Fix this slightly differently, by initializing end_pos to start_pos.

Fix LSN in keepalive messages, if no WAL has been sent yet
When a new connection is established to the safekeeper, the 'end_pos' field is initially set to Lsn::INVALID (i.e 0/0). If there is no WAL to send to the client, we send KeepAlive messages with Lsn::INVALID. That confuses the pageserver: it thinks that safekeeper is lagging very much behind the tip of the branch, and will reconnect to a different safekeeper. Then the same thing happens with the new safekeeper, until some WAL is streamed which sets 'end_pos' to a valid value. To fix, use 'start_pos' rather than 'end_pos' in the keepalive messages. When the safekeeper has sent all the WAL it has available, they are equal. When the safekeeper has some WAL to send, it will send an XLogData message rather than KeepAlive. If it did send a KeepAlive even when there was some WAL to send too, I think 'start_pos' was a more correct value anyway. Fixes https://github.com/neondatabase/neon/issues/3972
2026-02-12 23:20:37 +00:00 · 2023-05-03 20:53:31 +03:00 · 2023-05-03 17:06:18 +03:00
167 changed files with 4143 additions and 4316 deletions
--- a/.github/actions/allure-report-generate/action.yml
+++ b/.github/actions/allure-report-generate/action.yml
@@ -1,184 +0,0 @@
-name: 'Create Allure report'
-description: 'Generate Allure report from uploaded by actions/allure-report-store tests results'
-
-outputs:
-  report-url:
-    description: 'Allure report URL'
-    value: ${{ steps.generate-report.outputs.report-url }}
-  report-json-url:
-    description: 'Allure report JSON URL'
-    value: ${{ steps.generate-report.outputs.report-json-url }}
-
-runs:
-  using: "composite"
-
-  steps:
-    # We're using some of env variables quite offen, so let's set them once.
-    #
-    # It would be nice to have them set in common runs.env[0] section, but it doesn't work[1]
-    #
-    # - [0] https://docs.github.com/en/actions/creating-actions/metadata-syntax-for-github-actions#runsenv
-    # - [1] https://github.com/neondatabase/neon/pull/3907#discussion_r1154703456
-    #
-    - name: Set variables
-      shell: bash -euxo pipefail {0}
-      run: |
-        PR_NUMBER=$(jq --raw-output .pull_request.number "$GITHUB_EVENT_PATH" || true)
-        if [ "${PR_NUMBER}" != "null" ]; then
-          BRANCH_OR_PR=pr-${PR_NUMBER}
-        elif [ "${GITHUB_REF_NAME}" = "main" ] || [ "${GITHUB_REF_NAME}" = "release" ]; then
-          # Shortcut for special branches
-          BRANCH_OR_PR=${GITHUB_REF_NAME}
-        else
-          BRANCH_OR_PR=branch-$(printf "${GITHUB_REF_NAME}" | tr -c "[:alnum:]._-" "-")
-        fi
-
-        LOCK_FILE=reports/${BRANCH_OR_PR}/lock.txt
-
-        WORKDIR=/tmp/${BRANCH_OR_PR}-$(date +%s)
-        mkdir -p ${WORKDIR}
-
-        echo "BRANCH_OR_PR=${BRANCH_OR_PR}" >> $GITHUB_ENV
-        echo "LOCK_FILE=${LOCK_FILE}"       >> $GITHUB_ENV
-        echo "WORKDIR=${WORKDIR}"           >> $GITHUB_ENV
-        echo "BUCKET=${BUCKET}"             >> $GITHUB_ENV
-      env:
-        BUCKET: neon-github-public-dev
-
-    # TODO: We can replace with a special docker image with Java and Allure pre-installed
-    - uses: actions/setup-java@v3
-      with:
-        distribution: 'temurin'
-        java-version: '17'
-
-    - name: Install Allure
-      shell: bash -euxo pipefail {0}
-      run: |
-        if ! which allure; then
-          ALLURE_ZIP=allure-${ALLURE_VERSION}.zip
-          wget -q https://github.com/allure-framework/allure2/releases/download/${ALLURE_VERSION}/${ALLURE_ZIP}
-          echo "${ALLURE_ZIP_MD5}  ${ALLURE_ZIP}" | md5sum -c
-          unzip -q ${ALLURE_ZIP}
-          echo "$(pwd)/allure-${ALLURE_VERSION}/bin" >> $GITHUB_PATH
-          rm -f ${ALLURE_ZIP}
-        fi
-      env:
-        ALLURE_VERSION: 2.22.0
-        ALLURE_ZIP_MD5: d5c9f0989b896482536956340a7d5ec9
-
-    # Potentially we could have several running build for the same key (for example, for the main branch), so we use improvised lock for this
-    - name: Acquire lock
-      shell: bash -euxo pipefail {0}
-      run: |
-        LOCK_TIMEOUT=300 # seconds
-
-        LOCK_CONTENT="${GITHUB_RUN_ID}-${GITHUB_RUN_ATTEMPT}"
-        echo ${LOCK_CONTENT} > ${WORKDIR}/lock.txt
-
-        # Do it up to 5 times to avoid race condition
-        for _ in $(seq 1 5); do
-          for i in $(seq 1 ${LOCK_TIMEOUT}); do
-            LOCK_ACQUIRED=$(aws s3api head-object --bucket neon-github-public-dev --key ${LOCK_FILE} | jq --raw-output '.LastModified' || true)
-            # `date --date="..."` is supported only by gnu date (i.e. it doesn't work on BSD/macOS)
-            if [ -z "${LOCK_ACQUIRED}" ] || [ "$(( $(date +%s) - $(date --date="${LOCK_ACQUIRED}" +%s) ))" -gt "${LOCK_TIMEOUT}" ]; then
-              break
-            fi
-            sleep 1
-          done
-
-          aws s3 mv --only-show-errors ${WORKDIR}/lock.txt "s3://${BUCKET}/${LOCK_FILE}"
-
-          # Double-check that exactly THIS run has acquired the lock
-          aws s3 cp --only-show-errors "s3://${BUCKET}/${LOCK_FILE}" ./lock.txt
-          if [ "$(cat lock.txt)" = "${LOCK_CONTENT}" ]; then
-            break
-          fi
-        done
-
-    - name: Generate and publish final Allure report
-      id: generate-report
-      shell: bash -euxo pipefail {0}
-      run: |
-        REPORT_PREFIX=reports/${BRANCH_OR_PR}
-        RAW_PREFIX=reports-raw/${BRANCH_OR_PR}/${GITHUB_RUN_ID}
-
-        # Get previously uploaded data for this run
-        ZSTD_NBTHREADS=0
-
-        S3_FILEPATHS=$(aws s3api list-objects-v2 --bucket ${BUCKET} --prefix ${RAW_PREFIX}/ | jq --raw-output  '.Contents[].Key')
-        if [ -z "$S3_FILEPATHS" ]; then
-          # There's no previously uploaded data for this $GITHUB_RUN_ID
-          exit 0
-        fi
-        for S3_FILEPATH in ${S3_FILEPATHS}; do
-          time aws s3 cp --only-show-errors "s3://${BUCKET}/${S3_FILEPATH}" "${WORKDIR}"
-
-          archive=${WORKDIR}/$(basename $S3_FILEPATH)
-          mkdir -p ${archive%.tar.zst}
-          time tar -xf ${archive} -C ${archive%.tar.zst}
-          rm -f ${archive}
-        done
-
-        # Get history trend
-        time aws s3 cp --recursive --only-show-errors "s3://${BUCKET}/${REPORT_PREFIX}/latest/history" "${WORKDIR}/latest/history" || true
-
-        # Generate report
-        time allure generate --clean --output ${WORKDIR}/report ${WORKDIR}/*
-
-        # Replace a logo link with a redirect to the latest version of the report
-        sed -i 's|<a href="." class=|<a href="https://'${BUCKET}'.s3.amazonaws.com/'${REPORT_PREFIX}'/latest/index.html?nocache='"'+Date.now()+'"'" class=|g' ${WORKDIR}/report/app.js
-
-        # Upload a history and the final report (in this particular order to not to have duplicated history in 2 places)
-        time aws s3 mv --recursive --only-show-errors "${WORKDIR}/report/history" "s3://${BUCKET}/${REPORT_PREFIX}/latest/history"
-        time aws s3 mv --recursive --only-show-errors "${WORKDIR}/report" "s3://${BUCKET}/${REPORT_PREFIX}/${GITHUB_RUN_ID}"
-
-        REPORT_URL=https://${BUCKET}.s3.amazonaws.com/${REPORT_PREFIX}/${GITHUB_RUN_ID}/index.html
-
-        # Generate redirect
-        cat <<EOF > ${WORKDIR}/index.html
-          <!DOCTYPE html>
-
-          <meta charset="utf-8">
-          <title>Redirecting to ${REPORT_URL}</title>
-          <meta http-equiv="refresh" content="0; URL=${REPORT_URL}">
-        EOF
-        time aws s3 cp --only-show-errors ${WORKDIR}/index.html "s3://${BUCKET}/${REPORT_PREFIX}/latest/index.html"
-
-        echo "report-url=${REPORT_URL}"                                   >> $GITHUB_OUTPUT
-        echo "report-json-url=${REPORT_URL%/index.html}/data/suites.json" >> $GITHUB_OUTPUT
-
-    - name: Release lock
-      if: always()
-      shell: bash -euxo pipefail {0}
-      run: |
-        aws s3 cp --only-show-errors "s3://${BUCKET}/${LOCK_FILE}" ./lock.txt || exit 0
-
-        if [ "$(cat lock.txt)" = "${GITHUB_RUN_ID}-${GITHUB_RUN_ATTEMPT}" ]; then
-          aws s3 rm "s3://${BUCKET}/${LOCK_FILE}"
-        fi
-
-    - name: Cleanup
-      if: always()
-      shell: bash -euxo pipefail {0}
-      run: |
-        if [ -d "${WORKDIR}" ]; then
-          rm -rf ${WORKDIR}
-        fi
-
-    - uses: actions/github-script@v6
-      if: always()
-      env:
-        REPORT_URL: ${{ steps.generate-report.outputs.report-url }}
-        COMMIT_SHA: ${{ github.event.pull_request.head.sha || github.sha }}
-      with:
-        script: |
-          const { REPORT_URL, COMMIT_SHA } = process.env
-
-          await github.rest.repos.createCommitStatus({
-            owner: context.repo.owner,
-            repo: context.repo.repo,
-            sha: `${COMMIT_SHA}`,
-            state: 'success',
-            target_url: `${REPORT_URL}`,
-            context: 'Allure report',
-          })
--- a/.github/actions/allure-report-store/action.yml
+++ b/.github/actions/allure-report-store/action.yml
@@ -1,72 +0,0 @@
-name: 'Store Allure results'
-description: 'Upload test results to be used by actions/allure-report-generate'
-
-inputs:
-  report-dir:
-    description: 'directory with test results generated by tests'
-    required: true
-  unique-key:
-    description: 'string to distinguish different results in the same run'
-    required: true
-
-runs:
-  using: "composite"
-
-  steps:
-    - name: Set variables
-      shell: bash -euxo pipefail {0}
-      run: |
-        PR_NUMBER=$(jq --raw-output .pull_request.number "$GITHUB_EVENT_PATH" || true)
-        if [ "${PR_NUMBER}" != "null" ]; then
-          BRANCH_OR_PR=pr-${PR_NUMBER}
-        elif [ "${GITHUB_REF_NAME}" = "main" ] || [ "${GITHUB_REF_NAME}" = "release" ]; then
-          # Shortcut for special branches
-          BRANCH_OR_PR=${GITHUB_REF_NAME}
-        else
-          BRANCH_OR_PR=branch-$(printf "${GITHUB_REF_NAME}" | tr -c "[:alnum:]._-" "-")
-        fi
-
-        echo "BRANCH_OR_PR=${BRANCH_OR_PR}" >> $GITHUB_ENV
-        echo "REPORT_DIR=${REPORT_DIR}"     >> $GITHUB_ENV
-      env:
-        REPORT_DIR: ${{ inputs.report-dir }}
-
-    - name: Upload test results
-      shell: bash -euxo pipefail {0}
-      run: |
-        REPORT_PREFIX=reports/${BRANCH_OR_PR}
-        RAW_PREFIX=reports-raw/${BRANCH_OR_PR}/${GITHUB_RUN_ID}
-
-        # Add metadata
-        cat <<EOF > ${REPORT_DIR}/executor.json
-          {
-            "name": "GitHub Actions",
-            "type": "github",
-            "url": "https://${BUCKET}.s3.amazonaws.com/${REPORT_PREFIX}/latest/index.html",
-            "buildOrder": ${GITHUB_RUN_ID},
-            "buildName": "GitHub Actions Run #${GITHUB_RUN_NUMBER}/${GITHUB_RUN_ATTEMPT}",
-            "buildUrl": "${GITHUB_SERVER_URL}/${GITHUB_REPOSITORY}/actions/runs/${GITHUB_RUN_ID}/attempts/${GITHUB_RUN_ATTEMPT}",
-            "reportUrl": "https://${BUCKET}.s3.amazonaws.com/${REPORT_PREFIX}/${GITHUB_RUN_ID}/index.html",
-            "reportName": "Allure Report"
-          }
-        EOF
-
-        cat <<EOF > ${REPORT_DIR}/environment.properties
-          COMMIT_SHA=${COMMIT_SHA}
-        EOF
-
-        ARCHIVE="${UNIQUE_KEY}-${GITHUB_RUN_ATTEMPT}-$(date +%s).tar.zst"
-        ZSTD_NBTHREADS=0
-
-        time tar -C ${REPORT_DIR} -cf ${ARCHIVE} --zstd .
-        time aws s3 mv --only-show-errors ${ARCHIVE} "s3://${BUCKET}/${RAW_PREFIX}/${ARCHIVE}"
-      env:
-        UNIQUE_KEY: ${{ inputs.unique-key }}
-        COMMIT_SHA: ${{ github.event.pull_request.head.sha || github.sha }}
-        BUCKET: neon-github-public-dev
-
-    - name: Cleanup
-      if: always()
-      shell: bash -euxo pipefail {0}
-      run: |
-        rm -rf ${REPORT_DIR}
--- a/.github/actions/allure-report/action.yml
+++ b/.github/actions/allure-report/action.yml
@@ -0,0 +1,254 @@
+name: 'Create Allure report'
+description: 'Create and publish Allure report'
+
+inputs:
+  action:
+    desctiption: 'generate or store'
+    required: true
+  build_type:
+    description: '`build_type` from run-python-test-set action'
+    required: true
+  test_selection:
+    description: '`test_selector` from run-python-test-set action'
+    required: false
+outputs:
+  report-url:
+    description: 'Allure report URL'
+    value: ${{ steps.generate-report.outputs.report-url }}
+  report-json-url:
+    description: 'Allure report JSON URL'
+    value: ${{ steps.generate-report.outputs.report-json-url }}
+
+runs:
+  using: "composite"
+
+  steps:
+    # We're using some of env variables quite offen, so let's set them once.
+    #
+    # It would be nice to have them set in common runs.env[0] section, but it doesn't work[1]
+    #
+    # - [0] https://docs.github.com/en/actions/creating-actions/metadata-syntax-for-github-actions#runsenv
+    # - [1] https://github.com/neondatabase/neon/pull/3907#discussion_r1154703456
+    #
+    - name: Set common environment variables
+      shell: bash -euxo pipefail {0}
+      run: |
+        echo "BUILD_TYPE=${BUILD_TYPE}"   >> $GITHUB_ENV
+        echo "BUCKET=${BUCKET}"           >> $GITHUB_ENV
+        echo "TEST_OUTPUT=${TEST_OUTPUT}" >> $GITHUB_ENV
+      env:
+        BUILD_TYPE: ${{ inputs.build_type }}
+        BUCKET: neon-github-public-dev
+        TEST_OUTPUT: /tmp/test_output
+
+    - name: Validate input parameters
+      shell: bash -euxo pipefail {0}
+      run: |
+        if [ "${{ inputs.action }}" != "store" ] && [ "${{ inputs.action }}" != "generate" ]; then
+          echo >&2 "Unknown inputs.action type '${{ inputs.action }}'; allowed 'generate' or 'store' only"
+          exit 1
+        fi
+
+        if [ -z "${{ inputs.test_selection }}" ] && [ "${{ inputs.action }}" == "store" ]; then
+          echo >&2 "inputs.test_selection must be set for 'store' action"
+          exit 2
+        fi
+
+    - name: Calculate variables
+      id: calculate-vars
+      shell: bash -euxo pipefail {0}
+      run: |
+        # TODO: for manually triggered workflows (via workflow_dispatch) we need to have a separate key
+
+        pr_number=$(jq --raw-output .pull_request.number "$GITHUB_EVENT_PATH" || true)
+        if [ "${pr_number}" != "null" ]; then
+          key=pr-${pr_number}
+        elif [ "${GITHUB_REF_NAME}" = "main" ]; then
+          # Shortcut for a special branch
+          key=main
+        elif [ "${GITHUB_REF_NAME}" = "release" ]; then
+          # Shortcut for a special branch
+          key=release
+        else
+          key=branch-$(printf "${GITHUB_REF_NAME}" | tr -c "[:alnum:]._-" "-")
+        fi
+        echo "KEY=${key}" >> $GITHUB_OUTPUT
+
+        # Sanitize test selection to remove `/` and any other special characters
+        # Use printf instead of echo to avoid having `\n` at the end of the string
+        test_selection=$(printf "${{ inputs.test_selection }}" | tr -c "[:alnum:]._-" "-" )
+        echo "TEST_SELECTION=${test_selection}" >> $GITHUB_OUTPUT
+
+    - uses: actions/setup-java@v3
+      if: ${{ inputs.action == 'generate' }}
+      with:
+        distribution: 'temurin'
+        java-version: '17'
+
+    - name: Install Allure
+      if: ${{ inputs.action == 'generate' }}
+      shell: bash -euxo pipefail {0}
+      run: |
+        if ! which allure; then
+          ALLURE_ZIP=allure-${ALLURE_VERSION}.zip
+          wget -q https://github.com/allure-framework/allure2/releases/download/${ALLURE_VERSION}/${ALLURE_ZIP}
+          echo "${ALLURE_ZIP_MD5}  ${ALLURE_ZIP}" | md5sum -c
+          unzip -q ${ALLURE_ZIP}
+          echo "$(pwd)/allure-${ALLURE_VERSION}/bin" >> $GITHUB_PATH
+          rm -f ${ALLURE_ZIP}
+        fi
+      env:
+        ALLURE_VERSION: 2.21.0
+        ALLURE_ZIP_MD5: c8db4dd8e2a7882583d569ed2c82879c
+
+    - name: Upload Allure results
+      if: ${{ inputs.action == 'store' }}
+      env:
+        REPORT_PREFIX: reports/${{ steps.calculate-vars.outputs.KEY }}/${{ inputs.build_type }}
+        RAW_PREFIX: reports-raw/${{ steps.calculate-vars.outputs.KEY }}/${{ inputs.build_type }}
+        TEST_SELECTION: ${{ steps.calculate-vars.outputs.TEST_SELECTION }}
+      shell: bash -euxo pipefail {0}
+      run: |
+        # Add metadata
+        cat <<EOF > $TEST_OUTPUT/allure/results/executor.json
+          {
+            "name": "GitHub Actions",
+            "type": "github",
+            "url": "https://${BUCKET}.s3.amazonaws.com/${REPORT_PREFIX}/latest/index.html",
+            "buildOrder": ${GITHUB_RUN_ID},
+            "buildName": "GitHub Actions Run #${{ github.run_number }}/${GITHUB_RUN_ATTEMPT}",
+            "buildUrl": "${GITHUB_SERVER_URL}/${GITHUB_REPOSITORY}/actions/runs/${GITHUB_RUN_ID}/attempts/${GITHUB_RUN_ATTEMPT}",
+            "reportUrl": "https://${BUCKET}.s3.amazonaws.com/${REPORT_PREFIX}/${GITHUB_RUN_ID}/index.html",
+            "reportName": "Allure Report"
+          }
+        EOF
+        cat <<EOF > $TEST_OUTPUT/allure/results/environment.properties
+          TEST_SELECTION=${{ inputs.test_selection }}
+          BUILD_TYPE=${BUILD_TYPE}
+        EOF
+
+        ARCHIVE="${GITHUB_RUN_ID}-${TEST_SELECTION}-${GITHUB_RUN_ATTEMPT}-$(date +%s).tar.zst"
+        ZSTD_NBTHREADS=0
+
+        tar -C ${TEST_OUTPUT}/allure/results -cf ${ARCHIVE} --zstd .
+        aws s3 mv --only-show-errors ${ARCHIVE} "s3://${BUCKET}/${RAW_PREFIX}/${ARCHIVE}"
+
+    # Potentially we could have several running build for the same key (for example for the main branch), so we use improvised lock for this
+    - name: Acquire Allure lock
+      if: ${{ inputs.action == 'generate' }}
+      shell: bash -euxo pipefail {0}
+      env:
+        LOCK_FILE: reports/${{ steps.calculate-vars.outputs.KEY }}/lock.txt
+        TEST_SELECTION: ${{ steps.calculate-vars.outputs.TEST_SELECTION }}
+      run: |
+        LOCK_TIMEOUT=300 # seconds
+
+        for _ in $(seq 1 5); do
+          for i in $(seq 1 ${LOCK_TIMEOUT}); do
+            LOCK_ADDED=$(aws s3api head-object --bucket neon-github-public-dev --key ${LOCK_FILE} | jq --raw-output '.LastModified' || true)
+            # `date --date="..."` is supported only by gnu date (i.e. it doesn't work on BSD/macOS)
+            if [ -z "${LOCK_ADDED}" ] || [ "$(( $(date +%s) - $(date --date="${LOCK_ADDED}" +%s) ))" -gt "${LOCK_TIMEOUT}" ]; then
+              break
+            fi
+            sleep 1
+          done
+          echo "${GITHUB_RUN_ID}-${GITHUB_RUN_ATTEMPT}-${TEST_SELECTION}" > lock.txt
+          aws s3 mv --only-show-errors lock.txt "s3://${BUCKET}/${LOCK_FILE}"
+
+          # A double-check that exactly WE have acquired the lock
+          aws s3 cp --only-show-errors "s3://${BUCKET}/${LOCK_FILE}" ./lock.txt
+          if [ "$(cat lock.txt)" = "${GITHUB_RUN_ID}-${GITHUB_RUN_ATTEMPT}-${TEST_SELECTION}" ]; then
+            break
+          fi
+        done
+
+    - name: Generate and publish final Allure report
+      if: ${{ inputs.action == 'generate' }}
+      id: generate-report
+      env:
+        REPORT_PREFIX: reports/${{ steps.calculate-vars.outputs.KEY }}/${{ inputs.build_type }}
+        RAW_PREFIX: reports-raw/${{ steps.calculate-vars.outputs.KEY }}/${{ inputs.build_type }}
+      shell: bash -euxo pipefail {0}
+      run: |
+        # Get previously uploaded data for this run
+        ZSTD_NBTHREADS=0
+
+        s3_filepaths=$(aws s3api list-objects-v2 --bucket ${BUCKET} --prefix ${RAW_PREFIX}/${GITHUB_RUN_ID}- | jq --raw-output  '.Contents[].Key')
+        if [ -z "$s3_filepaths" ]; then
+          # There's no previously uploaded data for this run
+          exit 0
+        fi
+        for s3_filepath in ${s3_filepaths}; do
+          aws s3 cp --only-show-errors "s3://${BUCKET}/${s3_filepath}" "${TEST_OUTPUT}/allure/"
+
+          archive=${TEST_OUTPUT}/allure/$(basename $s3_filepath)
+          mkdir -p ${archive%.tar.zst}
+          tar -xf ${archive} -C ${archive%.tar.zst}
+          rm -f ${archive}
+        done
+
+        # Get history trend
+        aws s3 cp --recursive --only-show-errors "s3://${BUCKET}/${REPORT_PREFIX}/latest/history" "${TEST_OUTPUT}/allure/latest/history" || true
+
+        # Generate report
+        allure generate --clean --output $TEST_OUTPUT/allure/report $TEST_OUTPUT/allure/*
+
+        # Replace a logo link with a redirect to the latest version of the report
+        sed -i 's|<a href="." class=|<a href="https://'${BUCKET}'.s3.amazonaws.com/'${REPORT_PREFIX}'/latest/index.html" class=|g' $TEST_OUTPUT/allure/report/app.js
+
+        # Upload a history and the final report (in this particular order to not to have duplicated history in 2 places)
+        aws s3 mv --recursive --only-show-errors "${TEST_OUTPUT}/allure/report/history" "s3://${BUCKET}/${REPORT_PREFIX}/latest/history"
+        aws s3 mv --recursive --only-show-errors "${TEST_OUTPUT}/allure/report" "s3://${BUCKET}/${REPORT_PREFIX}/${GITHUB_RUN_ID}"
+
+        REPORT_URL=https://${BUCKET}.s3.amazonaws.com/${REPORT_PREFIX}/${GITHUB_RUN_ID}/index.html
+
+        # Generate redirect
+        cat <<EOF > ${TEST_OUTPUT}/allure/index.html
+          <!DOCTYPE html>
+
+          <meta charset="utf-8">
+          <title>Redirecting to ${REPORT_URL}</title>
+          <meta http-equiv="refresh" content="0; URL=${REPORT_URL}">
+        EOF
+        aws s3 cp --only-show-errors ${TEST_OUTPUT}/allure/index.html "s3://${BUCKET}/${REPORT_PREFIX}/latest/index.html"
+
+        echo "[Allure Report](${REPORT_URL})" >> ${GITHUB_STEP_SUMMARY}
+        echo "report-url=${REPORT_URL}" >> $GITHUB_OUTPUT
+        echo "report-json-url=${REPORT_URL%/index.html}/data/suites.json" >> $GITHUB_OUTPUT
+
+    - name: Release Allure lock
+      if: ${{ inputs.action == 'generate' && always() }}
+      shell: bash -euxo pipefail {0}
+      env:
+        LOCK_FILE: reports/${{ steps.calculate-vars.outputs.KEY }}/lock.txt
+        TEST_SELECTION: ${{ steps.calculate-vars.outputs.TEST_SELECTION }}
+      run: |
+        aws s3 cp --only-show-errors "s3://${BUCKET}/${LOCK_FILE}" ./lock.txt || exit 0
+
+        if [ "$(cat lock.txt)" = "${GITHUB_RUN_ID}-${GITHUB_RUN_ATTEMPT}-${TEST_SELECTION}" ]; then
+          aws s3 rm "s3://${BUCKET}/${LOCK_FILE}"
+        fi
+
+    - name: Cleanup
+      if: always()
+      shell: bash -euxo pipefail {0}
+      run: |
+        rm -rf ${TEST_OUTPUT}/allure
+
+    - uses: actions/github-script@v6
+      if: ${{ inputs.action == 'generate' && always() }}
+      env:
+        REPORT_URL: ${{ steps.generate-report.outputs.report-url }}
+        SHA: ${{ github.event.pull_request.head.sha || github.sha }}
+      with:
+        script: |
+          const { REPORT_URL, BUILD_TYPE, SHA } = process.env
+
+          await github.rest.repos.createCommitStatus({
+            owner: context.repo.owner,
+            repo: context.repo.repo,
+            sha: `${SHA}`,
+            state: 'success',
+            target_url: `${REPORT_URL}`,
+            context: `Allure report / ${BUILD_TYPE}`,
+          })
--- a/.github/actions/run-python-test-set/action.yml
+++ b/.github/actions/run-python-test-set/action.yml
@@ -197,13 +197,14 @@ runs:
      uses: ./.github/actions/upload
      with:
        name: compatibility-snapshot-${{ inputs.build_type }}-pg14-${{ github.run_id }}
-        # Directory is created by test_compatibility.py::test_create_snapshot, keep the path in sync with the test
-        path: /tmp/test_output/compatibility_snapshot_pg14/
+        # The path includes a test name (test_create_snapshot) and directory that the test creates (compatibility_snapshot_pg14), keep the path in sync with the test
+        path: /tmp/test_output/test_create_snapshot/compatibility_snapshot_pg14/
        prefix: latest

