nomerge: simulate run-kind=storage-rc-pr

.github/actionlint.yml

@@ -32,6 +32,3 @@ config-variables:
   - NEON_DEV_AWS_ACCOUNT_ID
   - NEON_PROD_AWS_ACCOUNT_ID
   - AWS_ECR_REGION
-  - BENCHMARK_LARGE_OLTP_PROJECTID
-  - SLACK_ON_CALL_DEVPROD_STREAM
-  - SLACK_RUST_CHANNEL_ID

.github/actions/neon-branch-create/action.yml

@@ -84,13 +84,7 @@ runs:
           --header "Authorization: Bearer ${API_KEY}"
           )
 
-        role_name=$(echo "$roles" | jq --raw-output '
-          (.roles | map(select(.protected == false))) as $roles |
-          if any($roles[]; .name == "neondb_owner")
-          then "neondb_owner"
-          else $roles[0].name
-          end
-        ')
+        role_name=$(echo $roles | jq --raw-output '.roles[] | select(.protected == false) | .name')
         echo "role_name=${role_name}" >> $GITHUB_OUTPUT
       env:
         API_HOST: ${{ inputs.api_host }}
@@ -113,13 +107,13 @@ runs:
             )
 
           if [ -z "${reset_password}" ]; then
-            sleep $i
+            sleep 1
             continue
           fi
 
           password=$(echo $reset_password | jq --raw-output '.role.password')
           if [ "${password}" == "null" ]; then
-            sleep $i # increasing backoff
+            sleep 1
             continue
           fi
 

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

@@ -44,11 +44,6 @@ inputs:
     description: 'Postgres version to use for tests'
     required: false
     default: 'v16'
-  sanitizers:
-    description: 'enabled or disabled'
-    required: false
-    default: 'disabled'
-    type: string
   benchmark_durations:
     description: 'benchmark durations JSON'
     required: false
@@ -64,7 +59,7 @@ runs:
       if: inputs.build_type != 'remote'
       uses: ./.github/actions/download
       with:
-        name: neon-${{ runner.os }}-${{ runner.arch }}-${{ inputs.build_type }}${{ inputs.sanitizers == 'enabled' && '-sanitized' || '' }}-artifact
+        name: neon-${{ runner.os }}-${{ runner.arch }}-${{ inputs.build_type }}-artifact
         path: /tmp/neon
         aws-oicd-role-arn: ${{ inputs.aws-oicd-role-arn }}
 
@@ -117,7 +112,6 @@ runs:
         ALLOW_FORWARD_COMPATIBILITY_BREAKAGE: contains(github.event.pull_request.labels.*.name, 'forward compatibility breakage')
         RERUN_FAILED: ${{ inputs.rerun_failed }}
         PG_VERSION: ${{ inputs.pg_version }}
-        SANITIZERS: ${{ inputs.sanitizers }}
       shell: bash -euxo pipefail {0}
       run: |
         # PLATFORM will be embedded in the perf test report

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

@@ -280,7 +280,7 @@ jobs:
       - name: Upload Neon artifact
         uses: ./.github/actions/upload
         with:
-          name: neon-${{ runner.os }}-${{ runner.arch }}-${{ inputs.build-type }}${{ inputs.sanitizers == 'enabled' && '-sanitized' || '' }}-artifact
+          name: neon-${{ runner.os }}-${{ runner.arch }}-${{ inputs.build-type }}-artifact
           path: /tmp/neon
           aws-oicd-role-arn: ${{ vars.DEV_AWS_OIDC_ROLE_ARN }}
 
@@ -347,7 +347,6 @@ jobs:
           real_s3_region: eu-central-1
           rerun_failed: true
           pg_version: ${{ matrix.pg_version }}
-          sanitizers: ${{ inputs.sanitizers }}
           aws-oicd-role-arn: ${{ vars.DEV_AWS_OIDC_ROLE_ARN }}
           # `--session-timeout` is equal to (timeout-minutes - 10 minutes) * 60 seconds.
           # Attempt to stop tests gracefully to generate test reports
@@ -360,6 +359,7 @@ jobs:
           PAGESERVER_VIRTUAL_FILE_IO_ENGINE: tokio-epoll-uring
           PAGESERVER_GET_VECTORED_CONCURRENT_IO: sidecar-task
           USE_LFC: ${{ matrix.lfc_state == 'with-lfc' && 'true' || 'false' }}
+          SANITIZERS: ${{ inputs.sanitizers }}
 
       # Temporary disable this step until we figure out why it's so flaky
       # Ref https://github.com/neondatabase/neon/issues/4540

.github/workflows/_meta.yml

@@ -19,7 +19,7 @@ on:
         description: "Tag of the last compute release"
         value: ${{ jobs.tags.outputs.compute }}
       run-kind:
-        description: "The kind of run we're currently in. Will be one of `push-main`, `storage-release`, `compute-release`, `proxy-release`, `storage-rc-pr`, `compute-rc-pr`,  `proxy-rc-pr`, `pr`, or `workflow-dispatch`"
+        description: "The kind of run we're currently in. Will be one of `pr`, `push-main`, `storage-rc`, `storage-release`, `proxy-rc`, `proxy-release`, `compute-rc`, `compute-release` or `merge_queue`"
         value: ${{ jobs.tags.outputs.run-kind }}
 
 permissions: {}
@@ -46,7 +46,7 @@ jobs:
         env:
           RUN_KIND: >-
             ${{
-              false
+              'storage-rc-pr'
               || (inputs.github-event-name == 'push'         && github.ref_name == 'main')            && 'push-main'
               || (inputs.github-event-name == 'push'         && github.ref_name == 'release')         && 'storage-release'
               || (inputs.github-event-name == 'push'         && github.ref_name == 'release-compute') && 'compute-release'
@@ -55,7 +55,6 @@ jobs:
               || (inputs.github-event-name == 'pull_request' && github.base_ref == 'release-compute') && 'compute-rc-pr'
               || (inputs.github-event-name == 'pull_request' && github.base_ref == 'release-proxy')   && 'proxy-rc-pr'
               || (inputs.github-event-name == 'pull_request')                                         && 'pr'
-              || (inputs.github-event-name == 'workflow_dispatch')                                    && 'workflow-dispatch'
               || 'unknown'
             }}
         run: |
@@ -86,9 +85,6 @@ jobs:
             BUILD_AND_TEST_RUN_ID=$(gh run list -b $CURRENT_BRANCH -c $CURRENT_SHA -w 'Build and Test' -L 1 --json databaseId --jq '.[].databaseId')
             echo "tag=$BUILD_AND_TEST_RUN_ID" | tee -a $GITHUB_OUTPUT
             ;;
-          workflow-dispatch)
-            echo "tag=$GITHUB_RUN_ID" | tee -a $GITHUB_OUTPUT
-            ;;
           *)
             echo "Unexpected RUN_KIND ('${RUN_KIND}'), failing to assign build-tag!"
             exit 1

.github/workflows/benchmarking.yml

@@ -141,8 +141,6 @@ jobs:
           --ignore test_runner/performance/test_physical_replication.py
           --ignore test_runner/performance/test_perf_ingest_using_pgcopydb.py
           --ignore test_runner/performance/test_cumulative_statistics_persistence.py
-          --ignore test_runner/performance/test_perf_many_relations.py
-          --ignore test_runner/performance/test_perf_oltp_large_tenant.py
       env:
         BENCHMARK_CONNSTR: ${{ steps.create-neon-project.outputs.dsn }}
         VIP_VAP_ACCESS_TOKEN: "${{ secrets.VIP_VAP_ACCESS_TOKEN }}"

.github/workflows/build_and_test.yml

@@ -692,15 +692,15 @@ jobs:
                                              neondatabase/neon-test-extensions-${{ matrix.version.pg }}:${{ needs.meta.outputs.build-tag }}-${{ matrix.version.debian }}-x64 \
                                              neondatabase/neon-test-extensions-${{ matrix.version.pg }}:${{ needs.meta.outputs.build-tag }}-${{ matrix.version.debian }}-arm64
 
-  vm-compute-node-image-arch:
+  vm-compute-node-image:
     needs: [ check-permissions, meta, compute-node-image ]
     if: ${{ contains(fromJSON('["push-main", "pr", "compute-release", "compute-rc-pr"]'), needs.meta.outputs.run-kind) }}
-    runs-on: ${{ fromJson(format('["self-hosted", "{0}"]', matrix.arch == 'arm64' && 'large-arm64' || 'large')) }}
+    runs-on: [ self-hosted, large ]
     strategy:
       fail-fast: false
       matrix:
-        arch: [ amd64, arm64 ]
         version:
+          # see the comment for `compute-node-image-arch` job
           - pg: v14
             debian: bullseye
           - pg: v15
@@ -717,7 +717,7 @@ jobs:
 
       - name: Downloading vm-builder
         run: |
-          curl -fL https://github.com/neondatabase/autoscaling/releases/download/$VM_BUILDER_VERSION/vm-builder-${{ matrix.arch }} -o vm-builder
+          curl -fL https://github.com/neondatabase/autoscaling/releases/download/$VM_BUILDER_VERSION/vm-builder-amd64 -o vm-builder
           chmod +x vm-builder
 
       - uses: neondatabase/dev-actions/set-docker-config-dir@6094485bf440001c94a94a3f9e221e81ff6b6193
@@ -738,37 +738,12 @@ jobs:
             -size=2G \
             -spec=compute/vm-image-spec-${{ matrix.version.debian }}.yaml \
             -src=neondatabase/compute-node-${{ matrix.version.pg }}:${{ needs.meta.outputs.build-tag }} \
-            -dst=neondatabase/vm-compute-node-${{ matrix.version.pg }}:${{ needs.meta.outputs.build-tag }}-${{ matrix.arch }} \
-            -target-arch=linux/${{ matrix.arch }}
+            -dst=neondatabase/vm-compute-node-${{ matrix.version.pg }}:${{ needs.meta.outputs.build-tag }} \
+            -target-arch=linux/amd64
 
       - name: Pushing vm-compute-node image
         run: |
-          docker push neondatabase/vm-compute-node-${{ matrix.version.pg }}:${{ needs.meta.outputs.build-tag }}-${{ matrix.arch }}
-
-  vm-compute-node-image:
-    needs: [ vm-compute-node-image-arch, meta ]
-    if: ${{ contains(fromJSON('["push-main", "pr", "compute-release", "compute-rc-pr"]'), needs.meta.outputs.run-kind) }}
-    runs-on: ubuntu-22.04
-    strategy:
-      matrix:
-        version:
-          # see the comment for `compute-node-image-arch` job
-          - pg: v14
-          - pg: v15
-          - pg: v16
-          - pg: v17
-    steps:
-      - uses: docker/login-action@v3
-        with:
-          username: ${{ secrets.NEON_DOCKERHUB_USERNAME }}
-          password: ${{ secrets.NEON_DOCKERHUB_PASSWORD }}
-
-      - name: Create multi-arch compute-node image
-        run: |
-          docker buildx imagetools create -t neondatabase/vm-compute-node-${{ matrix.version.pg }}:${{ needs.meta.outputs.build-tag }} \
-                                             neondatabase/vm-compute-node-${{ matrix.version.pg }}:${{ needs.meta.outputs.build-tag }}-amd64 \
-                                             neondatabase/vm-compute-node-${{ matrix.version.pg }}:${{ needs.meta.outputs.build-tag }}-arm64
-
+          docker push neondatabase/vm-compute-node-${{ matrix.version.pg }}:${{ needs.meta.outputs.build-tag }}
 
   test-images:
     needs: [ check-permissions, meta, neon-image, compute-node-image ]
@@ -1061,7 +1036,7 @@ jobs:
           exit 1
 
   deploy:
-    needs: [ check-permissions, push-neon-image-dev, push-compute-image-dev, push-neon-image-prod, push-compute-image-prod, meta, build-and-test-locally, trigger-custom-extensions-build-and-wait ]
+    needs: [ check-permissions, push-neon-image-prod, push-compute-image-prod, meta, build-and-test-locally, trigger-custom-extensions-build-and-wait ]
     # `!failure() && !cancelled()` is required because the workflow depends on the job that can be skipped: `push-neon-image-prod` and `push-compute-image-prod`
     if: ${{ contains(fromJSON('["push-main", "storage-release", "proxy-release", "compute-release"]'), needs.meta.outputs.run-kind) && !failure() && !cancelled() }}
     permissions:
@@ -1175,7 +1150,7 @@ jobs:
               -f deployPgSniRouter=false \
               -f deployProxy=false \
               -f deployStorage=true \
-              -f deployStorageBroker=false \
+              -f deployStorageBroker=true \
               -f deployStorageController=true \
               -f branch=main \
               -f dockerTag=${{needs.meta.outputs.build-tag}} \
@@ -1183,7 +1158,7 @@ jobs:
 
             gh workflow --repo neondatabase/infra run deploy-prod.yml --ref main \
               -f deployStorage=true \
-              -f deployStorageBroker=false \
+              -f deployStorageBroker=true \
               -f deployStorageController=true \
               -f branch=main \
               -f dockerTag=${{needs.meta.outputs.build-tag}}
@@ -1231,7 +1206,7 @@ jobs:
           payload: |
             channel: ${{ vars.SLACK_STORAGE_CHANNEL_ID }}
             text: |
-              🔴 <!subteam^S06CJ87UMNY|@oncall-storage>: deploy job on release branch had unexpected status "${{ needs.deploy.result }}" <${{ github.server_url }}/${{ github.repository }}/actions/runs/${{ github.run_id }}|GitHub Run>.
+              🔴 @oncall-storage: deploy job on release branch had unexpected status "${{ needs.deploy.result }}" <${{ github.server_url }}/${{ github.repository }}/actions/runs/${{ github.run_id }}|GitHub Run>.
 
   # The job runs on `release` branch and copies compatibility data and Neon artifact from the last *release PR* to the latest directory
   promote-compatibility-data:

.github/workflows/cargo-deny.yml

@@ -7,7 +7,7 @@ on:
         required: false
         type: string
   schedule:
-    - cron: '0 10 * * *'
+    - cron: '0 0 * * *'
 
 jobs:
   cargo-deny:
@@ -50,9 +50,8 @@ jobs:
           method: chat.postMessage
           token: ${{ secrets.SLACK_BOT_TOKEN }}
           payload: |
-            channel: ${{ vars.SLACK_ON_CALL_DEVPROD_STREAM }}
+            channel: ${{ vars.SLACK_CICD_CHANNEL_ID }}
             text: |
               Periodic cargo-deny on ${{ matrix.ref }}: ${{ job.status }}
               <${{ github.server_url }}/${{ github.repository }}/actions/runs/${{ github.run_id }}|GitHub Run>
-              Fixing the problem should be fairly straight forward from the logs. If not, <#${{ vars.SLACK_RUST_CHANNEL_ID }}> is there to help.
-              Pinging <!subteam^S0838JPSH32|@oncall-devprod>.
+              Pinging @oncall-devprod.

.github/workflows/force-test-extensions-upgrade.yml

@@ -52,9 +52,8 @@ jobs:
       - name: Test extension upgrade
         timeout-minutes: 20
         env:
-          NEW_COMPUTE_TAG: latest
-          OLD_COMPUTE_TAG: ${{ steps.get-last-compute-release-tag.outputs.tag }}
-          TEST_EXTENSIONS_TAG: ${{ steps.get-last-compute-release-tag.outputs.tag }}
+          NEWTAG: latest
+          OLDTAG: ${{ steps.get-last-compute-release-tag.outputs.tag }}
           PG_VERSION: ${{ matrix.pg-version }}
           FORCE_ALL_UPGRADE_TESTS: true
         run: ./docker-compose/test_extensions_upgrade.sh

.github/workflows/large_oltp_benchmark.yml

@@ -1,147 +0,0 @@
-name: large oltp benchmark
-
-on:
-  # uncomment to run on push for debugging your PR
-  push:
-    branches: [ bodobolero/synthetic_oltp_workload ]
-
-  schedule:
-    # * is a special character in YAML so you have to quote this string
-    #          ┌───────────── minute (0 - 59)
-    #          │ ┌───────────── hour (0 - 23)
-    #          │ │  ┌───────────── day of the month (1 - 31)
-    #          │ │  │ ┌───────────── month (1 - 12 or JAN-DEC)
-    #          │ │  │ │ ┌───────────── day of the week (0 - 6 or SUN-SAT)
-    - cron:   '0 15 * * *' # run once a day, timezone is utc, avoid conflict with other benchmarks
-  workflow_dispatch: # adds ability to run this manually
-
-defaults:
-  run:
-    shell: bash -euxo pipefail {0}
-
-concurrency:
-  # Allow only one workflow globally because we need dedicated resources which only exist once
-  group: large-oltp-bench-workflow
-  cancel-in-progress: true
-
-jobs:
-  oltp:
-    strategy:
-      fail-fast: false # allow other variants to continue even if one fails
-      matrix:
-        include:
-          - target: new_branch 
-            custom_scripts: insert_webhooks.sql@2 select_any_webhook_with_skew.sql@4 select_recent_webhook.sql@4 
-          - target: reuse_branch 
-            custom_scripts: insert_webhooks.sql@2 select_any_webhook_with_skew.sql@4 select_recent_webhook.sql@4 
-      max-parallel: 1 # we want to run each stripe size sequentially to be able to compare the results
-    permissions:
-      contents: write
-      statuses: write
-      id-token: write # aws-actions/configure-aws-credentials
-    env:
-      TEST_PG_BENCH_DURATIONS_MATRIX: "1h" # todo update to > 1 h 
-      TEST_PGBENCH_CUSTOM_SCRIPTS: ${{ matrix.custom_scripts }}
-      POSTGRES_DISTRIB_DIR: /tmp/neon/pg_install
-      PG_VERSION: 16 # pre-determined by pre-determined project
-      TEST_OUTPUT: /tmp/test_output
-      BUILD_TYPE: remote
-      SAVE_PERF_REPORT: ${{ github.ref_name == 'main' }}
-      PLATFORM: ${{ matrix.target }}
-
-    runs-on: [ self-hosted, us-east-2, x64 ]
-    container:
-      image: neondatabase/build-tools:pinned-bookworm
-      credentials:
-        username: ${{ secrets.NEON_DOCKERHUB_USERNAME }}
-        password: ${{ secrets.NEON_DOCKERHUB_PASSWORD }}
-      options: --init
-
-    # Increase timeout to 8h, default timeout is 6h
-    timeout-minutes: 480
-
-    steps:
-    - uses: actions/checkout@v4
-
-    - name: Configure AWS credentials # necessary to download artefacts
-      uses: aws-actions/configure-aws-credentials@v4
-      with:
-        aws-region: eu-central-1
-        role-to-assume: ${{ vars.DEV_AWS_OIDC_ROLE_ARN }}
-        role-duration-seconds: 18000 # 5 hours is currently max associated with IAM role
-
-    - name: Download Neon artifact
-      uses: ./.github/actions/download
-      with:
-        name: neon-${{ runner.os }}-${{ runner.arch }}-release-artifact
-        path: /tmp/neon/
-        prefix: latest
-        aws-oicd-role-arn: ${{ vars.DEV_AWS_OIDC_ROLE_ARN }}
-
-    - name: Create Neon Branch for large tenant
-      if: ${{ matrix.target == 'new_branch' }}
-      id: create-neon-branch-oltp-target
-      uses: ./.github/actions/neon-branch-create
-      with:
-          project_id: ${{ vars.BENCHMARK_LARGE_OLTP_PROJECTID }}
-          api_key: ${{ secrets.NEON_STAGING_API_KEY }}
-
-    - name: Set up Connection String
-      id: set-up-connstr
-      run: |
-          case "${{ matrix.target }}" in
-              new_branch)
-              CONNSTR=${{ steps.create-neon-branch-oltp-target.outputs.dsn }}
-              ;;
-              reuse_branch)
-              CONNSTR=${{ secrets.BENCHMARK_LARGE_OLTP_REUSE_CONNSTR }}
-              ;;
-              *)
-              echo >&2 "Unknown target=${{ matrix.target }}"
-              exit 1
-              ;;
-          esac
-
-          echo "connstr=${CONNSTR}" >> $GITHUB_OUTPUT
-
-    - name: Benchmark pgbench with custom-scripts
-      uses: ./.github/actions/run-python-test-set
-      with:
-        build_type: ${{ env.BUILD_TYPE }}
-        test_selection: performance
-        run_in_parallel: false
-        save_perf_report: ${{ env.SAVE_PERF_REPORT }}
-        extra_params: -m remote_cluster --timeout 21600 -k test_perf_oltp_large_tenant
-        pg_version: ${{ env.PG_VERSION }}
-        aws-oicd-role-arn: ${{ vars.DEV_AWS_OIDC_ROLE_ARN }}
-      env:
-        BENCHMARK_CONNSTR: ${{ steps.set-up-connstr.outputs.connstr }}
-        VIP_VAP_ACCESS_TOKEN: "${{ secrets.VIP_VAP_ACCESS_TOKEN }}"
-        PERF_TEST_RESULT_CONNSTR: "${{ secrets.PERF_TEST_RESULT_CONNSTR }}"
-
-    - name: Delete Neon Branch for large tenant
-      if: ${{ always() && matrix.target == 'new_branch' }}
-      uses: ./.github/actions/neon-branch-delete
-      with:
-        project_id: ${{ vars.BENCHMARK_LARGE_OLTP_PROJECTID }}
-        branch_id: ${{ steps.create-neon-branch-oltp-target.outputs.branch_id }}
-        api_key: ${{ secrets.NEON_STAGING_API_KEY }}
-
-    - name: Create Allure report
-      id: create-allure-report
-      if: ${{ !cancelled() }}
-      uses: ./.github/actions/allure-report-generate
-      with:
-        aws-oicd-role-arn: ${{ vars.DEV_AWS_OIDC_ROLE_ARN }}
-  
-    - name: Post to a Slack channel
-      if: ${{ github.event.schedule && failure() }}
-      uses: slackapi/slack-github-action@v1
-      with:
-        channel-id: "C06KHQVQ7U3" # on-call-qa-staging-stream
-        slack-message: |
-          Periodic large oltp perf testing: ${{ job.status }}
-          <${{ github.server_url }}/${{ github.repository }}/actions/runs/${{ github.run_id }}|GitHub Run>
-          <${{ steps.create-allure-report.outputs.report-url }}|Allure report>
-      env:
-        SLACK_BOT_TOKEN: ${{ secrets.SLACK_BOT_TOKEN }}

.github/workflows/periodic_pagebench.yml

@@ -3,12 +3,12 @@ name: Periodic pagebench performance test on dedicated EC2 machine in eu-central
 on:
   schedule:
     # * is a special character in YAML so you have to quote this string
-    #        ┌───────────── minute (0 - 59)
-    #        │   ┌───────────── hour (0 - 23)
-    #        │   │ ┌───────────── day of the month (1 - 31)
-    #        │   │ │ ┌───────────── month (1 - 12 or JAN-DEC)
-    #        │   │ │ │ ┌───────────── day of the week (0 - 6 or SUN-SAT)
-    - cron: '0 */3 * * *' # Runs every 3 hours
+    #          ┌───────────── minute (0 - 59)
+    #          │ ┌───────────── hour (0 - 23)
+    #          │ │ ┌───────────── day of the month (1 - 31)
+    #          │ │ │ ┌───────────── month (1 - 12 or JAN-DEC)
+    #          │ │ │ │ ┌───────────── day of the week (0 - 6 or SUN-SAT)
+    - cron:  '0 18 * * *' # Runs at 6 PM UTC every day
   workflow_dispatch: # Allows manual triggering of the workflow
     inputs:
       commit_hash:
@@ -78,10 +78,8 @@ jobs:
       run: |
         if [ -z "$INPUT_COMMIT_HASH" ]; then
           echo "COMMIT_HASH=$(curl -s https://api.github.com/repos/neondatabase/neon/commits/main | jq -r '.sha')" >> $GITHUB_ENV
-          echo "COMMIT_HASH_TYPE=latest" >> $GITHUB_ENV
         else
           echo "COMMIT_HASH=$INPUT_COMMIT_HASH" >> $GITHUB_ENV
-          echo "COMMIT_HASH_TYPE=manual" >> $GITHUB_ENV
         fi
 
     - name: Start Bench with run_id
@@ -91,7 +89,7 @@ jobs:
         -H 'accept: application/json' \
         -H 'Content-Type: application/json' \
         -H "Authorization: Bearer $API_KEY" \
-        -d "{\"neonRepoCommitHash\": \"${COMMIT_HASH}\", \"neonRepoCommitHashType\": \"${COMMIT_HASH_TYPE}\"}"
+        -d "{\"neonRepoCommitHash\": \"${COMMIT_HASH}\"}"
 
     - name: Poll Test Status
       id: poll_step

CODEOWNERS

@@ -1,8 +1,8 @@
 # Autoscaling
 /libs/vm_monitor/ @neondatabase/autoscaling
 
-# DevProd & PerfCorr
-/.github/ @neondatabase/developer-productivity @neondatabase/performance-correctness
+# DevProd
+/.github/ @neondatabase/developer-productivity
 
 # Compute
 /pgxn/ @neondatabase/compute

Cargo.lock

@@ -783,28 +783,6 @@ dependencies = [
  "tracing",
 ]
 
-[[package]]
-name = "axum-extra"
-version = "0.10.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "460fc6f625a1f7705c6cf62d0d070794e94668988b1c38111baeec177c715f7b"
-dependencies = [
- "axum",
- "axum-core",
- "bytes",
- "futures-util",
- "headers",
- "http 1.1.0",
- "http-body 1.0.0",
- "http-body-util",
- "mime",
- "pin-project-lite",
- "serde",
- "tower 0.5.2",
- "tower-layer",
- "tower-service",
-]
-
 [[package]]
 name = "azure_core"
 version = "0.21.0"
@@ -947,9 +925,9 @@ checksum = "0ea22880d78093b0cbe17c89f64a7d457941e65759157ec6cb31a31d652b05e5"
 
 [[package]]
 name = "base64"
-version = "0.21.7"
+version = "0.21.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "9d297deb1925b89f2ccc13d7635fa0714f12c87adce1c75356b39ca9b7178567"
+checksum = "3f1e31e207a6b8fb791a38ea3105e6cb541f55e4d029902d3039a4ad07cc4105"
 
 [[package]]
 name = "base64"
@@ -1127,9 +1105,9 @@ checksum = "37b2a672a2cb129a2e41c10b1224bb368f9f37a2b16b612598138befd7b37eb5"
 
 [[package]]
 name = "cc"
-version = "1.2.16"
+version = "1.1.30"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "be714c154be609ec7f5dad223a33bf1482fff90472de28f7362806e6d4832b8c"
+checksum = "b16803a61b81d9eabb7eae2588776c4c1e584b738ede45fdbb4c972cec1e9945"
 dependencies = [
  "jobserver",
  "libc",
@@ -1327,7 +1305,6 @@ dependencies = [
  "aws-sdk-s3",
  "aws-smithy-types",
  "axum",
- "axum-extra",
  "base64 0.13.1",
  "bytes",
  "camino",
@@ -1339,7 +1316,6 @@ dependencies = [
  "flate2",
  "futures",
  "http 1.1.0",
- "jsonwebtoken",
  "metrics",
  "nix 0.27.1",
  "notify",
@@ -2321,7 +2297,7 @@ name = "framed-websockets"
 version = "0.1.0"
 source = "git+https://github.com/neondatabase/framed-websockets#34eff3d6f8cfccbc5f35e4f65314ff7328621127"
 dependencies = [
- "base64 0.21.7",
+ "base64 0.21.1",
  "bytemuck",
  "bytes",
  "futures-core",
@@ -2434,9 +2410,9 @@ checksum = "f90f7dce0722e95104fcb095585910c0977252f286e354b5e3bd38902cd99988"
 
 [[package]]
 name = "futures-timer"
-version = "3.0.3"
+version = "3.0.2"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "f288b0a4f20f9a56b5d1da57e2227c661b7b16168e2f72365f57b63326e29b24"
+checksum = "e64b03909df88034c26dc1547e8970b91f98bdb65165d6a4e9110d94263dbb2c"
 
 [[package]]
 name = "futures-util"
@@ -2539,27 +2515,6 @@ version = "0.3.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "d2fabcfbdc87f4758337ca535fb41a6d701b65693ce38287d856d1674551ec9b"
 
-[[package]]
-name = "governor"
-version = "0.8.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "842dc78579ce01e6a1576ad896edc92fca002dd60c9c3746b7fc2bec6fb429d0"
-dependencies = [
- "cfg-if",
- "dashmap 6.1.0",
- "futures-sink",
- "futures-timer",
- "futures-util",
- "no-std-compat",
- "nonzero_ext",
- "parking_lot 0.12.1",
- "portable-atomic",
- "quanta",
- "rand 0.8.5",
- "smallvec",
- "spinning_top",
-]
-
 [[package]]
 name = "group"
 version = "0.12.1"
@@ -2677,7 +2632,7 @@ version = "7.5.4"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "765c9198f173dd59ce26ff9f95ef0aafd0a0fe01fb9d72841bc5066a4c06511d"
 dependencies = [
- "base64 0.21.7",
+ "base64 0.21.1",
  "byteorder",
  "crossbeam-channel",
  "flate2",
@@ -2685,30 +2640,6 @@ dependencies = [
  "num-traits",
 ]
 
-[[package]]
-name = "headers"
-version = "0.4.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "322106e6bd0cba2d5ead589ddb8150a13d7c4217cf80d7c4f682ca994ccc6aa9"
-dependencies = [
- "base64 0.21.7",
- "bytes",
- "headers-core",
- "http 1.1.0",
- "httpdate",
- "mime",
- "sha1",
-]
-
-[[package]]
-name = "headers-core"
-version = "0.3.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "54b4a22553d4242c49fddb9ba998a99962b5cc6f22cb5a3482bec22522403ce4"
-dependencies = [
- "http 1.1.0",
-]
-
 [[package]]
 name = "heck"
 version = "0.5.0"
@@ -2846,10 +2777,12 @@ name = "http-utils"
 version = "0.1.0"
 dependencies = [
  "anyhow",
+ "backtrace",
  "bytes",
  "fail",
- "futures",
+ "flate2",
  "hyper 0.14.30",
+ "inferno 0.12.0",
  "itertools 0.10.5",
  "jemalloc_pprof",
  "metrics",
@@ -2862,7 +2795,6 @@ dependencies = [
  "serde_path_to_error",
  "thiserror 1.0.69",
  "tokio",
- "tokio-rustls 0.26.0",
  "tokio-stream",
  "tokio-util",
  "tracing",
@@ -3349,9 +3281,9 @@ checksum = "49f1f14873335454500d59611f1cf4a4b0f786f9ac11f4312a78e4cf2566695b"
 
 [[package]]
 name = "jemalloc_pprof"
-version = "0.7.0"
+version = "0.6.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "5622af6d21ff86ed7797ef98e11b8f302da25ec69a7db9f6cde8e2e1c8df9992"
+checksum = "1a883828bd6a4b957cd9f618886ff19e5f3ebd34e06ba0e855849e049fef32fb"
 dependencies = [
  "anyhow",
  "libc",
@@ -3435,7 +3367,7 @@ version = "9.2.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "5c7ea04a7c5c055c175f189b6dc6ba036fd62306b58c66c9f6389036c503a3f4"
 dependencies = [
- "base64 0.21.7",
+ "base64 0.21.1",
  "js-sys",
  "pem",
  "ring",
@@ -3550,9 +3482,9 @@ dependencies = [
 
 [[package]]
 name = "mappings"
-version = "0.7.0"
+version = "0.6.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "e434981a332777c2b3062652d16a55f8e74fa78e6b1882633f0d77399c84fc2a"
+checksum = "ce9229c438fbf1c333926e2053c4c091feabbd40a1b590ec62710fea2384af9e"
 dependencies = [
  "anyhow",
  "libc",
@@ -3793,12 +3725,6 @@ dependencies = [
  "memoffset 0.9.0",
 ]
 
-[[package]]
-name = "no-std-compat"
-version = "0.4.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "b93853da6d84c2e3c7d730d6473e8817692dd89be387eb01b94d7f108ecb5b8c"
-
 [[package]]
 name = "nom"
 version = "7.1.3"
@@ -3809,12 +3735,6 @@ dependencies = [
  "minimal-lexical",
 ]
 
-[[package]]
-name = "nonzero_ext"
-version = "0.3.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "38bf9645c8b145698bb0b18a4637dcacbc421ea49bef2317e4fd8065a387cf21"
-
 [[package]]
 name = "notify"
 version = "8.0.0"
@@ -4064,7 +3984,7 @@ dependencies = [
  "opentelemetry-http",
  "opentelemetry-proto",
  "opentelemetry_sdk",
- "prost 0.13.3",
+ "prost",
  "reqwest",
  "thiserror 1.0.69",
 ]
@@ -4077,7 +3997,7 @@ checksum = "a6e05acbfada5ec79023c85368af14abd0b307c015e9064d249b2a950ef459a6"
 dependencies = [
  "opentelemetry",
  "opentelemetry_sdk",
- "prost 0.13.3",
+ "prost",
  "tonic",
 ]
 
@@ -4191,7 +4111,6 @@ dependencies = [
  "pageserver_api",
  "pageserver_client",
  "rand 0.8.5",
- "reqwest",
  "serde",
  "serde_json",
  "tokio",
@@ -4281,9 +4200,6 @@ dependencies = [
  "remote_storage",
  "reqwest",
  "rpds",
- "rustls 0.23.18",
- "rustls-pemfile 2.1.1",
- "rustls-pki-types",
  "scopeguard",
  "send-future",
  "serde",
@@ -4302,7 +4218,6 @@ dependencies = [
  "tokio-epoll-uring",
  "tokio-io-timeout",
  "tokio-postgres",
- "tokio-rustls 0.26.0",
  "tokio-stream",
  "tokio-tar",
  "tokio-util",
@@ -4310,7 +4225,6 @@ dependencies = [
  "tracing",
  "url",
  "utils",
- "uuid",
  "wal_decoder",
  "walkdir",
  "workspace_hack",
@@ -4393,9 +4307,9 @@ dependencies = [
 
 [[package]]
 name = "papaya"
-version = "0.2.0"
+version = "0.1.8"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "aab21828b6b5952fdadd6c377728ffae53ec3a21b2febc47319ab65741f7e2fd"
+checksum = "dc7c76487f7eaa00a0fc1d7f88dc6b295aec478d11b0fc79f857b62c2874124c"
 dependencies = [
  "equivalent",
  "seize",
@@ -4523,7 +4437,7 @@ version = "3.0.3"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "1b8fcc794035347fb64beda2d3b462595dd2753e3f268d89c5aae77e8cf2c310"
 dependencies = [
- "base64 0.21.7",
+ "base64 0.21.1",
  "serde",
 ]
 
@@ -4677,12 +4591,6 @@ dependencies = [
  "never-say-never",
 ]
 
-[[package]]
-name = "portable-atomic"
-version = "1.10.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "280dc24453071f1b63954171985a0b0d30058d287960968b9b2aca264c8d4ee6"
-
 [[package]]
 name = "postgres"
 version = "0.19.7"
@@ -4837,10 +4745,8 @@ dependencies = [
  "nix 0.26.4",
  "once_cell",
  "parking_lot 0.12.1",
- "prost 0.12.6",
- "prost-build 0.12.6",
- "prost-derive 0.12.6",
- "sha2",
+ "protobuf",
+ "protobuf-codegen-pure",
  "smallvec",
  "symbolic-demangle",
  "tempfile",
@@ -4849,17 +4755,15 @@ dependencies = [
 
 [[package]]
 name = "pprof_util"
-version = "0.7.0"
+version = "0.6.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "9fa015c78eed2130951e22c58d2095849391e73817ab2e74f71b0b9f63dd8416"
+checksum = "65c568b3f8c1c37886ae07459b1946249e725c315306b03be5632f84c239f781"
 dependencies = [
  "anyhow",
- "backtrace",
  "flate2",
- "inferno 0.12.0",
  "num",
  "paste",
- "prost 0.13.3",
+ "prost",
 ]
 
 [[package]]
@@ -4952,16 +4856,6 @@ dependencies = [
  "thiserror 1.0.69",
 ]
 
