from(self)
+        RequestContextBuilder::extend(self)
            .scope(Scope::new_secondary_tenant(tenant_shard_id))
-            .attached_child()
+            .build()
    }

    #[cfg(test)]
    pub fn with_scope_unit_test(&self) -> Self {
-        RequestContextBuilder::from(self)
-            .task_kind(TaskKind::UnitTest)
+        RequestContextBuilder::new(TaskKind::UnitTest)
            .scope(Scope::new_unit_test())
-            .attached_child()
+            .build()
    }

    pub fn with_scope_debug_tools(&self) -> Self {
-        RequestContextBuilder::from(self)
-            .task_kind(TaskKind::DebugTool)
+        RequestContextBuilder::new(TaskKind::DebugTool)
            .scope(Scope::new_debug_tools())
-            .attached_child()
+            .build()
    }

    pub fn task_kind(&self) -> TaskKind {
@@ -565,76 +504,4 @@ impl RequestContext {
            Scope::DebugTools { io_size_metrics } => io_size_metrics,
        }
    }
-
-    pub(crate) fn ondemand_download_wait_observe(&self, duration: Duration) {
-        if duration == Duration::ZERO {
-            return;
-        }
-
-        match &self.scope {
-            Scope::Timeline { arc_arc } => arc_arc
-                .wait_ondemand_download_time
-                .observe(self.task_kind, duration),
-            _ => {
-                use once_cell::sync::Lazy;
-                use std::sync::Mutex;
-                use std::time::Duration;
-                use utils::rate_limit::RateLimit;
-                static LIMIT: Lazy<Mutex<RateLimit>> =
-                    Lazy::new(|| Mutex::new(RateLimit::new(Duration::from_secs(1))));
-                let mut guard = LIMIT.lock().unwrap();
-                guard.call2(|rate_limit_stats| {
-                    warn!(
-                        %rate_limit_stats,
-                        backtrace=%std::backtrace::Backtrace::force_capture(),
-                        "ondemand downloads should always happen within timeline scope",
-                    );
-                });
-            }
-        }
-    }
-
-    pub(crate) fn perf_follows_from(&self, from: &RequestContext) {
-        if let (Some(span), Some(from_span)) = (&self.perf_span, &from.perf_span) {
-            span.inner().follows_from(from_span.inner());
-        }
-    }
-
-    pub(crate) fn has_perf_span(&self) -> bool {
-        self.perf_span.is_some()
-    }
 }
-
-/// [`Future`] extension trait that allow for creating performance
-/// spans on sampled requests
-pub(crate) trait PerfInstrumentFutureExt<'a>: Future + Send {
-    /// Instrument this future with a new performance span when the
-    /// provided request context indicates the originator request
-    /// was sampled. Otherwise, just box the future and return it as is.
-    fn maybe_perf_instrument<Fn>(
-        self,
-        ctx: &RequestContext,
-        make_span: Fn,
-    ) -> BoxFuture<'a, Self::Output>
-    where
-        Self: Sized + 'a,
-        Fn: FnOnce(&Span) -> Span,
-    {
-        match &ctx.perf_span {
-            Some(perf_span) => {
-                assert!(ctx.perf_span_dispatch.is_some());
-                let dispatcher = ctx.perf_span_dispatch.as_ref().unwrap();
-
-                let new_span =
-                    tracing::dispatcher::with_default(dispatcher, || make_span(perf_span.inner()));
-
-                let new_perf_span = PerfSpan::new(new_span, dispatcher.clone());
-                self.instrument(new_perf_span).boxed()
-            }
-            None => self.boxed(),
-        }
-    }
-}
-
-// Implement the trait for all types that satisfy the trait bounds
-impl<'a, T: Future + Send + 'a> PerfInstrumentFutureExt<'a> for T {}
--- a/pageserver/src/http/openapi_spec.yml
+++ b/pageserver/src/http/openapi_spec.yml
@@ -212,12 +212,6 @@ paths:
              schema:
                type: string
                format: date-time
-        "412":
-          description: No timestamp is found for given LSN, e.g. if there had been no commits till LSN
-          content:
-            application/json:
-              schema:
-                $ref: "#/components/schemas/PreconditionFailedError"

  /v1/tenant/{tenant_id}/timeline/{timeline_id}/get_lsn_by_timestamp:
    parameters:
--- a/pageserver/src/http/routes.rs
+++ b/pageserver/src/http/routes.rs
@@ -67,15 +67,15 @@ use crate::tenant::mgr::{
 };
 use crate::tenant::remote_timeline_client::index::GcCompactionState;
 use crate::tenant::remote_timeline_client::{
-    download_index_part, download_tenant_manifest, list_remote_tenant_shards, list_remote_timelines,
+    download_index_part, list_remote_tenant_shards, list_remote_timelines,
 };
 use crate::tenant::secondary::SecondaryController;
 use crate::tenant::size::ModelInputs;
 use crate::tenant::storage_layer::{IoConcurrency, LayerAccessStatsReset, LayerName};
 use crate::tenant::timeline::offload::{OffloadError, offload_timeline};
 use crate::tenant::timeline::{
-    CompactFlags, CompactOptions, CompactRequest, CompactionError, MarkInvisibleRequest, Timeline,
-    WaitLsnTimeout, WaitLsnWaiter, import_pgdata,
+    CompactFlags, CompactOptions, CompactRequest, CompactionError, Timeline, WaitLsnTimeout,
+    WaitLsnWaiter, import_pgdata,
 };
 use crate::tenant::{
    GetTimelineError, LogicalSizeCalculationCause, OffloadedTimeline, PageReconstructError,
@@ -445,9 +445,6 @@ async fn build_timeline_info_common(

    let (pitr_history_size, within_ancestor_pitr) = timeline.get_pitr_history_stats();

-    // Externally, expose the lowest LSN that can be used to create a branch.
-    // Internally we distinguish between the planned GC cutoff (PITR point) and the "applied" GC cutoff (where we
-    // actually trimmed data to), which can pass each other when PITR is changed.
    let min_readable_lsn = std::cmp::max(
        timeline.get_gc_cutoff_lsn(),
        *timeline.get_applied_gc_cutoff_lsn(),
@@ -464,6 +461,7 @@ async fn build_timeline_info_common(
        initdb_lsn,
        last_record_lsn,
        prev_record_lsn: Some(timeline.get_prev_record_lsn()),
+        _unused: Default::default(), // Unused, for legacy decode only
        min_readable_lsn,
        applied_gc_cutoff_lsn: *timeline.get_applied_gc_cutoff_lsn(),
        current_logical_size: current_logical_size.size_dont_care_about_accuracy(),
@@ -989,7 +987,7 @@ async fn get_lsn_by_timestamp_handler(
    if !tenant_shard_id.is_shard_zero() {
        // Requires SLRU contents, which are only stored on shard zero
        return Err(ApiError::BadRequest(anyhow!(
-            "Lsn calculations by timestamp are only available on shard zero"
+            "Size calculations are only available on shard zero"
        )));
    }

@@ -1064,7 +1062,7 @@ async fn get_timestamp_of_lsn_handler(
    if !tenant_shard_id.is_shard_zero() {
        // Requires SLRU contents, which are only stored on shard zero
        return Err(ApiError::BadRequest(anyhow!(
-            "Timestamp calculations by lsn are only available on shard zero"
+            "Size calculations are only available on shard zero"
        )));
    }

@@ -1090,8 +1088,8 @@ async fn get_timestamp_of_lsn_handler(
            .to_string();
            json_response(StatusCode::OK, time)
        }
-        None => Err(ApiError::PreconditionFailed(
-            format!("Timestamp for lsn {} not found", lsn).into(),
+        None => Err(ApiError::NotFound(
+            anyhow::anyhow!("Timestamp for lsn {} not found", lsn).into(),
        )),
    }
 }
@@ -2337,31 +2335,21 @@ async fn timeline_compact_handler(
 }

 async fn timeline_mark_invisible_handler(
-    mut request: Request<Body>,
+    request: Request<Body>,
    _cancel: CancellationToken,
 ) -> Result<Response<Body>, ApiError> {
    let tenant_shard_id: TenantShardId = parse_request_param(&request, "tenant_shard_id")?;
    let timeline_id: TimelineId = parse_request_param(&request, "timeline_id")?;
    check_permission(&request, Some(tenant_shard_id.tenant_id))?;

-    let compact_request = json_request_maybe::<Option<MarkInvisibleRequest>>(&mut request).await?;
-
    let state = get_state(&request);

-    let visibility = match compact_request {
-        Some(req) => match req.is_visible {
-            Some(true) => TimelineVisibilityState::Visible,
-            Some(false) | None => TimelineVisibilityState::Invisible,
-        },
-        None => TimelineVisibilityState::Invisible,
-    };
-
    async {
        let tenant = state
            .tenant_manager
            .get_attached_tenant_shard(tenant_shard_id)?;
        let timeline = tenant.get_timeline(timeline_id, true)?;
-        timeline.remote_client.schedule_index_upload_for_timeline_invisible_state(visibility).map_err(ApiError::InternalServerError)?;
+        timeline.remote_client.schedule_index_upload_for_timeline_invisible_state(TimelineVisibilityState::Invisible).map_err(ApiError::InternalServerError)?;
        json_response(StatusCode::OK, ())
    }
    .instrument(info_span!("manual_timeline_mark_invisible", tenant_id = %tenant_shard_id.tenant_id, shard_id = %tenant_shard_id.shard_slug(), %timeline_id))
@@ -2697,12 +2685,11 @@ async fn getpage_at_lsn_handler_inner(
    let lsn: Option<Lsn> = parse_query_param(&request, "lsn")?;

    async {
+        let ctx = RequestContext::new(TaskKind::MgmtRequest, DownloadBehavior::Download);
+        // Enable read path debugging
        let timeline = active_timeline_of_active_tenant(&state.tenant_manager, tenant_shard_id, timeline_id).await?;
-        let ctx = RequestContextBuilder::new(TaskKind::MgmtRequest)
-            .download_behavior(DownloadBehavior::Download)
-            .scope(context::Scope::new_timeline(&timeline))
-            .read_path_debug(true)
-            .root();
+        let ctx = RequestContextBuilder::extend(&ctx).read_path_debug(true)
+        .scope(context::Scope::new_timeline(&timeline)).build();

        // Use last_record_lsn if no lsn is provided
        let lsn = lsn.unwrap_or_else(|| timeline.get_last_record_lsn());
@@ -2911,22 +2898,9 @@ async fn tenant_scan_remote_handler(
            };
        }

-        let result =
-            download_tenant_manifest(&state.remote_storage, &tenant_shard_id, generation, &cancel)
-                .instrument(info_span!("download_tenant_manifest",
-                            tenant_id=%tenant_shard_id.tenant_id,
-                            shard_id=%tenant_shard_id.shard_slug()))
-                .await;
-        let stripe_size = match result {
-            Ok((manifest, _, _)) => manifest.stripe_size,
-            Err(DownloadError::NotFound) => None,
-            Err(err) => return Err(ApiError::InternalServerError(anyhow!(err))),
-        };
-
        response.shards.push(TenantScanRemoteStorageShard {
            tenant_shard_id,
            generation: generation.into(),
-            stripe_size,
        });
    }

@@ -3202,8 +3176,7 @@ async fn list_aux_files(
        timeline.gate.enter().map_err(|_| ApiError::Cancelled)?,
    );

-    let ctx = RequestContext::new(TaskKind::MgmtRequest, DownloadBehavior::Download)
-        .with_scope_timeline(&timeline);
+    let ctx = RequestContext::new(TaskKind::MgmtRequest, DownloadBehavior::Download);
    let files = timeline
        .list_aux_files(body.lsn, &ctx, io_concurrency)
        .await?;
@@ -3381,11 +3354,11 @@ async fn put_tenant_timeline_import_basebackup(

        let broker_client = state.broker_client.clone();

-        let mut body = StreamReader::new(
-            request
-                .into_body()
-                .map(|res| res.map_err(|error| std::io::Error::other(anyhow::anyhow!(error)))),
-        );
+        let mut body = StreamReader::new(request.into_body().map(|res| {
+            res.map_err(|error| {
+                std::io::Error::new(std::io::ErrorKind::Other, anyhow::anyhow!(error))
+            })
+        }));

        tenant.wait_to_become_active(ACTIVE_TENANT_TIMEOUT).await?;

@@ -3447,19 +3420,18 @@ async fn put_tenant_timeline_import_wal(

    check_permission(&request, Some(tenant_id))?;

+    let ctx = RequestContext::new(TaskKind::MgmtRequest, DownloadBehavior::Warn);
+
    let span = info_span!("import_wal", tenant_id=%tenant_id, timeline_id=%timeline_id, start_lsn=%start_lsn, end_lsn=%end_lsn);
    async move {
        let state = get_state(&request);

        let timeline = active_timeline_of_active_tenant(&state.tenant_manager, TenantShardId::unsharded(tenant_id), timeline_id).await?;
-        let ctx = RequestContextBuilder::new(TaskKind::MgmtRequest)
-            .download_behavior(DownloadBehavior::Warn)
-            .scope(context::Scope::new_timeline(&timeline))
-            .root();
+        let ctx = RequestContextBuilder::extend(&ctx).scope(context::Scope::new_timeline(&timeline)).build();

        let mut body = StreamReader::new(request.into_body().map(|res| {
            res.map_err(|error| {
-                std::io::Error::other( anyhow::anyhow!(error))
+                std::io::Error::new(std::io::ErrorKind::Other, anyhow::anyhow!(error))
            })
        }));