-    - name: Upload test results
+    - name: Create Allure report
      if: ${{ !cancelled() }}
-      uses: ./.github/actions/allure-report-store
+      uses: ./.github/actions/allure-report
      with:
-        report-dir: /tmp/test_output/allure/results
-        unique-key: ${{ inputs.build_type }}
+        action: store
+        build_type: ${{ inputs.build_type }}
+        test_selection: ${{ inputs.test_selection }}
--- a/.github/ansible/.gitignore
+++ b/.github/ansible/.gitignore
@@ -0,0 +1,5 @@
+neon_install.tar.gz
+.neon_current_version
+
+collections/*
+!collections/.keep
--- a/.github/ansible/ansible.cfg
+++ b/.github/ansible/ansible.cfg
@@ -0,0 +1,12 @@
+[defaults]
+
+localhost_warning = False
+host_key_checking = False
+timeout = 30
+
+[ssh_connection]
+ssh_args   = -F ./ansible.ssh.cfg
+# teleport doesn't support sftp yet https://github.com/gravitational/teleport/issues/7127
+# and scp neither worked for me
+transfer_method = piped
+pipelining = True
--- a/.github/ansible/ansible.ssh.cfg
+++ b/.github/ansible/ansible.ssh.cfg
@@ -0,0 +1,15 @@
+# Remove this once https://github.com/gravitational/teleport/issues/10918 is fixed
+# (use pre 8.5 option name to cope with old ssh in CI)
+PubkeyAcceptedKeyTypes +ssh-rsa-cert-v01@openssh.com
+
+Host tele.zenith.tech
+    User admin
+    Port 3023
+    StrictHostKeyChecking no
+    UserKnownHostsFile /dev/null
+
+Host * !tele.zenith.tech
+    User admin
+    StrictHostKeyChecking no
+    UserKnownHostsFile /dev/null
+    ProxyJump tele.zenith.tech
--- a/.github/ansible/collections/.keep
+++ b/.github/ansible/collections/.keep
--- a/.github/ansible/deploy.yaml
+++ b/.github/ansible/deploy.yaml
@@ -0,0 +1,211 @@
+- name: Upload Neon binaries
+  hosts: storage
+  gather_facts: False
+  remote_user: "{{ remote_user }}"
+
+  tasks:
+
+    - name: get latest version of Neon binaries
+      register: current_version_file
+      set_fact:
+        current_version: "{{ lookup('file', '.neon_current_version') | trim }}"
+      tags:
+      - pageserver
+      - safekeeper
+
+    - name: inform about versions
+      debug:
+        msg: "Version to deploy - {{ current_version }}"
+      tags:
+      - pageserver
+      - safekeeper
+
+    - name: upload and extract Neon binaries to /usr/local
+      ansible.builtin.unarchive:
+        owner: root
+        group: root
+        src: neon_install.tar.gz
+        dest: /usr/local
+      become: true
+      tags:
+      - pageserver
+      - safekeeper
+      - binaries
+      - putbinaries
+
+- name: Deploy pageserver
+  hosts: pageservers
+  gather_facts: False
+  remote_user: "{{ remote_user }}"
+
+  tasks:
+
+    - name: upload init script
+      when: console_mgmt_base_url is defined
+      ansible.builtin.template:
+        src: scripts/init_pageserver.sh
+        dest: /tmp/init_pageserver.sh
+        owner: root
+        group: root
+        mode: '0755'
+      become: true
+      tags:
+      - pageserver
+
+    - name: init pageserver
+      shell:
+        cmd: /tmp/init_pageserver.sh
+      args:
+        creates: "/storage/pageserver/data/tenants"
+      environment:
+        NEON_REPO_DIR: "/storage/pageserver/data"
+        LD_LIBRARY_PATH: "/usr/local/v14/lib"
+      become: true
+      tags:
+      - pageserver
+
+    - name: read the existing remote pageserver config
+      ansible.builtin.slurp:
+        src: /storage/pageserver/data/pageserver.toml
+      register: _remote_ps_config
+      tags:
+      - pageserver
+
+    - name: parse the existing pageserver configuration
+      ansible.builtin.set_fact:
+        _existing_ps_config: "{{ _remote_ps_config['content'] | b64decode | sivel.toiletwater.from_toml }}"
+      tags:
+      - pageserver
+
+    - name: construct the final pageserver configuration dict
+      ansible.builtin.set_fact:
+        pageserver_config: "{{ pageserver_config_stub | combine({'id': _existing_ps_config.id }) }}"
+      tags:
+      - pageserver
+
+    - name: template the pageserver config
+      template:
+        src: templates/pageserver.toml.j2
+        dest: /storage/pageserver/data/pageserver.toml
+      become: true
+      tags:
+      - pageserver
+
+    # used in `pageserver.service` template
+    - name: learn current availability_zone
+      shell:
+        cmd: "curl -s http://169.254.169.254/latest/meta-data/placement/availability-zone"
+      register: ec2_availability_zone
+
+    - set_fact: 
+        ec2_availability_zone={{ ec2_availability_zone.stdout }}
+
+    - name: upload systemd service definition
+      ansible.builtin.template:
+        src: systemd/pageserver.service
+        dest: /etc/systemd/system/pageserver.service
+        owner: root
+        group: root
+        mode: '0644'
+      become: true
+      tags:
+      - pageserver
+
+    - name: start systemd service
+      ansible.builtin.systemd:
+        daemon_reload: yes
+        name: pageserver
+        enabled: yes
+        state: restarted
+      become: true
+      tags:
+      - pageserver
+
+    - name: post version to console
+      when: console_mgmt_base_url is defined
+      shell:
+        cmd: |
+          INSTANCE_ID=$(curl -s http://169.254.169.254/latest/meta-data/instance-id)
+          curl -sfS -H "Authorization: Bearer {{ CONSOLE_API_TOKEN }}" {{ console_mgmt_base_url }}/management/api/v2/pageservers/$INSTANCE_ID | jq '.version = {{ current_version }}' > /tmp/new_version
+          curl -sfS -H "Authorization: Bearer {{ CONSOLE_API_TOKEN }}" -H "Content-Type: application/json" -X POST -d@/tmp/new_version {{ console_mgmt_base_url }}/management/api/v2/pageservers
+      tags:
+      - pageserver
+
+- name: Deploy safekeeper
+  hosts: safekeepers
+  gather_facts: False
+  remote_user: "{{ remote_user }}"
+
+  tasks:
+
+    - name: upload init script
+      when: console_mgmt_base_url is defined
+      ansible.builtin.template:
+        src: scripts/init_safekeeper.sh
+        dest: /tmp/init_safekeeper.sh
+        owner: root
+        group: root
+        mode: '0755'
+      become: true
+      tags:
+      - safekeeper
+
+    - name: init safekeeper
+      shell:
+        cmd: /tmp/init_safekeeper.sh
+      args:
+        creates: "/storage/safekeeper/data/safekeeper.id"
+      environment:
+        NEON_REPO_DIR: "/storage/safekeeper/data"
+        LD_LIBRARY_PATH: "/usr/local/v14/lib"
+      become: true
+      tags:
+      - safekeeper
+
+    # used in `safekeeper.service` template
+    - name: learn current availability_zone
+      shell:
+        cmd: "curl -s http://169.254.169.254/latest/meta-data/placement/availability-zone"
+      register: ec2_availability_zone
+
+    - set_fact: 
+        ec2_availability_zone={{ ec2_availability_zone.stdout }}
+
+    # in the future safekeepers should discover pageservers byself
+    # but currently use first pageserver that was discovered
+    - name: set first pageserver var for safekeepers
+      set_fact:
+        first_pageserver: "{{ hostvars[groups['pageservers'][0]]['inventory_hostname'] }}"
+      tags:
+      - safekeeper
+
+    - name: upload systemd service definition
+      ansible.builtin.template:
+        src: systemd/safekeeper.service
+        dest: /etc/systemd/system/safekeeper.service
+        owner: root
+        group: root
+        mode: '0644'
+      become: true
+      tags:
+      - safekeeper
+
+    - name: start systemd service
+      ansible.builtin.systemd:
+        daemon_reload: yes
+        name: safekeeper
+        enabled: yes
+        state: restarted
+      become: true
+      tags:
+      - safekeeper
+
+    - name: post version to console
+      when: console_mgmt_base_url is defined
+      shell:
+        cmd: |
+          INSTANCE_ID=$(curl -s http://169.254.169.254/latest/meta-data/instance-id)
+          curl -sfS -H "Authorization: Bearer {{ CONSOLE_API_TOKEN }}" {{ console_mgmt_base_url }}/management/api/v2/safekeepers/$INSTANCE_ID | jq '.version = {{ current_version }}' > /tmp/new_version
+          curl -sfS -H "Authorization: Bearer {{ CONSOLE_API_TOKEN }}" -H "Content-Type: application/json" -X POST -d@/tmp/new_version {{ console_mgmt_base_url }}/management/api/v2/safekeepers
+      tags:
+      - safekeeper
--- a/.github/ansible/get_binaries.sh
+++ b/.github/ansible/get_binaries.sh
@@ -0,0 +1,42 @@
+#!/bin/bash
+
+set -e
+
+if [ -n "${DOCKER_TAG}" ]; then
+  # Verson is DOCKER_TAG but without prefix
+  VERSION=$(echo $DOCKER_TAG | sed 's/^.*-//g')
+else
+  echo "Please set DOCKER_TAG environment variable"
+  exit 1
+fi
+
+
+# do initial cleanup
+rm -rf neon_install postgres_install.tar.gz neon_install.tar.gz .neon_current_version
+mkdir neon_install
+
+# retrieve binaries from docker image
+echo "getting binaries from docker image"
+docker pull --quiet neondatabase/neon:${DOCKER_TAG}
+ID=$(docker create neondatabase/neon:${DOCKER_TAG})
+docker cp ${ID}:/data/postgres_install.tar.gz .
+tar -xzf postgres_install.tar.gz -C neon_install
+mkdir neon_install/bin/
+docker cp ${ID}:/usr/local/bin/pageserver neon_install/bin/
+docker cp ${ID}:/usr/local/bin/pageserver_binutils neon_install/bin/
+docker cp ${ID}:/usr/local/bin/safekeeper neon_install/bin/
+docker cp ${ID}:/usr/local/bin/storage_broker neon_install/bin/
+docker cp ${ID}:/usr/local/bin/proxy neon_install/bin/
+docker cp ${ID}:/usr/local/v14/bin/ neon_install/v14/bin/
+docker cp ${ID}:/usr/local/v15/bin/ neon_install/v15/bin/
+docker cp ${ID}:/usr/local/v14/lib/ neon_install/v14/lib/
+docker cp ${ID}:/usr/local/v15/lib/ neon_install/v15/lib/
+docker rm -vf ${ID}
+
+# store version to file (for ansible playbooks) and create binaries tarball
+echo ${VERSION} > neon_install/.neon_current_version
+echo ${VERSION} > .neon_current_version
+tar -czf neon_install.tar.gz -C neon_install .
+
+# do final cleaup
+rm -rf neon_install postgres_install.tar.gz
--- a/.github/ansible/prod.ap-southeast-1.hosts.yaml
+++ b/.github/ansible/prod.ap-southeast-1.hosts.yaml
@@ -0,0 +1,48 @@
+storage:
+  vars:
+    bucket_name: neon-prod-storage-ap-southeast-1
+    bucket_region: ap-southeast-1
+    console_mgmt_base_url: http://neon-internal-api.aws.neon.tech
+    broker_endpoint: http://storage-broker-lb.epsilon.ap-southeast-1.internal.aws.neon.tech:50051
+    pageserver_config_stub:
+      pg_distrib_dir: /usr/local
+      metric_collection_endpoint: http://neon-internal-api.aws.neon.tech/billing/api/v1/usage_events
+      metric_collection_interval: 10min
+      disk_usage_based_eviction:
+        max_usage_pct: 85 # TODO: decrease to 80 after all pageservers are below 80
+        min_avail_bytes: 0
+        period: "10s"
+      tenant_config:
+        eviction_policy:
+          kind: "LayerAccessThreshold"
+          period: "10m"
+          threshold: &default_eviction_threshold "24h"
+        evictions_low_residence_duration_metric_threshold: *default_eviction_threshold
+      remote_storage:
+        bucket_name: "{{ bucket_name }}"
+        bucket_region: "{{ bucket_region }}"
+        prefix_in_bucket: "pageserver/v1"
+    safekeeper_s3_prefix: safekeeper/v1/wal
+    hostname_suffix: ""
+    remote_user: ssm-user
+    ansible_aws_ssm_region: ap-southeast-1
+    ansible_aws_ssm_bucket_name: neon-prod-storage-ap-southeast-1
+    console_region_id: aws-ap-southeast-1
+    sentry_environment: production
+
+  children:
+    pageservers:
+      hosts:
+        pageserver-0.ap-southeast-1.aws.neon.tech:
+          ansible_host:  i-064de8ea28bdb495b
+        pageserver-1.ap-southeast-1.aws.neon.tech:
+          ansible_host:  i-0b180defcaeeb6b93
+
+    safekeepers:
+      hosts:
+        safekeeper-0.ap-southeast-1.aws.neon.tech:
+          ansible_host:  i-0d6f1dc5161eef894
+        safekeeper-2.ap-southeast-1.aws.neon.tech:
+          ansible_host:  i-04fb63634e4679eb9
+        safekeeper-3.ap-southeast-1.aws.neon.tech:
+          ansible_host:  i-05481f3bc88cfc2d4
--- a/.github/ansible/prod.eu-central-1.hosts.yaml
+++ b/.github/ansible/prod.eu-central-1.hosts.yaml
@@ -0,0 +1,50 @@
+storage:
+  vars:
+    bucket_name: neon-prod-storage-eu-central-1
+    bucket_region: eu-central-1
+    console_mgmt_base_url: http://neon-internal-api.aws.neon.tech
+    broker_endpoint: http://storage-broker-lb.gamma.eu-central-1.internal.aws.neon.tech:50051
+    pageserver_config_stub:
+      pg_distrib_dir: /usr/local
+      metric_collection_endpoint: http://neon-internal-api.aws.neon.tech/billing/api/v1/usage_events
+      metric_collection_interval: 10min
+      disk_usage_based_eviction:
+        max_usage_pct: 85 # TODO: decrease to 80 after all pageservers are below 80
+        min_avail_bytes: 0
+        period: "10s"
+      tenant_config:
+        eviction_policy:
+          kind: "LayerAccessThreshold"
+          period: "10m"
+          threshold: &default_eviction_threshold "24h"
+        evictions_low_residence_duration_metric_threshold: *default_eviction_threshold
+      remote_storage:
+        bucket_name: "{{ bucket_name }}"
+        bucket_region: "{{ bucket_region }}"
+        prefix_in_bucket: "pageserver/v1"
+    safekeeper_s3_prefix: safekeeper/v1/wal
+    hostname_suffix: ""
+    remote_user: ssm-user
+    ansible_aws_ssm_region: eu-central-1
+    ansible_aws_ssm_bucket_name: neon-prod-storage-eu-central-1
+    console_region_id: aws-eu-central-1
+    sentry_environment: production
+
+  children:
+    pageservers:
+      hosts:
+        pageserver-0.eu-central-1.aws.neon.tech:
+          ansible_host:  i-0cd8d316ecbb715be
+        pageserver-1.eu-central-1.aws.neon.tech:
+          ansible_host:  i-090044ed3d383fef0
+        pageserver-2.eu-central-1.aws.neon.tech:
+          ansible_host:  i-033584edf3f4b6742
+
+    safekeepers:
+      hosts:
+        safekeeper-0.eu-central-1.aws.neon.tech:
+          ansible_host:  i-0b238612d2318a050
+        safekeeper-1.eu-central-1.aws.neon.tech:
+          ansible_host:  i-07b9c45e5c2637cd4
+        safekeeper-2.eu-central-1.aws.neon.tech:
+          ansible_host:  i-020257302c3c93d88
--- a/.github/ansible/prod.us-east-1.hosts.yaml
+++ b/.github/ansible/prod.us-east-1.hosts.yaml
@@ -0,0 +1,50 @@
+storage:
+  vars:
+    bucket_name: neon-prod-storage-us-east-1
+    bucket_region: us-east-1
+    console_mgmt_base_url: http://neon-internal-api.aws.neon.tech
+    broker_endpoint: http://storage-broker-lb.theta.us-east-1.internal.aws.neon.tech:50051
+    pageserver_config_stub:
+      pg_distrib_dir: /usr/local
+      metric_collection_endpoint: http://neon-internal-api.aws.neon.tech/billing/api/v1/usage_events
+      metric_collection_interval: 10min
+      disk_usage_based_eviction:
+        max_usage_pct: 85 # TODO: decrease to 80 after all pageservers are below 80
+        min_avail_bytes: 0
+        period: "10s"
+      tenant_config:
+        eviction_policy:
+          kind: "LayerAccessThreshold"
+          period: "10m"
+          threshold: &default_eviction_threshold "24h"
+        evictions_low_residence_duration_metric_threshold: *default_eviction_threshold
+      remote_storage:
+        bucket_name: "{{ bucket_name }}"
+        bucket_region: "{{ bucket_region }}"
+        prefix_in_bucket: "pageserver/v1"
+    safekeeper_s3_prefix: safekeeper/v1/wal
+    hostname_suffix: ""
+    remote_user: ssm-user
+    ansible_aws_ssm_region: us-east-1
+    ansible_aws_ssm_bucket_name: neon-prod-storage-us-east-1
+    console_region_id: aws-us-east-1
+    sentry_environment: production
+
+  children:
+    pageservers:
+      hosts:
+        pageserver-0.us-east-1.aws.neon.tech:
+          ansible_host: i-085222088b0d2e0c7
+        pageserver-1.us-east-1.aws.neon.tech:
+          ansible_host: i-0969d4f684d23a21e
+        pageserver-2.us-east-1.aws.neon.tech:
+          ansible_host: i-05dee87895da58dad
+
+    safekeepers:
+      hosts:
+        safekeeper-0.us-east-1.aws.neon.tech:
+          ansible_host: i-04ce739e88793d864
+        safekeeper-1.us-east-1.aws.neon.tech:
+          ansible_host: i-0e9e6c9227fb81410
+        safekeeper-2.us-east-1.aws.neon.tech:
+          ansible_host: i-072f4dd86a327d52f
--- a/.github/ansible/prod.us-east-2.hosts.yaml
+++ b/.github/ansible/prod.us-east-2.hosts.yaml
@@ -0,0 +1,51 @@
+storage:
+  vars:
+    bucket_name: neon-prod-storage-us-east-2
+    bucket_region: us-east-2
+    console_mgmt_base_url: http://neon-internal-api.aws.neon.tech
+    broker_endpoint: http://storage-broker-lb.delta.us-east-2.internal.aws.neon.tech:50051
+    pageserver_config_stub:
+      pg_distrib_dir: /usr/local
+      metric_collection_endpoint: http://neon-internal-api.aws.neon.tech/billing/api/v1/usage_events
+      metric_collection_interval: 10min
+      disk_usage_based_eviction:
+        max_usage_pct: 85 # TODO: decrease to 80 after all pageservers are below 80
+        min_avail_bytes: 0
+        period: "10s"
+      tenant_config:
+        eviction_policy:
+          kind: "LayerAccessThreshold"
+          period: "10m"
+          threshold: &default_eviction_threshold "24h"
+        evictions_low_residence_duration_metric_threshold: *default_eviction_threshold
+      remote_storage:
+        bucket_name: "{{ bucket_name }}"
+        bucket_region: "{{ bucket_region }}"
+        prefix_in_bucket: "pageserver/v1"
+    safekeeper_s3_prefix: safekeeper/v1/wal
+    hostname_suffix: ""
+    remote_user: ssm-user
+    ansible_aws_ssm_region: us-east-2
+    ansible_aws_ssm_bucket_name: neon-prod-storage-us-east-2
+    console_region_id: aws-us-east-2
+    sentry_environment: production
+
+  children:
+    pageservers:
+      hosts:
+        pageserver-0.us-east-2.aws.neon.tech:
+          ansible_host:  i-062227ba7f119eb8c
+        pageserver-1.us-east-2.aws.neon.tech:
+          ansible_host:  i-0b3ec0afab5968938
+        pageserver-2.us-east-2.aws.neon.tech:
+          ansible_host:  i-0d7a1c4325e71421d
+
+    safekeepers:
+      hosts:
+        safekeeper-0.us-east-2.aws.neon.tech:
+          ansible_host:  i-0e94224750c57d346
+        safekeeper-1.us-east-2.aws.neon.tech:
+          ansible_host:  i-06d113fb73bfddeb0
+        safekeeper-2.us-east-2.aws.neon.tech:
+          ansible_host:  i-09f66c8e04afff2e8
+
--- a/.github/ansible/prod.us-west-2.hosts.yaml
+++ b/.github/ansible/prod.us-west-2.hosts.yaml
@@ -0,0 +1,72 @@
+storage:
+  vars:
+    bucket_name: neon-prod-storage-us-west-2
+    bucket_region: us-west-2
+    console_mgmt_base_url: http://neon-internal-api.aws.neon.tech
+    broker_endpoint: http://storage-broker-lb.eta.us-west-2.internal.aws.neon.tech:50051
+    pageserver_config_stub:
+      pg_distrib_dir: /usr/local
+      metric_collection_endpoint: http://neon-internal-api.aws.neon.tech/billing/api/v1/usage_events
+      metric_collection_interval: 10min
+      disk_usage_based_eviction:
+        max_usage_pct: 85 # TODO: decrease to 80 after all pageservers are below 80
+        min_avail_bytes: 0
+        period: "10s"
+      tenant_config:
+        eviction_policy:
+          kind: "LayerAccessThreshold"
+          period: "10m"
+          threshold: &default_eviction_threshold "24h"
+        evictions_low_residence_duration_metric_threshold: *default_eviction_threshold
+      remote_storage:
+        bucket_name: "{{ bucket_name }}"
+        bucket_region: "{{ bucket_region }}"
+        prefix_in_bucket: "pageserver/v1"
+    safekeeper_s3_prefix: safekeeper/v1/wal
+    hostname_suffix: ""
+    remote_user: ssm-user
+    ansible_aws_ssm_region: us-west-2
+    ansible_aws_ssm_bucket_name: neon-prod-storage-us-west-2
+    console_region_id: aws-us-west-2-new
+    sentry_environment: production
+
+  children:
+    pageservers:
+      hosts:
+        pageserver-0.us-west-2.aws.neon.tech:
+          ansible_host: i-0d9f6dfae0e1c780d
+        pageserver-1.us-west-2.aws.neon.tech:
+          ansible_host: i-0c834be1dddba8b3f
+        pageserver-2.us-west-2.aws.neon.tech:
+          ansible_host: i-051642d372c0a4f32
+        pageserver-3.us-west-2.aws.neon.tech:
+          ansible_host: i-00c3844beb9ad1c6b
+        pageserver-4.us-west-2.aws.neon.tech:
+          ansible_host: i-013263dd1c239adcc
+        pageserver-5.us-west-2.aws.neon.tech:
+          ansible_host: i-00ca6417c7bf96820
+        pageserver-6.us-west-2.aws.neon.tech:
+          ansible_host: i-01cdf7d2bc1433b6a
+        pageserver-7.us-west-2.aws.neon.tech:
+          ansible_host: i-02eec9b40617db5bc
+
+    safekeepers:
+      hosts:
+        safekeeper-0.us-west-2.aws.neon.tech:
+          ansible_host: i-00719d8a74986fda6
+        safekeeper-1.us-west-2.aws.neon.tech:
+          ansible_host: i-074682f9d3c712e7c
+        safekeeper-2.us-west-2.aws.neon.tech:
+          ansible_host: i-042b7efb1729d7966
+        safekeeper-3.us-west-2.aws.neon.tech:
+          ansible_host: i-089f6b9ef426dff76
+        safekeeper-4.us-west-2.aws.neon.tech:
+          ansible_host: i-0fe6bf912c4710c82
+        safekeeper-5.us-west-2.aws.neon.tech:
+          ansible_host: i-0a83c1c46d2b4e409
+        safekeeper-6.us-west-2.aws.neon.tech:
+          ansible_host: i-0fef5317b8fdc9f8d
+        safekeeper-7.us-west-2.aws.neon.tech:
+          ansible_host: i-0be739190d4289bf9
+        safekeeper-8.us-west-2.aws.neon.tech:
+          ansible_host: i-00e851803669e5cfe                    
--- a/.github/ansible/scripts/init_pageserver.sh
+++ b/.github/ansible/scripts/init_pageserver.sh
@@ -0,0 +1,37 @@
+#!/bin/sh
+
+# fetch params from meta-data service
+INSTANCE_ID=$(curl -s http://169.254.169.254/latest/meta-data/instance-id)
+AZ_ID=$(curl -s http://169.254.169.254/latest/meta-data/placement/availability-zone)
+INSTANCE_TYPE=$(curl -s http://169.254.169.254/latest/meta-data/instance-type)
+DISK_SIZE=$(df -B1 /storage | tail -1 | awk '{print $2}')
+
+# store fqdn hostname in var
+HOST=$(hostname -f)
+
+
+cat <<EOF | tee /tmp/payload
+{
+  "version": 1,
+  "host": "${HOST}",
+  "port": 6400,
+  "region_id": "{{ console_region_id }}",
+  "instance_id": "${INSTANCE_ID}",
+  "http_host": "${HOST}",
+  "http_port": 9898,
+  "active": false,
+  "availability_zone_id": "${AZ_ID}",
+  "disk_size": ${DISK_SIZE},
+  "instance_type": "${INSTANCE_TYPE}"
+}
+EOF
+
+# check if pageserver already registered or not
+if ! curl -sf -H "Authorization: Bearer {{ CONSOLE_API_TOKEN }}" {{ console_mgmt_base_url }}/management/api/v2/pageservers/${INSTANCE_ID} -o /dev/null; then
+
+    # not registered, so register it now
+    ID=$(curl -sf -X POST -H "Authorization: Bearer {{ CONSOLE_API_TOKEN }}" -H "Content-Type: application/json" {{ console_mgmt_base_url }}/management/api/v2/pageservers -d@/tmp/payload | jq -r '.id')
+
+    # init pageserver
+    sudo -u pageserver /usr/local/bin/pageserver -c "id=${ID}" -c "pg_distrib_dir='/usr/local'" --init -D /storage/pageserver/data
+fi
--- a/.github/ansible/scripts/init_safekeeper.sh
+++ b/.github/ansible/scripts/init_safekeeper.sh
@@ -0,0 +1,31 @@
+#!/bin/sh
+
+# fetch params from meta-data service
+INSTANCE_ID=$(curl -s http://169.254.169.254/latest/meta-data/instance-id)
+AZ_ID=$(curl -s http://169.254.169.254/latest/meta-data/placement/availability-zone)
+
+# store fqdn hostname in var
+HOST=$(hostname -f)
+
+
+cat <<EOF | tee /tmp/payload
+{
+  "version": 1,
+  "host": "${HOST}",
+  "port": 6500,
+  "http_port": 7676,
+  "region_id": "{{ console_region_id }}",
+  "instance_id": "${INSTANCE_ID}",
+  "availability_zone_id": "${AZ_ID}",
+  "active": false
+}
+EOF
+
+# check if safekeeper already registered or not
+if ! curl -sf -H "Authorization: Bearer {{ CONSOLE_API_TOKEN }}" {{ console_mgmt_base_url }}/management/api/v2/safekeepers/${INSTANCE_ID} -o /dev/null; then
+
+    # not registered, so register it now
+    ID=$(curl -sf -X POST -H "Authorization: Bearer {{ CONSOLE_API_TOKEN }}" -H "Content-Type: application/json" {{ console_mgmt_base_url }}/management/api/v2/safekeepers -d@/tmp/payload | jq -r '.id')
+    # init safekeeper
+    sudo -u safekeeper /usr/local/bin/safekeeper --id ${ID} --init -D /storage/safekeeper/data
+fi
--- a/.github/ansible/ssm_config
+++ b/.github/ansible/ssm_config
@@ -0,0 +1,2 @@
+ansible_connection: aws_ssm
+ansible_python_interpreter: /usr/bin/python3
--- a/.github/ansible/staging.eu-central-1.hosts.yaml
+++ b/.github/ansible/staging.eu-central-1.hosts.yaml
@@ -0,0 +1,47 @@
+storage:
+  vars:
+    bucket_name: neon-dev-storage-eu-central-1
+    bucket_region: eu-central-1
+    # We only register/update storage in one preview console and manually copy to other instances
+    console_mgmt_base_url: http://neon-internal-api.helium.aws.neon.build
+    broker_endpoint: http://storage-broker-lb.alpha.eu-central-1.internal.aws.neon.build:50051
+    pageserver_config_stub:
+      pg_distrib_dir: /usr/local
+      metric_collection_endpoint: http://neon-internal-api.helium.aws.neon.build/billing/api/v1/usage_events
+      metric_collection_interval: 10min
+      disk_usage_based_eviction:
+        max_usage_pct: 80
+        min_avail_bytes: 0
+        period: "10s"
+      tenant_config:
+        eviction_policy:
+          kind: "LayerAccessThreshold"
+          period: "20m"
+          threshold: &default_eviction_threshold "20m"
+      evictions_low_residence_duration_metric_threshold: *default_eviction_threshold
+      remote_storage:
+        bucket_name: "{{ bucket_name }}"
+        bucket_region: "{{ bucket_region }}"
+        prefix_in_bucket: "pageserver/v1"
+    safekeeper_s3_prefix: safekeeper/v1/wal
+    hostname_suffix: ""
+    remote_user: ssm-user
+    ansible_aws_ssm_region: eu-central-1
+    ansible_aws_ssm_bucket_name: neon-dev-storage-eu-central-1
+    console_region_id: aws-eu-central-1
+    sentry_environment: staging
+
+  children:
+    pageservers:
+      hosts:
+        pageserver-0.eu-central-1.aws.neon.build:
+          ansible_host: i-011f93ec26cfba2d4
+
+    safekeepers:
+      hosts:
+        safekeeper-0.eu-central-1.aws.neon.build:
+          ansible_host: i-0ff026d27babf8ddd
+        safekeeper-1.eu-central-1.aws.neon.build:
+          ansible_host: i-03983a49ee54725d9
+        safekeeper-2.eu-central-1.aws.neon.build:
+          ansible_host: i-0bd025ecdb61b0db3
--- a/.github/ansible/staging.eu-west-1.hosts.yaml
+++ b/.github/ansible/staging.eu-west-1.hosts.yaml
@@ -0,0 +1,60 @@
+storage:
+  vars:
+    bucket_name: neon-dev-storage-eu-west-1
+    bucket_region: eu-west-1
+    console_mgmt_base_url: http://neon-internal-api.aws.neon.build
+    broker_endpoint: http://storage-broker-lb.zeta.eu-west-1.internal.aws.neon.build:50051
+    pageserver_config_stub:
+      pg_distrib_dir: /usr/local
+      metric_collection_endpoint: http://neon-internal-api.aws.neon.build/billing/api/v1/usage_events
+      metric_collection_interval: 10min
+      disk_usage_based_eviction:
+        max_usage_pct: 80
+        min_avail_bytes: 0
+        period: "10s"
+      tenant_config:
+        eviction_policy:
+          kind: "LayerAccessThreshold"
+          period: "20m"
+          threshold: &default_eviction_threshold "20m"
+        evictions_low_residence_duration_metric_threshold: *default_eviction_threshold
+      remote_storage:
+        bucket_name: "{{ bucket_name }}"
+        bucket_region: "{{ bucket_region }}"
+        prefix_in_bucket: "pageserver/v1"
+    safekeeper_s3_prefix: safekeeper/v1/wal
+    hostname_suffix: ""
+    remote_user: ssm-user
+    ansible_aws_ssm_region: eu-west-1
+    ansible_aws_ssm_bucket_name: neon-dev-storage-eu-west-1
+    console_region_id: aws-eu-west-1
+    sentry_environment: staging
+
+  children:
+    pageservers:
+      hosts:
+        pageserver-0.eu-west-1.aws.neon.build:
+          ansible_host: i-01d496c5041c7f34c
+        pageserver-1.eu-west-1.aws.neon.build:
+          ansible_host: i-0e8013e239ce3928c
+
+    safekeepers:
+      hosts:
+        safekeeper-0.eu-west-1.aws.neon.build:
+          ansible_host: i-05226ef85722831bf
+        safekeeper-1.eu-west-1.aws.neon.build:
+          ansible_host: i-06969ee1bf2958bfc
+        safekeeper-2.eu-west-1.aws.neon.build:
+          ansible_host: i-087892e9625984a0b
+        safekeeper-3.eu-west-1.aws.neon.build:
+          ansible_host: i-0a6f91660e99e8891
+        safekeeper-4.eu-west-1.aws.neon.build:
+          ansible_host: i-0012e309e28e7c249
+        safekeeper-5.eu-west-1.aws.neon.build:
+          ansible_host: i-085a2b1193287b32e
+        safekeeper-6.eu-west-1.aws.neon.build:
+          ansible_host: i-0c713248465ed0fbd
+        safekeeper-7.eu-west-1.aws.neon.build:
+          ansible_host: i-02ad231aed2a80b7a
+        safekeeper-8.eu-west-1.aws.neon.build:
+          ansible_host: i-0dbbd8ffef66efda8
--- a/.github/ansible/staging.us-east-2.hosts.yaml
+++ b/.github/ansible/staging.us-east-2.hosts.yaml
@@ -0,0 +1,56 @@
+storage:
+  vars:
+    bucket_name: neon-staging-storage-us-east-2
+    bucket_region: us-east-2
+    console_mgmt_base_url: http://neon-internal-api.aws.neon.build
+    broker_endpoint: http://storage-broker-lb.beta.us-east-2.internal.aws.neon.build:50051
+    pageserver_config_stub:
+      pg_distrib_dir: /usr/local
+      metric_collection_endpoint: http://neon-internal-api.aws.neon.build/billing/api/v1/usage_events
+      metric_collection_interval: 10min
+      disk_usage_based_eviction:
+        max_usage_pct: 80
+        min_avail_bytes: 0
+        period: "10s"
+      tenant_config:
+        eviction_policy:
+          kind: "LayerAccessThreshold"
+          period: "20m"
+          threshold: &default_eviction_threshold "20m"
+        evictions_low_residence_duration_metric_threshold: *default_eviction_threshold
+      remote_storage:
+        bucket_name: "{{ bucket_name }}"
+        bucket_region: "{{ bucket_region }}"
+        prefix_in_bucket: "pageserver/v1"
+    safekeeper_s3_prefix: safekeeper/v1/wal
+    hostname_suffix: ""
+    remote_user: ssm-user
+    ansible_aws_ssm_region: us-east-2
+    ansible_aws_ssm_bucket_name: neon-staging-storage-us-east-2
+    console_region_id: aws-us-east-2
+    sentry_environment: staging
+
+  children:
+    pageservers:
+      hosts:
+        pageserver-0.us-east-2.aws.neon.build:
+          ansible_host: i-0c3e70929edb5d691
+        pageserver-1.us-east-2.aws.neon.build:
+          ansible_host: i-0565a8b4008aa3f40
+        pageserver-2.us-east-2.aws.neon.build:
+          ansible_host: i-01e31cdf7e970586a
+        pageserver-3.us-east-2.aws.neon.build:
+          ansible_host: i-0602a0291365ef7cc
+        pageserver-99.us-east-2.aws.neon.build:
+          ansible_host: i-0c39491109bb88824
+
+    safekeepers:
+      hosts:
+        safekeeper-0.us-east-2.aws.neon.build:
+          ansible_host: i-027662bd552bf5db0
+        safekeeper-2.us-east-2.aws.neon.build:
+          ansible_host: i-0de0b03a51676a6ce
+        safekeeper-3.us-east-2.aws.neon.build:
+          ansible_host: i-05f8ba2cda243bd18
+        safekeeper-99.us-east-2.aws.neon.build:
+          ansible_host: i-0d61b6a2ea32028d5
--- a/.github/ansible/systemd/pageserver.service
+++ b/.github/ansible/systemd/pageserver.service
@@ -0,0 +1,18 @@
+[Unit]
+Description=Neon pageserver
+After=network.target auditd.service
+
+[Service]
+Type=simple
+User=pageserver
+Environment=RUST_BACKTRACE=1 NEON_REPO_DIR=/storage/pageserver LD_LIBRARY_PATH=/usr/local/v14/lib SENTRY_DSN={{ SENTRY_URL_PAGESERVER }} SENTRY_ENVIRONMENT={{ sentry_environment }}
+ExecStart=/usr/local/bin/pageserver -c "pg_distrib_dir='/usr/local'" -c "listen_pg_addr='0.0.0.0:6400'" -c "listen_http_addr='0.0.0.0:9898'" -c "broker_endpoint='{{ broker_endpoint }}'" -c "availability_zone='{{ ec2_availability_zone }}'" -D /storage/pageserver/data
+ExecReload=/bin/kill -HUP $MAINPID
+KillMode=mixed
+KillSignal=SIGINT
+Restart=on-failure
+TimeoutSec=10
+LimitNOFILE=30000000
+
+[Install]
+WantedBy=multi-user.target
--- a/.github/ansible/systemd/safekeeper.service
+++ b/.github/ansible/systemd/safekeeper.service
@@ -0,0 +1,18 @@
+[Unit]
+Description=Neon safekeeper
+After=network.target auditd.service
+
+[Service]
+Type=simple
+User=safekeeper
+Environment=RUST_BACKTRACE=1 NEON_REPO_DIR=/storage/safekeeper/data LD_LIBRARY_PATH=/usr/local/v14/lib SENTRY_DSN={{ SENTRY_URL_SAFEKEEPER }} SENTRY_ENVIRONMENT={{ sentry_environment }}
+ExecStart=/usr/local/bin/safekeeper -l {{ inventory_hostname }}{{ hostname_suffix }}:6500 --listen-http {{ inventory_hostname }}{{ hostname_suffix }}:7676 -D /storage/safekeeper/data --broker-endpoint={{ broker_endpoint }} --remote-storage='{bucket_name="{{bucket_name}}", bucket_region="{{bucket_region}}", prefix_in_bucket="{{ safekeeper_s3_prefix }}"}' --availability-zone={{ ec2_availability_zone }}
+ExecReload=/bin/kill -HUP $MAINPID
+KillMode=mixed
+KillSignal=SIGINT
+Restart=on-failure
+TimeoutSec=10
+LimitNOFILE=30000000
+
+[Install]
+WantedBy=multi-user.target
--- a/.github/ansible/templates/pageserver.toml.j2
+++ b/.github/ansible/templates/pageserver.toml.j2
@@ -0,0 +1 @@
+{{ pageserver_config | sivel.toiletwater.to_toml }}
--- a/.github/helm-values/dev-eu-central-1-alpha.neon-storage-broker.yaml
+++ b/.github/helm-values/dev-eu-central-1-alpha.neon-storage-broker.yaml
@@ -0,0 +1,52 @@
+# Helm chart values for neon-storage-broker
+podLabels:
+  neon_env: staging
+  neon_service: storage-broker
+
+# Use L4 LB
+service:
+  # service.annotations -- Annotations to add to the service
+  annotations:
+    service.beta.kubernetes.io/aws-load-balancer-type: external  # use newer AWS Load Balancer Controller
+    service.beta.kubernetes.io/aws-load-balancer-nlb-target-type: ip
+    service.beta.kubernetes.io/aws-load-balancer-scheme: internal  # deploy LB to private subnet
+    # assign service to this name at external-dns
+    external-dns.alpha.kubernetes.io/hostname: storage-broker-lb.alpha.eu-central-1.internal.aws.neon.build
+  # service.type -- Service type
+  type: LoadBalancer
+  # service.port -- broker listen port
+  port: 50051
+
+ingress:
+  enabled: false
+
+metrics:
+  enabled: false
+
+extraManifests:
+  - apiVersion: operator.victoriametrics.com/v1beta1
+    kind: VMServiceScrape
+    metadata:
+      name: "{{ include \"neon-storage-broker.fullname\" . }}"
+      labels:
+        helm.sh/chart: neon-storage-broker-{{ .Chart.Version }}
+        app.kubernetes.io/name: neon-storage-broker
+        app.kubernetes.io/instance: neon-storage-broker
+        app.kubernetes.io/version: "{{ .Chart.AppVersion }}"
+        app.kubernetes.io/managed-by: Helm
+      namespace: "{{ .Release.Namespace }}"
+    spec:
+      selector:
+        matchLabels:
+          app.kubernetes.io/name: "neon-storage-broker"
+      endpoints:
+        - port: broker
+          path: /metrics
+          interval: 10s
+          scrapeTimeout: 10s
+      namespaceSelector:
+        matchNames:
+          - "{{ .Release.Namespace }}"
+
+settings:
+  sentryEnvironment: "staging"
--- a/.github/helm-values/dev-eu-central-1-alpha.pg-sni-router.yaml
+++ b/.github/helm-values/dev-eu-central-1-alpha.pg-sni-router.yaml
@@ -0,0 +1,19 @@
+useCertManager: true
+
+replicaCount: 3
+
+exposedService:
+  # exposedService.port -- Exposed Service proxy port
+  port: 4432
+  annotations:
+    external-dns.alpha.kubernetes.io/hostname: "*.snirouter.alpha.eu-central-1.internal.aws.neon.build"
+
+settings:
+  domain: "*.snirouter.alpha.eu-central-1.internal.aws.neon.build"
+  sentryEnvironment: "staging"
+
+imagePullSecrets:
+  - name: docker-hub-neon
+
+metrics:
+  enabled: false
--- a/.github/helm-values/dev-eu-west-1-zeta.neon-proxy-scram.yaml
+++ b/.github/helm-values/dev-eu-west-1-zeta.neon-proxy-scram.yaml
@@ -0,0 +1,76 @@
+# Helm chart values for neon-proxy-scram.
+# This is a YAML-formatted file.
+
+deploymentStrategy:
+  type: RollingUpdate
+  rollingUpdate:
+    maxSurge: 100%
+    maxUnavailable: 50%
+
+# Delay the kill signal by 5 minutes (5 * 60)
+# The pod(s) will stay in Terminating, keeps the existing connections
+# but doesn't receive new ones
+containerLifecycle:
+  preStop:
+    exec:
+      command: ["/bin/sh", "-c", "sleep 300"]
+terminationGracePeriodSeconds: 604800
+
+image:
+  repository: neondatabase/neon
+
+settings:
+  authBackend: "console"
+  authEndpoint: "http://neon-internal-api.aws.neon.build/management/api/v2"
+  domain: "*.eu-west-1.aws.neon.build"
+  otelExporterOtlpEndpoint: "https://otel-collector.zeta.eu-west-1.internal.aws.neon.build"
+  sentryEnvironment: "staging"
+  wssPort: 8443
+  metricCollectionEndpoint: "http://neon-internal-api.aws.neon.build/billing/api/v1/usage_events"
+  metricCollectionInterval: "1min"
+
+# -- Additional labels for neon-proxy pods
+podLabels:
+  neon_service: proxy-scram
+  neon_env: dev
+  neon_region: eu-west-1
+
+exposedService:
+  annotations:
+    service.beta.kubernetes.io/aws-load-balancer-type: external
+    service.beta.kubernetes.io/aws-load-balancer-nlb-target-type: ip
+    service.beta.kubernetes.io/aws-load-balancer-scheme: internet-facing
+    external-dns.alpha.kubernetes.io/hostname: eu-west-1.aws.neon.build
+  httpsPort: 443
+
+#metrics:
+#  enabled: true
+#  serviceMonitor:
+#    enabled: true
+#    selector:
+#      release: kube-prometheus-stack
+
+extraManifests:
+  - apiVersion: operator.victoriametrics.com/v1beta1
+    kind: VMServiceScrape
+    metadata:
+      name: "{{ include \"neon-proxy.fullname\" . }}"
+      labels:
+        helm.sh/chart: neon-proxy-{{ .Chart.Version }}
+        app.kubernetes.io/name: neon-proxy
+        app.kubernetes.io/instance: "{{ include \"neon-proxy.fullname\" . }}"
+        app.kubernetes.io/version: "{{ .Chart.AppVersion }}"
+        app.kubernetes.io/managed-by: Helm
+      namespace: "{{ .Release.Namespace }}"
+    spec:
+      selector:
+        matchLabels:
+          app.kubernetes.io/name: "neon-proxy"
+      endpoints:
+        - port: http
+          path: /metrics
+          interval: 10s
+          scrapeTimeout: 10s
+      namespaceSelector:
+        matchNames:
+          - "{{ .Release.Namespace }}"
--- a/.github/helm-values/dev-eu-west-1-zeta.neon-storage-broker.yaml
+++ b/.github/helm-values/dev-eu-west-1-zeta.neon-storage-broker.yaml
@@ -0,0 +1,52 @@
+# Helm chart values for neon-storage-broker
+podLabels:
+  neon_env: staging
+  neon_service: storage-broker
+
+# Use L4 LB
+service:
+  # service.annotations -- Annotations to add to the service
+  annotations:
+    service.beta.kubernetes.io/aws-load-balancer-type: external  # use newer AWS Load Balancer Controller
+    service.beta.kubernetes.io/aws-load-balancer-nlb-target-type: ip
+    service.beta.kubernetes.io/aws-load-balancer-scheme: internal  # deploy LB to private subnet
+    # assign service to this name at external-dns
+    external-dns.alpha.kubernetes.io/hostname: storage-broker-lb.zeta.eu-west-1.internal.aws.neon.build
+  # service.type -- Service type
+  type: LoadBalancer
+  # service.port -- broker listen port
+  port: 50051
+
+ingress:
+  enabled: false
+
+metrics:
+  enabled: false
+
+extraManifests:
+  - apiVersion: operator.victoriametrics.com/v1beta1
+    kind: VMServiceScrape
+    metadata:
+      name: "{{ include \"neon-storage-broker.fullname\" . }}"
+      labels:
+        helm.sh/chart: neon-storage-broker-{{ .Chart.Version }}
+        app.kubernetes.io/name: neon-storage-broker
+        app.kubernetes.io/instance: neon-storage-broker
+        app.kubernetes.io/version: "{{ .Chart.AppVersion }}"
+        app.kubernetes.io/managed-by: Helm
+      namespace: "{{ .Release.Namespace }}"
+    spec:
+      selector:
+        matchLabels:
+          app.kubernetes.io/name: "neon-storage-broker"
+      endpoints:
+        - port: broker
+          path: /metrics
+          interval: 10s
+          scrapeTimeout: 10s
+      namespaceSelector:
+        matchNames:
+          - "{{ .Release.Namespace }}"
+
+settings:
+  sentryEnvironment: "staging"
--- a/.github/helm-values/dev-eu-west-1-zeta.pg-sni-router.yaml
+++ b/.github/helm-values/dev-eu-west-1-zeta.pg-sni-router.yaml
@@ -0,0 +1,19 @@
+useCertManager: true
+
+replicaCount: 3
+
+exposedService:
+  # exposedService.port -- Exposed Service proxy port
+  port: 4432
+  annotations:
+    external-dns.alpha.kubernetes.io/hostname: "*.snirouter.zeta.eu-west-1.internal.aws.neon.build"
+
+settings:
+  domain: "*.snirouter.zeta.eu-west-1.internal.aws.neon.build"
+  sentryEnvironment: "staging"
+
+imagePullSecrets:
+  - name: docker-hub-neon
+
+metrics:
+  enabled: false
--- a/.github/helm-values/dev-us-east-2-beta.neon-proxy-link.yaml
+++ b/.github/helm-values/dev-us-east-2-beta.neon-proxy-link.yaml
@@ -0,0 +1,68 @@
+# Helm chart values for neon-proxy-link.
+# This is a YAML-formatted file.
+
+image:
+  repository: neondatabase/neon
+
+settings:
+  authBackend: "link"
+  authEndpoint: "https://console.stage.neon.tech/authenticate_proxy_request/"
+  uri: "https://console.stage.neon.tech/psql_session/"
+  domain: "pg.neon.build"
+  otelExporterOtlpEndpoint: "https://otel-collector.beta.us-east-2.internal.aws.neon.build"
+  sentryEnvironment: "staging"
+  metricCollectionEndpoint: "http://neon-internal-api.aws.neon.build/billing/api/v1/usage_events"
+  metricCollectionInterval: "1min"
+
+# -- Additional labels for neon-proxy-link pods