-[[package]]
-name = "prost"
-version = "0.12.6"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "deb1435c188b76130da55f17a466d252ff7b1418b2ad3e037d127b94e3411f29"
-dependencies = [
- "bytes",
- "prost-derive 0.12.6",
-]
-
 [[package]]
 name = "prost"
 version = "0.13.3"
@@ -4969,28 +4863,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "7b0487d90e047de87f984913713b85c601c05609aad5b0df4b4573fbf69aa13f"
 dependencies = [
  "bytes",
- "prost-derive 0.13.3",
-]
-
-[[package]]
-name = "prost-build"
-version = "0.12.6"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "22505a5c94da8e3b7c2996394d1c933236c4d743e81a410bcca4e6989fc066a4"
-dependencies = [
- "bytes",
- "heck",
- "itertools 0.10.5",
- "log",
- "multimap",
- "once_cell",
- "petgraph",
- "prettyplease",
- "prost 0.12.6",
- "prost-types 0.12.6",
- "regex",
- "syn 2.0.90",
- "tempfile",
+ "prost-derive",
 ]
 
 [[package]]
@@ -5007,26 +4880,13 @@ dependencies = [
  "once_cell",
  "petgraph",
  "prettyplease",
- "prost 0.13.3",
- "prost-types 0.13.3",
+ "prost",
+ "prost-types",
  "regex",
  "syn 2.0.90",
  "tempfile",
 ]
 
-[[package]]
-name = "prost-derive"
-version = "0.12.6"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "81bddcdb20abf9501610992b6759a4c888aef7d1a7247ef75e2404275ac24af1"
-dependencies = [
- "anyhow",
- "itertools 0.10.5",
- "proc-macro2",
- "quote",
- "syn 2.0.90",
-]
-
 [[package]]
 name = "prost-derive"
 version = "0.13.3"
@@ -5040,22 +4900,38 @@ dependencies = [
  "syn 2.0.90",
 ]
 
-[[package]]
-name = "prost-types"
-version = "0.12.6"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "9091c90b0a32608e984ff2fa4091273cbdd755d54935c51d520887f4a1dbd5b0"
-dependencies = [
- "prost 0.12.6",
-]
-
 [[package]]
 name = "prost-types"
 version = "0.13.3"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "4759aa0d3a6232fb8dbdb97b61de2c20047c68aca932c7ed76da9d788508d670"
 dependencies = [
- "prost 0.13.3",
+ "prost",
+]
+
+[[package]]
+name = "protobuf"
+version = "2.28.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "106dd99e98437432fed6519dedecfade6a06a73bb7b2a1e019fdd2bee5778d94"
+
+[[package]]
+name = "protobuf-codegen"
+version = "2.28.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "033460afb75cf755fcfc16dfaed20b86468082a2ea24e05ac35ab4a099a017d6"
+dependencies = [
+ "protobuf",
+]
+
+[[package]]
+name = "protobuf-codegen-pure"
+version = "2.28.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "95a29399fc94bcd3eeaa951c715f7bea69409b2445356b00519740bcd6ddd865"
+dependencies = [
+ "protobuf",
+ "protobuf-codegen",
 ]
 
 [[package]]
@@ -5176,21 +5052,6 @@ dependencies = [
  "zerocopy",
 ]
 
-[[package]]
-name = "quanta"
-version = "0.12.5"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "3bd1fe6824cea6538803de3ff1bc0cf3949024db3d43c9643024bfb33a807c0e"
-dependencies = [
- "crossbeam-utils",
- "libc",
- "once_cell",
- "raw-cpuid",
- "wasi 0.11.0+wasi-snapshot-preview1",
- "web-sys",
- "winapi",
-]
-
 [[package]]
 name = "quick-xml"
 version = "0.26.0"
@@ -5321,15 +5182,6 @@ dependencies = [
  "num-traits",
 ]
 
-[[package]]
-name = "raw-cpuid"
-version = "11.3.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "c6928fa44c097620b706542d428957635951bade7143269085389d42c8a4927e"
-dependencies = [
- "bitflags 2.8.0",
-]
-
 [[package]]
 name = "rayon"
 version = "1.7.0"
@@ -5664,16 +5516,16 @@ dependencies = [
 
 [[package]]
 name = "ring"
-version = "0.17.13"
+version = "0.17.6"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "70ac5d832aa16abd7d1def883a8545280c20a60f523a370aa3a9617c2b8550ee"
+checksum = "684d5e6e18f669ccebf64a92236bb7db9a34f07be010e3627368182027180866"
 dependencies = [
  "cc",
- "cfg-if",
  "getrandom 0.2.11",
  "libc",
+ "spin",
  "untrusted",
- "windows-sys 0.52.0",
+ "windows-sys 0.48.0",
 ]
 
 [[package]]
@@ -5900,7 +5752,7 @@ version = "1.0.2"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "d194b56d58803a43635bdc398cd17e383d6f71f9182b9a192c127ca42494a59b"
 dependencies = [
- "base64 0.21.7",
+ "base64 0.21.1",
 ]
 
 [[package]]
@@ -5909,15 +5761,15 @@ version = "2.1.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "f48172685e6ff52a556baa527774f61fcaa884f59daf3375c62a3f1cd2549dab"
 dependencies = [
- "base64 0.21.7",
+ "base64 0.21.1",
  "rustls-pki-types",
 ]
 
 [[package]]
 name = "rustls-pki-types"
-version = "1.11.0"
+version = "1.10.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "917ce264624a4b4db1c364dcc35bfca9ded014d0a958cd47ad3e960e988ea51c"
+checksum = "16f1201b3c9a7ee8039bcadc17b7e605e2945b27eee7631788c1bd2b0643674b"
 
 [[package]]
 name = "rustls-webpki"
@@ -6148,9 +6000,9 @@ dependencies = [
 
 [[package]]
 name = "seize"
-version = "0.5.0"
+version = "0.4.9"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "e4b8d813387d566f627f3ea1b914c068aac94c40ae27ec43f5f33bde65abefe7"
+checksum = "d84b0c858bdd30cb56f5597f8b3bf702ec23829e652cc636a1e5a7b9de46ae93"
 dependencies = [
  "libc",
  "windows-sys 0.52.0",
@@ -6543,15 +6395,6 @@ version = "0.9.8"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "6980e8d7511241f8acf4aebddbb1ff938df5eebe98691418c4468d0b72a96a67"
 
-[[package]]
-name = "spinning_top"
-version = "0.3.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "d96d2d1d716fb500937168cc09353ffdc7a012be8475ac7308e1bdf0e3923300"
-dependencies = [
- "lock_api",
-]
-
 [[package]]
 name = "spki"
 version = "0.6.0"
@@ -6603,7 +6446,7 @@ dependencies = [
  "metrics",
  "once_cell",
  "parking_lot 0.12.1",
- "prost 0.13.3",
+ "prost",
  "rustls 0.23.18",
  "tokio",
  "tonic",
@@ -6621,7 +6464,6 @@ dependencies = [
  "bytes",
  "chrono",
  "clap",
- "clashmap",
  "control_plane",
  "cron",
  "diesel",
@@ -6629,7 +6471,6 @@ dependencies = [
  "diesel_migrations",
  "fail",
  "futures",
- "governor",
  "hex",
  "http-utils",
  "humantime",
@@ -7376,7 +7217,7 @@ dependencies = [
  "hyper-util",
  "percent-encoding",
  "pin-project",
- "prost 0.13.3",
+ "prost",
  "rustls-native-certs 0.8.0",
  "rustls-pemfile 2.1.1",
  "tokio",
@@ -7396,8 +7237,8 @@ checksum = "9557ce109ea773b399c9b9e5dca39294110b74f1f342cb347a80d1fce8c26a11"
 dependencies = [
  "prettyplease",
  "proc-macro2",
- "prost-build 0.13.3",
- "prost-types 0.13.3",
+ "prost-build",
+ "prost-types",
  "quote",
  "syn 2.0.90",
 ]
@@ -7444,12 +7285,10 @@ version = "0.6.2"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "403fa3b783d4b626a8ad51d766ab03cb6d2dbfc46b1c5d4448395e6628dc9697"
 dependencies = [
- "base64 0.22.1",
  "bitflags 2.8.0",
  "bytes",
  "http 1.1.0",
  "http-body 1.0.0",
- "mime",
  "pin-project-lite",
  "tower-layer",
  "tower-service",
@@ -7803,6 +7642,7 @@ dependencies = [
  "anyhow",
  "arc-swap",
  "async-compression",
+ "backtrace",
  "bincode",
  "byteorder",
  "bytes",
@@ -7930,7 +7770,7 @@ dependencies = [
  "pageserver_api",
  "postgres_ffi",
  "pprof",
- "prost 0.13.3",
+ "prost",
  "remote_storage",
  "serde",
  "serde_json",
@@ -8356,7 +8196,7 @@ dependencies = [
  "ahash",
  "anyhow",
  "base64 0.13.1",
- "base64 0.21.7",
+ "base64 0.21.1",
  "base64ct",
  "bytes",
  "camino",
@@ -8387,7 +8227,6 @@ dependencies = [
  "hyper-util",
  "indexmap 1.9.3",
  "indexmap 2.0.1",
- "itertools 0.10.5",
  "itertools 0.12.1",
  "lazy_static",
  "libc",
@@ -8406,7 +8245,7 @@ dependencies = [
  "parquet",
  "prettyplease",
  "proc-macro2",
- "prost 0.13.3",
+ "prost",
  "quote",
  "rand 0.8.5",
  "regex",

Cargo.toml

@@ -53,6 +53,7 @@ anyhow = { version = "1.0", features = ["backtrace"] }
 arc-swap = "1.6"
 async-compression = { version = "0.4.0", features = ["tokio", "gzip", "zstd"] }
 atomic-take = "1.1.0"
+backtrace = "0.3.74"
 flate2 = "1.0.26"
 assert-json-diff = "2"
 async-stream = "0.3"
@@ -67,7 +68,6 @@ aws-credential-types = "1.2.0"
 aws-sigv4 = { version = "1.2", features = ["sign-http"] }
 aws-types = "1.3"
 axum = { version = "0.8.1", features = ["ws"] }
-axum-extra = { version = "0.10.0", features = ["typed-header"] }
 base64 = "0.13.0"
 bincode = "1.3"
 bindgen = "0.71"
@@ -95,7 +95,6 @@ futures = "0.3"
 futures-core = "0.3"
 futures-util = "0.3"
 git-version = "0.3"
-governor = "0.8"
 hashbrown = "0.14"
 hashlink = "0.9.1"
 hdrhistogram = "7.5.2"
@@ -114,10 +113,11 @@ hyper-util = "0.1"
 tokio-tungstenite = "0.21.0"
 indexmap = "2"
 indoc = "2"
+inferno = "0.12.0"
 ipnet = "2.10.0"
 itertools = "0.10"
 itoa = "1.0.11"
-jemalloc_pprof = { version = "0.7", features = ["symbolize", "flamegraph"] }
+jemalloc_pprof = "0.6"
 jsonwebtoken = "9"
 lasso = "0.7"
 libc = "0.2"
@@ -139,7 +139,7 @@ parquet = { version = "53", default-features = false, features = ["zstd"] }
 parquet_derive = "53"
 pbkdf2 = { version = "0.12.1", features = ["simple", "std"] }
 pin-project-lite = "0.2"
-pprof = { version = "0.14", features = ["criterion", "flamegraph", "frame-pointer", "prost-codec"] }
+pprof = { version = "0.14", features = ["criterion", "flamegraph", "frame-pointer", "protobuf", "protobuf-codec"] }
 procfs = "0.16"
 prometheus = {version = "0.13", default-features=false, features = ["process"]} # removes protobuf dependency
 prost = "0.13"
@@ -155,7 +155,6 @@ rpds = "0.13"
 rustc-hash = "1.1.0"
 rustls = { version = "0.23.16", default-features = false }
 rustls-pemfile = "2"
-rustls-pki-types = "1.11"
 scopeguard = "1.1"
 sysinfo = "0.29.2"
 sd-notify = "0.4.1"
@@ -193,7 +192,7 @@ toml = "0.8"
 toml_edit = "0.22"
 tonic = {version = "0.12.3", default-features = false, features = ["channel", "tls", "tls-roots"]}
 tower = { version = "0.5.2", default-features = false }
-tower-http = { version = "0.6.2", features = ["auth", "request-id", "trace"] }
+tower-http = { version = "0.6.2", features = ["request-id", "trace"] }
 
 # This revision uses opentelemetry 0.27. There's no tag for it.
 tower-otel = { git = "https://github.com/mattiapenati/tower-otel", rev = "56a7321053bcb72443888257b622ba0d43a11fcd" }

Makefile

@@ -11,16 +11,15 @@ ICU_PREFIX_DIR := /usr/local/icu
 #
 BUILD_TYPE ?= debug
 WITH_SANITIZERS ?= no
-PG_CFLAGS = -fsigned-char
 ifeq ($(BUILD_TYPE),release)
 	PG_CONFIGURE_OPTS = --enable-debug --with-openssl
-	PG_CFLAGS += -O2 -g3 $(CFLAGS)
+	PG_CFLAGS = -O2 -g3 $(CFLAGS)
 	PG_LDFLAGS = $(LDFLAGS)
 	# Unfortunately, `--profile=...` is a nightly feature
 	CARGO_BUILD_FLAGS += --release
 else ifeq ($(BUILD_TYPE),debug)
 	PG_CONFIGURE_OPTS = --enable-debug --with-openssl --enable-cassert --enable-depend
-	PG_CFLAGS += -O0 -g3 $(CFLAGS)
+	PG_CFLAGS = -O0 -g3 $(CFLAGS)
 	PG_LDFLAGS = $(LDFLAGS)
 else
 	$(error Bad build type '$(BUILD_TYPE)', see Makefile for options)
@@ -160,8 +159,6 @@ postgres-%: postgres-configure-% \
 	$(MAKE) -C $(POSTGRES_INSTALL_DIR)/build/$*/contrib/pg_visibility install
 	+@echo "Compiling pageinspect $*"
 	$(MAKE) -C $(POSTGRES_INSTALL_DIR)/build/$*/contrib/pageinspect install
-	+@echo "Compiling pg_trgm $*"
-	$(MAKE) -C $(POSTGRES_INSTALL_DIR)/build/$*/contrib/pg_trgm install
 	+@echo "Compiling amcheck $*"
 	$(MAKE) -C $(POSTGRES_INSTALL_DIR)/build/$*/contrib/amcheck install
 	+@echo "Compiling test_decoding $*"

compute/compute-node.Dockerfile

@@ -162,7 +162,7 @@ FROM build-deps AS pg-build
 ARG PG_VERSION
 COPY vendor/postgres-${PG_VERSION:?} postgres
 RUN cd postgres && \
-    export CONFIGURE_CMD="./configure CFLAGS='-O2 -g3 -fsigned-char' --enable-debug --with-openssl --with-uuid=ossp \
+    export CONFIGURE_CMD="./configure CFLAGS='-O2 -g3' --enable-debug --with-openssl --with-uuid=ossp \
     --with-icu --with-libxml --with-libxslt --with-lz4" && \
     if [ "${PG_VERSION:?}" != "v14" ]; then \
         # zstd is available only from PG15
@@ -1484,7 +1484,7 @@ WORKDIR /ext-src
 COPY compute/patches/pg_duckdb_v031.patch .
 COPY compute/patches/duckdb_v120.patch .
 # pg_duckdb build requires source dir to be a git repo to get submodules
-# allow neon_superuser to execute some functions that in pg_duckdb are available to superuser only:
+# allow neon_superuser to execute some functions that in pg_duckdb are available to superuser only: 
 # - extension management function duckdb.install_extension()
 # - access to duckdb.extensions table and its sequence
 RUN git clone --depth 1 --branch v0.3.1 https://github.com/duckdb/pg_duckdb.git pg_duckdb-src && \
@@ -1499,8 +1499,8 @@ ARG PG_VERSION
 COPY --from=pg_duckdb-src /ext-src/ /ext-src/
 WORKDIR /ext-src/pg_duckdb-src
 RUN make install -j $(getconf _NPROCESSORS_ONLN) && \
-    echo 'trusted = true' >> /usr/local/pgsql/share/extension/pg_duckdb.control
-
+    echo 'trusted = true' >> /usr/local/pgsql/share/extension/pg_duckdb.control 
+        
 #########################################################################################
 #
 # Layer "pg_repack"
@@ -1758,15 +1758,15 @@ ARG TARGETARCH
 # test_runner/regress/test_compute_metrics.py
 # See comment on the top of the file regading `echo`, `-e` and `\n`
 RUN if [ "$TARGETARCH" = "amd64" ]; then\
-        postgres_exporter_sha256='59aa4a7bb0f7d361f5e05732f5ed8c03cc08f78449cef5856eadec33a627694b';\
+        postgres_exporter_sha256='027e75dda7af621237ff8f5ac66b78a40b0093595f06768612b92b1374bd3105';\
         pgbouncer_exporter_sha256='c9f7cf8dcff44f0472057e9bf52613d93f3ffbc381ad7547a959daa63c5e84ac';\
         sql_exporter_sha256='38e439732bbf6e28ca4a94d7bc3686d3fa1abdb0050773d5617a9efdb9e64d08';\
     else\
-        postgres_exporter_sha256='d1dedea97f56c6d965837bfd1fbb3e35a3b4a4556f8cccee8bd513d8ee086124';\
+        postgres_exporter_sha256='131a376d25778ff9701a4c81f703f179e0b58db5c2c496e66fa43f8179484786';\
         pgbouncer_exporter_sha256='217c4afd7e6492ae904055bc14fe603552cf9bac458c063407e991d68c519da3';\
         sql_exporter_sha256='11918b00be6e2c3a67564adfdb2414fdcbb15a5db76ea17d1d1a944237a893c6';\
     fi\
-    && curl -sL https://github.com/prometheus-community/postgres_exporter/releases/download/v0.17.1/postgres_exporter-0.17.1.linux-${TARGETARCH}.tar.gz\
+    && curl -sL https://github.com/prometheus-community/postgres_exporter/releases/download/v0.16.0/postgres_exporter-0.16.0.linux-${TARGETARCH}.tar.gz\
      | tar xzf - --strip-components=1 -C.\
     && curl -sL https://github.com/prometheus-community/pgbouncer_exporter/releases/download/v0.10.2/pgbouncer_exporter-0.10.2.linux-${TARGETARCH}.tar.gz\
      | tar xzf - --strip-components=1 -C.\
@@ -1933,7 +1933,6 @@ RUN apt update && \
         locales \
         procps \
         ca-certificates \
-        rsyslog \
         $VERSION_INSTALLS && \
     apt clean && rm -rf /var/lib/apt/lists/* /tmp/* /var/tmp/* && \
     localedef -i en_US -c -f UTF-8 -A /usr/share/locale/locale.alias en_US.UTF-8
@@ -1979,13 +1978,6 @@ COPY --from=sql_exporter_preprocessor --chmod=0644 /home/nonroot/compute/etc/neo
 # Make the libraries we built available
 RUN echo '/usr/local/lib' >> /etc/ld.so.conf && /sbin/ldconfig
 
-# rsyslog config permissions
-# directory for rsyslogd pid file
-RUN mkdir /var/run/rsyslogd && \
-    chown -R postgres:postgres /var/run/rsyslogd && \
-    chown -R postgres:postgres /etc/rsyslog.d/
-
-
 ENV LANG=en_US.utf8
 USER postgres
 ENTRYPOINT ["/usr/local/bin/compute_ctl"]

compute/etc/neon_collector.jsonnet

@@ -29,7 +29,6 @@
     import 'sql_exporter/lfc_approximate_working_set_size.libsonnet',
     import 'sql_exporter/lfc_approximate_working_set_size_windows.libsonnet',
     import 'sql_exporter/lfc_cache_size_limit.libsonnet',
-    import 'sql_exporter/lfc_chunk_size.libsonnet',
     import 'sql_exporter/lfc_hits.libsonnet',
     import 'sql_exporter/lfc_misses.libsonnet',
     import 'sql_exporter/lfc_used.libsonnet',

compute/etc/sql_exporter/db_total_size.sql

@@ -1,5 +1 @@
-SELECT sum(pg_database_size(datname)) AS total
-FROM pg_database
--- Ignore invalid databases, as we will likely have problems with
--- getting their size from the Pageserver.
-WHERE datconnlimit != -2;
+SELECT sum(pg_database_size(datname)) AS total FROM pg_database;

compute/etc/sql_exporter/lfc_chunk_size.libsonnet

@@ -1,10 +0,0 @@
-{
-  metric_name: 'lfc_chunk_size',
-  type: 'gauge',
-  help: 'LFC chunk size, measured in 8KiB pages',
-  key_labels: null,
-  values: [
-    'lfc_chunk_size_pages',
-  ],
-  query: importstr 'sql_exporter/lfc_chunk_size.sql',
-}

compute/etc/sql_exporter/lfc_chunk_size.sql

@@ -1 +0,0 @@
-SELECT lfc_value AS lfc_chunk_size_pages FROM neon.neon_lfc_stats WHERE lfc_key = 'file_cache_chunk_size_pages';

compute/etc/sql_exporter/pg_stats_userdb.sql

@@ -1,20 +1,10 @@
 -- We export stats for 10 non-system databases. Without this limit it is too
 -- easy to abuse the system by creating lots of databases.
 
-SELECT pg_database_size(datname) AS db_size,
-  deadlocks,
-  tup_inserted AS inserted,
-  tup_updated AS updated,
-  tup_deleted AS deleted,
-  datname
+SELECT pg_database_size(datname) AS db_size, deadlocks, tup_inserted AS inserted,
+  tup_updated AS updated, tup_deleted AS deleted, datname
 FROM pg_stat_database
 WHERE datname IN (
   SELECT datname FROM pg_database
-  -- Ignore invalid databases, as we will likely have problems with
-  -- getting their size from the Pageserver.
-  WHERE datconnlimit != -2
-    AND datname <> 'postgres'
-    AND NOT datistemplate
-  ORDER BY oid
-  LIMIT 10
+  WHERE datname <> 'postgres' AND NOT datistemplate ORDER BY oid LIMIT 10
 );

compute/vm-image-spec-bookworm.yaml

@@ -39,10 +39,6 @@ commands:
     user: nobody
     sysvInitAction: respawn
     shell: '/bin/sql_exporter -config.file=/etc/sql_exporter_autoscaling.yml -web.listen-address=:9499'
-  - name: rsyslogd
-    user: postgres
-    sysvInitAction: respawn
-    shell: '/usr/sbin/rsyslogd -n -i /var/run/rsyslogd/rsyslogd.pid -f /etc/compute_rsyslog.conf'
 shutdownHook: |
   su -p postgres --session-command '/usr/local/bin/pg_ctl stop -D /var/db/postgres/compute/pgdata -m fast --wait -t 10'
 files:
@@ -58,7 +54,7 @@ files:
       # regardless of hostname (ALL)
       #
       # Also allow it to shut down the VM. The fast_import job does that when it's finished.
-      postgres ALL=(root) NOPASSWD: /neonvm/bin/resize-swap, /neonvm/bin/set-disk-quota, /neonvm/bin/poweroff, /usr/sbin/rsyslogd
+      postgres ALL=(root) NOPASSWD: /neonvm/bin/resize-swap, /neonvm/bin/set-disk-quota, /neonvm/bin/poweroff
   - filename: cgconfig.conf
     content: |
       # Configuration for cgroups in VM compute nodes
@@ -73,12 +69,6 @@ files:
           }
           memory {}
       }
-# Create dummy rsyslog config, because it refuses to start without at least one action configured.
-# compute_ctl will rewrite this file with the actual configuration, if needed.
-  - filename: compute_rsyslog.conf
-    content: |
-      *.*    /dev/null
-      $IncludeConfig /etc/rsyslog.d/*.conf
 build: |
   # Build cgroup-tools
   #
@@ -142,12 +132,6 @@ merge: |
   RUN set -e \
       && chmod 0644 /etc/cgconfig.conf
 
-
-  COPY compute_rsyslog.conf /etc/compute_rsyslog.conf
-  RUN chmod 0666 /etc/compute_rsyslog.conf
-  RUN chmod 0666 /var/log/
-
-
   COPY --from=libcgroup-builder /libcgroup-install/bin/*  /usr/bin/
   COPY --from=libcgroup-builder /libcgroup-install/lib/*  /usr/lib/
   COPY --from=libcgroup-builder /libcgroup-install/sbin/* /usr/sbin/

compute/vm-image-spec-bullseye.yaml

@@ -39,10 +39,6 @@ commands:
     user: nobody
     sysvInitAction: respawn
     shell: '/bin/sql_exporter -config.file=/etc/sql_exporter_autoscaling.yml -web.listen-address=:9499'
-  - name: rsyslogd
-    user: postgres
-    sysvInitAction: respawn
-    shell: '/usr/sbin/rsyslogd -n -i /var/run/rsyslogd/rsyslogd.pid -f /etc/compute_rsyslog.conf'
 shutdownHook: |
   su -p postgres --session-command '/usr/local/bin/pg_ctl stop -D /var/db/postgres/compute/pgdata -m fast --wait -t 10'
 files:
@@ -58,7 +54,7 @@ files:
       # regardless of hostname (ALL)
       #
       # Also allow it to shut down the VM. The fast_import job does that when it's finished.
-      postgres ALL=(root) NOPASSWD: /neonvm/bin/resize-swap, /neonvm/bin/set-disk-quota, /neonvm/bin/poweroff, /usr/sbin/rsyslogd
+      postgres ALL=(root) NOPASSWD: /neonvm/bin/resize-swap, /neonvm/bin/set-disk-quota, /neonvm/bin/poweroff
   - filename: cgconfig.conf
     content: |
       # Configuration for cgroups in VM compute nodes
@@ -73,12 +69,6 @@ files:
           }
           memory {}
       }
-# Create dummy rsyslog config, because it refuses to start without at least one action configured.
-# compute_ctl will rewrite this file with the actual configuration, if needed.
-  - filename: compute_rsyslog.conf
-    content: |
-      *.*    /dev/null
-      $IncludeConfig /etc/rsyslog.d/*.conf
 build: |
   # Build cgroup-tools
   #
@@ -138,11 +128,6 @@ merge: |
   RUN set -e \
       && chmod 0644 /etc/cgconfig.conf
 
-  COPY compute_rsyslog.conf /etc/compute_rsyslog.conf
-  RUN chmod 0666 /etc/compute_rsyslog.conf
-  RUN chmod 0666 /var/log/
-
-
   COPY --from=libcgroup-builder /libcgroup-install/bin/*  /usr/bin/
   COPY --from=libcgroup-builder /libcgroup-install/lib/*  /usr/lib/
   COPY --from=libcgroup-builder /libcgroup-install/sbin/* /usr/sbin/

compute_tools/Cargo.toml

@@ -17,7 +17,6 @@ aws-sdk-kms.workspace = true
 aws-smithy-types.workspace = true
 anyhow.workspace = true
 axum = { workspace = true, features = [] }
-axum-extra.workspace = true
 camino.workspace = true
 chrono.workspace = true
 cfg-if.workspace = true
@@ -26,7 +25,6 @@ fail.workspace = true
 flate2.workspace = true
 futures.workspace = true
 http.workspace = true
-jsonwebtoken.workspace = true
 metrics.workspace = true
 nix.workspace = true
 notify.workspace = true

compute_tools/src/bin/compute_ctl.rs

@@ -33,27 +33,39 @@
 //!             -b /usr/local/bin/postgres \
 //!             -r http://pg-ext-s3-gateway \
 //! ```
+use std::collections::HashMap;
 use std::ffi::OsString;
 use std::fs::File;
 use std::path::Path;
 use std::process::exit;
-use std::sync::mpsc;
+use std::str::FromStr;
+use std::sync::atomic::Ordering;
+use std::sync::{Arc, Condvar, Mutex, RwLock, mpsc};
 use std::thread;
 use std::time::Duration;
 
 use anyhow::{Context, Result};
+use chrono::Utc;
 use clap::Parser;
-use compute_api::responses::ComputeCtlConfig;
+use compute_api::responses::{ComputeCtlConfig, ComputeStatus};
 use compute_api::spec::ComputeSpec;
-use compute_tools::compute::{ComputeNode, ComputeNodeParams, forward_termination_signal};
+use compute_tools::compute::{
+    ComputeNode, ComputeState, PG_PID, ParsedSpec, forward_termination_signal,
+};
+use compute_tools::configurator::launch_configurator;
+use compute_tools::disk_quota::set_disk_quota;
 use compute_tools::extension_server::get_pg_version_string;
+use compute_tools::http::server::Server;
 use compute_tools::logger::*;
+use compute_tools::lsn_lease::launch_lsn_lease_bg_task_for_static;
+use compute_tools::monitor::launch_monitor;
 use compute_tools::params::*;
 use compute_tools::spec::*;
+use compute_tools::swap::resize_swap;
 use rlimit::{Resource, setrlimit};
 use signal_hook::consts::{SIGINT, SIGQUIT, SIGTERM};
 use signal_hook::iterator::Signals;
-use tracing::{error, info};
+use tracing::{error, info, warn};
 use url::Url;
 use utils::failpoint_support;
 
@@ -152,41 +164,29 @@ fn main() -> Result<()> {
     // enable core dumping for all child processes
     setrlimit(Resource::CORE, rlimit::INFINITY, rlimit::INFINITY)?;
 
-    let connstr = Url::parse(&cli.connstr).context("cannot parse connstr as a URL")?;
+    let (pg_handle, start_pg_result) = {
+        // Enter startup tracing context
+        let _startup_context_guard = startup_context_from_env();
 
-    let cli_spec = try_spec_from_cli(&cli)?;
+        let cli_spec = try_spec_from_cli(&cli)?;
 
-    let compute_node = ComputeNode::new(
-        ComputeNodeParams {
-            compute_id: cli.compute_id,
-            connstr,
-            pgdata: cli.pgdata.clone(),
-            pgbin: cli.pgbin.clone(),
-            pgversion: get_pg_version_string(&cli.pgbin),
-            external_http_port: cli.external_http_port,
-            internal_http_port: cli.internal_http_port,
-            ext_remote_storage: cli.remote_ext_config.clone(),
-            resize_swap_on_bind: cli.resize_swap_on_bind,
-            set_disk_quota_for_fs: cli.set_disk_quota_for_fs,
-            #[cfg(target_os = "linux")]
-            filecache_connstr: cli.filecache_connstr,
-            #[cfg(target_os = "linux")]
-            cgroup: cli.cgroup,
-            #[cfg(target_os = "linux")]
-            vm_monitor_addr: cli.vm_monitor_addr,
-            build_tag,
+        let compute = wait_spec(build_tag, &cli, cli_spec)?;
 
-            live_config_allowed: cli_spec.live_config_allowed,
-        },
-        cli_spec.spec,
-        cli_spec.compute_ctl_config,
-    )?;
+        start_postgres(&cli, compute)?
 
-    let exit_code = compute_node.run()?;
+        // Startup is finished, exit the startup tracing span
+    };
+
+    // PostgreSQL is now running, if startup was successful. Wait until it exits.
+    let wait_pg_result = wait_postgres(pg_handle)?;
+
+    let delay_exit = cleanup_after_postgres_exit(start_pg_result)?;
+
+    maybe_delay_exit(delay_exit);
 
     scenario.teardown();
 
-    deinit_and_exit(exit_code);
+    deinit_and_exit(wait_pg_result);
 }
 
 async fn init() -> Result<String> {
@@ -207,6 +207,56 @@ async fn init() -> Result<String> {
     Ok(build_tag)
 }
 
+fn startup_context_from_env() -> Option<opentelemetry::ContextGuard> {
+    // Extract OpenTelemetry context for the startup actions from the
+    // TRACEPARENT and TRACESTATE env variables, and attach it to the current
+    // tracing context.
+    //
+    // This is used to propagate the context for the 'start_compute' operation
+    // from the neon control plane. This allows linking together the wider
+    // 'start_compute' operation that creates the compute container, with the
+    // startup actions here within the container.
+    //
+    // There is no standard for passing context in env variables, but a lot of
+    // tools use TRACEPARENT/TRACESTATE, so we use that convention too. See
+    // https://github.com/open-telemetry/opentelemetry-specification/issues/740
+    //
+    // Switch to the startup context here, and exit it once the startup has
+    // completed and Postgres is up and running.
+    //
+    // If this pod is pre-created without binding it to any particular endpoint
+    // yet, this isn't the right place to enter the startup context. In that
+    // case, the control plane should pass the tracing context as part of the
+    // /configure API call.
+    //
+    // NOTE: This is supposed to only cover the *startup* actions. Once
+    // postgres is configured and up-and-running, we exit this span. Any other
+    // actions that are performed on incoming HTTP requests, for example, are
+    // performed in separate spans.
+    //
+    // XXX: If the pod is restarted, we perform the startup actions in the same
+    // context as the original startup actions, which probably doesn't make
+    // sense.
+    let mut startup_tracing_carrier: HashMap<String, String> = HashMap::new();
+    if let Ok(val) = std::env::var("TRACEPARENT") {
+        startup_tracing_carrier.insert("traceparent".to_string(), val);
+    }
+    if let Ok(val) = std::env::var("TRACESTATE") {
+        startup_tracing_carrier.insert("tracestate".to_string(), val);
+    }
+    if !startup_tracing_carrier.is_empty() {
+        use opentelemetry::propagation::TextMapPropagator;
+        use opentelemetry_sdk::propagation::TraceContextPropagator;
+        let guard = TraceContextPropagator::new()
+            .extract(&startup_tracing_carrier)
+            .attach();
+        info!("startup tracing context attached");
+        Some(guard)
+    } else {
+        None
+    }
+}
+
 fn try_spec_from_cli(cli: &Cli) -> Result<CliSpecParams> {
     // First, try to get cluster spec from the cli argument
     if let Some(ref spec_json) = cli.spec_json {
@@ -257,7 +307,357 @@ struct CliSpecParams {
     live_config_allowed: bool,
 }
 
