Update release.yml

Update release.yml

.github/workflows/release.yml

@@ -6,36 +6,34 @@ on:
       - '[0-9]+.[0-9]+.[0-9]+'
   workflow_dispatch:
 
+concurrency:
+  group: release-${{ github.ref }}
+  cancel-in-progress: true
+
 permissions:
   contents: read
   packages: write
 
 env:
   CARGO_TERM_COLOR: always
+  BINARY_NAME: telemt
 
 jobs:
-  build:
+# ==========================
-    name: Build ${{ matrix.target }}
+# GNU / glibc
+# ==========================
+  build-gnu:
+    name: GNU ${{ matrix.target }}
     runs-on: ubuntu-latest
-    permissions:
-      contents: read
 
     strategy:
       fail-fast: false
       matrix:
         include:
           - target: x86_64-unknown-linux-gnu
-            artifact_name: telemt
+            asset: telemt-x86_64-linux-gnu
-            asset_name: telemt-x86_64-linux-gnu
           - target: aarch64-unknown-linux-gnu
-            artifact_name: telemt
+            asset: telemt-aarch64-linux-gnu
-            asset_name: telemt-aarch64-linux-gnu
-          - target: x86_64-unknown-linux-musl
-            artifact_name: telemt
-            asset_name: telemt-x86_64-linux-musl
-          - target: aarch64-unknown-linux-musl
-            artifact_name: telemt
-            asset_name: telemt-aarch64-linux-musl
 
     steps:
       - uses: actions/checkout@v4
@@ -43,12 +41,20 @@ jobs:
       - uses: dtolnay/rust-toolchain@v1
         with:
           toolchain: stable
-          targets: ${{ matrix.target }}
+          targets: |
+            x86_64-unknown-linux-gnu
+            aarch64-unknown-linux-gnu
 
-      - name: Install cross-compilation tools
+      - name: Install deps
         run: |
           sudo apt-get update
-          sudo apt-get install -y gcc-aarch64-linux-gnu
+          sudo apt-get install -y \
+            build-essential \
+            clang \
+            lld \
+            pkg-config \
+            gcc-aarch64-linux-gnu \
+            g++-aarch64-linux-gnu
 
       - uses: actions/cache@v4
         with:
@@ -56,41 +62,183 @@ jobs:
             ~/.cargo/registry
             ~/.cargo/git
             target
-          key: ${{ runner.os }}-${{ matrix.target }}-cargo-${{ hashFiles('**/Cargo.lock') }}
+          key: gnu-${{ matrix.target }}-${{ hashFiles('**/Cargo.lock') }}
-          restore-keys: |
-            ${{ runner.os }}-${{ matrix.target }}-cargo-
 
-      - name: Install cross
+      - name: Build
-        run: cargo install cross --git https://github.com/cross-rs/cross
-
-      - name: Build Release
-        env:
-          RUSTFLAGS: ${{ contains(matrix.target, 'musl') && '-C target-feature=+crt-static' || '' }}
-        run: cross build --release --target ${{ matrix.target }}
-
-      - name: Package binary
         run: |
-          cd target/${{ matrix.target }}/release
+          if [ "${{ matrix.target }}" = "aarch64-unknown-linux-gnu" ]; then
-          tar -czvf ${{ matrix.asset_name }}.tar.gz ${{ matrix.artifact_name }}
+            export CC=aarch64-linux-gnu-gcc
-          sha256sum ${{ matrix.asset_name }}.tar.gz > ${{ matrix.asset_name }}.sha256
+            export CXX=aarch64-linux-gnu-g++
+            export CC_aarch64_unknown_linux_gnu=aarch64-linux-gnu-gcc
+            export CXX_aarch64_unknown_linux_gnu=aarch64-linux-gnu-g++
+            export RUSTFLAGS="-C linker=aarch64-linux-gnu-gcc"
+          else
+            export CC=clang
+            export CXX=clang++
+            export CC_x86_64_unknown_linux_gnu=clang
+            export CXX_x86_64_unknown_linux_gnu=clang++
+            export RUSTFLAGS="-C linker=clang -C link-arg=-fuse-ld=lld"
+          fi
+
+          cargo build --release --target ${{ matrix.target }}
+
+      - name: Package
+        run: |
+          mkdir -p dist
+          BIN=target/${{ matrix.target }}/release/${{ env.BINARY_NAME }}
+
+          cp "$BIN" dist/${{ env.BINARY_NAME }}-${{ matrix.target }}
+
+          cd dist
+          tar -czf ${{ matrix.asset }}.tar.gz ${{ env.BINARY_NAME }}-${{ matrix.target }}
+          sha256sum ${{ matrix.asset }}.tar.gz > ${{ matrix.asset }}.sha256
 
       - uses: actions/upload-artifact@v4
         with:
-          name: ${{ matrix.asset_name }}
+          name: ${{ matrix.asset }}
           path: |
-            target/${{ matrix.target }}/release/${{ matrix.asset_name }}.tar.gz
+            dist/${{ matrix.asset }}.tar.gz
-            target/${{ matrix.target }}/release/${{ matrix.asset_name }}.sha256
+            dist/${{ matrix.asset }}.sha256
 
-  build-docker-image:
+# ==========================
-    needs: build
+# MUSL
+# ==========================
+  build-musl:
+    name: MUSL ${{ matrix.target }}
     runs-on: ubuntu-latest
-    permissions:
+
-      contents: read
+    container:
-      packages: write
+      image: rust:slim-bookworm
+
+    strategy:
+      fail-fast: false
+      matrix:
+        include:
+          - target: x86_64-unknown-linux-musl
+            asset: telemt-x86_64-linux-musl
+          - target: aarch64-unknown-linux-musl
+            asset: telemt-aarch64-linux-musl
 
     steps:
       - uses: actions/checkout@v4
 
+      - name: Install deps
+        run: |
+          apt-get update
+          apt-get install -y \
+            musl-tools \
+            pkg-config \
+            curl
+
+      # 💾 cache toolchain
+      - uses: actions/cache@v4
+        if: matrix.target == 'aarch64-unknown-linux-musl'
+        with:
+          path: ~/.musl-aarch64
+          key: musl-toolchain-aarch64-v1
+
+      # 🔥 надёжная установка
+      - name: Install aarch64 musl toolchain
+        if: matrix.target == 'aarch64-unknown-linux-musl'
+        run: |
+          set -e
+
+          TOOLCHAIN_DIR="$HOME/.musl-aarch64"
+          ARCHIVE="aarch64-linux-musl-cross.tgz"
+
+          if [ -x "$TOOLCHAIN_DIR/bin/aarch64-linux-musl-gcc" ]; then
+            echo "✅ musl toolchain already installed"
+          else
+            echo "⬇️ downloading musl toolchain..."
+
+            download() {
+              url="$1"
+              echo "→ trying $url"
+              curl -fL \
+                --retry 5 \
+                --retry-delay 3 \
+                --connect-timeout 10 \
+                --max-time 120 \
+                -o "$ARCHIVE" "$url" && return 0
+              return 1
+            }
+
+            download "https://musl.cc/$ARCHIVE" || \
+            download "https://more.musl.cc/$ARCHIVE" || \
+            { echo "❌ failed to download musl toolchain"; exit 1; }
+
+            mkdir -p "$TOOLCHAIN_DIR"
+            tar -xzf "$ARCHIVE" --strip-components=1 -C "$TOOLCHAIN_DIR"
+          fi
+
+          echo "$TOOLCHAIN_DIR/bin" >> $GITHUB_PATH
+
+      - name: Add rust target
+        run: rustup target add ${{ matrix.target }}
+
+      - uses: actions/cache@v4
+        with:
+          path: |
+            /usr/local/cargo/registry
+            /usr/local/cargo/git
+            target
+          key: musl-${{ matrix.target }}-${{ hashFiles('**/Cargo.lock') }}
+
+      - name: Build
+        run: |
+          if [ "${{ matrix.target }}" = "aarch64-unknown-linux-musl" ]; then
+            export CC=aarch64-linux-musl-gcc
+            export CC_aarch64_unknown_linux_musl=aarch64-linux-musl-gcc
+            export RUSTFLAGS="-C target-feature=+crt-static -C linker=aarch64-linux-musl-gcc"
+          else
+            export CC=musl-gcc
+            export CC_x86_64_unknown_linux_musl=musl-gcc
+            export RUSTFLAGS="-C target-feature=+crt-static"
+          fi
+
+          cargo build --release --target ${{ matrix.target }}
+
+      - name: Package
+        run: |
+          mkdir -p dist
+          BIN=target/${{ matrix.target }}/release/${{ env.BINARY_NAME }}
+
+          cp "$BIN" dist/${{ env.BINARY_NAME }}-${{ matrix.target }}
+
+          cd dist
+          tar -czf ${{ matrix.asset }}.tar.gz ${{ env.BINARY_NAME }}-${{ matrix.target }}
+          sha256sum ${{ matrix.asset }}.tar.gz > ${{ matrix.asset }}.sha256
+
+      - uses: actions/upload-artifact@v4
+        with:
+          name: ${{ matrix.asset }}
+          path: |
+            dist/${{ matrix.asset }}.tar.gz
+            dist/${{ matrix.asset }}.sha256
+
+# ==========================
+# Docker
+# ==========================
+  docker:
+    name: Docker
+    runs-on: ubuntu-latest
+    needs: [build-gnu, build-musl]
+    continue-on-error: true
+
+    steps:
+      - uses: actions/checkout@v4
+
+      - uses: actions/download-artifact@v4
+        with:
+          path: artifacts
+
+      - name: Extract binaries
+        run: |
+          mkdir dist
+          find artifacts -name "*.tar.gz" -exec tar -xzf {} -C dist \;
+
+          cp dist/telemt-x86_64-unknown-linux-musl dist/telemt || true
+
       - uses: docker/setup-qemu-action@v3
       - uses: docker/setup-buildx-action@v3
 
@@ -105,35 +253,43 @@ jobs:
         id: vars
         run: echo "VERSION=${GITHUB_REF#refs/tags/}" >> $GITHUB_OUTPUT
 
-      - name: Build and push
+      - name: Build & Push
         uses: docker/build-push-action@v6
         with:
           context: .
           push: true
+          platforms: linux/amd64,linux/arm64
           tags: |
             ghcr.io/${{ github.repository }}:${{ steps.vars.outputs.VERSION }}
             ghcr.io/${{ github.repository }}:latest
+          build-args: |
+            BINARY=dist/telemt
 
+# ==========================
+# Release
+# ==========================
   release:
-    name: Create Release
+    name: Release
-    needs: build
     runs-on: ubuntu-latest
+    needs: [build-gnu, build-musl]
+
     permissions:
       contents: write
 
     steps:
-      - uses: actions/checkout@v4
-        with:
-          fetch-depth: 0
-
       - uses: actions/download-artifact@v4
         with:
           path: artifacts
 
+      - name: Flatten artifacts
+        run: |
+          mkdir dist
+          find artifacts -type f -exec cp {} dist/ \;
+
       - name: Create Release
         uses: softprops/action-gh-release@v2
         with:
-          files: artifacts/**/*
+          files: dist/*
           generate_release_notes: true
           draft: false
           prerelease: ${{ contains(github.ref, '-rc') || contains(github.ref, '-beta') || contains(github.ref, '-alpha') }}

CODE_OF_CONDUCT.md

@@ -1,17 +1,19 @@
-# TELEMT Code of Conduct
+# Code of Conduct
 
 ## 1. Purpose
 
 Telemt exists to solve technical problems.
 
-It is not a platform for ideology, politics, or personal agendas.
+Telemt is open to contributors who want to learn, improve and build meaningful systems together.
 
-All interaction here is defined by systems, constraints, and outcomes.
+It is a place for building, testing, reasoning, documenting, and improving systems.
 
-Technology has consequences.
+Discussions that advance this work are in scope. Discussions that divert it are not.
-Responsibility is inherent.
+
+Technology has consequences. Responsibility is inherent.
 
 > **Zweck bestimmt die Form.**
+
 > Purpose defines form.
 
 ---
@@ -24,20 +26,24 @@ Responsibility is inherent.
 * **Clarity over noise**
   Communication is structured, concise, and relevant.
 
-* **Independence**
+* **Openness with standards**
-  Telemt does not represent any state, ideology, or organization.
+  Participation is open. The work remains disciplined.
 
-* **Open participation**
+* **Independence of judgment**
-  Access is open. Standards are not.
+  Claims are evaluated on technical merit, not affiliation or posture.
 
 * **Responsibility over capability**
   Capability does not justify careless use.
 
 * **Cooperation over friction**
-  Progress is achieved through coordination and mutual support.
+  Progress depends on coordination, mutual support, and honest review.
 
-> **Fakten sind nicht verhandelbar.**
+* **Good intent, rigorous method**
-> Facts are not negotiable.
+  Assume good intent, but require rigor.
+
+> **Aussagen gelten nach ihrer Begründung.**
+
+> Claims are weighed by evidence.
 
 ---
 
@@ -53,7 +59,12 @@ Participants are expected to:
 * Help others reach correct and reproducible outcomes
 * Act in a way that improves the system as a whole
 
+Precision is learned.
+
+New contributors are welcome. They are expected to grow into these standards. Existing contributors are expected to make that growth possible.
+
 > **Wer behauptet, belegt.**
+
 > Whoever claims, proves.
 
 ---
@@ -62,22 +73,25 @@ Participants are expected to:
 
 The following is not allowed:
 
-* Personal attacks, insults, harassment, intimidation
+* Personal attacks, insults, harassment, or intimidation
-* Political discourse, propaganda, ideological conflict
+* Repeatedly derailing discussion away from Telemt’s purpose
-* Off-topic or disruptive discussion
 * Spam, flooding, or repeated low-quality input
 * Misinformation presented as fact
-* Attempts to degrade or destabilize Telemt
+* Attempts to degrade, destabilize, or exhaust Telemt or its participants
-* Use of Telemt or its space to enable harm
+* Use of Telemt or its spaces to enable harm
+
+Telemt is not a venue for disputes that displace technical work.
+Such discussions may be closed, removed, or redirected.
 
 > **Störung ist kein Beitrag.**
+
 > Disruption is not contribution.
 
 ---
 
 ## 5. Security and Misuse
 
-Telemt is intended for lawful and responsible use.
+Telemt is intended for responsible use.
 
 * Do not use it to plan, coordinate, or execute harm
 * Do not publish vulnerabilities without responsible disclosure
@@ -86,11 +100,24 @@ Telemt is intended for lawful and responsible use.
 Security is both technical and behavioral.
 
 > **Verantwortung endet nicht am Code.**
+
 > Responsibility does not end at the code.
 
 ---
 
-## 6. Scope
+## 6. Openness
+
+Telemt is open to contributors of different backgrounds, experience levels, and working styles.
+
+Standards are public, legible, and applied to the work itself.
+
+Questions are welcome. Careful disagreement is welcome. Honest correction is welcome.
+
+Gatekeeping by obscurity, status signaling, or hostility is not.
+
+---
+
+## 7. Scope
 
 This Code of Conduct applies to all official spaces:
 
@@ -100,31 +127,43 @@ This Code of Conduct applies to all official spaces:
 
 ---
 
-## 7. Enforcement
+## 8. Maintainer Stewardship
 
-Maintainers may act to preserve the integrity of Telemt:
+Maintainers are responsible for final decisions in matters of conduct, scope, and direction.
 
-* Remove content
+This responsibility is stewardship: preserving continuity, protecting signal, maintaining standards, and keeping Telemt workable for others.
-* Lock discussions
-* Reject contributions
-* Restrict or ban participants
 
-Actions are taken to maintain function, continuity, and signal quality.
+Judgment should be exercised with restraint, consistency, and institutional responsibility.
+
+Not every decision requires extended debate.
+Not every intervention requires public explanation.
+
+All decisions are expected to serve the durability, clarity, and integrity of Telemt.
 
 > **Ordnung ist Voraussetzung der Funktion.**
+
 > Order is the precondition of function.
 
 ---
 
-## 8. Maintainer Authority
+## 9. Enforcement
 
-Maintainers have final authority in interpretation and enforcement.
+Maintainers may act to preserve the integrity of Telemt, including by:
 
-Authority exists to ensure continuity, consistency, and technical direction.
+* Removing content
+* Locking discussions
+* Rejecting contributions
+* Restricting or banning participants
+
+Actions are taken to maintain function, continuity, and signal quality.
+
+Where possible, correction is preferred to exclusion.
+
+Where necessary, exclusion is preferred to decay.
 
 ---
 
-## 9. Final
+## 10. Final
 
 Telemt is built on discipline, structure, and shared intent.
 
@@ -132,19 +171,23 @@ Signal over noise.
 Facts over opinion.
 Systems over rhetoric.
 
-Work here is collective.
+Work is collective.
 Outcomes are shared.
 Responsibility is distributed.
 
+Precision is learned.
+Rigor is expected.
+Help is part of the work.
+
 > **Ordnung ist Voraussetzung der Freiheit.**
 
-If you contribute — contribute with precision.
+If you contribute — contribute with care.
 If you speak — speak with substance.
 If you engage — engage constructively.
 
 ---
 
-## 10. After All
+## 11. After All
 
 Systems outlive intentions.
 
@@ -152,12 +195,14 @@ What is built will be used.
 What is released will propagate.
 What is maintained will define the future state.
 
-There is no neutral infrastructure.
+There is no neutral infrastructure, only infrastructure shaped well or poorly.
 
 > **Jedes System trägt Verantwortung.**
+
 > Every system carries responsibility.
 
 Stability requires discipline.
 Freedom requires structure.
+Trust requires honesty.
 
 In the end, the system reflects its contributors.

Cargo.lock

@@ -69,15 +69,15 @@ checksum = "4b46cbb362ab8752921c97e041f5e366ee6297bd428a31275b9fcf1e380f7299"
 
 [[package]]
 name = "anstyle"
-version = "1.0.13"
+version = "1.0.14"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "5192cca8006f1fd4f7237516f40fa183bb07f8fbdfedaa0036de5ea9b0b45e78"
+checksum = "940b3a0ca603d1eade50a4846a2afffd5ef57a9feac2c0e2ec2e14f9ead76000"
 
 [[package]]
 name = "anyhow"
-version = "1.0.101"
+version = "1.0.102"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "5f0e0fee31ef5ed1ba1316088939cea399010ed7731dba877ed44aeb407a75ea"
+checksum = "7f202df86484c868dbad7eaa557ef785d5c66295e41b460ef922eca0723b842c"
 
 [[package]]
 name = "arc-swap"
@@ -147,7 +147,7 @@ checksum = "9035ad2d096bed7955a320ee7e2230574d28fd3c3a0f186cbea1ff3c7eed5dbb"
 dependencies = [
  "proc-macro2",
  "quote",
- "syn 2.0.114",
+ "syn 2.0.117",
 ]
 
 [[package]]
@@ -197,9 +197,9 @@ checksum = "bef38d45163c2f1dde094a7dfd33ccf595c92905c8f8f4fdc18d06fb1037718a"
 
 [[package]]
 name = "bitflags"
-version = "2.10.0"
+version = "2.11.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "812e12b5285cc515a9c72a5c1d3b6d46a19dac5acfef5265968c166106e31dd3"
+checksum = "843867be96c8daad0d758b57df9392b6d8d271134fce549de6ce169ff98a92af"
 
 [[package]]
 name = "blake3"
@@ -235,9 +235,9 @@ dependencies = [
 
 [[package]]
 name = "bumpalo"
-version = "3.19.1"
+version = "3.20.2"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "5dd9dc738b7a8311c7ade152424974d8115f2cdad61e8dab8dac9f2362298510"
+checksum = "5d20789868f4b01b2f2caec9f5c4e0213b41e3e5702a50157d699ae31ced2fcb"
 
 [[package]]
 name = "byte_string"
@@ -268,9 +268,9 @@ dependencies = [
 
 [[package]]
 name = "cc"
-version = "1.2.55"
+version = "1.2.57"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "47b26a0954ae34af09b50f0de26458fa95369a0d478d8236d3f93082b219bd29"
+checksum = "7a0dd1ca384932ff3641c8718a02769f1698e7563dc6974ffd03346116310423"
 dependencies = [
  "find-msvc-tools",
  "shlex",
@@ -320,9 +320,9 @@ dependencies = [
 
 [[package]]
 name = "chrono"
-version = "0.4.43"
+version = "0.4.44"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "fac4744fb15ae8337dc853fee7fb3f4e48c0fbaa23d0afe49c447b4fab126118"
+checksum = "c673075a2e0e5f4a1dde27ce9dee1ea4558c7ffe648f576438a20ca1d2acc4b0"
 dependencies = [
  "iana-time-zone",
  "js-sys",
@@ -372,18 +372,18 @@ dependencies = [
 
 [[package]]
 name = "clap"
-version = "4.5.58"
+version = "4.6.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "63be97961acde393029492ce0be7a1af7e323e6bae9511ebfac33751be5e6806"
+checksum = "b193af5b67834b676abd72466a96c1024e6a6ad978a1f484bd90b85c94041351"
 dependencies = [
  "clap_builder",
 ]
 
 [[package]]
 name = "clap_builder"
-version = "4.5.58"
+version = "4.6.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "7f13174bda5dfd69d7e947827e5af4b0f2f94a4a3ee92912fba07a66150f21e2"
+checksum = "714a53001bf66416adb0e2ef5ac857140e7dc3a0c48fb28b2f10762fc4b5069f"
 dependencies = [
  "anstyle",
  "clap_lex",
@@ -391,9 +391,9 @@ dependencies = [
 
 [[package]]
 name = "clap_lex"
-version = "1.0.0"
+version = "1.1.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "3a822ea5bc7590f9d40f1ba12c0dc3c2760f3482c6984db1573ad11031420831"
+checksum = "c8d4a3bb8b1e0c1050499d1815f5ab16d04f0959b233085fb31653fbfc9d98f9"
 
 [[package]]
 name = "const-oid"
@@ -622,7 +622,7 @@ checksum = "97369cbbc041bc366949bc74d34658d6cda5621039731c6310521892a3a20ae0"
 dependencies = [
  "proc-macro2",
  "quote",
- "syn 2.0.114",
+ "syn 2.0.117",
 ]
 
 [[package]]
@@ -642,7 +642,7 @@ checksum = "0b0713d5c1d52e774c5cd7bb8b043d7c0fc4f921abfb678556140bfbe6ab2364"
 dependencies = [
  "proc-macro2",
  "quote",
- "syn 2.0.114",
+ "syn 2.0.117",
 ]
 
 [[package]]
@@ -670,7 +670,7 @@ dependencies = [
  "heck",
  "proc-macro2",
  "quote",
- "syn 2.0.114",
+ "syn 2.0.117",
 ]
 
 [[package]]
@@ -750,9 +750,9 @@ dependencies = [
 
 [[package]]
 name = "futures"
-version = "0.3.31"
+version = "0.3.32"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "65bc07b1a8bc7c85c5f2e110c476c7389b4554ba72af57d8445ea63a576b0876"
+checksum = "8b147ee9d1f6d097cef9ce628cd2ee62288d963e16fb287bd9286455b241382d"
 dependencies = [
  "futures-channel",
  "futures-core",
@@ -765,9 +765,9 @@ dependencies = [
 