--- a/pageserver/src/lib.rs
+++ b/pageserver/src/lib.rs
@@ -55,9 +55,6 @@ pub const DEFAULT_PG_VERSION: u32 = 16;
 pub const IMAGE_FILE_MAGIC: u16 = 0x5A60;
 pub const DELTA_FILE_MAGIC: u16 = 0x5A61;

-// Target used for performance traces.
-pub const PERF_TRACE_TARGET: &str = "P";
-
 static ZERO_PAGE: bytes::Bytes = bytes::Bytes::from_static(&[0u8; 8192]);

 pub use crate::metrics::preinitialize_metrics;
--- a/pageserver/src/metrics.rs
+++ b/pageserver/src/metrics.rs
@@ -1,8 +1,10 @@
 use std::collections::HashMap;
 use std::num::NonZeroUsize;
 use std::os::fd::RawFd;
+use std::pin::Pin;
 use std::sync::atomic::AtomicU64;
 use std::sync::{Arc, Mutex};
+use std::task::{Context, Poll};
 use std::time::{Duration, Instant};

 use enum_map::{Enum as _, EnumMap};
@@ -21,13 +23,13 @@ use pageserver_api::config::{
 };
 use pageserver_api::models::InMemoryLayerInfo;
 use pageserver_api::shard::TenantShardId;
+use pin_project_lite::pin_project;
 use postgres_backend::{QueryError, is_expected_io_error};
 use pq_proto::framed::ConnectionError;
 use strum::{EnumCount, IntoEnumIterator as _, VariantNames};
 use strum_macros::{IntoStaticStr, VariantNames};
 use utils::id::TimelineId;

-use crate::config;
 use crate::config::PageServerConf;
 use crate::context::{PageContentKind, RequestContext};
 use crate::pgdatadir_mapping::DatadirModificationStats;
@@ -497,100 +499,6 @@ pub(crate) static WAIT_LSN_IN_PROGRESS_GLOBAL_MICROS: Lazy<IntCounter> = Lazy::n
    .expect("failed to define a metric")
 });

-pub(crate) mod wait_ondemand_download_time {
-    use super::*;
-    const WAIT_ONDEMAND_DOWNLOAD_TIME_BUCKETS: &[f64] = &[
-        0.01, 0.02, 0.03, 0.04, 0.05, 0.06, 0.07, 0.08, 0.09, // 10 ms - 100ms
-        0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, // 100ms to 1s
-        1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0, 9.0, // 1s to 10s
-        10.0, 20.0, 30.0, 40.0, 50.0, 60.0, // 10s to 1m
-    ];
-
-    /// The task kinds for which we want to track wait times for on-demand downloads.
-    /// Other task kinds' wait times are accumulated in label value `unknown`.
-    pub(crate) const WAIT_ONDEMAND_DOWNLOAD_METRIC_TASK_KINDS: [TaskKind; 2] = [
-        TaskKind::PageRequestHandler,
-        TaskKind::WalReceiverConnectionHandler,
-    ];
-
-    pub(crate) static WAIT_ONDEMAND_DOWNLOAD_TIME_GLOBAL: Lazy<Vec<Histogram>> = Lazy::new(|| {
-        let histo = register_histogram_vec!(
-            "pageserver_wait_ondemand_download_seconds_global",
-            "Observations are individual tasks' wait times for on-demand downloads. \
-         If N tasks coalesce on an on-demand download, and it takes 10s, than we observe N * 10s.",
-            &["task_kind"],
-            WAIT_ONDEMAND_DOWNLOAD_TIME_BUCKETS.into(),
-        )
-        .expect("failed to define a metric");
-        WAIT_ONDEMAND_DOWNLOAD_METRIC_TASK_KINDS
-            .iter()
-            .map(|task_kind| histo.with_label_values(&[task_kind.into()]))
-            .collect::<Vec<_>>()
-    });
-
-    pub(crate) static WAIT_ONDEMAND_DOWNLOAD_TIME_SUM: Lazy<CounterVec> = Lazy::new(|| {
-        register_counter_vec!(
-            // use a name that _could_ be evolved into a per-timeline histogram later
-            "pageserver_wait_ondemand_download_seconds_sum",
-            "Like `pageserver_wait_ondemand_download_seconds_global` but per timeline",
-            &["tenant_id", "shard_id", "timeline_id", "task_kind"],
-        )
-        .unwrap()
-    });
-
-    pub struct WaitOndemandDownloadTimeSum {
-        counters: [Counter; WAIT_ONDEMAND_DOWNLOAD_METRIC_TASK_KINDS.len()],
-    }
-
-    impl WaitOndemandDownloadTimeSum {
-        pub(crate) fn new(tenant_id: &str, shard_id: &str, timeline_id: &str) -> Self {
-            let counters = WAIT_ONDEMAND_DOWNLOAD_METRIC_TASK_KINDS
-                .iter()
-                .map(|task_kind| {
-                    WAIT_ONDEMAND_DOWNLOAD_TIME_SUM
-                        .get_metric_with_label_values(&[
-                            tenant_id,
-                            shard_id,
-                            timeline_id,
-                            task_kind.into(),
-                        ])
-                        .unwrap()
-                })
-                .collect::<Vec<_>>();
-            Self {
-                counters: counters.try_into().unwrap(),
-            }
-        }
-        pub(crate) fn observe(&self, task_kind: TaskKind, duration: Duration) {
-            let maybe = WAIT_ONDEMAND_DOWNLOAD_METRIC_TASK_KINDS
-                .iter()
-                .enumerate()
-                .find(|(_, kind)| **kind == task_kind);
-            let Some((idx, _)) = maybe else {
-                return;
-            };
-            WAIT_ONDEMAND_DOWNLOAD_TIME_GLOBAL[idx].observe(duration.as_secs_f64());
-            let counter = &self.counters[idx];
-            counter.inc_by(duration.as_secs_f64());
-        }
-    }
-
-    pub(crate) fn shutdown_timeline(tenant_id: &str, shard_id: &str, timeline_id: &str) {
-        for task_kind in WAIT_ONDEMAND_DOWNLOAD_METRIC_TASK_KINDS {
-            let _ = WAIT_ONDEMAND_DOWNLOAD_TIME_SUM.remove_label_values(&[
-                tenant_id,
-                shard_id,
-                timeline_id,
-                task_kind.into(),
-            ]);
-        }
-    }
-
-    pub(crate) fn preinitialize_global_metrics() {
-        Lazy::force(&WAIT_ONDEMAND_DOWNLOAD_TIME_GLOBAL);
-    }
-}
-
 static LAST_RECORD_LSN: Lazy<IntGaugeVec> = Lazy::new(|| {
    register_int_gauge_vec!(
        "pageserver_last_record_lsn",
@@ -1340,13 +1248,13 @@ pub(crate) static STORAGE_IO_TIME_METRIC: Lazy<StorageIoTime> = Lazy::new(Storag

 #[derive(Clone, Copy)]
 #[repr(usize)]
-pub(crate) enum StorageIoSizeOperation {
+enum StorageIoSizeOperation {
    Read,
    Write,
 }

 impl StorageIoSizeOperation {
-    pub(crate) const VARIANTS: &'static [&'static str] = &["read", "write"];
+    const VARIANTS: &'static [&'static str] = &["read", "write"];

    fn as_str(&self) -> &'static str {
        Self::VARIANTS[*self as usize]
@@ -1354,7 +1262,7 @@ impl StorageIoSizeOperation {
 }

 // Needed for the https://neonprod.grafana.net/d/5uK9tHL4k/picking-tenant-for-relocation?orgId=1
-pub(crate) static STORAGE_IO_SIZE: Lazy<UIntGaugeVec> = Lazy::new(|| {
+static STORAGE_IO_SIZE: Lazy<UIntGaugeVec> = Lazy::new(|| {
    register_uint_gauge_vec!(
        "pageserver_io_operations_bytes_total",
        "Total amount of bytes read/written in IO operations",
@@ -2406,18 +2314,13 @@ impl RemoteOpFileKind {
    }
 }

-pub(crate) static REMOTE_TIMELINE_CLIENT_COMPLETION_LATENCY: Lazy<HistogramVec> = Lazy::new(|| {
+pub(crate) static REMOTE_OPERATION_TIME: Lazy<HistogramVec> = Lazy::new(|| {
    register_histogram_vec!(
-        "pageserver_remote_timeline_client_seconds_global",
-        "Time spent on remote timeline client operations. \
-        Grouped by task_kind, file_kind, operation_kind and status. \
-        The task_kind is \
-          - for layer downloads, populated from RequestContext (primary objective of having the label) \
-          - for index downloads, set to 'unknown' \
-          - for any upload operation, set to 'RemoteUploadTask' \
-        This keeps dimensionality at bay. \
+        "pageserver_remote_operation_seconds",
+        "Time spent on remote storage operations. \
+        Grouped by tenant, timeline, operation_kind and status. \
        Does not account for time spent waiting in remote timeline client's queues.",
-        &["task_kind", "file_kind", "op_kind", "status"]
+        &["file_kind", "op_kind", "status"]
    )
    .expect("failed to define a metric")
 });
@@ -2979,7 +2882,6 @@ pub(crate) struct TimelineMetrics {
    pub storage_io_size: StorageIoSizeMetrics,
    pub wait_lsn_in_progress_micros: GlobalAndPerTenantIntCounter,
    pub wait_lsn_start_finish_counterpair: IntCounterPair,
-    pub wait_ondemand_download_time: wait_ondemand_download_time::WaitOndemandDownloadTimeSum,
    shutdown: std::sync::atomic::AtomicBool,
 }

@@ -3125,13 +3027,6 @@ impl TimelineMetrics {
            .get_metric_with_label_values(&[&tenant_id, &shard_id, &timeline_id])
            .unwrap();

-        let wait_ondemand_download_time =
-            wait_ondemand_download_time::WaitOndemandDownloadTimeSum::new(
-                &tenant_id,
-                &shard_id,
-                &timeline_id,
-            );
-
        TimelineMetrics {
            tenant_id,
            shard_id,
@@ -3165,7 +3060,6 @@ impl TimelineMetrics {
            wal_records_received,
            wait_lsn_in_progress_micros,
            wait_lsn_start_finish_counterpair,
-            wait_ondemand_download_time,
            shutdown: std::sync::atomic::AtomicBool::default(),
        }
    }
@@ -3358,8 +3252,6 @@ impl TimelineMetrics {
                .remove_label_values(&mut res, &[tenant_id, shard_id, timeline_id]);
        }

-        wait_ondemand_download_time::shutdown_timeline(tenant_id, shard_id, timeline_id);
-
        let _ = SMGR_QUERY_STARTED_PER_TENANT_TIMELINE.remove_label_values(&[
            SmgrQueryType::GetPageAtLsn.into(),
            tenant_id,
@@ -3481,18 +3373,13 @@ impl RemoteTimelineClientMetrics {

    pub fn remote_operation_time(
        &self,
-        task_kind: Option<TaskKind>,
        file_kind: &RemoteOpFileKind,
        op_kind: &RemoteOpKind,
        status: &'static str,
    ) -> Histogram {
-        REMOTE_TIMELINE_CLIENT_COMPLETION_LATENCY
-            .get_metric_with_label_values(&[
-                task_kind.as_ref().map(|tk| tk.into()).unwrap_or("unknown"),
-                file_kind.as_str(),
-                op_kind.as_str(),
-                status,
-            ])
+        let key = (file_kind.as_str(), op_kind.as_str(), status);
+        REMOTE_OPERATION_TIME
+            .get_metric_with_label_values(&[key.0, key.1, key.2])
            .unwrap()
    }

@@ -3737,26 +3624,54 @@ impl Drop for RemoteTimelineClientMetrics {

 /// Wrapper future that measures the time spent by a remote storage operation,
 /// and records the time and success/failure as a prometheus metric.
-pub(crate) trait MeasureRemoteOp<O, E>: Sized + Future<Output = Result<O, E>> {
-    async fn measure_remote_op(
+pub(crate) trait MeasureRemoteOp: Sized {
+    fn measure_remote_op(
        self,
-        task_kind: Option<TaskKind>, // not all caller contexts have a RequestContext / TaskKind handy
        file_kind: RemoteOpFileKind,
        op: RemoteOpKind,
        metrics: Arc<RemoteTimelineClientMetrics>,
-    ) -> Result<O, E> {
+    ) -> MeasuredRemoteOp<Self> {
        let start = Instant::now();
-        let res = self.await;
-        let duration = start.elapsed();
-        let status = if res.is_ok() { &"success" } else { &"failure" };
-        metrics
-            .remote_operation_time(task_kind, &file_kind, &op, status)
-            .observe(duration.as_secs_f64());
-        res
+        MeasuredRemoteOp {
+            inner: self,
+            file_kind,
+            op,
+            start,
+            metrics,
+        }
    }
 }