-fn deinit_and_exit(exit_code: Option<i32>) -> ! {
+fn wait_spec(
+    build_tag: String,
+    cli: &Cli,
+    CliSpecParams {
+        spec,
+        live_config_allowed,
+        compute_ctl_config: _,
+    }: CliSpecParams,
+) -> Result<Arc<ComputeNode>> {
+    let mut new_state = ComputeState::new();
+    let spec_set;
+
+    if let Some(spec) = spec {
+        let pspec = ParsedSpec::try_from(spec).map_err(|msg| anyhow::anyhow!(msg))?;
+        info!("new pspec.spec: {:?}", pspec.spec);
+        new_state.pspec = Some(pspec);
+        spec_set = true;
+    } else {
+        spec_set = false;
+    }
+    let connstr = Url::parse(&cli.connstr).context("cannot parse connstr as a URL")?;
+    let conn_conf = postgres::config::Config::from_str(connstr.as_str())
+        .context("cannot build postgres config from connstr")?;
+    let tokio_conn_conf = tokio_postgres::config::Config::from_str(connstr.as_str())
+        .context("cannot build tokio postgres config from connstr")?;
+    let compute_node = ComputeNode {
+        compute_id: cli.compute_id.clone(),
+        connstr,
+        conn_conf,
+        tokio_conn_conf,
+        pgdata: cli.pgdata.clone(),
+        pgbin: cli.pgbin.clone(),
+        pgversion: get_pg_version_string(&cli.pgbin),
+        external_http_port: cli.external_http_port,
+        internal_http_port: cli.internal_http_port,
+        live_config_allowed,
+        state: Mutex::new(new_state),
+        state_changed: Condvar::new(),
+        ext_remote_storage: cli.remote_ext_config.clone(),
+        ext_download_progress: RwLock::new(HashMap::new()),
+        build_tag,
+    };
+    let compute = Arc::new(compute_node);
+
+    // If this is a pooled VM, prewarm before starting HTTP server and becoming
+    // available for binding. Prewarming helps Postgres start quicker later,
+    // because QEMU will already have its memory allocated from the host, and
+    // the necessary binaries will already be cached.
+    if !spec_set {
+        compute.prewarm_postgres()?;
+    }
+
+    // Launch the external HTTP server first, so that we can serve control plane
+    // requests while configuration is still in progress.
+    Server::External(cli.external_http_port).launch(&compute);
+
+    // The internal HTTP server could be launched later, but there isn't much
+    // sense in waiting.
+    Server::Internal(cli.internal_http_port).launch(&compute);
+
+    if !spec_set {
+        // No spec provided, hang waiting for it.
+        info!("no compute spec provided, waiting");
+
+        let mut state = compute.state.lock().unwrap();
+        while state.status != ComputeStatus::ConfigurationPending {
+            state = compute.state_changed.wait(state).unwrap();
+
+            if state.status == ComputeStatus::ConfigurationPending {
+                info!("got spec, continue configuration");
+                // Spec is already set by the http server handler.
+                break;
+            }
+        }
+
+        // Record for how long we slept waiting for the spec.
+        let now = Utc::now();
+        state.metrics.wait_for_spec_ms = now
+            .signed_duration_since(state.start_time)
+            .to_std()
+            .unwrap()
+            .as_millis() as u64;
+
+        // Reset start time, so that the total startup time that is calculated later will
+        // not include the time that we waited for the spec.
+        state.start_time = now;
+    }
+
+    launch_lsn_lease_bg_task_for_static(&compute);
+
+    Ok(compute)
+}
+
+fn start_postgres(
+    cli: &Cli,
+    compute: Arc<ComputeNode>,
+) -> Result<(Option<PostgresHandle>, StartPostgresResult)> {
+    // We got all we need, update the state.
+    let mut state = compute.state.lock().unwrap();
+
+    // Create a tracing span for the startup operation.
+    //
+    // We could otherwise just annotate the function with #[instrument], but if
+    // we're being configured from a /configure HTTP request, we want the
+    // startup to be considered part of the /configure request.
+    let _this_entered = {
+        // Temporarily enter the /configure request's span, so that the new span
+        // becomes its child.
+        let _parent_entered = state.startup_span.take().map(|p| p.entered());
+
+        tracing::info_span!("start_postgres")
+    }
+    .entered();
+
+    state.set_status(ComputeStatus::Init, &compute.state_changed);
+
+    info!(
+        "running compute with features: {:?}",
+        state.pspec.as_ref().unwrap().spec.features
+    );
+    // before we release the mutex, fetch some parameters for later.
+    let &ComputeSpec {
+        swap_size_bytes,
+        disk_quota_bytes,
+        #[cfg(target_os = "linux")]
+        disable_lfc_resizing,
+        ..
+    } = &state.pspec.as_ref().unwrap().spec;
+    drop(state);
+
+    // Launch remaining service threads
+    let _monitor_handle = launch_monitor(&compute);
+    let _configurator_handle = launch_configurator(&compute);
+
+    let mut prestartup_failed = false;
+    let mut delay_exit = false;
+
+    // Resize swap to the desired size if the compute spec says so
+    if let (Some(size_bytes), true) = (swap_size_bytes, cli.resize_swap_on_bind) {
+        // To avoid 'swapoff' hitting postgres startup, we need to run resize-swap to completion
+        // *before* starting postgres.
+        //
+        // In theory, we could do this asynchronously if SkipSwapon was enabled for VMs, but this
+        // carries a risk of introducing hard-to-debug issues - e.g. if postgres sometimes gets
+        // OOM-killed during startup because swap wasn't available yet.
+        match resize_swap(size_bytes) {
+            Ok(()) => {
+                let size_mib = size_bytes as f32 / (1 << 20) as f32; // just for more coherent display.
+                info!(%size_bytes, %size_mib, "resized swap");
+            }
+            Err(err) => {
+                let err = err.context("failed to resize swap");
+                error!("{err:#}");
+
+                // Mark compute startup as failed; don't try to start postgres, and report this
+                // error to the control plane when it next asks.
+                prestartup_failed = true;
+                compute.set_failed_status(err);
+                delay_exit = true;
+            }
+        }
+    }
+
+    // Set disk quota if the compute spec says so
+    if let (Some(disk_quota_bytes), Some(disk_quota_fs_mountpoint)) =
+        (disk_quota_bytes, cli.set_disk_quota_for_fs.as_ref())
+    {
+        match set_disk_quota(disk_quota_bytes, disk_quota_fs_mountpoint) {
+            Ok(()) => {
+                let size_mib = disk_quota_bytes as f32 / (1 << 20) as f32; // just for more coherent display.
+                info!(%disk_quota_bytes, %size_mib, "set disk quota");
+            }
+            Err(err) => {
+                let err = err.context("failed to set disk quota");
+                error!("{err:#}");
+
+                // Mark compute startup as failed; don't try to start postgres, and report this
+                // error to the control plane when it next asks.
+                prestartup_failed = true;
+                compute.set_failed_status(err);
+                delay_exit = true;
+            }
+        }
+    }
+
+    // Start Postgres
+    let mut pg = None;
+    if !prestartup_failed {
+        pg = match compute.start_compute() {
+            Ok(pg) => {
+                info!(postmaster_pid = %pg.0.id(), "Postgres was started");
+                Some(pg)
+            }
+            Err(err) => {
+                error!("could not start the compute node: {:#}", err);
+                compute.set_failed_status(err);
+                delay_exit = true;
+                None
+            }
+        };
+    } else {
+        warn!("skipping postgres startup because pre-startup step failed");
+    }
+
+    // Start the vm-monitor if directed to. The vm-monitor only runs on linux
+    // because it requires cgroups.
+    cfg_if::cfg_if! {
+        if #[cfg(target_os = "linux")] {
+            use std::env;
+            use tokio_util::sync::CancellationToken;
+
+            // This token is used internally by the monitor to clean up all threads
+            let token = CancellationToken::new();
+
+            // don't pass postgres connection string to vm-monitor if we don't want it to resize LFC
+            let pgconnstr = if disable_lfc_resizing.unwrap_or(false) {
+                None
+            } else {
+                Some(cli.filecache_connstr.clone())
+            };
+
+            let vm_monitor = if env::var_os("AUTOSCALING").is_some() {
+                let vm_monitor = tokio::spawn(vm_monitor::start(
+                    Box::leak(Box::new(vm_monitor::Args {
+                        cgroup: Some(cli.cgroup.clone()),
+                        pgconnstr,
+                        addr: cli.vm_monitor_addr.clone(),
+                    })),
+                    token.clone(),
+                ));
+                Some(vm_monitor)
+            } else {
+                None
+            };
+        }
+    }
+
+    Ok((
+        pg,
+        StartPostgresResult {
+            delay_exit,
+            compute,
+            #[cfg(target_os = "linux")]
+            token,
+            #[cfg(target_os = "linux")]
+            vm_monitor,
+        },
+    ))
+}
+
+type PostgresHandle = (std::process::Child, tokio::task::JoinHandle<Result<()>>);
+
+struct StartPostgresResult {
+    delay_exit: bool,
+    // passed through from WaitSpecResult
+    compute: Arc<ComputeNode>,
+
+    #[cfg(target_os = "linux")]
+    token: tokio_util::sync::CancellationToken,
+    #[cfg(target_os = "linux")]
+    vm_monitor: Option<tokio::task::JoinHandle<Result<()>>>,
+}
+
+fn wait_postgres(pg: Option<PostgresHandle>) -> Result<WaitPostgresResult> {
+    // Wait for the child Postgres process forever. In this state Ctrl+C will
+    // propagate to Postgres and it will be shut down as well.
+    let mut exit_code = None;
+    if let Some((mut pg, logs_handle)) = pg {
+        info!(postmaster_pid = %pg.id(), "Waiting for Postgres to exit");
+
+        let ecode = pg
+            .wait()
+            .expect("failed to start waiting on Postgres process");
+        PG_PID.store(0, Ordering::SeqCst);
+
+        // Process has exited. Wait for the log collecting task to finish.
+        let _ = tokio::runtime::Handle::current()
+            .block_on(logs_handle)
+            .map_err(|e| tracing::error!("log task panicked: {:?}", e));
+
+        info!("Postgres exited with code {}, shutting down", ecode);
+        exit_code = ecode.code()
+    }
+
+    Ok(WaitPostgresResult { exit_code })
+}
+
+struct WaitPostgresResult {
+    exit_code: Option<i32>,
+}
+
+fn cleanup_after_postgres_exit(
+    StartPostgresResult {
+        mut delay_exit,
+        compute,
+        #[cfg(target_os = "linux")]
+        vm_monitor,
+        #[cfg(target_os = "linux")]
+        token,
+    }: StartPostgresResult,
+) -> Result<bool> {
+    // Terminate the vm_monitor so it releases the file watcher on
+    // /sys/fs/cgroup/neon-postgres.
+    // Note: the vm-monitor only runs on linux because it requires cgroups.
+    cfg_if::cfg_if! {
+        if #[cfg(target_os = "linux")] {
+            if let Some(handle) = vm_monitor {
+                // Kills all threads spawned by the monitor
+                token.cancel();
+                // Kills the actual task running the monitor
+                handle.abort();
+            }
+        }
+    }
+
+    // Maybe sync safekeepers again, to speed up next startup
+    let compute_state = compute.state.lock().unwrap().clone();
+    let pspec = compute_state.pspec.as_ref().expect("spec must be set");
+    if matches!(pspec.spec.mode, compute_api::spec::ComputeMode::Primary) {
+        info!("syncing safekeepers on shutdown");
+        let storage_auth_token = pspec.storage_auth_token.clone();
+        let lsn = compute.sync_safekeepers(storage_auth_token)?;
+        info!("synced safekeepers at lsn {lsn}");
+    }
+
+    let mut state = compute.state.lock().unwrap();
+    if state.status == ComputeStatus::TerminationPending {
+        state.status = ComputeStatus::Terminated;
+        compute.state_changed.notify_all();
+        // we were asked to terminate gracefully, don't exit to avoid restart
+        delay_exit = true
+    }
+    drop(state);
+
+    if let Err(err) = compute.check_for_core_dumps() {
+        error!("error while checking for core dumps: {err:?}");
+    }
+
+    Ok(delay_exit)
+}
+
+fn maybe_delay_exit(delay_exit: bool) {
+    // If launch failed, keep serving HTTP requests for a while, so the cloud
+    // control plane can get the actual error.
+    if delay_exit {
+        info!("giving control plane 30s to collect the error before shutdown");
+        thread::sleep(Duration::from_secs(30));
+    }
+}
+
+fn deinit_and_exit(WaitPostgresResult { exit_code }: WaitPostgresResult) -> ! {
     // Shutdown trace pipeline gracefully, so that it has a chance to send any
     // pending traces before we exit. Shutting down OTEL tracing provider may
     // hang for quite some time, see, for example:

compute_tools/src/catalog.rs

@@ -58,14 +58,14 @@ pub async fn get_database_schema(
     compute: &Arc<ComputeNode>,
     dbname: &str,
 ) -> Result<impl Stream<Item = Result<bytes::Bytes, std::io::Error>> + use<>, SchemaDumpError> {
-    let pgbin = &compute.params.pgbin;
+    let pgbin = &compute.pgbin;
     let basepath = Path::new(pgbin).parent().unwrap();
     let pgdump = basepath.join("pg_dump");
 
     // Replace the DB in the connection string and disable it to parts.
     // This is the only option to handle DBs with special characters.
-    let conf = postgres_conf_for_db(&compute.params.connstr, dbname)
-        .map_err(|_| SchemaDumpError::Unexpected)?;
+    let conf =
+        postgres_conf_for_db(&compute.connstr, dbname).map_err(|_| SchemaDumpError::Unexpected)?;
     let host = conf
         .get_hosts()
         .first()

compute_tools/src/config.rs

@@ -1,16 +1,12 @@
-use anyhow::Result;
-use std::fmt::Write as FmtWrite;
 use std::fs::{File, OpenOptions};
 use std::io;
-use std::io::Write;
 use std::io::prelude::*;
 use std::path::Path;
 
-use compute_api::spec::{ComputeAudit, ComputeMode, ComputeSpec, GenericOption};
+use anyhow::Result;
+use compute_api::spec::{ComputeMode, ComputeSpec, GenericOption};
 
-use crate::pg_helpers::{
-    GenericOptionExt, GenericOptionsSearch, PgOptionsSerialize, escape_conf_value,
-};
+use crate::pg_helpers::{GenericOptionExt, PgOptionsSerialize, escape_conf_value};
 
 /// Check that `line` is inside a text file and put it there if it is not.
 /// Create file if it doesn't exist.
@@ -59,20 +55,10 @@ pub fn write_postgres_conf(
         writeln!(file, "neon.stripe_size={stripe_size}")?;
     }
     if !spec.safekeeper_connstrings.is_empty() {
-        let mut neon_safekeepers_value = String::new();
-        tracing::info!(
-            "safekeepers_connstrings is not zero, gen: {:?}",
-            spec.safekeepers_generation
-        );
-        // If generation is given, prepend sk list with g#number:
-        if let Some(generation) = spec.safekeepers_generation {
-            write!(neon_safekeepers_value, "g#{}:", generation)?;
-        }
-        neon_safekeepers_value.push_str(&spec.safekeeper_connstrings.join(","));
         writeln!(
             file,
             "neon.safekeepers={}",
-            escape_conf_value(&neon_safekeepers_value)
+            escape_conf_value(&spec.safekeeper_connstrings.join(","))
         )?;
     }
     if let Some(s) = &spec.tenant_id {
@@ -99,7 +85,6 @@ pub fn write_postgres_conf(
         writeln!(file, "lc_numeric='C.UTF-8'")?;
     }
 
-    writeln!(file, "neon.endpoint_type={}", spec.mode.to_type_str())?;
     match spec.mode {
         ComputeMode::Primary => {}
         ComputeMode::Static(lsn) => {
@@ -141,54 +126,6 @@ pub fn write_postgres_conf(
         writeln!(file, "# Managed by compute_ctl: end")?;
     }
 
-    // If audit logging is enabled, configure pgaudit.
-    //
-    // Note, that this is called after the settings from spec are written.
-    // This way we always override the settings from the spec
-    // and don't allow the user or the control plane admin to change them.
-    if let ComputeAudit::Hipaa = spec.audit_log_level {
-        writeln!(file, "# Managed by compute_ctl audit settings: begin")?;
-        // This log level is very verbose
-        // but this is necessary for HIPAA compliance.
-        writeln!(file, "pgaudit.log='all'")?;
-        writeln!(file, "pgaudit.log_parameter=on")?;
-        // Disable logging of catalog queries
-        // The catalog doesn't contain sensitive data, so we don't need to audit it.
-        writeln!(file, "pgaudit.log_catalog=off")?;
-        // Set log rotation to 5 minutes
-        // TODO: tune this after performance testing
-        writeln!(file, "pgaudit.log_rotation_age=5")?;
-
-        // Add audit shared_preload_libraries, if they are not present.
-        //
-        // The caller who sets the flag is responsible for ensuring that the necessary
-        // shared_preload_libraries are present in the compute image,
-        // otherwise the compute start will fail.
-        if let Some(libs) = spec.cluster.settings.find("shared_preload_libraries") {
-            let mut extra_shared_preload_libraries = String::new();
-            if !libs.contains("pgaudit") {
-                extra_shared_preload_libraries.push_str(",pgaudit");
-            }
-            if !libs.contains("pgauditlogtofile") {
-                extra_shared_preload_libraries.push_str(",pgauditlogtofile");
-            }
-            writeln!(
-                file,
-                "shared_preload_libraries='{}{}'",
-                libs, extra_shared_preload_libraries
-            )?;
-        } else {
-            // Typically, this should be unreacheable,
-            // because we always set at least some shared_preload_libraries in the spec
-            // but let's handle it explicitly anyway.
-            writeln!(
-                file,
-                "shared_preload_libraries='neon,pgaudit,pgauditlogtofile'"
-            )?;
-        }
-        writeln!(file, "# Managed by compute_ctl audit settings: end")?;
-    }
-
     writeln!(file, "neon.extension_server_port={}", extension_server_port)?;
 
     if spec.drop_subscriptions_before_start {

compute_tools/src/config_template/compute_audit_rsyslog_template.conf

@@ -1,10 +0,0 @@
-# Load imfile module to read log files
-module(load="imfile")
-
-# Input configuration for log files in the specified directory
-# Replace {log_directory} with the directory containing the log files
-input(type="imfile" File="{log_directory}/*.log" Tag="{tag}" Severity="info" Facility="local0")
-global(workDirectory="/var/log")
-
-# Forward logs to remote syslog server
-*.* @@{remote_endpoint}

compute_tools/src/extension_server.rs

@@ -202,24 +202,8 @@ pub async fn download_extension(
     // move contents of the libdir / sharedir in unzipped archive to the correct local paths
     for paths in [sharedir_paths, libdir_paths] {
         let (zip_dir, real_dir) = paths;
-
-        let dir = match std::fs::read_dir(&zip_dir) {
-            Ok(dir) => dir,
-            Err(e) => match e.kind() {
-                // In the event of a SQL-only extension, there would be nothing
-                // to move from the lib/ directory, so note that in the log and
-                // move on.
-                std::io::ErrorKind::NotFound => {
-                    info!("nothing to move from {}", zip_dir);
-                    continue;
-                }
-                _ => return Err(anyhow::anyhow!(e)),
-            },
-        };
-
         info!("mv {zip_dir:?}/*  {real_dir:?}");
-
-        for file in dir {
+        for file in std::fs::read_dir(zip_dir)? {
             let old_file = file?.path();
             let new_file =
                 Path::new(&real_dir).join(old_file.file_name().context("error parsing file")?);
@@ -269,31 +253,27 @@ pub fn create_control_files(remote_extensions: &RemoteExtSpec, pgbin: &str) {
     }
 }
 
-// Do request to extension storage proxy, e.g.,
+// Do request to extension storage proxy, i.e.
 // curl http://pg-ext-s3-gateway/latest/v15/extensions/anon.tar.zst
-// using HTTP GET and return the response body as bytes.
+// using HHTP GET
+// and return the response body as bytes
+//
 async fn download_extension_tar(ext_remote_storage: &str, ext_path: &str) -> Result<Bytes> {
     let uri = format!("{}/{}", ext_remote_storage, ext_path);
-    let filename = Path::new(ext_path)
-        .file_name()
-        .unwrap_or_else(|| std::ffi::OsStr::new("unknown"))
-        .to_str()
-        .unwrap_or("unknown")
-        .to_string();
 
-    info!("Downloading extension file '{}' from uri {}", filename, uri);
+    info!("Download extension {} from uri {}", ext_path, uri);
 
     match do_extension_server_request(&uri).await {
         Ok(resp) => {
             info!("Successfully downloaded remote extension data {}", ext_path);
             REMOTE_EXT_REQUESTS_TOTAL
-                .with_label_values(&[&StatusCode::OK.to_string(), &filename])
+                .with_label_values(&[&StatusCode::OK.to_string()])
                 .inc();
             Ok(resp)
         }
         Err((msg, status)) => {
             REMOTE_EXT_REQUESTS_TOTAL
-                .with_label_values(&[&status, &filename])
+                .with_label_values(&[&status])
                 .inc();
             bail!(msg);
         }

compute_tools/src/http/extract/mod.rs

@@ -1,9 +1,7 @@
 pub(crate) mod json;
 pub(crate) mod path;
 pub(crate) mod query;
-pub(crate) mod request_id;
 
 pub(crate) use json::Json;
 pub(crate) use path::Path;
 pub(crate) use query::Query;
-pub(crate) use request_id::RequestId;

compute_tools/src/http/extract/request_id.rs

@@ -1,86 +0,0 @@
-use std::{
-    fmt::Display,
-    ops::{Deref, DerefMut},
-};
-
-use axum::{extract::FromRequestParts, response::IntoResponse};
-use http::{StatusCode, request::Parts};
-
-use crate::http::{JsonResponse, headers::X_REQUEST_ID};
-
-/// Extract the request ID from the `X-Request-Id` header.
-#[derive(Debug, Clone, Default)]
-pub(crate) struct RequestId(pub String);
-
-#[derive(Debug)]
-/// Rejection used for [`RequestId`].
-///
-/// Contains one variant for each way the [`RequestId`] extractor can
-/// fail.
-pub(crate) enum RequestIdRejection {
-    /// The request is missing the header.
-    MissingRequestId,
-
-    /// The value of the header is invalid UTF-8.
-    InvalidUtf8,
-}
-
-impl RequestIdRejection {
-    pub fn status(&self) -> StatusCode {
-        match self {
-            RequestIdRejection::MissingRequestId => StatusCode::INTERNAL_SERVER_ERROR,
-            RequestIdRejection::InvalidUtf8 => StatusCode::BAD_REQUEST,
-        }
-    }
-
-    pub fn message(&self) -> String {
-        match self {
-            RequestIdRejection::MissingRequestId => "request ID is missing",
-            RequestIdRejection::InvalidUtf8 => "request ID is invalid UTF-8",
-        }
-        .to_string()
-    }
-}
-
-impl IntoResponse for RequestIdRejection {
-    fn into_response(self) -> axum::response::Response {
-        JsonResponse::error(self.status(), self.message())
-    }
-}
-
-impl<S> FromRequestParts<S> for RequestId
-where
-    S: Send + Sync,
-{
-    type Rejection = RequestIdRejection;
-
-    async fn from_request_parts(parts: &mut Parts, _state: &S) -> Result<Self, Self::Rejection> {
-        match parts.headers.get(X_REQUEST_ID) {
-            Some(value) => match value.to_str() {
-                Ok(request_id) => Ok(Self(request_id.to_string())),
-                Err(_) => Err(RequestIdRejection::InvalidUtf8),
-            },
-            None => Err(RequestIdRejection::MissingRequestId),
-        }
-    }
-}
-
-impl Deref for RequestId {
-    type Target = String;
-
-    fn deref(&self) -> &Self::Target {
-        &self.0
-    }
-}
-
-impl DerefMut for RequestId {
-    fn deref_mut(&mut self) -> &mut Self::Target {
-        &mut self.0
-    }
-}
-
-impl Display for RequestId {
-    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
-        f.write_str(&self.0)
-    }
-}

compute_tools/src/http/headers.rs

@@ -1,2 +0,0 @@
-/// Constant for `X-Request-Id` header.
-pub const X_REQUEST_ID: &str = "x-request-id";

compute_tools/src/http/middleware/authorize.rs

@@ -1,145 +0,0 @@
-use std::{collections::HashSet, net::SocketAddr};
-
-use anyhow::{Result, anyhow};
-use axum::{RequestExt, body::Body, extract::ConnectInfo};
-use axum_extra::{
-    TypedHeader,
-    headers::{Authorization, authorization::Bearer},
-};
-use futures::future::BoxFuture;
-use http::{Request, Response, StatusCode};
-use jsonwebtoken::{Algorithm, DecodingKey, TokenData, Validation, jwk::JwkSet};
-use serde::Deserialize;
-use tower_http::auth::AsyncAuthorizeRequest;
-use tracing::warn;
-
-use crate::http::{JsonResponse, extract::RequestId};
-
-#[derive(Clone, Debug, Deserialize)]
-pub(in crate::http) struct Claims {
-    compute_id: String,
-}
-
-#[derive(Clone, Debug)]
-pub(in crate::http) struct Authorize {
-    compute_id: String,
-    jwks: JwkSet,
-    validation: Validation,
-}
-
-impl Authorize {
-    pub fn new(compute_id: String, jwks: JwkSet) -> Self {
-        let mut validation = Validation::new(Algorithm::EdDSA);
-        // Nothing is currently required
-        validation.required_spec_claims = HashSet::new();
-        validation.validate_exp = true;
-        // Unused by the control plane
-        validation.validate_aud = false;
-        // Unused by the control plane
-        validation.validate_nbf = false;
-
-        Self {
-            compute_id,
-            jwks,
-            validation,
-        }
-    }
-}
-
-impl AsyncAuthorizeRequest<Body> for Authorize {
-    type RequestBody = Body;
-    type ResponseBody = Body;
-    type Future = BoxFuture<'static, Result<Request<Body>, Response<Self::ResponseBody>>>;
-
-    fn authorize(&mut self, mut request: Request<Body>) -> Self::Future {
-        let compute_id = self.compute_id.clone();
-        let jwks = self.jwks.clone();
-        let validation = self.validation.clone();
-
-        Box::pin(async move {
-            let request_id = request.extract_parts::<RequestId>().await.unwrap();
-
-            // TODO: Remove this check after a successful rollout
-            if jwks.keys.is_empty() {
-                warn!(%request_id, "Authorization has not been configured");
-
-                return Ok(request);
-            }
-
-            let connect_info = request
-                .extract_parts::<ConnectInfo<SocketAddr>>()
-                .await
-                .unwrap();
-
-            // In the event the request is coming from the loopback interface,
-            // allow all requests
-            if connect_info.ip().is_loopback() {
-                warn!(%request_id, "Bypassed authorization because request is coming from the loopback interface");
-
-                return Ok(request);
-            }
-
-            let TypedHeader(Authorization(bearer)) = request
-                .extract_parts::<TypedHeader<Authorization<Bearer>>>()
-                .await
-                .map_err(|_| {
-                    JsonResponse::error(StatusCode::BAD_REQUEST, "invalid authorization token")
-                })?;
-
-            let data = match Self::verify(&jwks, bearer.token(), &validation) {
-                Ok(claims) => claims,
-                Err(e) => return Err(JsonResponse::error(StatusCode::UNAUTHORIZED, e)),
-            };
-
-            if data.claims.compute_id != compute_id {
-                return Err(JsonResponse::error(
-                    StatusCode::UNAUTHORIZED,
-                    "invalid claims in authorization token",
-                ));
-            }
-
-            // Make claims available to any subsequent middleware or request
-            // handlers
-            request.extensions_mut().insert(data.claims);
-
-            Ok(request)
-        })
-    }
-}
-
-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,
-                Err(e) => {
-                    warn!(
-                        "Failed to construct decoding key from {}: {}",
-                        jwk.common.key_id.as_ref().unwrap(),
-                        e
-                    );
-
-                    continue;
-                }
-            };
-
-            match jsonwebtoken::decode::<Claims>(token, &decoding_key, validation) {
-                Ok(data) => return Ok(data),
-                Err(e) => {
-                    warn!(
-                        "Failed to decode authorization token using {}: {}",
-                        jwk.common.key_id.as_ref().unwrap(),
-                        e
-                    );
-
-                    continue;
-                }
-            }
-        }
-
-        Err(anyhow!("Failed to verify authorization token"))
-    }
-}

compute_tools/src/http/middleware/mod.rs

@@ -1 +0,0 @@
-pub(in crate::http) mod authorize;

compute_tools/src/http/mod.rs

@@ -7,8 +7,6 @@ use serde::Serialize;
 use tracing::error;
 
 mod extract;
-mod headers;
-mod middleware;
 mod routes;
 pub mod server;
 

compute_tools/src/http/routes/configure.rs

@@ -22,7 +22,7 @@ pub(in crate::http) async fn configure(
     State(compute): State<Arc<ComputeNode>>,
     request: Json<ConfigurationRequest>,
 ) -> Response {
-    if !compute.params.live_config_allowed {
+    if !compute.live_config_allowed {
         return JsonResponse::error(
             StatusCode::PRECONDITION_FAILED,
             "live configuration is not allowed for this compute node".to_string(),

compute_tools/src/http/routes/extension_server.rs

@@ -18,11 +18,11 @@ pub(in crate::http) struct ExtensionServerParams {
 /// Download a remote extension.
 pub(in crate::http) async fn download_extension(
     Path(filename): Path<String>,
-    ext_server_params: Query<ExtensionServerParams>,
+    params: Query<ExtensionServerParams>,
     State(compute): State<Arc<ComputeNode>>,
 ) -> Response {
     // Don't even try to download extensions if no remote storage is configured
-    if compute.params.ext_remote_storage.is_none() {
+    if compute.ext_remote_storage.is_none() {
         return JsonResponse::error(
             StatusCode::PRECONDITION_FAILED,
             "remote storage is not configured",
@@ -46,9 +46,9 @@ pub(in crate::http) async fn download_extension(
 
         remote_extensions.get_ext(
             &filename,
-            ext_server_params.is_library,
-            &compute.params.build_tag,
-            &compute.params.pgversion,
+            params.is_library,
+            &compute.build_tag,
+            &compute.pgversion,
         )
     };
 

compute_tools/src/http/server.rs

@@ -10,58 +10,48 @@ use axum::middleware::{self, Next};
 use axum::response::{IntoResponse, Response};
 use axum::routing::{get, post};
 use http::StatusCode;
-use jsonwebtoken::jwk::JwkSet;
 use tokio::net::TcpListener;
 use tower::ServiceBuilder;
-use tower_http::{
-    auth::AsyncRequireAuthorizationLayer, request_id::PropagateRequestIdLayer, trace::TraceLayer,
-};
-use tracing::{Span, error, info};
+use tower_http::request_id::PropagateRequestIdLayer;
+use tower_http::trace::TraceLayer;
+use tracing::{Span, debug, error, info};
 use uuid::Uuid;
 
-use super::{
-    headers::X_REQUEST_ID,
-    middleware::authorize::Authorize,
-    routes::{
-        check_writability, configure, database_schema, dbs_and_roles, extension_server, extensions,
-        grants, insights, metrics, metrics_json, status, terminate,
-    },
+use super::routes::{
+    check_writability, configure, database_schema, dbs_and_roles, extension_server, extensions,
+    grants, insights, metrics, metrics_json, status, terminate,
 };
 use crate::compute::ComputeNode;
 
+const X_REQUEST_ID: &str = "x-request-id";
+
 /// `compute_ctl` has two servers: internal and external. The internal server
 /// binds to the loopback interface and handles communication from clients on
 /// the compute. The external server is what receives communication from the
 /// control plane, the metrics scraper, etc. We make the distinction because
 /// certain routes in `compute_ctl` only need to be exposed to local processes
 /// like Postgres via the neon extension and local_proxy.
-#[derive(Clone, Debug)]
+#[derive(Clone, Copy, Debug)]
 pub enum Server {
-    Internal {
-        port: u16,
-    },
-    External {
-        port: u16,
-        jwks: JwkSet,
-        compute_id: String,
-    },
+    Internal(u16),
+    External(u16),
 }
 
 impl Display for Server {
     fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
         match self {
-            Server::Internal { .. } => f.write_str("internal"),
-            Server::External { .. } => f.write_str("external"),
+            Server::Internal(_) => f.write_str("internal"),
+            Server::External(_) => f.write_str("external"),
         }
     }
 }
 
-impl From<&Server> for Router<Arc<ComputeNode>> {
-    fn from(server: &Server) -> Self {
+impl From<Server> for Router<Arc<ComputeNode>> {
+    fn from(server: Server) -> Self {
         let mut router = Router::<Arc<ComputeNode>>::new();
 
         router = match server {
-            Server::Internal { .. } => {
+            Server::Internal(_) => {
                 router = router
                     .route(
                         "/extension_server/{*filename}",
@@ -79,71 +69,59 @@ impl From<&Server> for Router<Arc<ComputeNode>> {
 
                 router
             }
-            Server::External {
-                jwks, compute_id, ..
-            } => {
-                let unauthenticated_router =
-                    Router::<Arc<ComputeNode>>::new().route("/metrics", get(metrics::get_metrics));
-
-                let authenticated_router = Router::<Arc<ComputeNode>>::new()
-                    .route("/check_writability", post(check_writability::is_writable))
-                    .route("/configure", post(configure::configure))
-                    .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))
-                    .route("/metrics.json", get(metrics_json::get_metrics))
-                    .route("/status", get(status::get_status))
-                    .route("/terminate", post(terminate::terminate))
-                    .layer(AsyncRequireAuthorizationLayer::new(Authorize::new(
-                        compute_id.clone(),
-                        jwks.clone(),
-                    )));
-
-                router
-                    .merge(unauthenticated_router)
-                    .merge(authenticated_router)
-            }
+            Server::External(_) => router
+                .route("/check_writability", post(check_writability::is_writable))
+                .route("/configure", post(configure::configure))
+                .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))
+                .route("/metrics", get(metrics::get_metrics))
+                .route("/metrics.json", get(metrics_json::get_metrics))
+                .route("/status", get(status::get_status))
+                .route("/terminate", post(terminate::terminate)),
         };
 
-        router
-            .fallback(Server::handle_404)
-            .method_not_allowed_fallback(Server::handle_405)
-            .layer(
-                ServiceBuilder::new()
-                    .layer(tower_otel::trace::HttpLayer::server(tracing::Level::INFO))
-                    // Add this middleware since we assume the request ID exists
-                    .layer(middleware::from_fn(maybe_add_request_id_header))
-                    .layer(
-                        TraceLayer::new_for_http()
-                            .on_request(|request: &http::Request<_>, _span: &Span| {
-                                let request_id = request
+        router.fallback(Server::handle_404).method_not_allowed_fallback(Server::handle_405).layer(
+            ServiceBuilder::new()
+                // Add this middleware since we assume the request ID exists
+                .layer(middleware::from_fn(maybe_add_request_id_header))
+                .layer(
+                    TraceLayer::new_for_http()
+                        .on_request(|request: &http::Request<_>, _span: &Span| {
+                            let request_id = request
+                                .headers()
+                                .get(X_REQUEST_ID)
+                                .unwrap()
+                                .to_str()
+                                .unwrap();
+
+                            match request.uri().path() {
+                                "/metrics" => {
+                                    debug!(%request_id, "{} {}", request.method(), request.uri())
+                                }
+                                _ => info!(%request_id, "{} {}", request.method(), request.uri()),
+                            };
+                        })
+                        .on_response(
+                            |response: &http::Response<_>, latency: Duration, _span: &Span| {
+                                let request_id = response
                                     .headers()
                                     .get(X_REQUEST_ID)
                                     .unwrap()
                                     .to_str()
                                     .unwrap();
 
-                                info!(%request_id, "{} {}", request.method(), request.uri());
-                            })
-                            .on_response(
-                                |response: &http::Response<_>, latency: Duration, _span: &Span| {
-                                    let request_id = response
-                                        .headers()
-                                        .get(X_REQUEST_ID)
-                                        .unwrap()
-                                        .to_str()
-                                        .unwrap();
-
-                                    info!(
-                                        %request_id,
-                                        code = response.status().as_u16(),
-                                        latency = latency.as_millis()
-                                    );
-                                },
-                            ),
-                    )
-                    .layer(PropagateRequestIdLayer::x_request_id()),
-            )
+                                info!(
+                                    %request_id,
+                                    code = response.status().as_u16(),
+                                    latency = latency.as_millis()
+                                )
+                            },
+                        ),
+                )
+                .layer(PropagateRequestIdLayer::x_request_id()),
+        )
+            .layer(tower_otel::trace::HttpLayer::server(tracing::Level::INFO))
     }
 }
 
@@ -167,15 +145,15 @@ impl Server {
         match self {
             // TODO: Change this to Ipv6Addr::LOCALHOST when the GitHub runners
             // allow binding to localhost
-            Server::Internal { .. } => IpAddr::from(Ipv6Addr::UNSPECIFIED),
-            Server::External { .. } => IpAddr::from(Ipv6Addr::UNSPECIFIED),
+            Server::Internal(_) => IpAddr::from(Ipv6Addr::UNSPECIFIED),
+            Server::External(_) => IpAddr::from(Ipv6Addr::UNSPECIFIED),
         }
     }
 
-    fn port(&self) -> u16 {
+    fn port(self) -> u16 {
         match self {
-            Server::Internal { port, .. } => *port,
-            Server::External { port, .. } => *port,
+            Server::Internal(port) => port,
+            Server::External(port) => port,
         }
     }
 
@@ -202,9 +180,7 @@ impl Server {
             );
         }
 
-        let router = Router::from(&self)
-            .with_state(compute)
-            .into_make_service_with_connect_info::<SocketAddr>();
+        let router = Router::from(self).with_state(compute);
 
         if let Err(e) = axum::serve(listener, router).await {
             error!("compute_ctl {} HTTP server error: {}", self, e);

compute_tools/src/lib.rs

@@ -21,7 +21,6 @@ mod migration;
 pub mod monitor;
 pub mod params;
 pub mod pg_helpers;
-pub mod rsyslog;
 pub mod spec;
 mod spec_apply;
 pub mod swap;

compute_tools/src/logger.rs

@@ -1,5 +1,3 @@
-use std::collections::HashMap;
-use tracing::info;
 use tracing_subscriber::layer::SubscriberExt;
 use tracing_subscriber::prelude::*;
 