 [[package]]
 name = "futures-channel"
-version = "0.3.31"
+version = "0.3.32"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "2dff15bf788c671c1934e366d07e30c1814a8ef514e1af724a602e8a2fbe1b10"
+checksum = "07bbe89c50d7a535e539b8c17bc0b49bdb77747034daa8087407d655f3f7cc1d"
 dependencies = [
  "futures-core",
  "futures-sink",
@@ -775,15 +775,15 @@ dependencies = [
 
 [[package]]
 name = "futures-core"
-version = "0.3.31"
+version = "0.3.32"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "05f29059c0c2090612e8d742178b0580d2dc940c837851ad723096f87af6663e"
+checksum = "7e3450815272ef58cec6d564423f6e755e25379b217b0bc688e295ba24df6b1d"
 
 [[package]]
 name = "futures-executor"
-version = "0.3.31"
+version = "0.3.32"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "1e28d1d997f585e54aebc3f97d39e72338912123a67330d723fdbb564d646c9f"
+checksum = "baf29c38818342a3b26b5b923639e7b1f4a61fc5e76102d4b1981c6dc7a7579d"
 dependencies = [
  "futures-core",
  "futures-task",
@@ -792,38 +792,38 @@ dependencies = [
 
 [[package]]
 name = "futures-io"
-version = "0.3.31"
+version = "0.3.32"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "9e5c1b78ca4aae1ac06c48a526a655760685149f0d465d21f37abfe57ce075c6"
+checksum = "cecba35d7ad927e23624b22ad55235f2239cfa44fd10428eecbeba6d6a717718"
 
 [[package]]
 name = "futures-macro"
-version = "0.3.31"
+version = "0.3.32"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "162ee34ebcb7c64a8abebc059ce0fee27c2262618d7b60ed8faf72fef13c3650"
+checksum = "e835b70203e41293343137df5c0664546da5745f82ec9b84d40be8336958447b"
 dependencies = [
  "proc-macro2",
  "quote",
- "syn 2.0.114",
+ "syn 2.0.117",
 ]
 
 [[package]]
 name = "futures-sink"
-version = "0.3.31"
+version = "0.3.32"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "e575fab7d1e0dcb8d0c7bcf9a63ee213816ab51902e6d244a95819acacf1d4f7"
+checksum = "c39754e157331b013978ec91992bde1ac089843443c49cbc7f46150b0fad0893"
 
 [[package]]
 name = "futures-task"
-version = "0.3.31"
+version = "0.3.32"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "f90f7dce0722e95104fcb095585910c0977252f286e354b5e3bd38902cd99988"
+checksum = "037711b3d59c33004d3856fbdc83b99d4ff37a24768fa1be9ce3538a1cde4393"
 
 [[package]]
 name = "futures-util"
-version = "0.3.31"
+version = "0.3.32"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "9fa08315bb612088cc391249efdc3bc77536f16c91f6cf495e6fbe85b20a4a81"
+checksum = "389ca41296e6190b48053de0321d02a77f32f8a5d2461dd38762c0593805c6d6"
 dependencies = [
  "futures-channel",
  "futures-core",
@@ -833,7 +833,6 @@ dependencies = [
  "futures-task",
  "memchr",
  "pin-project-lite",
- "pin-utils",
  "slab",
 ]
 
@@ -869,20 +868,20 @@ dependencies = [
  "cfg-if",
  "js-sys",
  "libc",
- "r-efi",
+ "r-efi 5.3.0",
  "wasip2",
  "wasm-bindgen",
 ]
 
 [[package]]
 name = "getrandom"
-version = "0.4.1"
+version = "0.4.2"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "139ef39800118c7683f2fd3c98c1b23c09ae076556b435f8e9064ae108aaeeec"
+checksum = "0de51e6874e94e7bf76d726fc5d13ba782deca734ff60d5bb2fb2607c7406555"
 dependencies = [
  "cfg-if",
  "libc",
- "r-efi",
+ "r-efi 6.0.0",
  "wasip2",
  "wasip3",
 ]
@@ -1137,7 +1136,7 @@ dependencies = [
  "libc",
  "percent-encoding",
  "pin-project-lite",
- "socket2 0.6.2",
+ "socket2 0.6.3",
  "tokio",
  "tower-service",
  "tracing",
@@ -1304,7 +1303,7 @@ version = "0.11.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "bd5b3eaf1a28b758ac0faa5a4254e8ab2705605496f1b1f3fbbc3988ad73d199"
 dependencies = [
- "bitflags 2.10.0",
+ "bitflags 2.11.0",
  "inotify-sys",
  "libc",
 ]
@@ -1342,9 +1341,9 @@ dependencies = [
 
 [[package]]
 name = "ipnet"
-version = "2.11.0"
+version = "2.12.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "469fb0b9cefa57e3ef31275ee7cacb78f2fdca44e4765491884a2b119d4eb130"
+checksum = "d98f6fed1fde3f8c21bc40a1abb88dd75e67924f9cffc3ef95607bad8017f8e2"
 
 [[package]]
 name = "ipnetwork"
@@ -1387,15 +1386,15 @@ dependencies = [
 
 [[package]]
 name = "itoa"
-version = "1.0.17"
+version = "1.0.18"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "92ecc6618181def0457392ccd0ee51198e065e016d1d527a7ac1b6dc7c1f09d2"
+checksum = "8f42a60cbdf9a97f5d2305f08a87dc4e09308d1276d28c869c684d7777685682"
 
 [[package]]
 name = "js-sys"
-version = "0.3.85"
+version = "0.3.91"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "8c942ebf8e95485ca0d52d97da7c5a2c387d0e7f0ba4c35e93bfcaee045955b3"
+checksum = "b49715b7073f385ba4bc528e5747d02e66cb39c6146efb66b781f131f0fb399c"
 dependencies = [
  "once_cell",
  "wasm-bindgen",
@@ -1435,26 +1434,27 @@ checksum = "09edd9e8b54e49e587e4f6295a7d29c3ea94d469cb40ab8ca70b288248a81db2"
 
 [[package]]
 name = "libc"
-version = "0.2.181"
+version = "0.2.183"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "459427e2af2b9c839b132acb702a1c654d95e10f8c326bfc2ad11310e458b1c5"
+checksum = "b5b646652bf6661599e1da8901b3b9522896f01e736bad5f723fe7a3a27f899d"
 
 [[package]]
 name = "libredox"
-version = "0.1.12"
+version = "0.1.14"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "3d0b95e02c851351f877147b7deea7b1afb1df71b63aa5f8270716e0c5720616"
+checksum = "1744e39d1d6a9948f4f388969627434e31128196de472883b39f148769bfe30a"
 dependencies = [
- "bitflags 2.10.0",
+ "bitflags 2.11.0",
  "libc",
- "redox_syscall 0.7.1",
+ "plain",
+ "redox_syscall 0.7.3",
 ]
 
 [[package]]
 name = "linux-raw-sys"
-version = "0.11.0"
+version = "0.12.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "df1d3c3b53da64cf5760482273a98e575c651a67eec7f77df96b5b642de8f039"
+checksum = "32a66949e030da00e8c7d4434b251670a91556f4144941d37452769c25d58a53"
 
 [[package]]
 name = "litemap"
@@ -1585,7 +1585,7 @@ version = "0.28.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "ab2156c4fce2f8df6c499cc1c763e4394b7482525bf2a9701c9d79d215f519e4"
 dependencies = [
- "bitflags 2.10.0",
+ "bitflags 2.11.0",
  "cfg-if",
  "cfg_aliases 0.1.1",
  "libc",
@@ -1608,7 +1608,7 @@ version = "6.1.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "6205bd8bb1e454ad2e27422015fb5e4f2bcc7e08fa8f27058670d208324a4d2d"
 dependencies = [
- "bitflags 2.10.0",
+ "bitflags 2.11.0",
  "crossbeam-channel",
  "filetime",
  "fsevent-sys",
@@ -1627,7 +1627,7 @@ version = "8.2.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "4d3d07927151ff8575b7087f245456e549fea62edf0ec4e565a5ee50c8402bc3"
 dependencies = [
- "bitflags 2.10.0",
+ "bitflags 2.11.0",
  "fsevent-sys",
  "inotify 0.11.1",
  "kqueue",
@@ -1645,7 +1645,7 @@ version = "2.1.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "42b8cfee0e339a0337359f3c88165702ac6e600dc01c0cc9579a92d62b08477a"
 dependencies = [
- "bitflags 2.10.0",
+ "bitflags 2.11.0",
 ]
 
 [[package]]
@@ -1702,9 +1702,9 @@ dependencies = [
 
 [[package]]
 name = "once_cell"
-version = "1.21.3"
+version = "1.21.4"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "42f5e15c9953c5e4ccceeb2e7382a716482c34515315f7b03532b8b4e8393d2d"
+checksum = "9f7c3e4beb33f85d45ae3e3a1792185706c8e16d043238c593331cc7cd313b50"
 dependencies = [
  "critical-section",
  "portable-atomic",
@@ -1768,14 +1768,14 @@ checksum = "d9b20ed30f105399776b9c883e68e536ef602a16ae6f596d2c473591d6ad64c6"
 dependencies = [
  "proc-macro2",
  "quote",
- "syn 2.0.114",
+ "syn 2.0.117",
 ]
 
 [[package]]
 name = "pin-project-lite"
-version = "0.2.16"
+version = "0.2.17"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "3b3cff922bd51709b605d9ead9aa71031d81447142d828eb4a6eba76fe619f9b"
+checksum = "a89322df9ebe1c1578d689c92318e070967d1042b512afbe49518723f4e6d5cd"
 
 [[package]]
 name = "pin-utils"
@@ -1793,6 +1793,12 @@ dependencies = [
  "spki",
 ]
 
+[[package]]
+name = "plain"
+version = "0.2.3"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "b4596b6d070b27117e987119b4dac604f3c58cfb0b191112e24771b2faeac1a6"
+
 [[package]]
 name = "plotters"
 version = "0.3.7"
@@ -1881,7 +1887,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "479ca8adacdd7ce8f1fb39ce9ecccbfe93a3f1344b3d0d97f20bc0196208f62b"
 dependencies = [
  "proc-macro2",
- "syn 2.0.114",
+ "syn 2.0.117",
 ]
 
 [[package]]
@@ -1901,7 +1907,7 @@ checksum = "37566cb3fdacef14c0737f9546df7cfeadbfbc9fef10991038bf5015d0c80532"
 dependencies = [
  "bit-set",
  "bit-vec",
- "bitflags 2.10.0",
+ "bitflags 2.11.0",
  "num-traits",
  "rand",
  "rand_chacha",
@@ -1931,7 +1937,7 @@ dependencies = [
  "quinn-udp",
  "rustc-hash",
  "rustls",
- "socket2 0.6.2",
+ "socket2 0.6.3",
  "thiserror 2.0.18",
  "tokio",
  "tracing",
@@ -1940,9 +1946,9 @@ dependencies = [
 
 [[package]]
 name = "quinn-proto"
-version = "0.11.13"
+version = "0.11.14"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "f1906b49b0c3bc04b5fe5d86a77925ae6524a19b816ae38ce1e426255f1d8a31"
+checksum = "434b42fec591c96ef50e21e886936e66d3cc3f737104fdb9b737c40ffb94c098"
 dependencies = [
  "bytes",
  "getrandom 0.3.4",
@@ -1968,16 +1974,16 @@ dependencies = [
  "cfg_aliases 0.2.1",
  "libc",
  "once_cell",
- "socket2 0.6.2",
+ "socket2 0.6.3",
  "tracing",
  "windows-sys 0.60.2",
 ]
 
 [[package]]
 name = "quote"
-version = "1.0.44"
+version = "1.0.45"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "21b2ebcf727b7760c461f091f9f0f539b77b8e87f2fd88131e7f1b433b3cece4"
+checksum = "41f2619966050689382d2b44f664f4bc593e129785a36d6ee376ddf37259b924"
 dependencies = [
  "proc-macro2",
 ]
@@ -1988,6 +1994,12 @@ version = "5.3.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "69cdb34c158ceb288df11e18b4bd39de994f6657d83847bdffdbd7f346754b0f"
 
+[[package]]
+name = "r-efi"
+version = "6.0.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "f8dcc9c7d52a811697d2151c701e0d08956f92b0e24136cf4cf27b57a6a0d9bf"
+
 [[package]]
 name = "rand"
 version = "0.9.2"
@@ -2061,16 +2073,16 @@ version = "0.5.18"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "ed2bf2547551a7053d6fdfafda3f938979645c44812fbfcda098faae3f1a362d"
 dependencies = [
- "bitflags 2.10.0",
+ "bitflags 2.11.0",
 ]
 
 [[package]]
 name = "redox_syscall"
-version = "0.7.1"
+version = "0.7.3"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "35985aa610addc02e24fc232012c86fd11f14111180f902b67e2d5331f8ebf2b"
+checksum = "6ce70a74e890531977d37e532c34d45e9055d2409ed08ddba14529471ed0be16"
 dependencies = [
- "bitflags 2.10.0",
+ "bitflags 2.11.0",
 ]
 
 [[package]]
@@ -2098,9 +2110,9 @@ dependencies = [
 
 [[package]]
 name = "regex-syntax"
-version = "0.8.9"
+version = "0.8.10"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "a96887878f22d7bad8a3b6dc5b7440e0ada9a245242924394987b21cf2210a4c"
+checksum = "dc897dd8d9e8bd1ed8cdad82b5966c3e0ecae09fb1907d58efaa013543185d0a"
 
 [[package]]
 name = "reqwest"
@@ -2199,11 +2211,11 @@ dependencies = [
 
 [[package]]
 name = "rustix"
-version = "1.1.3"
+version = "1.1.4"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "146c9e247ccc180c1f61615433868c99f3de3ae256a30a43b49f67c2d9171f34"
+checksum = "b6fe4565b9518b83ef4f91bb47ce29620ca828bd32cb7e408f0062e9930ba190"
 dependencies = [
- "bitflags 2.10.0",
+ "bitflags 2.11.0",
  "errno",
  "libc",
  "linux-raw-sys",
@@ -2212,9 +2224,9 @@ dependencies = [
 
 [[package]]
 name = "rustls"
-version = "0.23.36"
+version = "0.23.37"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "c665f33d38cea657d9614f766881e4d510e0eda4239891eea56b4cadcf01801b"
+checksum = "758025cb5fccfd3bc2fd74708fd4682be41d99e5dff73c377c0646c6012c73a4"
 dependencies = [
  "once_cell",
  "ring",
@@ -2292,7 +2304,7 @@ checksum = "22f968c5ea23d555e670b449c1c5e7b2fc399fdaec1d304a17cd48e288abc107"
 dependencies = [
  "proc-macro2",
  "quote",
- "syn 2.0.114",
+ "syn 2.0.117",
 ]
 
 [[package]]
@@ -2338,7 +2350,7 @@ checksum = "d540f220d3187173da220f885ab66608367b6574e925011a9353e4badda91d79"
 dependencies = [
  "proc-macro2",
  "quote",
- "syn 2.0.114",
+ "syn 2.0.117",
 ]
 
 [[package]]
@@ -2426,7 +2438,7 @@ dependencies = [
  "serde_json",
  "serde_urlencoded",
  "shadowsocks-crypto",
- "socket2 0.6.2",
+ "socket2 0.6.3",
  "spin",
  "thiserror 2.0.18",
  "tokio",
@@ -2510,12 +2522,12 @@ dependencies = [
 
 [[package]]
 name = "socket2"
-version = "0.6.2"
+version = "0.6.3"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "86f4aa3ad99f2088c990dfa82d367e19cb29268ed67c574d10d0a4bfe71f07e0"
+checksum = "3a766e1110788c36f4fa1c2b71b387a7815aa65f88ce0229841826633d93723e"
 dependencies = [
  "libc",
- "windows-sys 0.60.2",
+ "windows-sys 0.61.2",
 ]
 
 [[package]]
@@ -2562,9 +2574,9 @@ dependencies = [
 
 [[package]]
 name = "syn"
-version = "2.0.114"
+version = "2.0.117"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "d4d107df263a3013ef9b1879b0df87d706ff80f65a86ea879bd9c31f9b307c2a"
+checksum = "e665b8803e7b1d2a727f4023456bbbbe74da67099c585258af0ad9c5013b9b99"
 dependencies = [
  "proc-macro2",
  "quote",
@@ -2600,7 +2612,7 @@ checksum = "728a70f3dbaf5bab7f0c4b1ac8d7ae5ea60a4b5549c8a5914361c99147a709d2"
 dependencies = [
  "proc-macro2",
  "quote",
- "syn 2.0.114",
+ "syn 2.0.117",
 ]
 
 [[package]]
@@ -2611,10 +2623,11 @@ checksum = "7b2093cf4c8eb1e67749a6762251bc9cd836b6fc171623bd0a9d324d37af2417"
 
 [[package]]
 name = "telemt"
-version = "3.3.20"
+version = "3.3.28"
 dependencies = [
  "aes",
  "anyhow",
+ "arc-swap",
  "base64",
  "bytes",
  "cbc",
@@ -2668,12 +2681,12 @@ dependencies = [
 
 [[package]]
 name = "tempfile"
-version = "3.25.0"
+version = "3.27.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "0136791f7c95b1f6dd99f9cc786b91bb81c3800b639b3478e561ddb7be95e5f1"
+checksum = "32497e9a4c7b38532efcdebeef879707aa9f794296a4f0244f6f69e9bc8574bd"
 dependencies = [
  "fastrand",
- "getrandom 0.4.1",
+ "getrandom 0.4.2",
  "once_cell",
  "rustix",
  "windows-sys 0.61.2",
@@ -2705,7 +2718,7 @@ checksum = "4fee6c4efc90059e10f81e6d42c60a18f76588c3d74cb83a0b242a2b6c7504c1"
 dependencies = [
  "proc-macro2",
  "quote",
- "syn 2.0.114",
+ "syn 2.0.117",
 ]
 
 [[package]]
@@ -2716,7 +2729,7 @@ checksum = "ebc4ee7f67670e9b64d05fa4253e753e016c6c95ff35b89b7941d6b856dec1d5"
 dependencies = [
  "proc-macro2",
  "quote",
- "syn 2.0.114",
+ "syn 2.0.117",
 ]
 
 [[package]]
@@ -2781,9 +2794,9 @@ dependencies = [
 
 [[package]]
 name = "tinyvec"
-version = "1.10.0"
+version = "1.11.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "bfa5fdc3bce6191a1dbc8c02d5c8bffcf557bafa17c124c5264a458f1b0613fa"
+checksum = "3e61e67053d25a4e82c844e8424039d9745781b3fc4f32b8d55ed50f5f667ef3"
 dependencies = [
  "tinyvec_macros",
 ]
@@ -2796,9 +2809,9 @@ checksum = "1f3ccbac311fea05f86f61904b462b55fb3df8837a366dfc601a0161d0532f20"
 
 [[package]]
 name = "tokio"
-version = "1.49.0"
+version = "1.50.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "72a2903cd7736441aac9df9d7688bd0ce48edccaadf181c3b90be801e81d3d86"
+checksum = "27ad5e34374e03cfffefc301becb44e9dc3c17584f414349ebe29ed26661822d"
 dependencies = [
  "bytes",
  "libc",
@@ -2806,7 +2819,7 @@ dependencies = [
  "parking_lot",
  "pin-project-lite",
  "signal-hook-registry",
- "socket2 0.6.2",
+ "socket2 0.6.3",
  "tokio-macros",
  "tracing",
  "windows-sys 0.61.2",
@@ -2814,13 +2827,13 @@ dependencies = [
 
 [[package]]
 name = "tokio-macros"
-version = "2.6.0"
+version = "2.6.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "af407857209536a95c8e56f8231ef2c2e2aff839b22e07a1ffcbc617e9db9fa5"
+checksum = "5c55a2eff8b69ce66c84f85e1da1c233edc36ceb85a2058d11b0d6a3c7e7569c"
 dependencies = [
  "proc-macro2",
  "quote",
- "syn 2.0.114",
+ "syn 2.0.117",
 ]
 
 [[package]]
@@ -2867,7 +2880,7 @@ dependencies = [
  "log",
  "once_cell",
  "pin-project",
- "socket2 0.6.2",
+ "socket2 0.6.3",
  "tokio",
  "windows-sys 0.60.2",
 ]
@@ -2951,7 +2964,7 @@ version = "0.6.8"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "d4e6559d53cc268e5031cd8429d05415bc4cb4aefc4aa5d6cc35fbf5b924a1f8"
 dependencies = [
- "bitflags 2.10.0",
+ "bitflags 2.11.0",
  "bytes",
  "futures-util",
  "http",
@@ -2994,7 +3007,7 @@ checksum = "7490cfa5ec963746568740651ac6781f701c9c5ea257c58e057f3ba8cf69e8da"
 dependencies = [
  "proc-macro2",
  "quote",
- "syn 2.0.114",
+ "syn 2.0.117",
 ]
 
 [[package]]
@@ -3020,9 +3033,9 @@ dependencies = [
 
 [[package]]
 name = "tracing-subscriber"
-version = "0.3.22"
+version = "0.3.23"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "2f30143827ddab0d256fd843b7a66d164e9f271cfa0dde49142c5ca0ca291f1e"
+checksum = "cb7f578e5945fb242538965c2d0b04418d38ec25c79d160cd279bf0731c8d319"
 dependencies = [
  "matchers",
  "nu-ansi-term",
@@ -3044,7 +3057,7 @@ checksum = "70977707304198400eb4835a78f6a9f928bf41bba420deb8fdb175cd965d77a7"
 dependencies = [
  "proc-macro2",
  "quote",
- "syn 2.0.114",
+ "syn 2.0.117",
 ]
 
 [[package]]
@@ -3067,9 +3080,9 @@ checksum = "eaea85b334db583fe3274d12b4cd1880032beab409c0d774be044d4480ab9a94"
 
 [[package]]
 name = "unicode-ident"
-version = "1.0.23"
+version = "1.0.24"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "537dd038a89878be9b64dd4bd1b260315c1bb94f4d784956b81e27a088d9a09e"
+checksum = "e6e4313cd5fcd3dad5cafa179702e2b244f760991f45397d14d4ebf38247da75"
 
 [[package]]
 name = "unicode-xid"
@@ -3117,7 +3130,7 @@ version = "1.22.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "a68d3c8f01c0cfa54a75291d83601161799e4a89a39e0929f4b0354d88757a37"
 dependencies = [
- "getrandom 0.4.1",
+ "getrandom 0.4.2",
  "js-sys",
  "wasm-bindgen",
 ]
@@ -3188,9 +3201,9 @@ dependencies = [
 
 [[package]]
 name = "wasm-bindgen"
-version = "0.2.108"
+version = "0.2.114"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "64024a30ec1e37399cf85a7ffefebdb72205ca1c972291c51512360d90bd8566"
+checksum = "6532f9a5c1ece3798cb1c2cfdba640b9b3ba884f5db45973a6f442510a87d38e"
 dependencies = [
  "cfg-if",
  "once_cell",
@@ -3201,9 +3214,9 @@ dependencies = [
 
 [[package]]
 name = "wasm-bindgen-futures"
-version = "0.4.58"
+version = "0.4.64"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "70a6e77fd0ae8029c9ea0063f87c46fde723e7d887703d74ad2616d792e51e6f"
+checksum = "e9c5522b3a28661442748e09d40924dfb9ca614b21c00d3fd135720e48b67db8"
 dependencies = [
  "cfg-if",
  "futures-util",
@@ -3215,9 +3228,9 @@ dependencies = [
 