-impl<Fut, O, E> MeasureRemoteOp<O, E> for Fut where Fut: Sized + Future<Output = Result<O, E>> {}
+impl<T: Sized> MeasureRemoteOp for T {}
+
+pin_project! {
+    pub(crate) struct MeasuredRemoteOp<F>
+    {
+        #[pin]
+        inner: F,
+        file_kind: RemoteOpFileKind,
+        op: RemoteOpKind,
+        start: Instant,
+        metrics: Arc<RemoteTimelineClientMetrics>,
+    }
+}
+
+impl<F: Future<Output = Result<O, E>>, O, E> Future for MeasuredRemoteOp<F> {
+    type Output = Result<O, E>;
+
+    fn poll(self: Pin<&mut Self>, cx: &mut Context<'_>) -> Poll<Self::Output> {
+        let this = self.project();
+        let poll_result = this.inner.poll(cx);
+        if let Poll::Ready(ref res) = poll_result {
+            let duration = this.start.elapsed();
+            let status = if res.is_ok() { &"success" } else { &"failure" };
+            this.metrics
+                .remote_operation_time(this.file_kind, this.op, status)
+                .observe(duration.as_secs_f64());
+        }
+        poll_result
+    }
+}

 pub mod tokio_epoll_uring {
    use std::collections::HashMap;
@@ -4192,33 +4107,9 @@ pub(crate) fn set_tokio_runtime_setup(setup: &str, num_threads: NonZeroUsize) {
        .set(u64::try_from(num_threads.get()).unwrap());
 }

-static PAGESERVER_CONFIG_IGNORED_ITEMS: Lazy<UIntGaugeVec> = Lazy::new(|| {
-    register_uint_gauge_vec!(
-        "pageserver_config_ignored_items",
-        "TOML items present in the on-disk configuration file but ignored by the pageserver config parser.\
-         The `item` label is the dot-separated path of the ignored item in the on-disk configuration file.\
-         The value for an unknown config item is always 1.\
-         There is a special label value \"\", which is 0, so that there is always a metric exposed (simplifies dashboards).",
-        &["item"]
-    )
-    .unwrap()
-});
-
-pub fn preinitialize_metrics(
-    conf: &'static PageServerConf,
-    ignored: config::ignored_fields::Paths,
-) {
+pub fn preinitialize_metrics(conf: &'static PageServerConf) {
    set_page_service_config_max_batch_size(&conf.page_service_pipelining);

-    PAGESERVER_CONFIG_IGNORED_ITEMS
-        .with_label_values(&[""])
-        .set(0);
-    for path in &ignored.paths {
-        PAGESERVER_CONFIG_IGNORED_ITEMS
-            .with_label_values(&[path])
-            .set(1);
-    }
-
    // Python tests need these and on some we do alerting.
    //
    // FIXME(4813): make it so that we have no top level metrics as this fn will easily fall out of
@@ -4304,5 +4195,4 @@ pub fn preinitialize_metrics(
    Lazy::force(&tokio_epoll_uring::THREAD_LOCAL_METRICS_STORAGE);

    tenant_throttling::preinitialize_global_metrics();
-    wait_ondemand_download_time::preinitialize_global_metrics();
 }
--- a/pageserver/src/page_service.rs
+++ b/pageserver/src/page_service.rs
@@ -9,7 +9,6 @@ use std::sync::Arc;
 use std::time::{Duration, Instant, SystemTime};
 use std::{io, str};

-use crate::PERF_TRACE_TARGET;
 use anyhow::{Context, bail};
 use async_compression::tokio::write::GzipEncoder;
 use bytes::Buf;
@@ -54,9 +53,7 @@ use utils::sync::spsc_fold;
 use crate::auth::check_permission;
 use crate::basebackup::BasebackupError;
 use crate::config::PageServerConf;
-use crate::context::{
-    DownloadBehavior, PerfInstrumentFutureExt, RequestContext, RequestContextBuilder,
-};
+use crate::context::{DownloadBehavior, RequestContext};
 use crate::metrics::{
    self, COMPUTE_COMMANDS_COUNTERS, ComputeCommandKind, LIVE_CONNECTIONS, SmgrOpTimer,
    TimelineMetrics,
@@ -103,7 +100,6 @@ pub fn spawn(
    conf: &'static PageServerConf,
    tenant_manager: Arc<TenantManager>,
    pg_auth: Option<Arc<SwappableJwtAuth>>,
-    perf_trace_dispatch: Option<Dispatch>,
    tcp_listener: tokio::net::TcpListener,
 ) -> Listener {
    let cancel = CancellationToken::new();
@@ -121,7 +117,6 @@ pub fn spawn(
            conf,
            tenant_manager,
            pg_auth,
-            perf_trace_dispatch,
            tcp_listener,
            conf.pg_auth_type,
            conf.page_service_pipelining.clone(),
@@ -178,7 +173,6 @@ pub async fn libpq_listener_main(
    conf: &'static PageServerConf,
    tenant_manager: Arc<TenantManager>,
    auth: Option<Arc<SwappableJwtAuth>>,
-    perf_trace_dispatch: Option<Dispatch>,
    listener: tokio::net::TcpListener,
    auth_type: AuthType,
    pipelining_config: PageServicePipeliningConfig,
@@ -211,12 +205,8 @@ pub async fn libpq_listener_main(
                // Connection established. Spawn a new task to handle it.
                debug!("accepted connection from {}", peer_addr);
                let local_auth = auth.clone();
-                let connection_ctx = RequestContextBuilder::from(&listener_ctx)
-                    .task_kind(TaskKind::PageRequestHandler)
-                    .download_behavior(DownloadBehavior::Download)
-                    .perf_span_dispatch(perf_trace_dispatch.clone())
-                    .detached_child();
-
+                let connection_ctx = listener_ctx
+                    .detached_child(TaskKind::PageRequestHandler, DownloadBehavior::Download);
                connection_handler_tasks.spawn(page_service_conn_main(
                    conf,
                    tenant_manager.clone(),
@@ -247,15 +237,6 @@ pub async fn libpq_listener_main(

 type ConnectionHandlerResult = anyhow::Result<()>;

-/// Perf root spans start at the per-request level, after shard routing.
-/// This struct carries connection-level information to the root perf span definition.
-#[derive(Clone)]
-struct ConnectionPerfSpanFields {
-    peer_addr: String,
-    application_name: Option<String>,
-    compute_mode: Option<String>,
-}
-
 #[instrument(skip_all, fields(peer_addr, application_name, compute_mode))]
 #[allow(clippy::too_many_arguments)]
 async fn page_service_conn_main(
@@ -280,12 +261,6 @@ async fn page_service_conn_main(
    let socket_fd = socket.as_raw_fd();

    let peer_addr = socket.peer_addr().context("get peer address")?;
-
-    let perf_span_fields = ConnectionPerfSpanFields {
-        peer_addr: peer_addr.to_string(),
-        application_name: None, // filled in later
-        compute_mode: None,     // filled in later
-    };
    tracing::Span::current().record("peer_addr", field::display(peer_addr));

    // setup read timeout of 10 minutes. the timeout is rather arbitrary for requirements:
@@ -329,7 +304,6 @@ async fn page_service_conn_main(
        tenant_manager,
        auth,
        pipelining_config,
-        perf_span_fields,
        connection_ctx,
        cancel.clone(),
        gate_guard,
@@ -374,8 +348,6 @@ struct PageServerHandler {
    /// `process_query` creates a child context from this one.
    connection_ctx: RequestContext,

-    perf_span_fields: ConnectionPerfSpanFields,
-
    cancel: CancellationToken,

    /// None only while pagestream protocol is being processed.
@@ -635,7 +607,6 @@ impl std::fmt::Display for BatchedPageStreamError {
 struct BatchedGetPageRequest {
    req: PagestreamGetPageRequest,
    timer: SmgrOpTimer,
-    ctx: RequestContext,
 }

 #[cfg(feature = "testing")]
@@ -721,13 +692,11 @@ impl BatchedFeMessage {
 }

 impl PageServerHandler {
-    #[allow(clippy::too_many_arguments)]
    pub fn new(
        conf: &'static PageServerConf,
        tenant_manager: Arc<TenantManager>,
        auth: Option<Arc<SwappableJwtAuth>>,
        pipelining_config: PageServicePipeliningConfig,
-        perf_span_fields: ConnectionPerfSpanFields,
        connection_ctx: RequestContext,
        cancel: CancellationToken,
        gate_guard: GateGuard,
@@ -737,7 +706,6 @@ impl PageServerHandler {
            auth,
            claims: None,
            connection_ctx,
-            perf_span_fields,
            timeline_handles: Some(TimelineHandles::new(tenant_manager)),
            cancel,
            pipelining_config,
@@ -775,7 +743,6 @@ impl PageServerHandler {
        tenant_id: TenantId,
        timeline_id: TimelineId,
        timeline_handles: &mut TimelineHandles,
-        conn_perf_span_fields: &ConnectionPerfSpanFields,
        cancel: &CancellationToken,
        ctx: &RequestContext,
        protocol_version: PagestreamProtocolVersion,
@@ -935,12 +902,10 @@ impl PageServerHandler {
                }

                let key = rel_block_to_key(req.rel, req.blkno);
-
-                let res = timeline_handles
+                let shard = match timeline_handles
                    .get(tenant_id, timeline_id, ShardSelector::Page(key))
-                    .await;
-
-                let shard = match res {
+                    .await
+                {
                    Ok(tl) => tl,
                    Err(e) => {
                        let span = mkspan!(before shard routing);
@@ -967,41 +932,6 @@ impl PageServerHandler {
                        }
                    }
                };
-
-                let ctx = if shard.is_get_page_request_sampled() {
-                    RequestContextBuilder::from(ctx)
-                        .root_perf_span(|| {
-                            info_span!(
-                            target: PERF_TRACE_TARGET,
-                            "GET_PAGE",
-                            peer_addr = conn_perf_span_fields.peer_addr,
-                            application_name = conn_perf_span_fields.application_name,
-                            compute_mode = conn_perf_span_fields.compute_mode,
-                            tenant_id = %tenant_id,
-                            shard_id = %shard.get_shard_identity().shard_slug(),
-                            timeline_id = %timeline_id,
-                            lsn = %req.hdr.request_lsn,
-                            request_id = %req.hdr.reqid,
-                            key = %key,
-                            )
-                        })
-                        .attached_child()
-                } else {
-                    ctx.attached_child()
-                };
-
-                // This ctx travels as part of the BatchedFeMessage through
-                // batching into the request handler.
-                // The request handler needs to do some per-request work
-                // (relsize check) before dispatching the batch as a single
-                // get_vectored call to the Timeline.
-                // This ctx will be used for the reslize check, whereas the
-                // get_vectored call will be a different ctx with separate
-                // perf span.
-                let ctx = ctx.with_scope_page_service_pagestream(&shard);
-
-                // Similar game for this `span`: we funnel it through so that
-                // request handler log messages contain the request-specific fields.
                let span = mkspan!(shard.tenant_shard_id.shard_slug());

                let timer = record_op_start_and_throttle(
@@ -1009,34 +939,19 @@ impl PageServerHandler {
                    metrics::SmgrQueryType::GetPageAtLsn,
                    received_at,
                )
-                .maybe_perf_instrument(&ctx, |current_perf_span| {
-                    info_span!(
-                        target: PERF_TRACE_TARGET,
-                        parent: current_perf_span,
-                        "THROTTLE",
-                    )
-                })
                .await?;

                // We're holding the Handle
-                // TODO: if we actually need to wait for lsn here, it delays the entire batch which doesn't need to wait
-                let res = Self::wait_or_get_last_lsn(
+                let effective_request_lsn = match Self::wait_or_get_last_lsn(
                    &shard,
                    req.hdr.request_lsn,
                    req.hdr.not_modified_since,
                    &shard.get_applied_gc_cutoff_lsn(),
-                    &ctx,
+                    ctx,
                )
-                .maybe_perf_instrument(&ctx, |current_perf_span| {
-                    info_span!(
-                        target: PERF_TRACE_TARGET,
-                        parent: current_perf_span,
-                        "WAIT_LSN",
-                    )
-                })
-                .await;
-
-                let effective_request_lsn = match res {
+                // TODO: if we actually need to wait for lsn here, it delays the entire batch which doesn't need to wait
+                .await
+                {
                    Ok(lsn) => lsn,
                    Err(e) => {
                        return respond_error!(span, e);
@@ -1046,7 +961,7 @@ impl PageServerHandler {
                    span,
                    shard: shard.downgrade(),
                    effective_request_lsn,
-                    pages: smallvec::smallvec![BatchedGetPageRequest { req, timer, ctx }],
+                    pages: smallvec::smallvec![BatchedGetPageRequest { req, timer }],
                }
            }
            #[cfg(feature = "testing")]
@@ -1599,14 +1514,12 @@ impl PageServerHandler {
        IO: AsyncRead + AsyncWrite + Send + Sync + Unpin + 'static,
    {
        let cancel = self.cancel.clone();
-
        let err = loop {
            let msg = Self::pagestream_read_message(
                &mut pgb_reader,
                tenant_id,
                timeline_id,
                &mut timeline_handles,
-                &self.perf_span_fields,
                &cancel,
                ctx,
                protocol_version,
@@ -1740,8 +1653,6 @@ impl PageServerHandler {
        // Batcher
        //