@@ -44,50 +42,3 @@ pub async fn init_tracing_and_logging(default_log_level: &str) -> anyhow::Result
 pub fn inlinify(s: &str) -> String {
     s.replace('\n', "\u{200B}")
 }
-
-pub fn startup_context_from_env() -> Option<opentelemetry::Context> {
-    // Extract OpenTelemetry context for the startup actions from the
-    // TRACEPARENT and TRACESTATE env variables, and attach it to the current
-    // tracing context.
-    //
-    // This is used to propagate the context for the 'start_compute' operation
-    // from the neon control plane. This allows linking together the wider
-    // 'start_compute' operation that creates the compute container, with the
-    // startup actions here within the container.
-    //
-    // There is no standard for passing context in env variables, but a lot of
-    // tools use TRACEPARENT/TRACESTATE, so we use that convention too. See
-    // https://github.com/open-telemetry/opentelemetry-specification/issues/740
-    //
-    // Switch to the startup context here, and exit it once the startup has
-    // completed and Postgres is up and running.
-    //
-    // If this pod is pre-created without binding it to any particular endpoint
-    // yet, this isn't the right place to enter the startup context. In that
-    // case, the control plane should pass the tracing context as part of the
-    // /configure API call.
-    //
-    // NOTE: This is supposed to only cover the *startup* actions. Once
-    // postgres is configured and up-and-running, we exit this span. Any other
-    // actions that are performed on incoming HTTP requests, for example, are
-    // performed in separate spans.
-    //
-    // XXX: If the pod is restarted, we perform the startup actions in the same
-    // context as the original startup actions, which probably doesn't make
-    // sense.
-    let mut startup_tracing_carrier: HashMap<String, String> = HashMap::new();
-    if let Ok(val) = std::env::var("TRACEPARENT") {
-        startup_tracing_carrier.insert("traceparent".to_string(), val);
-    }
-    if let Ok(val) = std::env::var("TRACESTATE") {
-        startup_tracing_carrier.insert("tracestate".to_string(), val);
-    }
-    if !startup_tracing_carrier.is_empty() {
-        use opentelemetry::propagation::TextMapPropagator;
-        use opentelemetry_sdk::propagation::TraceContextPropagator;
-        info!("got startup tracing context from env variables");
-        Some(TraceContextPropagator::new().extract(&startup_tracing_carrier))
-    } else {
-        None
-    }
-}

compute_tools/src/metrics.rs

@@ -54,7 +54,9 @@ pub(crate) static REMOTE_EXT_REQUESTS_TOTAL: Lazy<IntCounterVec> = Lazy::new(||
     register_int_counter_vec!(
         "compute_ctl_remote_ext_requests_total",
         "Total number of requests made by compute_ctl to download extensions from S3 proxy by status",
-        &["http_status", "filename"]
+        // Do not use any labels like extension name yet.
+        // We can add them later if needed.
+        &["http_status"]
     )
     .expect("failed to define a metric")
 });

compute_tools/src/monitor.rs

@@ -18,7 +18,7 @@ const MONITOR_CHECK_INTERVAL: Duration = Duration::from_millis(500);
 // should be handled gracefully.
 fn watch_compute_activity(compute: &ComputeNode) {
     // Suppose that `connstr` doesn't change
-    let connstr = compute.params.connstr.clone();
+    let connstr = compute.connstr.clone();
     let conf = compute.get_conn_conf(Some("compute_ctl:activity_monitor"));
 
     // During startup and configuration we connect to every Postgres database,

compute_tools/src/pg_helpers.rs

@@ -186,40 +186,15 @@ impl DatabaseExt for Database {
 /// Postgres SQL queries and DATABASE_URL.
 pub trait Escaping {
     fn pg_quote(&self) -> String;
-    fn pg_quote_dollar(&self) -> (String, String);
 }
 
 impl Escaping for PgIdent {
     /// This is intended to mimic Postgres quote_ident(), but for simplicity it
     /// always quotes provided string with `""` and escapes every `"`.
     /// **Not idempotent**, i.e. if string is already escaped it will be escaped again.
-    /// N.B. it's not useful for escaping identifiers that are used inside WHERE
-    /// clause, use `escape_literal()` instead.
     fn pg_quote(&self) -> String {
-        format!("\"{}\"", self.replace('"', "\"\""))
-    }
-
-    /// This helper is intended to be used for dollar-escaping strings for usage
-    /// inside PL/pgSQL procedures. In addition to dollar-escaping the string,
-    /// it also returns a tag that is intended to be used inside the outer
-    /// PL/pgSQL procedure. If you do not need an outer tag, just discard it.
-    /// Here we somewhat mimic the logic of Postgres' `pg_get_functiondef()`,
-    /// <https://github.com/postgres/postgres/blob/8b49392b270b4ac0b9f5c210e2a503546841e832/src/backend/utils/adt/ruleutils.c#L2924>
-    fn pg_quote_dollar(&self) -> (String, String) {
-        let mut tag: String = "".to_string();
-        let mut outer_tag = "x".to_string();
-
-        // Find the first suitable tag that is not present in the string.
-        // Postgres' max role/DB name length is 63 bytes, so even in the
-        // worst case it won't take long.
-        while self.contains(&format!("${tag}$")) || self.contains(&format!("${outer_tag}$")) {
-            tag += "x";
-            outer_tag = tag.clone() + "x";
-        }
-
-        let escaped = format!("${tag}${self}${tag}$");
-
-        (escaped, outer_tag)
+        let result = format!("\"{}\"", self.replace('"', "\"\""));
+        result
     }
 }
 
@@ -251,13 +226,10 @@ pub async fn get_existing_dbs_async(
     // invalid state. See:
     //   https://github.com/postgres/postgres/commit/a4b4cc1d60f7e8ccfcc8ff8cb80c28ee411ad9a9
     let rowstream = client
-        // We use a subquery instead of a fancy `datdba::regrole::text AS owner`,
-        // because the latter automatically wraps the result in double quotes,
-        // if the role name contains special characters.
         .query_raw::<str, &String, &[String; 0]>(
             "SELECT
                 datname AS name,
-                (SELECT rolname FROM pg_roles WHERE oid = datdba) AS owner,
+                datdba::regrole::text AS owner,
                 NOT datallowconn AS restrict_conn,
                 datconnlimit = - 2 AS invalid
             FROM

compute_tools/src/rsyslog.rs

@@ -1,77 +0,0 @@
-use std::process::Command;
-use std::{fs::OpenOptions, io::Write};
-
-use anyhow::{Context, Result};
-use tracing::info;
-
-fn get_rsyslog_pid() -> Option<String> {
-    let output = Command::new("pgrep")
-        .arg("rsyslogd")
-        .output()
-        .expect("Failed to execute pgrep");
-
-    if !output.stdout.is_empty() {
-        let pid = std::str::from_utf8(&output.stdout)
-            .expect("Invalid UTF-8 in process output")
-            .trim()
-            .to_string();
-        Some(pid)
-    } else {
-        None
-    }
-}
-
-// Restart rsyslogd to apply the new configuration.
-// This is necessary, because there is no other way to reload the rsyslog configuration.
-//
-// Rsyslogd shouldn't lose any messages, because of the restart,
-// because it tracks the last read position in the log files
-// and will continue reading from that position.
-// TODO: test it properly
-//
-fn restart_rsyslog() -> Result<()> {
-    let old_pid = get_rsyslog_pid().context("rsyslogd is not running")?;
-    info!("rsyslogd is running with pid: {}, restart it", old_pid);
-
-    // kill it to restart
-    let _ = Command::new("pkill")
-        .arg("rsyslogd")
-        .output()
-        .context("Failed to stop rsyslogd")?;
-
-    Ok(())
-}
-
-pub fn configure_audit_rsyslog(
-    log_directory: &str,
-    tag: &str,
-    remote_endpoint: &str,
-) -> Result<()> {
-    let config_content: String = format!(
-        include_str!("config_template/compute_audit_rsyslog_template.conf"),
-        log_directory = log_directory,
-        tag = tag,
-        remote_endpoint = remote_endpoint
-    );
-
-    info!("rsyslog config_content: {}", config_content);
-
-    let rsyslog_conf_path = "/etc/rsyslog.d/compute_audit_rsyslog.conf";
-    let mut file = OpenOptions::new()
-        .create(true)
-        .write(true)
-        .truncate(true)
-        .open(rsyslog_conf_path)?;
-
-    file.write_all(config_content.as_bytes())?;
-
-    info!(
-        "rsyslog configuration file {} added successfully. Starting rsyslogd",
-        rsyslog_conf_path
-    );
-
-    // start the service, using the configuration
-    restart_rsyslog()?;
-
-    Ok(())
-}

compute_tools/src/spec_apply.rs

@@ -6,22 +6,21 @@ use std::sync::Arc;
 
 use anyhow::{Context, Result};
 use compute_api::responses::ComputeStatus;
-use compute_api::spec::{ComputeAudit, ComputeFeature, ComputeSpec, Database, PgIdent, Role};
+use compute_api::spec::{ComputeFeature, ComputeSpec, Database, PgIdent, Role};
 use futures::future::join_all;
 use tokio::sync::RwLock;
 use tokio_postgres::Client;
 use tokio_postgres::error::SqlState;
 use tracing::{Instrument, debug, error, info, info_span, instrument, warn};
 
-use crate::compute::{ComputeNode, ComputeState};
+use crate::compute::{ComputeNode, ComputeState, construct_superuser_query};
 use crate::pg_helpers::{
-    DatabaseExt, Escaping, GenericOptionsSearch, RoleExt, get_existing_dbs_async,
+    DatabaseExt, Escaping, GenericOptionsSearch, RoleExt, escape_literal, get_existing_dbs_async,
     get_existing_roles_async,
 };
 use crate::spec_apply::ApplySpecPhase::{
-    CreateAndAlterDatabases, CreateAndAlterRoles, CreateAvailabilityCheck, CreateNeonSuperuser,
-    CreatePgauditExtension, CreatePgauditlogtofileExtension, CreateSchemaNeon,
-    DisablePostgresDBPgAudit, DropInvalidDatabases, DropRoles, FinalizeDropLogicalSubscriptions,
+    CreateAndAlterDatabases, CreateAndAlterRoles, CreateAvailabilityCheck, CreateSchemaNeon,
+    CreateSuperUser, DropInvalidDatabases, DropRoles, FinalizeDropLogicalSubscriptions,
     HandleNeonExtension, HandleOtherExtensions, RenameAndDeleteDatabases, RenameRoles,
     RunInEachDatabase,
 };
@@ -188,7 +187,7 @@ impl ComputeNode {
             }
 
             for phase in [
-                CreateNeonSuperuser,
+                CreateSuperUser,
                 DropInvalidDatabases,
                 RenameRoles,
                 CreateAndAlterRoles,
@@ -278,19 +277,6 @@ impl ComputeNode {
                 phases.push(FinalizeDropLogicalSubscriptions);
             }
 
-            // Keep DisablePostgresDBPgAudit phase at the end,
-            // so that all config operations are audit logged.
-            match spec.audit_log_level
-            {
-                ComputeAudit::Hipaa => {
-                    phases.push(CreatePgauditExtension);
-                    phases.push(CreatePgauditlogtofileExtension);
-                    phases.push(DisablePostgresDBPgAudit);
-                }
-                ComputeAudit::Log => { /* not implemented yet */ }
-                ComputeAudit::Disabled => {}
-            }
-
             for phase in phases {
                 debug!("Applying phase {:?}", &phase);
                 apply_operations(
@@ -469,7 +455,7 @@ pub enum PerDatabasePhase {
 
 #[derive(Clone, Debug)]
 pub enum ApplySpecPhase {
-    CreateNeonSuperuser,
+    CreateSuperUser,
     DropInvalidDatabases,
     RenameRoles,
     CreateAndAlterRoles,
@@ -477,9 +463,6 @@ pub enum ApplySpecPhase {
     CreateAndAlterDatabases,
     CreateSchemaNeon,
     RunInEachDatabase { db: DB, subphase: PerDatabasePhase },
-    CreatePgauditExtension,
-    CreatePgauditlogtofileExtension,
-    DisablePostgresDBPgAudit,
     HandleOtherExtensions,
     HandleNeonExtension,
     CreateAvailabilityCheck,
@@ -596,10 +579,14 @@ async fn get_operations<'a>(
     apply_spec_phase: &'a ApplySpecPhase,
 ) -> Result<Box<dyn Iterator<Item = Operation> + 'a + Send>> {
     match apply_spec_phase {
-        ApplySpecPhase::CreateNeonSuperuser => Ok(Box::new(once(Operation {
-            query: include_str!("sql/create_neon_superuser.sql").to_string(),
-            comment: None,
-        }))),
+        ApplySpecPhase::CreateSuperUser => {
+            let query = construct_superuser_query(spec);
+
+            Ok(Box::new(once(Operation {
+                query,
+                comment: None,
+            })))
+        }
         ApplySpecPhase::DropInvalidDatabases => {
             let mut ctx = ctx.write().await;
             let databases = &mut ctx.dbs;
@@ -733,15 +720,14 @@ async fn get_operations<'a>(
                         // We do not check whether the DB exists or not,
                         // Postgres will take care of it for us
                         "delete_db" => {
-                            let (db_name, outer_tag) = op.name.pg_quote_dollar();
                             // In Postgres we can't drop a database if it is a template.
                             // So we need to unset the template flag first, but it could
                             // be a retry, so we could've already dropped the database.
                             // Check that database exists first to make it idempotent.
                             let unset_template_query: String = format!(
                                 include_str!("sql/unset_template_for_drop_dbs.sql"),
-                                datname = db_name,
-                                outer_tag = outer_tag,
+                                datname_str = escape_literal(&op.name),
+                                datname = &op.name.pg_quote()
                             );
 
                             // Use FORCE to drop database even if there are active connections.
@@ -848,8 +834,6 @@ async fn get_operations<'a>(
                                 comment: None,
                             },
                             Operation {
-                                // ALL PRIVILEGES grants CREATE, CONNECT, and TEMPORARY on the database
-                                // (see https://www.postgresql.org/docs/current/ddl-priv.html)
                                 query: format!(
                                     "GRANT ALL PRIVILEGES ON DATABASE {} TO neon_superuser",
                                     db.name.pg_quote()
@@ -909,11 +893,9 @@ async fn get_operations<'a>(
                 PerDatabasePhase::DropLogicalSubscriptions => {
                     match &db {
                         DB::UserDB(db) => {
-                            let (db_name, outer_tag) = db.name.pg_quote_dollar();
                             let drop_subscription_query: String = format!(
                                 include_str!("sql/drop_subscriptions.sql"),
-                                datname_str = db_name,
-                                outer_tag = outer_tag,
+                                datname_str = escape_literal(&db.name),
                             );
 
                             let operations = vec![Operation {
@@ -952,7 +934,6 @@ async fn get_operations<'a>(
                                     DB::SystemDB => PgIdent::from("cloud_admin").pg_quote(),
                                     DB::UserDB(db) => db.owner.pg_quote(),
                                 };
-                                let (escaped_role, outer_tag) = op.name.pg_quote_dollar();
 
                                 Some(vec![
                                     // This will reassign all dependent objects to the db owner
@@ -967,9 +948,7 @@ async fn get_operations<'a>(
                                     Operation {
                                         query: format!(
                                             include_str!("sql/pre_drop_role_revoke_privileges.sql"),
-                                            // N.B. this has to be properly dollar-escaped with `pg_quote_dollar()`
-                                            role_name = escaped_role,
-                                            outer_tag = outer_tag,
+                                            role_name = quoted,
                                         ),
                                         comment: None,
                                     },
@@ -994,14 +973,12 @@ async fn get_operations<'a>(
                         DB::SystemDB => return Ok(Box::new(empty())),
                         DB::UserDB(db) => db,
                     };
-                    let (db_owner, outer_tag) = db.owner.pg_quote_dollar();
 
                     let operations = vec![
                         Operation {
                             query: format!(
                                 include_str!("sql/set_public_schema_owner.sql"),
-                                db_owner = db_owner,
-                                outer_tag = outer_tag,
+                                db_owner = db.owner.pg_quote()
                             ),
                             comment: None,
                         },
@@ -1121,25 +1098,6 @@ async fn get_operations<'a>(
             }
             Ok(Box::new(empty()))
         }
-        ApplySpecPhase::CreatePgauditExtension => Ok(Box::new(once(Operation {
-            query: String::from("CREATE EXTENSION IF NOT EXISTS pgaudit"),
-            comment: Some(String::from("create pgaudit extensions")),
-        }))),
-        ApplySpecPhase::CreatePgauditlogtofileExtension => Ok(Box::new(once(Operation {
-            query: String::from("CREATE EXTENSION IF NOT EXISTS pgauditlogtofile"),
-            comment: Some(String::from("create pgauditlogtofile extensions")),
-        }))),
-        // Disable pgaudit logging for postgres database.
-        // Postgres is neon system database used by monitors
-        // and compute_ctl tuning functions and thus generates a lot of noise.
-        // We do not consider data stored in this database as sensitive.
-        ApplySpecPhase::DisablePostgresDBPgAudit => {
-            let query = "ALTER DATABASE postgres SET pgaudit.log to 'none'";
-            Ok(Box::new(once(Operation {
-                query: query.to_string(),
-                comment: Some(query.to_string()),
-            })))
-        }
         ApplySpecPhase::HandleNeonExtension => {
             let operations = vec![
                 Operation {

compute_tools/src/sql/create_neon_superuser.sql

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

compute_tools/src/sql/drop_subscriptions.sql

@@ -1,4 +1,4 @@
-DO ${outer_tag}$
+DO $$
 DECLARE
     subname TEXT;
 BEGIN
@@ -9,4 +9,4 @@ BEGIN
         EXECUTE format('DROP SUBSCRIPTION %I;', subname);
     END LOOP;
 END;
-${outer_tag}$;
+$$;

compute_tools/src/sql/pre_drop_role_revoke_privileges.sql

@@ -1,7 +1,8 @@
-DO ${outer_tag}$
+SET SESSION ROLE neon_superuser;
+
+DO $$
 DECLARE
     schema TEXT;
-    grantor TEXT;
     revoke_query TEXT;
 BEGIN
     FOR schema IN
@@ -14,25 +15,14 @@ BEGIN
         -- ii) it's easy to add more schemas to the list if needed.
         WHERE schema_name IN ('public')
     LOOP
-        FOR grantor IN EXECUTE
-            format(
-                'SELECT DISTINCT rtg.grantor FROM information_schema.role_table_grants AS rtg WHERE grantee = %s',
-                -- N.B. this has to be properly dollar-escaped with `pg_quote_dollar()`
-                quote_literal({role_name})
-            )
-        LOOP
-            EXECUTE format('SET LOCAL ROLE %I', grantor);
+        revoke_query := format(
+            'REVOKE ALL PRIVILEGES ON ALL TABLES IN SCHEMA %I FROM {role_name} GRANTED BY neon_superuser;',
+            schema
+        );
 
-            revoke_query := format(
-                'REVOKE ALL PRIVILEGES ON ALL TABLES IN SCHEMA %I FROM %I GRANTED BY %I',
-                schema,
-                -- N.B. this has to be properly dollar-escaped with `pg_quote_dollar()`
-                {role_name},
-                grantor
-            );
-
-            EXECUTE revoke_query;
-        END LOOP;
+        EXECUTE revoke_query;
     END LOOP;
 END;
-${outer_tag}$;
+$$;
+
+RESET ROLE;

compute_tools/src/sql/set_public_schema_owner.sql

@@ -1,4 +1,5 @@
-DO ${outer_tag}$
+DO
+$$
     DECLARE
         schema_owner TEXT;
     BEGIN
@@ -15,8 +16,8 @@ DO ${outer_tag}$
 
             IF schema_owner = 'cloud_admin' OR schema_owner = 'zenith_admin'
             THEN
-                EXECUTE format('ALTER SCHEMA public OWNER TO %I', {db_owner});
+                ALTER SCHEMA public OWNER TO {db_owner};
             END IF;
         END IF;
     END
-${outer_tag}$;
+$$;

compute_tools/src/sql/unset_template_for_drop_dbs.sql

@@ -1,12 +1,12 @@
-DO ${outer_tag}$
+DO $$
     BEGIN
         IF EXISTS(
             SELECT 1
             FROM pg_catalog.pg_database
-            WHERE datname = {datname}
+            WHERE datname = {datname_str}
         )
         THEN
-            EXECUTE format('ALTER DATABASE %I is_template false', {datname});
+            ALTER DATABASE {datname} is_template false;
         END IF;
     END
-${outer_tag}$;
+$$;

compute_tools/tests/pg_helpers_tests.rs

@@ -61,23 +61,6 @@ test.escaping = 'here''s a backslash \\ and a quote '' and a double-quote " hoor
         assert_eq!(ident.pg_quote(), "\"\"\"name\"\";\\n select 1;\"");
     }
 
-    #[test]
-    fn ident_pg_quote_dollar() {
-        let test_cases = vec![
-            ("name", ("$$name$$", "x")),
-            ("name$$", ("$x$name$$$x$", "xx")),
-            ("name$$$", ("$x$name$$$$x$", "xx")),
-            ("name$$$$", ("$x$name$$$$$x$", "xx")),
-            ("name$x$", ("$xx$name$x$$xx$", "xxx")),
-        ];
-
-        for (input, expected) in test_cases {
-            let (escaped, tag) = PgIdent::from(input).pg_quote_dollar();
-            assert_eq!(escaped, expected.0);
-            assert_eq!(tag, expected.1);
-        }
-    }
-
     #[test]
     fn generic_options_search() {
         let generic_options: GenericOptions = Some(vec![

control_plane/src/bin/neon_local.rs

@@ -40,7 +40,6 @@ use pageserver_api::models::{ShardParameters, TimelineCreateRequest, TimelineInf
 use pageserver_api::shard::{ShardCount, ShardStripeSize, TenantShardId};
 use postgres_backend::AuthType;
 use postgres_connection::parse_host_port;
-use safekeeper_api::membership::SafekeeperGeneration;
 use safekeeper_api::{
     DEFAULT_HTTP_LISTEN_PORT as DEFAULT_SAFEKEEPER_HTTP_PORT,
     DEFAULT_PG_LISTEN_PORT as DEFAULT_SAFEKEEPER_PG_PORT,
@@ -597,15 +596,7 @@ struct EndpointStartCmdArgs {
     #[clap(long = "pageserver-id")]
     endpoint_pageserver_id: Option<NodeId>,
 
-    #[clap(
-        long,
-        help = "Safekeepers membership generation to prefix neon.safekeepers with. Normally neon_local sets it on its own, but this option allows to override. Non zero value forces endpoint to use membership configurations."
-    )]
-    safekeepers_generation: Option<u32>,
-    #[clap(
-        long,
-        help = "List of safekeepers endpoint will talk to. Normally neon_local chooses them on its own, but this option allows to override."
-    )]
+    #[clap(long)]
     safekeepers: Option<String>,
 
     #[clap(
@@ -626,9 +617,9 @@ struct EndpointStartCmdArgs {
     )]
     allow_multiple: bool,
 
-    #[clap(short = 't', long, value_parser= humantime::parse_duration, help = "timeout until we fail the command")]
-    #[arg(default_value = "90s")]
-    start_timeout: Duration,
+    #[clap(short = 't', long, help = "timeout until we fail the command")]
+    #[arg(default_value = "10s")]
+    start_timeout: humantime::Duration,
 }
 
 #[derive(clap::Args)]
@@ -963,7 +954,6 @@ fn handle_init(args: &InitCmdArgs) -> anyhow::Result<LocalEnv> {
                         id: pageserver_id,
                         listen_pg_addr: format!("127.0.0.1:{pg_port}"),
                         listen_http_addr: format!("127.0.0.1:{http_port}"),
-                        listen_https_addr: None,
                         pg_auth_type: AuthType::Trust,
                         http_auth_type: AuthType::Trust,
                         other: Default::default(),
@@ -978,7 +968,6 @@ fn handle_init(args: &InitCmdArgs) -> anyhow::Result<LocalEnv> {
             default_tenant_id: TenantId::from_array(std::array::from_fn(|_| 0)),
             storage_controller: None,
             control_plane_compute_hook_api: None,
-            generate_local_ssl_certs: false,
         }
     };
 
@@ -1361,7 +1350,6 @@ async fn handle_endpoint(subcmd: &EndpointCmd, env: &local_env::LocalEnv) -> Res
             let pageserver_id = args.endpoint_pageserver_id;
             let remote_ext_config = &args.remote_ext_config;
 
-            let safekeepers_generation = args.safekeepers_generation.map(SafekeeperGeneration::new);
             // If --safekeepers argument is given, use only the listed
             // safekeeper nodes; otherwise all from the env.
             let safekeepers = if let Some(safekeepers) = parse_safekeepers(&args.safekeepers)? {
@@ -1437,13 +1425,11 @@ async fn handle_endpoint(subcmd: &EndpointCmd, env: &local_env::LocalEnv) -> Res
             endpoint
                 .start(
                     &auth_token,
-                    safekeepers_generation,
                     safekeepers,
                     pageservers,
                     remote_ext_config.as_ref(),
                     stripe_size.0 as usize,
                     args.create_test_user,
-                    args.start_timeout,
                 )
                 .await?;
         }

control_plane/src/endpoint.rs

@@ -42,19 +42,17 @@ use std::path::PathBuf;
 use std::process::Command;
 use std::str::FromStr;
 use std::sync::Arc;
-use std::time::{Duration, Instant, SystemTime, UNIX_EPOCH};
+use std::time::{Duration, SystemTime, UNIX_EPOCH};
 
 use anyhow::{Context, Result, anyhow, bail};
 use compute_api::requests::ConfigurationRequest;
 use compute_api::responses::{ComputeCtlConfig, ComputeStatus, ComputeStatusResponse};
 use compute_api::spec::{
-    Cluster, ComputeAudit, ComputeFeature, ComputeMode, ComputeSpec, Database, PgIdent,
-    RemoteExtSpec, Role,
+    Cluster, ComputeFeature, ComputeMode, ComputeSpec, Database, PgIdent, RemoteExtSpec, Role,
 };
 use nix::sys::signal::{Signal, kill};
 use pageserver_api::shard::ShardStripeSize;
 use reqwest::header::CONTENT_TYPE;
-use safekeeper_api::membership::SafekeeperGeneration;
 use serde::{Deserialize, Serialize};
 use tracing::debug;
 use url::Host;
@@ -578,17 +576,14 @@ impl Endpoint {
         Ok(safekeeper_connstrings)
     }
 
-    #[allow(clippy::too_many_arguments)]
     pub async fn start(
         &self,
         auth_token: &Option<String>,
-        safekeepers_generation: Option<SafekeeperGeneration>,
         safekeepers: Vec<NodeId>,
         pageservers: Vec<(Host, u16)>,
         remote_ext_config: Option<&String>,
         shard_stripe_size: usize,
         create_test_user: bool,
-        start_timeout: Duration,
     ) -> Result<()> {
         if self.status() == EndpointStatus::Running {
             anyhow::bail!("The endpoint is already running");
@@ -660,7 +655,6 @@ impl Endpoint {
             timeline_id: Some(self.timeline_id),
             mode: self.mode,
             pageserver_connstring: Some(pageserver_connstring),
-            safekeepers_generation: safekeepers_generation.map(|g| g.into_inner()),
             safekeeper_connstrings,
             storage_auth_token: auth_token.clone(),
             remote_extensions,
@@ -669,7 +663,6 @@ impl Endpoint {
             local_proxy_config: None,
             reconfigure_concurrency: self.reconfigure_concurrency,
             drop_subscriptions_before_start: self.drop_subscriptions_before_start,
-            audit_log_level: ComputeAudit::Disabled,
         };
 
         // this strange code is needed to support respec() in tests
@@ -777,18 +770,17 @@ impl Endpoint {
         std::fs::write(pidfile_path, pid.to_string())?;
 
         // Wait for it to start
+        let mut attempt = 0;
         const ATTEMPT_INTERVAL: Duration = Duration::from_millis(100);
-        let start_at = Instant::now();
+        const MAX_ATTEMPTS: u32 = 10 * 90; // Wait up to 1.5 min
         loop {
+            attempt += 1;
             match self.get_status().await {
                 Ok(state) => {
                     match state.status {
                         ComputeStatus::Init => {
-                            if Instant::now().duration_since(start_at) > start_timeout {
-                                bail!(
-                                    "compute startup timed out {:?}; still in Init state",
-                                    start_timeout
-                                );
+                            if attempt == MAX_ATTEMPTS {
+                                bail!("compute startup timed out; still in Init state");
                             }
                             // keep retrying
                         }
@@ -815,11 +807,8 @@ impl Endpoint {
                     }
                 }
                 Err(e) => {
-                    if Instant::now().duration_since(start_at) > start_timeout {
-                        return Err(e).context(format!(
-                            "timed out {:?} waiting to connect to compute_ctl HTTP",
-                            start_timeout,
-                        ));
+                    if attempt == MAX_ATTEMPTS {
+                        return Err(e).context("timed out waiting to connect to compute_ctl HTTP");
                     }
                 }
             }

control_plane/src/local_env.rs

@@ -81,10 +81,6 @@ pub struct LocalEnv {
     // but deserialization into a generic toml object as `toml::Value::try_from` fails with an error.
     // https://toml.io/en/v1.0.0 does not contain a concept of "a table inside another table".
     pub branch_name_mappings: HashMap<String, Vec<(TenantId, TimelineId)>>,
-
-    /// Flag to generate SSL certificates for components that need it.
-    /// Also generates root CA certificate that is used to sign all other certificates.
-    pub generate_local_ssl_certs: bool,
 }
 
 /// On-disk state stored in `.neon/config`.
@@ -106,10 +102,6 @@ pub struct OnDiskConfig {
     pub control_plane_api: Option<Url>,
     pub control_plane_compute_hook_api: Option<Url>,
     branch_name_mappings: HashMap<String, Vec<(TenantId, TimelineId)>>,
-    // Note: skip serializing because in compat tests old storage controller fails
-    // to load new config file. May be removed after this field is in release branch.
-    #[serde(skip_serializing_if = "std::ops::Not::not")]
-    pub generate_local_ssl_certs: bool,
 }
 
 fn fail_if_pageservers_field_specified<'de, D>(_: D) -> Result<Vec<PageServerConf>, D::Error>
@@ -137,7 +129,6 @@ pub struct NeonLocalInitConf {
     pub safekeepers: Vec<SafekeeperConf>,
     pub control_plane_api: Option<Url>,
     pub control_plane_compute_hook_api: Option<Option<Url>>,
-    pub generate_local_ssl_certs: bool,
 }
 
 /// Broker config for cluster internal communication.
@@ -174,11 +165,6 @@ pub struct NeonStorageControllerConf {
 
     #[serde(with = "humantime_serde")]
     pub long_reconcile_threshold: Option<Duration>,
-
-    #[serde(default)]
-    pub use_https_pageserver_api: bool,
-
-    pub timelines_onto_safekeepers: bool,
 }
 
 impl NeonStorageControllerConf {
@@ -202,8 +188,6 @@ impl Default for NeonStorageControllerConf {
             max_secondary_lag_bytes: None,
             heartbeat_interval: Self::DEFAULT_HEARTBEAT_INTERVAL,
             long_reconcile_threshold: None,
-            use_https_pageserver_api: false,
-            timelines_onto_safekeepers: false,
         }
     }
 }
@@ -233,7 +217,6 @@ pub struct PageServerConf {
     pub id: NodeId,
     pub listen_pg_addr: String,
     pub listen_http_addr: String,
-    pub listen_https_addr: Option<String>,
     pub pg_auth_type: AuthType,
     pub http_auth_type: AuthType,
     pub no_sync: bool,
@@ -245,7 +228,6 @@ impl Default for PageServerConf {
             id: NodeId(0),
             listen_pg_addr: String::new(),
             listen_http_addr: String::new(),
-            listen_https_addr: None,
             pg_auth_type: AuthType::Trust,
             http_auth_type: AuthType::Trust,
             no_sync: false,
@@ -261,7 +243,6 @@ pub struct NeonLocalInitPageserverConf {
     pub id: NodeId,
     pub listen_pg_addr: String,
     pub listen_http_addr: String,
-    pub listen_https_addr: Option<String>,
     pub pg_auth_type: AuthType,
     pub http_auth_type: AuthType,
     #[serde(default, skip_serializing_if = "std::ops::Not::not")]
@@ -276,7 +257,6 @@ impl From<&NeonLocalInitPageserverConf> for PageServerConf {
             id,
             listen_pg_addr,
             listen_http_addr,
-            listen_https_addr,
             pg_auth_type,
             http_auth_type,
             no_sync,
@@ -286,7 +266,6 @@ impl From<&NeonLocalInitPageserverConf> for PageServerConf {
             id: *id,
             listen_pg_addr: listen_pg_addr.clone(),
             listen_http_addr: listen_http_addr.clone(),
-            listen_https_addr: listen_https_addr.clone(),
             pg_auth_type: *pg_auth_type,
             http_auth_type: *http_auth_type,
             no_sync: *no_sync,
@@ -431,41 +410,6 @@ impl LocalEnv {
         }
     }
 
-    pub fn ssl_ca_cert_path(&self) -> Option<PathBuf> {
-        if self.generate_local_ssl_certs {
-            Some(self.base_data_dir.join("rootCA.crt"))
-        } else {
-            None
-        }
-    }
-
-    pub fn ssl_ca_key_path(&self) -> Option<PathBuf> {
-        if self.generate_local_ssl_certs {
-            Some(self.base_data_dir.join("rootCA.key"))
-        } else {
-            None
-        }
-    }
-
-    pub fn generate_ssl_ca_cert(&self) -> anyhow::Result<()> {
-        let cert_path = self.ssl_ca_cert_path().unwrap();
-        let key_path = self.ssl_ca_key_path().unwrap();
-        if !fs::exists(cert_path.as_path())? {
-            generate_ssl_ca_cert(cert_path.as_path(), key_path.as_path())?;
-        }
-        Ok(())
-    }
-
-    pub fn generate_ssl_cert(&self, cert_path: &Path, key_path: &Path) -> anyhow::Result<()> {
-        self.generate_ssl_ca_cert()?;
-        generate_ssl_cert(
-            cert_path,
-            key_path,
-            self.ssl_ca_cert_path().unwrap().as_path(),
-            self.ssl_ca_key_path().unwrap().as_path(),
-        )
-    }
-
     /// Inspect the base data directory and extract the instance id and instance directory path
     /// for all storage controller instances
     pub async fn storage_controller_instances(&self) -> std::io::Result<Vec<(u8, PathBuf)>> {
@@ -575,7 +519,6 @@ impl LocalEnv {
                 control_plane_api,
                 control_plane_compute_hook_api,
                 branch_name_mappings,
-                generate_local_ssl_certs,
             } = on_disk_config;
             LocalEnv {
                 base_data_dir: repopath.to_owned(),
@@ -590,7 +533,6 @@ impl LocalEnv {
                 control_plane_api: control_plane_api.unwrap(),
                 control_plane_compute_hook_api,
                 branch_name_mappings,
-                generate_local_ssl_certs,
             }
         };
 
@@ -626,7 +568,6 @@ impl LocalEnv {
                 struct PageserverConfigTomlSubset {
                     listen_pg_addr: String,
                     listen_http_addr: String,
-                    listen_https_addr: Option<String>,
                     pg_auth_type: AuthType,
                     http_auth_type: AuthType,
                     #[serde(default)]
@@ -651,7 +592,6 @@ impl LocalEnv {
                 let PageserverConfigTomlSubset {
                     listen_pg_addr,
                     listen_http_addr,
-                    listen_https_addr,
                     pg_auth_type,
                     http_auth_type,
                     no_sync,
@@ -669,7 +609,6 @@ impl LocalEnv {
                     },
                     listen_pg_addr,
                     listen_http_addr,
-                    listen_https_addr,
                     pg_auth_type,
                     http_auth_type,
                     no_sync,
@@ -697,7 +636,6 @@ impl LocalEnv {
                 control_plane_api: Some(self.control_plane_api.clone()),
                 control_plane_compute_hook_api: self.control_plane_compute_hook_api.clone(),
                 branch_name_mappings: self.branch_name_mappings.clone(),
-                generate_local_ssl_certs: self.generate_local_ssl_certs,
             },
         )
     }
@@ -780,7 +718,6 @@ impl LocalEnv {
             safekeepers,
             control_plane_api,
             control_plane_compute_hook_api,
-            generate_local_ssl_certs,
         } = conf;
 
         // Find postgres binaries.
@@ -829,13 +766,8 @@ impl LocalEnv {
             control_plane_api: control_plane_api.unwrap(),
             control_plane_compute_hook_api: control_plane_compute_hook_api.unwrap_or_default(),
             branch_name_mappings: Default::default(),
-            generate_local_ssl_certs,
         };
 
-        if generate_local_ssl_certs {
-            env.generate_ssl_ca_cert()?;
-        }
-
         // create endpoints dir
         fs::create_dir_all(env.endpoints_path())?;
 