+podLabels:
+  neon_service: proxy
+  neon_env: dev
+  neon_region: us-east-2
+
+service:
+  type: LoadBalancer
+  annotations:
+    service.beta.kubernetes.io/aws-load-balancer-type: external
+    service.beta.kubernetes.io/aws-load-balancer-nlb-target-type: ip
+    service.beta.kubernetes.io/aws-load-balancer-scheme: internal
+    external-dns.alpha.kubernetes.io/hostname: neon-proxy-link-mgmt.beta.us-east-2.aws.neon.build
+
+exposedService:
+  annotations:
+    service.beta.kubernetes.io/aws-load-balancer-type: external
+    service.beta.kubernetes.io/aws-load-balancer-nlb-target-type: ip
+    service.beta.kubernetes.io/aws-load-balancer-scheme: internet-facing
+    external-dns.alpha.kubernetes.io/hostname: neon-proxy-link.beta.us-east-2.aws.neon.build
+
+#metrics:
+#  enabled: true
+#  serviceMonitor:
+#    enabled: true
+#    selector:
+#      release: kube-prometheus-stack
+
+extraManifests:
+  - apiVersion: operator.victoriametrics.com/v1beta1
+    kind: VMServiceScrape
+    metadata:
+      name: "{{ include \"neon-proxy.fullname\" . }}"
+      labels:
+        helm.sh/chart: neon-proxy-{{ .Chart.Version }}
+        app.kubernetes.io/name: neon-proxy
+        app.kubernetes.io/instance: "{{ include \"neon-proxy.fullname\" . }}"
+        app.kubernetes.io/version: "{{ .Chart.AppVersion }}"
+        app.kubernetes.io/managed-by: Helm
+      namespace: "{{ .Release.Namespace }}"
+    spec:
+      selector:
+        matchLabels:
+          app.kubernetes.io/name: "neon-proxy"
+      endpoints:
+        - port: http
+          path: /metrics
+          interval: 10s
+          scrapeTimeout: 10s
+      namespaceSelector:
+        matchNames:
+          - "{{ .Release.Namespace }}"
--- a/.github/helm-values/dev-us-east-2-beta.neon-proxy-scram-legacy.yaml
+++ b/.github/helm-values/dev-us-east-2-beta.neon-proxy-scram-legacy.yaml
@@ -0,0 +1,77 @@
+# Helm chart values for neon-proxy-scram.
+# This is a YAML-formatted file.
+
+deploymentStrategy:
+  type: RollingUpdate
+  rollingUpdate:
+    maxSurge: 100%
+    maxUnavailable: 50%
+
+# Delay the kill signal by 5 minutes (5 * 60)
+# The pod(s) will stay in Terminating, keeps the existing connections
+# but doesn't receive new ones
+containerLifecycle:
+  preStop:
+    exec:
+      command: ["/bin/sh", "-c", "sleep 300"]
+terminationGracePeriodSeconds: 604800
+
+
+image:
+  repository: neondatabase/neon
+
+settings:
+  authBackend: "console"
+  authEndpoint: "http://neon-internal-api.aws.neon.build/management/api/v2"
+  domain: "*.cloud.stage.neon.tech"
+  otelExporterOtlpEndpoint: "https://otel-collector.beta.us-east-2.internal.aws.neon.build"
+  sentryEnvironment: "staging"
+  wssPort: 8443
+  metricCollectionEndpoint: "http://neon-internal-api.aws.neon.build/billing/api/v1/usage_events"
+  metricCollectionInterval: "1min"
+
+# -- Additional labels for neon-proxy pods
+podLabels:
+  neon_service: proxy-scram-legacy
+  neon_env: dev
+  neon_region: us-east-2
+
+exposedService:
+  annotations:
+    service.beta.kubernetes.io/aws-load-balancer-type: external
+    service.beta.kubernetes.io/aws-load-balancer-nlb-target-type: ip
+    service.beta.kubernetes.io/aws-load-balancer-scheme: internet-facing
+    external-dns.alpha.kubernetes.io/hostname: neon-proxy-scram-legacy.beta.us-east-2.aws.neon.build
+  httpsPort: 443
+
+#metrics:
+#  enabled: true
+#  serviceMonitor:
+#    enabled: true
+#    selector:
+#      release: kube-prometheus-stack
+
+extraManifests:
+  - apiVersion: operator.victoriametrics.com/v1beta1
+    kind: VMServiceScrape
+    metadata:
+      name: "{{ include \"neon-proxy.fullname\" . }}"
+      labels:
+        helm.sh/chart: neon-proxy-{{ .Chart.Version }}
+        app.kubernetes.io/name: neon-proxy
+        app.kubernetes.io/instance: "{{ include \"neon-proxy.fullname\" . }}"
+        app.kubernetes.io/version: "{{ .Chart.AppVersion }}"
+        app.kubernetes.io/managed-by: Helm
+      namespace: "{{ .Release.Namespace }}"
+    spec:
+      selector:
+        matchLabels:
+          app.kubernetes.io/name: "neon-proxy"
+      endpoints:
+        - port: http
+          path: /metrics
+          interval: 10s
+          scrapeTimeout: 10s
+      namespaceSelector:
+        matchNames:
+          - "{{ .Release.Namespace }}"
--- a/.github/helm-values/dev-us-east-2-beta.neon-proxy-scram.yaml
+++ b/.github/helm-values/dev-us-east-2-beta.neon-proxy-scram.yaml
@@ -0,0 +1,78 @@
+# Helm chart values for neon-proxy-scram.
+# This is a YAML-formatted file.
+
+deploymentStrategy:
+  type: RollingUpdate
+  rollingUpdate:
+    maxSurge: 100%
+    maxUnavailable: 50%
+
+# Delay the kill signal by 5 minutes (5 * 60)
+# The pod(s) will stay in Terminating, keeps the existing connections
+# but doesn't receive new ones
+containerLifecycle:
+  preStop:
+    exec:
+      command: ["/bin/sh", "-c", "sleep 300"]
+terminationGracePeriodSeconds: 604800
+
+
+image:
+  repository: neondatabase/neon
+
+settings:
+  authBackend: "console"
+  authEndpoint: "http://neon-internal-api.aws.neon.build/management/api/v2"
+  domain: "*.us-east-2.aws.neon.build"
+  extraDomains: ["*.us-east-2.postgres.zenith.tech", "*.us-east-2.retooldb-staging.com"]
+  otelExporterOtlpEndpoint: "https://otel-collector.beta.us-east-2.internal.aws.neon.build"
+  sentryEnvironment: "staging"
+  wssPort: 8443
+  metricCollectionEndpoint: "http://neon-internal-api.aws.neon.build/billing/api/v1/usage_events"
+  metricCollectionInterval: "1min"
+
+# -- Additional labels for neon-proxy pods
+podLabels:
+  neon_service: proxy-scram
+  neon_env: dev
+  neon_region: us-east-2
+
+exposedService:
+  annotations:
+    service.beta.kubernetes.io/aws-load-balancer-type: external
+    service.beta.kubernetes.io/aws-load-balancer-nlb-target-type: ip
+    service.beta.kubernetes.io/aws-load-balancer-scheme: internet-facing
+    external-dns.alpha.kubernetes.io/hostname: us-east-2.aws.neon.build
+  httpsPort: 443
+
+#metrics:
+#  enabled: true
+#  serviceMonitor:
+#    enabled: true
+#    selector:
+#      release: kube-prometheus-stack
+
+extraManifests:
+  - apiVersion: operator.victoriametrics.com/v1beta1
+    kind: VMServiceScrape
+    metadata:
+      name: "{{ include \"neon-proxy.fullname\" . }}"
+      labels:
+        helm.sh/chart: neon-proxy-{{ .Chart.Version }}
+        app.kubernetes.io/name: neon-proxy
+        app.kubernetes.io/instance: "{{ include \"neon-proxy.fullname\" . }}"
+        app.kubernetes.io/version: "{{ .Chart.AppVersion }}"
+        app.kubernetes.io/managed-by: Helm
+      namespace: "{{ .Release.Namespace }}"
+    spec:
+      selector:
+        matchLabels:
+          app.kubernetes.io/name: "neon-proxy"
+      endpoints:
+        - port: http
+          path: /metrics
+          interval: 10s
+          scrapeTimeout: 10s
+      namespaceSelector:
+        matchNames:
+          - "{{ .Release.Namespace }}"
--- a/.github/helm-values/dev-us-east-2-beta.neon-storage-broker.yaml
+++ b/.github/helm-values/dev-us-east-2-beta.neon-storage-broker.yaml
@@ -0,0 +1,52 @@
+# Helm chart values for neon-storage-broker
+podLabels:
+  neon_env: staging
+  neon_service: storage-broker
+
+# Use L4 LB
+service:
+  # service.annotations -- Annotations to add to the service
+  annotations:
+    service.beta.kubernetes.io/aws-load-balancer-type: external  # use newer AWS Load Balancer Controller
+    service.beta.kubernetes.io/aws-load-balancer-nlb-target-type: ip
+    service.beta.kubernetes.io/aws-load-balancer-scheme: internal  # deploy LB to private subnet
+    # assign service to this name at external-dns
+    external-dns.alpha.kubernetes.io/hostname: storage-broker-lb.beta.us-east-2.internal.aws.neon.build
+  # service.type -- Service type
+  type: LoadBalancer
+  # service.port -- broker listen port
+  port: 50051
+
+ingress:
+  enabled: false
+
+metrics:
+  enabled: false
+
+extraManifests:
+  - apiVersion: operator.victoriametrics.com/v1beta1
+    kind: VMServiceScrape
+    metadata:
+      name: "{{ include \"neon-storage-broker.fullname\" . }}"
+      labels:
+        helm.sh/chart: neon-storage-broker-{{ .Chart.Version }}
+        app.kubernetes.io/name: neon-storage-broker
+        app.kubernetes.io/instance: neon-storage-broker
+        app.kubernetes.io/version: "{{ .Chart.AppVersion }}"
+        app.kubernetes.io/managed-by: Helm
+      namespace: "{{ .Release.Namespace }}"
+    spec:
+      selector:
+        matchLabels:
+          app.kubernetes.io/name: "neon-storage-broker"
+      endpoints:
+        - port: broker
+          path: /metrics
+          interval: 10s
+          scrapeTimeout: 10s
+      namespaceSelector:
+        matchNames:
+          - "{{ .Release.Namespace }}"
+
+settings:
+  sentryEnvironment: "staging"
--- a/.github/helm-values/dev-us-east-2-beta.pg-sni-router.yaml
+++ b/.github/helm-values/dev-us-east-2-beta.pg-sni-router.yaml
@@ -0,0 +1,19 @@
+useCertManager: true
+
+replicaCount: 3
+
+exposedService:
+  # exposedService.port -- Exposed Service proxy port
+  port: 4432
+  annotations:
+    external-dns.alpha.kubernetes.io/hostname: "*.snirouter.beta.us-east-2.internal.aws.neon.build"
+
+settings:
+  domain: "*.snirouter.beta.us-east-2.internal.aws.neon.build"
+  sentryEnvironment: "staging"
+
+imagePullSecrets:
+  - name: docker-hub-neon
+
+metrics:
+  enabled: false
--- a/.github/helm-values/preview-template.neon-proxy-scram.yaml
+++ b/.github/helm-values/preview-template.neon-proxy-scram.yaml
@@ -0,0 +1,67 @@
+# Helm chart values for neon-proxy-scram.
+# This is a YAML-formatted file.
+
+deploymentStrategy:
+  type: RollingUpdate
+  rollingUpdate:
+    maxSurge: 100%
+    maxUnavailable: 50%
+
+image:
+  repository: neondatabase/neon
+
+settings:
+  authBackend: "console"
+  authEndpoint: "http://neon-internal-api.${PREVIEW_NAME}.aws.neon.build/management/api/v2"
+  domain: "*.cloud.${PREVIEW_NAME}.aws.neon.build"
+  sentryEnvironment: "staging"
+  wssPort: 8443
+  metricCollectionEndpoint: "http://neon-internal-api.${PREVIEW_NAME}.aws.neon.build/billing/api/v1/usage_events"
+  metricCollectionInterval: "1min"
+
+# -- Additional labels for neon-proxy pods
+podLabels:
+  neon_service: proxy-scram
+  neon_env: test
+  neon_region: ${PREVIEW_NAME}.eu-central-1
+
+
+exposedService:
+  annotations:
+    service.beta.kubernetes.io/aws-load-balancer-type: external
+    service.beta.kubernetes.io/aws-load-balancer-nlb-target-type: ip
+    service.beta.kubernetes.io/aws-load-balancer-scheme: internet-facing
+    external-dns.alpha.kubernetes.io/hostname: cloud.${PREVIEW_NAME}.aws.neon.build
+  httpsPort: 443
+
+#metrics:
+#  enabled: true
+#  serviceMonitor:
+#    enabled: true
+#    selector:
+#      release: kube-prometheus-stack
+
+extraManifests:
+  - apiVersion: operator.victoriametrics.com/v1beta1
+    kind: VMServiceScrape
+    metadata:
+      name: "{{ include \"neon-proxy.fullname\" . }}"
+      labels:
+        helm.sh/chart: neon-proxy-{{ .Chart.Version }}
+        app.kubernetes.io/name: neon-proxy
+        app.kubernetes.io/instance: "{{ include \"neon-proxy.fullname\" . }}"
+        app.kubernetes.io/version: "{{ .Chart.AppVersion }}"
+        app.kubernetes.io/managed-by: Helm
+      namespace: "{{ .Release.Namespace }}"
+    spec:
+      selector:
+        matchLabels:
+          app.kubernetes.io/name: "neon-proxy"
+      endpoints:
+        - port: http
+          path: /metrics
+          interval: 10s
+          scrapeTimeout: 10s
+      namespaceSelector:
+        matchNames:
+          - "{{ .Release.Namespace }}"
--- a/.github/helm-values/prod-ap-southeast-1-epsilon.neon-proxy-scram.yaml
+++ b/.github/helm-values/prod-ap-southeast-1-epsilon.neon-proxy-scram.yaml
@@ -0,0 +1,77 @@
+# Helm chart values for neon-proxy-scram.
+# This is a YAML-formatted file.
+
+deploymentStrategy:
+  type: RollingUpdate
+  rollingUpdate:
+    maxSurge: 100%
+    maxUnavailable: 50%
+
+# Delay the kill signal by 5 minutes (5 * 60)
+# The pod(s) will stay in Terminating, keeps the existing connections
+# but doesn't receive new ones
+containerLifecycle:
+  preStop:
+    exec:
+      command: ["/bin/sh", "-c", "sleep 300"]
+terminationGracePeriodSeconds: 604800
+
+
+image:
+  repository: neondatabase/neon
+
+settings:
+  authBackend: "console"
+  authEndpoint: "http://neon-internal-api.aws.neon.tech/management/api/v2"
+  domain: "*.ap-southeast-1.aws.neon.tech"
+  extraDomains: ["*.ap-southeast-1.retooldb.com", "*.ap-southeast-1.postgres.vercel-storage.com"]
+  sentryEnvironment: "production"
+  wssPort: 8443
+  metricCollectionEndpoint: "http://neon-internal-api.aws.neon.tech/billing/api/v1/usage_events"
+  metricCollectionInterval: "10min"
+
+# -- Additional labels for neon-proxy pods
+podLabels:
+  neon_service: proxy-scram
+  neon_env: prod
+  neon_region: ap-southeast-1
+
+exposedService:
+  annotations:
+    service.beta.kubernetes.io/aws-load-balancer-type: external
+    service.beta.kubernetes.io/aws-load-balancer-nlb-target-type: ip
+    service.beta.kubernetes.io/aws-load-balancer-scheme: internet-facing
+    external-dns.alpha.kubernetes.io/hostname: ap-southeast-1.aws.neon.tech
+  httpsPort: 443
+
+#metrics:
+#  enabled: true
+#  serviceMonitor:
+#    enabled: true
+#    selector:
+#      release: kube-prometheus-stack
+
+extraManifests:
+  - apiVersion: operator.victoriametrics.com/v1beta1
+    kind: VMServiceScrape
+    metadata:
+      name: "{{ include \"neon-proxy.fullname\" . }}"
+      labels:
+        helm.sh/chart: neon-proxy-{{ .Chart.Version }}
+        app.kubernetes.io/name: neon-proxy
+        app.kubernetes.io/instance: "{{ include \"neon-proxy.fullname\" . }}"
+        app.kubernetes.io/version: "{{ .Chart.AppVersion }}"
+        app.kubernetes.io/managed-by: Helm
+      namespace: "{{ .Release.Namespace }}"
+    spec:
+      selector:
+        matchLabels:
+          app.kubernetes.io/name: "neon-proxy"
+      endpoints:
+        - port: http
+          path: /metrics
+          interval: 10s
+          scrapeTimeout: 10s
+      namespaceSelector:
+        matchNames:
+          - "{{ .Release.Namespace }}"
--- a/.github/helm-values/prod-ap-southeast-1-epsilon.neon-storage-broker.yaml
+++ b/.github/helm-values/prod-ap-southeast-1-epsilon.neon-storage-broker.yaml
@@ -0,0 +1,52 @@
+# Helm chart values for neon-storage-broker
+podLabels:
+  neon_env: production
+  neon_service: storage-broker
+
+# Use L4 LB
+service:
+  # service.annotations -- Annotations to add to the service
+  annotations:
+    service.beta.kubernetes.io/aws-load-balancer-type: external  # use newer AWS Load Balancer Controller
+    service.beta.kubernetes.io/aws-load-balancer-nlb-target-type: ip
+    service.beta.kubernetes.io/aws-load-balancer-scheme: internal  # deploy LB to private subnet
+    # assign service to this name at external-dns
+    external-dns.alpha.kubernetes.io/hostname: storage-broker-lb.epsilon.ap-southeast-1.internal.aws.neon.tech
+  # service.type -- Service type
+  type: LoadBalancer
+  # service.port -- broker listen port
+  port: 50051
+
+ingress:
+  enabled: false
+
+metrics:
+  enabled: false
+
+extraManifests:
+  - apiVersion: operator.victoriametrics.com/v1beta1
+    kind: VMServiceScrape
+    metadata:
+      name: "{{ include \"neon-storage-broker.fullname\" . }}"
+      labels:
+        helm.sh/chart: neon-storage-broker-{{ .Chart.Version }}
+        app.kubernetes.io/name: neon-storage-broker
+        app.kubernetes.io/instance: neon-storage-broker
+        app.kubernetes.io/version: "{{ .Chart.AppVersion }}"
+        app.kubernetes.io/managed-by: Helm
+      namespace: "{{ .Release.Namespace }}"
+    spec:
+      selector:
+        matchLabels:
+          app.kubernetes.io/name: "neon-storage-broker"
+      endpoints:
+        - port: broker
+          path: /metrics
+          interval: 10s
+          scrapeTimeout: 10s
+      namespaceSelector:
+        matchNames:
+          - "{{ .Release.Namespace }}"
+
+settings:
+  sentryEnvironment: "production"
--- a/.github/helm-values/prod-ap-southeast-1-epsilon.pg-sni-router.yaml
+++ b/.github/helm-values/prod-ap-southeast-1-epsilon.pg-sni-router.yaml
@@ -0,0 +1,19 @@
+useCertManager: true
+
+replicaCount: 3
+
+exposedService:
+  # exposedService.port -- Exposed Service proxy port
+  port: 4432
+  annotations:
+    external-dns.alpha.kubernetes.io/hostname: "*.snirouter.epsilon.ap-southeast-1.internal.aws.neon.tech"
+
+settings:
+  domain: "*.snirouter.epsilon.ap-southeast-1.internal.aws.neon.tech"
+  sentryEnvironment: "production"
+
+imagePullSecrets:
+  - name: docker-hub-neon
+
+metrics:
+  enabled: false
--- a/.github/helm-values/prod-eu-central-1-gamma.neon-proxy-scram.yaml
+++ b/.github/helm-values/prod-eu-central-1-gamma.neon-proxy-scram.yaml
@@ -0,0 +1,77 @@
+# Helm chart values for neon-proxy-scram.
+# This is a YAML-formatted file.
+
+deploymentStrategy:
+  type: RollingUpdate
+  rollingUpdate:
+    maxSurge: 100%
+    maxUnavailable: 50%
+
+# Delay the kill signal by 5 minutes (5 * 60)
+# The pod(s) will stay in Terminating, keeps the existing connections
+# but doesn't receive new ones
+containerLifecycle:
+  preStop:
+    exec:
+      command: ["/bin/sh", "-c", "sleep 300"]
+terminationGracePeriodSeconds: 604800
+
+
+image:
+  repository: neondatabase/neon
+
+settings:
+  authBackend: "console"
+  authEndpoint: "http://neon-internal-api.aws.neon.tech/management/api/v2"
+  domain: "*.eu-central-1.aws.neon.tech"
+  extraDomains: ["*.eu-central-1.retooldb.com", "*.eu-central-1.postgres.vercel-storage.com"]
+  sentryEnvironment: "production"
+  wssPort: 8443
+  metricCollectionEndpoint: "http://neon-internal-api.aws.neon.tech/billing/api/v1/usage_events"
+  metricCollectionInterval: "10min"
+
+# -- Additional labels for neon-proxy pods
+podLabels:
+  neon_service: proxy-scram
+  neon_env: prod
+  neon_region: eu-central-1
+
+exposedService:
+  annotations:
+    service.beta.kubernetes.io/aws-load-balancer-type: external
+    service.beta.kubernetes.io/aws-load-balancer-nlb-target-type: ip
+    service.beta.kubernetes.io/aws-load-balancer-scheme: internet-facing
+    external-dns.alpha.kubernetes.io/hostname: eu-central-1.aws.neon.tech
+  httpsPort: 443
+
+#metrics:
+#  enabled: true
+#  serviceMonitor:
+#    enabled: true
+#    selector:
+#      release: kube-prometheus-stack
+
+extraManifests:
+  - apiVersion: operator.victoriametrics.com/v1beta1
+    kind: VMServiceScrape
+    metadata:
+      name: "{{ include \"neon-proxy.fullname\" . }}"
+      labels:
+        helm.sh/chart: neon-proxy-{{ .Chart.Version }}
+        app.kubernetes.io/name: neon-proxy
+        app.kubernetes.io/instance: "{{ include \"neon-proxy.fullname\" . }}"
+        app.kubernetes.io/version: "{{ .Chart.AppVersion }}"
+        app.kubernetes.io/managed-by: Helm
+      namespace: "{{ .Release.Namespace }}"
+    spec:
+      selector:
+        matchLabels:
+          app.kubernetes.io/name: "neon-proxy"
+      endpoints:
+        - port: http
+          path: /metrics
+          interval: 10s
+          scrapeTimeout: 10s
+      namespaceSelector:
+        matchNames:
+          - "{{ .Release.Namespace }}"
--- a/.github/helm-values/prod-eu-central-1-gamma.neon-storage-broker.yaml
+++ b/.github/helm-values/prod-eu-central-1-gamma.neon-storage-broker.yaml
@@ -0,0 +1,52 @@
+# Helm chart values for neon-storage-broker
+podLabels:
+  neon_env: production
+  neon_service: storage-broker
+
+# Use L4 LB
+service:
+  # service.annotations -- Annotations to add to the service
+  annotations:
+    service.beta.kubernetes.io/aws-load-balancer-type: external  # use newer AWS Load Balancer Controller
+    service.beta.kubernetes.io/aws-load-balancer-nlb-target-type: ip
+    service.beta.kubernetes.io/aws-load-balancer-scheme: internal  # deploy LB to private subnet
+    # assign service to this name at external-dns
+    external-dns.alpha.kubernetes.io/hostname: storage-broker-lb.gamma.eu-central-1.internal.aws.neon.tech
+  # service.type -- Service type
+  type: LoadBalancer
+  # service.port -- broker listen port
+  port: 50051
+
+ingress:
+  enabled: false
+
+metrics:
+  enabled: false
+
+extraManifests:
+  - apiVersion: operator.victoriametrics.com/v1beta1
+    kind: VMServiceScrape
+    metadata:
+      name: "{{ include \"neon-storage-broker.fullname\" . }}"
+      labels:
+        helm.sh/chart: neon-storage-broker-{{ .Chart.Version }}
+        app.kubernetes.io/name: neon-storage-broker
+        app.kubernetes.io/instance: neon-storage-broker
+        app.kubernetes.io/version: "{{ .Chart.AppVersion }}"
+        app.kubernetes.io/managed-by: Helm
+      namespace: "{{ .Release.Namespace }}"
+    spec:
+      selector:
+        matchLabels:
+          app.kubernetes.io/name: "neon-storage-broker"
+      endpoints:
+        - port: broker
+          path: /metrics
+          interval: 10s
+          scrapeTimeout: 10s
+      namespaceSelector:
+        matchNames:
+          - "{{ .Release.Namespace }}"
+
+settings:
+  sentryEnvironment: "production"
--- a/.github/helm-values/prod-eu-central-1-gamma.pg-sni-router.yaml
+++ b/.github/helm-values/prod-eu-central-1-gamma.pg-sni-router.yaml
@@ -0,0 +1,19 @@
+useCertManager: true
+
+replicaCount: 3
+
+exposedService:
+  # exposedService.port -- Exposed Service proxy port
+  port: 4432
+  annotations:
+    external-dns.alpha.kubernetes.io/hostname: "*.snirouter.gamma.eu-central-1.internal.aws.neon.tech"
+
+settings:
+  domain: "*.snirouter.gamma.eu-central-1.internal.aws.neon.tech"
+  sentryEnvironment: "production"
+
+imagePullSecrets:
+  - name: docker-hub-neon
+
+metrics:
+  enabled: false
--- a/.github/helm-values/prod-us-east-1-theta.neon-proxy-scram.yaml
+++ b/.github/helm-values/prod-us-east-1-theta.neon-proxy-scram.yaml
@@ -0,0 +1,69 @@
+# Helm chart values for neon-proxy-scram.
+# This is a YAML-formatted file.
+
+deploymentStrategy:
+  type: RollingUpdate
+  rollingUpdate:
+    maxSurge: 100%
+    maxUnavailable: 50%
+
+# Delay the kill signal by 5 minutes (5 * 60)
+# The pod(s) will stay in Terminating, keeps the existing connections
+# but doesn't receive new ones
+containerLifecycle:
+  preStop:
+    exec:
+      command: ["/bin/sh", "-c", "sleep 300"]
+terminationGracePeriodSeconds: 604800
+
+image:
+  repository: neondatabase/neon
+
+settings:
+  authBackend: "console"
+  authEndpoint: "http://neon-internal-api.aws.neon.tech/management/api/v2"
+  domain: "*.us-east-1.aws.neon.tech"
+  # *.us-east-1.retooldb.com hasn't been delegated yet.
+  extraDomains: ["*.us-east-1.postgres.vercel-storage.com"]
+  sentryEnvironment: "production"
+  wssPort: 8443
+  metricCollectionEndpoint: "http://neon-internal-api.aws.neon.tech/billing/api/v1/usage_events"
+  metricCollectionInterval: "10min"
+
+podLabels:
+  neon_service: proxy-scram
+  neon_env: prod
+  neon_region: us-east-1
+
+exposedService:
+  annotations:
+    service.beta.kubernetes.io/aws-load-balancer-type: external
+    service.beta.kubernetes.io/aws-load-balancer-nlb-target-type: ip
+    service.beta.kubernetes.io/aws-load-balancer-scheme: internet-facing
+    external-dns.alpha.kubernetes.io/hostname: us-east-1.aws.neon.tech
+  httpsPort: 443
+
+extraManifests:
+  - apiVersion: operator.victoriametrics.com/v1beta1
+    kind: VMServiceScrape
+    metadata:
+      name: "{{ include \"neon-proxy.fullname\" . }}"
+      labels:
+        helm.sh/chart: neon-proxy-{{ .Chart.Version }}
+        app.kubernetes.io/name: neon-proxy
+        app.kubernetes.io/instance: "{{ include \"neon-proxy.fullname\" . }}"
+        app.kubernetes.io/version: "{{ .Chart.AppVersion }}"
+        app.kubernetes.io/managed-by: Helm
+      namespace: "{{ .Release.Namespace }}"
+    spec:
+      selector:
+        matchLabels:
+          app.kubernetes.io/name: "neon-proxy"
+      endpoints:
+        - port: http
+          path: /metrics
+          interval: 10s
+          scrapeTimeout: 10s
+      namespaceSelector:
+        matchNames:
+          - "{{ .Release.Namespace }}"
--- a/.github/helm-values/prod-us-east-1-theta.neon-storage-broker.yaml
+++ b/.github/helm-values/prod-us-east-1-theta.neon-storage-broker.yaml
@@ -0,0 +1,52 @@
+# Helm chart values for neon-storage-broker
+podLabels:
+  neon_env: production
+  neon_service: storage-broker
+
+# Use L4 LB
+service:
+  # service.annotations -- Annotations to add to the service
+  annotations:
+    service.beta.kubernetes.io/aws-load-balancer-type: external  # use newer AWS Load Balancer Controller
+    service.beta.kubernetes.io/aws-load-balancer-nlb-target-type: ip
+    service.beta.kubernetes.io/aws-load-balancer-scheme: internal  # deploy LB to private subnet
+    # assign service to this name at external-dns
+    external-dns.alpha.kubernetes.io/hostname: storage-broker-lb.theta.us-east-1.internal.aws.neon.tech
+  # service.type -- Service type
+  type: LoadBalancer
+  # service.port -- broker listen port
+  port: 50051
+
+ingress:
+  enabled: false
+
+metrics:
+  enabled: false
+
+extraManifests:
+  - apiVersion: operator.victoriametrics.com/v1beta1
+    kind: VMServiceScrape
+    metadata:
+      name: "{{ include \"neon-storage-broker.fullname\" . }}"
+      labels:
+        helm.sh/chart: neon-storage-broker-{{ .Chart.Version }}
+        app.kubernetes.io/name: neon-storage-broker
+        app.kubernetes.io/instance: neon-storage-broker
+        app.kubernetes.io/version: "{{ .Chart.AppVersion }}"
+        app.kubernetes.io/managed-by: Helm
+      namespace: "{{ .Release.Namespace }}"
+    spec:
+      selector:
+        matchLabels:
+          app.kubernetes.io/name: "neon-storage-broker"
+      endpoints:
+        - port: broker
+          path: /metrics
+          interval: 10s
+          scrapeTimeout: 10s
+      namespaceSelector:
+        matchNames:
+          - "{{ .Release.Namespace }}"
+
+settings:
+  sentryEnvironment: "production"
--- a/.github/helm-values/prod-us-east-1-theta.pg-sni-router.yaml
+++ b/.github/helm-values/prod-us-east-1-theta.pg-sni-router.yaml
@@ -0,0 +1,19 @@
+useCertManager: true
+
+replicaCount: 3
+
+exposedService:
+  # exposedService.port -- Exposed Service proxy port
+  port: 4432
+  annotations:
+    external-dns.alpha.kubernetes.io/hostname: "*.snirouter.theta.us-east-1.internal.aws.neon.tech"
+
+settings:
+  domain: "*.snirouter.theta.us-east-1.internal.aws.neon.tech"
+  sentryEnvironment: "production"
+
+imagePullSecrets:
+  - name: docker-hub-neon
+
+metrics:
+  enabled: false
--- a/.github/helm-values/prod-us-east-2-delta.neon-proxy-link.yaml
+++ b/.github/helm-values/prod-us-east-2-delta.neon-proxy-link.yaml
@@ -0,0 +1,58 @@
+# Helm chart values for neon-proxy-link.
+# This is a YAML-formatted file.
+
+image:
+  repository: neondatabase/neon
+
+settings:
+  authBackend: "link"
+  authEndpoint: "https://console.neon.tech/authenticate_proxy_request/"
+  uri: "https://console.neon.tech/psql_session/"
+  domain: "pg.neon.tech"
+  sentryEnvironment: "production"
+
+# -- Additional labels for zenith-proxy pods
+podLabels:
+  neon_service: proxy
+  neon_env: production
+  neon_region: us-east-2
+
+service:
+  type: LoadBalancer
+  annotations:
+    service.beta.kubernetes.io/aws-load-balancer-type: external
+    service.beta.kubernetes.io/aws-load-balancer-nlb-target-type: ip
+    service.beta.kubernetes.io/aws-load-balancer-scheme: internal
+    external-dns.alpha.kubernetes.io/hostname: neon-proxy-link-mgmt.delta.us-east-2.aws.neon.tech
+
+exposedService:
+  annotations:
+    service.beta.kubernetes.io/aws-load-balancer-type: external
+    service.beta.kubernetes.io/aws-load-balancer-nlb-target-type: ip
+    service.beta.kubernetes.io/aws-load-balancer-scheme: internet-facing
+    external-dns.alpha.kubernetes.io/hostname: neon-proxy-link.delta.us-east-2.aws.neon.tech
+
+extraManifests:
+  - apiVersion: operator.victoriametrics.com/v1beta1
+    kind: VMServiceScrape
+    metadata:
+      name: "{{ include \"neon-proxy.fullname\" . }}"
+      labels:
+        helm.sh/chart: neon-proxy-{{ .Chart.Version }}
+        app.kubernetes.io/name: neon-proxy
+        app.kubernetes.io/instance: "{{ include \"neon-proxy.fullname\" . }}"
+        app.kubernetes.io/version: "{{ .Chart.AppVersion }}"
+        app.kubernetes.io/managed-by: Helm
+      namespace: "{{ .Release.Namespace }}"
+    spec:
+      selector:
+        matchLabels:
+          app.kubernetes.io/name: "neon-proxy"
+      endpoints:
+        - port: http
+          path: /metrics
+          interval: 10s
+          scrapeTimeout: 10s
+      namespaceSelector:
+        matchNames:
+          - "{{ .Release.Namespace }}"
--- a/.github/helm-values/prod-us-east-2-delta.neon-proxy-scram.yaml
+++ b/.github/helm-values/prod-us-east-2-delta.neon-proxy-scram.yaml
@@ -0,0 +1,77 @@
+# Helm chart values for neon-proxy-scram.
+# This is a YAML-formatted file.
+
+deploymentStrategy:
+  type: RollingUpdate
+  rollingUpdate:
+    maxSurge: 100%
+    maxUnavailable: 50%
+
+# Delay the kill signal by 5 minutes (5 * 60)
+# The pod(s) will stay in Terminating, keeps the existing connections
+# but doesn't receive new ones
+containerLifecycle:
+  preStop:
+    exec:
+      command: ["/bin/sh", "-c", "sleep 300"]
+terminationGracePeriodSeconds: 604800
+
+
+image:
+  repository: neondatabase/neon
+
+settings:
+  authBackend: "console"
+  authEndpoint: "http://neon-internal-api.aws.neon.tech/management/api/v2"
+  domain: "*.us-east-2.aws.neon.tech"
+  extraDomains: ["*.us-east-2.retooldb.com", "*.us-east-2.postgres.vercel-storage.com"]
+  sentryEnvironment: "production"
+  wssPort: 8443
+  metricCollectionEndpoint: "http://neon-internal-api.aws.neon.tech/billing/api/v1/usage_events"
+  metricCollectionInterval: "10min"
+
+# -- Additional labels for neon-proxy pods
+podLabels:
+  neon_service: proxy-scram
+  neon_env: prod
+  neon_region: us-east-2
+
+exposedService:
+  annotations:
+    service.beta.kubernetes.io/aws-load-balancer-type: external
+    service.beta.kubernetes.io/aws-load-balancer-nlb-target-type: ip
+    service.beta.kubernetes.io/aws-load-balancer-scheme: internet-facing
+    external-dns.alpha.kubernetes.io/hostname: us-east-2.aws.neon.tech
+  httpsPort: 443
+
+#metrics:
+#  enabled: true
+#  serviceMonitor:
+#    enabled: true
+#    selector:
+#      release: kube-prometheus-stack
+
+extraManifests:
+  - apiVersion: operator.victoriametrics.com/v1beta1
+    kind: VMServiceScrape
+    metadata:
+      name: "{{ include \"neon-proxy.fullname\" . }}"
+      labels:
+        helm.sh/chart: neon-proxy-{{ .Chart.Version }}
+        app.kubernetes.io/name: neon-proxy
+        app.kubernetes.io/instance: "{{ include \"neon-proxy.fullname\" . }}"
+        app.kubernetes.io/version: "{{ .Chart.AppVersion }}"
+        app.kubernetes.io/managed-by: Helm
+      namespace: "{{ .Release.Namespace }}"
+    spec:
+      selector:
+        matchLabels:
+          app.kubernetes.io/name: "neon-proxy"
+      endpoints:
+        - port: http
+          path: /metrics
+          interval: 10s
+          scrapeTimeout: 10s
+      namespaceSelector:
+        matchNames:
+          - "{{ .Release.Namespace }}"
--- a/.github/helm-values/prod-us-east-2-delta.neon-storage-broker.yaml
+++ b/.github/helm-values/prod-us-east-2-delta.neon-storage-broker.yaml
@@ -0,0 +1,52 @@
+# Helm chart values for neon-storage-broker
+podLabels:
+  neon_env: production
+  neon_service: storage-broker
+
+# Use L4 LB
+service:
+  # service.annotations -- Annotations to add to the service
+  annotations:
+    service.beta.kubernetes.io/aws-load-balancer-type: external  # use newer AWS Load Balancer Controller
+    service.beta.kubernetes.io/aws-load-balancer-nlb-target-type: ip
+    service.beta.kubernetes.io/aws-load-balancer-scheme: internal  # deploy LB to private subnet
+    # assign service to this name at external-dns
+    external-dns.alpha.kubernetes.io/hostname: storage-broker-lb.delta.us-east-2.internal.aws.neon.tech
+  # service.type -- Service type
+  type: LoadBalancer
+  # service.port -- broker listen port
+  port: 50051
+
+ingress:
+  enabled: false
+
+metrics:
+  enabled: false
+
+extraManifests:
+  - apiVersion: operator.victoriametrics.com/v1beta1
+    kind: VMServiceScrape
+    metadata:
+      name: "{{ include \"neon-storage-broker.fullname\" . }}"
+      labels:
+        helm.sh/chart: neon-storage-broker-{{ .Chart.Version }}
+        app.kubernetes.io/name: neon-storage-broker
+        app.kubernetes.io/instance: neon-storage-broker
+        app.kubernetes.io/version: "{{ .Chart.AppVersion }}"
+        app.kubernetes.io/managed-by: Helm
+      namespace: "{{ .Release.Namespace }}"
+    spec:
+      selector:
+        matchLabels:
+          app.kubernetes.io/name: "neon-storage-broker"
+      endpoints:
+        - port: broker
+          path: /metrics
+          interval: 10s
+          scrapeTimeout: 10s
+      namespaceSelector:
+        matchNames:
+          - "{{ .Release.Namespace }}"
+
+settings:
+  sentryEnvironment: "production"
--- a/.github/helm-values/prod-us-east-2-delta.pg-sni-router.yaml
+++ b/.github/helm-values/prod-us-east-2-delta.pg-sni-router.yaml
@@ -0,0 +1,19 @@
+useCertManager: true
+
+replicaCount: 3
+
+exposedService:
+  # exposedService.port -- Exposed Service proxy port
+  port: 4432
+  annotations:
+    external-dns.alpha.kubernetes.io/hostname: "*.snirouter.delta.us-east-2.internal.aws.neon.tech"
+
+settings:
+  domain: "*.snirouter.delta.us-east-2.internal.aws.neon.tech"
+  sentryEnvironment: "production"
+
+imagePullSecrets:
+  - name: docker-hub-neon
+
+metrics:
+  enabled: false
--- a/.github/helm-values/prod-us-west-2-eta.neon-proxy-scram-legacy.yaml
+++ b/.github/helm-values/prod-us-west-2-eta.neon-proxy-scram-legacy.yaml
@@ -0,0 +1,76 @@
+# Helm chart values for neon-proxy-scram.
+# This is a YAML-formatted file.
+
+deploymentStrategy:
+  type: RollingUpdate
+  rollingUpdate:
+    maxSurge: 100%
+    maxUnavailable: 50%
+
+# Delay the kill signal by 5 minutes (5 * 60)
+# The pod(s) will stay in Terminating, keeps the existing connections
+# but doesn't receive new ones
+containerLifecycle:
+  preStop:
+    exec:
+      command: ["/bin/sh", "-c", "sleep 300"]
+terminationGracePeriodSeconds: 604800
+
+
+image:
+  repository: neondatabase/neon
+
+settings:
+  authBackend: "console"
+  authEndpoint: "http://neon-internal-api.aws.neon.tech/management/api/v2"
+  domain: "*.cloud.neon.tech"
+  sentryEnvironment: "production"
+  wssPort: 8443
+  metricCollectionEndpoint: "http://neon-internal-api.aws.neon.tech/billing/api/v1/usage_events"
+  metricCollectionInterval: "10min"
+
+# -- Additional labels for neon-proxy pods
+podLabels:
+  neon_service: proxy-scram
+  neon_env: prod
+  neon_region: us-west-2
+
+exposedService:
+  annotations:
+    service.beta.kubernetes.io/aws-load-balancer-type: external
+    service.beta.kubernetes.io/aws-load-balancer-nlb-target-type: ip
+    service.beta.kubernetes.io/aws-load-balancer-scheme: internet-facing
+    external-dns.alpha.kubernetes.io/hostname: neon-proxy-scram-legacy.eta.us-west-2.aws.neon.tech
+  httpsPort: 443
+
+#metrics:
+#  enabled: true
+#  serviceMonitor:
+#    enabled: true
+#    selector:
+#      release: kube-prometheus-stack
+
+extraManifests:
+  - apiVersion: operator.victoriametrics.com/v1beta1
+    kind: VMServiceScrape
+    metadata:
+      name: "{{ include \"neon-proxy.fullname\" . }}"
+      labels:
+        helm.sh/chart: neon-proxy-{{ .Chart.Version }}
+        app.kubernetes.io/name: neon-proxy
+        app.kubernetes.io/instance: "{{ include \"neon-proxy.fullname\" . }}"
+        app.kubernetes.io/version: "{{ .Chart.AppVersion }}"
+        app.kubernetes.io/managed-by: Helm
+      namespace: "{{ .Release.Namespace }}"
+    spec:
+      selector:
+        matchLabels:
+          app.kubernetes.io/name: "neon-proxy"
+      endpoints:
+        - port: http
+          path: /metrics
+          interval: 10s
+          scrapeTimeout: 10s
+      namespaceSelector:
+        matchNames:
+          - "{{ .Release.Namespace }}"
--- a/.github/helm-values/prod-us-west-2-eta.neon-proxy-scram.yaml
+++ b/.github/helm-values/prod-us-west-2-eta.neon-proxy-scram.yaml
@@ -0,0 +1,77 @@
+# Helm chart values for neon-proxy-scram.
+# This is a YAML-formatted file.
+
+deploymentStrategy:
+  type: RollingUpdate
+  rollingUpdate:
+    maxSurge: 100%
+    maxUnavailable: 50%
+
+# Delay the kill signal by 5 minutes (5 * 60)
+# The pod(s) will stay in Terminating, keeps the existing connections
+# but doesn't receive new ones
+containerLifecycle:
+  preStop:
+    exec:
+      command: ["/bin/sh", "-c", "sleep 300"]
+terminationGracePeriodSeconds: 604800
+
+
+image:
+  repository: neondatabase/neon
+
+settings:
+  authBackend: "console"
+  authEndpoint: "http://neon-internal-api.aws.neon.tech/management/api/v2"
+  domain: "*.us-west-2.aws.neon.tech"
+  extraDomains: ["*.us-west-2.retooldb.com", "*.us-west-2.postgres.vercel-storage.com"]
+  sentryEnvironment: "production"
+  wssPort: 8443
+  metricCollectionEndpoint: "http://neon-internal-api.aws.neon.tech/billing/api/v1/usage_events"
+  metricCollectionInterval: "10min"
+
+# -- Additional labels for neon-proxy pods
+podLabels:
+  neon_service: proxy-scram
+  neon_env: prod
+  neon_region: us-west-2
+
+exposedService:
+  annotations:
+    service.beta.kubernetes.io/aws-load-balancer-type: external
+    service.beta.kubernetes.io/aws-load-balancer-nlb-target-type: ip
+    service.beta.kubernetes.io/aws-load-balancer-scheme: internet-facing
+    external-dns.alpha.kubernetes.io/hostname: us-west-2.aws.neon.tech
+  httpsPort: 443
+
+#metrics:
+#  enabled: true
+#  serviceMonitor:
+#    enabled: true
+#    selector:
+#      release: kube-prometheus-stack
+
+extraManifests:
+  - apiVersion: operator.victoriametrics.com/v1beta1
+    kind: VMServiceScrape
+    metadata:
+      name: "{{ include \"neon-proxy.fullname\" . }}"
+      labels:
+        helm.sh/chart: neon-proxy-{{ .Chart.Version }}
+        app.kubernetes.io/name: neon-proxy
+        app.kubernetes.io/instance: "{{ include \"neon-proxy.fullname\" . }}"
+        app.kubernetes.io/version: "{{ .Chart.AppVersion }}"
+        app.kubernetes.io/managed-by: Helm
+      namespace: "{{ .Release.Namespace }}"
+    spec:
+      selector:
+        matchLabels:
+          app.kubernetes.io/name: "neon-proxy"
+      endpoints:
+        - port: http
+          path: /metrics
+          interval: 10s
+          scrapeTimeout: 10s
+      namespaceSelector:
+        matchNames:
+          - "{{ .Release.Namespace }}"
--- a/.github/helm-values/prod-us-west-2-eta.neon-storage-broker.yaml
+++ b/.github/helm-values/prod-us-west-2-eta.neon-storage-broker.yaml
@@ -0,0 +1,52 @@
+# Helm chart values for neon-storage-broker
+podLabels:
+  neon_env: production
+  neon_service: storage-broker
+
+# Use L4 LB
+service:
+  # service.annotations -- Annotations to add to the service
+  annotations:
+    service.beta.kubernetes.io/aws-load-balancer-type: external  # use newer AWS Load Balancer Controller
+    service.beta.kubernetes.io/aws-load-balancer-nlb-target-type: ip
+    service.beta.kubernetes.io/aws-load-balancer-scheme: internal  # deploy LB to private subnet
+    # assign service to this name at external-dns
+    external-dns.alpha.kubernetes.io/hostname: storage-broker-lb.eta.us-west-2.internal.aws.neon.tech
+  # service.type -- Service type
+  type: LoadBalancer
+  # service.port -- broker listen port
+  port: 50051
+
+ingress:
+  enabled: false
+
+metrics:
+  enabled: false
+
+extraManifests:
+  - apiVersion: operator.victoriametrics.com/v1beta1
+    kind: VMServiceScrape
+    metadata:
+      name: "{{ include \"neon-storage-broker.fullname\" . }}"
+      labels:
+        helm.sh/chart: neon-storage-broker-{{ .Chart.Version }}
+        app.kubernetes.io/name: neon-storage-broker
+        app.kubernetes.io/instance: neon-storage-broker
+        app.kubernetes.io/version: "{{ .Chart.AppVersion }}"
+        app.kubernetes.io/managed-by: Helm
+      namespace: "{{ .Release.Namespace }}"
+    spec:
+      selector:
+        matchLabels:
+          app.kubernetes.io/name: "neon-storage-broker"
+      endpoints:
+        - port: broker
+          path: /metrics
+          interval: 10s
+          scrapeTimeout: 10s
+      namespaceSelector:
+        matchNames:
+          - "{{ .Release.Namespace }}"
+
+settings:
+  sentryEnvironment: "production"
--- a/.github/helm-values/prod-us-west-2-eta.pg-sni-router.yaml
+++ b/.github/helm-values/prod-us-west-2-eta.pg-sni-router.yaml
@@ -0,0 +1,19 @@
+useCertManager: true
+
+replicaCount: 3
+
+exposedService:
+  # exposedService.port -- Exposed Service proxy port
+  port: 4432
+  annotations:
+    external-dns.alpha.kubernetes.io/hostname: "*.snirouter.eta.us-west-2.internal.aws.neon.tech"
+
+settings:
+  domain: "*.snirouter.eta.us-west-2.internal.aws.neon.tech"
+  sentryEnvironment: "production"
+
+imagePullSecrets:
+  - name: docker-hub-neon
+
+metrics:
+  enabled: false
--- a/.github/workflows/benchmarking.yml
+++ b/.github/workflows/benchmarking.yml
@@ -93,7 +93,10 @@ jobs:

    - name: Create Allure report
      if: ${{ !cancelled() }}
-      uses: ./.github/actions/allure-report-generate
+      uses: ./.github/actions/allure-report
+      with:
+        action: generate
+        build_type: ${{ env.BUILD_TYPE }}

    - name: Post to a Slack channel
      if: ${{ github.event.schedule && failure() }}
@@ -280,7 +283,10 @@ jobs:

    - name: Create Allure report
      if: ${{ !cancelled() }}
-      uses: ./.github/actions/allure-report-generate
+      uses: ./.github/actions/allure-report
+      with:
+        action: generate
+        build_type: ${{ env.BUILD_TYPE }}

    - name: Post to a Slack channel
      if: ${{ github.event.schedule && failure() }}
@@ -374,7 +380,10 @@ jobs:

    - name: Create Allure report
      if: ${{ !cancelled() }}
-      uses: ./.github/actions/allure-report-generate
+      uses: ./.github/actions/allure-report
+      with:
+        action: generate
+        build_type: ${{ env.BUILD_TYPE }}

    - name: Post to a Slack channel
      if: ${{ github.event.schedule && failure() }}
@@ -467,7 +476,10 @@ jobs:

    - name: Create Allure report
      if: ${{ !cancelled() }}
-      uses: ./.github/actions/allure-report-generate
+      uses: ./.github/actions/allure-report
+      with:
+        action: generate
+        build_type: ${{ env.BUILD_TYPE }}

    - name: Post to a Slack channel
      if: ${{ github.event.schedule && failure() }}
@@ -554,13 +566,16 @@ jobs:

    - name: Create Allure report
      if: ${{ !cancelled() }}
-      uses: ./.github/actions/allure-report-generate
+      uses: ./.github/actions/allure-report
+      with:
+        action: generate
+        build_type: ${{ env.BUILD_TYPE }}

    - name: Post to a Slack channel
      if: ${{ github.event.schedule && failure() }}
      uses: slackapi/slack-github-action@v1
      with:
        channel-id: "C033QLM5P7D" # dev-staging-stream
-        slack-message: "Periodic User example perf testing ${{ matrix.platform }}: ${{ job.status }}\n${{ github.server_url }}/${{ github.repository }}/actions/runs/${{ github.run_id }}"
+        slack-message: "Periodic TPC-H perf testing ${{ matrix.platform }}: ${{ job.status }}\n${{ github.server_url }}/${{ github.repository }}/actions/runs/${{ github.run_id }}"
      env:
        SLACK_BOT_TOKEN: ${{ secrets.SLACK_BOT_TOKEN }}
--- a/.github/workflows/build_and_test.yml
+++ b/.github/workflows/build_and_test.yml
@@ -330,7 +330,6 @@ jobs:
      fail-fast: false
      matrix:
        build_type: [ debug, release ]
-        pg_version: [ v14, v15 ]
    steps:
      - name: Checkout
        uses: actions/checkout@v3
@@ -351,12 +350,11 @@ jobs:
          real_s3_secret_access_key: "${{ secrets.AWS_SECRET_ACCESS_KEY_CI_TESTS_S3 }}"
          rerun_flaky: true
        env:
-          DEFAULT_PG_VERSION: ${{ matrix.pg_version }}
          TEST_RESULT_CONNSTR: ${{ secrets.REGRESS_TEST_RESULT_CONNSTR }}
          CHECK_ONDISK_DATA_COMPATIBILITY: nonempty

      - name: Merge and upload coverage data
-        if: matrix.build_type == 'debug' && matrix.pg_version == 'v14'
+        if: matrix.build_type == 'debug'
        uses: ./.github/actions/save-coverage-data

  benchmarks:
@@ -401,50 +399,79 @@ jobs:
    steps:
      - uses: actions/checkout@v3

-      - name: Create Allure report
+      - name: Create Allure report (debug)
        if: ${{ !cancelled() }}
-        id: create-allure-report
-        uses: ./.github/actions/allure-report-generate
+        id: create-allure-report-debug
+        uses: ./.github/actions/allure-report
+        with:
+          action: generate
+          build_type: debug
+
+      - name: Create Allure report (release)
+        if: ${{ !cancelled() }}
+        id: create-allure-report-release
+        uses: ./.github/actions/allure-report
+        with:
+          action: generate
+          build_type: release

      - uses: actions/github-script@v6
        if: >
          !cancelled() &&
-          github.event_name == 'pull_request'
+          github.event_name == 'pull_request' && (
+            steps.create-allure-report-debug.outputs.report-url ||
+            steps.create-allure-report-release.outputs.report-url
+          )
        with:
          # Retry script for 5XX server errors: https://github.com/actions/github-script#retries
          retries: 5
          script: |
-            const report = {
-              reportUrl:     "${{ steps.create-allure-report.outputs.report-url }}",
-              reportJsonUrl: "${{ steps.create-allure-report.outputs.report-json-url }}",
-            }
+            const reports = [{
+              buildType: "debug",
+              reportUrl: "${{ steps.create-allure-report-debug.outputs.report-url }}",
+              jsonUrl:   "${{ steps.create-allure-report-debug.outputs.report-json-url }}",
+            }, {
+              buildType: "release",
+              reportUrl: "${{ steps.create-allure-report-release.outputs.report-url }}",
+              jsonUrl:   "${{ steps.create-allure-report-release.outputs.report-json-url }}",
+            }]

            const script = require("./scripts/pr-comment-test-report.js")
            await script({
              github,
              context,
              fetch,
-              report,
+              reports,
            })

      - name: Store Allure test stat in the DB
-        if: ${{ !cancelled() && steps.create-allure-report.outputs.report-json-url }}
+        if: >
+          !cancelled() && (
+            steps.create-allure-report-debug.outputs.report-url ||
+            steps.create-allure-report-release.outputs.report-url
+          )
        env:
-          COMMIT_SHA: ${{ github.event.pull_request.head.sha || github.sha }}
-          REPORT_JSON_URL: ${{ steps.create-allure-report.outputs.report-json-url }}
+          SHA: ${{ github.event.pull_request.head.sha || github.sha }}
+          REPORT_JSON_URL_DEBUG: ${{ steps.create-allure-report-debug.outputs.report-json-url }}
+          REPORT_JSON_URL_RELEASE: ${{ steps.create-allure-report-release.outputs.report-json-url }}
          TEST_RESULT_CONNSTR: ${{ secrets.REGRESS_TEST_RESULT_CONNSTR }}
        run: |
          ./scripts/pysync

-          curl --fail --output suites.json "${REPORT_JSON_URL}"
-          export BUILD_TYPE=unified
-          export DATABASE_URL="$TEST_RESULT_CONNSTR"
+          for report_url in $REPORT_JSON_URL_DEBUG $REPORT_JSON_URL_RELEASE; do
+            if [ -z "$report_url" ]; then
+              continue
+            fi

-          poetry run python3 scripts/ingest_regress_test_result.py \
-            --revision ${COMMIT_SHA} \
-            --reference ${GITHUB_REF} \
-            --build-type ${BUILD_TYPE} \
-            --ingest suites.json
+            if [[ "$report_url" == "$REPORT_JSON_URL_DEBUG" ]]; then
+              BUILD_TYPE=debug
+            else
+              BUILD_TYPE=release
+            fi
+
+            curl --fail --output suites.json "${report_url}"
+            DATABASE_URL="$TEST_RESULT_CONNSTR" poetry run python3 scripts/ingest_regress_test_result.py --revision ${SHA} --reference ${GITHUB_REF} --build-type ${BUILD_TYPE} --ingest suites.json
+          done

  coverage-report:
    runs-on: [ self-hosted, gen3, small ]
@@ -606,7 +633,6 @@ jobs:
                           --cache-repo 369495373322.dkr.ecr.eu-central-1.amazonaws.com/cache
                           --context .
                           --build-arg GIT_VERSION=${{ github.sha }}
-                           --build-arg REPOSITORY=369495373322.dkr.ecr.eu-central-1.amazonaws.com
                           --destination 369495373322.dkr.ecr.eu-central-1.amazonaws.com/neon:${{needs.tag.outputs.build-tag}}
                           --destination neondatabase/neon:${{needs.tag.outputs.build-tag}}

@@ -692,7 +718,6 @@ jobs:
                           --cache-repo 369495373322.dkr.ecr.eu-central-1.amazonaws.com/cache
                           --context .
                           --build-arg GIT_VERSION=${{ github.sha }}
-                           --build-arg REPOSITORY=369495373322.dkr.ecr.eu-central-1.amazonaws.com
                           --dockerfile Dockerfile.compute-tools
                           --destination 369495373322.dkr.ecr.eu-central-1.amazonaws.com/compute-tools:${{needs.tag.outputs.build-tag}}
                           --destination neondatabase/compute-tools:${{needs.tag.outputs.build-tag}}
@@ -745,7 +770,6 @@ jobs:
                           --context .
                           --build-arg GIT_VERSION=${{ github.sha }}
                           --build-arg PG_VERSION=${{ matrix.version }}
-                           --build-arg REPOSITORY=369495373322.dkr.ecr.eu-central-1.amazonaws.com
                           --dockerfile Dockerfile.compute-node
                           --destination 369495373322.dkr.ecr.eu-central-1.amazonaws.com/compute-node-${{ matrix.version }}:${{needs.tag.outputs.build-tag}}
                           --destination neondatabase/compute-node-${{ matrix.version }}:${{needs.tag.outputs.build-tag}}
@@ -911,6 +935,42 @@ jobs:
      - name: Cleanup ECR folder
        run: rm -rf ~/.ecr

+  deploy-pr-test-new:
+    runs-on: [ self-hosted, gen3, small ]
+    container: 369495373322.dkr.ecr.eu-central-1.amazonaws.com/ansible:pinned
+    # We need both storage **and** compute images for deploy, because control plane picks the compute version based on the storage version.
+    # If it notices a fresh storage it may bump the compute version. And if compute image failed to build it may break things badly
+    needs: [ promote-images, tag, regress-tests ]
+    if: |
+      contains(github.event.pull_request.labels.*.name, 'deploy-test-storage') &&
+      github.event_name != 'workflow_dispatch'
+    defaults:
+      run:
+        shell: bash
+    strategy:
+      matrix:
+        target_region: [ eu-west-1 ]
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v3
+        with:
+          submodules: true
+          fetch-depth: 0
+
+      - name: Redeploy
+        run: |
+          export DOCKER_TAG=${{needs.tag.outputs.build-tag}}
+          cd "$(pwd)/.github/ansible"
+
+          ./get_binaries.sh
+
+          ansible-galaxy collection install sivel.toiletwater
+          ansible-playbook deploy.yaml -i staging.${{ matrix.target_region }}.hosts.yaml -e @ssm_config -e CONSOLE_API_TOKEN=${{ secrets.NEON_STAGING_API_KEY }} -e SENTRY_URL_PAGESERVER=${{ secrets.SENTRY_URL_PAGESERVER }} -e SENTRY_URL_SAFEKEEPER=${{ secrets.SENTRY_URL_SAFEKEEPER }}
+          rm -f neon_install.tar.gz .neon_current_version
+
+      - name: Cleanup ansible folder
+        run: rm -rf ~/.ansible
+
  deploy:
    runs-on: [ self-hosted, gen3, small ]
    container: 369495373322.dkr.ecr.eu-central-1.amazonaws.com/ansible:latest
@@ -935,12 +995,12 @@ jobs:

      - name: Trigger deploy workflow
        env:
-          GH_TOKEN: ${{ secrets.CI_ACCESS_TOKEN }}
+          GH_TOKEN: ${{ github.token }}
        run: |
          if [[ "$GITHUB_REF_NAME" == "main" ]]; then
-            gh workflow --repo neondatabase/aws run deploy-dev.yml --ref main -f branch=main -f dockerTag=${{needs.tag.outputs.build-tag}}
+            gh workflow run deploy-dev.yml --ref main -f branch=${{ github.sha }} -f dockerTag=${{needs.tag.outputs.build-tag}}
          elif [[ "$GITHUB_REF_NAME" == "release" ]]; then
-            gh workflow --repo neondatabase/aws run deploy-prod.yml --ref main -f branch=main -f dockerTag=${{needs.tag.outputs.build-tag}} -f disclamerAcknowledged=true
+            gh workflow run deploy-prod.yml --ref release -f branch=${{ github.sha }} -f dockerTag=${{needs.tag.outputs.build-tag}} -f disclamerAcknowledged=true
          else
            echo "GITHUB_REF_NAME (value '$GITHUB_REF_NAME') is not set to either 'main' or 'release'"
            exit 1
--- a/.github/workflows/deploy-dev.yml
+++ b/.github/workflows/deploy-dev.yml
@@ -0,0 +1,280 @@
+name: Neon Deploy dev
+
+on:
+  workflow_dispatch:
+    inputs:
+      dockerTag:
+        description: 'Docker tag to deploy'
+        required: true
+        type: string
+      branch:
+        description: 'Branch or commit used for deploy scripts and configs'
+        required: true
+        type: string
+        default: 'main'
+      deployStorage:
+        description: 'Deploy storage'
+        required: true
+        type: boolean
+        default: true
+      deployProxy:
+        description: 'Deploy proxy'
+        required: true
+        type: boolean
+        default: true
+      deployStorageBroker:
+        description: 'Deploy storage-broker'
+        required: true
+        type: boolean
+        default: true
+      deployPgSniRouter:
+        description: 'Deploy pg-sni-router'
+        required: true
+        type: boolean
+        default: true
+
+env:
+  AWS_ACCESS_KEY_ID: ${{ secrets.AWS_ACCESS_KEY_DEV }}
+  AWS_SECRET_ACCESS_KEY: ${{ secrets.AWS_SECRET_KEY_DEV }}
+
+concurrency:
+  group: deploy-dev
+  cancel-in-progress: false
+
+jobs:
+  deploy-storage-new:
+    runs-on: [ self-hosted, gen3, small ]
+    container:
+      image: 369495373322.dkr.ecr.eu-central-1.amazonaws.com/ansible:pinned
+      options: --user root --privileged
+    if: inputs.deployStorage
+    defaults:
+      run:
+        shell: bash
+    strategy:
+      matrix:
+        # TODO(sergey): Fix storage deploy in eu-central-1
+        target_region: [ eu-west-1, us-east-2]
+    environment:
+      name: dev-${{ matrix.target_region }}
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v3
+        with:
+          submodules: true
+          fetch-depth: 0
+          ref: ${{ inputs.branch }}
+
+      - name: Redeploy
+        run: |
+          export DOCKER_TAG=${{ inputs.dockerTag }}
+          cd "$(pwd)/.github/ansible"
+
+          ./get_binaries.sh
+
+          ansible-galaxy collection install sivel.toiletwater
+          ansible-playbook -v deploy.yaml -i staging.${{ matrix.target_region }}.hosts.yaml -e @ssm_config -e CONSOLE_API_TOKEN=${{ secrets.NEON_STAGING_API_KEY }} -e SENTRY_URL_PAGESERVER=${{ secrets.SENTRY_URL_PAGESERVER }} -e SENTRY_URL_SAFEKEEPER=${{ secrets.SENTRY_URL_SAFEKEEPER }}
+          rm -f neon_install.tar.gz .neon_current_version
+
+      - name: Cleanup ansible folder
+        run: rm -rf ~/.ansible
+
+  deploy-proxy-new:
+    runs-on: [ self-hosted, gen3, small ]
+    container: 369495373322.dkr.ecr.eu-central-1.amazonaws.com/ansible:pinned
+    if: inputs.deployProxy
+    defaults:
+      run:
+        shell: bash
+    strategy:
+      matrix:
+        include:
+          - target_region:  us-east-2
+            target_cluster: dev-us-east-2-beta
+            deploy_link_proxy: true
+            deploy_legacy_scram_proxy: true
+          - target_region:  eu-west-1
+            target_cluster: dev-eu-west-1-zeta
+            deploy_link_proxy: false
+            deploy_legacy_scram_proxy: false
+    environment:
+      name: dev-${{ matrix.target_region }}
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v3
+        with:
+          submodules: true
+          fetch-depth: 0
+          ref: ${{ inputs.branch }}
+  
+      - name: Configure AWS Credentials
+        uses: aws-actions/configure-aws-credentials@v1-node16
+        with:
+          role-to-assume: arn:aws:iam::369495373322:role/github-runner
+          aws-region: eu-central-1
+          role-skip-session-tagging: true
+          role-duration-seconds: 1800
+  
+      - name: Configure environment
+        run: |
+          helm repo add neondatabase https://neondatabase.github.io/helm-charts
+          aws --region ${{ matrix.target_region }} eks update-kubeconfig --name  ${{ matrix.target_cluster }}
+  
+      - name: Re-deploy scram proxy
+        run: |
+          DOCKER_TAG=${{ inputs.dockerTag }}
+          helm upgrade neon-proxy-scram neondatabase/neon-proxy --namespace neon-proxy --create-namespace --install --atomic -f .github/helm-values/${{ matrix.target_cluster }}.neon-proxy-scram.yaml --set image.tag=${DOCKER_TAG} --set settings.sentryUrl=${{ secrets.SENTRY_URL_PROXY }} --wait --timeout 15m0s
+  
+      - name: Re-deploy link proxy
+        if: matrix.deploy_link_proxy
+        run: |
+          DOCKER_TAG=${{ inputs.dockerTag }}
+          helm upgrade neon-proxy-link neondatabase/neon-proxy --namespace neon-proxy --create-namespace --install --atomic -f .github/helm-values/${{ matrix.target_cluster }}.neon-proxy-link.yaml --set image.tag=${DOCKER_TAG} --set settings.sentryUrl=${{ secrets.SENTRY_URL_PROXY }} --wait --timeout 15m0s
+  
+      - name: Re-deploy legacy scram proxy
+        if: matrix.deploy_legacy_scram_proxy
+        run: |
+          DOCKER_TAG=${{ inputs.dockerTag }}
+          helm upgrade neon-proxy-scram-legacy neondatabase/neon-proxy --namespace neon-proxy --create-namespace --install --atomic -f .github/helm-values/${{ matrix.target_cluster }}.neon-proxy-scram-legacy.yaml --set image.tag=${DOCKER_TAG} --set settings.sentryUrl=${{ secrets.SENTRY_URL_PROXY }} --wait --timeout 15m0s
+  
+      - name: Cleanup helm folder
+        run: rm -rf ~/.cache
+
+  deploy-preview-proxy-new:
+    runs-on: [ self-hosted, gen3, small ]
+    container: 369495373322.dkr.ecr.eu-central-1.amazonaws.com/ansible:pinned
+    if: inputs.deployProxy
+    defaults:
+      run:
+        shell: bash
+    strategy:
+      matrix:
+        include:
+          - target_region:  eu-central-1
+            target_cluster: dev-eu-central-1-alpha
+    environment:
+      name: dev-${{ matrix.target_region }}
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v3
+        with:
+          submodules: true
+          fetch-depth: 0
+          ref: ${{ inputs.branch }}
+  
+      - name: Configure AWS Credentials
+        uses: aws-actions/configure-aws-credentials@v1-node16
+        with:
+          role-to-assume: arn:aws:iam::369495373322:role/github-runner
+          aws-region: eu-central-1
+          role-skip-session-tagging: true
+          role-duration-seconds: 1800
+  
+      - name: Configure environment
+        run: |
+          helm repo add neondatabase https://neondatabase.github.io/helm-charts
+          aws --region ${{ matrix.target_region }} eks update-kubeconfig --name  ${{ matrix.target_cluster }}
+  
+      - name: Re-deploy preview proxies
+        run: |
+          DOCKER_TAG=${{ inputs.dockerTag }}
+          for PREVIEW_NAME in helium argon krypton xenon radon oganesson hydrogen nitrogen oxygen fluorine chlorine; do
+            export PREVIEW_NAME
+            envsubst <.github/helm-values/preview-template.neon-proxy-scram.yaml >preview-${PREVIEW_NAME}.neon-proxy-scram.yaml
+            helm upgrade neon-proxy-scram-${PREVIEW_NAME} neondatabase/neon-proxy --namespace neon-proxy-${PREVIEW_NAME} --create-namespace --install --atomic -f preview-${PREVIEW_NAME}.neon-proxy-scram.yaml --set image.tag=${DOCKER_TAG} --set settings.sentryUrl=${{ secrets.SENTRY_URL_PROXY }} --wait --timeout 15m0s
+          done
+
+      - name: Cleanup helm folder
+        run: rm -rf ~/.cache
+  
+  deploy-storage-broker-new:
+    runs-on: [ self-hosted, gen3, small ]
+    container: 369495373322.dkr.ecr.eu-central-1.amazonaws.com/ansible:pinned
+    if: inputs.deployStorageBroker
+    defaults:
+      run:
+        shell: bash
+    strategy:
+      matrix:
+        include:
+          - target_region:  us-east-2
+            target_cluster: dev-us-east-2-beta
+          - target_region:  eu-west-1
+            target_cluster: dev-eu-west-1-zeta
+          - target_region:  eu-central-1
+            target_cluster: dev-eu-central-1-alpha
+    environment:
+      name: dev-${{ matrix.target_region }}
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v3
+        with:
+          submodules: true
+          fetch-depth: 0
+          ref: ${{ inputs.branch }}
+  
+      - name: Configure AWS Credentials
+        uses: aws-actions/configure-aws-credentials@v1-node16
+        with:
+          role-to-assume: arn:aws:iam::369495373322:role/github-runner
+          aws-region: eu-central-1
+          role-skip-session-tagging: true
+          role-duration-seconds: 1800
+  
+      - name: Configure environment
+        run: |
+          helm repo add neondatabase https://neondatabase.github.io/helm-charts
+          aws --region ${{ matrix.target_region }} eks update-kubeconfig --name  ${{ matrix.target_cluster }}
+  
+      - name: Deploy storage-broker
+        run:
+          helm upgrade neon-storage-broker-lb neondatabase/neon-storage-broker --namespace neon-storage-broker-lb --create-namespace --install --atomic -f .github/helm-values/${{ matrix.target_cluster }}.neon-storage-broker.yaml --set image.tag=${{ inputs.dockerTag }} --set settings.sentryUrl=${{ secrets.SENTRY_URL_BROKER }} --wait --timeout 5m0s
+  
+      - name: Cleanup helm folder
+        run: rm -rf ~/.cache
+
+  deploy-pg-sni-router:
+    runs-on: [ self-hosted, gen3, small ]
+    container: 369495373322.dkr.ecr.eu-central-1.amazonaws.com/ansible:pinned
+    if: inputs.deployPgSniRouter
+    defaults:
+      run:
+        shell: bash
+    strategy:
+      matrix:
+        include:
+          - target_region:  us-east-2
+            target_cluster: dev-us-east-2-beta
+          - target_region:  eu-west-1
+            target_cluster: dev-eu-west-1-zeta
+          - target_region:  eu-central-1
+            target_cluster: dev-eu-central-1-alpha
+    environment:
+      name: dev-${{ matrix.target_region }}
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v3
+        with:
+          submodules: true
+          fetch-depth: 0
+          ref: ${{ inputs.branch }}
+  
+      - name: Configure AWS Credentials
+        uses: aws-actions/configure-aws-credentials@v1-node16
+        with:
+          role-to-assume: arn:aws:iam::369495373322:role/github-runner
+          aws-region: eu-central-1
+          role-skip-session-tagging: true
+          role-duration-seconds: 1800
+  
+      - name: Configure environment
+        run: |
+          helm repo add neondatabase https://neondatabase.github.io/helm-charts
+          aws --region ${{ matrix.target_region }} eks update-kubeconfig --name  ${{ matrix.target_cluster }}
+  
+      - name: Deploy pg-sni-router
+        run:
+          helm upgrade neon-pg-sni-router neondatabase/neon-pg-sni-router --namespace neon-pg-sni-router --create-namespace --install --atomic -f .github/helm-values/${{ matrix.target_cluster }}.pg-sni-router.yaml --set image.tag=${{ inputs.dockerTag }} --set settings.sentryUrl=${{ secrets.SENTRY_URL_BROKER }} --wait --timeout 15m0s
+  
+      - name: Cleanup helm folder
+        run: rm -rf ~/.cache
--- a/.github/workflows/deploy-prod.yml
+++ b/.github/workflows/deploy-prod.yml
@@ -0,0 +1,217 @@
+name: Neon Deploy prod
+
+on:
+  workflow_dispatch:
+    inputs:
+      dockerTag:
+        description: 'Docker tag to deploy'
+        required: true
+        type: string
+      branch:
+        description: 'Branch or commit used for deploy scripts and configs'
+        required: true
+        type: string
+        default: 'release'
+      deployStorage:
+        description: 'Deploy storage'
+        required: true
+        type: boolean
+        default: true
+      deployProxy:
+        description: 'Deploy proxy'
+        required: true
+        type: boolean
+        default: true
+      deployStorageBroker:
+        description: 'Deploy storage-broker'
+        required: true
+        type: boolean
+        default: true
+      deployPgSniRouter:
+        description: 'Deploy pg-sni-router'
+        required: true
+        type: boolean
+        default: true
+      disclamerAcknowledged:
+        description: 'I confirm that there is an emergency and I can not use regular release workflow'
+        required: true
+        type: boolean
+        default: false
+
+concurrency:
+  group: deploy-prod
+  cancel-in-progress: false
+
+jobs:
+  deploy-prod-new:
+    runs-on: prod
+    container:
+      image: 093970136003.dkr.ecr.eu-central-1.amazonaws.com/ansible:latest
+      options: --user root --privileged
+    if: inputs.deployStorage && inputs.disclamerAcknowledged
+    defaults:
+      run:
+        shell: bash
+    strategy:
+      matrix:
+        target_region: [ us-east-2, us-west-2, eu-central-1, ap-southeast-1, us-east-1 ]
+    environment:
+      name: prod-${{ matrix.target_region }}
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v3
+        with:
+          submodules: true
+          fetch-depth: 0
+          ref: ${{ inputs.branch }}
+
+      - name: Redeploy
+        run: |
+          export DOCKER_TAG=${{ inputs.dockerTag }}
+          cd "$(pwd)/.github/ansible"
+
+          ./get_binaries.sh
+
+          ansible-galaxy collection install sivel.toiletwater
+          ansible-playbook -v deploy.yaml -i prod.${{ matrix.target_region }}.hosts.yaml -e @ssm_config -e CONSOLE_API_TOKEN=${{ secrets.NEON_PRODUCTION_API_KEY }} -e SENTRY_URL_PAGESERVER=${{ secrets.SENTRY_URL_PAGESERVER }} -e SENTRY_URL_SAFEKEEPER=${{ secrets.SENTRY_URL_SAFEKEEPER }}
+          rm -f neon_install.tar.gz .neon_current_version
+
+  deploy-proxy-prod-new:
+    runs-on: prod
+    container: 093970136003.dkr.ecr.eu-central-1.amazonaws.com/ansible:latest
+    if: inputs.deployProxy && inputs.disclamerAcknowledged
+    defaults:
+      run:
+        shell: bash
+    strategy:
+      matrix:
+        include:
+          - target_region:  us-east-2
+            target_cluster: prod-us-east-2-delta
+            deploy_link_proxy: true
+            deploy_legacy_scram_proxy: false
+          - target_region:  us-west-2
+            target_cluster: prod-us-west-2-eta
+            deploy_link_proxy: false
+            deploy_legacy_scram_proxy: true
+          - target_region: eu-central-1
+            target_cluster: prod-eu-central-1-gamma
+            deploy_link_proxy: false
+            deploy_legacy_scram_proxy: false
+          - target_region: ap-southeast-1
+            target_cluster: prod-ap-southeast-1-epsilon
+            deploy_link_proxy: false
+            deploy_legacy_scram_proxy: false
+          - target_region: us-east-1
+            target_cluster: prod-us-east-1-theta
+            deploy_link_proxy: false
+            deploy_legacy_scram_proxy: false
+    environment:
+      name: prod-${{ matrix.target_region }}
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v3
+        with:
+          submodules: true
+          fetch-depth: 0
+          ref: ${{ inputs.branch }}
+
+      - name: Configure environment
+        run: |
+          helm repo add neondatabase https://neondatabase.github.io/helm-charts
+          aws --region ${{ matrix.target_region }} eks update-kubeconfig --name  ${{ matrix.target_cluster }}
+
+      - name: Re-deploy scram proxy
+        run: |
+          DOCKER_TAG=${{ inputs.dockerTag }}
+          helm upgrade neon-proxy-scram neondatabase/neon-proxy --namespace neon-proxy --create-namespace --install --atomic -f .github/helm-values/${{ matrix.target_cluster }}.neon-proxy-scram.yaml --set image.tag=${DOCKER_TAG} --set settings.sentryUrl=${{ secrets.SENTRY_URL_PROXY }} --wait --timeout 15m0s
+
+      - name: Re-deploy link proxy
+        if: matrix.deploy_link_proxy
+        run: |
+          DOCKER_TAG=${{ inputs.dockerTag }}
+          helm upgrade neon-proxy-link neondatabase/neon-proxy --namespace neon-proxy --create-namespace --install --atomic -f .github/helm-values/${{ matrix.target_cluster }}.neon-proxy-link.yaml --set image.tag=${DOCKER_TAG} --set settings.sentryUrl=${{ secrets.SENTRY_URL_PROXY }} --wait --timeout 15m0s
+
+      - name: Re-deploy legacy scram proxy
+        if: matrix.deploy_legacy_scram_proxy
+        run: |
+          DOCKER_TAG=${{ inputs.dockerTag }}
+          helm upgrade neon-proxy-scram-legacy neondatabase/neon-proxy --namespace neon-proxy --create-namespace --install --atomic -f .github/helm-values/${{ matrix.target_cluster }}.neon-proxy-scram-legacy.yaml --set image.tag=${DOCKER_TAG} --set settings.sentryUrl=${{ secrets.SENTRY_URL_PROXY }} --wait --timeout 15m0s
+
+  deploy-storage-broker-prod-new:
+    runs-on: prod
+    container: 093970136003.dkr.ecr.eu-central-1.amazonaws.com/ansible:latest
+    if: inputs.deployStorageBroker && inputs.disclamerAcknowledged
+    defaults:
+      run:
+        shell: bash
+    strategy:
+      matrix:
+        include:
+          - target_region:  us-east-2
+            target_cluster: prod-us-east-2-delta
+          - target_region:  us-west-2
+            target_cluster: prod-us-west-2-eta
+          - target_region: eu-central-1
+            target_cluster: prod-eu-central-1-gamma
+          - target_region: ap-southeast-1
+            target_cluster: prod-ap-southeast-1-epsilon
+          - target_region: us-east-1
+            target_cluster: prod-us-east-1-theta
+    environment:
+      name: prod-${{ matrix.target_region }}
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v3
+        with:
+          submodules: true
+          fetch-depth: 0
+          ref: ${{ inputs.branch }}
+
+      - name: Configure environment
+        run: |
+          helm repo add neondatabase https://neondatabase.github.io/helm-charts
+          aws --region ${{ matrix.target_region }} eks update-kubeconfig --name  ${{ matrix.target_cluster }}
+
+      - name: Deploy storage-broker
+        run:
+          helm upgrade neon-storage-broker-lb neondatabase/neon-storage-broker --namespace neon-storage-broker-lb --create-namespace --install --atomic -f .github/helm-values/${{ matrix.target_cluster }}.neon-storage-broker.yaml --set image.tag=${{ inputs.dockerTag }} --set settings.sentryUrl=${{ secrets.SENTRY_URL_BROKER }} --wait --timeout 5m0s
+
+  deploy-pg-sni-router:
+    runs-on: prod
+    container: 093970136003.dkr.ecr.eu-central-1.amazonaws.com/ansible:latest
+    if: inputs.deployPgSniRouter && inputs.disclamerAcknowledged
+    defaults:
+      run:
+        shell: bash
+    strategy:
+      matrix:
+        include:
+          - target_region:  us-east-2
+            target_cluster: prod-us-east-2-delta
+          - target_region:  us-west-2
+            target_cluster: prod-us-west-2-eta
+          - target_region: eu-central-1
+            target_cluster: prod-eu-central-1-gamma
+          - target_region: ap-southeast-1
+            target_cluster: prod-ap-southeast-1-epsilon
+          - target_region: us-east-1
+            target_cluster: prod-us-east-1-theta
+    environment:
+      name: prod-${{ matrix.target_region }}
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v3
+        with:
+          submodules: true
+          fetch-depth: 0
+          ref: ${{ inputs.branch }}
+  
+      - name: Configure environment
+        run: |
+          helm repo add neondatabase https://neondatabase.github.io/helm-charts
+          aws --region ${{ matrix.target_region }} eks update-kubeconfig --name  ${{ matrix.target_cluster }}
+  
+      - name: Deploy pg-sni-router
+        run:
+          helm upgrade neon-pg-sni-router neondatabase/neon-pg-sni-router --namespace neon-pg-sni-router --create-namespace --install --atomic -f .github/helm-values/${{ matrix.target_cluster }}.pg-sni-router.yaml --set image.tag=${{ inputs.dockerTag }} --set settings.sentryUrl=${{ secrets.SENTRY_URL_BROKER }} --wait --timeout 15m0s
--- a/Cargo.lock
+++ b/Cargo.lock
@@ -1032,7 +1032,6 @@ dependencies = [
 "serde",
 "serde_json",
 "serde_with",
- "utils",
 "workspace_hack",
 ]