-        let perf_span_fields = self.perf_span_fields.clone();
-
        let cancel_batcher = self.cancel.child_token();
        let (mut batch_tx, mut batch_rx) = spsc_fold::channel();
        let batcher = pipeline_stage!("batcher", cancel_batcher.clone(), move |cancel_batcher| {
@@ -1755,7 +1666,6 @@ impl PageServerHandler {
                        tenant_id,
                        timeline_id,
                        &mut timeline_handles,
-                        &perf_span_fields,
                        &cancel_batcher,
                        &ctx,
                        protocol_version,
@@ -2094,9 +2004,7 @@ impl PageServerHandler {

        let results = timeline
            .get_rel_page_at_lsn_batched(
-                requests
-                    .iter()
-                    .map(|p| (&p.req.rel, &p.req.blkno, p.ctx.attached_child())),
+                requests.iter().map(|p| (&p.req.rel, &p.req.blkno)),
                effective_lsn,
                io_concurrency,
                ctx,
@@ -2698,14 +2606,12 @@ where

        if let FeStartupPacket::StartupMessage { params, .. } = sm {
            if let Some(app_name) = params.get("application_name") {
-                self.perf_span_fields.application_name = Some(app_name.to_string());
                Span::current().record("application_name", field::display(app_name));
            }
            if let Some(options) = params.get("options") {
                let (config, _) = parse_options(options);
                for (key, value) in config {
                    if key == "neon.compute_mode" {
-                        self.perf_span_fields.compute_mode = Some(value.clone());
                        Span::current().record("compute_mode", field::display(value));
                    }
                }
--- a/pageserver/src/pgdatadir_mapping.rs
+++ b/pageserver/src/pgdatadir_mapping.rs
@@ -9,7 +9,6 @@
 use std::collections::{BTreeMap, HashMap, HashSet, hash_map};
 use std::ops::{ControlFlow, Range};

-use crate::PERF_TRACE_TARGET;
 use anyhow::{Context, ensure};
 use bytes::{Buf, Bytes, BytesMut};
 use enum_map::Enum;
@@ -32,7 +31,7 @@ use postgres_ffi::{BLCKSZ, Oid, RepOriginId, TimestampTz, TransactionId};
 use serde::{Deserialize, Serialize};
 use strum::IntoEnumIterator;
 use tokio_util::sync::CancellationToken;
-use tracing::{debug, info, info_span, trace, warn};
+use tracing::{debug, info, trace, warn};
 use utils::bin_ser::{BeSer, DeserializeError};
 use utils::lsn::Lsn;
 use utils::pausable_failpoint;
@@ -40,7 +39,7 @@ use wal_decoder::serialized_batch::{SerializedValueBatch, ValueMeta};

 use super::tenant::{PageReconstructError, Timeline};
 use crate::aux_file;
-use crate::context::{PerfInstrumentFutureExt, RequestContext, RequestContextBuilder};
+use crate::context::RequestContext;
 use crate::keyspace::{KeySpace, KeySpaceAccum};
 use crate::metrics::{
    RELSIZE_CACHE_ENTRIES, RELSIZE_CACHE_HITS, RELSIZE_CACHE_MISSES, RELSIZE_CACHE_MISSES_OLD,
@@ -210,9 +209,7 @@ impl Timeline {
                let pages: smallvec::SmallVec<[_; 1]> = smallvec::smallvec![(tag, blknum)];
                let res = self
                    .get_rel_page_at_lsn_batched(
-                        pages
-                            .iter()
-                            .map(|(tag, blknum)| (tag, blknum, ctx.attached_child())),
+                        pages.iter().map(|(tag, blknum)| (tag, blknum)),
                        effective_lsn,
                        io_concurrency.clone(),
                        ctx,
@@ -251,7 +248,7 @@ impl Timeline {
    /// The ordering of the returned vec corresponds to the ordering of `pages`.
    pub(crate) async fn get_rel_page_at_lsn_batched(
        &self,
-        pages: impl ExactSizeIterator<Item = (&RelTag, &BlockNumber, RequestContext)>,
+        pages: impl ExactSizeIterator<Item = (&RelTag, &BlockNumber)>,
        effective_lsn: Lsn,
        io_concurrency: IoConcurrency,
        ctx: &RequestContext,
@@ -265,11 +262,8 @@ impl Timeline {
        let mut result = Vec::with_capacity(pages.len());
        let result_slots = result.spare_capacity_mut();

-        let mut keys_slots: BTreeMap<Key, smallvec::SmallVec<[(usize, RequestContext); 1]>> =
-            BTreeMap::default();
-
-        let mut perf_instrument = false;
-        for (response_slot_idx, (tag, blknum, ctx)) in pages.enumerate() {
+        let mut keys_slots: BTreeMap<Key, smallvec::SmallVec<[usize; 1]>> = BTreeMap::default();
+        for (response_slot_idx, (tag, blknum)) in pages.enumerate() {
            if tag.relnode == 0 {
                result_slots[response_slot_idx].write(Err(PageReconstructError::Other(
                    RelationError::InvalidRelnode.into(),
@@ -280,16 +274,7 @@ impl Timeline {
            }

            let nblocks = match self
-                .get_rel_size(*tag, Version::Lsn(effective_lsn), &ctx)
-                .maybe_perf_instrument(&ctx, |crnt_perf_span| {
-                    info_span!(
-                        target: PERF_TRACE_TARGET,
-                        parent: crnt_perf_span,
-                        "GET_REL_SIZE",
-                        reltag=%tag,
-                        lsn=%effective_lsn,
-                    )
-                })
+                .get_rel_size(*tag, Version::Lsn(effective_lsn), ctx)
                .await
            {
                Ok(nblocks) => nblocks,
@@ -312,12 +297,8 @@ impl Timeline {

            let key = rel_block_to_key(*tag, *blknum);

-            if ctx.has_perf_span() {
-                perf_instrument = true;
-            }
-
            let key_slots = keys_slots.entry(key).or_default();
-            key_slots.push((response_slot_idx, ctx));
+            key_slots.push(response_slot_idx);
        }

        let keyspace = {
@@ -333,34 +314,16 @@ impl Timeline {
            acc.to_keyspace()
        };

-        let ctx = match perf_instrument {
-            true => RequestContextBuilder::from(ctx)
-                .root_perf_span(|| {
-                    info_span!(
-                        target: PERF_TRACE_TARGET,
-                        "GET_VECTORED",
-                        tenant_id = %self.tenant_shard_id.tenant_id,
-                        timeline_id = %self.timeline_id,
-                        lsn = %effective_lsn,
-                        shard = %self.tenant_shard_id.shard_slug(),
-                    )
-                })
-                .attached_child(),
-            false => ctx.attached_child(),
-        };
-
-        let res = self
-            .get_vectored(keyspace, effective_lsn, io_concurrency, &ctx)
-            .maybe_perf_instrument(&ctx, |current_perf_span| current_perf_span.clone())
-            .await;
-
-        match res {
+        match self
+            .get_vectored(keyspace, effective_lsn, io_concurrency, ctx)
+            .await
+        {
            Ok(results) => {
                for (key, res) in results {
                    let mut key_slots = keys_slots.remove(&key).unwrap().into_iter();
-                    let (first_slot, first_req_ctx) = key_slots.next().unwrap();
+                    let first_slot = key_slots.next().unwrap();

-                    for (slot, req_ctx) in key_slots {
+                    for slot in key_slots {
                        let clone = match &res {
                            Ok(buf) => Ok(buf.clone()),
                            Err(err) => Err(match err {
@@ -378,22 +341,17 @@ impl Timeline {
                        };

                        result_slots[slot].write(clone);
-                        // There is no standardized way to express that the batched span followed from N request spans.
-                        // So, abuse the system and mark the request contexts as follows_from the batch span, so we get
-                        // some linkage in our trace viewer. It allows us to answer: which GET_VECTORED did this GET_PAGE wait for.
-                        req_ctx.perf_follows_from(&ctx);
                        slots_filled += 1;
                    }

                    result_slots[first_slot].write(res);
-                    first_req_ctx.perf_follows_from(&ctx);
                    slots_filled += 1;
                }
            }
            Err(err) => {
                // this cannot really happen because get_vectored only errors globally on invalid LSN or too large batch size
                // (We enforce the max batch size outside of this function, in the code that constructs the batch request.)
-                for (slot, req_ctx) in keys_slots.values().flatten() {
+                for slot in keys_slots.values().flatten() {
                    // this whole `match` is a lot like `From<GetVectoredError> for PageReconstructError`
                    // but without taking ownership of the GetVectoredError
                    let err = match &err {
@@ -425,7 +383,6 @@ impl Timeline {
                        }
                    };

-                    req_ctx.perf_follows_from(&ctx);
                    result_slots[*slot].write(err);
                }

@@ -691,7 +648,7 @@ impl Timeline {
        Ok(buf.get_u32_le())
    }

-    /// Does the slru segment exist?
+    /// Get size of an SLRU segment
    pub(crate) async fn get_slru_segment_exists(
        &self,
        kind: SlruKind,
@@ -844,9 +801,9 @@ impl Timeline {
        .await
    }

-    /// Obtain the timestamp for the given lsn.
+    /// Obtain the possible timestamp range for the given lsn.
    ///
-    /// If the lsn has no timestamps (e.g. no commits), returns None.
+    /// If the lsn has no timestamps, returns None. returns `(min, max, median)` if it has timestamps.
    pub(crate) async fn get_timestamp_for_lsn(
        &self,
        probe_lsn: Lsn,
--- a/pageserver/src/task_mgr.rs
+++ b/pageserver/src/task_mgr.rs
@@ -219,7 +219,8 @@ pageserver_runtime!(MGMT_REQUEST_RUNTIME, "mgmt request worker");
 pageserver_runtime!(WALRECEIVER_RUNTIME, "walreceiver worker");
 pageserver_runtime!(BACKGROUND_RUNTIME, "background op worker");
 // Bump this number when adding a new pageserver_runtime!
-const NUM_MULTIPLE_RUNTIMES: NonZeroUsize = NonZeroUsize::new(4).unwrap();
+// SAFETY: it's obviously correct
+const NUM_MULTIPLE_RUNTIMES: NonZeroUsize = unsafe { NonZeroUsize::new_unchecked(4) };

 #[derive(Debug, Clone, Copy)]
 pub struct PageserverTaskId(u64);
--- a/pageserver/src/tenant.rs
+++ b/pageserver/src/tenant.rs
@@ -45,7 +45,6 @@ use remote_timeline_client::manifest::{
 };
 use remote_timeline_client::{
    FAILED_REMOTE_OP_RETRIES, FAILED_UPLOAD_WARN_THRESHOLD, UploadQueueNotReadyError,
-    download_tenant_manifest,
 };
 use secondary::heatmap::{HeatMapTenant, HeatMapTimeline};
 use storage_broker::BrokerClientChannel;
@@ -227,8 +226,7 @@ struct TimelinePreload {
 }

 pub(crate) struct TenantPreload {
-    /// The tenant manifest from remote storage, or None if no manifest was found.
-    tenant_manifest: Option<TenantManifest>,
+    tenant_manifest: TenantManifest,
    /// Map from timeline ID to a possible timeline preload. It is None iff the timeline is offloaded according to the manifest.
    timelines: HashMap<TimelineId, Option<TimelinePreload>>,
 }
@@ -284,15 +282,12 @@ pub struct Tenant {
    /// **Lock order**: if acquiring all (or a subset), acquire them in order `timelines`, `timelines_offloaded`, `timelines_creating`
    timelines_offloaded: Mutex<HashMap<TimelineId, Arc<OffloadedTimeline>>>,

-    /// The last tenant manifest known to be in remote storage. None if the manifest has not yet
-    /// been either downloaded or uploaded. Always Some after tenant attach.
+    /// Serialize writes of the tenant manifest to remote storage.  If there are concurrent operations
+    /// affecting the manifest, such as timeline deletion and timeline offload, they must wait for
+    /// each other (this could be optimized to coalesce writes if necessary).
    ///
-    /// Initially populated during tenant attach, updated via `maybe_upload_tenant_manifest`.
-    ///
-    /// Do not modify this directly. It is used to check whether a new manifest needs to be
-    /// uploaded. The manifest is constructed in `build_tenant_manifest`, and uploaded via
-    /// `maybe_upload_tenant_manifest`.
-    remote_tenant_manifest: tokio::sync::Mutex<Option<TenantManifest>>,
+    /// The contents of the Mutex are the last manifest we successfully uploaded
+    tenant_manifest_upload: tokio::sync::Mutex<Option<TenantManifest>>,

    // This mutex prevents creation of new timelines during GC.
    // Adding yet another mutex (in addition to `timelines`) is needed because holding
@@ -920,7 +915,6 @@ enum StartCreatingTimelineResult {
    Idempotent(Arc<Timeline>),
 }