@@ -919,80 +851,3 @@ fn generate_auth_keys(private_key_path: &Path, public_key_path: &Path) -> anyhow
     }
     Ok(())
 }
-
-fn generate_ssl_ca_cert(cert_path: &Path, key_path: &Path) -> anyhow::Result<()> {
-    // openssl req -x509 -newkey rsa:2048 -nodes -subj "/CN=Neon Local CA" -days 36500 \
-    // -out rootCA.crt -keyout rootCA.key
-    let keygen_output = Command::new("openssl")
-        .args([
-            "req", "-x509", "-newkey", "rsa:2048", "-nodes", "-days", "36500",
-        ])
-        .args(["-subj", "/CN=Neon Local CA"])
-        .args(["-out", cert_path.to_str().unwrap()])
-        .args(["-keyout", key_path.to_str().unwrap()])
-        .output()
-        .context("failed to generate CA certificate")?;
-    if !keygen_output.status.success() {
-        bail!(
-            "openssl failed: '{}'",
-            String::from_utf8_lossy(&keygen_output.stderr)
-        );
-    }
-    Ok(())
-}
-
-fn generate_ssl_cert(
-    cert_path: &Path,
-    key_path: &Path,
-    ca_cert_path: &Path,
-    ca_key_path: &Path,
-) -> anyhow::Result<()> {
-    // Generate Certificate Signing Request (CSR).
-    let mut csr_path = cert_path.to_path_buf();
-    csr_path.set_extension(".csr");
-
-    // openssl req -new -nodes -newkey rsa:2048 -keyout server.key -out server.csr \
-    // -subj "/CN=localhost" -addext "subjectAltName=DNS:localhost,IP:127.0.0.1"
-    let keygen_output = Command::new("openssl")
-        .args(["req", "-new", "-nodes"])
-        .args(["-newkey", "rsa:2048"])
-        .args(["-subj", "/CN=localhost"])
-        .args(["-addext", "subjectAltName=DNS:localhost,IP:127.0.0.1"])
-        .args(["-keyout", key_path.to_str().unwrap()])
-        .args(["-out", csr_path.to_str().unwrap()])
-        .output()
-        .context("failed to generate CSR")?;
-    if !keygen_output.status.success() {
-        bail!(
-            "openssl failed: '{}'",
-            String::from_utf8_lossy(&keygen_output.stderr)
-        );
-    }
-
-    // Sign CSR with CA key.
-    //
-    // openssl x509 -req -in server.csr -CA rootCA.crt -CAkey rootCA.key -CAcreateserial \
-    // -out server.crt -days 36500 -copy_extensions copyall
-    let keygen_output = Command::new("openssl")
-        .args(["x509", "-req"])
-        .args(["-in", csr_path.to_str().unwrap()])
-        .args(["-CA", ca_cert_path.to_str().unwrap()])
-        .args(["-CAkey", ca_key_path.to_str().unwrap()])
-        .arg("-CAcreateserial")
-        .args(["-out", cert_path.to_str().unwrap()])
-        .args(["-days", "36500"])
-        .args(["-copy_extensions", "copyall"])
-        .output()
-        .context("failed to sign CSR")?;
-    if !keygen_output.status.success() {
-        bail!(
-            "openssl failed: '{}'",
-            String::from_utf8_lossy(&keygen_output.stderr)
-        );
-    }
-
-    // Remove CSR file as it's not needed anymore.
-    fs::remove_file(csr_path)?;
-
-    Ok(())
-}

control_plane/src/pageserver.rs

@@ -21,7 +21,6 @@ use pageserver_api::shard::TenantShardId;
 use pageserver_client::mgmt_api;
 use postgres_backend::AuthType;
 use postgres_connection::{PgConnectionConfig, parse_host_port};
-use reqwest::Certificate;
 use utils::auth::{Claims, Scope};
 use utils::id::{NodeId, TenantId, TimelineId};
 use utils::lsn::Lsn;
@@ -50,29 +49,12 @@ impl PageServerNode {
         let (host, port) =
             parse_host_port(&conf.listen_pg_addr).expect("Unable to parse listen_pg_addr");
         let port = port.unwrap_or(5432);
-
-        let ssl_ca_cert = env.ssl_ca_cert_path().map(|ssl_ca_file| {
-            let buf = std::fs::read(ssl_ca_file).expect("SSL root CA file should exist");
-            Certificate::from_pem(&buf).expect("CA certificate should be valid")
-        });
-
-        let endpoint = if env.storage_controller.use_https_pageserver_api {
-            format!(
-                "https://{}",
-                conf.listen_https_addr.as_ref().expect(
-                    "listen https address should be specified if use_https_pageserver_api is on"
-                )
-            )
-        } else {
-            format!("http://{}", conf.listen_http_addr)
-        };
-
         Self {
             pg_connection_config: PgConnectionConfig::new_host_port(host, port),
             conf: conf.clone(),
             env: env.clone(),
             http_client: mgmt_api::Client::new(
-                endpoint,
+                format!("http://{}", conf.listen_http_addr),
                 {
                     match conf.http_auth_type {
                         AuthType::Trust => None,
@@ -83,9 +65,7 @@ impl PageServerNode {
                     }
                 }
                 .as_deref(),
-                ssl_ca_cert,
-            )
-            .expect("Client constructs with no errors"),
+            ),
         }
     }
 
@@ -240,13 +220,6 @@ impl PageServerNode {
             .context("write identity toml")?;
         drop(identity_toml);
 
-        if self.env.generate_local_ssl_certs {
-            self.env.generate_ssl_cert(
-                datadir.join("server.crt").as_path(),
-                datadir.join("server.key").as_path(),
-            )?;
-        }
-
         // TODO: invoke a TBD config-check command to validate that pageserver will start with the written config
 
         // Write metadata file, used by pageserver on startup to register itself with
@@ -257,15 +230,6 @@ impl PageServerNode {
             parse_host_port(&self.conf.listen_http_addr).expect("Unable to parse listen_http_addr");
         let http_port = http_port.unwrap_or(9898);
 
-        let https_port = match self.conf.listen_https_addr.as_ref() {
-            Some(https_addr) => {
-                let (_https_host, https_port) =
-                    parse_host_port(https_addr).expect("Unable to parse listen_https_addr");
-                Some(https_port.unwrap_or(9899))
-            }
-            None => None,
-        };
-
         // Intentionally hand-craft JSON: this acts as an implicit format compat test
         // in case the pageserver-side structure is edited, and reflects the real life
         // situation: the metadata is written by some other script.
@@ -276,7 +240,6 @@ impl PageServerNode {
                 postgres_port: self.pg_connection_config.port(),
                 http_host: "localhost".to_string(),
                 http_port,
-                https_port,
                 other: HashMap::from([(
                     "availability_zone_id".to_string(),
                     serde_json::json!(az_id),

control_plane/src/storage_controller.rs

@@ -12,10 +12,13 @@ use hyper0::Uri;
 use nix::unistd::Pid;
 use pageserver_api::controller_api::{
     NodeConfigureRequest, NodeDescribeResponse, NodeRegisterRequest, TenantCreateRequest,
-    TenantCreateResponse, TenantLocateResponse,
+    TenantCreateResponse, TenantLocateResponse, TenantShardMigrateRequest,
+    TenantShardMigrateResponse,
 };
-use pageserver_api::models::{TimelineCreateRequest, TimelineInfo};
-use pageserver_api::shard::TenantShardId;
+use pageserver_api::models::{
+    TenantShardSplitRequest, TenantShardSplitResponse, TimelineCreateRequest, TimelineInfo,
+};
+use pageserver_api::shard::{ShardStripeSize, TenantShardId};
 use pageserver_client::mgmt_api::ResponseErrorMessageExt;
 use postgres_backend::AuthType;
 use reqwest::Method;
@@ -534,14 +537,6 @@ impl StorageController {
             args.push("--start-as-candidate".to_string());
         }
 
-        if self.config.use_https_pageserver_api {
-            args.push("--use-https-pageserver-api".to_string());
-        }
-
-        if let Some(ssl_ca_file) = self.env.ssl_ca_cert_path() {
-            args.push(format!("--ssl-ca-file={}", ssl_ca_file.to_str().unwrap()));
-        }
-
         if let Some(private_key) = &self.private_key {
             let claims = Claims::new(None, Scope::PageServerApi);
             let jwt_token =
@@ -584,10 +579,6 @@ impl StorageController {
             self.env.base_data_dir.display()
         ));
 
-        if self.config.timelines_onto_safekeepers {
-            args.push("--timelines-onto-safekeepers".to_string());
-        }
-
         background_process::start_process(
             COMMAND,
             &instance_dir,
@@ -834,6 +825,41 @@ impl StorageController {
         .await
     }
 
+    #[instrument(skip(self))]
+    pub async fn tenant_migrate(
+        &self,
+        tenant_shard_id: TenantShardId,
+        node_id: NodeId,
+    ) -> anyhow::Result<TenantShardMigrateResponse> {
+        self.dispatch(
+            Method::PUT,
+            format!("control/v1/tenant/{tenant_shard_id}/migrate"),
+            Some(TenantShardMigrateRequest {
+                node_id,
+                migration_config: None,
+            }),
+        )
+        .await
+    }
+
+    #[instrument(skip(self), fields(%tenant_id, %new_shard_count))]
+    pub async fn tenant_split(
+        &self,
+        tenant_id: TenantId,
+        new_shard_count: u8,
+        new_stripe_size: Option<ShardStripeSize>,
+    ) -> anyhow::Result<TenantShardSplitResponse> {
+        self.dispatch(
+            Method::PUT,
+            format!("control/v1/tenant/{tenant_id}/shard_split"),
+            Some(TenantShardSplitRequest {
+                new_shard_count,
+                new_stripe_size,
+            }),
+        )
+        .await
+    }
+
     #[instrument(skip_all, fields(node_id=%req.node_id))]
     pub async fn node_register(&self, req: NodeRegisterRequest) -> anyhow::Result<()> {
         self.dispatch::<_, ()>(Method::POST, "control/v1/node".to_string(), Some(req))

control_plane/storcon_cli/src/main.rs

@@ -1,17 +1,16 @@
 use std::collections::{HashMap, HashSet};
-use std::path::PathBuf;
 use std::str::FromStr;
 use std::time::Duration;
 
 use clap::{Parser, Subcommand};
 use futures::StreamExt;
 use pageserver_api::controller_api::{
-    AvailabilityZone, MigrationConfig, NodeAvailabilityWrapper, NodeConfigureRequest,
-    NodeDescribeResponse, NodeRegisterRequest, NodeSchedulingPolicy, NodeShardResponse,
-    PlacementPolicy, SafekeeperDescribeResponse, SafekeeperSchedulingPolicyRequest,
-    ShardSchedulingPolicy, ShardsPreferredAzsRequest, ShardsPreferredAzsResponse,
-    SkSchedulingPolicy, TenantCreateRequest, TenantDescribeResponse, TenantPolicyRequest,
-    TenantShardMigrateRequest, TenantShardMigrateResponse,
+    AvailabilityZone, NodeAvailabilityWrapper, NodeConfigureRequest, NodeDescribeResponse,
+    NodeRegisterRequest, NodeSchedulingPolicy, NodeShardResponse, PlacementPolicy,
+    SafekeeperDescribeResponse, SafekeeperSchedulingPolicyRequest, ShardSchedulingPolicy,
+    ShardsPreferredAzsRequest, ShardsPreferredAzsResponse, SkSchedulingPolicy, TenantCreateRequest,
+    TenantDescribeResponse, TenantPolicyRequest, TenantShardMigrateRequest,
+    TenantShardMigrateResponse,
 };
 use pageserver_api::models::{
     EvictionPolicy, EvictionPolicyLayerAccessThreshold, LocationConfigSecondary, ShardParameters,
@@ -113,15 +112,6 @@ enum Command {
         tenant_shard_id: TenantShardId,
         #[arg(long)]
         node: NodeId,
-        #[arg(long, default_value_t = true, action = clap::ArgAction::Set)]
-        prewarm: bool,
-        #[arg(long, default_value_t = false, action = clap::ArgAction::Set)]
-        override_scheduler: bool,
-    },
-    /// Watch the location of a tenant shard evolve, e.g. while expecting it to migrate
-    TenantShardWatch {
-        #[arg(long)]
-        tenant_shard_id: TenantShardId,
     },
     /// Migrate the secondary location for a tenant shard to a specific pageserver.
     TenantShardMigrateSecondary {
@@ -279,10 +269,6 @@ struct Cli {
     /// a token with both scopes to use with this tool.
     jwt: Option<String>,
 
-    #[arg(long)]
-    /// Trusted root CA certificate to use in https APIs.
-    ssl_ca_file: Option<PathBuf>,
-
     #[command(subcommand)]
     command: Command,
 }
@@ -393,17 +379,9 @@ async fn main() -> anyhow::Result<()> {
 
     let storcon_client = Client::new(cli.api.clone(), cli.jwt.clone());
 
-    let ssl_ca_cert = match &cli.ssl_ca_file {
-        Some(ssl_ca_file) => {
-            let buf = tokio::fs::read(ssl_ca_file).await?;
-            Some(reqwest::Certificate::from_pem(&buf)?)
-        }
-        None => None,
-    };
-
     let mut trimmed = cli.api.to_string();
     trimmed.pop();
-    let vps_client = mgmt_api::Client::new(trimmed, cli.jwt.as_deref(), ssl_ca_cert)?;
+    let vps_client = mgmt_api::Client::new(trimmed, cli.jwt.as_deref());
 
     match cli.command {
         Command::NodeRegister {
@@ -641,43 +619,19 @@ async fn main() -> anyhow::Result<()> {
         Command::TenantShardMigrate {
             tenant_shard_id,
             node,
-            prewarm,
-            override_scheduler,
         } => {
-            let migration_config = MigrationConfig {
-                prewarm,
-                override_scheduler,
-                ..Default::default()
-            };
-
             let req = TenantShardMigrateRequest {
                 node_id: node,
-                origin_node_id: None,
-                migration_config,
+                migration_config: None,
             };
 
-            match storcon_client
+            storcon_client
                 .dispatch::<TenantShardMigrateRequest, TenantShardMigrateResponse>(
                     Method::PUT,
                     format!("control/v1/tenant/{tenant_shard_id}/migrate"),
                     Some(req),
                 )
-                .await
-            {
-                Err(mgmt_api::Error::ApiError(StatusCode::PRECONDITION_FAILED, msg)) => {
-                    anyhow::bail!(
-                        "Migration to {node} rejected, may require `--force` ({}) ",
-                        msg
-                    );
-                }
-                Err(e) => return Err(e.into()),
-                Ok(_) => {}
-            }
-
-            watch_tenant_shard(storcon_client, tenant_shard_id, Some(node)).await?;
-        }
-        Command::TenantShardWatch { tenant_shard_id } => {
-            watch_tenant_shard(storcon_client, tenant_shard_id, None).await?;
+                .await?;
         }
         Command::TenantShardMigrateSecondary {
             tenant_shard_id,
@@ -685,8 +639,7 @@ async fn main() -> anyhow::Result<()> {
         } => {
             let req = TenantShardMigrateRequest {
                 node_id: node,
-                origin_node_id: None,
-                migration_config: MigrationConfig::default(),
+                migration_config: None,
             };
 
             storcon_client
@@ -1152,8 +1105,7 @@ async fn main() -> anyhow::Result<()> {
                                 format!("control/v1/tenant/{}/migrate", mv.tenant_shard_id),
                                 Some(TenantShardMigrateRequest {
                                     node_id: mv.to,
-                                    origin_node_id: Some(mv.from),
-                                    migration_config: MigrationConfig::default(),
+                                    migration_config: None,
                                 }),
                             )
                             .await
@@ -1332,68 +1284,3 @@ async fn main() -> anyhow::Result<()> {
 
     Ok(())
 }
-
-static WATCH_INTERVAL: Duration = Duration::from_secs(5);
-
-async fn watch_tenant_shard(
-    storcon_client: Client,
-    tenant_shard_id: TenantShardId,
-    until_migrated_to: Option<NodeId>,
-) -> anyhow::Result<()> {
-    if let Some(until_migrated_to) = until_migrated_to {
-        println!(
-            "Waiting for tenant shard {} to be migrated to node {}",
-            tenant_shard_id, until_migrated_to
-        );
-    }
-
-    loop {
-        let desc = storcon_client
-            .dispatch::<(), TenantDescribeResponse>(
-                Method::GET,
-                format!("control/v1/tenant/{}", tenant_shard_id.tenant_id),
-                None,
-            )
-            .await?;
-
-        // Output the current state of the tenant shard
-        let shard = desc
-            .shards
-            .iter()
-            .find(|s| s.tenant_shard_id == tenant_shard_id)
-            .ok_or(anyhow::anyhow!("Tenant shard not found"))?;
-        let summary = format!(
-            "attached: {} secondary: {} {}",
-            shard
-                .node_attached
-                .map(|n| format!("{}", n))
-                .unwrap_or("none".to_string()),
-            shard
-                .node_secondary
-                .iter()
-                .map(|n| n.to_string())
-                .collect::<Vec<_>>()
-                .join(","),
-            if shard.is_reconciling {
-                "(reconciler active)"
-            } else {
-                "(reconciler idle)"
-            }
-        );
-        println!("{}", summary);
-
-        // Maybe drop out if we finished migration
-        if let Some(until_migrated_to) = until_migrated_to {
-            if shard.node_attached == Some(until_migrated_to) && !shard.is_reconciling {
-                println!(
-                    "Tenant shard {} is now on node {}",
-                    tenant_shard_id, until_migrated_to
-                );
-                break;
-            }
-        }
-
-        tokio::time::sleep(WATCH_INTERVAL).await;
-    }
-    Ok(())
-}

deny.toml

@@ -27,14 +27,6 @@ yanked = "warn"
 id = "RUSTSEC-2023-0071"
 reason = "the marvin attack only affects private key decryption, not public key signature verification"
 
-[[advisories.ignore]]
-id = "RUSTSEC-2024-0436"
-reason = "The paste crate is a build-only dependency with no runtime components. It is unlikely to have any security impact."
-
-[[advisories.ignore]]
-id = "RUSTSEC-2025-0014"
-reason = "The humantime is widely used and is not easy to replace right now. It is unmaintained, but it has no known vulnerabilities to care about. #11179"
-
 # This section is considered when running `cargo deny check licenses`
 # More documentation for the licenses section can be found here:
 # https://embarkstudios.github.io/cargo-deny/checks/licenses/cfg.html

docker-compose/test_extensions_upgrade.sh

@@ -6,11 +6,8 @@ generate_id() {
     local -n resvar=$1
     printf -v resvar '%08x%08x%08x%08x' $SRANDOM $SRANDOM $SRANDOM $SRANDOM
 }
-echo "${OLD_COMPUTE_TAG}"
-echo "${NEW_COMPUTE_TAG}"
-echo "${TEST_EXTENSIONS_TAG}"
-if [ -z "${OLD_COMPUTE_TAG:-}" ] || [ -z "${NEW_COMPUTE_TAG:-}" ] || [ -z "${TEST_EXTENSIONS_TAG:-}" ]; then
-  echo OLD_COMPUTE_TAG, NEW_COMPUTE_TAG and TEST_EXTENSIONS_TAG must be set
+if [ -z ${OLD_COMPUTE_TAG+x} ] || [ -z ${NEW_COMPUTE_TAG+x} ] || [ -z "${OLD_COMPUTE_TAG}" ] || [ -z "${NEW_COMPUTE_TAG}" ]; then
+  echo OLD_COMPUTE_TAG and NEW_COMPUTE_TAG must be defined
   exit 1
 fi
 export PG_VERSION=${PG_VERSION:-16}
@@ -85,7 +82,7 @@ EXTENSIONS='[
 {"extname": "pg_repack", "extdir": "pg_repack-src"}
 ]'
 EXTNAMES=$(echo ${EXTENSIONS} | jq -r '.[].extname' | paste -sd ' ' -)
-COMPUTE_TAG=${NEW_COMPUTE_TAG} docker compose --profile test-extensions up --quiet-pull --build -d
+COMPUTE_TAG=${NEW_COMPUTE_TAG} TEST_EXTENSIONS_TAG=${NEW_COMPUTE_TAG} docker compose --profile test-extensions up --quiet-pull --build -d
 wait_for_ready
 docker compose exec neon-test-extensions psql -c "DROP DATABASE IF EXISTS contrib_regression"
 docker compose exec neon-test-extensions psql -c "CREATE DATABASE contrib_regression"
@@ -93,7 +90,7 @@ create_extensions "${EXTNAMES}"
 query="select json_object_agg(extname,extversion) from pg_extension where extname in ('${EXTNAMES// /\',\'}')"
 new_vers=$(docker compose exec neon-test-extensions psql -Aqt -d contrib_regression -c "$query")
 docker compose --profile test-extensions down
-COMPUTE_TAG=${OLD_COMPUTE_TAG} docker compose --profile test-extensions up --quiet-pull --build -d --force-recreate
+COMPUTE_TAG=${OLD_COMPUTE_TAG} TEST_EXTENSIONS_TAG=${NEW_COMPUTE_TAG} docker compose --profile test-extensions up --quiet-pull --build -d --force-recreate
 wait_for_ready
 docker compose exec neon-test-extensions psql -c "DROP DATABASE IF EXISTS contrib_regression"
 docker compose exec neon-test-extensions psql -c "CREATE DATABASE contrib_regression"

docs/rfcs/041-rel-sparse-keyspace.md

@@ -1,201 +0,0 @@
-# Sparse Keyspace for Relation Directories
-
-## Summary
-
-This is an RFC describing a new storage strategy for storing relation directories.
-
-## Motivation
-
-Postgres maintains a directory structure for databases and relations. In Neon, we store these information
-by serializing the directory data in a single key (see `pgdatadir_mapping.rs`).
-
-```rust
-// DbDir:
-// 00 00000000 00000000 00000000 00   00000000
-
-// RelDir:
-// 00 SPCNODE  DBNODE   00000000 00   00000001 (Postgres never uses relfilenode 0)
-```
-
-We have a dedicated structure on the ingestion path to serialize the relation directory into this single key.
-
-```rust
-#[derive(Debug, Serialize, Deserialize, Default)]
-pub(crate) struct RelDirectory {
-    // Set of relations that exist. (relfilenode, forknum)
-    //
-    // TODO: Store it as a btree or radix tree or something else that spans multiple
-    // key-value pairs, if you have a lot of relations
-    pub(crate) rels: HashSet<(Oid, u8)>,
-}
-```
-
-The current codebase has the following three access patterns for the relation directory.
-
-1. Check if a relation exists.
-2. List all relations.
-3. Create/drop a relation.
-
-For (1), we currently have to get the reldir key, deserialize it, and check whether the relation exists in the
-hash set. For (2), we get the reldir key and the hash set. For (3), we need first to get
-and deserialize the key, add the new relation record to the hash set, and then serialize it and write it back.
-
-If we have 100k relations in a database, we would have a 100k-large hash set. Then, every
-relation created and dropped would have deserialized and serialized this 100k-large hash set. This makes the
-relation create/drop process to be quadratic. When we check if a relation exists in the ingestion path,
-we would have to deserialize this super big 100k-large key before checking if a single relation exists.
-
-In this RFC, we will propose a new way to store the reldir data in the sparse keyspace and propose how
-to seamlessly migrate users to use the new keyspace.
-
-The PoC patch is implemented in [PR10316](https://github.com/neondatabase/neon/pull/10316).
-
-## Key Mapping
-
-We will use the recently introduced sparse keyspace to store actual data. Sparse keyspace was proposed in
-[038-aux-file-v2.md](038-aux-file-v2.md). The original reldir has one single value of `HashSet<(Oid, u8)>`
-for each of the databases (identified as `spcnode, dbnode`). We encode the `Oid` (`relnode, forknum`),
-into the key.
-
-```plain
-(REL_DIR_KEY_PREFIX, spcnode, dbnode, relnode, forknum, 1) -> deleted
-(REL_DIR_KEY_PREFIX, spcnode, dbnode, relnode, forknum, 1) -> exists
-```
-
-Assume all reldir data are stored in this new keyspace; the 3 reldir operations we mentioned before can be
-implemented as follows.
-
-1. Check if a relation exists: check if the key maps to "exists".
-2. List all relations: scan the sprase keyspace over the `rel_dir_key_prefix`. Extract relnode and forknum from the key.
-3. Create/drop a relation: write "exists" or "deleted" to the corresponding key of the relation. The delete tombstone will
-   be removed during image layer generation upon compaction.
-
-Note that "exists" and "deleted" will be encoded as a single byte as two variants of an enum.
-The mapping is implemented as `rel_tag_sparse_key` in the PoC patch.
-
-## Changes to Sparse Keyspace
-
-Previously, we only used sparse keyspaces for the aux files, which did not carry over when branching. The reldir
-information needs to be preserved from the parent branch to the child branch. Therefore, the read path needs
-to be updated accordingly to accommodate such "inherited sparse keys". This is done in
-[PR#10313](https://github.com/neondatabase/neon/pull/10313).
-
-## Coexistence of the Old and New Keyspaces
-
-Migrating to the new keyspace will be done gradually: when we flip a config item to enable the new reldir keyspace, the
-ingestion path will start to write to the new keyspace and the old reldir data will be kept in the old one. The read
-path needs to combine the data from both keyspaces.
-
-Theoretically, we could do a rewrite at the startup time that scans all relation directories and copies that data into the
-new keyspace. However, this could take a long time, especially if we have thousands of tenants doing the migration
-process simultaneously after the pageserver restarts. Therefore, we propose the coexistence strategy so that the
-migration can happen seamlessly and imposes no potential downtime for the user.
-
-With the coexistence assumption, the 3 reldir operations will be implemented as follows:
-
-1. Check if a relation exists
-   - Check the new keyspace if the key maps to any value. If it maps to "exists" or "deleted", directly
-    return it to the user.
-   - Otherwise, deserialize the old reldir key and get the result.
-2. List all relations: scan the sparse keyspace over the `rel_dir_key_prefix` and deserialize the old reldir key.
-   Combine them to obtain the final result.
-3. Create/drop a relation: write "exists" or "deleted" to the corresponding key of the relation into the new keyspace.
-   - We assume no overwrite of relations will happen (i.e., the user won't create a relation at the same Oid). This will be implemented as a runtime check.
-   - For relation creation, we add `sparse_reldir_tableX -> exists` to the keyspace.
-   - For relation drop, we first check if the relation is recorded in the old keyspace. If yes, we deserialize the old reldir key,
-    remove the relation, and then write it back. Otherwise, we put `sparse_reldir_tableX -> deleted` to the keyspace.
-   - The delete tombstone will be removed during image layer generation upon compaction.
-
-This process ensures that the transition will not introduce any downtime and all new updates are written to the new keyspace. The total
-amount of data in the storage would be `O(relations_modifications)` and we can guarantee `O(current_relations)` after compaction.
-There could be some relations that exist in the old reldir key for a long time. Refer to the "Full Migration" section on how to deal
-with them. Plus, for relation modifications, it will have `O(old_relations)` complexity until we do the full migration, which gives
-us `O(1)` complexity after fully opt-in the sparse keyspace.
-
-The process also implies that a relation will only exists either in the old reldir key or in the new sparse keyspace. It is not possible
-to have a table to be recorded in the old reldir key while later having a delete tombstone for it in the sparse keyspace at any LSN.
-
-We will introduce a config item and an index_part record to record the current status of the migration process.
-
-- Config item `enable_reldir_v2`: controls whether the ingestion path writes the reldir info into the new keyspace.
-- `index_part.json` field `reldir_v2_status`: whether the timeline has written any key into the new reldir keyspace.
-
-If `enable_reldir_v2` is set to `true` and the timeline ingests the first key into the new reldir keyspace, it will update
-`index_part.json` to set `reldir_v2_status` to `Status::Migrating`. Even if `enable_reldir_v2` gets flipped back to
-`false` (i.e., when the pageserver restarts and such config isn't persisted), the read/write path will still
-read/write to the new keyspace to avoid data inconsistency. This also indicates that the migration is one-way only:
-once v2 is enabled, the user cannot go back to v1.
-
-## Next Steps
-
-### Full Migration
-
-This won't be implemented in the project's first phase but might be implemented in the future. Having both v1 and
-v2 existing in the system would force us to keep the code to deserialize the old reldir key forever. To entirely deprecate this
-code path, we must ensure the timeline has no old reldir data.
-
-We can trigger a special image layer generation process at the gc-horizon. The generated image layers will cover several keyspaces:
-the old reldir key in each of the databases, and the new reldir sparse keyspace. It will remove the old reldir key while
-copying them into the corresponding keys in the sparse keyspace in the resulting image. This special process happens in
-the background during compaction. For example, assume this special process is triggered at LSN 0/180. The `create_image_layers`
-process discovers the following keys at this LSN.
-
-```plain
-db1/reldir_key -> (table 1, table 2, table 3)
-...db1 rel keys
-db2/reldir_key -> (table 4, table 5, table 6)
-...db2 rel keys
-sparse_reldir_db2_table7 -> exists
-sparse_reldir_db1_table8 -> deleted
-```
-
-It will generate the following keys:
-
-```plain
-db1/reldir_key -> () # we have to keep the key because it is part of `collect_keyspace`.
-...db1 rel keys
-db2/reldir_key -> ()
-...db2 rel keys
-
--- start image layer for the sparse keyspace at sparse_reldir_prefix at LSN 0/180
-sparse_reldir_db1_table1 -> exists
-sparse_reldir_db1_table2 -> exists
-sparse_reldir_db1_table3 -> exists
-sparse_reldir_db2_table4 -> exists
-sparse_reldir_db2_table5 -> exists
-sparse_reldir_db2_table6 -> exists
-sparse_reldir_db2_table7 -> exists
--- end image layer for the sparse keyspace at sparse_reldir_prefix+1
-
-# The `sparse_reldir_db1_table8` key gets dropped as part of the image layer generation code for the sparse keyspace.
-# Note that the read path will stop reading if a key is not found in the image layer covering the key range so there
-# are no correctness issue.
-```
-
-We must verify that no pending modifications to the old reldir exists in the delta/image layers above the gc-horizon before
-we start this process (We can do a vectored read to get the full key history of the old reldir key and ensure there are no more images
-above the gc-horizon). Otherwise, it will violate the property that "a relation will only exists either in the old reldir key or
-in the new sparse keyspace". After we run this migration process, we can mark `reldir_v2_status` in the `index_part.json` to
-`Status::Migrated`, and the read path won't need to read from the old reldir anymore. Once the status is set to `Migrated`, we
-don't need to add the key into `collect_keyspace` and therefore all of them will be removed from all future image layers.
-
-The migration process can be proactively triggered across all attached/detached tenants to help us fully remove the old reldir code.
-
-### Consolidate Relation Size Keys
-
-We have relsize at the end of all relation nodes.
-
-```plain
-// RelSize:
-// 00 SPCNODE  DBNODE   RELNODE  FORK FFFFFFFF
-```
-
-This means that computing logical size requires us to do several single-key gets across the keyspace,
-potentially requiring downloading many layer files. We could consolidate them into a single
-keyspace, improving logical size calculation performance.
-
-### Migrate DBDir Keys
-
-We assume the number of databases created by the users will be small, and therefore, the current way
-of storing the database directory would be acceptable. In the future, we could also migrate DBDir keys into
-the sparse keyspace to support large amount of databases.

libs/compute_api/src/responses.rs

@@ -134,10 +134,8 @@ pub struct CatalogObjects {
     pub databases: Vec<Database>,
 }
 
-#[derive(Clone, Debug, Deserialize, Serialize)]
+#[derive(Debug, Deserialize, Serialize)]
 pub struct ComputeCtlConfig {
-    /// Set of JSON web keys that the compute can use to authenticate
-    /// communication from the control plane.
     pub jwks: JwkSet,
 }
 

libs/compute_api/src/spec.rs

@@ -101,17 +101,6 @@ pub struct ComputeSpec {
     pub timeline_id: Option<TimelineId>,
     pub pageserver_connstring: Option<String>,
 
-    /// Safekeeper membership config generation. It is put in
-    /// neon.safekeepers GUC and serves two purposes:
-    /// 1) Non zero value forces walproposer to use membership configurations.
-    /// 2) If walproposer wants to update list of safekeepers to connect to
-    ///    taking them from some safekeeper mconf, it should check what value
-    ///    is newer by comparing the generation.
-    ///
-    /// Note: it could be SafekeeperGeneration, but this needs linking
-    /// compute_ctl with postgres_ffi.
-    #[serde(default)]
-    pub safekeepers_generation: Option<u32>,
     #[serde(default)]
     pub safekeeper_connstrings: Vec<String>,
 
@@ -155,16 +144,6 @@ pub struct ComputeSpec {
     /// over the same replication content from publisher.
     #[serde(default)] // Default false
     pub drop_subscriptions_before_start: bool,
-
-    /// Log level for audit logging:
-    ///
-    /// Disabled - no audit logging. This is the default.
-    /// log - log masked statements to the postgres log using pgaudit extension
-    /// hipaa - log unmasked statements to the file using pgaudit and pgauditlogtofile extension
-    ///
-    /// Extensions should be present in shared_preload_libraries
-    #[serde(default)]
-    pub audit_log_level: ComputeAudit,
 }
 
 /// Feature flag to signal `compute_ctl` to enable certain experimental functionality.
@@ -272,29 +251,6 @@ pub enum ComputeMode {
     Replica,
 }
 
-impl ComputeMode {
-    /// Convert the compute mode to a string that can be used to identify the type of compute,
-    /// which means that if it's a static compute, the LSN will not be included.
-    pub fn to_type_str(&self) -> &'static str {
-        match self {
-            ComputeMode::Primary => "primary",
-            ComputeMode::Static(_) => "static",
-            ComputeMode::Replica => "replica",
-        }
-    }
-}
-
-/// Log level for audit logging
-/// Disabled, log, hipaa
-/// Default is Disabled
-#[derive(Clone, Debug, Default, Eq, PartialEq, Deserialize, Serialize)]
-pub enum ComputeAudit {
-    #[default]
-    Disabled,
-    Log,
-    Hipaa,
-}
-
 #[derive(Clone, Debug, Default, Deserialize, Serialize, PartialEq, Eq)]
 pub struct Cluster {
     pub cluster_id: Option<String>,

libs/http-utils/Cargo.toml

@@ -6,9 +6,11 @@ license.workspace = true
 
 [dependencies]
 anyhow.workspace = true
+backtrace.workspace = true
 bytes.workspace = true
+inferno.workspace = true
 fail.workspace = true
-futures.workspace = true
+flate2.workspace = true
 hyper0.workspace = true
 itertools.workspace = true
 jemalloc_pprof.workspace = true
@@ -22,7 +24,6 @@ serde_path_to_error.workspace = true
 thiserror.workspace = true
 tracing.workspace = true
 tokio.workspace = true
-tokio-rustls.workspace = true
 tokio-util.workspace = true
 url.workspace = true
 uuid.workspace = true

libs/http-utils/src/endpoint.rs

@@ -3,6 +3,8 @@ use std::io::Write as _;
 use std::str::FromStr;
 use std::time::Duration;
 
+use ::pprof::ProfilerGuardBuilder;
+use ::pprof::protos::Message as _;
 use anyhow::{Context, anyhow};
 use bytes::{Bytes, BytesMut};
 use hyper::header::{AUTHORIZATION, CONTENT_DISPOSITION, CONTENT_TYPE, HeaderName};
@@ -10,8 +12,7 @@ use hyper::http::HeaderValue;
 use hyper::{Body, Method, Request, Response};
 use metrics::{Encoder, IntCounter, TextEncoder, register_int_counter};
 use once_cell::sync::Lazy;
-use pprof::ProfilerGuardBuilder;
-use pprof::protos::Message as _;
+use regex::Regex;
 use routerify::ext::RequestExt;
 use routerify::{Middleware, RequestInfo, Router, RouterBuilder};
 use tokio::sync::{Mutex, Notify, mpsc};
@@ -21,6 +22,7 @@ use tracing::{Instrument, debug, info, info_span, warn};
 use utils::auth::{AuthError, Claims, SwappableJwtAuth};
 
 use crate::error::{ApiError, api_error_handler, route_error_handler};
+use crate::pprof;
 use crate::request::{get_query_param, parse_query_param};
 
 static SERVE_METRICS_COUNT: Lazy<IntCounter> = Lazy::new(|| {
@@ -399,10 +401,12 @@ pub async fn profile_cpu_handler(req: Request<Body>) -> Result<Response<Body>, A
     // Return the report in the requested format.
     match format {
         Format::Pprof => {
-            let body = report
+            let mut body = Vec::new();
+            report
                 .pprof()
                 .map_err(|err| ApiError::InternalServerError(err.into()))?
-                .encode_to_vec();
+                .write_to_vec(&mut body)
+                .map_err(|err| ApiError::InternalServerError(err.into()))?;
 
             Response::builder()
                 .status(200)
@@ -445,6 +449,20 @@ pub async fn profile_heap_handler(req: Request<Body>) -> Result<Response<Body>,
         Some(format) => return Err(ApiError::BadRequest(anyhow!("invalid format {format}"))),
     };
 