@@ -1106,7 +1105,6 @@ dependencies = [
 "anyhow",
 "clap 4.2.2",
 "comfy-table",
- "compute_api",
 "git-version",
 "nix",
 "once_cell",
@@ -3676,7 +3674,6 @@ dependencies = [
 "remote_storage",
 "reqwest",
 "safekeeper_api",
- "scopeguard",
 "serde",
 "serde_json",
 "serde_with",
--- a/2
+++ b/2
@@ -2,7 +2,7 @@
 ### The image itself is mainly used as a container for the binaries and for starting e2e tests with custom parameters.
 ### By default, the binaries inside the image have some mock parameters and can start, but are not intended to be used
 ### inside this image in the real deployments.
-ARG REPOSITORY=neondatabase
+ARG REPOSITORY=369495373322.dkr.ecr.eu-central-1.amazonaws.com
 ARG IMAGE=rust
 ARG TAG=pinned

--- a/Dockerfile.compute-node
+++ b/Dockerfile.compute-node
@@ -1,5 +1,5 @@
 ARG PG_VERSION
-ARG REPOSITORY=neondatabase
+ARG REPOSITORY=369495373322.dkr.ecr.eu-central-1.amazonaws.com
 ARG IMAGE=rust
 ARG TAG=pinned

@@ -393,28 +393,6 @@ RUN case "${PG_VERSION}" in \
    make install -j $(getconf _NPROCESSORS_ONLN) && \
    echo "trusted = true" >> /usr/local/pgsql/share/extension/pg_hint_plan.control

-#########################################################################################
-#
-# Layer "kq-imcx-pg-build"
-# compile kq_imcx extension
-#
-#########################################################################################
-FROM build-deps AS kq-imcx-pg-build
-COPY --from=pg-build /usr/local/pgsql/ /usr/local/pgsql/
-
-ENV PATH "/usr/local/pgsql/bin/:$PATH"
-RUN apt-get update && \
-    apt-get install -y git libgtk2.0-dev libpq-dev libpam-dev libxslt-dev libkrb5-dev cmake && \
-    wget https://github.com/ketteq-neon/postgres-exts/archive/e0bd1a9d9313d7120c1b9c7bb15c48c0dede4c4e.tar.gz -O kq_imcx.tar.gz && \
-    echo "dc93a97ff32d152d32737ba7e196d9687041cda15e58ab31344c2f2de8855336 kq_imcx.tar.gz" | sha256sum --check && \
-    mkdir kq_imcx-src && cd kq_imcx-src && tar xvzf ../kq_imcx.tar.gz --strip-components=1 -C . && \
-    mkdir build && \
-    cd build && \
-    cmake .. && \
-    make -j $(getconf _NPROCESSORS_ONLN) && \
-    make -j $(getconf _NPROCESSORS_ONLN) install && \
-    echo 'trusted = true' >> /usr/local/pgsql/share/extension/kq_imcx.control
-
 #########################################################################################
 #
 # Layer "rust extensions"
@@ -528,7 +506,6 @@ COPY --from=hll-pg-build /usr/local/pgsql/ /usr/local/pgsql/
 COPY --from=plpgsql-check-pg-build /usr/local/pgsql/ /usr/local/pgsql/
 COPY --from=timescaledb-pg-build /usr/local/pgsql/ /usr/local/pgsql/
 COPY --from=pg-hint-plan-pg-build /usr/local/pgsql/ /usr/local/pgsql/
-COPY --from=kq-imcx-pg-build /usr/local/pgsql/ /usr/local/pgsql/
 COPY pgxn/ pgxn/

 RUN make -j $(getconf _NPROCESSORS_ONLN) \
--- a/Dockerfile.compute-tools
+++ b/Dockerfile.compute-tools
@@ -1,6 +1,6 @@
 # First transient image to build compute_tools binaries
 # NB: keep in sync with rust image version in .github/workflows/build_and_test.yml
-ARG REPOSITORY=neondatabase
+ARG REPOSITORY=369495373322.dkr.ecr.eu-central-1.amazonaws.com
 ARG IMAGE=rust
 ARG TAG=pinned

--- a/README.md
+++ b/README.md
@@ -1,5 +1,3 @@
-[![Neon](https://user-images.githubusercontent.com/13738772/236813940-dcfdcb5b-69d3-449b-a686-013febe834d4.png)](https://neon.tech)
-
 # Neon

 Neon is a serverless open-source alternative to AWS Aurora Postgres. It separates storage and compute and substitutes the PostgreSQL storage layer by redistributing data across a cluster of nodes.
--- a/compute_tools/src/bin/compute_ctl.rs
+++ b/compute_tools/src/bin/compute_ctl.rs
@@ -30,7 +30,6 @@
 //!             -b /usr/local/bin/postgres
 //! ```
 //!
-use std::collections::HashMap;
 use std::fs::File;
 use std::panic;
 use std::path::Path;
@@ -68,54 +67,6 @@ fn main() -> Result<()> {
    let spec_json = matches.get_one::<String>("spec");
    let spec_path = matches.get_one::<String>("spec-path");

-    // 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);
-    }
-    let startup_context_guard = 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
-    };
-
    let compute_id = matches.get_one::<String>("compute-id");
    let control_plane_uri = matches.get_one::<String>("control-plane-uri");

@@ -197,6 +148,8 @@ fn main() -> Result<()> {

    // We got all we need, update the state.
    let mut state = compute.state.lock().unwrap();
+    let pspec = state.pspec.as_ref().expect("spec must be set");
+    let startup_tracing_context = pspec.spec.startup_tracing_context.clone();

    // Record for how long we slept waiting for the spec.
    state.metrics.wait_for_spec_ms = Utc::now()
@@ -212,6 +165,29 @@ fn main() -> Result<()> {
    compute.state_changed.notify_all();
    drop(state);

+    // Extract OpenTelemetry context for the startup actions from the spec, 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.
+    //
+    // Switch to the startup context here, and exit it once the startup has
+    // completed and Postgres is up and running.
+    //
+    // 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.
+    let startup_context_guard = if let Some(ref carrier) = startup_tracing_context {
+        use opentelemetry::propagation::TextMapPropagator;
+        use opentelemetry::sdk::propagation::TraceContextPropagator;
+        Some(TraceContextPropagator::new().extract(carrier).attach())
+    } else {
+        None
+    };
+
    // Launch remaining service threads
    let _monitor_handle = launch_monitor(&compute).expect("cannot launch compute monitor thread");
    let _configurator_handle =
--- a/compute_tools/src/compute.rs
+++ b/compute_tools/src/compute.rs
@@ -30,7 +30,7 @@ use utils::id::{TenantId, TimelineId};
 use utils::lsn::Lsn;

 use compute_api::responses::{ComputeMetrics, ComputeStatus};
-use compute_api::spec::{ComputeMode, ComputeSpec};
+use compute_api::spec::ComputeSpec;

 use crate::config;
 use crate::pg_helpers::*;
@@ -67,9 +67,8 @@ pub struct ComputeNode {
 pub struct ComputeState {
    pub start_time: DateTime<Utc>,
    pub status: ComputeStatus,
-    /// Timestamp of the last Postgres activity. It could be `None` if
-    /// compute wasn't used since start.
-    pub last_active: Option<DateTime<Utc>>,
+    /// Timestamp of the last Postgres activity
+    pub last_active: DateTime<Utc>,
    pub error: Option<String>,
    pub pspec: Option<ParsedSpec>,
    pub metrics: ComputeMetrics,
@@ -80,7 +79,7 @@ impl ComputeState {
        Self {
            start_time: Utc::now(),
            status: ComputeStatus::Empty,
-            last_active: None,
+            last_active: Utc::now(),
            error: None,
            pspec: None,
            metrics: ComputeMetrics::default(),
@@ -250,10 +249,38 @@ impl ComputeNode {
    /// safekeepers sync, basebackup, etc.
    #[instrument(skip(self, compute_state))]
    pub fn prepare_pgdata(&self, compute_state: &ComputeState) -> Result<()> {
+        #[derive(Clone)]
+        enum Replication {
+            Primary,
+            Static { lsn: Lsn },
+            HotStandby,
+        }
+
        let pspec = compute_state.pspec.as_ref().expect("spec must be set");
        let spec = &pspec.spec;
        let pgdata_path = Path::new(&self.pgdata);

+        let hot_replica = if let Some(option) = spec.cluster.settings.find_ref("hot_standby") {
+            if let Some(value) = &option.value {
+                anyhow::ensure!(option.vartype == "bool");
+                matches!(value.as_str(), "on" | "yes" | "true")
+            } else {
+                false
+            }
+        } else {
+            false
+        };
+
+        let replication = if hot_replica {
+            Replication::HotStandby
+        } else if let Some(lsn) = spec.cluster.settings.find("recovery_target_lsn") {
+            Replication::Static {
+                lsn: Lsn::from_str(&lsn)?,
+            }
+        } else {
+            Replication::Primary
+        };
+
        // Remove/create an empty pgdata directory and put configuration there.
        self.create_pgdata()?;
        config::write_postgres_conf(&pgdata_path.join("postgresql.conf"), &pspec.spec)?;
@@ -261,8 +288,8 @@ impl ComputeNode {
        // Syncing safekeepers is only safe with primary nodes: if a primary
        // is already connected it will be kicked out, so a secondary (standby)
        // cannot sync safekeepers.
-        let lsn = match spec.mode {
-            ComputeMode::Primary => {
+        let lsn = match &replication {
+            Replication::Primary => {
                info!("starting safekeepers syncing");
                let lsn = self
                    .sync_safekeepers(pspec.storage_auth_token.clone())
@@ -270,11 +297,11 @@ impl ComputeNode {
                info!("safekeepers synced at LSN {}", lsn);
                lsn
            }
-            ComputeMode::Static(lsn) => {
+            Replication::Static { lsn } => {
                info!("Starting read-only node at static LSN {}", lsn);
-                lsn
+                *lsn
            }
-            ComputeMode::Replica => {
+            Replication::HotStandby => {
                info!("Initializing standby from latest Pageserver LSN");
                Lsn(0)
            }
@@ -294,9 +321,9 @@ impl ComputeNode {
        // Update pg_hba.conf received with basebackup.
        update_pg_hba(pgdata_path)?;

-        match spec.mode {
-            ComputeMode::Primary | ComputeMode::Static(..) => {}
-            ComputeMode::Replica => {
+        match &replication {
+            Replication::Primary | Replication::Static { .. } => {}
+            Replication::HotStandby => {
                add_standby_signal(pgdata_path)?;
            }
        }
@@ -403,13 +430,11 @@ impl ComputeNode {
        self.pg_reload_conf(&mut client)?;

        // Proceed with post-startup configuration. Note, that order of operations is important.
-        if spec.mode == ComputeMode::Primary {
-            handle_roles(&spec, &mut client)?;
-            handle_databases(&spec, &mut client)?;
-            handle_role_deletions(&spec, self.connstr.as_str(), &mut client)?;
-            handle_grants(&spec, self.connstr.as_str(), &mut client)?;
-            handle_extensions(&spec, &mut client)?;
-        }
+        handle_roles(&spec, &mut client)?;
+        handle_databases(&spec, &mut client)?;
+        handle_role_deletions(&spec, self.connstr.as_str(), &mut client)?;
+        handle_grants(&spec, self.connstr.as_str(), &mut client)?;
+        handle_extensions(&spec, &mut client)?;

        // 'Close' connection
        drop(client);
@@ -442,9 +467,7 @@ impl ComputeNode {

        let pg = self.start_postgres(spec.storage_auth_token.clone())?;

-        if spec.spec.mode == ComputeMode::Primary {
-            self.apply_config(&compute_state)?;
-        }
+        self.apply_config(&compute_state)?;

        let startup_end_time = Utc::now();
        {
--- a/compute_tools/src/config.rs
+++ b/compute_tools/src/config.rs
@@ -6,7 +6,7 @@ use std::path::Path;
 use anyhow::Result;

 use crate::pg_helpers::PgOptionsSerialize;
-use compute_api::spec::{ComputeMode, ComputeSpec};
+use compute_api::spec::ComputeSpec;

 /// Check that `line` is inside a text file and put it there if it is not.
 /// Create file if it doesn't exist.
@@ -34,25 +34,17 @@ pub fn line_in_file(path: &Path, line: &str) -> Result<bool> {
 /// Create or completely rewrite configuration file specified by `path`
 pub fn write_postgres_conf(path: &Path, spec: &ComputeSpec) -> Result<()> {
    // File::create() destroys the file content if it exists.
-    let mut file = File::create(path)?;
+    let mut postgres_conf = File::create(path)?;

+    write_auto_managed_block(&mut postgres_conf, &spec.cluster.settings.as_pg_settings())?;
+
+    Ok(())
+}
+
+// Write Postgres config block wrapped with generated comment section
+fn write_auto_managed_block(file: &mut File, buf: &str) -> Result<()> {
    writeln!(file, "# Managed by compute_ctl: begin")?;
-
-    write!(file, "{}", &spec.cluster.settings.as_pg_settings())?;
-
-    match spec.mode {
-        ComputeMode::Primary => {}
-        ComputeMode::Static(lsn) => {
-            // hot_standby is 'on' by default, but let's be explicit
-            writeln!(file, "hot_standby=on")?;
-            writeln!(file, "recovery_target_lsn='{lsn}'")?;
-        }
-        ComputeMode::Replica => {
-            // hot_standby is 'on' by default, but let's be explicit
-            writeln!(file, "hot_standby=on")?;
-        }
-    }
-
+    writeln!(file, "{}", buf)?;
    writeln!(file, "# Managed by compute_ctl: end")?;

    Ok(())
--- a/compute_tools/src/http/openapi_spec.yaml
+++ b/compute_tools/src/http/openapi_spec.yaml
@@ -181,8 +181,8 @@ components:
    ComputeState:
      type: object
      required:
-        - start_time
        - status
+        - last_active
      properties:
        start_time:
          type: string
@@ -195,13 +195,11 @@ components:
          $ref: '#/components/schemas/ComputeStatus'
        last_active:
          type: string
-          description: |
-            The last detected compute activity timestamp in UTC and RFC3339 format.
-            It could be empty if compute was never used by user since start.
+          description: The last detected compute activity timestamp in UTC and RFC3339 format.
          example: "2022-10-12T07:20:50.52Z"
        error:
          type: string
-          description: Text of the error during compute startup or reconfiguration, if any.
+          description: Text of the error during compute startup, if any.
          example: ""
        tenant:
          type: string
@@ -224,12 +222,9 @@ components:
    ComputeStatus:
      type: string
      enum:
-        - empty
        - init
        - failed
        - running
-        - configuration_pending
-        - configuration
      example: running

    #
--- a/compute_tools/src/monitor.rs
+++ b/compute_tools/src/monitor.rs
@@ -74,7 +74,7 @@ fn watch_compute_activity(compute: &ComputeNode) {
                            // Found non-idle backend, so the last activity is NOW.
                            // Save it and exit the for loop. Also clear the idle backend
                            // `state_change` timestamps array as it doesn't matter now.
-                            last_active = Some(Utc::now());
+                            last_active = Utc::now();
                            idle_backs.clear();
                            break;
                        }
@@ -82,16 +82,15 @@ fn watch_compute_activity(compute: &ComputeNode) {

                    // Get idle backend `state_change` with the max timestamp.
                    if let Some(last) = idle_backs.iter().max() {
-                        last_active = Some(*last);
+                        last_active = *last;
                    }
                }

                // Update the last activity in the shared state if we got a more recent one.
                let mut state = compute.state.lock().unwrap();
-                // NB: `Some(<DateTime>)` is always greater than `None`.
                if last_active > state.last_active {
                    state.last_active = last_active;
-                    debug!("set the last compute activity time to: {:?}", last_active);
+                    debug!("set the last compute activity time to: {}", last_active);
                }
            }
            Err(e) => {
--- a/control_plane/Cargo.toml
+++ b/control_plane/Cargo.toml
@@ -30,5 +30,4 @@ postgres_connection.workspace = true
 storage_broker.workspace = true
 utils.workspace = true

-compute_api.workspace = true
 workspace_hack.workspace = true
--- a/control_plane/src/bin/neon_local.rs
+++ b/control_plane/src/bin/neon_local.rs
@@ -7,8 +7,8 @@
 //!
 use anyhow::{anyhow, bail, Context, Result};
 use clap::{value_parser, Arg, ArgAction, ArgMatches, Command};
-use compute_api::spec::ComputeMode;
 use control_plane::endpoint::ComputeControlPlane;
+use control_plane::endpoint::ComputeMode;
 use control_plane::local_env::LocalEnv;
 use control_plane::pageserver::PageServerNode;
 use control_plane::safekeeper::SafekeeperNode;
--- a/control_plane/src/endpoint.rs
+++ b/control_plane/src/endpoint.rs
@@ -22,8 +22,6 @@ use crate::local_env::LocalEnv;
 use crate::pageserver::PageServerNode;
 use crate::postgresql_conf::PostgresConf;

-use compute_api::spec::ComputeMode;
-
 // contents of a endpoint.json file
 #[serde_as]
 #[derive(Serialize, Deserialize, PartialEq, Eq, Clone, Debug)]
@@ -124,12 +122,26 @@ impl ComputeControlPlane {

 ///////////////////////////////////////////////////////////////////////////////

+#[serde_as]
+#[derive(Serialize, Deserialize, Debug, Clone, Copy, Eq, PartialEq)]
+pub enum ComputeMode {
+    // Regular read-write node
+    Primary,
+    // if recovery_target_lsn is provided, and we want to pin the node to a specific LSN
+    Static(#[serde_as(as = "DisplayFromStr")] Lsn),
+    // Hot standby; read-only replica.
+    // Future versions may want to distinguish between replicas with hot standby
+    // feedback and other kinds of replication configurations.
+    Replica,
+}
+
 #[derive(Debug)]
 pub struct Endpoint {
    /// used as the directory name
    name: String,
    pub tenant_id: TenantId,
    pub timeline_id: TimelineId,
+    // Some(lsn) if this is a read-only endpoint anchored at 'lsn'. None for the primary.
    pub mode: ComputeMode,

    // port and address of the Postgres server
--- a/libs/compute_api/Cargo.toml
+++ b/libs/compute_api/Cargo.toml
@@ -11,5 +11,4 @@ serde.workspace = true
 serde_with.workspace = true
 serde_json.workspace = true

-utils = { path = "../utils" }
 workspace_hack.workspace = true
--- a/libs/compute_api/src/responses.rs
+++ b/libs/compute_api/src/responses.rs
@@ -19,7 +19,7 @@ pub struct ComputeStatusResponse {
    pub timeline: Option<String>,
    pub status: ComputeStatus,
    #[serde(serialize_with = "rfc3339_serialize")]
-    pub last_active: Option<DateTime<Utc>>,
+    pub last_active: DateTime<Utc>,
    pub error: Option<String>,
 }

@@ -29,7 +29,7 @@ pub struct ComputeState {
    pub status: ComputeStatus,
    /// Timestamp of the last Postgres activity
    #[serde(serialize_with = "rfc3339_serialize")]
-    pub last_active: Option<DateTime<Utc>>,
+    pub last_active: DateTime<Utc>,
    pub error: Option<String>,
 }

@@ -54,15 +54,11 @@ pub enum ComputeStatus {
    Failed,
 }

-fn rfc3339_serialize<S>(x: &Option<DateTime<Utc>>, s: S) -> Result<S::Ok, S::Error>
+fn rfc3339_serialize<S>(x: &DateTime<Utc>, s: S) -> Result<S::Ok, S::Error>
 where
    S: Serializer,
 {
-    if let Some(x) = x {
-        x.to_rfc3339().serialize(s)
-    } else {
-        s.serialize_none()
-    }
+    x.to_rfc3339().serialize(s)
 }

 /// Response of the /metrics.json API
--- a/libs/compute_api/src/spec.rs
+++ b/libs/compute_api/src/spec.rs
@@ -3,9 +3,8 @@
 //! The spec.json file is used to pass information to 'compute_ctl'. It contains
 //! all the information needed to start up the right version of PostgreSQL,
 //! and connect it to the storage nodes.
-use serde::{Deserialize, Serialize};
-use serde_with::{serde_as, DisplayFromStr};
-use utils::lsn::Lsn;
+use serde::Deserialize;
+use std::collections::HashMap;

 /// String type alias representing Postgres identifier and
 /// intended to be used for DB / role names.
@@ -13,7 +12,6 @@ pub type PgIdent = String;

 /// Cluster spec or configuration represented as an optional number of
 /// delta operations + final cluster state description.
-#[serde_as]
 #[derive(Clone, Debug, Default, Deserialize)]
 pub struct ComputeSpec {
    pub format_version: f32,
@@ -26,25 +24,9 @@ pub struct ComputeSpec {
    pub cluster: Cluster,
    pub delta_operations: Option<Vec<DeltaOp>>,

-    #[serde(default)]
-    pub mode: ComputeMode,
-
    pub storage_auth_token: Option<String>,
-}

-#[serde_as]
-#[derive(Clone, Copy, Debug, Default, Eq, PartialEq, Deserialize, Serialize)]
-pub enum ComputeMode {
-    /// A read-write node
-    #[default]
-    Primary,
-    /// A read-only node, pinned at a particular LSN
-    Static(#[serde_as(as = "DisplayFromStr")] Lsn),
-    /// A read-only node that follows the tip of the branch in hot standby mode
-    ///
-    /// Future versions may want to distinguish between replicas with hot standby
-    /// feedback and other kinds of replication configurations.
-    Replica,
+    pub startup_tracing_context: Option<HashMap<String, String>>,
 }

 #[derive(Clone, Debug, Default, Deserialize)]
--- a/libs/pageserver_api/src/models.rs
+++ b/libs/pageserver_api/src/models.rs
@@ -48,33 +48,13 @@ pub enum TenantState {
 }

 impl TenantState {
-    pub fn attachment_status(&self) -> TenantAttachmentStatus {
-        use TenantAttachmentStatus::*;
+    pub fn has_in_progress_downloads(&self) -> bool {
        match self {
-            // The attach procedure writes the marker file before adding the Attaching tenant to the tenants map.
-            // So, technically, we can return Attached here.
-            // However, as soon as Console observes Attached, it will proceed with the Postgres-level health check.
-            // But, our attach task might still be fetching the remote timelines, etc.
-            // So, return `Maybe` while Attaching, making Console wait for the attach task to finish.
-            Self::Attaching => Maybe,
-            // tenant mgr startup distinguishes attaching from loading via marker file.
-            // If it's loading, there is no attach marker file, i.e., attach had finished in the past.
-            Self::Loading => Attached,
-            // We only reach Active after successful load / attach.
-            // So, call atttachment status Attached.
-            Self::Active => Attached,
-            // If the (initial or resumed) attach procedure fails, the tenant becomes Broken.
-            // However, it also becomes Broken if the regular load fails.
-            // We would need a separate TenantState variant to distinguish these cases.
-            // However, there's no practical difference from Console's perspective.
-            // It will run a Postgres-level health check as soon as it observes Attached.
-            // That will fail on Broken tenants.
-            // Console can then rollback the attach, or, wait for operator to fix the Broken tenant.
-            Self::Broken { .. } => Attached,
-            // Why is Stopping a Maybe case? Because, during pageserver shutdown,
-            // we set the Stopping state irrespective of whether the tenant
-            // has finished attaching or not.
-            Self::Stopping => Maybe,
+            Self::Loading => true,
+            Self::Attaching => true,
+            Self::Active => false,
+            Self::Stopping => false,
+            Self::Broken { .. } => false,
        }
    }

@@ -229,25 +209,16 @@ impl TenantConfigRequest {
    }
 }

-/// See [`TenantState::attachment_status`] and the OpenAPI docs for context.
-#[derive(Serialize, Deserialize, Clone)]
-#[serde(rename_all = "snake_case")]
-pub enum TenantAttachmentStatus {
-    Maybe,
-    Attached,
-}
-
 #[serde_as]
 #[derive(Serialize, Deserialize, Clone)]
 pub struct TenantInfo {
    #[serde_as(as = "DisplayFromStr")]
    pub id: TenantId,
-    // NB: intentionally not part of OpenAPI, we don't want to commit to a specific set of TenantState's
    pub state: TenantState,
    /// Sum of the size of all layer files.
    /// If a layer is present in both local FS and S3, it counts only once.
    pub current_physical_size: Option<u64>, // physical size is only included in `tenant_status` endpoint
-    pub attachment_status: TenantAttachmentStatus,
+    pub has_in_progress_downloads: Option<bool>,
 }

 /// This represents the output of the "timeline_detail" and "timeline_list" API calls.
@@ -720,7 +691,7 @@ mod tests {
            id: TenantId::generate(),
            state: TenantState::Active,
            current_physical_size: Some(42),
-            attachment_status: TenantAttachmentStatus::Attached,
+            has_in_progress_downloads: Some(false),
        };
        let expected_active = json!({
            "id": original_active.id.to_string(),
@@ -728,7 +699,7 @@ mod tests {
                "slug": "Active",
            },
            "current_physical_size": 42,
-            "attachment_status": "attached",
+            "has_in_progress_downloads": false,
        });

        let original_broken = TenantInfo {
@@ -738,7 +709,7 @@ mod tests {
                backtrace: "backtrace info".into(),
            },
            current_physical_size: Some(42),
-            attachment_status: TenantAttachmentStatus::Attached,
+            has_in_progress_downloads: Some(false),
        };
        let expected_broken = json!({
            "id": original_broken.id.to_string(),
@@ -750,7 +721,7 @@ mod tests {
                }
            },
            "current_physical_size": 42,
-            "attachment_status": "attached",
+            "has_in_progress_downloads": false,
        });

        assert_eq!(
--- a/libs/postgres_ffi/src/pg_constants.rs
+++ b/libs/postgres_ffi/src/pg_constants.rs
@@ -146,10 +146,6 @@ pub const XLH_UPDATE_OLD_ALL_VISIBLE_CLEARED: u8 = (1 << 0) as u8;
 pub const XLH_UPDATE_NEW_ALL_VISIBLE_CLEARED: u8 = (1 << 1) as u8;
 pub const XLH_DELETE_ALL_VISIBLE_CLEARED: u8 = (1 << 0) as u8;

-// From replication/message.h
-pub const XLOG_LOGICAL_MESSAGE: u8 = 0x00;
-
-// From rmgrlist.h
 pub const RM_XLOG_ID: u8 = 0;
 pub const RM_XACT_ID: u8 = 1;
 pub const RM_SMGR_ID: u8 = 2;
@@ -161,7 +157,6 @@ pub const RM_RELMAP_ID: u8 = 7;
 pub const RM_STANDBY_ID: u8 = 8;
 pub const RM_HEAP2_ID: u8 = 9;
 pub const RM_HEAP_ID: u8 = 10;
-pub const RM_LOGICALMSG_ID: u8 = 21;

 // from xlogreader.h
 pub const XLR_INFO_MASK: u8 = 0x0F;
--- a/libs/pq_proto/src/lib.rs
+++ b/libs/pq_proto/src/lib.rs
@@ -613,7 +613,7 @@ pub struct XLogDataBody<'a> {