 [[package]]
 name = "wasm-bindgen-macro"
-version = "0.2.108"
+version = "0.2.114"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "008b239d9c740232e71bd39e8ef6429d27097518b6b30bdf9086833bd5b6d608"
+checksum = "18a2d50fcf105fb33bb15f00e7a77b772945a2ee45dcf454961fd843e74c18e6"
 dependencies = [
  "quote",
  "wasm-bindgen-macro-support",
@@ -3225,22 +3238,22 @@ dependencies = [
 
 [[package]]
 name = "wasm-bindgen-macro-support"
-version = "0.2.108"
+version = "0.2.114"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "5256bae2d58f54820e6490f9839c49780dff84c65aeab9e772f15d5f0e913a55"
+checksum = "03ce4caeaac547cdf713d280eda22a730824dd11e6b8c3ca9e42247b25c631e3"
 dependencies = [
  "bumpalo",
  "proc-macro2",
  "quote",
- "syn 2.0.114",
+ "syn 2.0.117",
  "wasm-bindgen-shared",
 ]
 
 [[package]]
 name = "wasm-bindgen-shared"
-version = "0.2.108"
+version = "0.2.114"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "1f01b580c9ac74c8d8f0c0e4afb04eeef2acf145458e52c03845ee9cd23e3d12"
+checksum = "75a326b8c223ee17883a4251907455a2431acc2791c98c26279376490c378c16"
 dependencies = [
  "unicode-ident",
 ]
@@ -3273,7 +3286,7 @@ version = "0.244.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "47b807c72e1bac69382b3a6fb3dbe8ea4c0ed87ff5629b8685ae6b9a611028fe"
 dependencies = [
- "bitflags 2.10.0",
+ "bitflags 2.11.0",
  "hashbrown 0.15.5",
  "indexmap",
  "semver",
@@ -3281,9 +3294,9 @@ dependencies = [
 
 [[package]]
 name = "web-sys"
-version = "0.3.85"
+version = "0.3.91"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "312e32e551d92129218ea9a2452120f4aabc03529ef03e4d0d82fb2780608598"
+checksum = "854ba17bb104abfb26ba36da9729addc7ce7f06f5c0f90f3c391f8461cca21f9"
 dependencies = [
  "js-sys",
  "wasm-bindgen",
@@ -3353,7 +3366,7 @@ checksum = "053e2e040ab57b9dc951b72c264860db7eb3b0200ba345b4e4c3b14f67855ddf"
 dependencies = [
  "proc-macro2",
  "quote",
- "syn 2.0.114",
+ "syn 2.0.117",
 ]
 
 [[package]]
@@ -3364,7 +3377,7 @@ checksum = "3f316c4a2570ba26bbec722032c4099d8c8bc095efccdc15688708623367e358"
 dependencies = [
  "proc-macro2",
  "quote",
- "syn 2.0.114",
+ "syn 2.0.117",
 ]
 
 [[package]]
@@ -3615,9 +3628,9 @@ checksum = "d6bbff5f0aada427a1e5a6da5f1f98158182f26556f345ac9e04d36d0ebed650"
 
 [[package]]
 name = "winnow"
-version = "0.7.14"
+version = "0.7.15"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "5a5364e9d77fcdeeaa6062ced926ee3381faa2ee02d3eb83a5c27a8825540829"
+checksum = "df79d97927682d2fd8adb29682d1140b343be4ac0f08fd68b7765d9c059d3945"
 dependencies = [
  "memchr",
 ]
@@ -3662,7 +3675,7 @@ dependencies = [
  "heck",
  "indexmap",
  "prettyplease",
- "syn 2.0.114",
+ "syn 2.0.117",
  "wasm-metadata",
  "wit-bindgen-core",
  "wit-component",
@@ -3678,7 +3691,7 @@ dependencies = [
  "prettyplease",
  "proc-macro2",
  "quote",
- "syn 2.0.114",
+ "syn 2.0.117",
  "wit-bindgen-core",
  "wit-bindgen-rust",
 ]
@@ -3690,7 +3703,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "9d66ea20e9553b30172b5e831994e35fbde2d165325bec84fc43dbf6f4eb9cb2"
 dependencies = [
  "anyhow",
- "bitflags 2.10.0",
+ "bitflags 2.11.0",
  "indexmap",
  "log",
  "serde",
@@ -3762,28 +3775,28 @@ checksum = "b659052874eb698efe5b9e8cf382204678a0086ebf46982b79d6ca3182927e5d"
 dependencies = [
  "proc-macro2",
  "quote",
- "syn 2.0.114",
+ "syn 2.0.117",
  "synstructure 0.13.2",
 ]
 
 [[package]]
 name = "zerocopy"
-version = "0.8.39"
+version = "0.8.47"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "db6d35d663eadb6c932438e763b262fe1a70987f9ae936e60158176d710cae4a"
+checksum = "efbb2a062be311f2ba113ce66f697a4dc589f85e78a4aea276200804cea0ed87"
 dependencies = [
  "zerocopy-derive",
 ]
 
 [[package]]
 name = "zerocopy-derive"
-version = "0.8.39"
+version = "0.8.47"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "4122cd3169e94605190e77839c9a40d40ed048d305bfdc146e7df40ab0f3e517"
+checksum = "0e8bc7269b54418e7aeeef514aa68f8690b8c0489a06b0136e5f57c4c5ccab89"
 dependencies = [
  "proc-macro2",
  "quote",
- "syn 2.0.114",
+ "syn 2.0.117",
 ]
 
 [[package]]
@@ -3803,7 +3816,7 @@ checksum = "d71e5d6e06ab090c67b5e44993ec16b72dcbaabc526db883a360057678b48502"
 dependencies = [
  "proc-macro2",
  "quote",
- "syn 2.0.114",
+ "syn 2.0.117",
  "synstructure 0.13.2",
 ]
 
@@ -3824,7 +3837,7 @@ checksum = "85a5b4158499876c763cb03bc4e49185d3cccbabb15b33c627f7884f43db852e"
 dependencies = [
  "proc-macro2",
  "quote",
- "syn 2.0.114",
+ "syn 2.0.117",
 ]
 
 [[package]]
@@ -3857,11 +3870,11 @@ checksum = "eadce39539ca5cb3985590102671f2567e659fca9666581ad3411d59207951f3"
 dependencies = [
  "proc-macro2",
  "quote",
- "syn 2.0.114",
+ "syn 2.0.117",
 ]
 
 [[package]]
 name = "zmij"
-version = "1.0.20"
+version = "1.0.21"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "4de98dfa5d5b7fef4ee834d0073d560c9ca7b6c46a71d058c48db7960f8cfaf7"
+checksum = "b8848ee67ecc8aedbaf3e4122217aff892639231befc6a1b58d29fff4c2cabaa"

Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "telemt"
-version = "3.3.23"
+version = "3.3.28"
 edition = "2024"
 
 [dependencies]
@@ -41,6 +41,7 @@ tracing = "0.1"
 tracing-subscriber = { version = "0.3", features = ["env-filter"] }
 parking_lot = "0.12"
 dashmap = "5.5"
+arc-swap = "1.7"
 lru = "0.16"
 rand = "0.9"
 chrono = { version = "0.4", features = ["serde"] }

Dockerfile

@@ -1,3 +1,5 @@
+# syntax=docker/dockerfile:1
+
 # ==========================
 # Stage 1: Build
 # ==========================
@@ -5,39 +7,91 @@ FROM rust:1.88-slim-bookworm AS builder
 
 RUN apt-get update && apt-get install -y --no-install-recommends \
     pkg-config \
+    ca-certificates \
     && rm -rf /var/lib/apt/lists/*
 
 WORKDIR /build
 
+# Depcache
 COPY Cargo.toml Cargo.lock* ./
 RUN mkdir src && echo 'fn main() {}' > src/main.rs && \
     cargo build --release 2>/dev/null || true && \
     rm -rf src
 
+# Build
 COPY . .
 RUN cargo build --release && strip target/release/telemt
 
 # ==========================
-# Stage 2: Runtime
+# Stage 2: Compress (strip + UPX)
 # ==========================
-FROM debian:bookworm-slim
+FROM debian:12-slim AS minimal
+
+RUN apt-get update && apt-get install -y --no-install-recommends \
+    upx \
+    binutils \
+    && rm -rf /var/lib/apt/lists/*
+
+COPY --from=builder /build/target/release/telemt /telemt
+
+RUN strip /telemt || true
+RUN upx --best --lzma /telemt || true
+
+# ==========================
+# Stage 3: Debug base
+# ==========================
+FROM debian:12-slim AS debug-base
 
 RUN apt-get update && apt-get install -y --no-install-recommends \
     ca-certificates \
+    tzdata \
+    curl \
+    iproute2 \
+    busybox \
     && rm -rf /var/lib/apt/lists/*
 
-RUN useradd -r -s /usr/sbin/nologin telemt
+# ==========================
+# Stage 4: Debug image
+# ==========================
+FROM debug-base AS debug
 
 WORKDIR /app
 
-COPY --from=builder /build/target/release/telemt /app/telemt
+COPY --from=minimal /telemt /app/telemt
 COPY config.toml /app/config.toml
 
-RUN chown -R telemt:telemt /app
+USER root
-USER telemt
 
 EXPOSE 443
 EXPOSE 9090
+EXPOSE 9091
+
+ENTRYPOINT ["/app/telemt"]
+CMD ["config.toml"]
+
+# ==========================
+# Stage 5: Production (distroless)
+# ==========================
+FROM gcr.io/distroless/base-debian12 AS prod
+
+WORKDIR /app
+
+COPY --from=minimal /telemt /app/telemt
+COPY config.toml /app/config.toml
+
+# TLS + timezone + shell
+COPY --from=debug-base /etc/ssl/certs /etc/ssl/certs
+COPY --from=debug-base /usr/share/zoneinfo /usr/share/zoneinfo
+COPY --from=debug-base /bin/busybox /bin/busybox
+
+RUN ["/bin/busybox", "--install", "-s", "/bin"]
+
+# distroless user
+USER nonroot:nonroot
+
+EXPOSE 443
+EXPOSE 9090
+EXPOSE 9091
 
 ENTRYPOINT ["/app/telemt"]
 CMD ["config.toml"]

README.md

@@ -19,9 +19,9 @@
 
 ### 🇷🇺 RU
 
-#### Релиз 3.3.15 Semistable
+#### О релизах
 
-[3.3.15](https://github.com/telemt/telemt/releases/tag/3.3.15) по итогам работы в продакшн признан одним из самых стабильных и рекомендуется к использованию, когда cutting-edge фичи некритичны!
+[3.3.27](https://github.com/telemt/telemt/releases/tag/3.3.27) даёт баланс стабильности и передового функционала, а так же последние исправления по безопасности и багам
 
 Будем рады вашему фидбеку и предложениям по улучшению — особенно в части **API**, **статистики**, **UX**
 
@@ -40,9 +40,9 @@
 
 ### 🇬🇧 EN
 
-#### Release 3.3.15 Semistable
+#### About releases
 
-[3.3.15](https://github.com/telemt/telemt/releases/tag/3.3.15) is, based on the results of his work in production, recognized as one of the most stable and recommended for use when cutting-edge features are not so necessary!
+[3.3.27](https://github.com/telemt/telemt/releases/tag/3.3.27) provides a balance of stability and advanced functionality, as well as the latest security and bug fixes
 
 We are looking forward to your feedback and improvement proposals — especially regarding **API**, **statistics**, **UX**
 

docker-compose.yml

@@ -7,6 +7,7 @@ services:
     ports:
       - "443:443"
       - "127.0.0.1:9090:9090"
+      - "127.0.0.1:9091:9091"
     # Allow caching 'proxy-secret' in read-only container
     working_dir: /run/telemt
     volumes:

docs/QUICK_START_GUIDE.en.md

@@ -181,6 +181,8 @@ docker compose down
 docker build -t telemt:local .
 docker run --name telemt --restart unless-stopped \
   -p 443:443 \
+  -p 9090:9090 \
+  -p 9091:9091 \
   -e RUST_LOG=info \
   -v "$PWD/config.toml:/app/config.toml:ro" \
   --read-only \

docs/QUICK_START_GUIDE.ru.md

@@ -178,11 +178,13 @@ docker compose down
 > - По умолчанию публикуются порты 443:443, а контейнер запускается со сброшенными привилегиями (добавлена только `NET_BIND_SERVICE`)  
 > - Если вам действительно нужна сеть хоста (обычно это требуется только для некоторых конфигураций IPv6), раскомментируйте `network_mode: host`
 