+    // Functions and mappings to strip when symbolizing pprof profiles. If true,
+    // also remove child frames.
+    static STRIP_FUNCTIONS: Lazy<Vec<(Regex, bool)>> = Lazy::new(|| {
+        vec![
+            (Regex::new("^__rust").unwrap(), false),
+            (Regex::new("^_start$").unwrap(), false),
+            (Regex::new("^irallocx_prof").unwrap(), true),
+            (Regex::new("^prof_alloc_prep").unwrap(), true),
+            (Regex::new("^std::rt::lang_start").unwrap(), false),
+            (Regex::new("^std::sys::backtrace::__rust").unwrap(), false),
+        ]
+    });
+    const STRIP_MAPPINGS: &[&str] = &["libc", "libgcc", "pthread", "vdso"];
+
     // Obtain profiler handle.
     let mut prof_ctl = jemalloc_pprof::PROF_CTL
         .as_ref()
@@ -477,27 +495,45 @@ pub async fn profile_heap_handler(req: Request<Body>) -> Result<Response<Body>,
         }
 
         Format::Pprof => {
-            let data = tokio::task::spawn_blocking(move || prof_ctl.dump_pprof())
-                .await
-                .map_err(|join_err| ApiError::InternalServerError(join_err.into()))?
-                .map_err(ApiError::InternalServerError)?;
+            let data = tokio::task::spawn_blocking(move || {
+                let bytes = prof_ctl.dump_pprof()?;
+                // Symbolize the profile.
+                // TODO: consider moving this upstream to jemalloc_pprof and avoiding the
+                // serialization roundtrip.
+                let profile = pprof::decode(&bytes)?;
+                let profile = pprof::symbolize(profile)?;
+                let profile = pprof::strip_locations(profile, STRIP_MAPPINGS, &STRIP_FUNCTIONS);
+                pprof::encode(&profile)
+            })
+            .await
+            .map_err(|join_err| ApiError::InternalServerError(join_err.into()))?
+            .map_err(ApiError::InternalServerError)?;
             Response::builder()
                 .status(200)
                 .header(CONTENT_TYPE, "application/octet-stream")
-                .header(CONTENT_DISPOSITION, "attachment; filename=\"heap.pb.gz\"")
+                .header(CONTENT_DISPOSITION, "attachment; filename=\"heap.pb\"")
                 .body(Body::from(data))
                 .map_err(|err| ApiError::InternalServerError(err.into()))
         }
 
         Format::Svg => {
-            let svg = tokio::task::spawn_blocking(move || prof_ctl.dump_flamegraph())
-                .await
-                .map_err(|join_err| ApiError::InternalServerError(join_err.into()))?
-                .map_err(ApiError::InternalServerError)?;
+            let body = tokio::task::spawn_blocking(move || {
+                let bytes = prof_ctl.dump_pprof()?;
+                let profile = pprof::decode(&bytes)?;
+                let profile = pprof::symbolize(profile)?;
+                let profile = pprof::strip_locations(profile, STRIP_MAPPINGS, &STRIP_FUNCTIONS);
+                let mut opts = inferno::flamegraph::Options::default();
+                opts.title = "Heap inuse".to_string();
+                opts.count_name = "bytes".to_string();
+                pprof::flamegraph(profile, &mut opts)
+            })
+            .await
+            .map_err(|join_err| ApiError::InternalServerError(join_err.into()))?
+            .map_err(ApiError::InternalServerError)?;
             Response::builder()
                 .status(200)
                 .header(CONTENT_TYPE, "image/svg+xml")
-                .body(Body::from(svg))
+                .body(Body::from(body))
                 .map_err(|err| ApiError::InternalServerError(err.into()))
         }
     }

libs/http-utils/src/lib.rs

@@ -2,11 +2,11 @@ pub mod endpoint;
 pub mod error;
 pub mod failpoints;
 pub mod json;
+pub mod pprof;
 pub mod request;
-pub mod server;
 
 extern crate hyper0 as hyper;
 
 /// Current fast way to apply simple http routing in various Neon binaries.
 /// Re-exported for sake of uniform approach, that could be later replaced with better alternatives, if needed.
-pub use routerify::{RequestServiceBuilder, RouterBuilder, RouterService, ext::RequestExt};
+pub use routerify::{RouterBuilder, RouterService, ext::RequestExt};

libs/http-utils/src/pprof.rs

@@ -0,0 +1,238 @@
+use std::borrow::Cow;
+use std::collections::{HashMap, HashSet};
+use std::ffi::c_void;
+use std::io::Write as _;
+
+use anyhow::bail;
+use flate2::Compression;
+use flate2::write::{GzDecoder, GzEncoder};
+use itertools::Itertools as _;
+use pprof::protos::{Function, Line, Location, Message as _, Profile};
+use regex::Regex;
+
+/// Decodes a gzip-compressed Protobuf-encoded pprof profile.
+pub fn decode(bytes: &[u8]) -> anyhow::Result<Profile> {
+    let mut gz = GzDecoder::new(Vec::new());
+    gz.write_all(bytes)?;
+    Ok(Profile::parse_from_bytes(&gz.finish()?)?)
+}
+
+/// Encodes a pprof profile as gzip-compressed Protobuf.
+pub fn encode(profile: &Profile) -> anyhow::Result<Vec<u8>> {
+    let mut gz = GzEncoder::new(Vec::new(), Compression::default());
+    profile.write_to_writer(&mut gz)?;
+    Ok(gz.finish()?)
+}
+
+/// Symbolizes a pprof profile using the current binary.
+pub fn symbolize(mut profile: Profile) -> anyhow::Result<Profile> {
+    if !profile.function.is_empty() {
+        return Ok(profile); // already symbolized
+    }
+
+    // Collect function names.
+    let mut functions: HashMap<String, Function> = HashMap::new();
+    let mut strings: HashMap<String, i64> = profile
+        .string_table
+        .into_iter()
+        .enumerate()
+        .map(|(i, s)| (s, i as i64))
+        .collect();
+
+    // Helper to look up or register a string.
+    let mut string_id = |s: &str| -> i64 {
+        // Don't use .entry() to avoid unnecessary allocations.
+        if let Some(id) = strings.get(s) {
+            return *id;
+        }
+        let id = strings.len() as i64;
+        strings.insert(s.to_string(), id);
+        id
+    };
+
+    for loc in &mut profile.location {
+        if !loc.line.is_empty() {
+            continue;
+        }
+
+        // Resolve the line and function for each location.
+        backtrace::resolve(loc.address as *mut c_void, |symbol| {
+            let Some(symbol_name) = symbol.name() else {
+                return;
+            };
+
+            let function_name = format!("{symbol_name:#}");
+            let functions_len = functions.len();
+            let function_id = functions
+                .entry(function_name)
+                .or_insert_with_key(|function_name| {
+                    let function_id = functions_len as u64 + 1;
+                    let system_name = String::from_utf8_lossy(symbol_name.as_bytes());
+                    let filename = symbol
+                        .filename()
+                        .map(|path| path.to_string_lossy())
+                        .unwrap_or(Cow::Borrowed(""));
+                    Function {
+                        id: function_id,
+                        name: string_id(function_name),
+                        system_name: string_id(&system_name),
+                        filename: string_id(&filename),
+                        ..Default::default()
+                    }
+                })
+                .id;
+            loc.line.push(Line {
+                function_id,
+                line: symbol.lineno().unwrap_or(0) as i64,
+                ..Default::default()
+            });
+        });
+    }
+
+    // Store the resolved functions, and mark the mapping as resolved.
+    profile.function = functions.into_values().sorted_by_key(|f| f.id).collect();
+    profile.string_table = strings
+        .into_iter()
+        .sorted_by_key(|(_, i)| *i)
+        .map(|(s, _)| s)
+        .collect();
+
+    for mapping in &mut profile.mapping {
+        mapping.has_functions = true;
+        mapping.has_filenames = true;
+    }
+
+    Ok(profile)
+}
+
+/// Strips locations (stack frames) matching the given mappings (substring) or function names
+/// (regex). The function bool specifies whether child frames should be stripped as well.
+///
+/// The string definitions are left behind in the profile for simplicity, to avoid rewriting all
+/// string references.
+pub fn strip_locations(
+    mut profile: Profile,
+    mappings: &[&str],
+    functions: &[(Regex, bool)],
+) -> Profile {
+    // Strip mappings.
+    let mut strip_mappings: HashSet<u64> = HashSet::new();
+
+    profile.mapping.retain(|mapping| {
+        let Some(name) = profile.string_table.get(mapping.filename as usize) else {
+            return true;
+        };
+        if mappings.iter().any(|substr| name.contains(substr)) {
+            strip_mappings.insert(mapping.id);
+            return false;
+        }
+        true
+    });
+
+    // Strip functions.
+    let mut strip_functions: HashMap<u64, bool> = HashMap::new();
+
+    profile.function.retain(|function| {
+        let Some(name) = profile.string_table.get(function.name as usize) else {
+            return true;
+        };
+        for (regex, strip_children) in functions {
+            if regex.is_match(name) {
+                strip_functions.insert(function.id, *strip_children);
+                return false;
+            }
+        }
+        true
+    });
+
+    // Strip locations. The bool specifies whether child frames should be stripped too.
+    let mut strip_locations: HashMap<u64, bool> = HashMap::new();
+
+    profile.location.retain(|location| {
+        for line in &location.line {
+            if let Some(strip_children) = strip_functions.get(&line.function_id) {
+                strip_locations.insert(location.id, *strip_children);
+                return false;
+            }
+        }
+        if strip_mappings.contains(&location.mapping_id) {
+            strip_locations.insert(location.id, false);
+            return false;
+        }
+        true
+    });
+
+    // Strip sample locations.
+    for sample in &mut profile.sample {
+        // First, find the uppermost function with child removal and truncate the stack.
+        if let Some(truncate) = sample
+            .location_id
+            .iter()
+            .rposition(|id| strip_locations.get(id) == Some(&true))
+        {
+            sample.location_id.drain(..=truncate);
+        }
+        // Next, strip any individual frames without child removal.
+        sample
+            .location_id
+            .retain(|id| !strip_locations.contains_key(id));
+    }
+
+    profile
+}
+
+/// Generates an SVG flamegraph from a symbolized pprof profile.
+pub fn flamegraph(
+    profile: Profile,
+    opts: &mut inferno::flamegraph::Options,
+) -> anyhow::Result<Vec<u8>> {
+    if profile.mapping.iter().any(|m| !m.has_functions) {
+        bail!("profile not symbolized");
+    }
+
+    // Index locations, functions, and strings.
+    let locations: HashMap<u64, Location> =
+        profile.location.into_iter().map(|l| (l.id, l)).collect();
+    let functions: HashMap<u64, Function> =
+        profile.function.into_iter().map(|f| (f.id, f)).collect();
+    let strings = profile.string_table;
+
+    // Resolve stacks as function names, and sum sample values per stack. Also reverse the stack,
+    // since inferno expects it bottom-up.
+    let mut stacks: HashMap<Vec<&str>, i64> = HashMap::new();
+    for sample in profile.sample {
+        let mut stack = Vec::with_capacity(sample.location_id.len());
+        for location in sample.location_id.into_iter().rev() {
+            let Some(location) = locations.get(&location) else {
+                bail!("missing location {location}");
+            };
+            for line in location.line.iter().rev() {
+                let Some(function) = functions.get(&line.function_id) else {
+                    bail!("missing function {}", line.function_id);
+                };
+                let Some(name) = strings.get(function.name as usize) else {
+                    bail!("missing string {}", function.name);
+                };
+                stack.push(name.as_str());
+            }
+        }
+        let Some(&value) = sample.value.first() else {
+            bail!("missing value");
+        };
+        *stacks.entry(stack).or_default() += value;
+    }
+
+    // Construct stack lines for inferno.
+    let lines = stacks
+        .into_iter()
+        .map(|(stack, value)| (stack.into_iter().join(";"), value))
+        .map(|(stack, value)| format!("{stack} {value}"))
+        .sorted()
+        .collect_vec();
+
+    // Construct the flamegraph.
+    let mut bytes = Vec::new();
+    let lines = lines.iter().map(|line| line.as_str());
+    inferno::flamegraph::from_lines(opts, lines, &mut bytes)?;
+    Ok(bytes)
+}

libs/http-utils/src/server.rs

@@ -1,155 +0,0 @@
-use std::{error::Error, sync::Arc};
-
-use futures::StreamExt;
-use futures::stream::FuturesUnordered;
-use hyper0::Body;
-use hyper0::server::conn::Http;
-use routerify::{RequestService, RequestServiceBuilder};
-use tokio::io::{AsyncRead, AsyncWrite};
-use tokio_rustls::TlsAcceptor;
-use tokio_util::sync::CancellationToken;
-use tracing::{error, info};
-
-use crate::error::ApiError;
-
-/// A simple HTTP server over hyper library.
-/// You may want to use it instead of [`hyper0::server::Server`] because:
-/// 1. hyper0's Server was removed from hyper v1.
-///    It's recommended to replace hyepr0's Server with a manual loop, which is done here.
-/// 2. hyper0's Server doesn't support TLS out of the box, and there is no way
-///    to support it efficiently with the Accept trait that hyper0's Server uses.
-///    That's one of the reasons why it was removed from v1.
-///    <https://github.com/hyperium/hyper/blob/115339d3df50f20c8717680aa35f48858e9a6205/docs/ROADMAP.md#higher-level-client-and-server-problems>
-pub struct Server {
-    request_service: Arc<RequestServiceBuilder<Body, ApiError>>,
-    listener: tokio::net::TcpListener,
-    tls_acceptor: Option<TlsAcceptor>,
-}
-
-impl Server {
-    pub fn new(
-        request_service: Arc<RequestServiceBuilder<Body, ApiError>>,
-        listener: std::net::TcpListener,
-        tls_acceptor: Option<TlsAcceptor>,
-    ) -> anyhow::Result<Self> {
-        // Note: caller of from_std is responsible for setting nonblocking mode.
-        listener.set_nonblocking(true)?;
-        let listener = tokio::net::TcpListener::from_std(listener)?;
-
-        Ok(Self {
-            request_service,
-            listener,
-            tls_acceptor,
-        })
-    }
-
-    pub async fn serve(self, cancel: CancellationToken) -> anyhow::Result<()> {
-        fn suppress_io_error(err: &std::io::Error) -> bool {
-            use std::io::ErrorKind::*;
-            matches!(err.kind(), ConnectionReset | ConnectionAborted | BrokenPipe)
-        }
-        fn suppress_hyper_error(err: &hyper0::Error) -> bool {
-            if err.is_incomplete_message() || err.is_closed() || err.is_timeout() {
-                return true;
-            }
-            if let Some(inner) = err.source() {
-                if let Some(io) = inner.downcast_ref::<std::io::Error>() {
-                    return suppress_io_error(io);
-                }
-            }
-            false
-        }
-
-        let mut connections = FuturesUnordered::new();
-        loop {
-            tokio::select! {
-                stream = self.listener.accept() => {
-                    let (tcp_stream, remote_addr) = match stream {
-                        Ok(stream) => stream,
-                        Err(err) => {
-                            if !suppress_io_error(&err) {
-                                info!("Failed to accept TCP connection: {err:#}");
-                            }
-                            continue;
-                        }
-                    };
-
-                    let service = self.request_service.build(remote_addr);
-                    let tls_acceptor = self.tls_acceptor.clone();
-                    let cancel = cancel.clone();
-
-                    connections.push(tokio::spawn(
-                        async move {
-                            match tls_acceptor {
-                                Some(tls_acceptor) => {
-                                    // Handle HTTPS connection.
-                                    let tls_stream = tokio::select! {
-                                        tls_stream = tls_acceptor.accept(tcp_stream) => tls_stream,
-                                        _ = cancel.cancelled() => return,
-                                    };
-                                    let tls_stream = match tls_stream {
-                                        Ok(tls_stream) => tls_stream,
-                                        Err(err) => {
-                                            if !suppress_io_error(&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!("Failed to serve HTTPS connection: {err:#}");
-                                        }
-                                    }
-                                }
-                                None => {
-                                    // Handle HTTP connection.
-                                    if let Err(err) = Self::serve_connection(tcp_stream, service, cancel).await {
-                                        if !suppress_hyper_error(&err) {
-                                            info!("Failed to serve HTTP connection: {err:#}");
-                                        }
-                                    }
-                                }
-                            };
-                        }));
-                 }
-                Some(conn) = connections.next() => {
-                    if let Err(err) = conn {
-                        error!("Connection panicked: {err:#}");
-                    }
-                }
-                _ = cancel.cancelled() => {
-                    // Wait for graceful shutdown of all connections.
-                    while let Some(conn) = connections.next().await {
-                        if let Err(err) = conn {
-                            error!("Connection panicked: {err:#}");
-                        }
-                    }
-                    break;
-                }
-            }
-        }
-        Ok(())
-    }
-
-    /// Serves HTTP connection with graceful shutdown.
-    async fn serve_connection<I>(
-        io: I,
-        service: RequestService<Body, ApiError>,
-        cancel: CancellationToken,
-    ) -> Result<(), hyper0::Error>
-    where
-        I: AsyncRead + AsyncWrite + Unpin + Send + 'static,
-    {
-        let mut conn = Http::new().serve_connection(io, service).with_upgrades();
-
-        tokio::select! {
-            res = &mut conn => res,
-            _ = cancel.cancelled() => {
-                Pin::new(&mut conn).graceful_shutdown();
-                // Note: connection should still be awaited for graceful shutdown to complete.
-                conn.await
-            }
-        }
-    }
-}

libs/pageserver_api/src/config.rs

@@ -35,7 +35,6 @@ pub struct NodeMetadata {
     pub postgres_port: u16,
     pub http_host: String,
     pub http_port: u16,
-    pub https_port: Option<u16>,
 
     // Deployment tools may write fields to the metadata file beyond what we
     // use in this type: this type intentionally only names fields that require.
@@ -58,9 +57,6 @@ pub struct ConfigToml {
     // types mapped 1:1 into the runtime PageServerConfig type
     pub listen_pg_addr: String,
     pub listen_http_addr: String,
-    pub listen_https_addr: Option<String>,
-    pub ssl_key_file: Utf8PathBuf,
-    pub ssl_cert_file: Utf8PathBuf,
     pub availability_zone: Option<String>,
     #[serde(with = "humantime_serde")]
     pub wait_lsn_timeout: Duration,
@@ -127,10 +123,6 @@ pub struct ConfigToml {
     pub enable_read_path_debugging: Option<bool>,
     #[serde(skip_serializing_if = "Option::is_none")]
     pub validate_wal_contiguity: Option<bool>,
-    #[serde(skip_serializing_if = "Option::is_none")]
-    pub load_previous_heatmap: Option<bool>,
-    #[serde(skip_serializing_if = "Option::is_none")]
-    pub generate_unarchival_heatmap: Option<bool>,
 }
 
 #[derive(Debug, Clone, PartialEq, Eq, serde::Serialize, serde::Deserialize)]
@@ -425,9 +417,6 @@ pub mod defaults {
 
     pub const DEFAULT_WAL_RECEIVER_PROTOCOL: utils::postgres_client::PostgresClientProtocol =
         utils::postgres_client::PostgresClientProtocol::Vanilla;
-
-    pub const DEFAULT_SSL_KEY_FILE: &str = "server.key";
-    pub const DEFAULT_SSL_CERT_FILE: &str = "server.crt";
 }
 
 impl Default for ConfigToml {
@@ -437,9 +426,6 @@ impl Default for ConfigToml {
         Self {
             listen_pg_addr: (DEFAULT_PG_LISTEN_ADDR.to_string()),
             listen_http_addr: (DEFAULT_HTTP_LISTEN_ADDR.to_string()),
-            listen_https_addr: (None),
-            ssl_key_file: Utf8PathBuf::from(DEFAULT_SSL_KEY_FILE),
-            ssl_cert_file: Utf8PathBuf::from(DEFAULT_SSL_CERT_FILE),
             availability_zone: (None),
             wait_lsn_timeout: (humantime::parse_duration(DEFAULT_WAIT_LSN_TIMEOUT)
                 .expect("cannot parse default wait lsn timeout")),
@@ -537,8 +523,6 @@ impl Default for ConfigToml {
                 None
             },
             validate_wal_contiguity: None,
-            load_previous_heatmap: None,
-            generate_unarchival_heatmap: None,
         }
     }
 }

libs/pageserver_api/src/config/tests.rs

@@ -16,30 +16,6 @@ fn test_node_metadata_v1_backward_compatibilty() {
             postgres_port: 23,
             http_host: "localhost".to_string(),
             http_port: 42,
-            https_port: None,
-            other: HashMap::new(),
-        }
-    )
-}
-
-#[test]
-fn test_node_metadata_v2_backward_compatibilty() {
-    let v2 = serde_json::to_vec(&serde_json::json!({
-        "host": "localhost",
-        "port": 23,
-        "http_host": "localhost",
-        "http_port": 42,
-        "https_port": 123,
-    }));
-
-    assert_eq!(
-        serde_json::from_slice::<NodeMetadata>(&v2.unwrap()).unwrap(),
-        NodeMetadata {
-            postgres_host: "localhost".to_string(),
-            postgres_port: 23,
-            http_host: "localhost".to_string(),
-            http_port: 42,
-            https_port: Some(123),
             other: HashMap::new(),
         }
     )

libs/pageserver_api/src/controller_api.rs

@@ -182,66 +182,20 @@ pub struct TenantDescribeResponseShard {
 #[derive(Serialize, Deserialize, Debug)]
 pub struct TenantShardMigrateRequest {
     pub node_id: NodeId,
-
-    /// Optionally, callers may specify the node they are migrating _from_, and the server will
-    /// reject the request if the shard is no longer attached there: this enables writing safer
-    /// clients that don't risk fighting with some other movement of the shard.
     #[serde(default)]
-    pub origin_node_id: Option<NodeId>,
-
-    #[serde(default)]
-    pub migration_config: MigrationConfig,
+    pub migration_config: Option<MigrationConfig>,
 }
 
-#[derive(Serialize, Deserialize, Debug, PartialEq, Eq)]
+#[derive(Serialize, Deserialize, Debug)]
 pub struct MigrationConfig {
-    /// If true, the migration will be executed even if it is to a location with a sub-optimal scheduling
-    /// score: this is usually not what you want, and if you use this then you'll also need to set the
-    /// tenant's scheduling policy to Essential or Pause to avoid the optimiser reverting your migration.
-    ///
-    /// Default: false
-    #[serde(default)]
-    pub override_scheduler: bool,
-
-    /// If true, the migration will be done gracefully by creating a secondary location first and
-    /// waiting for it to warm up before cutting over.  If false, if there is no existing secondary
-    /// location at the destination, the tenant will be migrated immediately.  If the tenant's data
-    /// can't be downloaded within [`Self::secondary_warmup_timeout`], then the migration will go
-    /// ahead but run with a cold cache that can severely reduce performance until it warms up.
-    ///
-    /// When doing a graceful migration, the migration API returns as soon as it is started.
-    ///
-    /// Default: true
-    #[serde(default = "default_prewarm")]
-    pub prewarm: bool,
-
-    /// For non-prewarm migrations which will immediately enter a cutover to the new node: how long to wait
-    /// overall for secondary warmup before cutting over
     #[serde(default)]
     #[serde(with = "humantime_serde")]
     pub secondary_warmup_timeout: Option<Duration>,
-    /// For non-prewarm migrations which will immediately enter a cutover to the new node: how long to wait
-    /// within each secondary download poll call to pageserver.
     #[serde(default)]
     #[serde(with = "humantime_serde")]
     pub secondary_download_request_timeout: Option<Duration>,
 }
 
-fn default_prewarm() -> bool {
-    true
-}
-
-impl Default for MigrationConfig {
-    fn default() -> Self {
-        Self {
-            override_scheduler: false,
-            prewarm: default_prewarm(),
-            secondary_warmup_timeout: None,
-            secondary_download_request_timeout: None,
-        }
-    }
-}
-
 #[derive(Serialize, Clone, Debug)]
 #[serde(into = "NodeAvailabilityWrapper")]
 pub enum NodeAvailability {
@@ -489,7 +443,6 @@ pub struct SafekeeperDescribeResponse {
     pub host: String,
     pub port: i32,
     pub http_port: i32,
-    pub https_port: Option<i32>,
     pub availability_zone_id: String,
     pub scheduling_policy: SkSchedulingPolicy,
 }
@@ -534,43 +487,4 @@ mod test {
             err
         );
     }
-
-    /// Check that a minimal migrate request with no config results in the expected default settings
-    #[test]
-    fn test_migrate_request_decode_defaults() {
-        let json = r#"{
-            "node_id": 123
-        }"#;
-
-        let request: TenantShardMigrateRequest = serde_json::from_str(json).unwrap();
-        assert_eq!(request.node_id, NodeId(123));
-        assert_eq!(request.origin_node_id, None);
-        assert!(!request.migration_config.override_scheduler);
-        assert!(request.migration_config.prewarm);
-        assert_eq!(request.migration_config.secondary_warmup_timeout, None);
-        assert_eq!(
-            request.migration_config.secondary_download_request_timeout,
-            None
-        );
-    }
-
-    /// Check that a partially specified migration config results in the expected default settings
-    #[test]
-    fn test_migration_config_decode_defaults() {
-        // Specify just one field of the config
-        let json = r#"{
-        }"#;
-
-        let config: MigrationConfig = serde_json::from_str(json).unwrap();
-
-        // Check each field's expected default value
-        assert!(!config.override_scheduler);
-        assert!(config.prewarm);
-        assert_eq!(config.secondary_warmup_timeout, None);
-        assert_eq!(config.secondary_download_request_timeout, None);
-        assert_eq!(config.secondary_warmup_timeout, None);
-
-        // Consistency check that the Default impl agrees with our serde defaults
-        assert_eq!(MigrationConfig::default(), config);
-    }
 }

libs/pageserver_api/src/models.rs

@@ -274,31 +274,6 @@ pub struct TimelineCreateRequest {
     pub mode: TimelineCreateRequestMode,
 }
 
-/// Storage controller specific extensions to [`TimelineInfo`].
-#[derive(Serialize, Deserialize, Clone)]
-pub struct TimelineCreateResponseStorcon {
-    #[serde(flatten)]
-    pub timeline_info: TimelineInfo,
-
-    pub safekeepers: Option<SafekeepersInfo>,
-}
-
-/// Safekeepers as returned in timeline creation request to storcon or pushed to
-/// cplane in the post migration hook.
-#[derive(Serialize, Deserialize, Clone)]
-pub struct SafekeepersInfo {
-    pub tenant_id: TenantId,
-    pub timeline_id: TimelineId,
-    pub generation: u32,
-    pub safekeepers: Vec<SafekeeperInfo>,
-}
-
-#[derive(Serialize, Deserialize, Clone)]
-pub struct SafekeeperInfo {
-    pub id: NodeId,
-    pub hostname: String,
-}
-
 #[derive(Serialize, Deserialize, Clone)]
 #[serde(untagged)]
 pub enum TimelineCreateRequestMode {
@@ -1171,15 +1146,6 @@ pub struct TimelineArchivalConfigRequest {
     pub state: TimelineArchivalState,
 }
 
-#[derive(Serialize, Deserialize, PartialEq, Eq, Clone)]
-pub struct TimelinePatchIndexPartRequest {
-    pub rel_size_migration: Option<RelSizeMigration>,
-    pub gc_compaction_last_completed_lsn: Option<Lsn>,
-    pub applied_gc_cutoff_lsn: Option<Lsn>,
-    #[serde(default)]
-    pub force_index_update: bool,
-}
-
 #[derive(Debug, Serialize, Deserialize, Clone)]
 pub struct TimelinesInfoAndOffloaded {
     pub timelines: Vec<TimelineInfo>,
@@ -1199,21 +1165,6 @@ pub struct OffloadedTimelineInfo {
     pub archived_at: chrono::DateTime<chrono::Utc>,
 }
 
-#[derive(Debug, Clone, PartialEq, Eq, Serialize, Deserialize)]
-#[serde(rename_all = "camelCase")]
-pub enum RelSizeMigration {
-    /// The tenant is using the old rel_size format.
-    /// Note that this enum is persisted as `Option<RelSizeMigration>` in the index part, so
-    /// `None` is the same as `Some(RelSizeMigration::Legacy)`.
-    Legacy,
-    /// The tenant is migrating to the new rel_size format. Both old and new rel_size format are
-    /// persisted in the index part. The read path will read both formats and merge them.
-    Migrating,
-    /// The tenant has migrated to the new rel_size format. Only the new rel_size format is persisted
-    /// in the index part, and the read path will not read the old format.
-    Migrated,
-}
-
 /// This represents the output of the "timeline_detail" and "timeline_list" API calls.
 #[derive(Debug, Serialize, Deserialize, Clone)]
 pub struct TimelineInfo {
@@ -1292,11 +1243,7 @@ pub struct TimelineInfo {
     // Forward compatibility: a previous version of the pageserver will receive a JSON. serde::Deserialize does
     // not deny unknown fields by default so it's safe to set the field to some value, though it won't be
     // read.
-    /// Whether the timeline is archived.
     pub is_archived: Option<bool>,
-
-    /// The status of the rel_size migration.
-    pub rel_size_migration: Option<RelSizeMigration>,
 }
 
 #[derive(Debug, Clone, Serialize, Deserialize)]
@@ -1476,14 +1423,8 @@ pub struct TenantScanRemoteStorageResponse {
 #[derive(Serialize, Deserialize, Debug, Clone)]
 #[serde(rename_all = "snake_case")]
 pub enum TenantSorting {
-    /// Total size of layers on local disk for all timelines in a shard.
     ResidentSize,
-    /// The logical size of the largest timeline within a _tenant_ (not shard). Only tracked on
-    /// shard 0, contains the sum across all shards.
     MaxLogicalSize,
-    /// The logical size of the largest timeline within a _tenant_ (not shard), divided by number of
-    /// shards. Only tracked on shard 0, and estimates the per-shard logical size.
-    MaxLogicalSizePerShard,
 }
 
 impl Default for TenantSorting {
@@ -1513,20 +1454,14 @@ pub struct TopTenantShardsRequest {
 pub struct TopTenantShardItem {
     pub id: TenantShardId,
 
-    /// Total size of layers on local disk for all timelines in this shard.
+    /// Total size of layers on local disk for all timelines in this tenant
     pub resident_size: u64,
 
-    /// Total size of layers in remote storage for all timelines in this shard.
+    /// Total size of layers in remote storage for all timelines in this tenant
     pub physical_size: u64,
 
-    /// The largest logical size of a timeline within this _tenant_ (not shard). This is only
-    /// tracked on shard 0, and contains the sum of the logical size across all shards.
+    /// The largest logical size of a timeline within this tenant
     pub max_logical_size: u64,
-
-    /// The largest logical size of a timeline within this _tenant_ (not shard) divided by number of
-    /// shards. This is only tracked on shard 0, and is only an estimate as we divide it evenly by
-    /// shard count, rounded up.
-    pub max_logical_size_per_shard: u64,
 }
 
 #[derive(Serialize, Deserialize, Debug, Default)]

libs/postgres_ffi/src/lib.rs

@@ -396,14 +396,6 @@ pub mod waldecoder {
             self.lsn + self.inputbuf.remaining() as u64
         }
 
-        /// Returns the LSN up to which the WAL decoder has processed.
-        ///
-        /// If [`Self::poll_decode`] returned a record, then this will return
-        /// the end LSN of said record.
-        pub fn lsn(&self) -> Lsn {
-            self.lsn
-        }
-
         pub fn feed_bytes(&mut self, buf: &[u8]) {
             self.inputbuf.extend_from_slice(buf);
         }

libs/proxy/tokio-postgres2/src/cancel_query.rs

@@ -34,13 +34,8 @@ where
         .make_tls_connect(hostname)
         .map_err(|e| Error::tls(e.into()))?;
 
-    let socket = connect_socket::connect_socket(
-        config.host_addr,
-        &config.host,
-        config.port,
-        config.connect_timeout,
-    )
-    .await?;
+    let socket =
+        connect_socket::connect_socket(&config.host, config.port, config.connect_timeout).await?;
 
     cancel_query_raw::cancel_query_raw(socket, ssl_mode, tls, process_id, secret_key).await
 }

libs/proxy/tokio-postgres2/src/client.rs

@@ -1,6 +1,5 @@
 use std::collections::HashMap;
 use std::fmt;
-use std::net::IpAddr;
 use std::sync::Arc;
 use std::task::{Context, Poll};
 use std::time::Duration;
@@ -138,7 +137,6 @@ impl InnerClient {
 
 #[derive(Clone, Serialize, Deserialize)]
 pub struct SocketConfig {
-    pub host_addr: Option<IpAddr>,
     pub host: Host,
     pub port: u16,
     pub connect_timeout: Option<Duration>,
@@ -284,18 +282,6 @@ impl Client {
         simple_query::batch_execute(self.inner(), query).await
     }
 
-    pub async fn discard_all(&self) -> Result<ReadyForQueryStatus, Error> {
-        // clear the prepared statements that are about to be nuked from the postgres session
-        {
-            let mut typeinfo = self.inner.cached_typeinfo.lock();
-            typeinfo.typeinfo = None;
-            typeinfo.typeinfo_composite = None;
-            typeinfo.typeinfo_enum = None;
-        }
-
-        self.batch_execute("discard all").await
-    }
-
     /// Begins a new database transaction.
     ///
     /// The transaction will roll back by default - use the `commit` method to commit it.

libs/proxy/tokio-postgres2/src/config.rs

@@ -1,6 +1,5 @@
 //! Connection configuration.
 