 #[derive(Debug)]
 pub struct WalSndKeepAlive {
-    pub wal_end: u64, // current end of WAL on the server
+    pub sent_ptr: u64,
    pub timestamp: i64,
    pub request_reply: bool,
 }
@@ -924,7 +924,7 @@ impl<'a> BeMessage<'a> {
                buf.put_u8(b'd');
                write_body(buf, |buf| {
                    buf.put_u8(b'k');
-                    buf.put_u64(req.wal_end);
+                    buf.put_u64(req.sent_ptr);
                    buf.put_i64(req.timestamp);
                    buf.put_u8(u8::from(req.request_reply));
                });
--- a/libs/remote_storage/src/local_fs.rs
+++ b/libs/remote_storage/src/local_fs.rs
@@ -128,15 +128,6 @@ impl RemoteStorage for LocalFs {
        // We need this dance with sort of durable rename (without fsyncs)
        // to prevent partial uploads. This was really hit when pageserver shutdown
        // cancelled the upload and partial file was left on the fs
-        // NOTE: Because temp file suffix always the same this operation is racy.
-        // Two concurrent operations can lead to the following sequence:
-        // T1: write(temp)
-        // T2: write(temp) -> overwrites the content
-        // T1: rename(temp, dst) -> succeeds
-        // T2: rename(temp, dst) -> fails, temp no longet exists
-        // This can be solved by supplying unique temp suffix every time, but this situation
-        // is not normal in the first place, the error can help (and helped at least once)
-        // to discover bugs in upper level synchronization.
        let temp_file_path =
            path_with_suffix_extension(&target_file_path, LOCAL_FS_TEMP_FILE_SUFFIX);
        let mut destination = io::BufWriter::new(
--- a/libs/utils/scripts/restore_from_wal.sh
+++ b/libs/utils/scripts/restore_from_wal.sh
@@ -1,21 +1,21 @@
 #!/bin/bash
-
-set -euxo pipefail
-
 PG_BIN=$1
 WAL_PATH=$2
 DATA_DIR=$3
 PORT=$4
-SYSID=$(od -A n -j 24 -N 8 -t d8 "$WAL_PATH"/000000010000000000000002* | cut -c 3-)
-rm -fr "$DATA_DIR"
-env -i LD_LIBRARY_PATH="$PG_BIN"/../lib "$PG_BIN"/initdb -E utf8 -U cloud_admin -D "$DATA_DIR" --sysid="$SYSID"
-echo port="$PORT" >> "$DATA_DIR"/postgresql.conf
-REDO_POS=0x$("$PG_BIN"/pg_controldata -D "$DATA_DIR" | grep -F "REDO location"| cut -c 42-)
+SYSID=`od -A n -j 24 -N 8 -t d8 $WAL_PATH/000000010000000000000002* | cut -c 3-`
+rm -fr $DATA_DIR
+env -i LD_LIBRARY_PATH=$PG_BIN/../lib $PG_BIN/initdb -E utf8 -U cloud_admin -D $DATA_DIR --sysid=$SYSID
+echo port=$PORT >> $DATA_DIR/postgresql.conf
+REDO_POS=0x`$PG_BIN/pg_controldata -D $DATA_DIR | fgrep "REDO location"| cut -c 42-`
 declare -i WAL_SIZE=$REDO_POS+114
-"$PG_BIN"/pg_ctl -D "$DATA_DIR" -l logfile start
-"$PG_BIN"/pg_ctl -D "$DATA_DIR" -l logfile stop -m immediate
-cp "$DATA_DIR"/pg_wal/000000010000000000000001 .
-cp "$WAL_PATH"/* "$DATA_DIR"/pg_wal/
-for partial in "$DATA_DIR"/pg_wal/*.partial ; do mv "$partial" "${partial%.partial}" ; done
-dd if=000000010000000000000001 of="$DATA_DIR"/pg_wal/000000010000000000000001 bs=$WAL_SIZE count=1 conv=notrunc
+$PG_BIN/pg_ctl -D $DATA_DIR -l logfile start
+$PG_BIN/pg_ctl -D $DATA_DIR -l logfile stop -m immediate
+cp $DATA_DIR/pg_wal/000000010000000000000001 .
+cp $WAL_PATH/* $DATA_DIR/pg_wal/
+if [ -f $DATA_DIR/pg_wal/*.partial ]
+then
+	(cd $DATA_DIR/pg_wal ; for partial in \*.partial ; do mv $partial `basename $partial .partial` ; done)
+fi
+dd if=000000010000000000000001 of=$DATA_DIR/pg_wal/000000010000000000000001 bs=$WAL_SIZE count=1 conv=notrunc
 rm -f 000000010000000000000001
--- a/libs/utils/scripts/restore_from_wal_archive.sh
+++ b/libs/utils/scripts/restore_from_wal_archive.sh
@@ -0,0 +1,20 @@
+PG_BIN=$1
+WAL_PATH=$2
+DATA_DIR=$3
+PORT=$4
+SYSID=`od -A n -j 24 -N 8 -t d8 $WAL_PATH/000000010000000000000002* | cut -c 3-`
+rm -fr $DATA_DIR /tmp/pg_wals
+mkdir /tmp/pg_wals
+env -i LD_LIBRARY_PATH=$PG_BIN/../lib $PG_BIN/initdb -E utf8 -U cloud_admin -D $DATA_DIR --sysid=$SYSID
+echo port=$PORT >> $DATA_DIR/postgresql.conf
+REDO_POS=0x`$PG_BIN/pg_controldata -D $DATA_DIR | fgrep "REDO location"| cut -c 42-`
+declare -i WAL_SIZE=$REDO_POS+114
+cp $WAL_PATH/* /tmp/pg_wals
+if [ -f $DATA_DIR/pg_wal/*.partial ]
+then
+	(cd /tmp/pg_wals ; for partial in \*.partial ; do  mv $partial `basename $partial .partial` ; done)
+fi
+dd if=$DATA_DIR/pg_wal/000000010000000000000001 of=/tmp/pg_wals/000000010000000000000001 bs=$WAL_SIZE count=1 conv=notrunc
+echo > $DATA_DIR/recovery.signal
+rm -f $DATA_DIR/pg_wal/*
+echo "restore_command = 'cp /tmp/pg_wals/%f %p'" >> $DATA_DIR/postgresql.conf
--- a/libs/utils/src/lib.rs
+++ b/libs/utils/src/lib.rs
@@ -58,45 +58,24 @@ pub mod pageserver_feedback;

 pub mod tracing_span_assert;

-pub mod rate_limit;
-
-mod failpoint_macro_helpers {
-
-    /// use with fail::cfg("$name", "return(2000)")
-    ///
-    /// The effect is similar to a "sleep(2000)" action, i.e. we sleep for the
-    /// specified time (in milliseconds). The main difference is that we use async
-    /// tokio sleep function. Another difference is that we print lines to the log,
-    /// which can be useful in tests to check that the failpoint was hit.
-    #[macro_export]
-    macro_rules! failpoint_sleep_millis_async {
-        ($name:literal) => {{
-            // If the failpoint is used with a "return" action, set should_sleep to the
-            // returned value (as string). Otherwise it's set to None.
-            let should_sleep = (|| {
-                ::fail::fail_point!($name, |x| x);
-                ::std::option::Option::None
-            })();
-
-            // Sleep if the action was a returned value
-            if let ::std::option::Option::Some(duration_str) = should_sleep {
-                $crate::failpoint_sleep_helper($name, duration_str).await
-            }
-        }};
-    }
-
-    // Helper function used by the macro. (A function has nicer scoping so we
-    // don't need to decorate everything with "::")
-    pub async fn failpoint_sleep_helper(name: &'static str, duration_str: String) {
-        let millis = duration_str.parse::<u64>().unwrap();
-        let d = std::time::Duration::from_millis(millis);
-
-        tracing::info!("failpoint {:?}: sleeping for {:?}", name, d);
-        tokio::time::sleep(d).await;
-        tracing::info!("failpoint {:?}: sleep done", name);
-    }
+/// use with fail::cfg("$name", "return(2000)")
+#[macro_export]
+macro_rules! failpoint_sleep_millis_async {
+    ($name:literal) => {{
+        let should_sleep: Option<std::time::Duration> = (|| {
+            fail::fail_point!($name, |v: Option<_>| {
+                let millis = v.unwrap().parse::<u64>().unwrap();
+                Some(Duration::from_millis(millis))
+            });
+            None
+        })();
+        if let Some(d) = should_sleep {
+            tracing::info!("failpoint {:?}: sleeping for {:?}", $name, d);
+            tokio::time::sleep(d).await;
+            tracing::info!("failpoint {:?}: sleep done", $name);
+        }
+    }};
 }
-pub use failpoint_macro_helpers::failpoint_sleep_helper;

 /// This is a shortcut to embed git sha into binaries and avoid copying the same build script to all packages
 ///
--- a/libs/utils/src/rate_limit.rs
+++ b/libs/utils/src/rate_limit.rs
@@ -1,66 +0,0 @@
-//! A helper to rate limit operations.
-
-use std::time::{Duration, Instant};
-
-pub struct RateLimit {
-    last: Option<Instant>,
-    interval: Duration,
-}
-
-impl RateLimit {
-    pub fn new(interval: Duration) -> Self {
-        Self {
-            last: None,
-            interval,
-        }
-    }
-
-    /// Call `f` if the rate limit allows.
-    /// Don't call it otherwise.
-    pub fn call<F: FnOnce()>(&mut self, f: F) {
-        let now = Instant::now();
-        match self.last {
-            Some(last) if now - last <= self.interval => {
-                // ratelimit
-            }
-            _ => {
-                self.last = Some(now);
-                f();
-            }
-        }
-    }
-}
-
-#[cfg(test)]
-mod tests {
-    use std::sync::atomic::AtomicUsize;
-
-    #[test]
-    fn basics() {
-        use super::RateLimit;
-        use std::sync::atomic::Ordering::Relaxed;
-        use std::time::Duration;
-
-        let called = AtomicUsize::new(0);
-        let mut f = RateLimit::new(Duration::from_millis(100));
-
-        let cl = || {
-            called.fetch_add(1, Relaxed);
-        };
-
-        f.call(cl);
-        assert_eq!(called.load(Relaxed), 1);
-        f.call(cl);
-        assert_eq!(called.load(Relaxed), 1);
-        f.call(cl);
-        assert_eq!(called.load(Relaxed), 1);
-        std::thread::sleep(Duration::from_millis(100));
-        f.call(cl);
-        assert_eq!(called.load(Relaxed), 2);
-        f.call(cl);
-        assert_eq!(called.load(Relaxed), 2);
-        std::thread::sleep(Duration::from_millis(100));
-        f.call(cl);
-        assert_eq!(called.load(Relaxed), 3);
-    }
-}
--- a/pageserver/src/consumption_metrics.rs
+++ b/pageserver/src/consumption_metrics.rs
@@ -5,7 +5,7 @@
 //!
 use crate::context::{DownloadBehavior, RequestContext};
 use crate::task_mgr::{self, TaskKind, BACKGROUND_RUNTIME};
-use crate::tenant::{mgr, LogicalSizeCalculationCause};
+use crate::tenant::mgr;
 use anyhow;
 use chrono::Utc;
 use consumption_metrics::{idempotency_key, Event, EventChunk, EventType, CHUNK_SIZE};
@@ -164,8 +164,7 @@ pub async fn collect_metrics_iteration(
                    timeline_written_size,
                ));

-                let span = info_span!("collect_metrics_iteration", tenant_id = %timeline.tenant_id, timeline_id = %timeline.timeline_id);
-                match span.in_scope(|| timeline.get_current_logical_size(ctx)) {
+                match timeline.get_current_logical_size(ctx) {
                    // Only send timeline logical size when it is fully calculated.
                    Ok((size, is_exact)) if is_exact => {
                        current_metrics.push((
@@ -335,9 +334,7 @@ pub async fn calculate_synthetic_size_worker(

                    if let Ok(tenant) = mgr::get_tenant(tenant_id, true).await
                    {
-                        if let Err(e) = tenant.calculate_synthetic_size(
-                            LogicalSizeCalculationCause::ConsumptionMetricsSyntheticSize,
-                            ctx).await {
+                        if let Err(e) = tenant.calculate_synthetic_size(ctx).await {
                            error!("failed to calculate synthetic size for tenant {}: {}", tenant_id, e);
                        }
                    }
--- a/pageserver/src/http/openapi_spec.yml
+++ b/pageserver/src/http/openapi_spec.yml
@@ -340,29 +340,7 @@ paths:
          format: hex

    post:
-      description: |
-        Schedules attach operation to happen in the background for the given tenant.
-        As soon as the caller sends this request, it must assume the pageserver
-        starts writing to the tenant's S3 state unless it receives one of the
-        distinguished errors below that state otherwise.
-
-        If a client receives a not-distinguished response, e.g., a network timeout,
-        it MUST retry the /attach request and poll again for the tenant's
-        attachment status.
-
-        After the client has received a 202, it MUST poll the tenant's
-        attachment status (field `attachment_status`) to reach state `attached`.
-        If the `attachment_status` is missing, the client MUST retry the `/attach`
-        request (goto previous paragraph). This is a robustness measure in case the tenant
-        status endpoint is buggy, but the attach operation is ongoing.
-
-        There is no way to cancel an in-flight request.
-
-        In any case, the client
-        * MUST NOT ASSUME that the /attach request has been lost in the network,
-        * MUST NOT ASSUME that the request has been lost, based on the observation
-          that a subsequent tenant status request returns 404. The request may
-          still be in flight. It must be retried.
+      description: Schedules attach operation to happen in the background for given tenant
      responses:
        "202":
          description: Tenant attaching scheduled
@@ -888,27 +866,13 @@ components:
      type: object
      required:
        - id
-        - attachment_status
      properties:
        id:
          type: string
        current_physical_size:
          type: integer
-        attachment_status:
-          description: |
-            Status of this tenant's attachment to this pageserver.
-
-            - `maybe` means almost nothing, don't read anything into it
-              except for the fact that the pageserver _might_ be already
-              writing to the tenant's S3 state, so, DO NOT ATTACH the
-              tenant to any other pageserver, or we risk split-brain.
-            - `attached` means that the attach operation has completed,
-              maybe successfully, maybe not. Perform a health check at
-              the Postgres level to determine healthiness of the tenant.
-
-            See the tenant `/attach` endpoint for more information.
-          type: string
-          enum: [ "maybe", "attached" ]
+        has_in_progress_downloads:
+          type: boolean
    TenantCreateInfo:
      type: object
      properties:
--- a/pageserver/src/http/routes.rs
+++ b/pageserver/src/http/routes.rs
@@ -25,7 +25,7 @@ use crate::tenant::config::TenantConfOpt;
 use crate::tenant::mgr::{TenantMapInsertError, TenantStateError};
 use crate::tenant::size::ModelInputs;
 use crate::tenant::storage_layer::LayerAccessStatsReset;
-use crate::tenant::{LogicalSizeCalculationCause, PageReconstructError, Timeline};
+use crate::tenant::{PageReconstructError, Timeline};
 use crate::{config::PageServerConf, tenant::mgr};
 use utils::{
    auth::JwtAuth,
@@ -105,9 +105,6 @@ impl From<PageReconstructError> for ApiError {
            PageReconstructError::Cancelled => {
                ApiError::InternalServerError(anyhow::anyhow!("request was cancelled"))
            }
-            PageReconstructError::AncestorStopping(_) => {
-                ApiError::InternalServerError(anyhow::Error::new(pre))
-            }
            PageReconstructError::WalRedo(pre) => {
                ApiError::InternalServerError(anyhow::Error::new(pre))
            }
@@ -172,8 +169,6 @@ async fn build_timeline_info(
    include_non_incremental_logical_size: bool,
    ctx: &RequestContext,
 ) -> anyhow::Result<TimelineInfo> {
-    crate::tenant::debug_assert_current_span_has_tenant_and_timeline_id();
-
    let mut info = build_timeline_info_common(timeline, ctx)?;
    if include_non_incremental_logical_size {
        // XXX we should be using spawn_ondemand_logical_size_calculation here.
@@ -196,7 +191,6 @@ fn build_timeline_info_common(
    timeline: &Arc<Timeline>,
    ctx: &RequestContext,
 ) -> anyhow::Result<TimelineInfo> {
-    crate::tenant::debug_assert_current_span_has_tenant_and_timeline_id();
    let last_record_lsn = timeline.get_last_record_lsn();
    let (wal_source_connstr, last_received_msg_lsn, last_received_msg_ts) = {
        let guard = timeline.last_received_wal.lock().unwrap();
@@ -269,28 +263,25 @@ async fn timeline_create_handler(mut request: Request<Body>) -> Result<Response<

    let ctx = RequestContext::new(TaskKind::MgmtRequest, DownloadBehavior::Error);

-    async {
-        let tenant = mgr::get_tenant(tenant_id, true).await?;
-        match tenant.create_timeline(
-            new_timeline_id,
-            request_data.ancestor_timeline_id.map(TimelineId::from),
-            request_data.ancestor_start_lsn,
-            request_data.pg_version.unwrap_or(crate::DEFAULT_PG_VERSION),
-            &ctx,
-        )
-        .await {
-            Ok(Some(new_timeline)) => {
-                // Created. Construct a TimelineInfo for it.
-                let timeline_info = build_timeline_info_common(&new_timeline, &ctx)
-                    .map_err(ApiError::InternalServerError)?;
-                json_response(StatusCode::CREATED, timeline_info)
-            }
-            Ok(None) => json_response(StatusCode::CONFLICT, ()), // timeline already exists
-            Err(err) => Err(ApiError::InternalServerError(err)),
-        }
-    }
+    let tenant = mgr::get_tenant(tenant_id, true).await?;
+    match tenant.create_timeline(
+        new_timeline_id,
+        request_data.ancestor_timeline_id.map(TimelineId::from),
+        request_data.ancestor_start_lsn,
+        request_data.pg_version.unwrap_or(crate::DEFAULT_PG_VERSION),
+        &ctx,
+    )
    .instrument(info_span!("timeline_create", tenant = %tenant_id, new_timeline = ?request_data.new_timeline_id, timeline_id = %new_timeline_id, lsn=?request_data.ancestor_start_lsn, pg_version=?request_data.pg_version))
-    .await
+    .await {
+        Ok(Some(new_timeline)) => {
+            // Created. Construct a TimelineInfo for it.
+            let timeline_info = build_timeline_info_common(&new_timeline, &ctx)
+                .map_err(ApiError::InternalServerError)?;
+            json_response(StatusCode::CREATED, timeline_info)
+        }
+        Ok(None) => json_response(StatusCode::CONFLICT, ()), // timeline already exists
+        Err(err) => Err(ApiError::InternalServerError(err)),
+    }
 }

 async fn timeline_list_handler(request: Request<Body>) -> Result<Response<Body>, ApiError> {
@@ -312,7 +303,6 @@ async fn timeline_list_handler(request: Request<Body>) -> Result<Response<Body>,
                include_non_incremental_logical_size.unwrap_or(false),
                &ctx,
            )
-            .instrument(info_span!("build_timeline_info", timeline_id = %timeline.timeline_id))
            .await
            .context("Failed to convert tenant timeline {timeline_id} into the local one: {e:?}")
            .map_err(ApiError::InternalServerError)?;
@@ -477,7 +467,7 @@ async fn tenant_list_handler(request: Request<Body>) -> Result<Response<Body>, A
            id: *id,
            state: state.clone(),
            current_physical_size: None,
-            attachment_status: state.attachment_status(),
+            has_in_progress_downloads: Some(state.has_in_progress_downloads()),
        })
        .collect::<Vec<TenantInfo>>();

@@ -502,7 +492,7 @@ async fn tenant_status(request: Request<Body>) -> Result<Response<Body>, ApiErro
            id: tenant_id,
            state: state.clone(),
            current_physical_size: Some(current_physical_size),
-            attachment_status: state.attachment_status(),
+            has_in_progress_downloads: Some(state.has_in_progress_downloads()),
        })
    }
    .instrument(info_span!("tenant_status_handler", tenant = %tenant_id))
@@ -537,11 +527,7 @@ async fn tenant_size_handler(request: Request<Body>) -> Result<Response<Body>, A

    // this can be long operation
    let inputs = tenant
-        .gather_size_inputs(
-            retention_period,
-            LogicalSizeCalculationCause::TenantSizeHandler,
-            &ctx,
-        )
+        .gather_size_inputs(retention_period, &ctx)
        .await
        .map_err(ApiError::InternalServerError)?;

@@ -705,6 +691,16 @@ pub fn html_response(status: StatusCode, data: String) -> Result<Response<Body>,
    Ok(response)
 }

+// Helper function to standardize the error messages we produce on bad durations
+//
+// Intended to be used with anyhow's `with_context`, e.g.:
+//
+//   let value = result.with_context(bad_duration("name", &value))?;
+//
+fn bad_duration<'a>(field_name: &'static str, value: &'a str) -> impl 'a + Fn() -> String {
+    move || format!("Cannot parse `{field_name}` duration {value:?}")
+}
+
 async fn tenant_create_handler(mut request: Request<Body>) -> Result<Response<Body>, ApiError> {
    check_permission(&request, None)?;

@@ -712,7 +708,91 @@ async fn tenant_create_handler(mut request: Request<Body>) -> Result<Response<Bo

    let request_data: TenantCreateRequest = json_request(&mut request).await?;

-    let tenant_conf = TenantConfOpt::try_from(&request_data).map_err(ApiError::BadRequest)?;
+    let mut tenant_conf = TenantConfOpt::default();
+    if let Some(gc_period) = request_data.gc_period {
+        tenant_conf.gc_period = Some(
+            humantime::parse_duration(&gc_period)
+                .with_context(bad_duration("gc_period", &gc_period))
+                .map_err(ApiError::BadRequest)?,
+        );
+    }
+    tenant_conf.gc_horizon = request_data.gc_horizon;
+    tenant_conf.image_creation_threshold = request_data.image_creation_threshold;
+
+    if let Some(pitr_interval) = request_data.pitr_interval {
+        tenant_conf.pitr_interval = Some(
+            humantime::parse_duration(&pitr_interval)
+                .with_context(bad_duration("pitr_interval", &pitr_interval))
+                .map_err(ApiError::BadRequest)?,
+        );
+    }
+
+    if let Some(walreceiver_connect_timeout) = request_data.walreceiver_connect_timeout {
+        tenant_conf.walreceiver_connect_timeout = Some(
+            humantime::parse_duration(&walreceiver_connect_timeout)
+                .with_context(bad_duration(
+                    "walreceiver_connect_timeout",
+                    &walreceiver_connect_timeout,
+                ))
+                .map_err(ApiError::BadRequest)?,
+        );
+    }
+    if let Some(lagging_wal_timeout) = request_data.lagging_wal_timeout {
+        tenant_conf.lagging_wal_timeout = Some(
+            humantime::parse_duration(&lagging_wal_timeout)
+                .with_context(bad_duration("lagging_wal_timeout", &lagging_wal_timeout))
+                .map_err(ApiError::BadRequest)?,
+        );
+    }
+    if let Some(max_lsn_wal_lag) = request_data.max_lsn_wal_lag {
+        tenant_conf.max_lsn_wal_lag = Some(max_lsn_wal_lag);
+    }
+    if let Some(trace_read_requests) = request_data.trace_read_requests {
+        tenant_conf.trace_read_requests = Some(trace_read_requests);
+    }
+
+    tenant_conf.checkpoint_distance = request_data.checkpoint_distance;
+    if let Some(checkpoint_timeout) = request_data.checkpoint_timeout {
+        tenant_conf.checkpoint_timeout = Some(
+            humantime::parse_duration(&checkpoint_timeout)
+                .with_context(bad_duration("checkpoint_timeout", &checkpoint_timeout))
+                .map_err(ApiError::BadRequest)?,
+        );
+    }
+
+    tenant_conf.compaction_target_size = request_data.compaction_target_size;
+    tenant_conf.compaction_threshold = request_data.compaction_threshold;
+
+    if let Some(compaction_period) = request_data.compaction_period {
+        tenant_conf.compaction_period = Some(
+            humantime::parse_duration(&compaction_period)
+                .with_context(bad_duration("compaction_period", &compaction_period))
+                .map_err(ApiError::BadRequest)?,
+        );
+    }
+
+    if let Some(eviction_policy) = request_data.eviction_policy {
+        tenant_conf.eviction_policy = Some(
+            serde_json::from_value(eviction_policy)
+                .context("parse field `eviction_policy`")
+                .map_err(ApiError::BadRequest)?,
+        );
+    }
+
+    tenant_conf.min_resident_size_override = request_data.min_resident_size_override;
+
+    if let Some(evictions_low_residence_duration_metric_threshold) =
+        request_data.evictions_low_residence_duration_metric_threshold
+    {
+        tenant_conf.evictions_low_residence_duration_metric_threshold = Some(
+            humantime::parse_duration(&evictions_low_residence_duration_metric_threshold)
+                .with_context(bad_duration(
+                    "evictions_low_residence_duration_metric_threshold",
+                    &evictions_low_residence_duration_metric_threshold,
+                ))
+                .map_err(ApiError::BadRequest)?,
+        );
+    }

    let target_tenant_id = request_data
        .new_tenant_id
@@ -780,7 +860,85 @@ async fn update_tenant_config_handler(
    let tenant_id = request_data.tenant_id;
    check_permission(&request, Some(tenant_id))?;

-    let tenant_conf = TenantConfOpt::try_from(&request_data).map_err(ApiError::BadRequest)?;
+    let mut tenant_conf = TenantConfOpt::default();
+    if let Some(gc_period) = request_data.gc_period {
+        tenant_conf.gc_period = Some(
+            humantime::parse_duration(&gc_period)
+                .with_context(bad_duration("gc_period", &gc_period))
+                .map_err(ApiError::BadRequest)?,
+        );
+    }
+    tenant_conf.gc_horizon = request_data.gc_horizon;
+    tenant_conf.image_creation_threshold = request_data.image_creation_threshold;
+
+    if let Some(pitr_interval) = request_data.pitr_interval {
+        tenant_conf.pitr_interval = Some(
+            humantime::parse_duration(&pitr_interval)
+                .with_context(bad_duration("pitr_interval", &pitr_interval))
+                .map_err(ApiError::BadRequest)?,
+        );
+    }
+    if let Some(walreceiver_connect_timeout) = request_data.walreceiver_connect_timeout {
+        tenant_conf.walreceiver_connect_timeout = Some(
+            humantime::parse_duration(&walreceiver_connect_timeout)
+                .with_context(bad_duration(
+                    "walreceiver_connect_timeout",
+                    &walreceiver_connect_timeout,
+                ))
+                .map_err(ApiError::BadRequest)?,
+        );
+    }
+    if let Some(lagging_wal_timeout) = request_data.lagging_wal_timeout {
+        tenant_conf.lagging_wal_timeout = Some(
+            humantime::parse_duration(&lagging_wal_timeout)
+                .with_context(bad_duration("lagging_wal_timeout", &lagging_wal_timeout))
+                .map_err(ApiError::BadRequest)?,
+        );
+    }
+    tenant_conf.max_lsn_wal_lag = request_data.max_lsn_wal_lag;
+    tenant_conf.trace_read_requests = request_data.trace_read_requests;
+
+    tenant_conf.checkpoint_distance = request_data.checkpoint_distance;
+    if let Some(checkpoint_timeout) = request_data.checkpoint_timeout {
+        tenant_conf.checkpoint_timeout = Some(
+            humantime::parse_duration(&checkpoint_timeout)
+                .with_context(bad_duration("checkpoint_timeout", &checkpoint_timeout))
+                .map_err(ApiError::BadRequest)?,
+        );
+    }
+    tenant_conf.compaction_target_size = request_data.compaction_target_size;
+    tenant_conf.compaction_threshold = request_data.compaction_threshold;
+
+    if let Some(compaction_period) = request_data.compaction_period {
+        tenant_conf.compaction_period = Some(
+            humantime::parse_duration(&compaction_period)
+                .with_context(bad_duration("compaction_period", &compaction_period))
+                .map_err(ApiError::BadRequest)?,
+        );
+    }
+
+    if let Some(eviction_policy) = request_data.eviction_policy {
+        tenant_conf.eviction_policy = Some(
+            serde_json::from_value(eviction_policy)
+                .context("parse field `eviction_policy`")
+                .map_err(ApiError::BadRequest)?,
+        );
+    }
+
+    tenant_conf.min_resident_size_override = request_data.min_resident_size_override;
+
+    if let Some(evictions_low_residence_duration_metric_threshold) =
+        request_data.evictions_low_residence_duration_metric_threshold
+    {
+        tenant_conf.evictions_low_residence_duration_metric_threshold = Some(
+            humantime::parse_duration(&evictions_low_residence_duration_metric_threshold)
+                .with_context(bad_duration(
+                    "evictions_low_residence_duration_metric_threshold",
+                    &evictions_low_residence_duration_metric_threshold,
+                ))
+                .map_err(ApiError::BadRequest)?,
+        );
+    }

    let state = get_state(&request);
    mgr::set_new_tenant_config(state.conf, tenant_conf, tenant_id)
--- a/pageserver/src/keyspace.rs
+++ b/pageserver/src/keyspace.rs
@@ -5,7 +5,7 @@ use std::ops::Range;
 ///
 /// Represents a set of Keys, in a compact form.
 ///
-#[derive(Clone, Debug, Default)]
+#[derive(Clone, Debug)]
 pub struct KeySpace {
    /// Contiguous ranges of keys that belong to the key space. In key order,
    /// and with no overlap.
@@ -61,58 +61,6 @@ impl KeySpace {

        KeyPartitioning { parts }
    }
-
-    ///
-    /// Calculate logical size of delta layers: total size of all blocks covered by it's key range
-    ///
-    pub fn get_logical_size(&self, range: &Range<Key>) -> u64 {
-        let mut start_key = range.start;
-        let n_ranges = self.ranges.len();
-        let start_index = match self.ranges.binary_search_by_key(&start_key, |r| r.start) {
-            Ok(index) => index, // keyspace range starts with start_key
-            Err(index) => {
-                if index != 0 && self.ranges[index - 1].end > start_key {
-                    index - 1 // previous keyspace range overlaps with specified
-                } else if index == n_ranges {
-                    return 0; // no intersection with specified range
-                } else {
-                    start_key = self.ranges[index].start;
-                    index
-                }
-            }
-        };
-        let mut size = 0u64;
-        for i in start_index..n_ranges {
-            if self.ranges[i].start >= range.end {
-                break;
-            }
-            let end_key = if self.ranges[i].end < range.end {
-                self.ranges[i].end
-            } else {
-                range.end
-            };
-            let n_blocks = key_range_size(&(start_key..end_key));
-            if n_blocks != u32::MAX {
-                size += n_blocks as u64 * BLCKSZ as u64;
-            }
-            if i + 1 < n_ranges {
-                start_key = self.ranges[i + 1].start;
-            }
-        }
-        size
-    }
-
-    ///
-    /// Check if key space contains overlapping range
-    ///
-    pub fn overlaps(&self, range: &Range<Key>) -> bool {
-        match self.ranges.binary_search_by_key(&range.end, |r| r.start) {
-            Ok(0) => false,
-            Err(0) => false,
-            Ok(index) => self.ranges[index - 1].end > range.start,
-            Err(index) => self.ranges[index - 1].end > range.start,
-        }
-    }
 }

 ///
@@ -181,226 +129,3 @@ impl KeySpaceAccum {
        }
    }
 }
-
-///
-/// A helper object, to collect a set of keys and key ranges into a KeySpace
-/// object. Key ranges may be inserted in any order and can overlap.
-///
-#[derive(Clone, Debug, Default)]
-pub struct KeySpaceRandomAccum {
-    ranges: Vec<Range<Key>>,
-}
-
-impl KeySpaceRandomAccum {
-    pub fn new() -> Self {
-        Self { ranges: Vec::new() }
-    }
-
-    pub fn add_key(&mut self, key: Key) {
-        self.add_range(singleton_range(key))
-    }
-
-    pub fn add_range(&mut self, range: Range<Key>) {
-        self.ranges.push(range);
-    }
-
-    pub fn to_keyspace(mut self) -> KeySpace {
-        let mut ranges = Vec::new();
-        if !self.ranges.is_empty() {
-            self.ranges.sort_by_key(|r| r.start);
-            let mut start = self.ranges.first().unwrap().start;
-            let mut end = self.ranges.first().unwrap().end;
-            for r in self.ranges {
-                assert!(r.start >= start);
-                if r.start > end {
-                    ranges.push(start..end);
-                    start = r.start;
-                    end = r.end;
-                } else if r.end > end {
-                    end = r.end;
-                }
-            }
-            ranges.push(start..end);
-        }
-        KeySpace { ranges }
-    }
-}
-
-#[cfg(test)]
-mod tests {
-    use super::*;
-    use std::fmt::Write;
-
-    // Helper function to create a key range.
-    //
-    // Make the tests below less verbose.
-    fn kr(irange: Range<i128>) -> Range<Key> {
-        Key::from_i128(irange.start)..Key::from_i128(irange.end)
-    }
-
-    #[allow(dead_code)]
-    fn dump_keyspace(ks: &KeySpace) {
-        for r in ks.ranges.iter() {
-            println!("  {}..{}", r.start.to_i128(), r.end.to_i128());
-        }
-    }
-
-    fn assert_ks_eq(actual: &KeySpace, expected: Vec<Range<Key>>) {
-        if actual.ranges != expected {
-            let mut msg = String::new();
-
-            writeln!(msg, "expected:").unwrap();
-            for r in &expected {
-                writeln!(msg, "  {}..{}", r.start.to_i128(), r.end.to_i128()).unwrap();
-            }
-            writeln!(msg, "got:").unwrap();
-            for r in &actual.ranges {
-                writeln!(msg, "  {}..{}", r.start.to_i128(), r.end.to_i128()).unwrap();
-            }
-            panic!("{}", msg);
-        }
-    }
-
-    #[test]
-    fn keyspace_add_range() {
-        // two separate ranges
-        //
-        // #####
-        //         #####
-        let mut ks = KeySpaceRandomAccum::default();
-        ks.add_range(kr(0..10));
-        ks.add_range(kr(20..30));
-        assert_ks_eq(&ks.to_keyspace(), vec![kr(0..10), kr(20..30)]);
-
-        // two separate ranges, added in reverse order
-        //
-        //         #####
-        // #####
-        let mut ks = KeySpaceRandomAccum::default();
-        ks.add_range(kr(20..30));
-        ks.add_range(kr(0..10));
-
-        // add range that is adjacent to the end of an existing range
-        //
-        // #####
-        //      #####
-        ks.add_range(kr(0..10));
-        ks.add_range(kr(10..30));
-        assert_ks_eq(&ks.to_keyspace(), vec![kr(0..30)]);
-
-        // add range that is adjacent to the start of an existing range
-        //
-        //      #####
-        // #####
-        let mut ks = KeySpaceRandomAccum::default();
-        ks.add_range(kr(10..30));
-        ks.add_range(kr(0..10));
-        assert_ks_eq(&ks.to_keyspace(), vec![kr(0..30)]);
-
-        // add range that overlaps with the end of an existing range
-        //
-        // #####
-        //    #####
-        let mut ks = KeySpaceRandomAccum::default();
-        ks.add_range(kr(0..10));
-        ks.add_range(kr(5..30));
-        assert_ks_eq(&ks.to_keyspace(), vec![kr(0..30)]);
-
-        // add range that overlaps with the start of an existing range
-        //
-        //    #####
-        // #####
-        let mut ks = KeySpaceRandomAccum::default();
-        ks.add_range(kr(5..30));
-        ks.add_range(kr(0..10));
-        assert_ks_eq(&ks.to_keyspace(), vec![kr(0..30)]);
-
-        // add range that is fully covered by an existing range
-        //
-        // #########
-        //   #####
-        let mut ks = KeySpaceRandomAccum::default();
-        ks.add_range(kr(0..30));
-        ks.add_range(kr(10..20));
-        assert_ks_eq(&ks.to_keyspace(), vec![kr(0..30)]);
-
-        // add range that extends an existing range from both ends
-        //
-        //   #####
-        // #########
-        let mut ks = KeySpaceRandomAccum::default();
-        ks.add_range(kr(10..20));
-        ks.add_range(kr(0..30));
-        assert_ks_eq(&ks.to_keyspace(), vec![kr(0..30)]);
-
-        // add a range that overlaps with two existing ranges, joining them
-        //
-        // #####   #####
-        //    #######
-        let mut ks = KeySpaceRandomAccum::default();
-        ks.add_range(kr(0..10));
-        ks.add_range(kr(20..30));
-        ks.add_range(kr(5..25));
-        assert_ks_eq(&ks.to_keyspace(), vec![kr(0..30)]);
-    }
-
-    #[test]
-    fn keyspace_overlaps() {
-        let mut ks = KeySpaceRandomAccum::default();
-        ks.add_range(kr(10..20));
-        ks.add_range(kr(30..40));
-        let ks = ks.to_keyspace();
-
-        //        #####      #####
-        // xxxx
-        assert!(!ks.overlaps(&kr(0..5)));
-
-        //        #####      #####
-        //   xxxx
-        assert!(!ks.overlaps(&kr(5..9)));
-
-        //        #####      #####
-        //    xxxx
-        assert!(!ks.overlaps(&kr(5..10)));
-
-        //        #####      #####
-        //     xxxx
-        assert!(ks.overlaps(&kr(5..11)));
-
-        //        #####      #####
-        //        xxxx
-        assert!(ks.overlaps(&kr(10..15)));
-
-        //        #####      #####
-        //         xxxx
-        assert!(ks.overlaps(&kr(15..20)));
-
-        //        #####      #####
-        //           xxxx
-        assert!(ks.overlaps(&kr(15..25)));
-
-        //        #####      #####
-        //              xxxx
-        assert!(!ks.overlaps(&kr(22..28)));
-
-        //        #####      #####
-        //               xxxx
-        assert!(!ks.overlaps(&kr(25..30)));
-
-        //        #####      #####
-        //                      xxxx
-        assert!(ks.overlaps(&kr(35..35)));
-
-        //        #####      #####
-        //                        xxxx
-        assert!(!ks.overlaps(&kr(40..45)));
-
-        //        #####      #####
-        //                        xxxx
-        assert!(!ks.overlaps(&kr(45..50)));
-
-        //        #####      #####
-        //        xxxxxxxxxxx
-        assert!(ks.overlaps(&kr(0..30))); // XXXXX This fails currently!
-    }
-}
--- a/pageserver/src/metrics.rs
+++ b/pageserver/src/metrics.rs
@@ -1,9 +1,9 @@
 use metrics::core::{AtomicU64, GenericCounter};
 use metrics::{
    register_counter_vec, register_histogram, register_histogram_vec, register_int_counter,
-    register_int_counter_vec, register_int_gauge, register_int_gauge_vec, register_uint_gauge_vec,
-    Counter, CounterVec, Histogram, HistogramVec, IntCounter, IntCounterVec, IntGauge, IntGaugeVec,
-    UIntGauge, UIntGaugeVec,
+    register_int_counter_vec, register_int_gauge_vec, register_uint_gauge_vec, Counter, CounterVec,
+    Histogram, HistogramVec, IntCounter, IntCounterVec, IntGauge, IntGaugeVec, UIntGauge,
+    UIntGaugeVec,
 };
 use once_cell::sync::Lazy;
 use pageserver_api::models::TenantState;
@@ -30,7 +30,6 @@ const STORAGE_TIME_OPERATIONS: &[&str] = &[
    "create images",
    "init logical size",
    "logical size",
-    "imitate logical size",
    "load layer map",
    "gc",
 ];
@@ -187,16 +186,6 @@ static PERSISTENT_BYTES_WRITTEN: Lazy<IntCounterVec> = Lazy::new(|| {
    .expect("failed to define a metric")
 });