-**Запуск в Docker Compose**
+**Запуск без Docker Compose**
 ```bash
 docker build -t telemt:local .
 docker run --name telemt --restart unless-stopped \
   -p 443:443 \
+  -p 9090:9090 \
+  -p 9091:9091 \
   -e RUST_LOG=info \
   -v "$PWD/config.toml:/app/config.toml:ro" \
   --read-only \

install.sh

@@ -1,154 +1,190 @@
 #!/bin/sh
 set -eu
 
-# --- Global Configurations ---
 REPO="${REPO:-telemt/telemt}"
 BIN_NAME="${BIN_NAME:-telemt}"
 INSTALL_DIR="${INSTALL_DIR:-/bin}"
 CONFIG_DIR="${CONFIG_DIR:-/etc/telemt}"
 CONFIG_FILE="${CONFIG_FILE:-${CONFIG_DIR}/telemt.toml}"
 WORK_DIR="${WORK_DIR:-/opt/telemt}"
+TLS_DOMAIN="${TLS_DOMAIN:-petrovich.ru}"
 SERVICE_NAME="telemt"
 TEMP_DIR=""
 SUDO=""
+CONFIG_PARENT_DIR=""
+SERVICE_START_FAILED=0
 
-# --- Argument Parsing ---
 ACTION="install"
 TARGET_VERSION="${VERSION:-latest}"
 
 while [ $# -gt 0 ]; do
     case "$1" in
-        -h|--help)
+        -h|--help) ACTION="help"; shift ;;
-            ACTION="help"
-            shift
-            ;;
         uninstall|--uninstall)
-            [ "$ACTION" != "purge" ] && ACTION="uninstall"
+            if [ "$ACTION" != "purge" ]; then ACTION="uninstall"; fi
-            shift
+            shift ;;
-            ;;
+        purge|--purge) ACTION="purge"; shift ;;
-        --purge)
+        install|--install) ACTION="install"; shift ;;
-            ACTION="purge"
+        -*) printf '[ERROR] Unknown option: %s\n' "$1" >&2; exit 1 ;;
-            shift
-            ;;
-        install|--install)
-            ACTION="install"
-            shift
-            ;;
-        -*)
-            printf '[ERROR] Unknown option: %s\n' "$1" >&2
-            exit 1
-            ;;
         *)
-            if [ "$ACTION" = "install" ]; then
+            if [ "$ACTION" = "install" ]; then TARGET_VERSION="$1"
-                TARGET_VERSION="$1"
+            else printf '[WARNING] Ignoring extra argument: %s\n' "$1" >&2; fi
-            fi
+            shift ;;
-            shift
-            ;;
     esac
 done
 
-# --- Core Functions ---
+say() {
-say() { printf '[INFO] %s\n' "$*"; }
+    if [ "$#" -eq 0 ] || [ -z "${1:-}" ]; then
+        printf '\n'
+    else
+        printf '[INFO] %s\n' "$*"
+    fi
+}
 die() { printf '[ERROR] %s\n' "$*" >&2; exit 1; }
 
+write_root() { $SUDO sh -c 'cat > "$1"' _ "$1"; }
+
 cleanup() {
     if [ -n "${TEMP_DIR:-}" ] && [ -d "$TEMP_DIR" ]; then
         rm -rf -- "$TEMP_DIR"
     fi
 }
-
 trap cleanup EXIT INT TERM
 
 show_help() {
-    say "Usage: $0 [version | install | uninstall | --purge | --help]"
+    say "Usage: $0 [ <version> | install | uninstall | purge | --help ]"
-    say "  version      Install specific version (e.g. 1.0.0, default: latest)"
+    say "  <version>    Install specific version (e.g. 3.3.15, default: latest)"
-    say "  uninstall    Remove the binary and service (keeps config)"
+    say "  install      Install the latest version"
-    say "  --purge      Remove everything including configuration"
+    say "  uninstall    Remove the binary and service (keeps config and user)"
+    say "  purge        Remove everything including configuration, data, and user"
     exit 0
 }
 
-user_exists() {
+check_os_entity() {
-    if command -v getent >/dev/null 2>&1; then
+    if command -v getent >/dev/null 2>&1; then getent "$1" "$2" >/dev/null 2>&1
-        getent passwd "$1" >/dev/null 2>&1
+    else grep -q "^${2}:" "/etc/$1" 2>/dev/null; fi
+}
+
+normalize_path() {
+    printf '%s\n' "$1" | tr -s '/' | sed 's|/$||; s|^$|/|'
+}
+
+get_realpath() {
+    path_in="$1"
+    case "$path_in" in /*) ;; *) path_in="$(pwd)/$path_in" ;; esac
+
+    if command -v realpath >/dev/null 2>&1; then 
+        if realpath_out="$(realpath -m "$path_in" 2>/dev/null)"; then
+            printf '%s\n' "$realpath_out"
+            return
+        fi
+    fi
+    
+    if command -v readlink >/dev/null 2>&1; then
+        resolved_path="$(readlink -f "$path_in" 2>/dev/null || true)"
+        if [ -n "$resolved_path" ]; then
+            printf '%s\n' "$resolved_path"
+            return
+        fi
+    fi
+
+    d="${path_in%/*}"; b="${path_in##*/}"
+    if [ -z "$d" ]; then d="/"; fi
+    if [ "$d" = "$path_in" ]; then d="/"; b="$path_in"; fi
+
+    if [ -d "$d" ]; then
+        abs_d="$(cd "$d" >/dev/null 2>&1 && pwd || true)"
+        if [ -n "$abs_d" ]; then
+            if [ "$b" = "." ] || [ -z "$b" ]; then printf '%s\n' "$abs_d"
+            elif [ "$abs_d" = "/" ]; then printf '/%s\n' "$b"
+            else printf '%s/%s\n' "$abs_d" "$b"; fi
+        else
+            normalize_path "$path_in"
+        fi
     else
-        grep -q "^${1}:" /etc/passwd 2>/dev/null
+        normalize_path "$path_in"
     fi
 }
 
-group_exists() {
+get_svc_mgr() {
-    if command -v getent >/dev/null 2>&1; then
+    if command -v systemctl >/dev/null 2>&1 && [ -d /run/systemd/system ]; then echo "systemd"
-        getent group "$1" >/dev/null 2>&1
+    elif command -v rc-service >/dev/null 2>&1; then echo "openrc"
-    else
+    else echo "none"; fi
-        grep -q "^${1}:" /etc/group 2>/dev/null
-    fi
 }
 
 verify_common() {
-    [ -z "$BIN_NAME" ] && die "BIN_NAME cannot be empty."
+    [ -n "$BIN_NAME" ] || die "BIN_NAME cannot be empty."
-    [ -z "$INSTALL_DIR" ] && die "INSTALL_DIR cannot be empty."
+    [ -n "$INSTALL_DIR" ] || die "INSTALL_DIR cannot be empty."
-    [ -z "$CONFIG_DIR" ] && die "CONFIG_DIR cannot be empty."
+    [ -n "$CONFIG_DIR" ] || die "CONFIG_DIR cannot be empty."
+    [ -n "$CONFIG_FILE" ] || die "CONFIG_FILE cannot be empty."
+
+    case "${INSTALL_DIR}${CONFIG_DIR}${WORK_DIR}${CONFIG_FILE}" in
+        *[!a-zA-Z0-9_./-]*) die "Invalid characters in paths. Only alphanumeric, _, ., -, and / allowed." ;;
+    esac
+
+    case "$TARGET_VERSION" in *[!a-zA-Z0-9_.-]*) die "Invalid characters in version." ;; esac
+    case "$BIN_NAME" in *[!a-zA-Z0-9_-]*) die "Invalid characters in BIN_NAME." ;; esac
+
+    INSTALL_DIR="$(get_realpath "$INSTALL_DIR")"
+    CONFIG_DIR="$(get_realpath "$CONFIG_DIR")"
+    WORK_DIR="$(get_realpath "$WORK_DIR")"
+    CONFIG_FILE="$(get_realpath "$CONFIG_FILE")"
+
+    CONFIG_PARENT_DIR="${CONFIG_FILE%/*}"
+    if [ -z "$CONFIG_PARENT_DIR" ]; then CONFIG_PARENT_DIR="/"; fi
+    if [ "$CONFIG_PARENT_DIR" = "$CONFIG_FILE" ]; then CONFIG_PARENT_DIR="."; fi
 
     if [ "$(id -u)" -eq 0 ]; then
         SUDO=""
     else
-        if ! command -v sudo >/dev/null 2>&1; then
+        command -v sudo >/dev/null 2>&1 || die "This script requires root or sudo. Neither found."
-            die "This script requires root or sudo. Neither found."
-        fi
         SUDO="sudo"
-        say "sudo is available. Caching credentials..."
+        if ! sudo -n true 2>/dev/null; then
-        if ! sudo -v; then
+            if ! [ -t 0 ]; then
-            die "Failed to cache sudo credentials"
+                die "sudo requires a password, but no TTY detected. Aborting to prevent hang."
+            fi
         fi
     fi
 
-    case "${INSTALL_DIR}${CONFIG_DIR}${WORK_DIR}" in
+    if [ -n "$SUDO" ]; then
-        *[!a-zA-Z0-9_./-]*)
+        if $SUDO sh -c '[ -d "$1" ]' _ "$CONFIG_FILE"; then
-            die "Invalid characters in path variables. Only alphanumeric, _, ., -, and / are allowed."
+            die "Safety check failed: CONFIG_FILE '$CONFIG_FILE' is a directory."
-            ;;
+        fi
-    esac
+    elif [ -d "$CONFIG_FILE" ]; then
-
+        die "Safety check failed: CONFIG_FILE '$CONFIG_FILE' is a directory."
-    case "$BIN_NAME" in
+    fi
-        *[!a-zA-Z0-9_-]*) die "Invalid characters in BIN_NAME: $BIN_NAME" ;;
-    esac
-
-    for path in "$CONFIG_DIR" "$WORK_DIR"; do
-        check_path="$path"
-
-        while [ "$check_path" != "/" ] && [ "${check_path%"/"}" != "$check_path" ]; do
-            check_path="${check_path%"/"}"
-        done
-        [ -z "$check_path" ] && check_path="/"
 
+    for path in "$CONFIG_DIR" "$CONFIG_PARENT_DIR" "$WORK_DIR"; do
+        check_path="$(get_realpath "$path")"
         case "$check_path" in
-            /|/bin|/sbin|/usr|/usr/bin|/usr/local|/etc|/opt|/var|/home|/root|/tmp)
+            /|/bin|/sbin|/usr|/usr/bin|/usr/sbin|/usr/local|/usr/local/bin|/usr/local/sbin|/usr/local/etc|/usr/local/share|/etc|/var|/var/lib|/var/log|/var/run|/home|/root|/tmp|/lib|/lib64|/opt|/run|/boot|/dev|/sys|/proc)
-                die "Safety check failed: '$path' is a critical system directory."
+                die "Safety check failed: '$path' (resolved to '$check_path') is a critical system directory." ;;
-                ;;
         esac
     done
 
-    for cmd in uname grep find rm chown chmod mv head mktemp; do
+    check_install_dir="$(get_realpath "$INSTALL_DIR")"
+    case "$check_install_dir" in
+        /|/etc|/var|/home|/root|/tmp|/usr|/usr/local|/opt|/boot|/dev|/sys|/proc|/run)
+            die "Safety check failed: INSTALL_DIR '$INSTALL_DIR' is a critical system directory." ;;
+    esac
+
+    for cmd in id uname grep find rm chown chmod mv mktemp mkdir tr dd sed ps head sleep cat tar gzip rmdir; do
         command -v "$cmd" >/dev/null 2>&1 || die "Required command not found: $cmd"
     done
 }
 
 verify_install_deps() {
-    if ! command -v curl >/dev/null 2>&1 && ! command -v wget >/dev/null 2>&1; then
+    command -v curl >/dev/null 2>&1 || command -v wget >/dev/null 2>&1 || die "Neither curl nor wget is installed."
-        die "Neither curl nor wget is installed."
-    fi
-    command -v tar >/dev/null 2>&1 || die "Required command not found: tar"
-    command -v gzip >/dev/null 2>&1 || die "Required command not found: gzip"
     command -v cp >/dev/null 2>&1 || command -v install >/dev/null 2>&1 || die "Need cp or install"
 
     if ! command -v setcap >/dev/null 2>&1; then
-        say "setcap is missing. Installing required capability tools..."
         if command -v apk >/dev/null 2>&1; then
-            $SUDO apk add --no-cache libcap || die "Failed to install libcap"
+            $SUDO apk add --no-cache libcap-utils >/dev/null 2>&1 || $SUDO apk add --no-cache libcap >/dev/null 2>&1 || true
         elif command -v apt-get >/dev/null 2>&1; then
-            $SUDO apt-get update -qq && $SUDO apt-get install -y -qq libcap2-bin || die "Failed to install libcap2-bin"
+            $SUDO apt-get update -q >/dev/null 2>&1 || true
-        elif command -v dnf >/dev/null 2>&1 || command -v yum >/dev/null 2>&1; then
+            $SUDO apt-get install -y -q libcap2-bin >/dev/null 2>&1 || true
-            $SUDO ${YUM_CMD:-yum} install -y -q libcap || die "Failed to install libcap"
+        elif command -v dnf >/dev/null 2>&1; then $SUDO dnf install -y -q libcap >/dev/null 2>&1 || true
-        else
+        elif command -v yum >/dev/null 2>&1; then $SUDO yum install -y -q libcap >/dev/null 2>&1 || true
-            die "Cannot install 'setcap'. Package manager not found. Please install libcap manually."
         fi
     fi
 }
@@ -163,122 +199,96 @@ detect_arch() {
 }
 
 detect_libc() {
-    if command -v ldd >/dev/null 2>&1 && ldd --version 2>&1 | grep -qi musl; then
-        echo "musl"; return 0
-    fi
-
-    if grep -q '^ID=alpine' /etc/os-release 2>/dev/null || grep -q '^ID="alpine"' /etc/os-release 2>/dev/null; then
-        echo "musl"; return 0
-    fi
     for f in /lib/ld-musl-*.so.* /lib64/ld-musl-*.so.*; do
-        if [ -e "$f" ]; then
+        if [ -e "$f" ]; then echo "musl"; return 0; fi
-            echo "musl"; return 0
-        fi
     done
+    if grep -qE '^ID="?alpine"?' /etc/os-release 2>/dev/null; then echo "musl"; return 0; fi
+    if command -v ldd >/dev/null 2>&1 && (ldd --version 2>&1 || true) | grep -qi musl; then echo "musl"; return 0; fi
     echo "gnu"
 }
 
 fetch_file() {
-    fetch_url="$1"
+    if command -v curl >/dev/null 2>&1; then curl -fsSL "$1" -o "$2"
-    fetch_out="$2"
+    else wget -q -O "$2" "$1"; fi
-
-    if command -v curl >/dev/null 2>&1; then
-        curl -fsSL "$fetch_url" -o "$fetch_out" || return 1
-    elif command -v wget >/dev/null 2>&1; then
-        wget -qO "$fetch_out" "$fetch_url" || return 1
-    else
-        die "curl or wget required"
-    fi
 }
 
 ensure_user_group() {
-    nologin_bin="/bin/false"
+    nologin_bin="$(command -v nologin 2>/dev/null || command -v false 2>/dev/null || echo /bin/false)"
 
-    cmd_nologin="$(command -v nologin 2>/dev/null || true)"
+    if ! check_os_entity group telemt; then
-    if [ -n "$cmd_nologin" ] && [ -x "$cmd_nologin" ]; then
+        if command -v groupadd >/dev/null 2>&1; then $SUDO groupadd -r telemt
-        nologin_bin="$cmd_nologin"
+        elif command -v addgroup >/dev/null 2>&1; then $SUDO addgroup -S telemt
-    else
+        else die "Cannot create group"; fi
-        for bin in /sbin/nologin /usr/sbin/nologin; do
-            if [ -x "$bin" ]; then
-                nologin_bin="$bin"
-                break
-            fi
-        done
     fi
 
-    if ! group_exists telemt; then
+    if ! check_os_entity passwd telemt; then
-        if command -v groupadd >/dev/null 2>&1; then
-            $SUDO groupadd -r telemt || die "Failed to create group via groupadd"
-        elif command -v addgroup >/dev/null 2>&1; then
-            $SUDO addgroup -S telemt || die "Failed to create group via addgroup"
-        else
-            die "Cannot create group: neither groupadd nor addgroup found"
-        fi
-    fi
-
-    if ! user_exists telemt; then
         if command -v useradd >/dev/null 2>&1; then
-            $SUDO useradd -r -g telemt -d "$WORK_DIR" -s "$nologin_bin" -c "Telemt Proxy" telemt || die "Failed to create user via useradd"
+            $SUDO useradd -r -g telemt -d "$WORK_DIR" -s "$nologin_bin" -c "Telemt Proxy" telemt
         elif command -v adduser >/dev/null 2>&1; then
-            $SUDO adduser -S -D -H -h "$WORK_DIR" -s "$nologin_bin" -G telemt telemt || die "Failed to create user via adduser"
+            if adduser --help 2>&1 | grep -q -- '-S'; then
-        else
+                $SUDO adduser -S -D -H -h "$WORK_DIR" -s "$nologin_bin" -G telemt telemt
-            die "Cannot create user: neither useradd nor adduser found"
+            else
-        fi
+                $SUDO adduser --system --home "$WORK_DIR" --shell "$nologin_bin" --no-create-home --ingroup telemt --disabled-password telemt
+            fi
+        else die "Cannot create user"; fi
     fi
 }
 
 setup_dirs() {
-    say "Setting up directories..."
+    $SUDO mkdir -p "$WORK_DIR" "$CONFIG_DIR" "$CONFIG_PARENT_DIR" || die "Failed to create directories"
-    $SUDO mkdir -p "$WORK_DIR" "$CONFIG_DIR" || die "Failed to create directories"
+    
-    $SUDO chown telemt:telemt "$WORK_DIR" || die "Failed to set owner on WORK_DIR"
+    $SUDO chown telemt:telemt "$WORK_DIR" && $SUDO chmod 750 "$WORK_DIR"
-    $SUDO chmod 750 "$WORK_DIR" || die "Failed to set permissions on WORK_DIR"
+    $SUDO chown root:telemt "$CONFIG_DIR" && $SUDO chmod 750 "$CONFIG_DIR"
+    
+    if [ "$CONFIG_PARENT_DIR" != "$CONFIG_DIR" ] && [ "$CONFIG_PARENT_DIR" != "." ] && [ "$CONFIG_PARENT_DIR" != "/" ]; then
+        $SUDO chown root:telemt "$CONFIG_PARENT_DIR" && $SUDO chmod 750 "$CONFIG_PARENT_DIR"
+    fi
 }
 
 stop_service() {
-    say "Stopping service if running..."
+    svc="$(get_svc_mgr)"
-    if command -v systemctl >/dev/null 2>&1 && [ -d /run/systemd/system ]; then
+    if [ "$svc" = "systemd" ] && systemctl is-active --quiet "$SERVICE_NAME" 2>/dev/null; then
         $SUDO systemctl stop "$SERVICE_NAME" 2>/dev/null || true
-    elif command -v rc-service >/dev/null 2>&1; then
+    elif [ "$svc" = "openrc" ] && rc-service "$SERVICE_NAME" status >/dev/null 2>&1; then
         $SUDO rc-service "$SERVICE_NAME" stop 2>/dev/null || true
     fi
 }
 
 install_binary() {
-    bin_src="$1"
+    bin_src="$1"; bin_dst="$2"
-    bin_dst="$2"
+    if [ -e "$INSTALL_DIR" ] && [ ! -d "$INSTALL_DIR" ]; then
+        die "'$INSTALL_DIR' is not a directory."
+    fi
 
     $SUDO mkdir -p "$INSTALL_DIR" || die "Failed to create install directory"
     if command -v install >/dev/null 2>&1; then
         $SUDO install -m 0755 "$bin_src" "$bin_dst" || die "Failed to install binary"
     else
-        $SUDO rm -f "$bin_dst"
+        $SUDO rm -f "$bin_dst" 2>/dev/null || true
-        $SUDO cp "$bin_src" "$bin_dst" || die "Failed to copy binary"
+        $SUDO cp "$bin_src" "$bin_dst" && $SUDO chmod 0755 "$bin_dst" || die "Failed to copy binary"
-        $SUDO chmod 0755 "$bin_dst" || die "Failed to set permissions"
     fi
 
-    if [ ! -x "$bin_dst" ]; then
+    $SUDO sh -c '[ -x "$1" ]' _ "$bin_dst" || die "Binary not executable: $bin_dst"
-        die "Failed to install binary or it is not executable: $bin_dst"
-    fi
 
-    say "Granting network bind capabilities to bind port 443..."
+    if command -v setcap >/dev/null 2>&1; then
-    if ! $SUDO setcap cap_net_bind_service=+ep "$bin_dst" 2>/dev/null; then
+        $SUDO setcap cap_net_bind_service=+ep "$bin_dst" 2>/dev/null || true
-        say "[WARNING] Failed to apply setcap. The service will NOT be able to open port 443!"
-        say "[WARNING] This usually happens inside unprivileged Docker/LXC containers."
     fi
 }
 
 generate_secret() {
-    if command -v openssl >/dev/null 2>&1; then
+    secret="$(command -v openssl >/dev/null 2>&1 && openssl rand -hex 16 2>/dev/null || true)"
-        secret="$(openssl rand -hex 16 2>/dev/null)" && [ -n "$secret" ] && { echo "$secret"; return 0; }
+    if [ -z "$secret" ] || [ "${#secret}" -ne 32 ]; then
+        if command -v od >/dev/null 2>&1; then secret="$(dd if=/dev/urandom bs=16 count=1 2>/dev/null | od -An -tx1 | tr -d ' \n')"
+        elif command -v hexdump >/dev/null 2>&1; then secret="$(dd if=/dev/urandom bs=16 count=1 2>/dev/null | hexdump -e '1/1 "%02x"')"
+        elif command -v xxd >/dev/null 2>&1; then secret="$(dd if=/dev/urandom bs=16 count=1 2>/dev/null | xxd -p | tr -d '\n')"
+        fi
     fi
-    if command -v xxd >/dev/null 2>&1; then
+    if [ "${#secret}" -eq 32 ]; then echo "$secret"; else return 1; fi
-        secret="$(dd if=/dev/urandom bs=1 count=16 2>/dev/null | xxd -p | tr -d '\n')" && [ -n "$secret" ] && { echo "$secret"; return 0; }
-    fi
-    secret="$(dd if=/dev/urandom bs=1 count=16 2>/dev/null | od -An -tx1 | tr -d ' \n')" && [ -n "$secret" ] && { echo "$secret"; return 0; }
-    return 1
 }
 
 generate_config_content() {
+    escaped_tls_domain="$(printf '%s\n' "$TLS_DOMAIN" | tr -d '[:cntrl:]' | sed 's/\\/\\\\/g; s/"/\\"/g')"
+
     cat <<EOF
 [general]
 use_middle_proxy = false
@@ -297,7 +307,7 @@ listen = "127.0.0.1:9091"
 whitelist = ["127.0.0.1/32"]
 
 [censorship]
-tls_domain = "petrovich.ru"
+tls_domain = "${escaped_tls_domain}"
 
 [access.users]
 hello = "$1"
@@ -305,44 +315,38 @@ EOF
 }
 
 install_config() {
-    config_exists=0
-
     if [ -n "$SUDO" ]; then
-        $SUDO sh -c "[ -f '$CONFIG_FILE' ]" 2>/dev/null && config_exists=1 || true
+        if $SUDO sh -c '[ -f "$1" ]' _ "$CONFIG_FILE"; then
-    else
+            say "  -> Config already exists at $CONFIG_FILE. Skipping creation."
-        [ -f "$CONFIG_FILE" ] && config_exists=1 || true
+            return 0
-    fi
+        fi
-
+    elif [ -f "$CONFIG_FILE" ]; then
-    if [ "$config_exists" -eq 1 ]; then
+        say "  -> Config already exists at $CONFIG_FILE. Skipping creation."
-        say "Config already exists, skipping generation."
         return 0
     fi
 
-    toml_secret="$(generate_secret)" || die "Failed to generate secret"
+    toml_secret="$(generate_secret)" || die "Failed to generate secret."
-    say "Creating config at $CONFIG_FILE..."
 
-    tmp_conf="$(mktemp "${TEMP_DIR:-/tmp}/telemt_conf.XXXXXX")" || die "Failed to create temp config"
+    generate_config_content "$toml_secret" | write_root "$CONFIG_FILE" || die "Failed to install config"
-    generate_config_content "$toml_secret" > "$tmp_conf" || die "Failed to write temp config"
+    $SUDO chown root:telemt "$CONFIG_FILE" && $SUDO chmod 640 "$CONFIG_FILE"
 
-    $SUDO mv "$tmp_conf" "$CONFIG_FILE" || die "Failed to install config file"
+    say "  -> Config created successfully."
-    $SUDO chown root:telemt "$CONFIG_FILE" || die "Failed to set owner"
+    say "  -> Generated secret for default user 'hello': $toml_secret"
-    $SUDO chmod 640 "$CONFIG_FILE" || die "Failed to set config permissions"
-
-    say "Secret for user 'hello': $toml_secret"
 }
 
 generate_systemd_content() {
     cat <<EOF
 [Unit]
-Description=Telemt Proxy Service
+Description=Telemt
 After=network-online.target
+Wants=network-online.target
 
 [Service]
 Type=simple
 User=telemt
 Group=telemt
 WorkingDirectory=$WORK_DIR
-ExecStart=${INSTALL_DIR}/${BIN_NAME} ${CONFIG_FILE}
+ExecStart="${INSTALL_DIR}/${BIN_NAME}" "${CONFIG_FILE}"
 Restart=on-failure
 LimitNOFILE=65536
 AmbientCapabilities=CAP_NET_BIND_SERVICE
@@ -370,156 +374,183 @@ EOF
 }
 
 install_service() {
-    if command -v systemctl >/dev/null 2>&1 && [ -d /run/systemd/system ]; then
+    svc="$(get_svc_mgr)"
-        say "Installing systemd service..."
+    if [ "$svc" = "systemd" ]; then
-        tmp_svc="$(mktemp "${TEMP_DIR:-/tmp}/${SERVICE_NAME}.service.XXXXXX")" || die "Failed to create temp service"
+        generate_systemd_content | write_root "/etc/systemd/system/${SERVICE_NAME}.service"
-        generate_systemd_content > "$tmp_svc" || die "Failed to generate service content"
+        $SUDO chown root:root "/etc/systemd/system/${SERVICE_NAME}.service" && $SUDO chmod 644 "/etc/systemd/system/${SERVICE_NAME}.service"
 
-        $SUDO mv "$tmp_svc" "/etc/systemd/system/${SERVICE_NAME}.service" || die "Failed to move service file"
+        $SUDO systemctl daemon-reload || true
-        $SUDO chown root:root "/etc/systemd/system/${SERVICE_NAME}.service"
+        $SUDO systemctl enable "$SERVICE_NAME" || true
-        $SUDO chmod 644 "/etc/systemd/system/${SERVICE_NAME}.service"
+        
+        if ! $SUDO systemctl start "$SERVICE_NAME"; then
+            say "[WARNING] Failed to start service"
+            SERVICE_START_FAILED=1
+        fi
+    elif [ "$svc" = "openrc" ]; then
+        generate_openrc_content | write_root "/etc/init.d/${SERVICE_NAME}"
+        $SUDO chown root:root "/etc/init.d/${SERVICE_NAME}" && $SUDO chmod 0755 "/etc/init.d/${SERVICE_NAME}"
 
-        $SUDO systemctl daemon-reload || die "Failed to reload systemd"
+        $SUDO rc-update add "$SERVICE_NAME" default 2>/dev/null || true
-        $SUDO systemctl enable "$SERVICE_NAME" || die "Failed to enable service"
+        
-        $SUDO systemctl start "$SERVICE_NAME" || die "Failed to start service"
+        if ! $SUDO rc-service "$SERVICE_NAME" start 2>/dev/null; then
-
+            say "[WARNING] Failed to start service"
-    elif command -v rc-update >/dev/null 2>&1; then
+            SERVICE_START_FAILED=1
-        say "Installing OpenRC service..."
+        fi
-        tmp_svc="$(mktemp "${TEMP_DIR:-/tmp}/${SERVICE_NAME}.init.XXXXXX")" || die "Failed to create temp file"
-        generate_openrc_content > "$tmp_svc" || die "Failed to generate init content"
-
-        $SUDO mv "$tmp_svc" "/etc/init.d/${SERVICE_NAME}" || die "Failed to move service file"
-        $SUDO chown root:root "/etc/init.d/${SERVICE_NAME}"
-        $SUDO chmod 0755 "/etc/init.d/${SERVICE_NAME}"
-
-        $SUDO rc-update add "$SERVICE_NAME" default 2>/dev/null || die "Failed to register service"
-        $SUDO rc-service "$SERVICE_NAME" start 2>/dev/null || die "Failed to start OpenRC service"
     else
-        say "No service manager found. You can start it manually with:"
+        cmd="\"${INSTALL_DIR}/${BIN_NAME}\" \"${CONFIG_FILE}\""
-        if [ -n "$SUDO" ]; then
+        if [ -n "$SUDO" ]; then 
-            say "  sudo -u telemt ${INSTALL_DIR}/${BIN_NAME} ${CONFIG_FILE}"
+            say "  -> Service manager not found. Start manually: sudo -u telemt $cmd"