-use std::net::IpAddr;
 use std::time::Duration;
 use std::{fmt, str};
 
@@ -66,7 +65,6 @@ pub enum AuthKeys {
 /// Connection configuration.
 #[derive(Clone, PartialEq, Eq)]
 pub struct Config {
-    pub(crate) host_addr: Option<IpAddr>,
     pub(crate) host: Host,
     pub(crate) port: u16,
 
@@ -85,7 +83,6 @@ impl Config {
     /// Creates a new configuration.
     pub fn new(host: String, port: u16) -> Config {
         Config {
-            host_addr: None,
             host: Host::Tcp(host),
             port,
             password: None,
@@ -166,15 +163,6 @@ impl Config {
         self
     }
 
-    pub fn set_host_addr(&mut self, addr: IpAddr) -> &mut Config {
-        self.host_addr = Some(addr);
-        self
-    }
-
-    pub fn get_host_addr(&self) -> Option<IpAddr> {
-        self.host_addr
-    }
-
     /// Sets the SSL configuration.
     ///
     /// Defaults to `prefer`.

libs/proxy/tokio-postgres2/src/connect.rs

@@ -1,5 +1,3 @@
-use std::net::IpAddr;
-
 use postgres_protocol2::message::backend::Message;
 use tokio::net::TcpStream;
 use tokio::sync::mpsc;
@@ -27,14 +25,13 @@ where
         .make_tls_connect(hostname)
         .map_err(|e| Error::tls(e.into()))?;
 
-    match connect_once(config.host_addr, &config.host, config.port, tls, config).await {
+    match connect_once(&config.host, config.port, tls, config).await {
         Ok((client, connection)) => Ok((client, connection)),
         Err(e) => Err(e),
     }
 }
 
 async fn connect_once<T>(
-    host_addr: Option<IpAddr>,
     host: &Host,
     port: u16,
     tls: T,
@@ -43,7 +40,7 @@ async fn connect_once<T>(
 where
     T: TlsConnect<TcpStream>,
 {
-    let socket = connect_socket(host_addr, host, port, config.connect_timeout).await?;
+    let socket = connect_socket(host, port, config.connect_timeout).await?;
     let RawConnection {
         stream,
         parameters,
@@ -53,7 +50,6 @@ where
     } = connect_raw(socket, tls, config).await?;
 
     let socket_config = SocketConfig {
-        host_addr,
         host: host.clone(),
         port,
         connect_timeout: config.connect_timeout,

libs/proxy/tokio-postgres2/src/connect_socket.rs

@@ -1,6 +1,5 @@
 use std::future::Future;
 use std::io;
-use std::net::{IpAddr, SocketAddr};
 use std::time::Duration;
 
 use tokio::net::{self, TcpStream};
@@ -10,20 +9,15 @@ use crate::Error;
 use crate::config::Host;
 
 pub(crate) async fn connect_socket(
-    host_addr: Option<IpAddr>,
     host: &Host,
     port: u16,
     connect_timeout: Option<Duration>,
 ) -> Result<TcpStream, Error> {
     match host {
         Host::Tcp(host) => {
-            let addrs = match host_addr {
-                Some(addr) => vec![SocketAddr::new(addr, port)],
-                None => net::lookup_host((&**host, port))
-                    .await
-                    .map_err(Error::connect)?
-                    .collect(),
-            };
+            let addrs = net::lookup_host((&**host, port))
+                .await
+                .map_err(Error::connect)?;
 
             let mut last_err = None;
 

libs/safekeeper_api/src/membership.rs

@@ -131,14 +131,6 @@ impl Configuration {
         }
     }
 
-    pub fn new(members: MemberSet) -> Self {
-        Configuration {
-            generation: INITIAL_GENERATION,
-            members,
-            new_members: None,
-        }
-    }
-
     /// Is `sk_id` member of the configuration?
     pub fn contains(&self, sk_id: NodeId) -> bool {
         self.members.contains(sk_id) || self.new_members.as_ref().is_some_and(|m| m.contains(sk_id))

libs/safekeeper_api/src/models.rs

@@ -18,7 +18,7 @@ pub struct SafekeeperStatus {
     pub id: NodeId,
 }
 
-#[derive(Serialize, Deserialize, Clone)]
+#[derive(Serialize, Deserialize)]
 pub struct TimelineCreateRequest {
     pub tenant_id: TenantId,
     pub timeline_id: TimelineId,
@@ -283,7 +283,7 @@ pub struct SafekeeperUtilization {
 }
 
 /// pull_timeline request body.
-#[derive(Debug, Clone, Deserialize, Serialize)]
+#[derive(Debug, Deserialize, Serialize)]
 pub struct PullTimelineRequest {
     pub tenant_id: TenantId,
     pub timeline_id: TimelineId,

libs/utils/Cargo.toml

@@ -15,6 +15,7 @@ arc-swap.workspace = true
 sentry.workspace = true
 async-compression.workspace = true
 anyhow.workspace = true
+backtrace.workspace = true
 bincode.workspace = true
 bytes.workspace = true
 camino.workspace = true

libs/utils/src/sentry_init.rs

@@ -3,24 +3,20 @@ use std::env;
 
 use sentry::ClientInitGuard;
 pub use sentry::release_name;
-use tracing::{error, info};
 
 #[must_use]
 pub fn init_sentry(
     release_name: Option<Cow<'static, str>>,
     extra_options: &[(&str, &str)],
 ) -> Option<ClientInitGuard> {
-    let Ok(dsn) = env::var("SENTRY_DSN") else {
-        info!("not initializing Sentry, no SENTRY_DSN given");
-        return None;
-    };
+    let dsn = env::var("SENTRY_DSN").ok()?;
     let environment = env::var("SENTRY_ENVIRONMENT").unwrap_or_else(|_| "development".into());
 
     let guard = sentry::init((
         dsn,
         sentry::ClientOptions {
-            release: release_name.clone(),
-            environment: Some(environment.clone().into()),
+            release: release_name,
+            environment: Some(environment.into()),
             ..Default::default()
         },
     ));
@@ -29,19 +25,5 @@ pub fn init_sentry(
             scope.set_extra(key, value.into());
         }
     });
-
-    if let Some(dsn) = guard.dsn() {
-        info!(
-            "initialized Sentry for project {}, environment {}, release {} (using API {})",
-            dsn.project_id(),
-            environment,
-            release_name.unwrap_or(Cow::Borrowed("None")),
-            dsn.envelope_api_url(),
-        );
-    } else {
-        // This should panic during sentry::init(), but we may as well cover it.
-        error!("failed to initialize Sentry, invalid DSN");
-    }
-
     Some(guard)
 }

pageserver/Cargo.toml

@@ -48,9 +48,6 @@ pprof.workspace = true
 rand.workspace = true
 range-set-blaze = { version = "0.1.16", features = ["alloc"] }
 regex.workspace = true
-rustls-pemfile.workspace = true
-rustls-pki-types.workspace = true
-rustls.workspace = true
 scopeguard.workspace = true
 send-future.workspace = true
 serde.workspace = true
@@ -65,7 +62,6 @@ tokio = { workspace = true, features = ["process", "sync", "fs", "rt", "io-util"
 tokio-epoll-uring.workspace = true
 tokio-io-timeout.workspace = true
 tokio-postgres.workspace = true
-tokio-rustls.workspace = true
 tokio-stream.workspace = true
 tokio-util.workspace = true
 toml_edit = { workspace = true, features = [ "serde" ] }
@@ -102,7 +98,6 @@ criterion.workspace = true
 hex-literal.workspace = true
 tokio = { workspace = true, features = ["process", "sync", "fs", "rt", "io-util", "time", "test-util"] }
 indoc.workspace = true
-uuid.workspace = true
 
 [[bench]]
 name = "bench_layer_map"
@@ -120,10 +115,6 @@ harness = false
 name = "upload_queue"
 harness = false
 
-[[bench]]
-name = "bench_metrics"
-harness = false
-
 [[bin]]
 name = "test_helper_slow_client_reads"
 required-features = [ "testing" ]

pageserver/benches/bench_layer_map.rs

@@ -7,6 +7,7 @@ use std::time::Instant;
 
 use criterion::measurement::WallTime;
 use criterion::{BenchmarkGroup, Criterion, black_box, criterion_group, criterion_main};
+use pageserver::keyspace::{KeyPartitioning, KeySpace};
 use pageserver::tenant::layer_map::LayerMap;
 use pageserver::tenant::storage_layer::{LayerName, PersistentLayerDesc};
 use pageserver_api::key::Key;
@@ -71,6 +72,41 @@ fn uniform_query_pattern(layer_map: &LayerMap) -> Vec<(Key, Lsn)> {
         .collect()
 }
 
+// Construct a partitioning for testing get_difficulty map when we
+// don't have an exact result of `collect_keyspace` to work with.
+fn uniform_key_partitioning(layer_map: &LayerMap, _lsn: Lsn) -> KeyPartitioning {
+    let mut parts = Vec::new();
+
+    // We add a partition boundary at the start of each image layer,
+    // no matter what lsn range it covers. This is just the easiest
+    // thing to do. A better thing to do would be to get a real
+    // partitioning from some database. Even better, remove the need
+    // for key partitions by deciding where to create image layers
+    // directly based on a coverage-based difficulty map.
+    let mut keys: Vec<_> = layer_map
+        .iter_historic_layers()
+        .filter_map(|l| {
+            if l.is_incremental() {
+                None
+            } else {
+                let kr = l.get_key_range();
+                Some(kr.start.next())
+            }
+        })
+        .collect();
+    keys.sort();
+
+    let mut current_key = Key::from_hex("000000000000000000000000000000000000").unwrap();
+    for key in keys {
+        parts.push(KeySpace {
+            ranges: vec![current_key..key],
+        });
+        current_key = key;
+    }
+
+    KeyPartitioning { parts }
+}
+
 // Benchmark using metadata extracted from our performance test environment, from
 // a project where we have run pgbench many timmes. The pgbench database was initialized
 // between each test run.
@@ -112,6 +148,41 @@ fn bench_from_real_project(c: &mut Criterion) {
     // Choose uniformly distributed queries
     let queries: Vec<(Key, Lsn)> = uniform_query_pattern(&layer_map);
 
+    // Choose inputs for get_difficulty_map
+    let latest_lsn = layer_map
+        .iter_historic_layers()
+        .map(|l| l.get_lsn_range().end)
+        .max()
+        .unwrap();
+    let partitioning = uniform_key_partitioning(&layer_map, latest_lsn);
+
+    // Check correctness of get_difficulty_map
+    // TODO put this in a dedicated test outside of this mod
+    {
+        println!("running correctness check");
+
+        let now = Instant::now();
+        let result_bruteforce = layer_map.get_difficulty_map_bruteforce(latest_lsn, &partitioning);
+        assert!(result_bruteforce.len() == partitioning.parts.len());
+        println!("Finished bruteforce in {:?}", now.elapsed());
+
+        let now = Instant::now();
+        let result_fast = layer_map.get_difficulty_map(latest_lsn, &partitioning, None);
+        assert!(result_fast.len() == partitioning.parts.len());
+        println!("Finished fast in {:?}", now.elapsed());
+
+        // Assert results are equal. Manually iterate for easier debugging.
+        let zip = std::iter::zip(
+            &partitioning.parts,
+            std::iter::zip(result_bruteforce, result_fast),
+        );
+        for (_part, (bruteforce, fast)) in zip {
+            assert_eq!(bruteforce, fast);
+        }
+
+        println!("No issues found");
+    }
+
     // Define and name the benchmark function
     let mut group = c.benchmark_group("real_map");
     group.bench_function("uniform_queries", |b| {
@@ -121,6 +192,11 @@ fn bench_from_real_project(c: &mut Criterion) {
             }
         });
     });
+    group.bench_function("get_difficulty_map", |b| {
+        b.iter(|| {
+            layer_map.get_difficulty_map(latest_lsn, &partitioning, Some(3));
+        });
+    });
     group.finish();
 }
 

pageserver/benches/bench_metrics.rs

@@ -1,366 +0,0 @@
-use criterion::{BenchmarkId, Criterion, criterion_group, criterion_main};
-use utils::id::{TenantId, TimelineId};
-
-//
-// Demonstrates that repeat label values lookup is a multicore scalability bottleneck
-// that is worth avoiding.
-//
-criterion_group!(
-    label_values,
-    label_values::bench_naive_usage,
-    label_values::bench_cache_label_values_lookup
-);
-mod label_values {
-    use super::*;
-
-    pub fn bench_naive_usage(c: &mut Criterion) {
-        let mut g = c.benchmark_group("label_values__naive_usage");
-
-        for ntimelines in [1, 4, 8] {
-            g.bench_with_input(
-                BenchmarkId::new("ntimelines", ntimelines),
-                &ntimelines,
-                |b, ntimelines| {
-                    b.iter_custom(|iters| {
-                        let barrier = std::sync::Barrier::new(*ntimelines + 1);
-
-                        let timelines = (0..*ntimelines)
-                            .map(|_| {
-                                (
-                                    TenantId::generate().to_string(),
-                                    "0000".to_string(),
-                                    TimelineId::generate().to_string(),
-                                )
-                            })
-                            .collect::<Vec<_>>();
-
-                        let metric_vec = metrics::UIntGaugeVec::new(
-                            metrics::opts!("testmetric", "testhelp"),
-                            &["tenant_id", "shard_id", "timeline_id"],
-                        )
-                        .unwrap();
-
-                        std::thread::scope(|s| {
-                            for (tenant_id, shard_id, timeline_id) in &timelines {
-                                s.spawn(|| {
-                                    barrier.wait();
-                                    for _ in 0..iters {
-                                        metric_vec
-                                            .with_label_values(&[tenant_id, shard_id, timeline_id])
-                                            .inc();
-                                    }
-                                    barrier.wait();
-                                });
-                            }
-                            barrier.wait();
-                            let start = std::time::Instant::now();
-                            barrier.wait();
-                            start.elapsed()
-                        })
-                    })
-                },
-            );
-        }
-        g.finish();
-    }
-
-    pub fn bench_cache_label_values_lookup(c: &mut Criterion) {
-        let mut g = c.benchmark_group("label_values__cache_label_values_lookup");
-
-        for ntimelines in [1, 4, 8] {
-            g.bench_with_input(
-                BenchmarkId::new("ntimelines", ntimelines),
-                &ntimelines,
-                |b, ntimelines| {
-                    b.iter_custom(|iters| {
-                        let barrier = std::sync::Barrier::new(*ntimelines + 1);
-
-                        let timelines = (0..*ntimelines)
-                            .map(|_| {
-                                (
-                                    TenantId::generate().to_string(),
-                                    "0000".to_string(),
-                                    TimelineId::generate().to_string(),
-                                )
-                            })
-                            .collect::<Vec<_>>();
-
-                        let metric_vec = metrics::UIntGaugeVec::new(
-                            metrics::opts!("testmetric", "testhelp"),
-                            &["tenant_id", "shard_id", "timeline_id"],
-                        )
-                        .unwrap();
-
-                        std::thread::scope(|s| {
-                            for (tenant_id, shard_id, timeline_id) in &timelines {
-                                s.spawn(|| {
-                                    let metric = metric_vec.with_label_values(&[
-                                        tenant_id,
-                                        shard_id,
-                                        timeline_id,
-                                    ]);
-                                    barrier.wait();
-                                    for _ in 0..iters {
-                                        metric.inc();
-                                    }
-                                    barrier.wait();
-                                });
-                            }
-                            barrier.wait();
-                            let start = std::time::Instant::now();
-                            barrier.wait();
-                            start.elapsed()
-                        })
-                    })
-                },
-            );
-        }
-        g.finish();
-    }
-}
-
-//
-// Demonstrates that even a single metric can be a scalability bottleneck
-// if multiple threads in it concurrently but there's nothing we can do
-// about it without changing the metrics framework to use e.g. sharded counte atomics.
-//
-criterion_group!(
-    single_metric_multicore_scalability,
-    single_metric_multicore_scalability::bench,
-);
-mod single_metric_multicore_scalability {
-    use super::*;
-
-    pub fn bench(c: &mut Criterion) {
-        let mut g = c.benchmark_group("single_metric_multicore_scalability");
-
-        for nthreads in [1, 4, 8] {
-            g.bench_with_input(
-                BenchmarkId::new("nthreads", nthreads),
-                &nthreads,
-                |b, nthreads| {
-                    b.iter_custom(|iters| {
-                        let barrier = std::sync::Barrier::new(*nthreads + 1);
-
-                        let metric = metrics::UIntGauge::new("testmetric", "testhelp").unwrap();
-
-                        std::thread::scope(|s| {
-                            for _ in 0..*nthreads {
-                                s.spawn(|| {
-                                    barrier.wait();
-                                    for _ in 0..iters {
-                                        metric.inc();
-                                    }
-                                    barrier.wait();
-                                });
-                            }
-                            barrier.wait();
-                            let start = std::time::Instant::now();
-                            barrier.wait();
-                            start.elapsed()
-                        })
-                    })
-                },
-            );
-        }
-        g.finish();
-    }
-}
-
-//
-// Demonstrates that even if we cache label value, the propagation of such a cached metric value
-// by Clone'ing it is a scalability bottleneck.
-// The reason is that it's an Arc internally and thus there's contention on the reference count atomics.
-//
-// We can avoid that by having long-lived references per thread (= indirection).
-//
-criterion_group!(
-    propagation_of_cached_label_value,
-    propagation_of_cached_label_value::bench_naive,
-    propagation_of_cached_label_value::bench_long_lived_reference_per_thread,
-);
-mod propagation_of_cached_label_value {
-    use std::sync::Arc;
-
-    use super::*;
-
-    pub fn bench_naive(c: &mut Criterion) {
-        let mut g = c.benchmark_group("propagation_of_cached_label_value__naive");
-
-        for nthreads in [1, 4, 8] {
-            g.bench_with_input(
-                BenchmarkId::new("nthreads", nthreads),
-                &nthreads,
-                |b, nthreads| {
-                    b.iter_custom(|iters| {
-                        let barrier = std::sync::Barrier::new(*nthreads + 1);
-
-                        let metric = metrics::UIntGauge::new("testmetric", "testhelp").unwrap();
-
-                        std::thread::scope(|s| {
-                            for _ in 0..*nthreads {
-                                s.spawn(|| {
-                                    barrier.wait();
-                                    for _ in 0..iters {
-                                        // propagating the metric means we'd clone it into the child RequestContext
-                                        let propagated = metric.clone();
-                                        // simulate some work
-                                        criterion::black_box(propagated);
-                                    }
-                                    barrier.wait();
-                                });
-                            }
-                            barrier.wait();
-                            let start = std::time::Instant::now();
-                            barrier.wait();
-                            start.elapsed()
-                        })
-                    })
-                },
-            );
-        }
-        g.finish();
-    }
-
-    pub fn bench_long_lived_reference_per_thread(c: &mut Criterion) {
-        let mut g =
-            c.benchmark_group("propagation_of_cached_label_value__long_lived_reference_per_thread");
-
-        for nthreads in [1, 4, 8] {
-            g.bench_with_input(
-                BenchmarkId::new("nthreads", nthreads),
-                &nthreads,
-                |b, nthreads| {
-                    b.iter_custom(|iters| {
-                        let barrier = std::sync::Barrier::new(*nthreads + 1);
-
-                        let metric = metrics::UIntGauge::new("testmetric", "testhelp").unwrap();
-
-                        std::thread::scope(|s| {
-                            for _ in 0..*nthreads {
-                                s.spawn(|| {
-                                    // This is the technique.
-                                    let this_threads_metric_reference = Arc::new(metric.clone());
-
-                                    barrier.wait();
-                                    for _ in 0..iters {
-                                        // propagating the metric means we'd clone it into the child RequestContext
-                                        let propagated = Arc::clone(&this_threads_metric_reference);
-                                        // simulate some work (include the pointer chase!)
-                                        criterion::black_box(&*propagated);
-                                    }
-                                    barrier.wait();
-                                });
-                            }
-                            barrier.wait();
-                            let start = std::time::Instant::now();
-                            barrier.wait();
-                            start.elapsed()
-                        })
-                    })
-                },
-            );
-        }
-    }
-}
-
-criterion_main!(
-    label_values,
-    single_metric_multicore_scalability,
-    propagation_of_cached_label_value
-);
-
-/*
-RUST_BACKTRACE=full cargo bench --bench bench_metrics --  --discard-baseline --noplot
-
-Results on an im4gn.2xlarge instance
-
-label_values__naive_usage/ntimelines/1 time:   [178.71 ns 178.74 ns 178.76 ns]
-label_values__naive_usage/ntimelines/4 time:   [532.94 ns 539.59 ns 546.31 ns]
-label_values__naive_usage/ntimelines/8 time:   [1.1082 µs 1.1109 µs 1.1135 µs]
-label_values__cache_label_values_lookup/ntimelines/1 time:   [6.4116 ns 6.4119 ns 6.4123 ns]
-label_values__cache_label_values_lookup/ntimelines/4 time:   [6.3482 ns 6.3819 ns 6.4079 ns]
-label_values__cache_label_values_lookup/ntimelines/8 time:   [6.4213 ns 6.5279 ns 6.6293 ns]
-single_metric_multicore_scalability/nthreads/1 time:   [6.0102 ns 6.0104 ns 6.0106 ns]
-single_metric_multicore_scalability/nthreads/4 time:   [38.127 ns 38.275 ns 38.416 ns]
-single_metric_multicore_scalability/nthreads/8 time:   [73.698 ns 74.882 ns 75.864 ns]
-propagation_of_cached_label_value__naive/nthreads/1 time:   [14.424 ns 14.425 ns 14.426 ns]
-propagation_of_cached_label_value__naive/nthreads/4 time:   [100.71 ns 102.53 ns 104.35 ns]
-propagation_of_cached_label_value__naive/nthreads/8 time:   [211.50 ns 214.44 ns 216.87 ns]
-propagation_of_cached_label_value__long_lived_reference_per_thread/nthreads/1 time:   [14.135 ns 14.147 ns 14.160 ns]
-propagation_of_cached_label_value__long_lived_reference_per_thread/nthreads/4 time:   [14.243 ns 14.255 ns 14.268 ns]
-propagation_of_cached_label_value__long_lived_reference_per_thread/nthreads/8 time:   [14.470 ns 14.682 ns 14.895 ns]
-
-Results on an i3en.3xlarge instance
-
-label_values__naive_usage/ntimelines/1      time:   [117.32 ns 117.53 ns 117.74 ns]
-label_values__naive_usage/ntimelines/4      time:   [736.58 ns 741.12 ns 745.61 ns]
-label_values__naive_usage/ntimelines/8      time:   [1.4513 µs 1.4596 µs 1.4665 µs]
-label_values__cache_label_values_lookup/ntimelines/1      time:   [8.0964 ns 8.0979 ns 8.0995 ns]
-label_values__cache_label_values_lookup/ntimelines/4      time:   [8.1620 ns 8.2912 ns 8.4491 ns]
-label_values__cache_label_values_lookup/ntimelines/8      time:   [14.148 ns 14.237 ns 14.324 ns]
-single_metric_multicore_scalability/nthreads/1      time:   [8.0993 ns 8.1013 ns 8.1046 ns]
-single_metric_multicore_scalability/nthreads/4      time:   [80.039 ns 80.672 ns 81.297 ns]
-single_metric_multicore_scalability/nthreads/8      time:   [153.58 ns 154.23 ns 154.90 ns]
-propagation_of_cached_label_value__naive/nthreads/1     time:   [13.924 ns 13.926 ns 13.928 ns]
-propagation_of_cached_label_value__naive/nthreads/4     time:   [143.66 ns 145.27 ns 146.59 ns]
-propagation_of_cached_label_value__naive/nthreads/8     time:   [296.51 ns 297.90 ns 299.30 ns]
-propagation_of_cached_label_value__long_lived_reference_per_thread/nthreads/1     time:   [14.013 ns 14.149 ns 14.308 ns]
-propagation_of_cached_label_value__long_lived_reference_per_thread/nthreads/4     time:   [14.311 ns 14.625 ns 14.984 ns]
-propagation_of_cached_label_value__long_lived_reference_per_thread/nthreads/8     time:   [25.981 ns 26.227 ns 26.476 ns]
-
-Results on an Standard L16s v3 (16 vcpus, 128 GiB memory)  Intel(R) Xeon(R) Platinum 8370C CPU @ 2.80GHz
-
-label_values__naive_usage/ntimelines/1      time:   [101.63 ns 101.84 ns 102.06 ns]
-label_values__naive_usage/ntimelines/4      time:   [417.55 ns 424.73 ns 432.63 ns]
-label_values__naive_usage/ntimelines/8      time:   [874.91 ns 889.51 ns 904.25 ns]
-label_values__cache_label_values_lookup/ntimelines/1      time:   [5.7724 ns 5.7760 ns 5.7804 ns]
-label_values__cache_label_values_lookup/ntimelines/4      time:   [7.8878 ns 7.9401 ns 8.0034 ns]
-label_values__cache_label_values_lookup/ntimelines/8      time:   [7.2621 ns 7.6354 ns 8.0337 ns]
-single_metric_multicore_scalability/nthreads/1      time:   [5.7710 ns 5.7744 ns 5.7785 ns]
-single_metric_multicore_scalability/nthreads/4      time:   [66.629 ns 66.994 ns 67.336 ns]
-single_metric_multicore_scalability/nthreads/8      time:   [130.85 ns 131.98 ns 132.91 ns]
-propagation_of_cached_label_value__naive/nthreads/1     time:   [11.540 ns 11.546 ns 11.553 ns]
-propagation_of_cached_label_value__naive/nthreads/4     time:   [131.22 ns 131.90 ns 132.56 ns]
-propagation_of_cached_label_value__naive/nthreads/8     time:   [260.99 ns 262.75 ns 264.26 ns]
-propagation_of_cached_label_value__long_lived_reference_per_thread/nthreads/1     time:   [11.544 ns 11.550 ns 11.557 ns]
-propagation_of_cached_label_value__long_lived_reference_per_thread/nthreads/4     time:   [11.568 ns 11.642 ns 11.763 ns]
-propagation_of_cached_label_value__long_lived_reference_per_thread/nthreads/8     time:   [13.416 ns 14.121 ns 14.886 ns
-
-Results on an M4 MAX MacBook Pro   Total Number of Cores:	14 (10 performance and 4 efficiency)
-
-label_values__naive_usage/ntimelines/1      time:   [52.711 ns 53.026 ns 53.381 ns]
-label_values__naive_usage/ntimelines/4      time:   [323.99 ns 330.40 ns 337.53 ns]
-label_values__naive_usage/ntimelines/8      time:   [1.1615 µs 1.1998 µs 1.2399 µs]
-label_values__cache_label_values_lookup/ntimelines/1      time:   [1.6635 ns 1.6715 ns 1.6809 ns]
-label_values__cache_label_values_lookup/ntimelines/4      time:   [1.7786 ns 1.7876 ns 1.8028 ns]
-label_values__cache_label_values_lookup/ntimelines/8      time:   [1.8195 ns 1.8371 ns 1.8665 ns]
-single_metric_multicore_scalability/nthreads/1      time:   [1.7764 ns 1.7909 ns 1.8079 ns]
-single_metric_multicore_scalability/nthreads/4      time:   [33.875 ns 34.868 ns 35.923 ns]
-single_metric_multicore_scalability/nthreads/8      time:   [226.85 ns 235.30 ns 244.18 ns]
-propagation_of_cached_label_value__naive/nthreads/1     time:   [3.4337 ns 3.4491 ns 3.4660 ns]
-propagation_of_cached_label_value__naive/nthreads/4     time:   [69.486 ns 71.937 ns 74.472 ns]
-propagation_of_cached_label_value__naive/nthreads/8     time:   [434.87 ns 456.47 ns 477.84 ns]
-propagation_of_cached_label_value__long_lived_reference_per_thread/nthreads/1     time:   [3.3767 ns 3.3974 ns 3.4220 ns]
-propagation_of_cached_label_value__long_lived_reference_per_thread/nthreads/4     time:   [3.6105 ns 4.2355 ns 5.1463 ns]
-propagation_of_cached_label_value__long_lived_reference_per_thread/nthreads/8     time:   [4.0889 ns 4.9714 ns 6.0779 ns]
-
-Results on a Hetzner AX102 AMD Ryzen 9 7950X3D 16-Core Processor
-
-label_values__naive_usage/ntimelines/1      time:   [64.510 ns 64.559 ns 64.610 ns]
-label_values__naive_usage/ntimelines/4      time:   [309.71 ns 326.09 ns 342.32 ns]
-label_values__naive_usage/ntimelines/8      time:   [776.92 ns 819.35 ns 856.93 ns]
-label_values__cache_label_values_lookup/ntimelines/1      time:   [1.2855 ns 1.2943 ns 1.3021 ns]
-label_values__cache_label_values_lookup/ntimelines/4      time:   [1.3865 ns 1.4139 ns 1.4441 ns]
-label_values__cache_label_values_lookup/ntimelines/8      time:   [1.5311 ns 1.5669 ns 1.6046 ns]
-single_metric_multicore_scalability/nthreads/1      time:   [1.1927 ns 1.1981 ns 1.2049 ns]
-single_metric_multicore_scalability/nthreads/4      time:   [24.346 ns 25.439 ns 26.634 ns]
-single_metric_multicore_scalability/nthreads/8      time:   [58.666 ns 60.137 ns 61.486 ns]
-propagation_of_cached_label_value__naive/nthreads/1     time:   [2.7067 ns 2.7238 ns 2.7402 ns]
-propagation_of_cached_label_value__naive/nthreads/4     time:   [62.723 ns 66.214 ns 69.787 ns]
-propagation_of_cached_label_value__naive/nthreads/8     time:   [164.24 ns 170.10 ns 175.68 ns]
-propagation_of_cached_label_value__long_lived_reference_per_thread/nthreads/1     time:   [2.2915 ns 2.2960 ns 2.3012 ns]
-propagation_of_cached_label_value__long_lived_reference_per_thread/nthreads/4     time:   [2.5726 ns 2.6158 ns 2.6624 ns]
-propagation_of_cached_label_value__long_lived_reference_per_thread/nthreads/8     time:   [2.7068 ns 2.8243 ns 2.9824 ns]
-
-*/

pageserver/client/src/mgmt_api.rs

@@ -7,7 +7,7 @@ use http_utils::error::HttpErrorBody;
 use pageserver_api::models::*;
 use pageserver_api::shard::TenantShardId;
 pub use reqwest::Body as ReqwestBody;
-use reqwest::{Certificate, IntoUrl, Method, StatusCode};
+use reqwest::{IntoUrl, Method, StatusCode};
 use utils::id::{TenantId, TimelineId};
 use utils::lsn::Lsn;
 
@@ -38,9 +38,6 @@ pub enum Error {
 
     #[error("Cancelled")]
     Cancelled,
-
-    #[error("create client: {0}{}", .0.source().map(|e| format!(": {e}")).unwrap_or_default())]
-    CreateClient(reqwest::Error),
 }
 
 pub type Result<T> = std::result::Result<T, Error>;
@@ -72,17 +69,8 @@ pub enum ForceAwaitLogicalSize {
 }
 
 impl Client {
-    pub fn new(
-        mgmt_api_endpoint: String,
-        jwt: Option<&str>,
-        ssl_ca_cert: Option<Certificate>,
-    ) -> Result<Self> {
-        let mut http_client = reqwest::Client::builder();
-        if let Some(ssl_ca_cert) = ssl_ca_cert {
-            http_client = http_client.add_root_certificate(ssl_ca_cert);
-        }
-        let http_client = http_client.build().map_err(Error::CreateClient)?;
-        Ok(Self::from_client(http_client, mgmt_api_endpoint, jwt))
+    pub fn new(mgmt_api_endpoint: String, jwt: Option<&str>) -> Self {
+        Self::from_client(reqwest::Client::new(), mgmt_api_endpoint, jwt)
     }
 
     pub fn from_client(
@@ -113,10 +101,12 @@ impl Client {
         debug_assert!(path.starts_with('/'));
         let uri = format!("{}{}", self.mgmt_api_endpoint, path);
 
-        let mut req = self.client.request(Method::GET, uri);
-        if let Some(value) = &self.authorization_header {
-            req = req.header(reqwest::header::AUTHORIZATION, value);
-        }
+        let req = self.client.request(Method::GET, uri);
+        let req = if let Some(value) = &self.authorization_header {
+            req.header(reqwest::header::AUTHORIZATION, value)
+        } else {
+            req
+        };
         req.send().await.map_err(Error::ReceiveBody)
     }
 
@@ -490,7 +480,6 @@ impl Client {
         tenant_shard_id: TenantShardId,
         timeline_id: TimelineId,
         concurrency: Option<usize>,
-        recurse: bool,
     ) -> Result<()> {
         let mut path = reqwest::Url::parse(&format!(
             "{}/v1/tenant/{}/timeline/{}/download_heatmap_layers",
@@ -498,9 +487,6 @@ impl Client {
         ))
         .expect("Cannot build URL");
 
-        path.query_pairs_mut()
-            .append_pair("recurse", &format!("{}", recurse));
-
         if let Some(concurrency) = concurrency {
             path.query_pairs_mut()
                 .append_pair("concurrency", &format!("{}", concurrency));

pageserver/pagebench/Cargo.toml

@@ -15,7 +15,6 @@ hdrhistogram.workspace = true
 humantime.workspace = true
 humantime-serde.workspace = true
 rand.workspace = true
-reqwest.workspace=true
 serde.workspace = true
 serde_json.workspace = true
 tracing.workspace = true

pageserver/pagebench/src/cmd/aux_files.rs

@@ -36,8 +36,7 @@ async fn main_impl(args: Args) -> anyhow::Result<()> {
     let mgmt_api_client = Arc::new(pageserver_client::mgmt_api::Client::new(
         args.mgmt_api_endpoint.clone(),
         args.pageserver_jwt.as_deref(),
-        None, // TODO: support ssl_ca_file for https APIs in pagebench.
-    )?);
+    ));
 
     // discover targets
     let timelines: Vec<TenantTimelineId> = crate::util::cli::targets::discover(

pageserver/pagebench/src/cmd/basebackup.rs

@@ -77,8 +77,7 @@ async fn main_impl(
     let mgmt_api_client = Arc::new(pageserver_client::mgmt_api::Client::new(
         args.mgmt_api_endpoint.clone(),
         args.pageserver_jwt.as_deref(),
-        None, // TODO: support ssl_ca_file for https APIs in pagebench.
-    )?);
+    ));
 
     // discover targets
     let timelines: Vec<TenantTimelineId> = crate::util::cli::targets::discover(

pageserver/pagebench/src/cmd/getpage_latest_lsn.rs

@@ -125,8 +125,7 @@ async fn main_impl(
     let mgmt_api_client = Arc::new(pageserver_client::mgmt_api::Client::new(
         args.mgmt_api_endpoint.clone(),
         args.pageserver_jwt.as_deref(),
-        None, // TODO: support ssl_ca_file for https APIs in pagebench.
-    )?);
+    ));
 
     if let Some(engine_str) = &args.set_io_engine {
         mgmt_api_client.put_io_engine(engine_str).await?;

pageserver/pagebench/src/cmd/ondemand_download_churn.rs

@@ -83,8 +83,7 @@ async fn main_impl(args: Args) -> anyhow::Result<()> {
     let mgmt_api_client = Arc::new(pageserver_client::mgmt_api::Client::new(
         args.mgmt_api_endpoint.clone(),
         args.pageserver_jwt.as_deref(),
-        None, // TODO: support ssl_ca_file for https APIs in pagebench.
-    )?);
+    ));
 
     if let Some(engine_str) = &args.set_io_engine {
         mgmt_api_client.put_io_engine(engine_str).await?;

pageserver/pagebench/src/cmd/trigger_initial_size_calculation.rs

@@ -40,8 +40,7 @@ async fn main_impl(args: Args) -> anyhow::Result<()> {
     let mgmt_api_client = Arc::new(pageserver_client::mgmt_api::Client::new(
         args.mgmt_api_endpoint.clone(),
         args.pageserver_jwt.as_deref(),
-        None, // TODO: support ssl_ca_file for https APIs in pagebench.
-    )?);
+    ));
 
     // discover targets
     let timelines: Vec<TenantTimelineId> = crate::util::cli::targets::discover(

pageserver/src/basebackup.rs

@@ -33,9 +33,8 @@ use utils::lsn::Lsn;
 
 use crate::context::RequestContext;
 use crate::pgdatadir_mapping::Version;
+use crate::tenant::Timeline;
 use crate::tenant::storage_layer::IoConcurrency;
-use crate::tenant::timeline::GetVectoredError;
-use crate::tenant::{PageReconstructError, Timeline};
 
 #[derive(Debug, thiserror::Error)]
 pub enum BasebackupError {
@@ -43,26 +42,6 @@ pub enum BasebackupError {
     Server(#[from] anyhow::Error),
     #[error("basebackup client error {0:#} when {1}")]
     Client(#[source] io::Error, &'static str),
-    #[error("basebackup during shutdown")]
-    Shutdown,
-}
-
-impl From<PageReconstructError> for BasebackupError {
-    fn from(value: PageReconstructError) -> Self {
-        match value {
-            PageReconstructError::Cancelled => BasebackupError::Shutdown,
-            err => BasebackupError::Server(err.into()),
-        }
-    }
-}
-
-impl From<GetVectoredError> for BasebackupError {
-    fn from(value: GetVectoredError) -> Self {
-        match value {
-            GetVectoredError::Cancelled => BasebackupError::Shutdown,
-            err => BasebackupError::Server(err.into()),
-        }
-    }
 }
 
 /// Create basebackup with non-rel data in it.
@@ -148,7 +127,7 @@ where
             timeline
                 .gate
                 .enter()
-                .map_err(|_| BasebackupError::Shutdown)?,
+                .map_err(|e| BasebackupError::Server(e.into()))?,
         ),
     };
     basebackup
@@ -344,7 +323,8 @@ where
             let slru_partitions = self
                 .timeline
                 .get_slru_keyspace(Version::Lsn(self.lsn), self.ctx)
-                .await?
+                .await
+                .map_err(|e| BasebackupError::Server(e.into()))?
                 .partition(
                     self.timeline.get_shard_identity(),
                     Timeline::MAX_GET_VECTORED_KEYS * BLCKSZ as u64,
@@ -356,10 +336,11 @@ where
                 let blocks = self
                     .timeline
                     .get_vectored(part, self.lsn, self.io_concurrency.clone(), self.ctx)
-                    .await?;
+                    .await
+                    .map_err(|e| BasebackupError::Server(e.into()))?;
 
                 for (key, block) in blocks {
-                    let block = block?;
+                    let block = block.map_err(|e| BasebackupError::Server(e.into()))?;
                     slru_builder.add_block(&key, block).await?;
                 }
             }
@@ -368,8 +349,11 @@ where
 
         let mut min_restart_lsn: Lsn = Lsn::MAX;
         // Create tablespace directories
-        for ((spcnode, dbnode), has_relmap_file) in
-            self.timeline.list_dbdirs(self.lsn, self.ctx).await?
+        for ((spcnode, dbnode), has_relmap_file) in self
+            .timeline
+            .list_dbdirs(self.lsn, self.ctx)
+            .await
+            .map_err(|e| BasebackupError::Server(e.into()))?
         {
             self.add_dbdir(spcnode, dbnode, has_relmap_file).await?;
 