-pub(crate) static EVICTION_ITERATION_DURATION: Lazy<HistogramVec> = Lazy::new(|| {
-    register_histogram_vec!(
-        "pageserver_eviction_iteration_duration_seconds_global",
-        "Time spent on a single eviction iteration",
-        &["period_secs", "threshold_secs"],
-        STORAGE_OP_BUCKETS.into(),
-    )
-    .expect("failed to define a metric")
-});
-
 static EVICTIONS: Lazy<IntCounterVec> = Lazy::new(|| {
    register_int_counter_vec!(
        "pageserver_evictions",
@@ -489,65 +478,6 @@ pub static TENANT_TASK_EVENTS: Lazy<IntCounterVec> = Lazy::new(|| {
    .expect("Failed to register tenant_task_events metric")
 });

-pub static BACKGROUND_LOOP_PERIOD_OVERRUN_COUNT: Lazy<IntCounterVec> = Lazy::new(|| {
-    register_int_counter_vec!(
-        "pageserver_background_loop_period_overrun_count",
-        "Incremented whenever warn_when_period_overrun() logs a warning.",
-        &["task", "period"],
-    )
-    .expect("failed to define a metric")
-});
-
-// walreceiver metrics
-
-pub static WALRECEIVER_STARTED_CONNECTIONS: Lazy<IntCounter> = Lazy::new(|| {
-    register_int_counter!(
-        "pageserver_walreceiver_started_connections_total",
-        "Number of started walreceiver connections"
-    )
-    .expect("failed to define a metric")
-});
-
-pub static WALRECEIVER_ACTIVE_MANAGERS: Lazy<IntGauge> = Lazy::new(|| {
-    register_int_gauge!(
-        "pageserver_walreceiver_active_managers",
-        "Number of active walreceiver managers"
-    )
-    .expect("failed to define a metric")
-});
-
-pub static WALRECEIVER_SWITCHES: Lazy<IntCounterVec> = Lazy::new(|| {
-    register_int_counter_vec!(
-        "pageserver_walreceiver_switches_total",
-        "Number of walreceiver manager change_connection calls",
-        &["reason"]
-    )
-    .expect("failed to define a metric")
-});
-
-pub static WALRECEIVER_BROKER_UPDATES: Lazy<IntCounter> = Lazy::new(|| {
-    register_int_counter!(
-        "pageserver_walreceiver_broker_updates_total",
-        "Number of received broker updates in walreceiver"
-    )
-    .expect("failed to define a metric")
-});
-
-pub static WALRECEIVER_CANDIDATES_EVENTS: Lazy<IntCounterVec> = Lazy::new(|| {
-    register_int_counter_vec!(
-        "pageserver_walreceiver_candidates_events_total",
-        "Number of walreceiver candidate events",
-        &["event"]
-    )
-    .expect("failed to define a metric")
-});
-
-pub static WALRECEIVER_CANDIDATES_ADDED: Lazy<IntCounter> =
-    Lazy::new(|| WALRECEIVER_CANDIDATES_EVENTS.with_label_values(&["add"]));
-
-pub static WALRECEIVER_CANDIDATES_REMOVED: Lazy<IntCounter> =
-    Lazy::new(|| WALRECEIVER_CANDIDATES_EVENTS.with_label_values(&["remove"]));
-
 // Metrics collected on WAL redo operations
 //
 // We collect the time spent in actual WAL redo ('redo'), and time waiting
@@ -708,7 +638,6 @@ pub struct TimelineMetrics {
    pub compact_time_histo: StorageTimeMetrics,
    pub create_images_time_histo: StorageTimeMetrics,
    pub logical_size_histo: StorageTimeMetrics,
-    pub imitate_logical_size_histo: StorageTimeMetrics,
    pub load_layer_map_histo: StorageTimeMetrics,
    pub garbage_collect_histo: StorageTimeMetrics,
    pub last_record_gauge: IntGauge,
@@ -741,8 +670,6 @@ impl TimelineMetrics {
        let create_images_time_histo =
            StorageTimeMetrics::new("create images", &tenant_id, &timeline_id);
        let logical_size_histo = StorageTimeMetrics::new("logical size", &tenant_id, &timeline_id);
-        let imitate_logical_size_histo =
-            StorageTimeMetrics::new("imitate logical size", &tenant_id, &timeline_id);
        let load_layer_map_histo =
            StorageTimeMetrics::new("load layer map", &tenant_id, &timeline_id);
        let garbage_collect_histo = StorageTimeMetrics::new("gc", &tenant_id, &timeline_id);
@@ -779,7 +706,6 @@ impl TimelineMetrics {
            compact_time_histo,
            create_images_time_histo,
            logical_size_histo,
-            imitate_logical_size_histo,
            garbage_collect_histo,
            load_layer_map_histo,
            last_record_gauge,
@@ -1240,7 +1166,4 @@ pub fn preinitialize_metrics() {
    // Initialize it eagerly, so that our alert rule can distinguish absence of the metric from metric value 0.
    assert_eq!(UNEXPECTED_ONDEMAND_DOWNLOADS.get(), 0);
    UNEXPECTED_ONDEMAND_DOWNLOADS.reset();
-
-    // Same as above for this metric, but, it's a Vec-type metric for which we don't know all the labels.
-    BACKGROUND_LOOP_PERIOD_OVERRUN_COUNT.reset();
 }
--- a/pageserver/src/page_service.rs
+++ b/pageserver/src/page_service.rs
@@ -256,10 +256,7 @@ async fn page_service_conn_main(
    //
    // no write timeout is used, because the kernel is assumed to error writes after some time.
    let mut socket = tokio_io_timeout::TimeoutReader::new(socket);
-
-    // timeout should be lower, but trying out multiple days for
-    // <https://github.com/neondatabase/neon/issues/4205>
-    socket.set_timeout(Some(std::time::Duration::from_secs(60 * 60 * 24 * 3)));
+    socket.set_timeout(Some(std::time::Duration::from_secs(60 * 10)));
    let socket = std::pin::pin!(socket);

    // XXX: pgbackend.run() should take the connection_ctx,
--- a/pageserver/src/pgdatadir_mapping.rs
+++ b/pageserver/src/pgdatadir_mapping.rs
@@ -500,8 +500,6 @@ impl Timeline {
        cancel: CancellationToken,
        ctx: &RequestContext,
    ) -> Result<u64, CalculateLogicalSizeError> {
-        crate::tenant::debug_assert_current_span_has_tenant_and_timeline_id();
-
        // Fetch list of database dirs and iterate them
        let buf = self.get(DBDIR_KEY, lsn, ctx).await.context("read dbdir")?;
        let dbdir = DbDirectory::des(&buf).context("deserialize db directory")?;
--- a/pageserver/src/tenant.rs
+++ b/pageserver/src/tenant.rs
@@ -58,8 +58,6 @@ use crate::task_mgr::TaskKind;
 use crate::tenant::config::TenantConfOpt;
 use crate::tenant::metadata::load_metadata;
 use crate::tenant::remote_timeline_client::index::IndexPart;
-use crate::tenant::remote_timeline_client::MaybeDeletedIndexPart;
-use crate::tenant::remote_timeline_client::PersistIndexPartWithDeletedFlagError;
 use crate::tenant::storage_layer::DeltaLayer;
 use crate::tenant::storage_layer::ImageLayer;
 use crate::tenant::storage_layer::Layer;
@@ -97,10 +95,7 @@ mod timeline;

 pub mod size;

-pub(crate) use timeline::debug_assert_current_span_has_tenant_and_timeline_id;
-pub use timeline::{
-    LocalLayerInfoForDiskUsageEviction, LogicalSizeCalculationCause, PageReconstructError, Timeline,
-};
+pub use timeline::{LocalLayerInfoForDiskUsageEviction, PageReconstructError, Timeline};

 // re-export this function so that page_cache.rs can use it.
 pub use crate::tenant::ephemeral_file::writeback as writeback_ephemeral_file;
@@ -276,10 +271,7 @@ impl UninitializedTimeline<'_> {
            .await
            .context("Failed to flush after basebackup import")?;

-        // Initialize without loading the layer map. We started with an empty layer map, and already
-        // updated it for the layers that we created during the import.
-        let mut timelines = self.owning_tenant.timelines.lock().unwrap();
-        self.initialize_with_lock(ctx, &mut timelines, false, true)
+        self.initialize(ctx)
    }

    fn raw_timeline(&self) -> anyhow::Result<&Arc<Timeline>> {
@@ -596,19 +588,16 @@ impl Tenant {
    /// finishes. You can use wait_until_active() to wait for the task to
    /// complete.
    ///
-    pub(crate) fn spawn_attach(
+    pub fn spawn_attach(
        conf: &'static PageServerConf,
        tenant_id: TenantId,
        remote_storage: GenericRemoteStorage,
        ctx: &RequestContext,
-    ) -> anyhow::Result<Arc<Tenant>> {
+    ) -> Arc<Tenant> {
        // XXX: Attach should provide the config, especially during tenant migration.
        //      See https://github.com/neondatabase/neon/issues/1555
        let tenant_conf = TenantConfOpt::default();

-        Self::attach_idempotent_create_marker_file(conf, tenant_id)
-            .context("create attach marker file")?;
-
        let wal_redo_manager = Arc::new(PostgresRedoManager::new(conf, tenant_id));
        let tenant = Arc::new(Tenant::new(
            TenantState::Attaching,
@@ -641,46 +630,7 @@ impl Tenant {
                Ok(())
            },
        );
-        Ok(tenant)
-    }
-
-    fn attach_idempotent_create_marker_file(
-        conf: &'static PageServerConf,
-        tenant_id: TenantId,
-    ) -> anyhow::Result<()> {
-        // Create directory with marker file to indicate attaching state.
-        // The load_local_tenants() function in tenant::mgr relies on the marker file
-        // to determine whether a tenant has finished attaching.
-        let tenant_dir = conf.tenant_path(&tenant_id);
-        let marker_file = conf.tenant_attaching_mark_file_path(&tenant_id);
-        debug_assert_eq!(marker_file.parent().unwrap(), tenant_dir);
-        // TODO: should use tokio::fs here, but
-        // 1. caller is not async, for good reason (it holds tenants map lock)
-        // 2. we'd need to think about cancel safety. Turns out dropping a tokio::fs future
-        //    doesn't wait for the activity in the fs thread pool.
-        crashsafe::create_dir_all(&tenant_dir).context("create tenant directory")?;
-        match fs::OpenOptions::new()
-            .write(true)
-            .create_new(true)
-            .open(&marker_file)
-        {
-            Ok(_) => {}
-            Err(e) if e.kind() == std::io::ErrorKind::AlreadyExists => {
-                // Either this is a retry of attach or there is a concurrent task also doing attach for this tenant.
-                // We cannot distinguish this here.
-                // The caller is responsible for ensuring there's no concurrent attach for a tenant.
-                {} // fsync again, we don't know if that already happened
-            }
-            err => {
-                err.context("create tenant attaching marker file")?;
-                unreachable!("we covered the Ok() case above");
-            }
-        }
-        crashsafe::fsync_file_and_parent(&marker_file)
-            .context("fsync tenant attaching marker file and parent")?;
-        debug_assert!(tenant_dir.is_dir());
-        debug_assert!(marker_file.is_file());
-        Ok(())
+        tenant
    }

    ///
@@ -688,15 +638,26 @@ impl Tenant {
    ///
    #[instrument(skip_all, fields(tenant_id=%self.tenant_id))]
    async fn attach(self: &Arc<Tenant>, ctx: RequestContext) -> anyhow::Result<()> {
+        // Create directory with marker file to indicate attaching state.
+        // The load_local_tenants() function in tenant::mgr relies on the marker file
+        // to determine whether a tenant has finished attaching.
+        let tenant_dir = self.conf.tenant_path(&self.tenant_id);
        let marker_file = self.conf.tenant_attaching_mark_file_path(&self.tenant_id);
-        if !tokio::fs::try_exists(&marker_file)
-            .await
-            .context("check for existence of marker file")?
-        {
-            anyhow::bail!(
-                "implementation error: marker file should exist at beginning of this function"
-            );
+        debug_assert_eq!(marker_file.parent().unwrap(), tenant_dir);
+        if tenant_dir.exists() {
+            if !marker_file.is_file() {
+                anyhow::bail!(
+                    "calling Tenant::attach with a tenant directory that doesn't have the attaching marker file:\ntenant_dir: {}\nmarker_file: {}",
+                    tenant_dir.display(), marker_file.display());
+            }
+        } else {
+            crashsafe::create_dir_all(&tenant_dir).context("create tenant directory")?;
+            fs::File::create(&marker_file).context("create tenant attaching marker file")?;
+            crashsafe::fsync_file_and_parent(&marker_file)
+                .context("fsync tenant attaching marker file and parent")?;
        }
+        debug_assert!(tenant_dir.is_dir());
+        debug_assert!(marker_file.is_file());

        // Get list of remote timelines
        // download index files for every tenant timeline
@@ -734,9 +695,16 @@ impl Tenant {
                        .await
                        .context("download index file")?;

+                    let remote_metadata = index_part.parse_metadata().context("parse metadata")?;
+
                    debug!("finished index part download");

-                    Result::<_, anyhow::Error>::Ok((timeline_id, client, index_part))
+                    Result::<_, anyhow::Error>::Ok((
+                        timeline_id,
+                        client,
+                        index_part,
+                        remote_metadata,
+                    ))
                }
                .map(move |res| {
                    res.with_context(|| format!("download index part for timeline {timeline_id}"))
@@ -745,26 +713,17 @@ impl Tenant {
            );
        }
        // Wait for all the download tasks to complete & collect results.
-        let mut remote_index_and_client = HashMap::new();
+        let mut remote_clients = HashMap::new();
+        let mut index_parts = HashMap::new();
        let mut timeline_ancestors = HashMap::new();
        while let Some(result) = part_downloads.join_next().await {
            // NB: we already added timeline_id as context to the error
            let result: Result<_, anyhow::Error> = result.context("joinset task join")?;
-            let (timeline_id, client, index_part) = result?;
+            let (timeline_id, client, index_part, remote_metadata) = result?;
            debug!("successfully downloaded index part for timeline {timeline_id}");
-            match index_part {
-                MaybeDeletedIndexPart::IndexPart(index_part) => {
-                    timeline_ancestors.insert(
-                        timeline_id,
-                        index_part.parse_metadata().context("parse_metadata")?,
-                    );
-                    remote_index_and_client.insert(timeline_id, (index_part, client));
-                }
-                MaybeDeletedIndexPart::Deleted => {
-                    info!("timeline {} is deleted, skipping", timeline_id);
-                    continue;
-                }
-            }
+            timeline_ancestors.insert(timeline_id, remote_metadata);
+            index_parts.insert(timeline_id, index_part);
+            remote_clients.insert(timeline_id, client);
        }

        // For every timeline, download the metadata file, scan the local directory,
@@ -772,16 +731,12 @@ impl Tenant {
        // layer file.
        let sorted_timelines = tree_sort_timelines(timeline_ancestors)?;
        for (timeline_id, remote_metadata) in sorted_timelines {
-            let (index_part, remote_client) = remote_index_and_client
-                .remove(&timeline_id)
-                .expect("just put it in above");
-
            // TODO again handle early failure
            self.load_remote_timeline(
                timeline_id,
-                index_part,
+                index_parts.remove(&timeline_id).unwrap(),
                remote_metadata,
-                remote_client,
+                remote_clients.remove(&timeline_id).unwrap(),
                &ctx,
            )
            .await
@@ -873,15 +828,11 @@ impl Tenant {
    }

    /// Create a placeholder Tenant object for a broken tenant
-    pub fn create_broken_tenant(
-        conf: &'static PageServerConf,
-        tenant_id: TenantId,
-        reason: String,
-    ) -> Arc<Tenant> {
+    pub fn create_broken_tenant(conf: &'static PageServerConf, tenant_id: TenantId) -> Arc<Tenant> {
        let wal_redo_manager = Arc::new(PostgresRedoManager::new(conf, tenant_id));
        Arc::new(Tenant::new(
            TenantState::Broken {
-                reason,
+                reason: "create_broken_tenant".into(),
                backtrace: String::new(),
            },
            conf,
@@ -914,7 +865,7 @@ impl Tenant {
            Ok(conf) => conf,
            Err(e) => {
                error!("load tenant config failed: {:?}", e);
-                return Tenant::create_broken_tenant(conf, tenant_id, format!("{e:#}"));
+                return Tenant::create_broken_tenant(conf, tenant_id);
            }
        };

@@ -1091,13 +1042,21 @@ impl Tenant {
    /// Subroutine of `load_tenant`, to load an individual timeline
    ///
    /// NB: The parent is assumed to be already loaded!
-    #[instrument(skip_all, fields(timeline_id))]
+    #[instrument(skip(self, local_metadata, ctx), fields(timeline_id=%timeline_id))]
    async fn load_local_timeline(
        &self,
        timeline_id: TimelineId,
        local_metadata: TimelineMetadata,
        ctx: &RequestContext,
    ) -> anyhow::Result<()> {
+        let ancestor = if let Some(ancestor_timeline_id) = local_metadata.ancestor_timeline() {
+            let ancestor_timeline = self.get_timeline(ancestor_timeline_id, false)
+            .with_context(|| anyhow::anyhow!("cannot find ancestor timeline {ancestor_timeline_id} for timeline {timeline_id}"))?;
+            Some(ancestor_timeline)
+        } else {
+            None
+        };
+
        let remote_client = self.remote_storage.as_ref().map(|remote_storage| {
            RemoteTimelineClient::new(
                remote_storage.clone(),
@@ -1110,29 +1069,6 @@ impl Tenant {
        let remote_startup_data = match &remote_client {
            Some(remote_client) => match remote_client.download_index_file().await {
                Ok(index_part) => {
-                    let index_part = match index_part {
-                        MaybeDeletedIndexPart::IndexPart(index_part) => index_part,
-                        MaybeDeletedIndexPart::Deleted => {
-                            // TODO: we won't reach here if remote storage gets de-configured after start of the deletion operation.
-                            // Example:
-                            //  start deletion operation
-                            //  finishes upload of index part
-                            //  pageserver crashes
-                            //  remote storage gets de-configured
-                            //  pageserver starts
-                            //
-                            // We don't really anticipate remote storage to be de-configured, so, for now, this is fine.
-                            // Also, maybe we'll remove that option entirely in the future, see https://github.com/neondatabase/neon/issues/4099.
-                            info!("is_deleted is set on remote, resuming removal of local data originally done by timeline deletion handler");
-                            std::fs::remove_dir_all(
-                                self.conf.timeline_path(&timeline_id, &self.tenant_id),
-                            )
-                            .context("remove_dir_all")?;
-
-                            return Ok(());
-                        }
-                    };
-
                    let remote_metadata = index_part.parse_metadata().context("parse_metadata")?;
                    Some(RemoteStartupData {
                        index_part,
@@ -1148,14 +1084,6 @@ impl Tenant {
            None => None,
        };

-        let ancestor = if let Some(ancestor_timeline_id) = local_metadata.ancestor_timeline() {
-            let ancestor_timeline = self.get_timeline(ancestor_timeline_id, false)
-            .with_context(|| anyhow::anyhow!("cannot find ancestor timeline {ancestor_timeline_id} for timeline {timeline_id}"))?;
-            Some(ancestor_timeline)
-        } else {
-            None
-        };
-
        self.timeline_init_and_sync(
            timeline_id,
            remote_client,
@@ -1264,24 +1192,8 @@ impl Tenant {
            "Cannot create timelines on inactive tenant"
        );

-        if let Ok(existing) = self.get_timeline(new_timeline_id, false) {
+        if self.get_timeline(new_timeline_id, false).is_ok() {
            debug!("timeline {new_timeline_id} already exists");
-
-            if let Some(remote_client) = existing.remote_client.as_ref() {
-                // Wait for uploads to complete, so that when we return Ok, the timeline
-                // is known to be durable on remote storage. Just like we do at the end of
-                // this function, after we have created the timeline ourselves.
-                //
-                // We only really care that the initial version of `index_part.json` has
-                // been uploaded. That's enough to remember that the timeline
-                // exists. However, there is no function to wait specifically for that so
-                // we just wait for all in-progress uploads to finish.
-                remote_client
-                    .wait_completion()
-                    .await
-                    .context("wait for timeline uploads to complete")?;
-            }
-
            return Ok(None);
        }

@@ -1323,17 +1235,6 @@ impl Tenant {
            }
        };

-        if let Some(remote_client) = loaded_timeline.remote_client.as_ref() {
-            // Wait for the upload of the 'index_part.json` file to finish, so that when we return
-            // Ok, the timeline is durable in remote storage.
-            let kind = ancestor_timeline_id
-                .map(|_| "branched")
-                .unwrap_or("bootstrapped");
-            remote_client.wait_completion().await.with_context(|| {
-                format!("wait for {} timeline initial uploads to complete", kind)
-            })?;
-        }
-
        Ok(Some(loaded_timeline))
    }

@@ -1430,8 +1331,6 @@ impl Tenant {
        timeline_id: TimelineId,
        _ctx: &RequestContext,
    ) -> Result<(), DeleteTimelineError> {
-        timeline::debug_assert_current_span_has_tenant_and_timeline_id();
-
        // Transition the timeline into TimelineState::Stopping.
        // This should prevent new operations from starting.
        let timeline = {
@@ -1472,44 +1371,9 @@ impl Tenant {
        timeline.walreceiver.stop().await;
        debug!("wal receiver shutdown confirmed");

-        // Prevent new uploads from starting.
-        if let Some(remote_client) = timeline.remote_client.as_ref() {
-            let res = remote_client.stop();
-            match res {
-                Ok(()) => {}
-                Err(e) => match e {
-                    remote_timeline_client::StopError::QueueUninitialized => {
-                        // This case shouldn't happen currently because the
-                        // load and attach code bails out if _any_ of the timeline fails to fetch its IndexPart.
-                        // That is, before we declare the Tenant as Active.
-                        // But we only allow calls to delete_timeline on Active tenants.
-                        return Err(DeleteTimelineError::Other(anyhow::anyhow!("upload queue is uninitialized, likely the timeline was in Broken state prior to this call because it failed to fetch IndexPart during load or attach, check the logs")));
-                    }
-                },
-            }
-        }
-
-        // Stop & wait for the remaining timeline tasks, including upload tasks.
-        // NB: This and other delete_timeline calls do not run as a task_mgr task,
-        //     so, they are not affected by this shutdown_tasks() call.
        info!("waiting for timeline tasks to shutdown");
        task_mgr::shutdown_tasks(None, Some(self.tenant_id), Some(timeline_id)).await;

-        // Mark timeline as deleted in S3 so we won't pick it up next time
-        // during attach or pageserver restart.
-        // See comment in persist_index_part_with_deleted_flag.
-        if let Some(remote_client) = timeline.remote_client.as_ref() {
-            match remote_client.persist_index_part_with_deleted_flag().await {
-                // If we (now, or already) marked it successfully as deleted, we can proceed
-                Ok(()) | Err(PersistIndexPartWithDeletedFlagError::AlreadyDeleted(_)) => (),
-                // Bail out otherwise
-                Err(e @ PersistIndexPartWithDeletedFlagError::AlreadyInProgress(_))
-                | Err(e @ PersistIndexPartWithDeletedFlagError::Other(_)) => {
-                    return Err(DeleteTimelineError::Other(anyhow::anyhow!(e)));
-                }
-            }
-        }
-
        {
            // Grab the layer_removal_cs lock, and actually perform the deletion.
            //
@@ -1533,54 +1397,19 @@ impl Tenant {
            //     by the caller.

            let local_timeline_directory = self.conf.timeline_path(&timeline_id, &self.tenant_id);
-
-            fail::fail_point!("timeline-delete-before-rm", |_| {
-                Err(anyhow::anyhow!("failpoint: timeline-delete-before-rm"))?
-            });
-
-            // NB: This need not be atomic because the deleted flag in the IndexPart
-            // will be observed during tenant/timeline load. The deletion will be resumed there.
-            //
-            // For configurations without remote storage, we tolerate that we're not crash-safe here.
-            // The timeline may come up Active but with missing layer files, in such setups.
-            // See https://github.com/neondatabase/neon/pull/3919#issuecomment-1531726720
-            match std::fs::remove_dir_all(&local_timeline_directory) {
-                Err(e) if e.kind() == std::io::ErrorKind::NotFound => {
-                    // This can happen if we're called a second time, e.g.,
-                    // because of a previous failure/cancellation at/after
-                    // failpoint timeline-delete-after-rm.
-                    //
-                    // It can also happen if we race with tenant detach, because,
-                    // it doesn't grab the layer_removal_cs lock.
-                    //
-                    // For now, log and continue.
-                    // warn! level is technically not appropriate for the
-                    // first case because we should expect retries to happen.
-                    // But the error is so rare, it seems better to get attention if it happens.
-                    let tenant_state = self.current_state();
-                    warn!(
-                        timeline_dir=?local_timeline_directory,
-                        ?tenant_state,
-                        "timeline directory not found, proceeding anyway"
-                    );
-                    // continue with the rest of the deletion
-                }
-                res => res.with_context(|| {
-                    format!(
-                        "Failed to remove local timeline directory '{}'",
-                        local_timeline_directory.display()
-                    )
-                })?,
-            }
+            // XXX make this atomic so that, if we crash-mid-way, the timeline won't be picked up
+            // with some layers missing.
+            std::fs::remove_dir_all(&local_timeline_directory).with_context(|| {
+                format!(
+                    "Failed to remove local timeline directory '{}'",
+                    local_timeline_directory.display()
+                )
+            })?;

            info!("finished deleting layer files, releasing layer_removal_cs.lock()");
            drop(layer_removal_guard);
        }

-        fail::fail_point!("timeline-delete-after-rm", |_| {
-            Err(anyhow::anyhow!("failpoint: timeline-delete-after-rm"))?
-        });
-
        // Remove the timeline from the map.
        let mut timelines = self.timelines.lock().unwrap();
        let children_exist = timelines
@@ -2195,7 +2024,7 @@ impl Tenant {
                // made.
                break;
            }
-            let result = timeline.gc(ctx).await?;
+            let result = timeline.gc().await?;
            totals += result;
        }

@@ -2403,18 +2232,17 @@ impl Tenant {
            src_timeline.initdb_lsn,
            src_timeline.pg_version,
        );
-
-        let new_timeline = {
-            let mut timelines = self.timelines.lock().unwrap();
-            self.prepare_timeline(
+        let mut timelines = self.timelines.lock().unwrap();
+        let new_timeline = self
+            .prepare_timeline(
                dst_id,
                &metadata,
                timeline_uninit_mark,
                false,
                Some(Arc::clone(src_timeline)),
            )?
-            .initialize_with_lock(ctx, &mut timelines, true, true)?
-        };
+            .initialize_with_lock(ctx, &mut timelines, true, true)?;
+        drop(timelines);

        // Root timeline gets its layers during creation and uploads them along with the metadata.
        // A branch timeline though, when created, can get no writes for some time, hence won't get any layers created.
@@ -2524,8 +2352,6 @@ impl Tenant {
                )
            })?;

-        // Initialize the timeline without loading the layer map, because we already updated the layer
-        // map above, when we imported the datadir.
        let timeline = {
            let mut timelines = self.timelines.lock().unwrap();
            raw_timeline.initialize_with_lock(ctx, &mut timelines, false, true)?
@@ -2672,7 +2498,6 @@ impl Tenant {
        // `max_retention_period` overrides the cutoff that is used to calculate the size
        // (only if it is shorter than the real cutoff).
        max_retention_period: Option<u64>,
-        cause: LogicalSizeCalculationCause,
        ctx: &RequestContext,
    ) -> anyhow::Result<size::ModelInputs> {
        let logical_sizes_at_once = self
@@ -2694,7 +2519,6 @@ impl Tenant {
            logical_sizes_at_once,
            max_retention_period,
            &mut shared_cache,
-            cause,
            ctx,
        )
        .await
@@ -2704,12 +2528,8 @@ impl Tenant {
    /// This is periodically called by background worker.
    /// result is cached in tenant struct
    #[instrument(skip_all, fields(tenant_id=%self.tenant_id))]
-    pub async fn calculate_synthetic_size(
-        &self,
-        cause: LogicalSizeCalculationCause,
-        ctx: &RequestContext,
-    ) -> anyhow::Result<u64> {
-        let inputs = self.gather_size_inputs(None, cause, ctx).await?;
+    pub async fn calculate_synthetic_size(&self, ctx: &RequestContext) -> anyhow::Result<u64> {
+        let inputs = self.gather_size_inputs(None, ctx).await?;

        let size = inputs.calculate()?;

@@ -3469,26 +3289,14 @@ mod tests {
            .gc_iteration(Some(TIMELINE_ID), 0x10, Duration::ZERO, &ctx)
            .await?;

-        // The branchpoints should contain all timelines, even ones marked
-        // as Broken.
-        {
-            let branchpoints = &tline.gc_info.read().unwrap().retain_lsns;
-            assert_eq!(branchpoints.len(), 1);
-            assert_eq!(branchpoints[0], Lsn(0x40));
-        }
-
-        // You can read the key from the child branch even though the parent is
-        // Broken, as long as you don't need to access data from the parent.
        assert_eq!(
-            newtline.get(*TEST_KEY, Lsn(0x70), &ctx).await?,
-            TEST_IMG(&format!("foo at {}", Lsn(0x70)))
+            newtline.get(*TEST_KEY, Lsn(0x50), &ctx).await?,
+            TEST_IMG(&format!("foo at {}", Lsn(0x40)))
        );

-        // This needs to traverse to the parent, and fails.
-        let err = newtline.get(*TEST_KEY, Lsn(0x50), &ctx).await.unwrap_err();
-        assert!(err
-            .to_string()
-            .contains("will not become active. Current state: Broken"));
+        let branchpoints = &tline.gc_info.read().unwrap().retain_lsns;
+        assert_eq!(branchpoints.len(), 1);
+        assert_eq!(branchpoints[0], Lsn(0x40));

        Ok(())
    }
@@ -3754,7 +3562,7 @@ mod tests {
                .await?;
            tline.freeze_and_flush().await?;
            tline.compact(&ctx).await?;
-            tline.gc(&ctx).await?;
+            tline.gc().await?;
        }

        Ok(())
@@ -3826,7 +3634,7 @@ mod tests {
                .await?;
            tline.freeze_and_flush().await?;
            tline.compact(&ctx).await?;
-            tline.gc(&ctx).await?;
+            tline.gc().await?;
        }

        Ok(())
@@ -3910,7 +3718,7 @@ mod tests {
                .await?;
            tline.freeze_and_flush().await?;
            tline.compact(&ctx).await?;
-            tline.gc(&ctx).await?;
+            tline.gc().await?;
        }

        Ok(())
--- a/pageserver/src/tenant/config.rs
+++ b/pageserver/src/tenant/config.rs
@@ -8,8 +8,6 @@
 //! We cannot use global or default config instead, because wrong settings
 //! may lead to a data loss.
 //!
-use anyhow::Context;
-use pageserver_api::models::{TenantConfigRequest, TenantCreateRequest};
 use serde::{Deserialize, Serialize};
 use std::num::NonZeroU64;
 use std::time::Duration;
@@ -282,167 +280,6 @@ impl Default for TenantConf {
    }
 }

-// Helper function to standardize the error messages we produce on bad durations
-//
-// Intended to be used with anyhow's `with_context`, e.g.:
-//
-//   let value = result.with_context(bad_duration("name", &value))?;
-//
-fn bad_duration<'a>(field_name: &'static str, value: &'a str) -> impl 'a + Fn() -> String {
-    move || format!("Cannot parse `{field_name}` duration {value:?}")
-}
-
-impl TenantConfOpt {
-    #[allow(clippy::too_many_arguments)]
-    fn from_request(
-        checkpoint_distance: Option<u64>,
-        checkpoint_timeout: &Option<String>,
-        compaction_target_size: Option<u64>,
-        compaction_period: &Option<String>,
-        compaction_threshold: Option<usize>,
-        gc_horizon: Option<u64>,
-        gc_period: &Option<String>,
-        image_creation_threshold: Option<usize>,
-        pitr_interval: &Option<String>,
-        walreceiver_connect_timeout: &Option<String>,
-        lagging_wal_timeout: &Option<String>,
-        max_lsn_wal_lag: Option<NonZeroU64>,
-        trace_read_requests: Option<bool>,
-        eviction_policy: &Option<serde_json::Value>,
-        min_resident_size_override: Option<u64>,
-        evictions_low_residence_duration_metric_threshold: &Option<String>,
-    ) -> Result<Self, anyhow::Error> {
-        let mut tenant_conf = TenantConfOpt::default();
-
-        if let Some(gc_period) = &gc_period {
-            tenant_conf.gc_period = Some(
-                humantime::parse_duration(gc_period)
-                    .with_context(bad_duration("gc_period", gc_period))?,
-            );
-        }
-        tenant_conf.gc_horizon = gc_horizon;
-        tenant_conf.image_creation_threshold = image_creation_threshold;
-
-        if let Some(pitr_interval) = &pitr_interval {
-            tenant_conf.pitr_interval = Some(
-                humantime::parse_duration(pitr_interval)
-                    .with_context(bad_duration("pitr_interval", pitr_interval))?,
-            );
-        }
-
-        if let Some(walreceiver_connect_timeout) = &walreceiver_connect_timeout {
-            tenant_conf.walreceiver_connect_timeout = Some(
-                humantime::parse_duration(walreceiver_connect_timeout).with_context(
-                    bad_duration("walreceiver_connect_timeout", walreceiver_connect_timeout),
-                )?,
-            );
-        }
-        if let Some(lagging_wal_timeout) = &lagging_wal_timeout {
-            tenant_conf.lagging_wal_timeout = Some(
-                humantime::parse_duration(lagging_wal_timeout)
-                    .with_context(bad_duration("lagging_wal_timeout", lagging_wal_timeout))?,
-            );
-        }
-        if let Some(max_lsn_wal_lag) = max_lsn_wal_lag {
-            tenant_conf.max_lsn_wal_lag = Some(max_lsn_wal_lag);
-        }
-        if let Some(trace_read_requests) = trace_read_requests {
-            tenant_conf.trace_read_requests = Some(trace_read_requests);
-        }
-
-        tenant_conf.checkpoint_distance = checkpoint_distance;
-        if let Some(checkpoint_timeout) = &checkpoint_timeout {
-            tenant_conf.checkpoint_timeout = Some(
-                humantime::parse_duration(checkpoint_timeout)
-                    .with_context(bad_duration("checkpoint_timeout", checkpoint_timeout))?,
-            );
-        }
-
-        tenant_conf.compaction_target_size = compaction_target_size;
-        tenant_conf.compaction_threshold = compaction_threshold;
-
-        if let Some(compaction_period) = &compaction_period {
-            tenant_conf.compaction_period = Some(
-                humantime::parse_duration(compaction_period)
-                    .with_context(bad_duration("compaction_period", compaction_period))?,
-            );
-        }
-
-        if let Some(eviction_policy) = &eviction_policy {
-            tenant_conf.eviction_policy = Some(
-                serde::Deserialize::deserialize(eviction_policy)
-                    .context("parse field `eviction_policy`")?,
-            );
-        }
-
-        tenant_conf.min_resident_size_override = min_resident_size_override;
-
-        if let Some(evictions_low_residence_duration_metric_threshold) =
-            &evictions_low_residence_duration_metric_threshold
-        {
-            tenant_conf.evictions_low_residence_duration_metric_threshold = Some(
-                humantime::parse_duration(evictions_low_residence_duration_metric_threshold)
-                    .with_context(bad_duration(
-                        "evictions_low_residence_duration_metric_threshold",
-                        evictions_low_residence_duration_metric_threshold,
-                    ))?,
-            );
-        }
-
-        Ok(tenant_conf)
-    }
-}
-
-impl TryFrom<&'_ TenantCreateRequest> for TenantConfOpt {
-    type Error = anyhow::Error;
-
-    fn try_from(request_data: &TenantCreateRequest) -> Result<Self, Self::Error> {
-        Self::from_request(
-            request_data.checkpoint_distance,
-            &request_data.checkpoint_timeout,
-            request_data.compaction_target_size,
-            &request_data.compaction_period,
-            request_data.compaction_threshold,
-            request_data.gc_horizon,
-            &request_data.gc_period,
-            request_data.image_creation_threshold,
-            &request_data.pitr_interval,
-            &request_data.walreceiver_connect_timeout,
-            &request_data.lagging_wal_timeout,
-            request_data.max_lsn_wal_lag,
-            request_data.trace_read_requests,
-            &request_data.eviction_policy,
-            request_data.min_resident_size_override,
-            &request_data.evictions_low_residence_duration_metric_threshold,
-        )
-    }
-}
-
-impl TryFrom<&'_ TenantConfigRequest> for TenantConfOpt {
-    type Error = anyhow::Error;
-
-    fn try_from(request_data: &TenantConfigRequest) -> Result<Self, Self::Error> {
-        Self::from_request(
-            request_data.checkpoint_distance,
-            &request_data.checkpoint_timeout,
-            request_data.compaction_target_size,
-            &request_data.compaction_period,
-            request_data.compaction_threshold,
-            request_data.gc_horizon,
-            &request_data.gc_period,
-            request_data.image_creation_threshold,
-            &request_data.pitr_interval,
-            &request_data.walreceiver_connect_timeout,
-            &request_data.lagging_wal_timeout,
-            request_data.max_lsn_wal_lag,
-            request_data.trace_read_requests,
-            &request_data.eviction_policy,
-            request_data.min_resident_size_override,
-            &request_data.evictions_low_residence_duration_metric_threshold,
-        )
-    }
-}
-
 #[cfg(test)]
 mod tests {
    use super::*;
--- a/pageserver/src/tenant/metadata.rs
+++ b/pageserver/src/tenant/metadata.rs
@@ -12,7 +12,6 @@ use std::io::Write;
 use anyhow::{bail, ensure, Context};
 use serde::{Deserialize, Serialize};
 use tracing::info_span;
-use utils::bin_ser::SerializeError;
 use utils::{
    bin_ser::BeSer,
    id::{TenantId, TimelineId},
@@ -183,7 +182,7 @@ impl TimelineMetadata {
        }
    }