-#[allow(clippy::large_enum_variant, reason = "TODO")]
 enum TimelineInitAndSyncResult {
    ReadyToActivate(Arc<Timeline>),
    NeedsSpawnImportPgdata(TimelineInitAndSyncNeedsSpawnImportPgdata),
@@ -1007,7 +1001,6 @@ enum CreateTimelineCause {
    Delete,
 }

-#[allow(clippy::large_enum_variant, reason = "TODO")]
 enum LoadTimelineCause {
    Attach,
    Unoffload,
@@ -1361,41 +1354,36 @@ impl Tenant {
                    }
                }

-                fn make_broken_or_stopping(t: &Tenant, err: anyhow::Error) {
-                    t.state.send_modify(|state| match state {
-                        // TODO: the old code alluded to DeleteTenantFlow sometimes setting
-                        // TenantState::Stopping before we get here, but this may be outdated.
-                        // Let's find out with a testing assertion. If this doesn't fire, and the
-                        // logs don't show this happening in production, remove the Stopping cases.
-                        TenantState::Stopping{..} if cfg!(any(test, feature = "testing")) => {
-                            panic!("unexpected TenantState::Stopping during attach")
-                        }
-                        // If the tenant is cancelled, assume the error was caused by cancellation.
-                        TenantState::Attaching if t.cancel.is_cancelled() => {
-                            info!("attach cancelled, setting tenant state to Stopping: {err}");
-                            // NB: progress None tells `set_stopping` that attach has cancelled.
-                            *state = TenantState::Stopping { progress: None };
-                        }
-                        // According to the old code, DeleteTenantFlow may already have set this to
-                        // Stopping. Retain its progress.
-                        // TODO: there is no DeleteTenantFlow. Is this still needed? See above.
-                        TenantState::Stopping { progress } if t.cancel.is_cancelled() => {
-                            assert!(progress.is_some(), "concurrent attach cancellation");
-                            info!("attach cancelled, already Stopping: {err}");
-                        }
-                        // Mark the tenant as broken.
-                        TenantState::Attaching | TenantState::Stopping { .. } => {
-                            error!("attach failed, setting tenant state to Broken (was {state}): {err:?}");
-                            *state = TenantState::broken_from_reason(err.to_string())
-                        }
-                        // The attach task owns the tenant state until activated.
-                        state => panic!("invalid tenant state {state} during attach: {err:?}"),
-                    });
+                // Ideally we should use Tenant::set_broken_no_wait, but it is not supposed to be used when tenant is in loading state.
+                enum BrokenVerbosity {
+                    Error,
+                    Info
                }
+                let make_broken =
+                    |t: &Tenant, err: anyhow::Error, verbosity: BrokenVerbosity| {
+                        match verbosity {
+                            BrokenVerbosity::Info => {
+                                info!("attach cancelled, setting tenant state to Broken: {err}");
+                            },
+                            BrokenVerbosity::Error => {
+                                error!("attach failed, setting tenant state to Broken: {err:?}");
+                            }
+                        }
+                        t.state.send_modify(|state| {
+                            // The Stopping case is for when we have passed control on to DeleteTenantFlow:
+                            // if it errors, we will call make_broken when tenant is already in Stopping.
+                            assert!(
+                                matches!(*state, TenantState::Attaching | TenantState::Stopping { .. }),
+                                "the attach task owns the tenant state until activation is complete"
+                            );
+
+                            *state = TenantState::broken_from_reason(err.to_string());
+                        });
+                    };

                // TODO: should also be rejecting tenant conf changes that violate this check.
                if let Err(e) = crate::tenant::storage_layer::inmemory_layer::IndexEntry::validate_checkpoint_distance(tenant_clone.get_checkpoint_distance()) {
-                    make_broken_or_stopping(&tenant_clone, anyhow::anyhow!(e));
+                    make_broken(&tenant_clone, anyhow::anyhow!(e), BrokenVerbosity::Error);
                    return Ok(());
                }

@@ -1447,8 +1435,10 @@ impl Tenant {
                            // stayed in Activating for such a long time that shutdown found it in
                            // that state.
                            tracing::info!(state=%tenant_clone.current_state(), "Tenant shut down before activation");
-                            // Set the tenant to Stopping to signal `set_stopping` that we're done.
-                            make_broken_or_stopping(&tenant_clone, anyhow::anyhow!("Shut down while Attaching"));
+                            // Make the tenant broken so that set_stopping will not hang waiting for it to leave
+                            // the Attaching state.  This is an over-reaction (nothing really broke, the tenant is
+                            // just shutting down), but ensures progress.
+                            make_broken(&tenant_clone, anyhow::anyhow!("Shut down while Attaching"), BrokenVerbosity::Info);
                            return Ok(());
                        },
                    )
@@ -1467,7 +1457,7 @@ impl Tenant {
                        match res {
                            Ok(p) => Some(p),
                            Err(e) => {
-                                make_broken_or_stopping(&tenant_clone, anyhow::anyhow!(e));
+                                make_broken(&tenant_clone, anyhow::anyhow!(e), BrokenVerbosity::Error);
                                return Ok(());
                            }
                        }
@@ -1493,7 +1483,9 @@ impl Tenant {
                        info!("attach finished, activating");
                        tenant_clone.activate(broker_client, None, &ctx);
                    }
-                    Err(e) => make_broken_or_stopping(&tenant_clone, anyhow::anyhow!(e)),
+                    Err(e) => {
+                        make_broken(&tenant_clone, anyhow::anyhow!(e), BrokenVerbosity::Error);
+                    }
                }

                // If we are doing an opportunistic warmup attachment at startup, initialize
@@ -1533,27 +1525,28 @@ impl Tenant {
            cancel.clone(),
        )
        .await?;
-
-        let tenant_manifest = match download_tenant_manifest(
-            remote_storage,
-            &self.tenant_shard_id,
-            self.generation,
-            &cancel,
-        )
-        .await
-        {
-            Ok((tenant_manifest, _, _)) => Some(tenant_manifest),
-            Err(DownloadError::NotFound) => None,
-            Err(err) => return Err(err.into()),
-        };
+        let (offloaded_add, tenant_manifest) =
+            match remote_timeline_client::download_tenant_manifest(
+                remote_storage,
+                &self.tenant_shard_id,
+                self.generation,
+                &cancel,
+            )
+            .await
+            {
+                Ok((tenant_manifest, _generation, _manifest_mtime)) => (
+                    format!("{} offloaded", tenant_manifest.offloaded_timelines.len()),
+                    tenant_manifest,
+                ),
+                Err(DownloadError::NotFound) => {
+                    ("no manifest".to_string(), TenantManifest::empty())
+                }
+                Err(e) => Err(e)?,
+            };

        info!(
-            "found {} timelines ({} offloaded timelines)",
-            remote_timeline_ids.len(),
-            tenant_manifest
-                .as_ref()
-                .map(|m| m.offloaded_timelines.len())
-                .unwrap_or(0)
+            "found {} timelines, and {offloaded_add}",
+            remote_timeline_ids.len()
        );

        for k in other_keys {
@@ -1562,13 +1555,11 @@ impl Tenant {

        // Avoid downloading IndexPart of offloaded timelines.
        let mut offloaded_with_prefix = HashSet::new();
-        if let Some(tenant_manifest) = &tenant_manifest {
-            for offloaded in tenant_manifest.offloaded_timelines.iter() {
-                if remote_timeline_ids.remove(&offloaded.timeline_id) {
-                    offloaded_with_prefix.insert(offloaded.timeline_id);
-                } else {
-                    // We'll take care later of timelines in the manifest without a prefix
-                }
+        for offloaded in tenant_manifest.offloaded_timelines.iter() {
+            if remote_timeline_ids.remove(&offloaded.timeline_id) {
+                offloaded_with_prefix.insert(offloaded.timeline_id);
+            } else {
+                // We'll take care later of timelines in the manifest without a prefix
            }
        }

@@ -1642,14 +1633,12 @@ impl Tenant {

        let mut offloaded_timeline_ids = HashSet::new();
        let mut offloaded_timelines_list = Vec::new();
-        if let Some(tenant_manifest) = &preload.tenant_manifest {
-            for timeline_manifest in tenant_manifest.offloaded_timelines.iter() {
-                let timeline_id = timeline_manifest.timeline_id;
-                let offloaded_timeline =
-                    OffloadedTimeline::from_manifest(self.tenant_shard_id, timeline_manifest);
-                offloaded_timelines_list.push((timeline_id, Arc::new(offloaded_timeline)));
-                offloaded_timeline_ids.insert(timeline_id);
-            }
+        for timeline_manifest in preload.tenant_manifest.offloaded_timelines.iter() {
+            let timeline_id = timeline_manifest.timeline_id;
+            let offloaded_timeline =
+                OffloadedTimeline::from_manifest(self.tenant_shard_id, timeline_manifest);
+            offloaded_timelines_list.push((timeline_id, Arc::new(offloaded_timeline)));
+            offloaded_timeline_ids.insert(timeline_id);
        }
        // Complete deletions for offloaded timeline id's from manifest.
        // The manifest will be uploaded later in this function.
@@ -1807,21 +1796,15 @@ impl Tenant {
            .context("resume_deletion")
            .map_err(LoadLocalTimelineError::ResumeDeletion)?;
        }
+        let needs_manifest_upload =
+            offloaded_timelines_list.len() != preload.tenant_manifest.offloaded_timelines.len();
        {
            let mut offloaded_timelines_accessor = self.timelines_offloaded.lock().unwrap();
            offloaded_timelines_accessor.extend(offloaded_timelines_list.into_iter());
        }
-
-        // Stash the preloaded tenant manifest, and upload a new manifest if changed.
-        //
-        // NB: this must happen after the tenant is fully populated above. In particular the
-        // offloaded timelines, which are included in the manifest.
-        {
-            let mut guard = self.remote_tenant_manifest.lock().await;
-            assert!(guard.is_none(), "tenant manifest set before preload"); // first populated here
-            *guard = preload.tenant_manifest;
+        if needs_manifest_upload {
+            self.store_tenant_manifest().await?;
        }
-        self.maybe_upload_tenant_manifest().await?;

        // The local filesystem contents are a cache of what's in the remote IndexPart;
        // IndexPart is the source of truth.
@@ -2235,7 +2218,7 @@ impl Tenant {
        };

        // Upload new list of offloaded timelines to S3
-        self.maybe_upload_tenant_manifest().await?;
+        self.store_tenant_manifest().await?;

        // Activate the timeline (if it makes sense)
        if !(timeline.is_broken() || timeline.is_stopping()) {
@@ -3265,23 +3248,17 @@ impl Tenant {
    async fn housekeeping(&self) {
        // Call through to all timelines to freeze ephemeral layers as needed. This usually happens
        // during ingest, but we don't want idle timelines to hold open layers for too long.
-        //
-        // We don't do this if the tenant can't upload layers (i.e. it's in stale attachment mode).
-        // We don't run compaction in this case either, and don't want to keep flushing tiny L0
-        // layers that won't be compacted down.
-        if self.tenant_conf.load().location.may_upload_layers_hint() {
-            let timelines = self
-                .timelines
-                .lock()
-                .unwrap()
-                .values()
-                .filter(|tli| tli.is_active())
-                .cloned()
-                .collect_vec();
+        let timelines = self
+            .timelines
+            .lock()
+            .unwrap()
+            .values()
+            .filter(|tli| tli.is_active())
+            .cloned()
+            .collect_vec();

-            for timeline in timelines {
-                timeline.maybe_freeze_ephemeral_layer().await;
-            }
+        for timeline in timelines {
+            timeline.maybe_freeze_ephemeral_layer().await;
        }

        // Shut down walredo if idle.
@@ -3446,7 +3423,7 @@ impl Tenant {
            shutdown_mode
        };

-        match self.set_stopping(shutdown_progress).await {
+        match self.set_stopping(shutdown_progress, false, false).await {
            Ok(()) => {}
            Err(SetStoppingError::Broken) => {
                // assume that this is acceptable
@@ -3526,13 +3503,25 @@ impl Tenant {
    /// This function waits for the tenant to become active if it isn't already, before transitioning it into Stopping state.
    ///
    /// This function is not cancel-safe!
-    async fn set_stopping(&self, progress: completion::Barrier) -> Result<(), SetStoppingError> {
+    ///
+    /// `allow_transition_from_loading` is needed for the special case of loading task deleting the tenant.
+    /// `allow_transition_from_attaching` is needed for the special case of attaching deleted tenant.
+    async fn set_stopping(
+        &self,
+        progress: completion::Barrier,
+        _allow_transition_from_loading: bool,
+        allow_transition_from_attaching: bool,
+    ) -> Result<(), SetStoppingError> {
        let mut rx = self.state.subscribe();