-        else
+        else 
-            say "  su -s /bin/sh telemt -c '${INSTALL_DIR}/${BIN_NAME} ${CONFIG_FILE}'"
+            say "  -> Service manager not found. Start manually: su -s /bin/sh telemt -c '$cmd'"
         fi
     fi
 }
 
 kill_user_procs() {
-    say "Ensuring $BIN_NAME processes are killed..."
+    if command -v pkill >/dev/null 2>&1; then
-
+        $SUDO pkill -u telemt "$BIN_NAME" 2>/dev/null || true
-    if pkill_cmd="$(command -v pkill 2>/dev/null)"; then
-        $SUDO "$pkill_cmd" -u telemt "$BIN_NAME" 2>/dev/null || true
         sleep 1
-        $SUDO "$pkill_cmd" -9 -u telemt "$BIN_NAME" 2>/dev/null || true
+        $SUDO pkill -9 -u telemt "$BIN_NAME" 2>/dev/null || true
-    elif killall_cmd="$(command -v killall 2>/dev/null)"; then
+    else
-        $SUDO "$killall_cmd" "$BIN_NAME" 2>/dev/null || true
+        if command -v pgrep >/dev/null 2>&1; then
-        sleep 1
+            pids="$(pgrep -u telemt 2>/dev/null || true)"
-        $SUDO "$killall_cmd" -9 "$BIN_NAME" 2>/dev/null || true
+        else
+            pids="$(ps -u telemt -o pid= 2>/dev/null || true)"
+        fi
+        
+        if [ -n "$pids" ]; then
+            for pid in $pids; do
+                case "$pid" in ''|*[!0-9]*) continue ;; *) $SUDO kill "$pid" 2>/dev/null || true ;; esac
+            done
+            sleep 1
+            for pid in $pids; do
+                case "$pid" in ''|*[!0-9]*) continue ;; *) $SUDO kill -9 "$pid" 2>/dev/null || true ;; esac
+            done
+        fi
     fi
 }
 
 uninstall() {
-    purge_data=0
+    say "Starting uninstallation of $BIN_NAME..."
-    [ "$ACTION" = "purge" ] && purge_data=1
 
-    say "Uninstalling $BIN_NAME..."
+    say ">>> Stage 1: Stopping services"
     stop_service
 
-    if command -v systemctl >/dev/null 2>&1 && [ -d /run/systemd/system ]; then
+    say ">>> Stage 2: Removing service configuration"
+    svc="$(get_svc_mgr)"
+    if [ "$svc" = "systemd" ]; then
         $SUDO systemctl disable "$SERVICE_NAME" 2>/dev/null || true
         $SUDO rm -f "/etc/systemd/system/${SERVICE_NAME}.service"
-        $SUDO systemctl daemon-reload || true
+        $SUDO systemctl daemon-reload 2>/dev/null || true
-    elif command -v rc-update >/dev/null 2>&1; then
+    elif [ "$svc" = "openrc" ]; then
         $SUDO rc-update del "$SERVICE_NAME" 2>/dev/null || true
         $SUDO rm -f "/etc/init.d/${SERVICE_NAME}"
     fi
 
+    say ">>> Stage 3: Terminating user processes"
     kill_user_procs
 
+    say ">>> Stage 4: Removing binary"
     $SUDO rm -f "${INSTALL_DIR}/${BIN_NAME}"
 
-    $SUDO userdel telemt 2>/dev/null || $SUDO deluser telemt 2>/dev/null || true
+    if [ "$ACTION" = "purge" ]; then
-    $SUDO groupdel telemt 2>/dev/null || $SUDO delgroup telemt 2>/dev/null || true
+        say ">>> Stage 5: Purging configuration, data, and user"
-
-    if [ "$purge_data" -eq 1 ]; then
-        say "Purging configuration and data..."
         $SUDO rm -rf "$CONFIG_DIR" "$WORK_DIR"
+        $SUDO rm -f "$CONFIG_FILE"
+        if [ "$CONFIG_PARENT_DIR" != "$CONFIG_DIR" ] && [ "$CONFIG_PARENT_DIR" != "." ] && [ "$CONFIG_PARENT_DIR" != "/" ]; then
+            $SUDO rmdir "$CONFIG_PARENT_DIR" 2>/dev/null || true
+        fi
+        $SUDO userdel telemt 2>/dev/null || $SUDO deluser telemt 2>/dev/null || true
+        $SUDO groupdel telemt 2>/dev/null || $SUDO delgroup telemt 2>/dev/null || true
     else
-        say "Note: Configuration in $CONFIG_DIR was kept. Run with '--purge' to remove it."
+        say "Note: Configuration and user kept. Run with 'purge' to remove completely."
     fi
-
+    
-    say "Uninstallation complete."
+    printf '\n====================================================================\n'
+    printf '                    UNINSTALLATION COMPLETE\n'
+    printf '====================================================================\n\n'
     exit 0
 }
 
-# ============================================================================
-# Main Entry Point
-# ============================================================================
-
 case "$ACTION" in
-    help)
+    help) show_help ;;
-        show_help
+    uninstall|purge) verify_common; uninstall ;;
-        ;;
-    uninstall|purge)
-        verify_common
-        uninstall
-        ;;
     install)
-        say "Starting installation..."
+        say "Starting installation of $BIN_NAME (Version: $TARGET_VERSION)"
-        verify_common
-        verify_install_deps
 
-        ARCH="$(detect_arch)"
+        say ">>> Stage 1: Verifying environment and dependencies"
-        LIBC="$(detect_libc)"
+        verify_common; verify_install_deps
-        say "Detected system: $ARCH-linux-$LIBC"
 
+        if [ "$TARGET_VERSION" != "latest" ]; then 
+            TARGET_VERSION="${TARGET_VERSION#v}"
+        fi
+        
+        ARCH="$(detect_arch)"; LIBC="$(detect_libc)"
         FILE_NAME="${BIN_NAME}-${ARCH}-linux-${LIBC}.tar.gz"
-        FILE_NAME="$(printf '%s' "$FILE_NAME" | tr -d ' \t\n\r')"
+        
-
         if [ "$TARGET_VERSION" = "latest" ]; then
             DL_URL="https://github.com/${REPO}/releases/latest/download/${FILE_NAME}"
-        else
+        else 
             DL_URL="https://github.com/${REPO}/releases/download/${TARGET_VERSION}/${FILE_NAME}"
         fi
 
-        TEMP_DIR="$(mktemp -d)" || die "Failed to create temp directory"
+        say ">>> Stage 2: Downloading archive"
+        TEMP_DIR="$(mktemp -d)" || die "Temp directory creation failed"
         if [ -z "$TEMP_DIR" ] || [ ! -d "$TEMP_DIR" ]; then
-            die "Temp directory creation failed"
+            die "Temp directory is invalid or was not created"
         fi
 
-        say "Downloading from $DL_URL..."
+        fetch_file "$DL_URL" "${TEMP_DIR}/${FILE_NAME}" || die "Download failed"
-        fetch_file "$DL_URL" "${TEMP_DIR}/archive.tar.gz" || die "Download failed (check version or network)"
 
-        gzip -dc "${TEMP_DIR}/archive.tar.gz" | tar -xf - -C "$TEMP_DIR" || die "Extraction failed"
+        say ">>> Stage 3: Extracting archive"
+        if ! gzip -dc "${TEMP_DIR}/${FILE_NAME}" | tar -xf - -C "$TEMP_DIR" 2>/dev/null; then
+            die "Extraction failed (downloaded archive might be invalid or 404)."
+        fi
 
-        EXTRACTED_BIN="$(find "$TEMP_DIR" -type f -name "$BIN_NAME" -print 2>/dev/null | head -n 1)"
+        EXTRACTED_BIN="$(find "$TEMP_DIR" -type f -name "$BIN_NAME" -print 2>/dev/null | head -n 1 || true)"
-        [ -z "$EXTRACTED_BIN" ] && die "Binary '$BIN_NAME' not found in archive"
+        [ -n "$EXTRACTED_BIN" ] || die "Binary '$BIN_NAME' not found in archive"
 
-        ensure_user_group
+        say ">>> Stage 4: Setting up environment (User, Group, Directories)"
-        setup_dirs
+        ensure_user_group; setup_dirs; stop_service
-        stop_service
+        
-
+        say ">>> Stage 5: Installing binary"
-        say "Installing binary..."
         install_binary "$EXTRACTED_BIN" "${INSTALL_DIR}/${BIN_NAME}"
-
+        
+        say ">>> Stage 6: Generating configuration"
         install_config
+        
+        say ">>> Stage 7: Installing and starting service"
         install_service
 
-        say ""
+        if [ "${SERVICE_START_FAILED:-0}" -eq 1 ]; then
-        say "============================================="
+            printf '\n====================================================================\n'
-        say "Installation complete!"
+            printf '               INSTALLATION COMPLETED WITH WARNINGS\n'
-        say "============================================="
+            printf '====================================================================\n\n'
-        if command -v systemctl >/dev/null 2>&1 && [ -d /run/systemd/system ]; then
+            printf 'The service was installed but failed to start automatically.\n'
-            say "To check the logs, run:"
+            printf 'Please check the logs to determine the issue.\n\n'
-            say "  journalctl -u $SERVICE_NAME -f"
-            say ""
-        fi
-        say "To get user connection links, run:"
-        if command -v jq >/dev/null 2>&1; then
-            say "  curl -s http://127.0.0.1:9091/v1/users | jq -r '.data[] | \"User: \\(.username)\\n\\(.links.tls[0] // empty)\"'"
         else
-            say "  curl -s http://127.0.0.1:9091/v1/users"
+            printf '\n====================================================================\n'
-            say "  (Note: Install 'jq' package to see the links nicely formatted)"
+            printf '                      INSTALLATION SUCCESS\n'
+            printf '====================================================================\n\n'
         fi
+        
+        svc="$(get_svc_mgr)"
+        if [ "$svc" = "systemd" ]; then
+            printf 'To check the status of your proxy service, run:\n'
+            printf '  systemctl status %s\n\n' "$SERVICE_NAME"
+        elif [ "$svc" = "openrc" ]; then
+            printf 'To check the status of your proxy service, run:\n'
+            printf '  rc-service %s status\n\n' "$SERVICE_NAME"
+        fi
+        
+        printf 'To get your user connection links (for Telegram), run:\n'
+        if command -v jq >/dev/null 2>&1; then
+            printf '  curl -s http://127.0.0.1:9091/v1/users | jq -r '\''.data[] | "User: \\(.username)\\n\\(.links.tls[0] // empty)\\n"'\''\n'
+        else
+            printf '  curl -s http://127.0.0.1:9091/v1/users\n'
+            printf '  (Tip: Install '\''jq'\'' for a much cleaner output)\n'
+        fi
+        
+        printf '\n====================================================================\n'
         ;;
 esac

src/api/model.rs

@@ -206,6 +206,16 @@ pub(super) struct ZeroPoolData {
     pub(super) refill_failed_total: u64,
     pub(super) writer_restored_same_endpoint_total: u64,
     pub(super) writer_restored_fallback_total: u64,
+    pub(super) teardown_attempt_total_normal: u64,
+    pub(super) teardown_attempt_total_hard_detach: u64,
+    pub(super) teardown_success_total_normal: u64,
+    pub(super) teardown_success_total_hard_detach: u64,
+    pub(super) teardown_timeout_total: u64,
+    pub(super) teardown_escalation_total: u64,
+    pub(super) teardown_noop_total: u64,
+    pub(super) teardown_cleanup_side_effect_failures_total: u64,
+    pub(super) teardown_duration_count_total: u64,
+    pub(super) teardown_duration_sum_seconds_total: f64,
 }
 
 #[derive(Serialize, Clone)]
@@ -365,6 +375,7 @@ pub(super) struct MinimalMeRuntimeData {
     pub(super) me_reconnect_backoff_cap_ms: u64,
     pub(super) me_reconnect_fast_retry_count: u32,
     pub(super) me_pool_drain_ttl_secs: u64,
+    pub(super) me_instadrain: bool,
     pub(super) me_pool_drain_soft_evict_enabled: bool,
     pub(super) me_pool_drain_soft_evict_grace_secs: u64,
     pub(super) me_pool_drain_soft_evict_per_writer: u8,

src/api/runtime_min.rs

@@ -4,6 +4,9 @@ use std::time::{SystemTime, UNIX_EPOCH};
 use serde::Serialize;
 
 use crate::config::ProxyConfig;
+use crate::stats::{
+    MeWriterCleanupSideEffectStep, MeWriterTeardownMode, MeWriterTeardownReason, Stats,
+};
 
 use super::ApiShared;
 
@@ -98,6 +101,50 @@ pub(super) struct RuntimeMeQualityCountersData {
     pub(super) reconnect_success_total: u64,
 }
 
+#[derive(Serialize)]
+pub(super) struct RuntimeMeQualityTeardownAttemptData {
+    pub(super) reason: &'static str,
+    pub(super) mode: &'static str,
+    pub(super) total: u64,
+}
+
+#[derive(Serialize)]
+pub(super) struct RuntimeMeQualityTeardownSuccessData {
+    pub(super) mode: &'static str,
+    pub(super) total: u64,
+}
+
+#[derive(Serialize)]
+pub(super) struct RuntimeMeQualityTeardownSideEffectData {
+    pub(super) step: &'static str,
+    pub(super) total: u64,
+}
+
+#[derive(Serialize)]
+pub(super) struct RuntimeMeQualityTeardownDurationBucketData {
+    pub(super) le_seconds: &'static str,
+    pub(super) total: u64,
+}
+
+#[derive(Serialize)]
+pub(super) struct RuntimeMeQualityTeardownDurationData {
+    pub(super) mode: &'static str,
+    pub(super) count: u64,
+    pub(super) sum_seconds: f64,
+    pub(super) buckets: Vec<RuntimeMeQualityTeardownDurationBucketData>,
+}
+
+#[derive(Serialize)]
+pub(super) struct RuntimeMeQualityTeardownData {
+    pub(super) attempts: Vec<RuntimeMeQualityTeardownAttemptData>,
+    pub(super) success: Vec<RuntimeMeQualityTeardownSuccessData>,
+    pub(super) timeout_total: u64,
+    pub(super) escalation_total: u64,
+    pub(super) noop_total: u64,
+    pub(super) cleanup_side_effect_failures: Vec<RuntimeMeQualityTeardownSideEffectData>,
+    pub(super) duration: Vec<RuntimeMeQualityTeardownDurationData>,
+}
+
 #[derive(Serialize)]
 pub(super) struct RuntimeMeQualityRouteDropData {
     pub(super) no_conn_total: u64,
@@ -107,6 +154,25 @@ pub(super) struct RuntimeMeQualityRouteDropData {
     pub(super) queue_full_high_total: u64,
 }
 
+#[derive(Serialize)]
+pub(super) struct RuntimeMeQualityFamilyStateData {
+    pub(super) family: &'static str,
+    pub(super) state: &'static str,
+    pub(super) state_since_epoch_secs: u64,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub(super) suppressed_until_epoch_secs: Option<u64>,
+    pub(super) fail_streak: u32,
+    pub(super) recover_success_streak: u32,
+}
+
+#[derive(Serialize)]
+pub(super) struct RuntimeMeQualityDrainGateData {
+    pub(super) route_quorum_ok: bool,
+    pub(super) redundancy_ok: bool,
+    pub(super) block_reason: &'static str,
+    pub(super) updated_at_epoch_secs: u64,
+}
+
 #[derive(Serialize)]
 pub(super) struct RuntimeMeQualityDcRttData {
     pub(super) dc: i16,
@@ -120,7 +186,10 @@ pub(super) struct RuntimeMeQualityDcRttData {
 #[derive(Serialize)]
 pub(super) struct RuntimeMeQualityPayload {
     pub(super) counters: RuntimeMeQualityCountersData,
+    pub(super) teardown: RuntimeMeQualityTeardownData,
     pub(super) route_drops: RuntimeMeQualityRouteDropData,
+    pub(super) family_states: Vec<RuntimeMeQualityFamilyStateData>,
+    pub(super) drain_gate: RuntimeMeQualityDrainGateData,
     pub(super) dc_rtt: Vec<RuntimeMeQualityDcRttData>,
 }
 
@@ -362,6 +431,19 @@ pub(super) async fn build_runtime_me_quality_data(shared: &ApiShared) -> Runtime
     };
 
     let status = pool.api_status_snapshot().await;
+    let family_states = pool
+        .api_family_state_snapshot()
+        .into_iter()
+        .map(|entry| RuntimeMeQualityFamilyStateData {
+            family: entry.family,
+            state: entry.state,
+            state_since_epoch_secs: entry.state_since_epoch_secs,
+            suppressed_until_epoch_secs: entry.suppressed_until_epoch_secs,
+            fail_streak: entry.fail_streak,
+            recover_success_streak: entry.recover_success_streak,
+        })
+        .collect();
+    let drain_gate_snapshot = pool.api_drain_gate_snapshot();
     RuntimeMeQualityData {
         enabled: true,
         reason: None,
@@ -375,6 +457,7 @@ pub(super) async fn build_runtime_me_quality_data(shared: &ApiShared) -> Runtime
                 reconnect_attempt_total: shared.stats.get_me_reconnect_attempts(),
                 reconnect_success_total: shared.stats.get_me_reconnect_success(),
             },
+            teardown: build_runtime_me_teardown_data(shared),
             route_drops: RuntimeMeQualityRouteDropData {
                 no_conn_total: shared.stats.get_me_route_drop_no_conn(),
                 channel_closed_total: shared.stats.get_me_route_drop_channel_closed(),
@@ -382,6 +465,13 @@ pub(super) async fn build_runtime_me_quality_data(shared: &ApiShared) -> Runtime
                 queue_full_base_total: shared.stats.get_me_route_drop_queue_full_base(),
                 queue_full_high_total: shared.stats.get_me_route_drop_queue_full_high(),
             },
+            family_states,
+            drain_gate: RuntimeMeQualityDrainGateData {
+                route_quorum_ok: drain_gate_snapshot.route_quorum_ok,
+                redundancy_ok: drain_gate_snapshot.redundancy_ok,
+                block_reason: drain_gate_snapshot.block_reason,
+                updated_at_epoch_secs: drain_gate_snapshot.updated_at_epoch_secs,
+            },
             dc_rtt: status
                 .dcs
                 .into_iter()
@@ -398,6 +488,81 @@ pub(super) async fn build_runtime_me_quality_data(shared: &ApiShared) -> Runtime
     }
 }
 
+fn build_runtime_me_teardown_data(shared: &ApiShared) -> RuntimeMeQualityTeardownData {
+    let attempts = MeWriterTeardownReason::ALL
+        .iter()
+        .copied()
+        .flat_map(|reason| {
+            MeWriterTeardownMode::ALL
+                .iter()
+                .copied()
+                .map(move |mode| RuntimeMeQualityTeardownAttemptData {
+                    reason: reason.as_str(),
+                    mode: mode.as_str(),
+                    total: shared.stats.get_me_writer_teardown_attempt_total(reason, mode),
+                })
+        })
+        .collect();
+
+    let success = MeWriterTeardownMode::ALL
+        .iter()
+        .copied()
+        .map(|mode| RuntimeMeQualityTeardownSuccessData {
+            mode: mode.as_str(),
+            total: shared.stats.get_me_writer_teardown_success_total(mode),
+        })
+        .collect();
+
+    let cleanup_side_effect_failures = MeWriterCleanupSideEffectStep::ALL
+        .iter()
+        .copied()
+        .map(|step| RuntimeMeQualityTeardownSideEffectData {
+            step: step.as_str(),
+            total: shared
+                .stats
+                .get_me_writer_cleanup_side_effect_failures_total(step),
+        })
+        .collect();
+
+    let duration = MeWriterTeardownMode::ALL
+        .iter()
+        .copied()
+        .map(|mode| {
+            let count = shared.stats.get_me_writer_teardown_duration_count(mode);
+            let mut buckets: Vec<RuntimeMeQualityTeardownDurationBucketData> = Stats::me_writer_teardown_duration_bucket_labels()
+                .iter()
+                .enumerate()
+                .map(|(bucket_idx, label)| RuntimeMeQualityTeardownDurationBucketData {
+                    le_seconds: label,
+                    total: shared
+                        .stats
+                        .get_me_writer_teardown_duration_bucket_total(mode, bucket_idx),
+                })
+                .collect();
+            buckets.push(RuntimeMeQualityTeardownDurationBucketData {
+                le_seconds: "+Inf",
+                total: count,
+            });
+            RuntimeMeQualityTeardownDurationData {
+                mode: mode.as_str(),
+                count,
+                sum_seconds: shared.stats.get_me_writer_teardown_duration_sum_seconds(mode),
+                buckets,
+            }
+        })
+        .collect();
+
+    RuntimeMeQualityTeardownData {
+        attempts,
+        success,
+        timeout_total: shared.stats.get_me_writer_teardown_timeout_total(),
+        escalation_total: shared.stats.get_me_writer_teardown_escalation_total(),
+        noop_total: shared.stats.get_me_writer_teardown_noop_total(),
+        cleanup_side_effect_failures,
+        duration,
+    }
+}
+
 pub(super) async fn build_runtime_upstream_quality_data(
     shared: &ApiShared,
 ) -> RuntimeUpstreamQualityData {

src/api/runtime_stats.rs

@@ -1,9 +1,9 @@
 use std::time::{Duration, Instant, SystemTime, UNIX_EPOCH};
 
 use crate::config::ApiConfig;
-use crate::stats::Stats;
+use crate::stats::{MeWriterTeardownMode, Stats};
-use crate::transport::UpstreamRouteKind;
 use crate::transport::upstream::IpPreference;
+use crate::transport::UpstreamRouteKind;
 
 use super::ApiShared;
 use super::model::{
@@ -106,6 +106,29 @@ pub(super) fn build_zero_all_data(stats: &Stats, configured_users: usize) -> Zer
             refill_failed_total: stats.get_me_refill_failed_total(),
             writer_restored_same_endpoint_total: stats.get_me_writer_restored_same_endpoint_total(),
             writer_restored_fallback_total: stats.get_me_writer_restored_fallback_total(),
+            teardown_attempt_total_normal: stats
+                .get_me_writer_teardown_attempt_total_by_mode(MeWriterTeardownMode::Normal),
+            teardown_attempt_total_hard_detach: stats
+                .get_me_writer_teardown_attempt_total_by_mode(MeWriterTeardownMode::HardDetach),
+            teardown_success_total_normal: stats
+                .get_me_writer_teardown_success_total(MeWriterTeardownMode::Normal),
+            teardown_success_total_hard_detach: stats
+                .get_me_writer_teardown_success_total(MeWriterTeardownMode::HardDetach),
+            teardown_timeout_total: stats.get_me_writer_teardown_timeout_total(),
+            teardown_escalation_total: stats.get_me_writer_teardown_escalation_total(),
+            teardown_noop_total: stats.get_me_writer_teardown_noop_total(),
+            teardown_cleanup_side_effect_failures_total: stats
+                .get_me_writer_cleanup_side_effect_failures_total_all(),
+            teardown_duration_count_total: stats
+                .get_me_writer_teardown_duration_count(MeWriterTeardownMode::Normal)
+                .saturating_add(
+                    stats.get_me_writer_teardown_duration_count(MeWriterTeardownMode::HardDetach),
+                ),
+            teardown_duration_sum_seconds_total: stats
+                .get_me_writer_teardown_duration_sum_seconds(MeWriterTeardownMode::Normal)
+                + stats.get_me_writer_teardown_duration_sum_seconds(
+                    MeWriterTeardownMode::HardDetach,
+                ),
         },
         desync: ZeroDesyncData {
             secure_padding_invalid_total: stats.get_secure_padding_invalid(),
@@ -429,6 +452,7 @@ async fn get_minimal_payload_cached(
         me_reconnect_backoff_cap_ms: runtime.me_reconnect_backoff_cap_ms,
         me_reconnect_fast_retry_count: runtime.me_reconnect_fast_retry_count,
         me_pool_drain_ttl_secs: runtime.me_pool_drain_ttl_secs,
+        me_instadrain: runtime.me_instadrain,
         me_pool_drain_soft_evict_enabled: runtime.me_pool_drain_soft_evict_enabled,
         me_pool_drain_soft_evict_grace_secs: runtime.me_pool_drain_soft_evict_grace_secs,
         me_pool_drain_soft_evict_per_writer: runtime.me_pool_drain_soft_evict_per_writer,

src/cli.rs

@@ -198,8 +198,15 @@ desync_all_full = false
 update_every = 43200
 hardswap = false
 me_pool_drain_ttl_secs = 90
+me_instadrain = false
+me_pool_drain_threshold = 32
+me_pool_drain_soft_evict_grace_secs = 10
+me_pool_drain_soft_evict_per_writer = 2
+me_pool_drain_soft_evict_budget_per_core = 16
+me_pool_drain_soft_evict_cooldown_ms = 1000
+me_bind_stale_mode = "never"
 me_pool_min_fresh_ratio = 0.8
-me_reinit_drain_timeout_secs = 120
+me_reinit_drain_timeout_secs = 90
 
 [network]
 ipv4 = true
@@ -261,7 +268,7 @@ fn generate_systemd_unit(exe_path: &Path, config_path: &Path) -> String {
     format!(
 r#"[Unit]
 Description=Telemt MTProxy
-Documentation=https://github.com/nicepkg/telemt
+Documentation=https://github.com/telemt/telemt
 After=network-online.target
 Wants=network-online.target
 

src/config/defaults.rs

@@ -40,10 +40,10 @@ const DEFAULT_ME_ROUTE_HYBRID_MAX_WAIT_MS: u64 = 3000;
 const DEFAULT_ME_ROUTE_BLOCKING_SEND_TIMEOUT_MS: u64 = 250;
 const DEFAULT_ME_C2ME_SEND_TIMEOUT_MS: u64 = 4000;
 const DEFAULT_ME_POOL_DRAIN_SOFT_EVICT_ENABLED: bool = true;
-const DEFAULT_ME_POOL_DRAIN_SOFT_EVICT_GRACE_SECS: u64 = 30;
+const DEFAULT_ME_POOL_DRAIN_SOFT_EVICT_GRACE_SECS: u64 = 10;
-const DEFAULT_ME_POOL_DRAIN_SOFT_EVICT_PER_WRITER: u8 = 1;
+const DEFAULT_ME_POOL_DRAIN_SOFT_EVICT_PER_WRITER: u8 = 2;
-const DEFAULT_ME_POOL_DRAIN_SOFT_EVICT_BUDGET_PER_CORE: u16 = 8;
+const DEFAULT_ME_POOL_DRAIN_SOFT_EVICT_BUDGET_PER_CORE: u16 = 16;