@@ -378,7 +362,8 @@ where
             let rels = self
                 .timeline
                 .list_rels(spcnode, dbnode, Version::Lsn(self.lsn), self.ctx)
-                .await?;
+                .await
+                .map_err(|e| BasebackupError::Server(e.into()))?;
             for &rel in rels.iter() {
                 // Send init fork as main fork to provide well formed empty
                 // contents of UNLOGGED relations. Postgres copies it in
@@ -406,7 +391,8 @@ where
         let aux_files = self
             .timeline
             .list_aux_files(self.lsn, self.ctx, self.io_concurrency.clone())
-            .await?;
+            .await
+            .map_err(|e| BasebackupError::Server(e.into()))?;
         let aux_scan_time = start_time.elapsed();
         let aux_estimated_size = aux_files
             .values()
@@ -465,14 +451,16 @@ where
         for xid in self
             .timeline
             .list_twophase_files(self.lsn, self.ctx)
-            .await?
+            .await
+            .map_err(|e| BasebackupError::Server(e.into()))?
         {
             self.add_twophase_file(xid).await?;
         }
         let repl_origins = self
             .timeline
             .get_replorigins(self.lsn, self.ctx, self.io_concurrency.clone())
-            .await?;
+            .await
+            .map_err(|e| BasebackupError::Server(e.into()))?;
         let n_origins = repl_origins.len();
         if n_origins != 0 {
             //
@@ -517,7 +505,8 @@ where
         let nblocks = self
             .timeline
             .get_rel_size(src, Version::Lsn(self.lsn), self.ctx)
-            .await?;
+            .await
+            .map_err(|e| BasebackupError::Server(e.into()))?;
 
         // If the relation is empty, create an empty file
         if nblocks == 0 {
@@ -543,7 +532,8 @@ where
                     // TODO: investigate using get_vectored for the entire startblk..endblk range.
                     // But this code path is not on the critical path for most basebackups (?).
                     .get(rel_block_to_key(src, blknum), self.lsn, self.ctx)
-                    .await?;
+                    .await
+                    .map_err(|e| BasebackupError::Server(e.into()))?;
                 segment_data.extend_from_slice(&img[..]);
             }
 
@@ -577,7 +567,8 @@ where
             let img = self
                 .timeline
                 .get_relmap_file(spcnode, dbnode, Version::Lsn(self.lsn), self.ctx)
-                .await?;
+                .await
+                .map_err(|e| BasebackupError::Server(e.into()))?;
 
             if img.len()
                 != dispatch_pgversion!(self.timeline.pg_version, pgv::bindings::SIZEOF_RELMAPFILE)
@@ -631,7 +622,8 @@ where
                 && self
                     .timeline
                     .list_rels(spcnode, dbnode, Version::Lsn(self.lsn), self.ctx)
-                    .await?
+                    .await
+                    .map_err(|e| BasebackupError::Server(e.into()))?
                     .is_empty()
             {
                 return Ok(());
@@ -682,7 +674,8 @@ where
         let img = self
             .timeline
             .get_twophase_file(xid, self.lsn, self.ctx)
-            .await?;
+            .await
+            .map_err(|e| BasebackupError::Server(e.into()))?;
 
         let mut buf = BytesMut::new();
         buf.extend_from_slice(&img[..]);

pageserver/src/bin/pageserver.rs

@@ -25,12 +25,11 @@ use pageserver::task_mgr::{
 };
 use pageserver::tenant::{TenantSharedResources, mgr, secondary};
 use pageserver::{
-    CancellableTask, ConsumptionMetricsTasks, HttpEndpointListener, HttpsEndpointListener, http,
-    page_cache, page_service, task_mgr, virtual_file,
+    CancellableTask, ConsumptionMetricsTasks, HttpEndpointListener, http, page_cache, page_service,
+    task_mgr, virtual_file,
 };
 use postgres_backend::AuthType;
 use remote_storage::GenericRemoteStorage;
-use rustls_pki_types::{CertificateDer, PrivateKeyDer};
 use tokio::signal::unix::SignalKind;
 use tokio::time::Instant;
 use tokio_util::sync::CancellationToken;
@@ -344,15 +343,8 @@ fn start_pageserver(
     info!("Starting pageserver http handler on {http_addr}");
     let http_listener = tcp_listener::bind(http_addr)?;
 
-    let https_listener = match conf.listen_https_addr.as_ref() {
-        Some(https_addr) => {
-            info!("Starting pageserver https handler on {https_addr}");
-            Some(tcp_listener::bind(https_addr)?)
-        }
-        None => None,
-    };
-
     let pg_addr = &conf.listen_pg_addr;
+
     info!("Starting pageserver pg protocol handler on {pg_addr}");
     let pageserver_listener = tcp_listener::bind(pg_addr)?;
 
@@ -583,8 +575,9 @@ fn start_pageserver(
 
     // Start up the service to handle HTTP mgmt API request. We created the
     // listener earlier already.
-    let (http_endpoint_listener, https_endpoint_listener) = {
+    let http_endpoint_listener = {
         let _rt_guard = MGMT_REQUEST_RUNTIME.enter(); // for hyper
+        let cancel = CancellationToken::new();
 
         let router_state = Arc::new(
             http::routes::State::new(
@@ -599,51 +592,22 @@ fn start_pageserver(
             )
             .context("Failed to initialize router state")?,
         );
-
         let router = http::make_router(router_state, launch_ts, http_auth.clone())?
             .build()
             .map_err(|err| anyhow!(err))?;
+        let service = http_utils::RouterService::new(router).unwrap();
+        let server = hyper0::Server::from_tcp(http_listener)?
+            .serve(service)
+            .with_graceful_shutdown({
+                let cancel = cancel.clone();
+                async move { cancel.clone().cancelled().await }
+            });
 
-        let service =
-            Arc::new(http_utils::RequestServiceBuilder::new(router).map_err(|err| anyhow!(err))?);
-
-        let http_task = {
-            let server =
-                http_utils::server::Server::new(Arc::clone(&service), http_listener, None)?;
-            let cancel = CancellationToken::new();
-
-            let task = MGMT_REQUEST_RUNTIME.spawn(task_mgr::exit_on_panic_or_error(
-                "http endpoint listener",
-                server.serve(cancel.clone()),
-            ));
-            HttpEndpointListener(CancellableTask { task, cancel })
-        };
-
-        let https_task = match https_listener {
-            Some(https_listener) => {
-                let certs = load_certs(&conf.ssl_cert_file)?;
-                let key = load_private_key(&conf.ssl_key_file)?;
-
-                let server_config = rustls::ServerConfig::builder()
-                    .with_no_client_auth()
-                    .with_single_cert(certs, key)?;
-
-                let tls_acceptor = tokio_rustls::TlsAcceptor::from(Arc::new(server_config));
-
-                let server =
-                    http_utils::server::Server::new(service, https_listener, Some(tls_acceptor))?;
-                let cancel = CancellationToken::new();
-
-                let task = MGMT_REQUEST_RUNTIME.spawn(task_mgr::exit_on_panic_or_error(
-                    "https endpoint listener",
-                    server.serve(cancel.clone()),
-                ));
-                Some(HttpsEndpointListener(CancellableTask { task, cancel }))
-            }
-            None => None,
-        };
-
-        (http_task, https_task)
+        let task = MGMT_REQUEST_RUNTIME.spawn(task_mgr::exit_on_panic_or_error(
+            "http endpoint listener",
+            server,
+        ));
+        HttpEndpointListener(CancellableTask { task, cancel })
     };
 
     let consumption_metrics_tasks = {
@@ -719,7 +683,6 @@ fn start_pageserver(
         shutdown_pageserver.cancel();
         pageserver::shutdown_pageserver(
             http_endpoint_listener,
-            https_endpoint_listener,
             page_service,
             consumption_metrics_tasks,
             disk_usage_eviction_task,
@@ -734,25 +697,6 @@ fn start_pageserver(
     })
 }
 
-fn load_certs(filename: &Utf8Path) -> std::io::Result<Vec<CertificateDer<'static>>> {
-    let file = std::fs::File::open(filename)?;
-    let mut reader = std::io::BufReader::new(file);
-
-    rustls_pemfile::certs(&mut reader).collect()
-}
-
-fn load_private_key(filename: &Utf8Path) -> anyhow::Result<PrivateKeyDer<'static>> {
-    let file = std::fs::File::open(filename)?;
-    let mut reader = std::io::BufReader::new(file);
-
-    let key = rustls_pemfile::private_key(&mut reader)?;
-
-    key.ok_or(anyhow::anyhow!(
-        "no private key found in {}",
-        filename.as_str(),
-    ))
-}
-
 async fn create_remote_storage_client(
     conf: &'static PageServerConf,
 ) -> anyhow::Result<GenericRemoteStorage> {

pageserver/src/config.rs

@@ -53,11 +53,6 @@ pub struct PageServerConf {
     pub listen_pg_addr: String,
     /// Example (default): 127.0.0.1:9898
     pub listen_http_addr: String,
-    /// Example: 127.0.0.1:9899
-    pub listen_https_addr: Option<String>,
-
-    pub ssl_key_file: Utf8PathBuf,
-    pub ssl_cert_file: Utf8PathBuf,
 
     /// Current availability zone. Used for traffic metrics.
     pub availability_zone: Option<String>,
@@ -199,13 +194,6 @@ pub struct PageServerConf {
     /// Interpreted protocol feature: if enabled, validate that the logical WAL received from
     /// safekeepers does not have gaps.
     pub validate_wal_contiguity: bool,
-
-    /// When set, the previously written to disk heatmap is loaded on tenant attach and used
-    /// to avoid clobbering the heatmap from new, cold, attached locations.
-    pub load_previous_heatmap: bool,
-
-    /// When set, include visible layers in the next uploaded heatmaps of an unarchived timeline.
-    pub generate_unarchival_heatmap: bool,
 }
 
 /// Token for authentication to safekeepers
@@ -322,9 +310,6 @@ impl PageServerConf {
         let pageserver_api::config::ConfigToml {
             listen_pg_addr,
             listen_http_addr,
-            listen_https_addr,
-            ssl_key_file,
-            ssl_cert_file,
             availability_zone,
             wait_lsn_timeout,
             wal_redo_timeout,
@@ -373,8 +358,6 @@ impl PageServerConf {
             get_vectored_concurrent_io,
             enable_read_path_debugging,
             validate_wal_contiguity,
-            load_previous_heatmap,
-            generate_unarchival_heatmap,
         } = config_toml;
 
         let mut conf = PageServerConf {
@@ -383,9 +366,6 @@ impl PageServerConf {
             // ------------------------------------------------------------
             listen_pg_addr,
             listen_http_addr,
-            listen_https_addr,
-            ssl_key_file,
-            ssl_cert_file,
             availability_zone,
             wait_lsn_timeout,
             wal_redo_timeout,
@@ -467,8 +447,6 @@ impl PageServerConf {
             no_sync: no_sync.unwrap_or(false),
             enable_read_path_debugging: enable_read_path_debugging.unwrap_or(false),
             validate_wal_contiguity: validate_wal_contiguity.unwrap_or(false),
-            load_previous_heatmap: load_previous_heatmap.unwrap_or(true),
-            generate_unarchival_heatmap: generate_unarchival_heatmap.unwrap_or(true),
         };
 
         // ------------------------------------------------------------
@@ -502,9 +480,7 @@ impl PageServerConf {
     #[cfg(test)]
     pub fn test_repo_dir(test_name: &str) -> Utf8PathBuf {
         let test_output_dir = std::env::var("TEST_OUTPUT").unwrap_or("../tmp_check".into());
-
-        let test_id = uuid::Uuid::new_v4();
-        Utf8PathBuf::from(format!("{test_output_dir}/test_{test_name}_{test_id}"))
+        Utf8PathBuf::from(format!("{test_output_dir}/test_{test_name}"))
     }
 
     pub fn dummy_conf(repo_dir: Utf8PathBuf) -> Self {
@@ -517,8 +493,6 @@ impl PageServerConf {
             metric_collection_interval: Duration::from_secs(60),
             synthetic_size_calculation_interval: Duration::from_secs(60),
             background_task_maximum_delay: Duration::ZERO,
-            load_previous_heatmap: Some(true),
-            generate_unarchival_heatmap: Some(true),
             ..Default::default()
         };
         PageServerConf::parse_and_validate(NodeId(0), config_toml, &repo_dir).unwrap()

pageserver/src/context.rs

@@ -89,112 +89,16 @@
 //! [`RequestContext`] argument. Functions in the middle of the call chain
 //! only need to pass it on.
 
-use std::sync::Arc;
-
-use once_cell::sync::Lazy;
-use tracing::warn;
-use utils::{id::TimelineId, shard::TenantShardId};
-
-use crate::{
-    metrics::{StorageIoSizeMetrics, TimelineMetrics},
-    task_mgr::TaskKind,
-    tenant::Timeline,
-};
+use crate::task_mgr::TaskKind;
 
 // The main structure of this module, see module-level comment.
+#[derive(Debug)]
 pub struct RequestContext {
     task_kind: TaskKind,
     download_behavior: DownloadBehavior,
     access_stats_behavior: AccessStatsBehavior,
     page_content_kind: PageContentKind,
     read_path_debug: bool,
-    scope: Scope,
-}
-
-#[derive(Clone)]
-pub(crate) enum Scope {
-    Global {
-        io_size_metrics: &'static crate::metrics::StorageIoSizeMetrics,
-    },
-    SecondaryTenant {
-        io_size_metrics: &'static crate::metrics::StorageIoSizeMetrics,
-    },
-    SecondaryTimeline {
-        io_size_metrics: crate::metrics::StorageIoSizeMetrics,
-    },
-    Timeline {
-        // We wrap the `Arc<TimelineMetrics>`s inside another Arc to avoid child
-        // context creation contending for the ref counters of the Arc<TimelineMetrics>,
-        // which are shared among all tasks that operate on the timeline, especially
-        // concurrent page_service connections.
-        #[allow(clippy::redundant_allocation)]
-        arc_arc: Arc<Arc<TimelineMetrics>>,
-    },
-    #[cfg(test)]
-    UnitTest {
-        io_size_metrics: &'static crate::metrics::StorageIoSizeMetrics,
-    },
-}
-
-static GLOBAL_IO_SIZE_METRICS: Lazy<crate::metrics::StorageIoSizeMetrics> =
-    Lazy::new(|| crate::metrics::StorageIoSizeMetrics::new("*", "*", "*"));
-
-impl Scope {
-    pub(crate) fn new_global() -> Self {
-        Scope::Global {
-            io_size_metrics: &GLOBAL_IO_SIZE_METRICS,
-        }
-    }
-    /// NB: this allocates, so, use only at relatively long-lived roots, e.g., at start
-    /// of a compaction iteration.
-    pub(crate) fn new_timeline(timeline: &Timeline) -> Self {
-        Scope::Timeline {
-            arc_arc: Arc::new(Arc::clone(&timeline.metrics)),
-        }
-    }
-    pub(crate) fn new_page_service_pagestream(
-        timeline_handle: &crate::tenant::timeline::handle::Handle<
-            crate::page_service::TenantManagerTypes,
-        >,
-    ) -> Self {
-        Scope::Timeline {
-            arc_arc: Arc::clone(&timeline_handle.metrics),
-        }
-    }
-    pub(crate) fn new_secondary_timeline(
-        tenant_shard_id: &TenantShardId,
-        timeline_id: &TimelineId,
-    ) -> Self {
-        // TODO(https://github.com/neondatabase/neon/issues/11156): secondary timelines have no infrastructure for metrics lifecycle.
-
-        let tenant_id = tenant_shard_id.tenant_id.to_string();
-        let shard_id = tenant_shard_id.shard_slug().to_string();
-        let timeline_id = timeline_id.to_string();
-
-        let io_size_metrics =
-            crate::metrics::StorageIoSizeMetrics::new(&tenant_id, &shard_id, &timeline_id);
-        Scope::SecondaryTimeline { io_size_metrics }
-    }
-    pub(crate) fn new_secondary_tenant(_tenant_shard_id: &TenantShardId) -> Self {
-        // Before propagating metrics via RequestContext, the labels were inferred from file path.
-        // The only user of VirtualFile at tenant scope is the heatmap download & read.
-        // The inferred labels for the path of the heatmap file on local disk were that of the global metric (*,*,*).
-        // Thus, we do the same here, and extend that for anything secondary-tenant scoped.
-        //
-        // If we want to have (tenant_id, shard_id, '*') labels for secondary tenants in the future,
-        // we will need to think about the metric lifecycle, i.e., remove them during secondary tenant shutdown,
-        // like we do for attached timelines. (We don't have attached-tenant-scoped usage of VirtualFile
-        // at this point, so, we were able to completely side-step tenant-scoped stuff there).
-        Scope::SecondaryTenant {
-            io_size_metrics: &GLOBAL_IO_SIZE_METRICS,
-        }
-    }
-    #[cfg(test)]
-    pub(crate) fn new_unit_test() -> Self {
-        Scope::UnitTest {
-            io_size_metrics: &GLOBAL_IO_SIZE_METRICS,
-        }
-    }
 }
 
 /// The kind of access to the page cache.
@@ -253,7 +157,6 @@ impl RequestContextBuilder {
                 access_stats_behavior: AccessStatsBehavior::Update,
                 page_content_kind: PageContentKind::Unknown,
                 read_path_debug: false,
-                scope: Scope::new_global(),
             },
         }
     }
@@ -268,16 +171,10 @@ impl RequestContextBuilder {
                 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(),
             },
         }
     }
 
-    pub fn task_kind(mut self, k: TaskKind) -> Self {
-        self.inner.task_kind = k;
-        self
-    }
-
     /// Configure the DownloadBehavior of the context: whether to
     /// download missing layers, and/or warn on the download.
     pub fn download_behavior(mut self, b: DownloadBehavior) -> Self {
@@ -302,11 +199,6 @@ impl RequestContextBuilder {
         self
     }
 
-    pub(crate) fn scope(mut self, s: Scope) -> Self {
-        self.inner.scope = s;
-        self
-    }
-
     pub fn build(self) -> RequestContext {
         self.inner
     }
@@ -389,50 +281,7 @@ impl RequestContext {
     }
 
     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::extend(self)
-            .scope(Scope::new_timeline(timeline))
-            .build()
-    }
-
-    pub(crate) fn with_scope_page_service_pagestream(
-        &self,
-        timeline_handle: &crate::tenant::timeline::handle::Handle<
-            crate::page_service::TenantManagerTypes,
-        >,
-    ) -> Self {
-        RequestContextBuilder::extend(self)
-            .scope(Scope::new_page_service_pagestream(timeline_handle))
-            .build()
-    }
-
-    pub fn with_scope_secondary_timeline(
-        &self,
-        tenant_shard_id: &TenantShardId,
-        timeline_id: &TimelineId,
-    ) -> Self {
-        RequestContextBuilder::extend(self)
-            .scope(Scope::new_secondary_timeline(tenant_shard_id, timeline_id))
-            .build()
-    }
-
-    pub fn with_scope_secondary_tenant(&self, tenant_shard_id: &TenantShardId) -> Self {
-        RequestContextBuilder::extend(self)
-            .scope(Scope::new_secondary_tenant(tenant_shard_id))
-            .build()
-    }
-
-    #[cfg(test)]
-    pub fn with_scope_unit_test(&self) -> Self {
-        RequestContextBuilder::new(TaskKind::UnitTest)
-            .scope(Scope::new_unit_test())
-            .build()
+        Self::new(task_kind, download_behavior)
     }
 
     pub fn task_kind(&self) -> TaskKind {
@@ -454,38 +303,4 @@ impl RequestContext {
     pub(crate) fn read_path_debug(&self) -> bool {
         self.read_path_debug
     }
-
-    pub(crate) fn io_size_metrics(&self) -> &StorageIoSizeMetrics {
-        match &self.scope {
-            Scope::Global { io_size_metrics } => {
-                let is_unit_test = cfg!(test);
-                let is_regress_test_build = cfg!(feature = "testing");
-                if is_unit_test || is_regress_test_build {
-                    panic!("all VirtualFile instances are timeline-scoped");
-                } else {
-                    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(),
-                            "all VirtualFile instances are timeline-scoped",
-                        );
-                    });
-
-                    io_size_metrics
-                }
-            }
-            Scope::Timeline { arc_arc } => &arc_arc.storage_io_size,
-            Scope::SecondaryTimeline { io_size_metrics } => io_size_metrics,
-            Scope::SecondaryTenant { io_size_metrics } => io_size_metrics,
-            #[cfg(test)]
-            Scope::UnitTest { io_size_metrics } => io_size_metrics,
-        }
-    }
 }

pageserver/src/controller_upcall_client.rs

@@ -181,7 +181,7 @@ impl ControlPlaneGenerationsApi for ControllerUpcallClient {
                         listen_pg_port: m.postgres_port,
                         listen_http_addr: m.http_host,
                         listen_http_port: m.http_port,
-                        listen_https_port: m.https_port,
+                        listen_https_port: None, // TODO: Support https.
                         availability_zone_id: az_id.expect("Checked above"),
                     })
                 }

pageserver/src/http/openapi_spec.yml

@@ -842,12 +842,6 @@ paths:
         required: false
         schema:
           type: integer
-      - name: recurse
-        description: When set, will recurse with the downloads into ancestor timelines
-        in: query
-        required: false
-        schema:
-          type: boolean
     post:
       description: |
         Download all layers in the specified timeline's heatmap. The `tenant_shard_id` parameter

pageserver/src/http/routes.rs

@@ -37,8 +37,7 @@ use pageserver_api::models::{
     TenantShardSplitResponse, TenantSorting, TenantState, TenantWaitLsnRequest,
     TimelineArchivalConfigRequest, TimelineCreateRequest, TimelineCreateRequestMode,
     TimelineCreateRequestModeImportPgdata, TimelineGcRequest, TimelineInfo,
-    TimelinePatchIndexPartRequest, TimelinesInfoAndOffloaded, TopTenantShardItem,
-    TopTenantShardsRequest, TopTenantShardsResponse,
+    TimelinesInfoAndOffloaded, TopTenantShardItem, TopTenantShardsRequest, TopTenantShardsResponse,
 };
 use pageserver_api::shard::{ShardCount, TenantShardId};
 use remote_storage::{DownloadError, GenericRemoteStorage, TimeTravelError};
@@ -55,7 +54,6 @@ use utils::id::{TenantId, TimelineId};
 use utils::lsn::Lsn;
 
 use crate::config::PageServerConf;
-use crate::context;
 use crate::context::{DownloadBehavior, RequestContext, RequestContextBuilder};
 use crate::deletion_queue::DeletionQueueClient;
 use crate::pgdatadir_mapping::LsnForTimestamp;
@@ -65,7 +63,6 @@ use crate::tenant::mgr::{
     GetActiveTenantError, GetTenantError, TenantManager, TenantMapError, TenantMapInsertError,
     TenantSlot, TenantSlotError, TenantSlotUpsertError, TenantStateError, UpsertLocationError,
 };
-use crate::tenant::remote_timeline_client::index::GcCompactionState;
 use crate::tenant::remote_timeline_client::{
     download_index_part, list_remote_tenant_shards, list_remote_timelines,
 };
@@ -484,7 +481,6 @@ async fn build_timeline_info_common(
 
         state,
         is_archived: Some(is_archived),
-        rel_size_migration: Some(timeline.get_rel_size_v2_status()),
 
         walreceiver_status,
     };
@@ -861,75 +857,6 @@ async fn timeline_archival_config_handler(
     json_response(StatusCode::OK, ())
 }
 
-/// This API is used to patch the index part of a timeline. You must ensure such patches are safe to apply. Use this API as an emergency
-/// measure only.
-///
-/// Some examples of safe patches:
-/// - Increase the gc_cutoff and gc_compaction_cutoff to a larger value in case of a bug that didn't bump the cutoff and cause read errors.
-/// - Force set the index part to use reldir v2 (migrating/migrated).
-///
-/// Some examples of unsafe patches:
-/// - Force set the index part from v2 to v1 (legacy). This will cause the code path to ignore anything written to the new keyspace and cause
-///   errors.
-/// - Decrease the gc_cutoff without validating the data really exists. It will cause read errors in the background.
-async fn timeline_patch_index_part_handler(
-    mut 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")?;
-
-    let request_data: TimelinePatchIndexPartRequest = json_request(&mut request).await?;
-    check_permission(&request, None)?; // require global permission for this request
-    let state = get_state(&request);
-
-    async {
-        let timeline =
-            active_timeline_of_active_tenant(&state.tenant_manager, tenant_shard_id, timeline_id)
-                .await?;
-
-        if let Some(rel_size_migration) = request_data.rel_size_migration {
-            timeline
-                .update_rel_size_v2_status(rel_size_migration)
-                .map_err(ApiError::InternalServerError)?;
-        }
-
-        if let Some(gc_compaction_last_completed_lsn) =
-            request_data.gc_compaction_last_completed_lsn
-        {
-            timeline
-                .update_gc_compaction_state(GcCompactionState {
-                    last_completed_lsn: gc_compaction_last_completed_lsn,
-                })
-                .map_err(ApiError::InternalServerError)?;
-        }
-
-        if let Some(applied_gc_cutoff_lsn) = request_data.applied_gc_cutoff_lsn {
-            {
-                let guard = timeline.applied_gc_cutoff_lsn.lock_for_write();
-                guard.store_and_unlock(applied_gc_cutoff_lsn);
-            }
-        }
-
-        if request_data.force_index_update {
-            timeline
-                .remote_client
-                .force_schedule_index_upload()
-                .context("force schedule index upload")
-                .map_err(ApiError::InternalServerError)?;
-        }
-
-        Ok::<_, ApiError>(())
-    }
-    .instrument(info_span!("timeline_patch_index_part",
-                tenant_id = %tenant_shard_id.tenant_id,
-                shard_id = %tenant_shard_id.shard_slug(),
-                %timeline_id))
-    .await?;
-
-    json_response(StatusCode::OK, ())
-}
-
 async fn timeline_detail_handler(
     request: Request<Body>,
     _cancel: CancellationToken,
@@ -954,13 +881,12 @@ async fn timeline_detail_handler(
         tenant.wait_to_become_active(ACTIVE_TENANT_TIMEOUT).await?;
 
         let timeline = tenant.get_timeline(timeline_id, false)?;
-        let ctx = &ctx.with_scope_timeline(&timeline);
 
         let timeline_info = build_timeline_info(
             &timeline,
             include_non_incremental_logical_size.unwrap_or(false),
             force_await_initial_logical_size.unwrap_or(false),
-            ctx,
+            &ctx,
         )
         .await
         .context("get local timeline info")
@@ -1004,8 +930,7 @@ async fn get_lsn_by_timestamp_handler(
     let timeline =
         active_timeline_of_active_tenant(&state.tenant_manager, tenant_shard_id, timeline_id)
             .await?;
-    let ctx = RequestContext::new(TaskKind::MgmtRequest, DownloadBehavior::Download)
-        .with_scope_timeline(&timeline);
+    let ctx = RequestContext::new(TaskKind::MgmtRequest, DownloadBehavior::Download);
     let result = timeline
         .find_lsn_for_timestamp(timestamp_pg, &cancel, &ctx)
         .await?;
@@ -1077,8 +1002,7 @@ async fn get_timestamp_of_lsn_handler(
     let timeline =
         active_timeline_of_active_tenant(&state.tenant_manager, tenant_shard_id, timeline_id)
             .await?;
-    let ctx = RequestContext::new(TaskKind::MgmtRequest, DownloadBehavior::Download)
-        .with_scope_timeline(&timeline);
+    let ctx = RequestContext::new(TaskKind::MgmtRequest, DownloadBehavior::Download);
     let result = timeline.get_timestamp_for_lsn(lsn, &ctx).await?;
 
     match result {
@@ -1433,8 +1357,7 @@ async fn timeline_layer_scan_disposable_keys(
         active_timeline_of_active_tenant(&state.tenant_manager, tenant_shard_id, timeline_id)
             .await?;
 
-    let ctx = RequestContext::new(TaskKind::MgmtRequest, DownloadBehavior::Download)
-        .with_scope_timeline(&timeline);
+    let ctx = RequestContext::new(TaskKind::MgmtRequest, DownloadBehavior::Download);
 
     let guard = timeline.layers.read().await;
     let Some(layer) = guard.try_get_from_key(&layer_name.clone().into()) else {
@@ -1512,7 +1435,6 @@ async fn timeline_download_heatmap_layers_handler(
 
     let desired_concurrency =
         parse_query_param(&request, "concurrency")?.unwrap_or(DEFAULT_CONCURRENCY);
-    let recurse = parse_query_param(&request, "recurse")?.unwrap_or(false);
 
     check_permission(&request, Some(tenant_shard_id.tenant_id))?;
 
@@ -1520,8 +1442,7 @@ async fn timeline_download_heatmap_layers_handler(
     let timeline =
         active_timeline_of_active_tenant(&state.tenant_manager, tenant_shard_id, timeline_id)
             .await?;
-    let ctx = RequestContext::new(TaskKind::MgmtRequest, DownloadBehavior::Download)
-        .with_scope_timeline(&timeline);
+    let ctx = RequestContext::new(TaskKind::MgmtRequest, DownloadBehavior::Download);
 
     let max_concurrency = get_config(&request)
         .remote_storage_config
@@ -1530,7 +1451,9 @@ async fn timeline_download_heatmap_layers_handler(
         .unwrap_or(DEFAULT_MAX_CONCURRENCY);
     let concurrency = std::cmp::min(max_concurrency, desired_concurrency);
 
-    timeline.start_heatmap_layers_download(concurrency, recurse, &ctx)?;
+    timeline
+        .start_heatmap_layers_download(concurrency, &ctx)
+        .await?;
 
     json_response(StatusCode::ACCEPTED, ())
 }
@@ -1569,8 +1492,7 @@ async fn layer_download_handler(
     let timeline =
         active_timeline_of_active_tenant(&state.tenant_manager, tenant_shard_id, timeline_id)
             .await?;
-    let ctx = RequestContext::new(TaskKind::MgmtRequest, DownloadBehavior::Download)
-        .with_scope_timeline(&timeline);
+    let ctx = RequestContext::new(TaskKind::MgmtRequest, DownloadBehavior::Download);
     let downloaded = timeline
         .download_layer(&layer_name, &ctx)
         .await
@@ -2306,8 +2228,8 @@ async fn timeline_compact_handler(
         .unwrap_or(false);
 
     async {
+        let ctx = RequestContext::new(TaskKind::MgmtRequest, DownloadBehavior::Download);
         let timeline = active_timeline_of_active_tenant(&state.tenant_manager, tenant_shard_id, timeline_id).await?;
-        let ctx = RequestContext::new(TaskKind::MgmtRequest, DownloadBehavior::Download).with_scope_timeline(&timeline);
         if scheduled {
             let tenant = state
                 .tenant_manager
@@ -2414,8 +2336,8 @@ async fn timeline_checkpoint_handler(
         parse_query_param::<_, bool>(&request, "wait_until_uploaded")?.unwrap_or(false);
 
     async {
+        let ctx = RequestContext::new(TaskKind::MgmtRequest, DownloadBehavior::Download);
         let timeline = active_timeline_of_active_tenant(&state.tenant_manager, tenant_shard_id, timeline_id).await?;
-        let ctx = RequestContext::new(TaskKind::MgmtRequest, DownloadBehavior::Download).with_scope_timeline(&timeline);
         if wait_until_flushed {
             timeline.freeze_and_flush().await
         } else {
@@ -2470,8 +2392,7 @@ async fn timeline_download_remote_layers_handler_post(
     let timeline =
         active_timeline_of_active_tenant(&state.tenant_manager, tenant_shard_id, timeline_id)
             .await?;
-    let ctx = RequestContext::new(TaskKind::MgmtRequest, DownloadBehavior::Download)
-        .with_scope_timeline(&timeline);
+    let ctx = RequestContext::new(TaskKind::MgmtRequest, DownloadBehavior::Download);
     match timeline.spawn_download_all_remote_layers(body, &ctx).await {
         Ok(st) => json_response(StatusCode::ACCEPTED, st),
         Err(st) => json_response(StatusCode::CONFLICT, st),
@@ -2554,7 +2475,6 @@ async fn timeline_detach_ancestor_handler(
         tracing::info!("all timeline upload queues are drained");
 
         let timeline = tenant.get_timeline(timeline_id, true)?;
-        let ctx = &ctx.with_scope_timeline(&timeline);
 
         let progress = timeline
             .prepare_to_detach_from_ancestor(&tenant, options, ctx)
@@ -2661,9 +2581,8 @@ async fn getpage_at_lsn_handler_inner(
     async {
         let ctx = RequestContext::new(TaskKind::MgmtRequest, DownloadBehavior::Download);
         // Enable read path debugging
+        let ctx = RequestContextBuilder::extend(&ctx).read_path_debug(true).build();
         let timeline = active_timeline_of_active_tenant(&state.tenant_manager, tenant_shard_id, timeline_id).await?;
-        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());
@@ -2697,8 +2616,8 @@ async fn timeline_collect_keyspace(
     let at_lsn: Option<Lsn> = parse_query_param(&request, "at_lsn")?;
 
     async {
+        let ctx = RequestContext::new(TaskKind::MgmtRequest, DownloadBehavior::Download);
         let timeline = active_timeline_of_active_tenant(&state.tenant_manager, tenant_shard_id, timeline_id).await?;
-        let ctx = RequestContext::new(TaskKind::MgmtRequest, DownloadBehavior::Download).with_scope_timeline(&timeline);
         let at_lsn = at_lsn.unwrap_or_else(|| timeline.get_last_record_lsn());
         let (dense_ks, sparse_ks) = timeline
             .collect_keyspace(at_lsn, &ctx)
@@ -3223,7 +3142,6 @@ async fn post_top_tenants(
         match order_by {
             TenantSorting::ResidentSize => sizes.resident_size,
             TenantSorting::MaxLogicalSize => sizes.max_logical_size,
-            TenantSorting::MaxLogicalSizePerShard => sizes.max_logical_size_per_shard,
         }
     }
 
@@ -3336,7 +3254,7 @@ async fn put_tenant_timeline_import_basebackup(
 
         tenant.wait_to_become_active(ACTIVE_TENANT_TIMEOUT).await?;
 
-        let (timeline, timeline_ctx) = tenant
+        let timeline = tenant
             .create_empty_timeline(timeline_id, base_lsn, pg_version, &ctx)
             .map_err(ApiError::InternalServerError)
             .await?;
@@ -3355,13 +3273,7 @@ async fn put_tenant_timeline_import_basebackup(
         info!("importing basebackup");
 
         timeline
-            .import_basebackup_from_tar(
-                tenant.clone(),
-                &mut body,
-                base_lsn,
-                broker_client,
-                &timeline_ctx,
-            )
+            .import_basebackup_from_tar(tenant.clone(), &mut body, base_lsn, broker_client, &ctx)
             .await
             .map_err(ApiError::InternalServerError)?;
 
@@ -3401,7 +3313,6 @@ async fn put_tenant_timeline_import_wal(
         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::extend(&ctx).scope(context::Scope::new_timeline(&timeline)).build();
 
         let mut body = StreamReader::new(request.into_body().map(|res| {
             res.map_err(|error| {
@@ -3718,10 +3629,6 @@ pub fn make_router(
             "/v1/tenant/:tenant_shard_id/timeline/:timeline_id/get_timestamp_of_lsn",
             |r| api_handler(r, get_timestamp_of_lsn_handler),
         )
-        .post(
-            "/v1/tenant/:tenant_shard_id/timeline/:timeline_id/patch_index_part",
-            |r| api_handler(r, timeline_patch_index_part_handler),
-        )
         .post(
             "/v1/tenant/:tenant_shard_id/timeline/:timeline_id/lsn_lease",
             |r| api_handler(r, lsn_lease_handler),

pageserver/src/lib.rs

@@ -64,7 +64,6 @@ pub struct CancellableTask {
     pub cancel: CancellationToken,
 }
 pub struct HttpEndpointListener(pub CancellableTask);
-pub struct HttpsEndpointListener(pub CancellableTask);
 pub struct ConsumptionMetricsTasks(pub CancellableTask);
 pub struct DiskUsageEvictionTask(pub CancellableTask);
 impl CancellableTask {
@@ -78,7 +77,6 @@ impl CancellableTask {
 #[allow(clippy::too_many_arguments)]
 pub async fn shutdown_pageserver(
     http_listener: HttpEndpointListener,
-    https_listener: Option<HttpsEndpointListener>,
     page_service: page_service::Listener,
     consumption_metrics_worker: ConsumptionMetricsTasks,
     disk_usage_eviction_task: Option<DiskUsageEvictionTask>,
@@ -215,15 +213,6 @@ pub async fn shutdown_pageserver(
     )
     .await;
 
-    if let Some(https_listener) = https_listener {
-        timed(
-            https_listener.0.shutdown(),
-            "shutdown https",
-            Duration::from_secs(1),
-        )
-        .await;
-    }
-
     // Shut down the HTTP endpoint last, so that you can still check the server's
     // status while it's shutting down.
     // FIXME: We should probably stop accepting commands like attach/detach earlier.