        // cannot stop before we're done activating, so wait out until we're done activating
        rx.wait_for(|state| match state {
+            TenantState::Attaching if allow_transition_from_attaching => true,
            TenantState::Activating(_) | TenantState::Attaching => {
-                info!("waiting for {state} to turn Active|Broken|Stopping");
+                info!(
+                    "waiting for {} to turn Active|Broken|Stopping",
+                    <&'static str>::from(state)
+                );
                false
            }
            TenantState::Active | TenantState::Broken { .. } | TenantState::Stopping { .. } => true,
@@ -3543,24 +3532,25 @@ impl Tenant {
        // we now know we're done activating, let's see whether this task is the winner to transition into Stopping
        let mut err = None;
        let stopping = self.state.send_if_modified(|current_state| match current_state {
-            TenantState::Activating(_) | TenantState::Attaching => {
-                unreachable!("we ensured above that we're done with activation, and, there is no re-activation")
+            TenantState::Activating(_) => {
+                unreachable!("1we ensured above that we're done with activation, and, there is no re-activation")
+            }
+            TenantState::Attaching => {
+                if !allow_transition_from_attaching {
+                    unreachable!("2we ensured above that we're done with activation, and, there is no re-activation")
+                };
+                *current_state = TenantState::Stopping { progress };
+                true
            }
            TenantState::Active => {
                // FIXME: due to time-of-check vs time-of-use issues, it can happen that new timelines
                // are created after the transition to Stopping. That's harmless, as the Timelines
                // won't be accessible to anyone afterwards, because the Tenant is in Stopping state.
-                *current_state = TenantState::Stopping { progress: Some(progress) };
+                *current_state = TenantState::Stopping { progress };
                // Continue stopping outside the closure. We need to grab timelines.lock()
                // and we plan to turn it into a tokio::sync::Mutex in a future patch.
                true
            }
-            TenantState::Stopping { progress: None } => {
-                // An attach was cancelled, and the attach transitioned the tenant from Attaching to
-                // Stopping(None) to let us know it exited. Register our progress and continue.
-                *current_state = TenantState::Stopping { progress: Some(progress) };
-                true
-            }
            TenantState::Broken { reason, .. } => {
                info!(
                    "Cannot set tenant to Stopping state, it is in Broken state due to: {reason}"
@@ -3568,7 +3558,7 @@ impl Tenant {
                err = Some(SetStoppingError::Broken);
                false
            }
-            TenantState::Stopping { progress: Some(progress) } => {
+            TenantState::Stopping { progress } => {
                info!("Tenant is already in Stopping state");
                err = Some(SetStoppingError::AlreadyStopping(progress.clone()));
                false
@@ -3693,7 +3683,7 @@ impl Tenant {
                        }
                    }
                }
-                TenantState::Active => {
+                TenantState::Active { .. } => {
                    return Ok(());
                }
                TenantState::Broken { reason, .. } => {
@@ -4069,20 +4059,18 @@ impl Tenant {

    /// Generate an up-to-date TenantManifest based on the state of this Tenant.
    fn build_tenant_manifest(&self) -> TenantManifest {
-        // Collect the offloaded timelines, and sort them for deterministic output.
-        let offloaded_timelines = self
-            .timelines_offloaded
-            .lock()
-            .unwrap()
-            .values()
-            .map(|tli| tli.manifest())
-            .sorted_by_key(|m| m.timeline_id)
-            .collect_vec();
+        let timelines_offloaded = self.timelines_offloaded.lock().unwrap();
+
+        let mut timeline_manifests = timelines_offloaded
+            .iter()
+            .map(|(_timeline_id, offloaded)| offloaded.manifest())
+            .collect::<Vec<_>>();
+        // Sort the manifests so that our output is deterministic
+        timeline_manifests.sort_by_key(|timeline_manifest| timeline_manifest.timeline_id);

        TenantManifest {
            version: LATEST_TENANT_MANIFEST_VERSION,
-            stripe_size: Some(self.get_shard_stripe_size()),
-            offloaded_timelines,
+            offloaded_timelines: timeline_manifests,
        }
    }

@@ -4211,9 +4199,9 @@ impl Tenant {
            self.cancel.child_token(),
        );

-        let timeline_ctx = RequestContextBuilder::from(ctx)
+        let timeline_ctx = RequestContextBuilder::extend(ctx)
            .scope(context::Scope::new_timeline(&timeline))
-            .detached_child();
+            .build();

        Ok((timeline, timeline_ctx))
    }
@@ -4305,7 +4293,7 @@ impl Tenant {
            timelines: Mutex::new(HashMap::new()),
            timelines_creating: Mutex::new(HashSet::new()),
            timelines_offloaded: Mutex::new(HashMap::new()),
-            remote_tenant_manifest: Default::default(),
+            tenant_manifest_upload: Default::default(),
            gc_cs: tokio::sync::Mutex::new(()),
            walredo_mgr,
            remote_storage,
@@ -4401,7 +4389,10 @@ impl Tenant {
        .to_string();

        fail::fail_point!("tenant-config-before-write", |_| {
-            Err(std::io::Error::other("tenant-config-before-write"))
+            Err(std::io::Error::new(
+                std::io::ErrorKind::Other,
+                "tenant-config-before-write",
+            ))
        });

        // Convert the config to a toml file.
@@ -5535,35 +5526,27 @@ impl Tenant {
            .unwrap_or(0)
    }

-    /// Builds a new tenant manifest, and uploads it if it differs from the last-known tenant
-    /// manifest in `Self::remote_tenant_manifest`.
-    ///
-    /// TODO: instead of requiring callers to remember to call `maybe_upload_tenant_manifest` after
-    /// changing any `Tenant` state that's included in the manifest, consider making the manifest
-    /// the authoritative source of data with an API that automatically uploads on changes. Revisit
-    /// this when the manifest is more widely used and we have a better idea of the data model.
-    pub(crate) async fn maybe_upload_tenant_manifest(&self) -> Result<(), TenantManifestError> {
-        // Multiple tasks may call this function concurrently after mutating the Tenant runtime
-        // state, affecting the manifest generated by `build_tenant_manifest`. We use an async mutex
-        // to serialize these callers. `eq_ignoring_version` acts as a slightly inefficient but
-        // simple coalescing mechanism.
+    /// Serialize and write the latest TenantManifest to remote storage.
+    pub(crate) async fn store_tenant_manifest(&self) -> Result<(), TenantManifestError> {
+        // Only one manifest write may be done at at time, and the contents of the manifest
+        // must be loaded while holding this lock. This makes it safe to call this function
+        // from anywhere without worrying about colliding updates.
        let mut guard = tokio::select! {
-            guard = self.remote_tenant_manifest.lock() => guard,
-            _ = self.cancel.cancelled() => return Err(TenantManifestError::Cancelled),
+            g = self.tenant_manifest_upload.lock() => {
+                g
+            },
+            _ = self.cancel.cancelled() => {
+                return Err(TenantManifestError::Cancelled);
+            }
        };

-        // Build a new manifest.
        let manifest = self.build_tenant_manifest();
-
-        // Check if the manifest has changed. We ignore the version number here, to avoid
-        // uploading every manifest on version number bumps.
-        if let Some(old) = guard.as_ref() {
-            if manifest.eq_ignoring_version(old) {
-                return Ok(());
-            }
+        if Some(&manifest) == (*guard).as_ref() {
+            // Optimisation: skip uploads that don't change anything.
+            return Ok(());
        }

-        // Upload the manifest. Remote storage does no retries internally, so retry here.
+        // Remote storage does no retries internally, so wrap it
        match backoff::retry(
            || async {
                upload_tenant_manifest(
@@ -5575,7 +5558,7 @@ impl Tenant {
                )
                .await
            },
-            |_| self.cancel.is_cancelled(),
+            |_e| self.cancel.is_cancelled(),
            FAILED_UPLOAD_WARN_THRESHOLD,
            FAILED_REMOTE_OP_RETRIES,
            "uploading tenant manifest",
@@ -8733,21 +8716,6 @@ mod tests {
                Lsn(0x20),
                Value::WalRecord(NeonWalRecord::wal_init("i")),
            ),
-            (
-                get_key(4),
-                Lsn(0x30),
-                Value::WalRecord(NeonWalRecord::wal_append_conditional("j", "i")),
-            ),
-            (
-                get_key(5),
-                Lsn(0x20),
-                Value::WalRecord(NeonWalRecord::wal_init("1")),
-            ),
-            (
-                get_key(5),
-                Lsn(0x30),
-                Value::WalRecord(NeonWalRecord::wal_append_conditional("j", "2")),
-            ),
        ];
        let image1 = vec![(get_key(1), "0x10".into())];

@@ -8778,18 +8746,8 @@ mod tests {

        // Need to remove the limit of "Neon WAL redo requires base image".

-        assert_eq!(
-            tline.get(get_key(3), Lsn(0x50), &ctx).await?,
-            Bytes::from_static(b"c")
-        );
-        assert_eq!(
-            tline.get(get_key(4), Lsn(0x50), &ctx).await?,
-            Bytes::from_static(b"ij")
-        );
-
-        // Manual testing required: currently, read errors will panic the process in debug mode. So we
-        // cannot enable this assertion in the unit test.
-        // assert!(tline.get(get_key(5), Lsn(0x50), &ctx).await.is_err());
+        // assert_eq!(tline.get(get_key(3), Lsn(0x50), &ctx).await?, Bytes::new());
+        // assert_eq!(tline.get(get_key(4), Lsn(0x50), &ctx).await?, Bytes::new());

        Ok(())
    }
--- a/pageserver/src/tenant/blob_io.rs
+++ b/pageserver/src/tenant/blob_io.rs
@@ -15,7 +15,7 @@
 //! len >= 128: 1CCCXXXX XXXXXXXX XXXXXXXX XXXXXXXX
 //!
 use std::cmp::min;
-use std::io::Error;
+use std::io::{Error, ErrorKind};

 use async_compression::Level;
 use bytes::{BufMut, BytesMut};
@@ -331,7 +331,10 @@ impl<const BUFFERED: bool> BlobWriter<BUFFERED> {
                    return (
                        (
                            io_buf.slice_len(),
-                            Err(Error::other(format!("blob too large ({len} bytes)"))),
+                            Err(Error::new(
+                                ErrorKind::Other,
+                                format!("blob too large ({len} bytes)"),
+                            )),
                        ),
                        srcbuf,
                    );
--- a/pageserver/src/tenant/block_io.rs
+++ b/pageserver/src/tenant/block_io.rs
@@ -216,8 +216,12 @@ impl<'a> FileBlockReader<'a> {
        match cache
            .read_immutable_buf(self.file_id, blknum, ctx)
            .await
-            .map_err(|e| std::io::Error::other(format!("Failed to read immutable buf: {e:#}")))?
-        {
+            .map_err(|e| {
+                std::io::Error::new(
+                    std::io::ErrorKind::Other,
+                    format!("Failed to read immutable buf: {e:#}"),
+                )
+            })? {
            ReadBufResult::Found(guard) => Ok(guard.into()),
            ReadBufResult::NotFound(write_guard) => {
                // Read the page from disk into the buffer
--- a/pageserver/src/tenant/layer_map/layer_coverage.rs
+++ b/pageserver/src/tenant/layer_map/layer_coverage.rs
@@ -53,7 +53,7 @@ impl<Value: Clone> LayerCoverage<Value> {
    ///
    /// Complexity: O(log N)
    fn add_node(&mut self, key: i128) {
-        let value = match self.nodes.range(..=key).next_back() {
+        let value = match self.nodes.range(..=key).last() {
            Some((_, Some(v))) => Some(v.clone()),
            Some((_, None)) => None,
            None => None,
--- a/pageserver/src/tenant/mgr.rs
+++ b/pageserver/src/tenant/mgr.rs
@@ -58,7 +58,7 @@ use crate::{InitializationOrder, TEMP_FILE_SUFFIX};

 /// For a tenant that appears in TenantsMap, it may either be
 /// - `Attached`: has a full Tenant object, is elegible to service
-///   reads and ingest WAL.
+///    reads and ingest WAL.
 /// - `Secondary`: is only keeping a local cache warm.
 ///
 /// Secondary is a totally distinct state rather than being a mode of a `Tenant`, because
--- a/pageserver/src/tenant/remote_timeline_client.rs
+++ b/pageserver/src/tenant/remote_timeline_client.rs
@@ -642,7 +642,6 @@ impl RemoteTimelineClient {
            cancel,
        )
        .measure_remote_op(
-            Option::<TaskKind>::None,
            RemoteOpFileKind::Index,
            RemoteOpKind::Download,
            Arc::clone(&self.metrics),
@@ -740,7 +739,6 @@ impl RemoteTimelineClient {
                ctx,
            )
            .measure_remote_op(
-                Some(ctx.task_kind()),
                RemoteOpFileKind::Layer,
                RemoteOpKind::Download,
                Arc::clone(&self.metrics),
@@ -2177,7 +2175,6 @@ impl RemoteTimelineClient {
                        &self.cancel,
                    )
                    .measure_remote_op(
-                        Some(TaskKind::RemoteUploadTask),
                        RemoteOpFileKind::Layer,
                        RemoteOpKind::Upload,
                        Arc::clone(&self.metrics),
@@ -2194,7 +2191,6 @@ impl RemoteTimelineClient {
                        &self.cancel,
                    )
                    .measure_remote_op(
-                        Some(TaskKind::RemoteUploadTask),
                        RemoteOpFileKind::Index,
                        RemoteOpKind::Upload,
                        Arc::clone(&self.metrics),
--- a/pageserver/src/tenant/remote_timeline_client/index.rs
+++ b/pageserver/src/tenant/remote_timeline_client/index.rs
@@ -130,7 +130,7 @@ impl IndexPart {
    /// Version history
    /// - 2: added `deleted_at`
    /// - 3: no longer deserialize `timeline_layers` (serialized format is the same, but timeline_layers
-    ///   is always generated from the keys of `layer_metadata`)
+    ///      is always generated from the keys of `layer_metadata`)
    /// - 4: timeline_layers is fully removed.
    /// - 5: lineage was added
    /// - 6: last_aux_file_policy is added.
--- a/pageserver/src/tenant/remote_timeline_client/manifest.rs
+++ b/pageserver/src/tenant/remote_timeline_client/manifest.rs
@@ -1,33 +1,21 @@
 use chrono::NaiveDateTime;
-use pageserver_api::shard::ShardStripeSize;
 use serde::{Deserialize, Serialize};
 use utils::id::TimelineId;
 use utils::lsn::Lsn;