-const DEFAULT_ME_POOL_DRAIN_SOFT_EVICT_COOLDOWN_MS: u64 = 5000;
+const DEFAULT_ME_POOL_DRAIN_SOFT_EVICT_COOLDOWN_MS: u64 = 1000;
 const DEFAULT_USER_MAX_UNIQUE_IPS_WINDOW_SECS: u64 = 30;
 const DEFAULT_ACCEPT_PERMIT_TIMEOUT_MS: u64 = 250;
 const DEFAULT_UPSTREAM_CONNECT_RETRY_ATTEMPTS: u32 = 2;
@@ -65,6 +65,10 @@ pub(crate) fn default_tls_domain() -> String {
     "petrovich.ru".to_string()
 }
 
+pub(crate) fn default_tls_fetch_scope() -> String {
+    String::new()
+}
+
 pub(crate) fn default_mask_port() -> u16 {
     443
 }
@@ -606,15 +610,19 @@ pub(crate) fn default_proxy_secret_len_max() -> usize {
 }
 
 pub(crate) fn default_me_reinit_drain_timeout_secs() -> u64 {
-    120
+    90
 }
 
 pub(crate) fn default_me_pool_drain_ttl_secs() -> u64 {
     90
 }
 
+pub(crate) fn default_me_instadrain() -> bool {
+    false
+}
+
 pub(crate) fn default_me_pool_drain_threshold() -> u64 {
-    128
+    32
 }
 
 pub(crate) fn default_me_pool_drain_soft_evict_enabled() -> bool {

src/config/hot_reload.rs

@@ -56,6 +56,7 @@ pub struct HotFields {
     pub me_reinit_coalesce_window_ms: u64,
     pub hardswap:                bool,
     pub me_pool_drain_ttl_secs:  u64,
+    pub me_instadrain: bool,
     pub me_pool_drain_threshold: u64,
     pub me_pool_drain_soft_evict_enabled: bool,
     pub me_pool_drain_soft_evict_grace_secs: u64,
@@ -143,6 +144,7 @@ impl HotFields {
             me_reinit_coalesce_window_ms: cfg.general.me_reinit_coalesce_window_ms,
             hardswap:                cfg.general.hardswap,
             me_pool_drain_ttl_secs:  cfg.general.me_pool_drain_ttl_secs,
+            me_instadrain: cfg.general.me_instadrain,
             me_pool_drain_threshold: cfg.general.me_pool_drain_threshold,
             me_pool_drain_soft_evict_enabled: cfg.general.me_pool_drain_soft_evict_enabled,
             me_pool_drain_soft_evict_grace_secs: cfg.general.me_pool_drain_soft_evict_grace_secs,
@@ -477,6 +479,7 @@ fn overlay_hot_fields(old: &ProxyConfig, new: &ProxyConfig) -> ProxyConfig {
     cfg.general.me_reinit_coalesce_window_ms = new.general.me_reinit_coalesce_window_ms;
     cfg.general.hardswap = new.general.hardswap;
     cfg.general.me_pool_drain_ttl_secs = new.general.me_pool_drain_ttl_secs;
+    cfg.general.me_instadrain = new.general.me_instadrain;
     cfg.general.me_pool_drain_threshold = new.general.me_pool_drain_threshold;
     cfg.general.me_pool_drain_soft_evict_enabled = new.general.me_pool_drain_soft_evict_enabled;
     cfg.general.me_pool_drain_soft_evict_grace_secs =
@@ -620,6 +623,7 @@ fn warn_non_hot_changes(old: &ProxyConfig, new: &ProxyConfig, non_hot_changed: b
     }
     if old.censorship.tls_domain != new.censorship.tls_domain
         || old.censorship.tls_domains != new.censorship.tls_domains
+        || old.censorship.tls_fetch_scope != new.censorship.tls_fetch_scope
         || old.censorship.mask != new.censorship.mask
         || old.censorship.mask_host != new.censorship.mask_host
         || old.censorship.mask_port != new.censorship.mask_port
@@ -869,6 +873,12 @@ fn log_changes(
             old_hot.me_pool_drain_ttl_secs, new_hot.me_pool_drain_ttl_secs,
         );
     }
+    if old_hot.me_instadrain != new_hot.me_instadrain {
+        info!(
+            "config reload: me_instadrain: {} → {}",
+            old_hot.me_instadrain, new_hot.me_instadrain,
+        );
+    }
 
     if old_hot.me_pool_drain_threshold != new_hot.me_pool_drain_threshold {
         info!(

src/config/load.rs

@@ -646,6 +646,11 @@ impl ProxyConfig {
                 "general.me_route_backpressure_base_timeout_ms must be > 0".to_string(),
             ));
         }
+        if config.general.me_route_backpressure_base_timeout_ms > 5000 {
+            return Err(ProxyError::Config(
+                "general.me_route_backpressure_base_timeout_ms must be within [1, 5000]".to_string(),
+            ));
+        }
 
         if config.general.me_route_backpressure_high_timeout_ms
             < config.general.me_route_backpressure_base_timeout_ms
@@ -654,6 +659,11 @@ impl ProxyConfig {
                 "general.me_route_backpressure_high_timeout_ms must be >= general.me_route_backpressure_base_timeout_ms".to_string(),
             ));
         }
+        if config.general.me_route_backpressure_high_timeout_ms > 5000 {
+            return Err(ProxyError::Config(
+                "general.me_route_backpressure_high_timeout_ms must be within [1, 5000]".to_string(),
+            ));
+        }
 
         if !(1..=100).contains(&config.general.me_route_backpressure_high_watermark_pct) {
             return Err(ProxyError::Config(
@@ -804,6 +814,9 @@ impl ProxyConfig {
             config.censorship.mask_host = Some(config.censorship.tls_domain.clone());
         }
 
+        // Normalize optional TLS fetch scope: whitespace-only values disable scoped routing.
+        config.censorship.tls_fetch_scope = config.censorship.tls_fetch_scope.trim().to_string();
+
         // Merge primary + extra TLS domains, deduplicate (primary always first).
         if !config.censorship.tls_domains.is_empty() {
             let mut all = Vec::with_capacity(1 + config.censorship.tls_domains.len());
@@ -1668,6 +1681,47 @@ mod tests {
         let _ = std::fs::remove_file(path_valid);
     }
 
+    #[test]
+    fn me_route_backpressure_base_timeout_ms_out_of_range_is_rejected() {
+        let toml = r#"
+            [general]
+            me_route_backpressure_base_timeout_ms = 5001
+
+            [censorship]
+            tls_domain = "example.com"
+
+            [access.users]
+            user = "00000000000000000000000000000000"
+        "#;
+        let dir = std::env::temp_dir();
+        let path = dir.join("telemt_me_route_backpressure_base_timeout_ms_out_of_range_test.toml");
+        std::fs::write(&path, toml).unwrap();
+        let err = ProxyConfig::load(&path).unwrap_err().to_string();
+        assert!(err.contains("general.me_route_backpressure_base_timeout_ms must be within [1, 5000]"));
+        let _ = std::fs::remove_file(path);
+    }
+
+    #[test]
+    fn me_route_backpressure_high_timeout_ms_out_of_range_is_rejected() {
+        let toml = r#"
+            [general]
+            me_route_backpressure_base_timeout_ms = 100
+            me_route_backpressure_high_timeout_ms = 5001
+
+            [censorship]
+            tls_domain = "example.com"
+
+            [access.users]
+            user = "00000000000000000000000000000000"
+        "#;
+        let dir = std::env::temp_dir();
+        let path = dir.join("telemt_me_route_backpressure_high_timeout_ms_out_of_range_test.toml");
+        std::fs::write(&path, toml).unwrap();
+        let err = ProxyConfig::load(&path).unwrap_err().to_string();
+        assert!(err.contains("general.me_route_backpressure_high_timeout_ms must be within [1, 5000]"));
+        let _ = std::fs::remove_file(path);
+    }
+
     #[test]
     fn me_route_no_writer_wait_ms_out_of_range_is_rejected() {
         let toml = r#"
@@ -2030,6 +2084,45 @@ mod tests {
         let _ = std::fs::remove_file(path);
     }
 
+    #[test]
+    fn force_close_default_matches_drain_ttl() {
+        let toml = r#"
+            [censorship]
+            tls_domain = "example.com"
+
+            [access.users]
+            user = "00000000000000000000000000000000"
+        "#;
+        let dir = std::env::temp_dir();
+        let path = dir.join("telemt_force_close_default_test.toml");
+        std::fs::write(&path, toml).unwrap();
+        let cfg = ProxyConfig::load(&path).unwrap();
+        assert_eq!(cfg.general.me_reinit_drain_timeout_secs, 90);
+        assert_eq!(cfg.general.effective_me_pool_force_close_secs(), 90);
+        let _ = std::fs::remove_file(path);
+    }
+
+    #[test]
+    fn force_close_zero_uses_runtime_safety_fallback() {
+        let toml = r#"
+            [general]
+            me_reinit_drain_timeout_secs = 0
+
+            [censorship]
+            tls_domain = "example.com"
+
+            [access.users]
+            user = "00000000000000000000000000000000"
+        "#;
+        let dir = std::env::temp_dir();
+        let path = dir.join("telemt_force_close_zero_fallback_test.toml");
+        std::fs::write(&path, toml).unwrap();
+        let cfg = ProxyConfig::load(&path).unwrap();
+        assert_eq!(cfg.general.me_reinit_drain_timeout_secs, 0);
+        assert_eq!(cfg.general.effective_me_pool_force_close_secs(), 300);
+        let _ = std::fs::remove_file(path);
+    }
+
     #[test]
     fn force_close_bumped_when_below_drain_ttl() {
         let toml = r#"
@@ -2051,6 +2144,59 @@ mod tests {
         let _ = std::fs::remove_file(path);
     }
 
+    #[test]
+    fn tls_fetch_scope_default_is_empty() {
+        let toml = r#"
+            [censorship]
+            tls_domain = "example.com"
+
+            [access.users]
+            user = "00000000000000000000000000000000"
+        "#;
+        let dir = std::env::temp_dir();
+        let path = dir.join("telemt_tls_fetch_scope_default_test.toml");
+        std::fs::write(&path, toml).unwrap();
+        let cfg = ProxyConfig::load(&path).unwrap();
+        assert!(cfg.censorship.tls_fetch_scope.is_empty());
+        let _ = std::fs::remove_file(path);
+    }
+
+    #[test]
+    fn tls_fetch_scope_is_trimmed_during_load() {
+        let toml = r#"
+            [censorship]
+            tls_domain = "example.com"
+            tls_fetch_scope = "  me  "
+
+            [access.users]
+            user = "00000000000000000000000000000000"
+        "#;
+        let dir = std::env::temp_dir();
+        let path = dir.join("telemt_tls_fetch_scope_trim_test.toml");
+        std::fs::write(&path, toml).unwrap();
+        let cfg = ProxyConfig::load(&path).unwrap();
+        assert_eq!(cfg.censorship.tls_fetch_scope, "me");
+        let _ = std::fs::remove_file(path);
+    }
+
+    #[test]
+    fn tls_fetch_scope_whitespace_becomes_empty() {
+        let toml = r#"
+            [censorship]
+            tls_domain = "example.com"
+            tls_fetch_scope = "   "
+
+            [access.users]
+            user = "00000000000000000000000000000000"
+        "#;
+        let dir = std::env::temp_dir();
+        let path = dir.join("telemt_tls_fetch_scope_blank_test.toml");
+        std::fs::write(&path, toml).unwrap();
+        let cfg = ProxyConfig::load(&path).unwrap();
+        assert!(cfg.censorship.tls_fetch_scope.is_empty());
+        let _ = std::fs::remove_file(path);
+    }
+
     #[test]
     fn invalid_ad_tag_is_disabled_during_load() {
         let toml = r#"

src/config/types.rs

@@ -135,8 +135,8 @@ impl MeSocksKdfPolicy {
 #[derive(Debug, Clone, Copy, PartialEq, Eq, Serialize, Deserialize, Default)]
 #[serde(rename_all = "lowercase")]
 pub enum MeBindStaleMode {
-    Never,
     #[default]
+    Never,
     Ttl,
     Always,
 }
@@ -812,6 +812,10 @@ pub struct GeneralConfig {
     #[serde(default = "default_me_pool_drain_ttl_secs")]
     pub me_pool_drain_ttl_secs: u64,
 
+    /// Force-remove any draining writer on the next cleanup tick, regardless of age/deadline.
+    #[serde(default = "default_me_instadrain")]
+    pub me_instadrain: bool,
+
     /// Maximum allowed number of draining ME writers before oldest ones are force-closed in batches.
     /// Set to 0 to disable threshold-based draining cleanup and keep timeout-only behavior.
     #[serde(default = "default_me_pool_drain_threshold")]
@@ -851,7 +855,7 @@ pub struct GeneralConfig {
     pub me_pool_min_fresh_ratio: f32,
 
     /// Drain timeout in seconds for stale ME writers after endpoint map changes.
-    /// Set to 0 to keep stale writers draining indefinitely (no force-close).
+    /// Set to 0 to use the runtime safety fallback timeout.
     #[serde(default = "default_me_reinit_drain_timeout_secs")]
     pub me_reinit_drain_timeout_secs: u64,
 
@@ -1036,6 +1040,7 @@ impl Default for GeneralConfig {
             me_secret_atomic_snapshot: default_me_secret_atomic_snapshot(),
             proxy_secret_len_max: default_proxy_secret_len_max(),
             me_pool_drain_ttl_secs: default_me_pool_drain_ttl_secs(),
+            me_instadrain: default_me_instadrain(),
             me_pool_drain_threshold: default_me_pool_drain_threshold(),
             me_pool_drain_soft_evict_enabled: default_me_pool_drain_soft_evict_enabled(),
             me_pool_drain_soft_evict_grace_secs: default_me_pool_drain_soft_evict_grace_secs(),
@@ -1081,8 +1086,13 @@ impl GeneralConfig {
 
     /// Resolve force-close timeout for stale writers.
     /// `me_reinit_drain_timeout_secs` remains backward-compatible alias.
+    /// A configured `0` uses the runtime safety fallback (300s).
     pub fn effective_me_pool_force_close_secs(&self) -> u64 {
-        self.me_reinit_drain_timeout_secs
+        if self.me_reinit_drain_timeout_secs == 0 {
+            300
+        } else {
+            self.me_reinit_drain_timeout_secs
+        }
     }
 }
 
@@ -1316,6 +1326,11 @@ pub struct AntiCensorshipConfig {
     #[serde(default)]
     pub tls_domains: Vec<String>,
 
+    /// Upstream scope used for TLS front metadata fetches.
+    /// Empty value keeps default upstream routing behavior.
+    #[serde(default = "default_tls_fetch_scope")]
+    pub tls_fetch_scope: String,
+
     #[serde(default = "default_true")]
     pub mask: bool,
 
@@ -1373,6 +1388,7 @@ impl Default for AntiCensorshipConfig {
         Self {
             tls_domain: default_tls_domain(),
             tls_domains: Vec::new(),
+            tls_fetch_scope: default_tls_fetch_scope(),
             mask: default_true(),
             mask_host: None,
             mask_port: default_mask_port(),

src/maestro/me_startup.rs

@@ -237,6 +237,7 @@ pub(crate) async fn initialize_me_pool(
                     config.general.me_adaptive_floor_max_warm_writers_global,
                     config.general.hardswap,
                     config.general.me_pool_drain_ttl_secs,
+                    config.general.me_instadrain,
                     config.general.me_pool_drain_threshold,
                     config.general.me_pool_drain_soft_evict_enabled,
                     config.general.me_pool_drain_soft_evict_grace_secs,
@@ -331,18 +332,76 @@ pub(crate) async fn initialize_me_pool(
                                             "Middle-End pool initialized successfully"
                                         );
 
-                                        let pool_health = pool_bg.clone();
+                                            // ── Supervised background tasks ──────────────────
-                                        let rng_health = rng_bg.clone();
+                                            // Each task runs inside a nested tokio::spawn so
-                                        let min_conns = pool_size;
+                                            // that a panic is caught via JoinHandle and the
-                                        tokio::spawn(async move {
+                                            // outer loop restarts the task automatically.
-                                            crate::transport::middle_proxy::me_health_monitor(
+                                            let pool_health = pool_bg.clone();
-                                                pool_health,
+                                            let rng_health = rng_bg.clone();
-                                                rng_health,
+                                            let min_conns = pool_size;
-                                                min_conns,
+                                            tokio::spawn(async move {
-                                            )
+                                                loop {
-                                            .await;
+                                                    let p = pool_health.clone();
-                                        });
+                                                    let r = rng_health.clone();
-                                        break;
+                                                    let res = tokio::spawn(async move {
+                                                        crate::transport::middle_proxy::me_health_monitor(
+                                                            p, r, min_conns,
+                                                        )
+                                                        .await;
+                                                    })
+                                                    .await;
+                                                    match res {
+                                                        Ok(()) => warn!("me_health_monitor exited unexpectedly, restarting"),
+                                                        Err(e) => {
+                                                            error!(error = %e, "me_health_monitor panicked, restarting in 1s");
+                                                            tokio::time::sleep(Duration::from_secs(1)).await;
+                                                        }
+                                                    }
+                                                }
+                                            });
+                                            let pool_drain_enforcer = pool_bg.clone();
+                                            tokio::spawn(async move {
+                                                loop {
+                                                    let p = pool_drain_enforcer.clone();
+                                                    let res = tokio::spawn(async move {
+                                                        crate::transport::middle_proxy::me_drain_timeout_enforcer(p).await;
+                                                    })
+                                                    .await;
+                                                    match res {
+                                                        Ok(()) => warn!("me_drain_timeout_enforcer exited unexpectedly, restarting"),
+                                                        Err(e) => {
+                                                            error!(error = %e, "me_drain_timeout_enforcer panicked, restarting in 1s");
+                                                            tokio::time::sleep(Duration::from_secs(1)).await;
+                                                        }
+                                                    }
+                                                }
+                                            });
+                                            let pool_watchdog = pool_bg.clone();
+                                            tokio::spawn(async move {
+                                                loop {
+                                                    let p = pool_watchdog.clone();
+                                                    let res = tokio::spawn(async move {
+                                                        crate::transport::middle_proxy::me_zombie_writer_watchdog(p).await;
+                                                    })
+                                                    .await;
+                                                    match res {
+                                                        Ok(()) => warn!("me_zombie_writer_watchdog exited unexpectedly, restarting"),
+                                                        Err(e) => {
+                                                            error!(error = %e, "me_zombie_writer_watchdog panicked, restarting in 1s");
+                                                            tokio::time::sleep(Duration::from_secs(1)).await;
+                                                        }
+                                                    }
+                                                }
+                                            });
+                                            // CRITICAL: keep the current-thread runtime
+                                            // alive. Without this, block_on() returns,
+                                            // the Runtime is dropped, and ALL spawned
+                                            // background tasks (health monitor, drain
+                                            // enforcer, zombie watchdog) are silently
+                                            // cancelled — causing the draining-writer
+                                            // leak that brought us here.
+                                            std::future::pending::<()>().await;
+                                            unreachable!();
                                     }
                                     Err(e) => {
                                         startup_tracker_bg.set_me_last_error(Some(e.to_string())).await;
@@ -400,16 +459,65 @@ pub(crate) async fn initialize_me_pool(
                                     "Middle-End pool initialized successfully"
                                 );
 
-                                let pool_clone = pool.clone();
+                                    // ── Supervised background tasks ──────────────────
-                                let rng_clone = rng.clone();
+                                    let pool_clone = pool.clone();
-                                let min_conns = pool_size;
+                                    let rng_clone = rng.clone();
-                                tokio::spawn(async move {
+                                    let min_conns = pool_size;
-                                    crate::transport::middle_proxy::me_health_monitor(
+                                    tokio::spawn(async move {
-                                        pool_clone, rng_clone, min_conns,
+                                        loop {
-                                    )
+                                            let p = pool_clone.clone();
-                                    .await;
+                                            let r = rng_clone.clone();
-                                });
+                                            let res = tokio::spawn(async move {
-
+                                                crate::transport::middle_proxy::me_health_monitor(
+                                                    p, r, min_conns,
+                                                )
+                                                .await;
+                                            })
+                                            .await;
+                                            match res {
+                                                Ok(()) => warn!("me_health_monitor exited unexpectedly, restarting"),
+                                                Err(e) => {
+                                                    error!(error = %e, "me_health_monitor panicked, restarting in 1s");
+                                                    tokio::time::sleep(Duration::from_secs(1)).await;
+                                                }
+                                            }
+                                        }
+                                    });
+                                    let pool_drain_enforcer = pool.clone();
+                                    tokio::spawn(async move {
+                                        loop {
+                                            let p = pool_drain_enforcer.clone();
+                                            let res = tokio::spawn(async move {
+                                                crate::transport::middle_proxy::me_drain_timeout_enforcer(p).await;
+                                            })
+                                            .await;