-    pub fn to_bytes(&self) -> Result<Vec<u8>, SerializeError> {
+    pub fn to_bytes(&self) -> anyhow::Result<Vec<u8>> {
        let body_bytes = self.body.ser()?;
        let metadata_size = METADATA_HDR_SIZE + body_bytes.len();
        let hdr = TimelineMetadataHeader {
--- a/pageserver/src/tenant/mgr.rs
+++ b/pageserver/src/tenant/mgr.rs
@@ -186,20 +186,10 @@ pub fn schedule_local_tenant_processing(
    let tenant = if conf.tenant_attaching_mark_file_path(&tenant_id).exists() {
        info!("tenant {tenant_id} has attaching mark file, resuming its attach operation");
        if let Some(remote_storage) = remote_storage {
-            match Tenant::spawn_attach(conf, tenant_id, remote_storage, ctx) {
-                Ok(tenant) => tenant,
-                Err(e) => {
-                    error!("Failed to spawn_attach tenant {tenant_id}, reason: {e:#}");
-                    Tenant::create_broken_tenant(conf, tenant_id, format!("{e:#}"))
-                }
-            }
+            Tenant::spawn_attach(conf, tenant_id, remote_storage, ctx)
        } else {
            warn!("tenant {tenant_id} has attaching mark file, but pageserver has no remote storage configured");
-            Tenant::create_broken_tenant(
-                conf,
-                tenant_id,
-                "attaching mark file present but no remote storage configured".to_string(),
-            )
+            Tenant::create_broken_tenant(conf, tenant_id)
        }
    } else {
        info!("tenant {tenant_id} is assumed to be loadable, starting load operation");
@@ -476,8 +466,7 @@ pub async fn attach_tenant(
            "Cannot attach tenant {tenant_id}, local tenant directory already exists"
        );

-        let tenant =
-            Tenant::spawn_attach(conf, tenant_id, remote_storage, ctx).context("spawn_attach")?;
+        let tenant = Tenant::spawn_attach(conf, tenant_id, remote_storage, ctx);
        vacant_entry.insert(tenant);
        Ok(())
    })
--- a/pageserver/src/tenant/remote_timeline_client.rs
+++ b/pageserver/src/tenant/remote_timeline_client.rs
@@ -204,11 +204,8 @@ mod download;
 pub mod index;
 mod upload;

-use anyhow::Context;
-use chrono::{NaiveDateTime, Utc};
 // re-export these
 pub use download::{is_temp_download_file, list_remote_timelines};
-use scopeguard::ScopeGuard;

 use std::sync::atomic::{AtomicU32, Ordering};
 use std::sync::{Arc, Mutex};
@@ -216,7 +213,7 @@ use std::sync::{Arc, Mutex};
 use remote_storage::{DownloadError, GenericRemoteStorage};
 use std::ops::DerefMut;
 use tokio::runtime::Runtime;
-use tracing::{debug, error, info, warn};
+use tracing::{debug, info, warn};
 use tracing::{info_span, Instrument};
 use utils::lsn::Lsn;

@@ -243,7 +240,6 @@ use utils::id::{TenantId, TimelineId};
 use self::index::IndexPart;

 use super::storage_layer::LayerFileName;
-use super::upload_queue::SetDeletedFlagProgress;

 // Occasional network issues and such can cause remote operations to fail, and
 // that's expected. If a download fails, we log it at info-level, and retry.
@@ -257,30 +253,6 @@ const FAILED_DOWNLOAD_RETRIES: u32 = 10;
 // retries. Uploads and deletions are retried forever, though.
 const FAILED_UPLOAD_WARN_THRESHOLD: u32 = 3;

-pub enum MaybeDeletedIndexPart {
-    IndexPart(IndexPart),
-    Deleted,
-}
-
-/// Errors that can arise when calling [`RemoteTimelineClient::stop`].
-#[derive(Debug, thiserror::Error)]
-pub enum StopError {
-    /// Returned if the upload queue was never initialized.
-    /// See [`RemoteTimelineClient::init_upload_queue`] and [`RemoteTimelineClient::init_upload_queue_for_empty_remote`].
-    #[error("queue is not initialized")]
-    QueueUninitialized,
-}
-
-#[derive(Debug, thiserror::Error)]
-pub enum PersistIndexPartWithDeletedFlagError {
-    #[error("another task is already setting the deleted_flag, started at {0:?}")]
-    AlreadyInProgress(NaiveDateTime),
-    #[error("the deleted_flag was already set, value is {0:?}")]
-    AlreadyDeleted(NaiveDateTime),
-    #[error(transparent)]
-    Other(#[from] anyhow::Error),
-}
-
 /// A client for accessing a timeline's data in remote storage.
 ///
 /// This takes care of managing the number of connections, and balancing them
@@ -395,7 +367,7 @@ impl RemoteTimelineClient {
    //

    /// Download index file
-    pub async fn download_index_file(&self) -> Result<MaybeDeletedIndexPart, DownloadError> {
+    pub async fn download_index_file(&self) -> Result<IndexPart, DownloadError> {
        let _unfinished_gauge_guard = self.metrics.call_begin(
            &RemoteOpFileKind::Index,
            &RemoteOpKind::Download,
@@ -404,7 +376,7 @@ impl RemoteTimelineClient {
            },
        );

-        let index_part = download::download_index_part(
+        download::download_index_part(
            self.conf,
            &self.storage_impl,
            self.tenant_id,
@@ -417,13 +389,7 @@ impl RemoteTimelineClient {
            RemoteOpKind::Download,
            Arc::clone(&self.metrics),
        )
-        .await?;
-
-        if index_part.deleted_at.is_some() {
-            Ok(MaybeDeletedIndexPart::Deleted)
-        } else {
-            Ok(MaybeDeletedIndexPart::IndexPart(index_part))
-        }
+        .await
    }

    /// Download a (layer) file from `path`, into local filesystem.
@@ -658,116 +624,6 @@ impl RemoteTimelineClient {
        Ok(())
    }

-    /// Set the deleted_at field in the remote index file.
-    ///
-    /// This fails if the upload queue has not been `stop()`ed.
-    ///
-    /// The caller is responsible for calling `stop()` AND for waiting
-    /// for any ongoing upload tasks to finish after `stop()` has succeeded.
-    /// Check method [`RemoteTimelineClient::stop`] for details.
-    pub(crate) async fn persist_index_part_with_deleted_flag(
-        self: &Arc<Self>,
-    ) -> Result<(), PersistIndexPartWithDeletedFlagError> {
-        let index_part_with_deleted_at = {
-            let mut locked = self.upload_queue.lock().unwrap();
-
-            // We must be in stopped state because otherwise
-            // we can have inprogress index part upload that can overwrite the file
-            // with missing is_deleted flag that we going to set below
-            let stopped = match &mut *locked {
-                UploadQueue::Uninitialized => {
-                    return Err(anyhow::anyhow!("is not Stopped but Uninitialized").into())
-                }
-                UploadQueue::Initialized(_) => {
-                    return Err(anyhow::anyhow!("is not Stopped but Initialized").into())
-                }
-                UploadQueue::Stopped(stopped) => stopped,
-            };
-
-            match stopped.deleted_at {
-                SetDeletedFlagProgress::NotRunning => (), // proceed
-                SetDeletedFlagProgress::InProgress(at) => {
-                    return Err(PersistIndexPartWithDeletedFlagError::AlreadyInProgress(at));
-                }
-                SetDeletedFlagProgress::Successful(at) => {
-                    return Err(PersistIndexPartWithDeletedFlagError::AlreadyDeleted(at));
-                }
-            };
-            let deleted_at = Utc::now().naive_utc();
-            stopped.deleted_at = SetDeletedFlagProgress::InProgress(deleted_at);
-
-            let mut index_part = IndexPart::new(
-                stopped.latest_files.clone(),
-                stopped.last_uploaded_consistent_lsn,
-                stopped
-                    .latest_metadata
-                    .to_bytes()
-                    .context("serialize metadata")?,
-            );
-            index_part.deleted_at = Some(deleted_at);
-            index_part
-        };
-
-        let undo_deleted_at = scopeguard::guard(Arc::clone(self), |self_clone| {
-            let mut locked = self_clone.upload_queue.lock().unwrap();
-            let stopped = match &mut *locked {
-                UploadQueue::Uninitialized | UploadQueue::Initialized(_) => unreachable!(
-                    "there's no way out of Stopping, and we checked it's Stopping above: {:?}",
-                    locked.as_str(),
-                ),
-                UploadQueue::Stopped(stopped) => stopped,
-            };
-            stopped.deleted_at = SetDeletedFlagProgress::NotRunning;
-        });
-
-        // Have a failpoint that can use the `pause` failpoint action.
-        // We don't want to block the executor thread, hence, spawn_blocking + await.
-        #[cfg(feature = "testing")]
-        tokio::task::spawn_blocking({
-            let current = tracing::Span::current();
-            move || {
-                let _entered = current.entered();
-                tracing::info!(
-                    "at failpoint persist_index_part_with_deleted_flag_after_set_before_upload_pause"
-                );
-                fail::fail_point!(
-                    "persist_index_part_with_deleted_flag_after_set_before_upload_pause"
-                );
-            }
-        })
-        .await
-        .expect("spawn_blocking");
-
-        upload::upload_index_part(
-            self.conf,
-            &self.storage_impl,
-            self.tenant_id,
-            self.timeline_id,
-            &index_part_with_deleted_at,
-        )
-        .await?;
-
-        // all good, disarm the guard and mark as success
-        ScopeGuard::into_inner(undo_deleted_at);
-        {
-            let mut locked = self.upload_queue.lock().unwrap();
-            let stopped = match &mut *locked {
-                UploadQueue::Uninitialized | UploadQueue::Initialized(_) => unreachable!(
-                    "there's no way out of Stopping, and we checked it's Stopping above: {:?}",
-                    locked.as_str(),
-                ),
-                UploadQueue::Stopped(stopped) => stopped,
-            };
-            stopped.deleted_at = SetDeletedFlagProgress::Successful(
-                index_part_with_deleted_at
-                    .deleted_at
-                    .expect("we set it above"),
-            );
-        }
-
-        Ok(())
-    }
-
    ///
    /// Pick next tasks from the queue, and start as many of them as possible without violating
    /// the ordering constraints.
@@ -885,13 +741,8 @@ impl RemoteTimelineClient {
            // upload finishes or times out soon enough.
            if task_mgr::is_shutdown_requested() {
                info!("upload task cancelled by shutdown request");
-                match self.stop() {
-                    Ok(()) => {}
-                    Err(StopError::QueueUninitialized) => {
-                        unreachable!("we never launch an upload task if the queue is uninitialized, and once it is initialized, we never go back")
-                    }
-                }
                self.calls_unfinished_metric_end(&task.op);
+                self.stop();
                return;
            }

@@ -1095,48 +946,32 @@ impl RemoteTimelineClient {
        self.metrics.call_end(&file_kind, &op_kind, track_bytes);
    }

-    /// Close the upload queue for new operations and cancel queued operations.
-    /// In-progress operations will still be running after this function returns.
-    /// Use `task_mgr::shutdown_tasks(None, Some(self.tenant_id), Some(timeline_id))`
-    /// to wait for them to complete, after calling this function.
-    pub fn stop(&self) -> Result<(), StopError> {
+    fn stop(&self) {
        // Whichever *task* for this RemoteTimelineClient grabs the mutex first will transition the queue
        // into stopped state, thereby dropping all off the queued *ops* which haven't become *tasks* yet.
        // The other *tasks* will come here and observe an already shut down queue and hence simply wrap up their business.
        let mut guard = self.upload_queue.lock().unwrap();
-        match &mut *guard {
-            UploadQueue::Uninitialized => Err(StopError::QueueUninitialized),
+        match &*guard {
+            UploadQueue::Uninitialized => panic!(
+                "callers are responsible for ensuring this is only called on initialized queue"
+            ),
            UploadQueue::Stopped(_) => {
                // nothing to do
                info!("another concurrent task already shut down the queue");
-                Ok(())
            }
-            UploadQueue::Initialized(UploadQueueInitialized {
-                latest_files,
-                latest_metadata,
-                last_uploaded_consistent_lsn,
-                ..
-            }) => {
+            UploadQueue::Initialized(qi) => {
                info!("shutting down upload queue");

                // Replace the queue with the Stopped state, taking ownership of the old
                // Initialized queue. We will do some checks on it, and then drop it.
                let qi = {
-                    // take or clone what we need
-                    let latest_files = std::mem::take(latest_files);
-                    let last_uploaded_consistent_lsn = *last_uploaded_consistent_lsn;
-                    // this could be Copy
-                    let latest_metadata = latest_metadata.clone();
-
-                    let stopped = UploadQueueStopped {
-                        latest_files,
-                        last_uploaded_consistent_lsn,
-                        latest_metadata,
-                        deleted_at: SetDeletedFlagProgress::NotRunning,
-                    };
-
-                    let upload_queue =
-                        std::mem::replace(&mut *guard, UploadQueue::Stopped(stopped));
+                    let last_uploaded_consistent_lsn = qi.last_uploaded_consistent_lsn;
+                    let upload_queue = std::mem::replace(
+                        &mut *guard,
+                        UploadQueue::Stopped(UploadQueueStopped {
+                            last_uploaded_consistent_lsn,
+                        }),
+                    );
                    if let UploadQueue::Initialized(qi) = upload_queue {
                        qi
                    } else {
@@ -1144,8 +979,6 @@ impl RemoteTimelineClient {
                    }
                };

-                assert!(qi.latest_files.is_empty(), "do not use this anymore");
-
                // consistency check
                assert_eq!(
                    qi.num_inprogress_layer_uploads
@@ -1169,7 +1002,6 @@ impl RemoteTimelineClient {

                // We're done.
                drop(guard);
-                Ok(())
            }
        }
    }
@@ -1408,11 +1240,7 @@ mod tests {
        }

        // Download back the index.json, and check that the list of files is correct
-        let index_part = match runtime.block_on(client.download_index_file())? {
-            MaybeDeletedIndexPart::IndexPart(index_part) => index_part,
-            MaybeDeletedIndexPart::Deleted => panic!("unexpectedly got deleted index part"),
-        };
-
+        let index_part = runtime.block_on(client.download_index_file())?;
        assert_file_list(
            &index_part.timeline_layers,
            &[
--- a/pageserver/src/tenant/remote_timeline_client/index.rs
+++ b/pageserver/src/tenant/remote_timeline_client/index.rs
@@ -4,7 +4,6 @@

 use std::collections::{HashMap, HashSet};

-use chrono::NaiveDateTime;
 use serde::{Deserialize, Serialize};
 use serde_with::{serde_as, DisplayFromStr};

@@ -56,10 +55,6 @@ pub struct IndexPart {
    #[serde(default)]
    version: usize,

-    #[serde(default)]
-    #[serde(skip_serializing_if = "Option::is_none")]
-    pub deleted_at: Option<NaiveDateTime>,
-
    /// Layer names, which are stored on the remote storage.
    ///
    /// Additional metadata can might exist in `layer_metadata`.
@@ -83,7 +78,7 @@ impl IndexPart {
    /// used to understand later versions.
    ///
    /// Version is currently informative only.
-    const LATEST_VERSION: usize = 2;
+    const LATEST_VERSION: usize = 1;
    pub const FILE_NAME: &'static str = "index_part.json";

    pub fn new(
@@ -106,7 +101,6 @@ impl IndexPart {
            layer_metadata,
            disk_consistent_lsn,
            metadata_bytes,
-            deleted_at: None,
        }
    }

@@ -162,7 +156,6 @@ mod tests {
            ]),
            disk_consistent_lsn: "0/16960E8".parse::<Lsn>().unwrap(),
            metadata_bytes: [113,11,159,210,0,54,0,4,0,0,0,0,1,105,96,232,1,0,0,0,0,1,105,96,112,0,0,0,0,0,0,0,0,0,0,0,0,0,1,105,96,112,0,0,0,0,1,105,96,112,0,0,0,14,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0].to_vec(),
-            deleted_at: None,
        };

        let part = serde_json::from_str::<IndexPart>(example).unwrap();
@@ -199,7 +192,6 @@ mod tests {
            ]),
            disk_consistent_lsn: "0/16960E8".parse::<Lsn>().unwrap(),
            metadata_bytes: [112,11,159,210,0,54,0,4,0,0,0,0,1,105,96,232,1,0,0,0,0,1,105,96,112,0,0,0,0,0,0,0,0,0,0,0,0,0,1,105,96,112,0,0,0,0,1,105,96,112,0,0,0,14,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0].to_vec(),
-            deleted_at: None,
        };

        let part = serde_json::from_str::<IndexPart>(example).unwrap();
@@ -244,7 +236,6 @@ mod tests {
                0, 0,
            ]
            .to_vec(),
-            deleted_at: None,
        };

        let empty_layers_parsed = serde_json::from_str::<IndexPart>(empty_layers_json).unwrap();
--- a/pageserver/src/tenant/remote_timeline_client/upload.rs
+++ b/pageserver/src/tenant/remote_timeline_client/upload.rs
@@ -19,12 +19,9 @@ pub(super) async fn upload_index_part<'a>(
    timeline_id: TimelineId,
    index_part: &'a IndexPart,
 ) -> anyhow::Result<()> {
-    tracing::trace!("uploading new index part");
-
    fail_point!("before-upload-index", |_| {
        bail!("failpoint before-upload-index")
    });
-
    let index_part_bytes = serde_json::to_vec(&index_part)
        .context("Failed to serialize index part file into bytes")?;
    let index_part_size = index_part_bytes.len();
@@ -34,7 +31,6 @@ pub(super) async fn upload_index_part<'a>(
        .metadata_path(timeline_id, tenant_id)
        .with_file_name(IndexPart::FILE_NAME);
    let storage_path = conf.remote_path(&index_part_path)?;
-
    storage
        .upload_storage_object(Box::new(index_part_bytes), index_part_size, &storage_path)
        .await
--- a/pageserver/src/tenant/size.rs
+++ b/pageserver/src/tenant/size.rs
@@ -11,7 +11,7 @@ use tokio_util::sync::CancellationToken;
 use crate::context::RequestContext;
 use crate::pgdatadir_mapping::CalculateLogicalSizeError;

-use super::{LogicalSizeCalculationCause, Tenant};
+use super::Tenant;
 use crate::tenant::Timeline;
 use utils::id::TimelineId;
 use utils::lsn::Lsn;
@@ -126,7 +126,6 @@ pub(super) async fn gather_inputs(
    limit: &Arc<Semaphore>,
    max_retention_period: Option<u64>,
    logical_size_cache: &mut HashMap<(TimelineId, Lsn), u64>,
-    cause: LogicalSizeCalculationCause,
    ctx: &RequestContext,
 ) -> anyhow::Result<ModelInputs> {
    // refresh is needed to update gc related pitr_cutoff and horizon_cutoff
@@ -319,15 +318,7 @@ pub(super) async fn gather_inputs(

    // We left the 'size' field empty in all of the Segments so far.
    // Now find logical sizes for all of the points that might need or benefit from them.
-    fill_logical_sizes(
-        &timelines,
-        &mut segments,
-        limit,
-        logical_size_cache,
-        cause,
-        ctx,
-    )
-    .await?;
+    fill_logical_sizes(&timelines, &mut segments, limit, logical_size_cache, ctx).await?;

    Ok(ModelInputs {
        segments,
@@ -345,7 +336,6 @@ async fn fill_logical_sizes(
    segments: &mut [SegmentMeta],
    limit: &Arc<Semaphore>,
    logical_size_cache: &mut HashMap<(TimelineId, Lsn), u64>,
-    cause: LogicalSizeCalculationCause,
    ctx: &RequestContext,
 ) -> anyhow::Result<()> {
    let timeline_hash: HashMap<TimelineId, Arc<Timeline>> = HashMap::from_iter(
@@ -383,17 +373,13 @@ async fn fill_logical_sizes(
                let timeline = Arc::clone(timeline_hash.get(&timeline_id).unwrap());
                let parallel_size_calcs = Arc::clone(limit);
                let ctx = ctx.attached_child();
-                joinset.spawn(
-                    calculate_logical_size(
-                        parallel_size_calcs,
-                        timeline,
-                        lsn,
-                        cause,
-                        ctx,
-                        cancel.child_token(),
-                    )
-                    .in_current_span(),
-                );
+                joinset.spawn(calculate_logical_size(
+                    parallel_size_calcs,
+                    timeline,
+                    lsn,
+                    ctx,
+                    cancel.child_token(),
+                ));
            }
            e.insert(cached_size);
        }
@@ -496,7 +482,6 @@ async fn calculate_logical_size(
    limit: Arc<tokio::sync::Semaphore>,
    timeline: Arc<crate::tenant::Timeline>,
    lsn: utils::lsn::Lsn,
-    cause: LogicalSizeCalculationCause,
    ctx: RequestContext,
    cancel: CancellationToken,
 ) -> Result<TimelineAtLsnSizeResult, RecvError> {
@@ -505,7 +490,7 @@ async fn calculate_logical_size(
        .expect("global semaphore should not had been closed");

    let size_res = timeline
-        .spawn_ondemand_logical_size_calculation(lsn, cause, ctx, cancel)
+        .spawn_ondemand_logical_size_calculation(lsn, ctx, cancel)
        .instrument(info_span!("spawn_ondemand_logical_size_calculation"))
        .await?;
    Ok(TimelineAtLsnSizeResult(timeline, lsn, size_res))
--- a/pageserver/src/tenant/storage_layer.rs
+++ b/pageserver/src/tenant/storage_layer.rs
@@ -13,9 +13,9 @@ use crate::task_mgr::TaskKind;
 use crate::walrecord::NeonWalRecord;
 use anyhow::Result;
 use bytes::Bytes;
+use either::Either;
 use enum_map::EnumMap;
 use enumset::EnumSet;
-use once_cell::sync::Lazy;
 use pageserver_api::models::LayerAccessKind;
 use pageserver_api::models::{
    HistoricLayerInfo, LayerResidenceEvent, LayerResidenceEventReason, LayerResidenceStatus,
@@ -23,10 +23,8 @@ use pageserver_api::models::{
 use std::ops::Range;
 use std::path::PathBuf;
 use std::sync::{Arc, Mutex};
-use std::time::{Duration, SystemTime, UNIX_EPOCH};
-use tracing::warn;
+use std::time::{SystemTime, UNIX_EPOCH};
 use utils::history_buffer::HistoryBufferWithDropCounter;
-use utils::rate_limit::RateLimit;

 use utils::{
    id::{TenantId, TimelineId},
@@ -39,8 +37,6 @@ pub use image_layer::{ImageLayer, ImageLayerWriter};
 pub use inmemory_layer::InMemoryLayer;
 pub use remote_layer::RemoteLayer;

-use super::layer_map::BatchedUpdates;
-
 pub fn range_overlaps<T>(a: &Range<T>, b: &Range<T>) -> bool
 where
    T: PartialOrd<T>,
@@ -162,82 +158,41 @@ impl LayerAccessStatFullDetails {
 }

 impl LayerAccessStats {
-    /// Create an empty stats object.
-    ///
-    /// The caller is responsible for recording a residence event
-    /// using [`record_residence_event`] before calling `latest_activity`.
-    /// If they don't, [`latest_activity`] will return `None`.
-    pub(crate) fn empty_will_record_residence_event_later() -> Self {
-        LayerAccessStats(Mutex::default())
+    pub(crate) fn for_loading_layer(status: LayerResidenceStatus) -> Self {
+        let new = LayerAccessStats(Mutex::new(LayerAccessStatsLocked::default()));
+        new.record_residence_event(status, LayerResidenceEventReason::LayerLoad);
+        new
    }

-    /// Create an empty stats object and record a [`LayerLoad`] event with the given residence status.
-    ///
-    /// See [`record_residence_event`] for why you need to do this while holding the layer map lock.
-    pub(crate) fn for_loading_layer<L>(
-        layer_map_lock_held_witness: &BatchedUpdates<'_, L>,
-        status: LayerResidenceStatus,
-    ) -> Self
-    where
-        L: ?Sized + Layer,
-    {
+    pub(crate) fn for_new_layer_file() -> Self {
        let new = LayerAccessStats(Mutex::new(LayerAccessStatsLocked::default()));
        new.record_residence_event(
-            layer_map_lock_held_witness,
-            status,
-            LayerResidenceEventReason::LayerLoad,
+            LayerResidenceStatus::Resident,
+            LayerResidenceEventReason::LayerCreate,
        );
        new
    }

    /// Creates a clone of `self` and records `new_status` in the clone.
-    ///
-    /// The `new_status` is not recorded in `self`.
-    ///
-    /// See [`record_residence_event`] for why you need to do this while holding the layer map lock.
-    pub(crate) fn clone_for_residence_change<L>(
+    /// The `new_status` is not recorded in `self`
+    pub(crate) fn clone_for_residence_change(
        &self,
-        layer_map_lock_held_witness: &BatchedUpdates<'_, L>,
        new_status: LayerResidenceStatus,
-    ) -> LayerAccessStats
-    where
-        L: ?Sized + Layer,
-    {
+    ) -> LayerAccessStats {
        let clone = {
            let inner = self.0.lock().unwrap();
            inner.clone()
        };
        let new = LayerAccessStats(Mutex::new(clone));
-        new.record_residence_event(
-            layer_map_lock_held_witness,
-            new_status,
-            LayerResidenceEventReason::ResidenceChange,
-        );
+        new.record_residence_event(new_status, LayerResidenceEventReason::ResidenceChange);
        new
    }

-    /// Record a change in layer residency.
-    ///
-    /// Recording the event must happen while holding the layer map lock to
-    /// ensure that latest-activity-threshold-based layer eviction (eviction_task.rs)
-    /// can do an "imitate access" to this layer, before it observes `now-latest_activity() > threshold`.
-    ///
-    /// If we instead recorded the residence event with a timestamp from before grabbing the layer map lock,
-    /// the following race could happen:
-    ///
-    /// - Compact: Write out an L1 layer from several L0 layers. This records residence event LayerCreate with the current timestamp.
-    /// - Eviction: imitate access logical size calculation. This accesses the L0 layers because the L1 layer is not yet in the layer map.
-    /// - Compact: Grab layer map lock, add the new L1 to layer map and remove the L0s, release layer map lock.
-    /// - Eviction: observes the new L1 layer whose only activity timestamp is the LayerCreate event.
-    ///
-    pub(crate) fn record_residence_event<L>(
+    fn record_residence_event(
        &self,
-        _layer_map_lock_held_witness: &BatchedUpdates<'_, L>,
        status: LayerResidenceStatus,
        reason: LayerResidenceEventReason,
-    ) where
-        L: ?Sized + Layer,
-    {
+    ) {
        let mut locked = self.0.lock().unwrap();
        locked.iter_mut().for_each(|inner| {
            inner
@@ -300,37 +255,24 @@ impl LayerAccessStats {
        ret
    }

-    /// Get the latest access timestamp, falling back to latest residence event.
-    ///
-    /// This function can only return `None` if there has not yet been a call to the
-    /// [`record_residence_event`] method. That would generally be considered an
-    /// implementation error. This function logs a rate-limited warning in that case.
-    ///
-    /// TODO: use type system to avoid the need for `fallback`.
-    /// The approach in https://github.com/neondatabase/neon/pull/3775
-    /// could be used to enforce that a residence event is recorded
-    /// before a layer is added to the layer map. We could also have
-    /// a layer wrapper type that holds the LayerAccessStats, and ensure
-    /// that that type can only be produced by inserting into the layer map.
-    pub(crate) fn latest_activity(&self) -> Option<SystemTime> {
+    fn most_recent_access_or_residence_event(
+        &self,
+    ) -> Either<LayerAccessStatFullDetails, LayerResidenceEvent> {
        let locked = self.0.lock().unwrap();
        let inner = &locked.for_eviction_policy;
        match inner.last_accesses.recent() {
-            Some(a) => Some(a.when),
+            Some(a) => Either::Left(*a),
            None => match inner.last_residence_changes.recent() {
-                Some(e) => Some(e.timestamp),
-                None => {
-                    static WARN_RATE_LIMIT: Lazy<Mutex<(usize, RateLimit)>> =
-                        Lazy::new(|| Mutex::new((0, RateLimit::new(Duration::from_secs(10)))));
-                    let mut guard = WARN_RATE_LIMIT.lock().unwrap();
-                    guard.0 += 1;
-                    let occurences = guard.0;
-                    guard.1.call(move || {
-                        warn!(parent: None, occurences, "latest_activity not available, this is an implementation bug, using fallback value");
-                    });
-                    None
-                }
-            },
+                Some(e) => Either::Right(e.clone()),
+                None => unreachable!("constructors for LayerAccessStats ensure that there's always a residence change event"),
+            }
+        }
+    }
+
+    pub(crate) fn latest_activity(&self) -> SystemTime {
+        match self.most_recent_access_or_residence_event() {
+            Either::Left(mra) => mra.when,
+            Either::Right(re) => re.timestamp,
        }
    }
 }
--- a/pageserver/src/tenant/storage_layer/delta_layer.rs
+++ b/pageserver/src/tenant/storage_layer/delta_layer.rs
@@ -57,7 +57,7 @@ use utils::{

 use super::{
    DeltaFileName, Layer, LayerAccessStats, LayerAccessStatsReset, LayerFileName, LayerIter,
-    LayerKeyIter, PathOrConf,
+    LayerKeyIter, LayerResidenceStatus, PathOrConf,
 };

 ///
@@ -637,7 +637,7 @@ impl DeltaLayer {
            key_range: summary.key_range,
            lsn_range: summary.lsn_range,
            file_size: metadata.len(),
-            access_stats: LayerAccessStats::empty_will_record_residence_event_later(),
+            access_stats: LayerAccessStats::for_loading_layer(LayerResidenceStatus::Resident),
            inner: RwLock::new(DeltaLayerInner {
                loaded: false,
                file: None,
@@ -808,7 +808,7 @@ impl DeltaLayerWriterInner {
            key_range: self.key_start..key_end,
            lsn_range: self.lsn_range.clone(),
            file_size: metadata.len(),
-            access_stats: LayerAccessStats::empty_will_record_residence_event_later(),
+            access_stats: LayerAccessStats::for_new_layer_file(),
            inner: RwLock::new(DeltaLayerInner {
                loaded: false,
                file: None,
--- a/pageserver/src/tenant/storage_layer/image_layer.rs
+++ b/pageserver/src/tenant/storage_layer/image_layer.rs
@@ -53,7 +53,7 @@ use utils::{
 };

 use super::filename::{ImageFileName, LayerFileName};
-use super::{Layer, LayerAccessStatsReset, LayerIter, PathOrConf};
+use super::{Layer, LayerAccessStatsReset, LayerIter, LayerResidenceStatus, PathOrConf};

 ///
 /// Header stored in the beginning of the file
@@ -438,7 +438,7 @@ impl ImageLayer {
            key_range: summary.key_range,
            lsn: summary.lsn,
            file_size: metadata.len(),
-            access_stats: LayerAccessStats::empty_will_record_residence_event_later(),
+            access_stats: LayerAccessStats::for_loading_layer(LayerResidenceStatus::Resident),
            inner: RwLock::new(ImageLayerInner {
                file: None,
                loaded: false,
@@ -598,7 +598,7 @@ impl ImageLayerWriterInner {
            key_range: self.key_range.clone(),
            lsn: self.lsn,
            file_size: metadata.len(),
-            access_stats: LayerAccessStats::empty_will_record_residence_event_later(),
+            access_stats: LayerAccessStats::for_new_layer_file(),
            inner: RwLock::new(ImageLayerInner {
                loaded: false,
                file: None,
--- a/pageserver/src/tenant/storage_layer/remote_layer.rs
+++ b/pageserver/src/tenant/storage_layer/remote_layer.rs
@@ -4,7 +4,6 @@
 use crate::config::PageServerConf;
 use crate::context::RequestContext;
 use crate::repository::Key;
-use crate::tenant::layer_map::BatchedUpdates;
 use crate::tenant::remote_timeline_client::index::LayerFileMetadata;
 use crate::tenant::storage_layer::{Layer, ValueReconstructResult, ValueReconstructState};
 use anyhow::{bail, Result};
@@ -247,15 +246,11 @@ impl RemoteLayer {
    }

    /// Create a Layer struct representing this layer, after it has been downloaded.
-    pub fn create_downloaded_layer<L>(
+    pub fn create_downloaded_layer(
        &self,
-        layer_map_lock_held_witness: &BatchedUpdates<'_, L>,
        conf: &'static PageServerConf,
        file_size: u64,
-    ) -> Arc<dyn PersistentLayer>
-    where
-        L: ?Sized + Layer,
-    {
+    ) -> Arc<dyn PersistentLayer> {
        if self.is_delta {
            let fname = DeltaFileName {
                key_range: self.key_range.clone(),
@@ -267,10 +262,8 @@ impl RemoteLayer {
                self.tenantid,
                &fname,
                file_size,
-                self.access_stats.clone_for_residence_change(
-                    layer_map_lock_held_witness,
-                    LayerResidenceStatus::Resident,
-                ),
+                self.access_stats
+                    .clone_for_residence_change(LayerResidenceStatus::Resident),
            ))
        } else {
            let fname = ImageFileName {
@@ -283,10 +276,8 @@ impl RemoteLayer {
                self.tenantid,
                &fname,
                file_size,
-                self.access_stats.clone_for_residence_change(
-                    layer_map_lock_held_witness,
-                    LayerResidenceStatus::Resident,
-                ),
+                self.access_stats
+                    .clone_for_residence_change(LayerResidenceStatus::Resident),
            ))
        }
    }
--- a/pageserver/src/tenant/tasks.rs
+++ b/pageserver/src/tenant/tasks.rs
@@ -259,7 +259,6 @@ pub(crate) async fn random_init_delay(
    }
 }

-/// Attention: the `task` and `period` beocme labels of a pageserver-wide prometheus metric.
 pub(crate) fn warn_when_period_overrun(elapsed: Duration, period: Duration, task: &str) {
    // Duration::ZERO will happen because it's the "disable [bgtask]" value.
    if elapsed >= period && period != Duration::ZERO {
@@ -272,8 +271,5 @@ pub(crate) fn warn_when_period_overrun(elapsed: Duration, period: Duration, task
            task,
            "task iteration took longer than the configured period"
        );
-        crate::metrics::BACKGROUND_LOOP_PERIOD_OVERRUN_COUNT
-            .with_label_values(&[task, &format!("{}", period.as_secs())])
-            .inc();
    }
 }
--- a/Show More
+++ b/Show More
				`@@ -0,0 +1 @@`
				`{{ pageserver_config \| sivel.toiletwater.to_toml }}`