-/// Tenant shard manifest, stored in remote storage. Contains offloaded timelines and other tenant
-/// shard-wide information that must be persisted in remote storage.
-///
-/// The manifest is always updated on tenant attach, and as needed.
-#[derive(Clone, Debug, Serialize, Deserialize, PartialEq, Eq)]
+/// Tenant-shard scoped manifest
+#[derive(Clone, Serialize, Deserialize, PartialEq, Eq)]
 pub struct TenantManifest {
-    /// The manifest version. Incremented on manifest format changes, even non-breaking ones.
-    /// Manifests must generally always be backwards and forwards compatible for one release, to
-    /// allow release rollbacks.
+    /// Debugging aid describing the version of this manifest.
+    /// Can also be used for distinguishing breaking changes later on.
    pub version: usize,

-    /// This tenant's stripe size. This is only advisory, and used to recover tenant data from
-    /// remote storage. The autoritative source is the storage controller. If None, assume the
-    /// original default value of 32768 blocks (256 MB).
-    #[serde(skip_serializing_if = "Option::is_none")]
-    pub stripe_size: Option<ShardStripeSize>,
-
    /// The list of offloaded timelines together with enough information
    /// to not have to actually load them.
    ///
    /// Note: the timelines mentioned in this list might be deleted, i.e.
    /// we don't hold an invariant that the references aren't dangling.
    /// Existence of index-part.json is the actual indicator of timeline existence.
-    #[serde(default)]
    pub offloaded_timelines: Vec<OffloadedTimelineManifest>,
 }

@@ -36,7 +24,7 @@ pub struct TenantManifest {
 /// Very similar to [`pageserver_api::models::OffloadedTimelineInfo`],
 /// but the two datastructures serve different needs, this is for a persistent disk format
 /// that must be backwards compatible, while the other is only for informative purposes.
-#[derive(Clone, Debug, Serialize, Deserialize, Copy, PartialEq, Eq)]
+#[derive(Clone, Serialize, Deserialize, Copy, PartialEq, Eq)]
 pub struct OffloadedTimelineManifest {
    pub timeline_id: TimelineId,
    /// Whether the timeline has a parent it has been branched off from or not
@@ -47,166 +35,20 @@ pub struct OffloadedTimelineManifest {
    pub archived_at: NaiveDateTime,
 }

-/// The newest manifest version. This should be incremented on changes, even non-breaking ones. We
-/// do not use deny_unknown_fields, so new fields are not breaking.
-///
-/// 1: initial version
-/// 2: +stripe_size
-///
-/// When adding new versions, also add a parse_vX test case below.
-pub const LATEST_TENANT_MANIFEST_VERSION: usize = 2;
+pub const LATEST_TENANT_MANIFEST_VERSION: usize = 1;

 impl TenantManifest {
-    /// Returns true if the manifests are equal, ignoring the version number. This avoids
-    /// re-uploading all manifests just because the version number is bumped.
-    pub fn eq_ignoring_version(&self, other: &Self) -> bool {
-        // Fast path: if the version is equal, just compare directly.
-        if self.version == other.version {
-            return self == other;
+    pub(crate) fn empty() -> Self {
+        Self {
+            version: LATEST_TENANT_MANIFEST_VERSION,
+            offloaded_timelines: vec![],
        }
-
-        // We could alternatively just clone and modify the version here.
-        let Self {
-            version: _, // ignore version
-            stripe_size,
-            offloaded_timelines,
-        } = self;
-
-        stripe_size == &other.stripe_size && offloaded_timelines == &other.offloaded_timelines
    }
-
-    /// Decodes a manifest from JSON.
    pub fn from_json_bytes(bytes: &[u8]) -> Result<Self, serde_json::Error> {
-        serde_json::from_slice(bytes)
+        serde_json::from_slice::<Self>(bytes)
    }

-    /// Encodes a manifest as JSON.
-    pub fn to_json_bytes(&self) -> serde_json::Result<Vec<u8>> {
+    pub(crate) fn to_json_bytes(&self) -> serde_json::Result<Vec<u8>> {
        serde_json::to_vec(self)
    }
 }
-
-#[cfg(test)]
-mod tests {
-    use std::str::FromStr;
-
-    use utils::id::TimelineId;
-
-    use super::*;
-
-    /// Empty manifests should be parsed. Version is required.
-    #[test]
-    fn parse_empty() -> anyhow::Result<()> {
-        let json = r#"{
-             "version": 0
-         }"#;
-        let expected = TenantManifest {
-            version: 0,
-            stripe_size: None,
-            offloaded_timelines: Vec::new(),
-        };
-        assert_eq!(expected, TenantManifest::from_json_bytes(json.as_bytes())?);
-        Ok(())
-    }
-
-    /// Unknown fields should be ignored, for forwards compatibility.
-    #[test]
-    fn parse_unknown_fields() -> anyhow::Result<()> {
-        let json = r#"{
-             "version": 1,
-             "foo": "bar"
-         }"#;
-        let expected = TenantManifest {
-            version: 1,
-            stripe_size: None,
-            offloaded_timelines: Vec::new(),
-        };
-        assert_eq!(expected, TenantManifest::from_json_bytes(json.as_bytes())?);
-        Ok(())
-    }
-
-    /// v1 manifests should be parsed, for backwards compatibility.
-    #[test]
-    fn parse_v1() -> anyhow::Result<()> {
-        let json = r#"{
-             "version": 1,
-             "offloaded_timelines": [
-                 {
-                     "timeline_id": "5c4df612fd159e63c1b7853fe94d97da",
-                     "archived_at": "2025-03-07T11:07:11.373105434"
-                 },
-                 {
-                     "timeline_id": "f3def5823ad7080d2ea538d8e12163fa",
-                     "ancestor_timeline_id": "5c4df612fd159e63c1b7853fe94d97da",
-                     "ancestor_retain_lsn": "0/1F79038",
-                     "archived_at": "2025-03-05T11:10:22.257901390"
-                 }
-             ]
-         }"#;
-        let expected = TenantManifest {
-            version: 1,
-            stripe_size: None,
-            offloaded_timelines: vec![
-                OffloadedTimelineManifest {
-                    timeline_id: TimelineId::from_str("5c4df612fd159e63c1b7853fe94d97da")?,
-                    ancestor_timeline_id: None,
-                    ancestor_retain_lsn: None,
-                    archived_at: NaiveDateTime::from_str("2025-03-07T11:07:11.373105434")?,
-                },
-                OffloadedTimelineManifest {
-                    timeline_id: TimelineId::from_str("f3def5823ad7080d2ea538d8e12163fa")?,
-                    ancestor_timeline_id: Some(TimelineId::from_str(
-                        "5c4df612fd159e63c1b7853fe94d97da",
-                    )?),
-                    ancestor_retain_lsn: Some(Lsn::from_str("0/1F79038")?),
-                    archived_at: NaiveDateTime::from_str("2025-03-05T11:10:22.257901390")?,
-                },
-            ],
-        };
-        assert_eq!(expected, TenantManifest::from_json_bytes(json.as_bytes())?);
-        Ok(())
-    }
-
-    /// v2 manifests should be parsed, for backwards compatibility.
-    #[test]
-    fn parse_v2() -> anyhow::Result<()> {
-        let json = r#"{
-             "version": 2,
-             "stripe_size": 32768,
-             "offloaded_timelines": [
-                 {
-                     "timeline_id": "5c4df612fd159e63c1b7853fe94d97da",
-                     "archived_at": "2025-03-07T11:07:11.373105434"
-                 },
-                 {
-                     "timeline_id": "f3def5823ad7080d2ea538d8e12163fa",
-                     "ancestor_timeline_id": "5c4df612fd159e63c1b7853fe94d97da",
-                     "ancestor_retain_lsn": "0/1F79038",
-                     "archived_at": "2025-03-05T11:10:22.257901390"
-                 }
-             ]
-         }"#;
-        let expected = TenantManifest {
-            version: 2,
-            stripe_size: Some(ShardStripeSize(32768)),
-            offloaded_timelines: vec![
-                OffloadedTimelineManifest {
-                    timeline_id: TimelineId::from_str("5c4df612fd159e63c1b7853fe94d97da")?,
-                    ancestor_timeline_id: None,
-                    ancestor_retain_lsn: None,
-                    archived_at: NaiveDateTime::from_str("2025-03-07T11:07:11.373105434")?,
-                },
-                OffloadedTimelineManifest {
-                    timeline_id: TimelineId::from_str("f3def5823ad7080d2ea538d8e12163fa")?,
-                    ancestor_timeline_id: Some(TimelineId::from_str(
-                        "5c4df612fd159e63c1b7853fe94d97da",
-                    )?),
-                    ancestor_retain_lsn: Some(Lsn::from_str("0/1F79038")?),
-                    archived_at: NaiveDateTime::from_str("2025-03-05T11:10:22.257901390")?,
-                },
-            ],
-        };
-        assert_eq!(expected, TenantManifest::from_json_bytes(json.as_bytes())?);
-        Ok(())
-    }
-}
--- a/pageserver/src/tenant/remote_timeline_client/upload.rs
+++ b/pageserver/src/tenant/remote_timeline_client/upload.rs
@@ -61,7 +61,6 @@ pub(crate) async fn upload_index_part(
        .await
        .with_context(|| format!("upload index part for '{tenant_shard_id} / {timeline_id}'"))
 }
-
 /// Serializes and uploads the given tenant manifest data to the remote storage.
 pub(crate) async fn upload_tenant_manifest(
    storage: &GenericRemoteStorage,
@@ -77,14 +76,16 @@ pub(crate) async fn upload_tenant_manifest(
    });
    pausable_failpoint!("before-upload-manifest-pausable");

-    let serialized = Bytes::from(tenant_manifest.to_json_bytes()?);
-    let tenant_manifest_size = serialized.len();
-    let remote_path = remote_tenant_manifest_path(tenant_shard_id, generation);
+    let serialized = tenant_manifest.to_json_bytes()?;
+    let serialized = Bytes::from(serialized);

+    let tenant_manifest_site = serialized.len();
+
+    let remote_path = remote_tenant_manifest_path(tenant_shard_id, generation);
    storage
        .upload_storage_object(
            futures::stream::once(futures::future::ready(Ok(serialized))),
-            tenant_manifest_size,
+            tenant_manifest_site,
            &remote_path,
            cancel,
        )
--- a/pageserver/src/tenant/secondary.rs
+++ b/pageserver/src/tenant/secondary.rs
@@ -167,17 +167,10 @@ impl SecondaryTenant {

        self.validate_metrics();

-        // Metrics are subtracted from and/or removed eagerly.
-        // Deletions are done in the background via [`BackgroundPurges::spawn`].
        let tenant_id = self.tenant_shard_id.tenant_id.to_string();
        let shard_id = format!("{}", self.tenant_shard_id.shard_slug());
        let _ = SECONDARY_RESIDENT_PHYSICAL_SIZE.remove_label_values(&[&tenant_id, &shard_id]);
        let _ = SECONDARY_HEATMAP_TOTAL_SIZE.remove_label_values(&[&tenant_id, &shard_id]);
-
-        self.detail
-            .lock()
-            .unwrap()
-            .drain_timelines(&self.tenant_shard_id, &self.resident_size_metric);
    }

    pub(crate) fn set_config(&self, config: &SecondaryLocationConfig) {
--- a/pageserver/src/tenant/secondary/downloader.rs
+++ b/pageserver/src/tenant/secondary/downloader.rs
@@ -4,7 +4,6 @@ use std::str::FromStr;
 use std::sync::Arc;
 use std::time::{Duration, Instant, SystemTime};

-use crate::metrics::{STORAGE_IO_SIZE, StorageIoSizeOperation};
 use camino::Utf8PathBuf;
 use chrono::format::{DelayedFormat, StrftimeItems};
 use futures::Future;
@@ -125,53 +124,15 @@ impl OnDiskState {
    }
 }

+#[derive(Debug, Clone, Default)]
 pub(super) struct SecondaryDetailTimeline {
    on_disk_layers: HashMap<LayerName, OnDiskState>,