+                                            match res {
+                                                Ok(()) => warn!("me_drain_timeout_enforcer exited unexpectedly, restarting"),
+                                                Err(e) => {
+                                                    error!(error = %e, "me_drain_timeout_enforcer panicked, restarting in 1s");
+                                                    tokio::time::sleep(Duration::from_secs(1)).await;
+                                                }
+                                            }
+                                        }
+                                    });
+                                    let pool_watchdog = pool.clone();
+                                    tokio::spawn(async move {
+                                        loop {
+                                            let p = pool_watchdog.clone();
+                                            let res = tokio::spawn(async move {
+                                                crate::transport::middle_proxy::me_zombie_writer_watchdog(p).await;
+                                            })
+                                            .await;
+                                            match res {
+                                                Ok(()) => warn!("me_zombie_writer_watchdog exited unexpectedly, restarting"),
+                                                Err(e) => {
+                                                    error!(error = %e, "me_zombie_writer_watchdog panicked, restarting in 1s");
+                                                    tokio::time::sleep(Duration::from_secs(1)).await;
+                                                }
+                                            }
+                                        }
+                                    });
+    
                                 break Some(pool);
                             }
                             Err(e) => {

src/maestro/tls_bootstrap.rs

@@ -38,12 +38,15 @@ pub(crate) async fn bootstrap_tls_front(
             .clone()
             .unwrap_or_else(|| config.censorship.tls_domain.clone());
         let mask_unix_sock = config.censorship.mask_unix_sock.clone();
+        let tls_fetch_scope = (!config.censorship.tls_fetch_scope.is_empty())
+            .then(|| config.censorship.tls_fetch_scope.clone());
         let fetch_timeout = Duration::from_secs(5);
 
         let cache_initial = cache.clone();
         let domains_initial = tls_domains.to_vec();
         let host_initial = mask_host.clone();
         let unix_sock_initial = mask_unix_sock.clone();
+        let scope_initial = tls_fetch_scope.clone();
         let upstream_initial = upstream_manager.clone();
         tokio::spawn(async move {
             let mut join = tokio::task::JoinSet::new();
@@ -51,6 +54,7 @@ pub(crate) async fn bootstrap_tls_front(
                 let cache_domain = cache_initial.clone();
                 let host_domain = host_initial.clone();
                 let unix_sock_domain = unix_sock_initial.clone();
+                let scope_domain = scope_initial.clone();
                 let upstream_domain = upstream_initial.clone();
                 join.spawn(async move {
                     match crate::tls_front::fetcher::fetch_real_tls(
@@ -59,6 +63,7 @@ pub(crate) async fn bootstrap_tls_front(
                         &domain,
                         fetch_timeout,
                         Some(upstream_domain),
+                        scope_domain.as_deref(),
                         proxy_protocol,
                         unix_sock_domain.as_deref(),
                     )
@@ -100,6 +105,7 @@ pub(crate) async fn bootstrap_tls_front(
         let domains_refresh = tls_domains.to_vec();
         let host_refresh = mask_host.clone();
         let unix_sock_refresh = mask_unix_sock.clone();
+        let scope_refresh = tls_fetch_scope.clone();
         let upstream_refresh = upstream_manager.clone();
         tokio::spawn(async move {
             loop {
@@ -112,6 +118,7 @@ pub(crate) async fn bootstrap_tls_front(
                     let cache_domain = cache_refresh.clone();
                     let host_domain = host_refresh.clone();
                     let unix_sock_domain = unix_sock_refresh.clone();
+                    let scope_domain = scope_refresh.clone();
                     let upstream_domain = upstream_refresh.clone();
                     join.spawn(async move {
                         match crate::tls_front::fetcher::fetch_real_tls(
@@ -120,6 +127,7 @@ pub(crate) async fn bootstrap_tls_front(
                             &domain,
                             fetch_timeout,
                             Some(upstream_domain),
+                            scope_domain.as_deref(),
                             proxy_protocol,
                             unix_sock_domain.as_deref(),
                         )

src/metrics.rs

@@ -16,7 +16,9 @@ use tracing::{info, warn, debug};
 use crate::config::ProxyConfig;
 use crate::ip_tracker::UserIpTracker;
 use crate::stats::beobachten::BeobachtenStore;
-use crate::stats::Stats;
+use crate::stats::{
+    MeWriterCleanupSideEffectStep, MeWriterTeardownMode, MeWriterTeardownReason, Stats,
+};
 use crate::transport::{ListenOptions, create_listener};
 
 pub async fn serve(
@@ -1692,6 +1694,57 @@ async fn render_metrics(stats: &Stats, config: &ProxyConfig, ip_tracker: &UserIp
         }
     );
 
+    let _ = writeln!(
+        out,
+        "# HELP telemt_me_writer_close_signal_drop_total Close-signal drops for already-removed ME writers"
+    );
+    let _ = writeln!(out, "# TYPE telemt_me_writer_close_signal_drop_total counter");
+    let _ = writeln!(
+        out,
+        "telemt_me_writer_close_signal_drop_total {}",
+        if me_allows_normal {
+            stats.get_me_writer_close_signal_drop_total()
+        } else {
+            0
+        }
+    );
+
+    let _ = writeln!(
+        out,
+        "# HELP telemt_me_writer_close_signal_channel_full_total Close-signal drops caused by full writer command channels"
+    );
+    let _ = writeln!(
+        out,
+        "# TYPE telemt_me_writer_close_signal_channel_full_total counter"
+    );
+    let _ = writeln!(
+        out,
+        "telemt_me_writer_close_signal_channel_full_total {}",
+        if me_allows_normal {
+            stats.get_me_writer_close_signal_channel_full_total()
+        } else {
+            0
+        }
+    );
+
+    let _ = writeln!(
+        out,
+        "# HELP telemt_me_draining_writers_reap_progress_total Draining-writer removals processed by reap cleanup"
+    );
+    let _ = writeln!(
+        out,
+        "# TYPE telemt_me_draining_writers_reap_progress_total counter"
+    );
+    let _ = writeln!(
+        out,
+        "telemt_me_draining_writers_reap_progress_total {}",
+        if me_allows_normal {
+            stats.get_me_draining_writers_reap_progress_total()
+        } else {
+            0
+        }
+    );
+
     let _ = writeln!(out, "# HELP telemt_me_writer_removed_total Total ME writer removals");
     let _ = writeln!(out, "# TYPE telemt_me_writer_removed_total counter");
     let _ = writeln!(
@@ -1719,6 +1772,169 @@ async fn render_metrics(stats: &Stats, config: &ProxyConfig, ip_tracker: &UserIp
         }
     );
 
+    let _ = writeln!(
+        out,
+        "# HELP telemt_me_writer_teardown_attempt_total ME writer teardown attempts by reason and mode"
+    );
+    let _ = writeln!(out, "# TYPE telemt_me_writer_teardown_attempt_total counter");
+    for reason in MeWriterTeardownReason::ALL {
+        for mode in MeWriterTeardownMode::ALL {
+            let _ = writeln!(
+                out,
+                "telemt_me_writer_teardown_attempt_total{{reason=\"{}\",mode=\"{}\"}} {}",
+                reason.as_str(),
+                mode.as_str(),
+                if me_allows_normal {
+                    stats.get_me_writer_teardown_attempt_total(reason, mode)
+                } else {
+                    0
+                }
+            );
+        }
+    }
+
+    let _ = writeln!(
+        out,
+        "# HELP telemt_me_writer_teardown_success_total ME writer teardown successes by mode"
+    );
+    let _ = writeln!(out, "# TYPE telemt_me_writer_teardown_success_total counter");
+    for mode in MeWriterTeardownMode::ALL {
+        let _ = writeln!(
+            out,
+            "telemt_me_writer_teardown_success_total{{mode=\"{}\"}} {}",
+            mode.as_str(),
+            if me_allows_normal {
+                stats.get_me_writer_teardown_success_total(mode)
+            } else {
+                0
+            }
+        );
+    }
+
+    let _ = writeln!(
+        out,
+        "# HELP telemt_me_writer_teardown_timeout_total Teardown operations that timed out"
+    );
+    let _ = writeln!(out, "# TYPE telemt_me_writer_teardown_timeout_total counter");
+    let _ = writeln!(
+        out,
+        "telemt_me_writer_teardown_timeout_total {}",
+        if me_allows_normal {
+            stats.get_me_writer_teardown_timeout_total()
+        } else {
+            0
+        }
+    );
+
+    let _ = writeln!(
+        out,
+        "# HELP telemt_me_writer_teardown_escalation_total Watchdog teardown escalations to hard detach"
+    );
+    let _ = writeln!(
+        out,
+        "# TYPE telemt_me_writer_teardown_escalation_total counter"
+    );
+    let _ = writeln!(
+        out,
+        "telemt_me_writer_teardown_escalation_total {}",
+        if me_allows_normal {
+            stats.get_me_writer_teardown_escalation_total()
+        } else {
+            0
+        }
+    );
+
+    let _ = writeln!(
+        out,
+        "# HELP telemt_me_writer_teardown_noop_total Teardown operations that became no-op"
+    );
+    let _ = writeln!(out, "# TYPE telemt_me_writer_teardown_noop_total counter");
+    let _ = writeln!(
+        out,
+        "telemt_me_writer_teardown_noop_total {}",
+        if me_allows_normal {
+            stats.get_me_writer_teardown_noop_total()
+        } else {
+            0
+        }
+    );
+
+    let _ = writeln!(
+        out,
+        "# HELP telemt_me_writer_teardown_duration_seconds ME writer teardown latency histogram by mode"
+    );
+    let _ = writeln!(
+        out,
+        "# TYPE telemt_me_writer_teardown_duration_seconds histogram"
+    );
+    let bucket_labels = Stats::me_writer_teardown_duration_bucket_labels();
+    for mode in MeWriterTeardownMode::ALL {
+        for (bucket_idx, label) in bucket_labels.iter().enumerate() {
+            let _ = writeln!(
+                out,
+                "telemt_me_writer_teardown_duration_seconds_bucket{{mode=\"{}\",le=\"{}\"}} {}",
+                mode.as_str(),
+                label,
+                if me_allows_normal {
+                    stats.get_me_writer_teardown_duration_bucket_total(mode, bucket_idx)
+                } else {
+                    0
+                }
+            );
+        }
+        let _ = writeln!(
+            out,
+            "telemt_me_writer_teardown_duration_seconds_bucket{{mode=\"{}\",le=\"+Inf\"}} {}",
+            mode.as_str(),
+            if me_allows_normal {
+                stats.get_me_writer_teardown_duration_count(mode)
+            } else {
+                0
+            }
+        );
+        let _ = writeln!(
+            out,
+            "telemt_me_writer_teardown_duration_seconds_sum{{mode=\"{}\"}} {:.6}",
+            mode.as_str(),
+            if me_allows_normal {
+                stats.get_me_writer_teardown_duration_sum_seconds(mode)
+            } else {
+                0.0
+            }
+        );
+        let _ = writeln!(
+            out,
+            "telemt_me_writer_teardown_duration_seconds_count{{mode=\"{}\"}} {}",
+            mode.as_str(),
+            if me_allows_normal {
+                stats.get_me_writer_teardown_duration_count(mode)
+            } else {
+                0
+            }
+        );
+    }
+
+    let _ = writeln!(
+        out,
+        "# HELP telemt_me_writer_cleanup_side_effect_failures_total Failed cleanup side effects by step"
+    );
+    let _ = writeln!(
+        out,
+        "# TYPE telemt_me_writer_cleanup_side_effect_failures_total counter"
+    );
+    for step in MeWriterCleanupSideEffectStep::ALL {
+        let _ = writeln!(
+            out,
+            "telemt_me_writer_cleanup_side_effect_failures_total{{step=\"{}\"}} {}",
+            step.as_str(),
+            if me_allows_normal {
+                stats.get_me_writer_cleanup_side_effect_failures_total(step)
+            } else {
+                0
+            }
+        );
+    }
+
     let _ = writeln!(out, "# HELP telemt_me_refill_triggered_total Immediate ME refill runs started");
     let _ = writeln!(out, "# TYPE telemt_me_refill_triggered_total counter");
     let _ = writeln!(
@@ -2124,6 +2340,24 @@ mod tests {
         assert!(output.contains("# TYPE telemt_me_rpc_proxy_req_signal_sent_total counter"));
         assert!(output.contains("# TYPE telemt_me_idle_close_by_peer_total counter"));
         assert!(output.contains("# TYPE telemt_me_writer_removed_total counter"));
+        assert!(output.contains("# TYPE telemt_me_writer_teardown_attempt_total counter"));
+        assert!(output.contains("# TYPE telemt_me_writer_teardown_success_total counter"));
+        assert!(output.contains("# TYPE telemt_me_writer_teardown_timeout_total counter"));
+        assert!(output.contains("# TYPE telemt_me_writer_teardown_escalation_total counter"));
+        assert!(output.contains("# TYPE telemt_me_writer_teardown_noop_total counter"));
+        assert!(output.contains(
+            "# TYPE telemt_me_writer_teardown_duration_seconds histogram"
+        ));
+        assert!(output.contains(
+            "# TYPE telemt_me_writer_cleanup_side_effect_failures_total counter"
+        ));
+        assert!(output.contains("# TYPE telemt_me_writer_close_signal_drop_total counter"));
+        assert!(output.contains(
+            "# TYPE telemt_me_writer_close_signal_channel_full_total counter"
+        ));
+        assert!(output.contains(
+            "# TYPE telemt_me_draining_writers_reap_progress_total counter"
+        ));
         assert!(output.contains("# TYPE telemt_pool_drain_soft_evict_total counter"));
         assert!(output.contains("# TYPE telemt_pool_drain_soft_evict_writer_total counter"));
         assert!(output.contains(

src/stats/mod.rs

@@ -19,6 +19,137 @@ use tracing::debug;
 use crate::config::{MeTelemetryLevel, MeWriterPickMode};
 use self::telemetry::TelemetryPolicy;
 
+const ME_WRITER_TEARDOWN_MODE_COUNT: usize = 2;
+const ME_WRITER_TEARDOWN_REASON_COUNT: usize = 11;
+const ME_WRITER_CLEANUP_SIDE_EFFECT_STEP_COUNT: usize = 2;
+const ME_WRITER_TEARDOWN_DURATION_BUCKET_COUNT: usize = 12;
+const ME_WRITER_TEARDOWN_DURATION_BUCKET_BOUNDS_MICROS: [u64; ME_WRITER_TEARDOWN_DURATION_BUCKET_COUNT] = [
+    1_000,
+    5_000,
+    10_000,
+    25_000,
+    50_000,
+    100_000,
+    250_000,
+    500_000,
+    1_000_000,
+    2_500_000,
+    5_000_000,
+    10_000_000,
+];
+const ME_WRITER_TEARDOWN_DURATION_BUCKET_LABELS: [&str; ME_WRITER_TEARDOWN_DURATION_BUCKET_COUNT] = [
+    "0.001",
+    "0.005",
+    "0.01",
+    "0.025",
+    "0.05",
+    "0.1",
+    "0.25",
+    "0.5",
+    "1",
+    "2.5",
+    "5",
+    "10",
+];
+
+#[derive(Clone, Copy, Debug, PartialEq, Eq, Hash)]
+#[repr(u8)]
+pub enum MeWriterTeardownMode {
+    Normal = 0,
+    HardDetach = 1,
+}
+
+impl MeWriterTeardownMode {
+    pub const ALL: [Self; ME_WRITER_TEARDOWN_MODE_COUNT] =
+        [Self::Normal, Self::HardDetach];
+
+    pub const fn as_str(self) -> &'static str {
+        match self {
+            Self::Normal => "normal",
+            Self::HardDetach => "hard_detach",
+        }
+    }
+
+    const fn idx(self) -> usize {
+        self as usize
+    }
+}
+
+#[derive(Clone, Copy, Debug, PartialEq, Eq, Hash)]
+#[repr(u8)]
+pub enum MeWriterTeardownReason {
+    ReaderExit = 0,
+    WriterTaskExit = 1,
+    PingSendFail = 2,
+    SignalSendFail = 3,
+    RouteChannelClosed = 4,
+    CloseRpcChannelClosed = 5,
+    PruneClosedWriter = 6,
+    ReapTimeoutExpired = 7,
+    ReapThresholdForce = 8,
+    ReapEmpty = 9,
+    WatchdogStuckDraining = 10,
+}
+
+impl MeWriterTeardownReason {
+    pub const ALL: [Self; ME_WRITER_TEARDOWN_REASON_COUNT] = [
+        Self::ReaderExit,
+        Self::WriterTaskExit,
+        Self::PingSendFail,
+        Self::SignalSendFail,
+        Self::RouteChannelClosed,
+        Self::CloseRpcChannelClosed,
+        Self::PruneClosedWriter,
+        Self::ReapTimeoutExpired,
+        Self::ReapThresholdForce,
+        Self::ReapEmpty,
+        Self::WatchdogStuckDraining,
+    ];
+
+    pub const fn as_str(self) -> &'static str {
+        match self {
+            Self::ReaderExit => "reader_exit",
+            Self::WriterTaskExit => "writer_task_exit",
+            Self::PingSendFail => "ping_send_fail",
+            Self::SignalSendFail => "signal_send_fail",
+            Self::RouteChannelClosed => "route_channel_closed",
+            Self::CloseRpcChannelClosed => "close_rpc_channel_closed",
+            Self::PruneClosedWriter => "prune_closed_writer",
+            Self::ReapTimeoutExpired => "reap_timeout_expired",
+            Self::ReapThresholdForce => "reap_threshold_force",
+            Self::ReapEmpty => "reap_empty",
+            Self::WatchdogStuckDraining => "watchdog_stuck_draining",
+        }
+    }
+
+    const fn idx(self) -> usize {
+        self as usize
+    }
+}
+
+#[derive(Clone, Copy, Debug, PartialEq, Eq, Hash)]
+#[repr(u8)]
+pub enum MeWriterCleanupSideEffectStep {
+    CloseSignalChannelFull = 0,
+    CloseSignalChannelClosed = 1,
+}
+
+impl MeWriterCleanupSideEffectStep {
+    pub const ALL: [Self; ME_WRITER_CLEANUP_SIDE_EFFECT_STEP_COUNT] =
+        [Self::CloseSignalChannelFull, Self::CloseSignalChannelClosed];
+
+    pub const fn as_str(self) -> &'static str {
+        match self {
+            Self::CloseSignalChannelFull => "close_signal_channel_full",
+            Self::CloseSignalChannelClosed => "close_signal_channel_closed",
+        }
+    }
+
+    const fn idx(self) -> usize {
+        self as usize
+    }
+}
+
 // ============= Stats =============
 
 #[derive(Default)]
@@ -123,8 +254,23 @@ pub struct Stats {
     pool_drain_soft_evict_total: AtomicU64,
     pool_drain_soft_evict_writer_total: AtomicU64,
     pool_stale_pick_total: AtomicU64,
+    me_writer_close_signal_drop_total: AtomicU64,
+    me_writer_close_signal_channel_full_total: AtomicU64,
+    me_draining_writers_reap_progress_total: AtomicU64,
     me_writer_removed_total: AtomicU64,
     me_writer_removed_unexpected_total: AtomicU64,
+    me_writer_teardown_attempt_total:
+        [[AtomicU64; ME_WRITER_TEARDOWN_MODE_COUNT]; ME_WRITER_TEARDOWN_REASON_COUNT],
+    me_writer_teardown_success_total: [AtomicU64; ME_WRITER_TEARDOWN_MODE_COUNT],
+    me_writer_teardown_timeout_total: AtomicU64,
+    me_writer_teardown_escalation_total: AtomicU64,
+    me_writer_teardown_noop_total: AtomicU64,
+    me_writer_cleanup_side_effect_failures_total:
+        [AtomicU64; ME_WRITER_CLEANUP_SIDE_EFFECT_STEP_COUNT],
+    me_writer_teardown_duration_bucket_hits:
+        [[AtomicU64; ME_WRITER_TEARDOWN_DURATION_BUCKET_COUNT + 1]; ME_WRITER_TEARDOWN_MODE_COUNT],
+    me_writer_teardown_duration_sum_micros: [AtomicU64; ME_WRITER_TEARDOWN_MODE_COUNT],
+    me_writer_teardown_duration_count: [AtomicU64; ME_WRITER_TEARDOWN_MODE_COUNT],
     me_refill_triggered_total: AtomicU64,
     me_refill_skipped_inflight_total: AtomicU64,
     me_refill_failed_total: AtomicU64,
@@ -734,6 +880,24 @@ impl Stats {
             self.pool_stale_pick_total.fetch_add(1, Ordering::Relaxed);
         }
     }
+    pub fn increment_me_writer_close_signal_drop_total(&self) {
+        if self.telemetry_me_allows_normal() {
+            self.me_writer_close_signal_drop_total
+                .fetch_add(1, Ordering::Relaxed);
+        }
+    }
+    pub fn increment_me_writer_close_signal_channel_full_total(&self) {
+        if self.telemetry_me_allows_normal() {
+            self.me_writer_close_signal_channel_full_total
+                .fetch_add(1, Ordering::Relaxed);
+        }
+    }
+    pub fn increment_me_draining_writers_reap_progress_total(&self) {
+        if self.telemetry_me_allows_normal() {
+            self.me_draining_writers_reap_progress_total
+                .fetch_add(1, Ordering::Relaxed);
+        }
+    }
     pub fn increment_me_writer_removed_total(&self) {
         if self.telemetry_me_allows_debug() {
             self.me_writer_removed_total.fetch_add(1, Ordering::Relaxed);
@@ -744,6 +908,74 @@ impl Stats {
             self.me_writer_removed_unexpected_total.fetch_add(1, Ordering::Relaxed);
         }
     }
+    pub fn increment_me_writer_teardown_attempt_total(
+        &self,
+        reason: MeWriterTeardownReason,
+        mode: MeWriterTeardownMode,
+    ) {
+        if self.telemetry_me_allows_normal() {
+            self.me_writer_teardown_attempt_total[reason.idx()][mode.idx()]
+                .fetch_add(1, Ordering::Relaxed);
+        }
+    }
+    pub fn increment_me_writer_teardown_success_total(&self, mode: MeWriterTeardownMode) {
+        if self.telemetry_me_allows_normal() {
+            self.me_writer_teardown_success_total[mode.idx()].fetch_add(1, Ordering::Relaxed);
+        }
+    }
+    pub fn increment_me_writer_teardown_timeout_total(&self) {
+        if self.telemetry_me_allows_normal() {
+            self.me_writer_teardown_timeout_total
+                .fetch_add(1, Ordering::Relaxed);
+        }
+    }
+    pub fn increment_me_writer_teardown_escalation_total(&self) {
+        if self.telemetry_me_allows_normal() {
+            self.me_writer_teardown_escalation_total
+                .fetch_add(1, Ordering::Relaxed);
+        }
+    }
+    pub fn increment_me_writer_teardown_noop_total(&self) {
+        if self.telemetry_me_allows_normal() {
+            self.me_writer_teardown_noop_total
+                .fetch_add(1, Ordering::Relaxed);
+        }
+    }
+    pub fn increment_me_writer_cleanup_side_effect_failures_total(
+        &self,
+        step: MeWriterCleanupSideEffectStep,
+    ) {
+        if self.telemetry_me_allows_normal() {
+            self.me_writer_cleanup_side_effect_failures_total[step.idx()]
+                .fetch_add(1, Ordering::Relaxed);
+        }
+    }
+    pub fn observe_me_writer_teardown_duration(
+        &self,
+        mode: MeWriterTeardownMode,
+        duration: Duration,
+    ) {
+        if !self.telemetry_me_allows_normal() {
+            return;
+        }
+        let duration_micros = duration.as_micros().min(u64::MAX as u128) as u64;
+        let mut bucket_idx = ME_WRITER_TEARDOWN_DURATION_BUCKET_COUNT;
+        for (idx, upper_bound_micros) in ME_WRITER_TEARDOWN_DURATION_BUCKET_BOUNDS_MICROS
+            .iter()
+            .copied()
+            .enumerate()
+        {
+            if duration_micros <= upper_bound_micros {
+                bucket_idx = idx;
+                break;
+            }
+        }
+        self.me_writer_teardown_duration_bucket_hits[mode.idx()][bucket_idx]
+            .fetch_add(1, Ordering::Relaxed);
+        self.me_writer_teardown_duration_sum_micros[mode.idx()]
+            .fetch_add(duration_micros, Ordering::Relaxed);
+        self.me_writer_teardown_duration_count[mode.idx()].fetch_add(1, Ordering::Relaxed);
+    }
     pub fn increment_me_refill_triggered_total(&self) {
         if self.telemetry_me_allows_debug() {
             self.me_refill_triggered_total.fetch_add(1, Ordering::Relaxed);
@@ -1259,12 +1491,96 @@ impl Stats {
     pub fn get_pool_stale_pick_total(&self) -> u64 {