    /// We remember when layers were evicted, to prevent re-downloading them.
    pub(super) evicted_at: HashMap<LayerName, SystemTime>,
-
-    ctx: RequestContext,
-}
-
-impl Clone for SecondaryDetailTimeline {
-    fn clone(&self) -> Self {
-        Self {
-            on_disk_layers: self.on_disk_layers.clone(),
-            evicted_at: self.evicted_at.clone(),
-            // This is a bit awkward. The downloader code operates on a snapshot
-            // of the secondary list to avoid locking it for extended periods of time.
-            // No particularly strong reason to chose [`RequestContext::detached_child`],
-            // but makes more sense than [`RequestContext::attached_child`].
-            ctx: self
-                .ctx
-                .detached_child(self.ctx.task_kind(), self.ctx.download_behavior()),
-        }
-    }
-}
-
-impl std::fmt::Debug for SecondaryDetailTimeline {
-    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
-        f.debug_struct("SecondaryDetailTimeline")
-            .field("on_disk_layers", &self.on_disk_layers)
-            .field("evicted_at", &self.evicted_at)
-            .finish()
-    }
 }

 impl SecondaryDetailTimeline {
-    pub(super) fn empty(ctx: RequestContext) -> Self {
-        SecondaryDetailTimeline {
-            on_disk_layers: Default::default(),
-            evicted_at: Default::default(),
-            ctx,
-        }
-    }
-
-    pub(super) fn context(&self) -> &RequestContext {
-        &self.ctx
-    }
-
    pub(super) fn remove_layer(
        &mut self,
        name: &LayerName,
@@ -297,50 +258,18 @@ impl SecondaryDetail {

    pub(super) fn remove_timeline(
        &mut self,
-        tenant_shard_id: &TenantShardId,
        timeline_id: &TimelineId,
        resident_metric: &UIntGauge,
    ) {
        let removed = self.timelines.remove(timeline_id);
        if let Some(removed) = removed {
-            Self::clear_timeline_metrics(tenant_shard_id, timeline_id, removed, resident_metric);
-        }
-    }
-
-    pub(super) fn drain_timelines(
-        &mut self,
-        tenant_shard_id: &TenantShardId,
-        resident_metric: &UIntGauge,
-    ) {
-        for (timeline_id, removed) in self.timelines.drain() {
-            Self::clear_timeline_metrics(tenant_shard_id, &timeline_id, removed, resident_metric);
-        }
-    }
-
-    fn clear_timeline_metrics(
-        tenant_shard_id: &TenantShardId,
-        timeline_id: &TimelineId,
-        detail: SecondaryDetailTimeline,
-        resident_metric: &UIntGauge,
-    ) {
-        resident_metric.sub(
-            detail
-                .on_disk_layers
-                .values()
-                .map(|l| l.metadata.file_size)
-                .sum(),
-        );
-
-        let shard_id = format!("{}", tenant_shard_id.shard_slug());
-        let tenant_id = tenant_shard_id.tenant_id.to_string();
-        let timeline_id = timeline_id.to_string();
-        for op in StorageIoSizeOperation::VARIANTS {
-            let _ = STORAGE_IO_SIZE.remove_label_values(&[
-                op,
-                tenant_id.as_str(),
-                shard_id.as_str(),
-                timeline_id.as_str(),
-            ]);
+            resident_metric.sub(
+                removed
+                    .on_disk_layers
+                    .values()
+                    .map(|l| l.metadata.file_size)
+                    .sum(),
+            );
        }
    }

@@ -798,7 +727,6 @@ impl<'a> TenantDownloader<'a> {
                        last_heatmap,
                        timeline,
                        &self.secondary_state.resident_size_metric,
-                        ctx,
                    )
                    .await;

@@ -846,6 +774,7 @@ impl<'a> TenantDownloader<'a> {

        // Download the layers in the heatmap
        for timeline in heatmap.timelines {
+            let ctx = &ctx.with_scope_secondary_timeline(tenant_shard_id, &timeline.timeline_id);
            let timeline_state = timeline_states
                .remove(&timeline.timeline_id)
                .expect("Just populated above");
@@ -988,11 +917,7 @@ impl<'a> TenantDownloader<'a> {
            for delete_timeline in &delete_timelines {
                // We haven't removed from disk yet, but optimistically remove from in-memory state: if removal
                // from disk fails that will be a fatal error.
-                detail.remove_timeline(
-                    self.secondary_state.get_tenant_shard_id(),
-                    delete_timeline,
-                    &self.secondary_state.resident_size_metric,
-                );
+                detail.remove_timeline(delete_timeline, &self.secondary_state.resident_size_metric);
            }
        }

@@ -1088,6 +1013,7 @@ impl<'a> TenantDownloader<'a> {
        timeline: HeatMapTimeline,
        timeline_state: SecondaryDetailTimeline,
        deadline: Instant,
+        ctx: &RequestContext,
    ) -> (Result<(), UpdateError>, Vec<HeatMapLayer>) {
        // Accumulate updates to the state
        let mut touched = Vec::new();
@@ -1118,12 +1044,7 @@ impl<'a> TenantDownloader<'a> {
            }

            match self
-                .download_layer(
-                    tenant_shard_id,
-                    &timeline_id,
-                    layer,
-                    timeline_state.context(),
-                )
+                .download_layer(tenant_shard_id, &timeline_id, layer, ctx)
                .await
            {
                Ok(Some(layer)) => touched.push(layer),
@@ -1234,16 +1155,13 @@ impl<'a> TenantDownloader<'a> {
        tracing::debug!(timeline_id=%timeline_id, "Downloading layers, {} in heatmap", timeline.hot_layers().count());

        let (result, touched) = self
-            .download_timeline_layers(tenant_shard_id, timeline, timeline_state, deadline)
+            .download_timeline_layers(tenant_shard_id, timeline, timeline_state, deadline, ctx)
            .await;

        // Write updates to state to record layers we just downloaded or touched, irrespective of whether the overall result was successful
        {
            let mut detail = self.secondary_state.detail.lock().unwrap();
-            let timeline_detail = detail.timelines.entry(timeline_id).or_insert_with(|| {
-                let ctx = ctx.with_scope_secondary_timeline(tenant_shard_id, &timeline_id);
-                SecondaryDetailTimeline::empty(ctx)
-            });
+            let timeline_detail = detail.timelines.entry(timeline_id).or_default();

            tracing::info!("Wrote timeline_detail for {} touched layers", touched.len());
            touched.into_iter().for_each(|t| {
@@ -1377,12 +1295,10 @@ async fn init_timeline_state(
    last_heatmap: Option<&HeatMapTimeline>,
    heatmap: &HeatMapTimeline,
    resident_metric: &UIntGauge,
-    ctx: &RequestContext,
 ) -> SecondaryDetailTimeline {
-    let ctx = ctx.with_scope_secondary_timeline(tenant_shard_id, &heatmap.timeline_id);
-    let mut detail = SecondaryDetailTimeline::empty(ctx);
-
    let timeline_path = conf.timeline_path(tenant_shard_id, &heatmap.timeline_id);
+    let mut detail = SecondaryDetailTimeline::default();
+
    let mut dir = match tokio::fs::read_dir(&timeline_path).await {
        Ok(d) => d,
        Err(e) => {
--- a/pageserver/src/tenant/storage_layer.rs
+++ b/pageserver/src/tenant/storage_layer.rs
@@ -13,13 +13,13 @@ pub mod merge_iterator;
 use std::cmp::Ordering;
 use std::collections::hash_map::Entry;
 use std::collections::{BinaryHeap, HashMap};
+use std::future::Future;
 use std::ops::Range;
 use std::pin::Pin;
 use std::sync::Arc;
 use std::sync::atomic::AtomicUsize;
 use std::time::{Duration, SystemTime, UNIX_EPOCH};

-use crate::PERF_TRACE_TARGET;
 pub use batch_split_writer::{BatchLayerWriter, SplitDeltaLayerWriter, SplitImageLayerWriter};
 use bytes::Bytes;
 pub use delta_layer::{DeltaLayer, DeltaLayerWriter, ValueRef};
@@ -34,7 +34,7 @@ use pageserver_api::key::Key;
 use pageserver_api::keyspace::{KeySpace, KeySpaceRandomAccum};
 use pageserver_api::record::NeonWalRecord;
 use pageserver_api::value::Value;
-use tracing::{Instrument, info_span, trace};
+use tracing::{Instrument, trace};
 use utils::lsn::Lsn;
 use utils::sync::gate::GateGuard;

@@ -43,9 +43,7 @@ use super::PageReconstructError;
 use super::layer_map::InMemoryLayerDesc;
 use super::timeline::{GetVectoredError, ReadPath};
 use crate::config::PageServerConf;
-use crate::context::{
-    AccessStatsBehavior, PerfInstrumentFutureExt, RequestContext, RequestContextBuilder,
-};
+use crate::context::{AccessStatsBehavior, RequestContext};

 pub fn range_overlaps<T>(a: &Range<T>, b: &Range<T>) -> bool
 where
@@ -876,37 +874,13 @@ impl ReadableLayer {
    ) -> Result<(), GetVectoredError> {
        match self {
            ReadableLayer::PersistentLayer(layer) => {
-                let ctx = RequestContextBuilder::from(ctx)
-                    .perf_span(|crnt_perf_span| {
-                        info_span!(
-                            target: PERF_TRACE_TARGET,
-                            parent: crnt_perf_span,
-                            "PLAN_LAYER",
-                            layer = %layer
-                        )
-                    })
-                    .attached_child();
-
                layer
-                    .get_values_reconstruct_data(keyspace, lsn_range, reconstruct_state, &ctx)
-                    .maybe_perf_instrument(&ctx, |crnt_perf_span| crnt_perf_span.clone())
+                    .get_values_reconstruct_data(keyspace, lsn_range, reconstruct_state, ctx)
                    .await
            }
            ReadableLayer::InMemoryLayer(layer) => {
-                let ctx = RequestContextBuilder::from(ctx)
-                    .perf_span(|crnt_perf_span| {
-                        info_span!(
-                            target: PERF_TRACE_TARGET,
-                            parent: crnt_perf_span,
-                            "PLAN_LAYER",
-                            layer = %layer
-                        )
-                    })
-                    .attached_child();
-
                layer
-                    .get_values_reconstruct_data(keyspace, lsn_range, reconstruct_state, &ctx)
-                    .maybe_perf_instrument(&ctx, |crnt_perf_span| crnt_perf_span.clone())
+                    .get_values_reconstruct_data(keyspace, lsn_range, reconstruct_state, ctx)
                    .await
            }
        }
--- a/pageserver/src/tenant/storage_layer/batch_split_writer.rs
+++ b/pageserver/src/tenant/storage_layer/batch_split_writer.rs
@@ -366,7 +366,7 @@ impl SplitDeltaLayerWriter {
                )
                .await?;
                let (start_key, prev_delta_writer) =
-                    self.inner.replace((key, next_delta_writer)).unwrap();
+                    std::mem::replace(&mut self.inner, Some((key, next_delta_writer))).unwrap();
                self.batches.add_unfinished_delta_writer(
                    prev_delta_writer,
                    start_key..key,
--- a/pageserver/src/tenant/storage_layer/delta_layer.rs
+++ b/pageserver/src/tenant/storage_layer/delta_layer.rs
@@ -896,9 +896,9 @@ impl DeltaLayerInner {
    where
        Reader: BlockReader + Clone,
    {
-        let ctx = RequestContextBuilder::from(ctx)
+        let ctx = RequestContextBuilder::extend(ctx)
            .page_content_kind(PageContentKind::DeltaLayerBtreeNode)
-            .attached_child();
+            .build();

        for range in keyspace.ranges.iter() {
            let mut range_end_handled = false;
@@ -1105,9 +1105,9 @@ impl DeltaLayerInner {
                    all_keys.push(entry);
                    true
                },
-                &RequestContextBuilder::from(ctx)
+                &RequestContextBuilder::extend(ctx)
                    .page_content_kind(PageContentKind::DeltaLayerBtreeNode)
-                    .attached_child(),
+                    .build(),
            )
            .await?;
        if let Some(last) = all_keys.last_mut() {
--- a/pageserver/src/tenant/storage_layer/image_layer.rs
+++ b/pageserver/src/tenant/storage_layer/image_layer.rs
@@ -481,9 +481,9 @@ impl ImageLayerInner {
        let tree_reader =
            DiskBtreeReader::new(self.index_start_blk, self.index_root_blk, block_reader);

-        let ctx = RequestContextBuilder::from(ctx)
+        let ctx = RequestContextBuilder::extend(ctx)
            .page_content_kind(PageContentKind::ImageLayerBtreeNode)
-            .attached_child();
+            .build();

        for range in keyspace.ranges.iter() {
            let mut range_end_handled = false;
--- a/Show More
+++ b/Show More
Author	SHA1	Message	Date
Conrad Ludgate	b0411e612a	use flag-bearer	2025-04-02 12:51:27 +01:00
Conrad Ludgate	0579533ac8	remove dynamic limiter internal Box<dyn>	2025-04-02 12:45:04 +01:00
Conrad Ludgate	99c0da9607	proxy: simplify dynamic limiter impl	2025-04-02 12:40:45 +01:00
				`@@ -1 +0,0 @@`
				`SELECT lfc_value AS lfc_used_pages FROM neon.neon_lfc_stats WHERE lfc_key = 'file_cache_used_pages';`