         self.pool_stale_pick_total.load(Ordering::Relaxed)
     }
+    pub fn get_me_writer_close_signal_drop_total(&self) -> u64 {
+        self.me_writer_close_signal_drop_total.load(Ordering::Relaxed)
+    }
+    pub fn get_me_writer_close_signal_channel_full_total(&self) -> u64 {
+        self.me_writer_close_signal_channel_full_total
+            .load(Ordering::Relaxed)
+    }
+    pub fn get_me_draining_writers_reap_progress_total(&self) -> u64 {
+        self.me_draining_writers_reap_progress_total
+            .load(Ordering::Relaxed)
+    }
     pub fn get_me_writer_removed_total(&self) -> u64 {
         self.me_writer_removed_total.load(Ordering::Relaxed)
     }
     pub fn get_me_writer_removed_unexpected_total(&self) -> u64 {
         self.me_writer_removed_unexpected_total.load(Ordering::Relaxed)
     }
+    pub fn get_me_writer_teardown_attempt_total(
+        &self,
+        reason: MeWriterTeardownReason,
+        mode: MeWriterTeardownMode,
+    ) -> u64 {
+        self.me_writer_teardown_attempt_total[reason.idx()][mode.idx()]
+            .load(Ordering::Relaxed)
+    }
+    pub fn get_me_writer_teardown_attempt_total_by_mode(&self, mode: MeWriterTeardownMode) -> u64 {
+        MeWriterTeardownReason::ALL
+            .iter()
+            .copied()
+            .map(|reason| self.get_me_writer_teardown_attempt_total(reason, mode))
+            .sum()
+    }
+    pub fn get_me_writer_teardown_success_total(&self, mode: MeWriterTeardownMode) -> u64 {
+        self.me_writer_teardown_success_total[mode.idx()].load(Ordering::Relaxed)
+    }
+    pub fn get_me_writer_teardown_timeout_total(&self) -> u64 {
+        self.me_writer_teardown_timeout_total.load(Ordering::Relaxed)
+    }
+    pub fn get_me_writer_teardown_escalation_total(&self) -> u64 {
+        self.me_writer_teardown_escalation_total
+            .load(Ordering::Relaxed)
+    }
+    pub fn get_me_writer_teardown_noop_total(&self) -> u64 {
+        self.me_writer_teardown_noop_total.load(Ordering::Relaxed)
+    }
+    pub fn get_me_writer_cleanup_side_effect_failures_total(
+        &self,
+        step: MeWriterCleanupSideEffectStep,
+    ) -> u64 {
+        self.me_writer_cleanup_side_effect_failures_total[step.idx()]
+            .load(Ordering::Relaxed)
+    }
+    pub fn get_me_writer_cleanup_side_effect_failures_total_all(&self) -> u64 {
+        MeWriterCleanupSideEffectStep::ALL
+            .iter()
+            .copied()
+            .map(|step| self.get_me_writer_cleanup_side_effect_failures_total(step))
+            .sum()
+    }
+    pub fn me_writer_teardown_duration_bucket_labels(
+    ) -> &'static [&'static str; ME_WRITER_TEARDOWN_DURATION_BUCKET_COUNT] {
+        &ME_WRITER_TEARDOWN_DURATION_BUCKET_LABELS
+    }
+    pub fn get_me_writer_teardown_duration_bucket_hits(
+        &self,
+        mode: MeWriterTeardownMode,
+        bucket_idx: usize,
+    ) -> u64 {
+        self.me_writer_teardown_duration_bucket_hits[mode.idx()][bucket_idx]
+            .load(Ordering::Relaxed)
+    }
+    pub fn get_me_writer_teardown_duration_bucket_total(
+        &self,
+        mode: MeWriterTeardownMode,
+        bucket_idx: usize,
+    ) -> u64 {
+        let capped_idx = bucket_idx.min(ME_WRITER_TEARDOWN_DURATION_BUCKET_COUNT);
+        let mut total = 0u64;
+        for idx in 0..=capped_idx {
+            total = total.saturating_add(self.get_me_writer_teardown_duration_bucket_hits(mode, idx));
+        }
+        total
+    }
+    pub fn get_me_writer_teardown_duration_count(&self, mode: MeWriterTeardownMode) -> u64 {
+        self.me_writer_teardown_duration_count[mode.idx()].load(Ordering::Relaxed)
+    }
+    pub fn get_me_writer_teardown_duration_sum_seconds(&self, mode: MeWriterTeardownMode) -> f64 {
+        self.me_writer_teardown_duration_sum_micros[mode.idx()].load(Ordering::Relaxed) as f64
+            / 1_000_000.0
+    }
     pub fn get_me_refill_triggered_total(&self) -> u64 {
         self.me_refill_triggered_total.load(Ordering::Relaxed)
     }
@@ -1768,6 +2084,79 @@ mod tests {
         assert_eq!(stats.get_me_keepalive_sent(), 0);
         assert_eq!(stats.get_me_route_drop_queue_full(), 0);
     }
+
+    #[test]
+    fn test_teardown_counters_and_duration() {
+        let stats = Stats::new();
+        stats.increment_me_writer_teardown_attempt_total(
+            MeWriterTeardownReason::ReaderExit,
+            MeWriterTeardownMode::Normal,
+        );
+        stats.increment_me_writer_teardown_success_total(MeWriterTeardownMode::Normal);
+        stats.observe_me_writer_teardown_duration(
+            MeWriterTeardownMode::Normal,
+            Duration::from_millis(3),
+        );
+        stats.increment_me_writer_cleanup_side_effect_failures_total(
+            MeWriterCleanupSideEffectStep::CloseSignalChannelFull,
+        );
+
+        assert_eq!(
+            stats.get_me_writer_teardown_attempt_total(
+                MeWriterTeardownReason::ReaderExit,
+                MeWriterTeardownMode::Normal
+            ),
+            1
+        );
+        assert_eq!(
+            stats.get_me_writer_teardown_success_total(MeWriterTeardownMode::Normal),
+            1
+        );
+        assert_eq!(
+            stats.get_me_writer_teardown_duration_count(MeWriterTeardownMode::Normal),
+            1
+        );
+        assert!(
+            stats.get_me_writer_teardown_duration_sum_seconds(MeWriterTeardownMode::Normal) > 0.0
+        );
+        assert_eq!(
+            stats.get_me_writer_cleanup_side_effect_failures_total(
+                MeWriterCleanupSideEffectStep::CloseSignalChannelFull
+            ),
+            1
+        );
+    }
+
+    #[test]
+    fn test_teardown_counters_respect_me_silent() {
+        let stats = Stats::new();
+        stats.apply_telemetry_policy(TelemetryPolicy {
+            core_enabled: true,
+            user_enabled: true,
+            me_level: MeTelemetryLevel::Silent,
+        });
+        stats.increment_me_writer_teardown_attempt_total(
+            MeWriterTeardownReason::ReaderExit,
+            MeWriterTeardownMode::Normal,
+        );
+        stats.increment_me_writer_teardown_timeout_total();
+        stats.observe_me_writer_teardown_duration(
+            MeWriterTeardownMode::Normal,
+            Duration::from_millis(1),
+        );
+        assert_eq!(
+            stats.get_me_writer_teardown_attempt_total(
+                MeWriterTeardownReason::ReaderExit,
+                MeWriterTeardownMode::Normal
+            ),
+            0
+        );
+        assert_eq!(stats.get_me_writer_teardown_timeout_total(), 0);
+        assert_eq!(
+            stats.get_me_writer_teardown_duration_count(MeWriterTeardownMode::Normal),
+            0
+        );
+    }
     
     #[test]
     fn test_replay_checker_basic() {

src/tls_front/fetcher.rs

@@ -399,15 +399,17 @@ async fn connect_tcp_with_upstream(
     port: u16,
     connect_timeout: Duration,
     upstream: Option<std::sync::Arc<crate::transport::UpstreamManager>>,
+    scope: Option<&str>,
 ) -> Result<UpstreamStream> {
     if let Some(manager) = upstream {
         if let Some(addr) = resolve_socket_addr(host, port) {
-            match manager.connect(addr, None, None).await {
+            match manager.connect(addr, None, scope).await {
                 Ok(stream) => return Ok(stream),
                 Err(e) => {
                     warn!(
                         host = %host,
                         port = port,
+                        scope = ?scope,
                         error = %e,
                         "Upstream connect failed, using direct connect"
                     );
@@ -416,12 +418,13 @@ async fn connect_tcp_with_upstream(
         } else if let Ok(mut addrs) = tokio::net::lookup_host((host, port)).await
             && let Some(addr) = addrs.find(|a| a.is_ipv4())
         {
-            match manager.connect(addr, None, None).await {
+            match manager.connect(addr, None, scope).await {
                 Ok(stream) => return Ok(stream),
                 Err(e) => {
                     warn!(
                         host = %host,
                         port = port,
+                        scope = ?scope,
                         error = %e,
                         "Upstream connect failed, using direct connect"
                     );
@@ -542,6 +545,7 @@ async fn fetch_via_raw_tls(
     sni: &str,
     connect_timeout: Duration,
     upstream: Option<std::sync::Arc<crate::transport::UpstreamManager>>,
+    scope: Option<&str>,
     proxy_protocol: u8,
     unix_sock: Option<&str>,
 ) -> Result<TlsFetchResult> {
@@ -578,7 +582,7 @@ async fn fetch_via_raw_tls(
     #[cfg(not(unix))]
     let _ = unix_sock;
 
-    let stream = connect_tcp_with_upstream(host, port, connect_timeout, upstream).await?;
+    let stream = connect_tcp_with_upstream(host, port, connect_timeout, upstream, scope).await?;
     fetch_via_raw_tls_stream(stream, sni, connect_timeout, proxy_protocol).await
 }
 
@@ -682,6 +686,7 @@ async fn fetch_via_rustls(
     sni: &str,
     connect_timeout: Duration,
     upstream: Option<std::sync::Arc<crate::transport::UpstreamManager>>,
+    scope: Option<&str>,
     proxy_protocol: u8,
     unix_sock: Option<&str>,
 ) -> Result<TlsFetchResult> {
@@ -717,7 +722,7 @@ async fn fetch_via_rustls(
     #[cfg(not(unix))]
     let _ = unix_sock;
 
-    let stream = connect_tcp_with_upstream(host, port, connect_timeout, upstream).await?;
+    let stream = connect_tcp_with_upstream(host, port, connect_timeout, upstream, scope).await?;
     fetch_via_rustls_stream(stream, host, sni, proxy_protocol).await
 }
 
@@ -733,6 +738,7 @@ pub async fn fetch_real_tls(
     sni: &str,
     connect_timeout: Duration,
     upstream: Option<std::sync::Arc<crate::transport::UpstreamManager>>,
+    scope: Option<&str>,
     proxy_protocol: u8,
     unix_sock: Option<&str>,
 ) -> Result<TlsFetchResult> {
@@ -742,6 +748,7 @@ pub async fn fetch_real_tls(
         sni,
         connect_timeout,
         upstream.clone(),
+        scope,
         proxy_protocol,
         unix_sock,
     )
@@ -760,6 +767,7 @@ pub async fn fetch_real_tls(
         sni,
         connect_timeout,
         upstream,
+        scope,
         proxy_protocol,
         unix_sock,
     )

src/transport/middle_proxy/config_updater.rs

@@ -298,6 +298,7 @@ async fn run_update_cycle(
     pool.update_runtime_reinit_policy(
         cfg.general.hardswap,
         cfg.general.me_pool_drain_ttl_secs,
+        cfg.general.me_instadrain,
         cfg.general.me_pool_drain_threshold,
         cfg.general.me_pool_drain_soft_evict_enabled,
         cfg.general.me_pool_drain_soft_evict_grace_secs,
@@ -530,6 +531,7 @@ pub async fn me_config_updater(
                 pool.update_runtime_reinit_policy(
                     cfg.general.hardswap,
                     cfg.general.me_pool_drain_ttl_secs,
+                    cfg.general.me_instadrain,
                     cfg.general.me_pool_drain_threshold,
                     cfg.general.me_pool_drain_soft_evict_enabled,
                     cfg.general.me_pool_drain_soft_evict_grace_secs,

src/transport/middle_proxy/health.rs

@@ -10,8 +10,10 @@ use tracing::{debug, info, warn};
 use crate::config::MeFloorMode;
 use crate::crypto::SecureRandom;
 use crate::network::IpFamily;
+use crate::stats::MeWriterTeardownReason;
 
 use super::MePool;
+use super::pool::{MeFamilyRuntimeState, MeWriter};
 
 const JITTER_FRAC_NUM: u64 = 2; // jitter up to 50% of backoff
 #[allow(dead_code)]
@@ -30,6 +32,35 @@ const HEALTH_DRAIN_CLOSE_BUDGET_MIN: usize = 16;
 const HEALTH_DRAIN_CLOSE_BUDGET_MAX: usize = 256;
 const HEALTH_DRAIN_SOFT_EVICT_BUDGET_MIN: usize = 8;
 const HEALTH_DRAIN_SOFT_EVICT_BUDGET_MAX: usize = 256;
+const HEALTH_DRAIN_REAP_OPPORTUNISTIC_INTERVAL_SECS: u64 = 1;
+const HEALTH_DRAIN_TIMEOUT_ENFORCER_INTERVAL_SECS: u64 = 1;
+const FAMILY_SUPPRESS_FAIL_STREAK_THRESHOLD: u32 = 6;
+const FAMILY_SUPPRESS_WINDOW_SECS: u64 = 120;
+const FAMILY_RECOVER_PROBE_INTERVAL_SECS: u64 = 5;
+const FAMILY_RECOVER_SUCCESS_STREAK_REQUIRED: u32 = 3;
+
+#[derive(Debug, Clone)]
+struct FamilyCircuitState {
+    state: MeFamilyRuntimeState,
+    state_since_at: Instant,
+    suppressed_until: Option<Instant>,
+    next_probe_at: Instant,
+    fail_streak: u32,
+    recover_success_streak: u32,
+}
+
+impl FamilyCircuitState {
+    fn new(now: Instant) -> Self {
+        Self {
+            state: MeFamilyRuntimeState::Healthy,
+            state_since_at: now,
+            suppressed_until: None,
+            next_probe_at: now,
+            fail_streak: 0,
+            recover_success_streak: 0,
+        }
+    }
+}
 
 #[derive(Debug, Clone)]
 struct DcFloorPlanEntry {
@@ -69,6 +100,25 @@ pub async fn me_health_monitor(pool: Arc<MePool>, rng: Arc<SecureRandom>, _min_c
     let mut floor_warn_next_allowed: HashMap<(i32, IpFamily), Instant> = HashMap::new();
     let mut drain_warn_next_allowed: HashMap<u64, Instant> = HashMap::new();
     let mut drain_soft_evict_next_allowed: HashMap<u64, Instant> = HashMap::new();
+    let mut family_v4_circuit = FamilyCircuitState::new(Instant::now());
+    let mut family_v6_circuit = FamilyCircuitState::new(Instant::now());
+    let init_epoch_secs = MePool::now_epoch_secs();
+    pool.set_family_runtime_state(
+        IpFamily::V4,
+        family_v4_circuit.state,
+        init_epoch_secs,
+        0,
+        family_v4_circuit.fail_streak,
+        family_v4_circuit.recover_success_streak,
+    );
+    pool.set_family_runtime_state(
+        IpFamily::V6,
+        family_v6_circuit.state,
+        init_epoch_secs,
+        0,
+        family_v6_circuit.fail_streak,
+        family_v6_circuit.recover_success_streak,
+    );
     let mut degraded_interval = true;
     loop {
         let interval = if degraded_interval {
@@ -84,7 +134,9 @@ pub async fn me_health_monitor(pool: Arc<MePool>, rng: Arc<SecureRandom>, _min_c
             &mut drain_soft_evict_next_allowed,
         )
         .await;
-        let v4_degraded = check_family(
+        let now = Instant::now();
+        let now_epoch_secs = MePool::now_epoch_secs();
+        let v4_degraded_raw = check_family(
             IpFamily::V4,
             &pool,
             &rng,
@@ -99,29 +151,252 @@ pub async fn me_health_monitor(pool: Arc<MePool>, rng: Arc<SecureRandom>, _min_c
             &mut adaptive_idle_since,
             &mut adaptive_recover_until,
             &mut floor_warn_next_allowed,
+            &mut drain_warn_next_allowed,
+            &mut drain_soft_evict_next_allowed,
         )
         .await;
-        let v6_degraded = check_family(
+        let v4_degraded = apply_family_circuit_result(
-            IpFamily::V6,
             &pool,
-            &rng,
+            IpFamily::V4,
-            &mut backoff,
+            &mut family_v4_circuit,
-            &mut next_attempt,
+            Some(v4_degraded_raw),
-            &mut inflight,
+            false,
-            &mut outage_backoff,
+            now,
-            &mut outage_next_attempt,
+            now_epoch_secs,
-            &mut single_endpoint_outage,
+        );
-            &mut shadow_rotate_deadline,
+
-            &mut idle_refresh_next_attempt,
+        let v6_check_ran = should_run_family_check(&mut family_v6_circuit, now);
-            &mut adaptive_idle_since,
+        let v6_degraded_raw = if v6_check_ran {
-            &mut adaptive_recover_until,
+            check_family(
-            &mut floor_warn_next_allowed,
+                IpFamily::V6,
-        )
+                &pool,
-        .await;
+                &rng,
+                &mut backoff,
+                &mut next_attempt,
+                &mut inflight,
+                &mut outage_backoff,
+                &mut outage_next_attempt,
+                &mut single_endpoint_outage,
+                &mut shadow_rotate_deadline,
+                &mut idle_refresh_next_attempt,
+                &mut adaptive_idle_since,
+                &mut adaptive_recover_until,
+                &mut floor_warn_next_allowed,
+                &mut drain_warn_next_allowed,
+                &mut drain_soft_evict_next_allowed,
+            )
+            .await
+        } else {
+            false
+        };
+        let v6_degraded = apply_family_circuit_result(
+            &pool,
+            IpFamily::V6,
+            &mut family_v6_circuit,
+            if v6_check_ran {
+                Some(v6_degraded_raw)
+            } else {
+                None
+            },
+            true,
+            now,
+            now_epoch_secs,
+        );
         degraded_interval = v4_degraded || v6_degraded;
     }
 }
 
+pub async fn me_drain_timeout_enforcer(pool: Arc<MePool>) {
+    let mut drain_warn_next_allowed: HashMap<u64, Instant> = HashMap::new();
+    let mut drain_soft_evict_next_allowed: HashMap<u64, Instant> = HashMap::new();
+    loop {
+        tokio::time::sleep(Duration::from_secs(
+            HEALTH_DRAIN_TIMEOUT_ENFORCER_INTERVAL_SECS,
+        ))
+        .await;
+        reap_draining_writers(
+            &pool,
+            &mut drain_warn_next_allowed,
+            &mut drain_soft_evict_next_allowed,
+        )
+        .await;
+    }
+}
+
+fn should_run_family_check(circuit: &mut FamilyCircuitState, now: Instant) -> bool {
+    match circuit.state {
+        MeFamilyRuntimeState::Suppressed => {
+            if now < circuit.next_probe_at {
+                return false;
+            }
+            circuit.next_probe_at =
+                now + Duration::from_secs(FAMILY_RECOVER_PROBE_INTERVAL_SECS);
+            true
+        }
+        _ => true,
+    }
+}
+
+fn apply_family_circuit_result(
+    pool: &Arc<MePool>,
+    family: IpFamily,
+    circuit: &mut FamilyCircuitState,
+    degraded: Option<bool>,
+    allow_suppress: bool,
+    now: Instant,
+    now_epoch_secs: u64,
+) -> bool {
+    let Some(degraded) = degraded else {
+        // Preserve suppression state when probe tick is intentionally skipped.
+        return false;
+    };
+
+    let previous_state = circuit.state;
+    match circuit.state {
+        MeFamilyRuntimeState::Suppressed => {
+            if degraded {
+                circuit.fail_streak = circuit.fail_streak.saturating_add(1);
+                circuit.recover_success_streak = 0;
+                let until = now + Duration::from_secs(FAMILY_SUPPRESS_WINDOW_SECS);
+                circuit.suppressed_until = Some(until);
+                circuit.state_since_at = now;
+                warn!(
+                    ?family,
+                    fail_streak = circuit.fail_streak,
+                    suppress_secs = FAMILY_SUPPRESS_WINDOW_SECS,
+                    "ME family remains suppressed due to ongoing failures"
+                );
+            } else {
+                circuit.fail_streak = 0;
+                circuit.recover_success_streak = 1;
+                circuit.state = MeFamilyRuntimeState::Recovering;
+            }
+        }
+        MeFamilyRuntimeState::Recovering => {
+            if degraded {
+                circuit.fail_streak = circuit.fail_streak.saturating_add(1);
+                if allow_suppress {
+                    circuit.state = MeFamilyRuntimeState::Suppressed;
+                    let until = now + Duration::from_secs(FAMILY_SUPPRESS_WINDOW_SECS);
+                    circuit.suppressed_until = Some(until);
+                    circuit.next_probe_at =
+                        now + Duration::from_secs(FAMILY_RECOVER_PROBE_INTERVAL_SECS);
+                    warn!(
+                        ?family,
+                        fail_streak = circuit.fail_streak,
+                        suppress_secs = FAMILY_SUPPRESS_WINDOW_SECS,
+                        "ME family temporarily suppressed after repeated degradation"
+                    );
+                } else {
+                    circuit.state = MeFamilyRuntimeState::Degraded;
+                }
+            } else {
+                circuit.recover_success_streak = circuit.recover_success_streak.saturating_add(1);
+                if circuit.recover_success_streak >= FAMILY_RECOVER_SUCCESS_STREAK_REQUIRED {
+                    circuit.fail_streak = 0;
+                    circuit.recover_success_streak = 0;
+                    circuit.suppressed_until = None;
+                    circuit.state = MeFamilyRuntimeState::Healthy;
+                    info!(
+                        ?family,
+                        "ME family suppression lifted after stable recovery probes"
+                    );
+                }
+            }
+        }
+        _ => {
+            if degraded {
+                circuit.fail_streak = circuit.fail_streak.saturating_add(1);
+                circuit.recover_success_streak = 0;
+                circuit.state = MeFamilyRuntimeState::Degraded;
+                if allow_suppress && circuit.fail_streak >= FAMILY_SUPPRESS_FAIL_STREAK_THRESHOLD {
+                    circuit.state = MeFamilyRuntimeState::Suppressed;
+                    let until = now + Duration::from_secs(FAMILY_SUPPRESS_WINDOW_SECS);
+                    circuit.suppressed_until = Some(until);
+                    circuit.next_probe_at =
+                        now + Duration::from_secs(FAMILY_RECOVER_PROBE_INTERVAL_SECS);
+                    warn!(
+                        ?family,
+                        fail_streak = circuit.fail_streak,
+                        suppress_secs = FAMILY_SUPPRESS_WINDOW_SECS,
+                        "ME family temporarily suppressed after repeated degradation"
+                    );
+                }
+            } else {
+                circuit.fail_streak = 0;
+                circuit.recover_success_streak = 0;
+                circuit.suppressed_until = None;
+                circuit.state = MeFamilyRuntimeState::Healthy;
+            }
+        }
+    }
+
+    if previous_state != circuit.state {
+        circuit.state_since_at = now;
+    }
+
+    let suppressed_until_epoch_secs = circuit
+        .suppressed_until
+        .and_then(|until| {
+            if until > now {
+                Some(
+                    now_epoch_secs
+                        .saturating_add(until.saturating_duration_since(now).as_secs()),
+                )
+            } else {
+                None
+            }
+        })
+        .unwrap_or(0);
+    let state_since_epoch_secs = if previous_state == circuit.state {
+        pool.family_runtime_state_since_epoch_secs(family)
+    } else {
+        now_epoch_secs
+    };
+    pool.set_family_runtime_state(
+        family,
+        circuit.state,
+        state_since_epoch_secs,
+        suppressed_until_epoch_secs,
+        circuit.fail_streak,
+        circuit.recover_success_streak,
+    );
+
+    !matches!(circuit.state, MeFamilyRuntimeState::Suppressed) && degraded
+}
+
+fn draining_writer_timeout_expired(
+    pool: &MePool,
+    writer: &MeWriter,
+    now_epoch_secs: u64,
+    drain_ttl_secs: u64,
+) -> bool {
+    if pool
+        .me_instadrain
+        .load(std::sync::atomic::Ordering::Relaxed)
+    {
+        return true;
+    }
+
+    let deadline_epoch_secs = writer
+        .drain_deadline_epoch_secs
+        .load(std::sync::atomic::Ordering::Relaxed);
+    if deadline_epoch_secs != 0 {
+        return now_epoch_secs >= deadline_epoch_secs;
+    }
+
+    if drain_ttl_secs == 0 {
+        return false;
+    }
+    let drain_started_at_epoch_secs = writer
+        .draining_started_at_epoch_secs
+        .load(std::sync::atomic::Ordering::Relaxed);
+    if drain_started_at_epoch_secs == 0 {
+        return false;
+    }
+    now_epoch_secs.saturating_sub(drain_started_at_epoch_secs) > drain_ttl_secs
+}
+
 pub(super) async fn reap_draining_writers(
     pool: &Arc<MePool>,
     warn_next_allowed: &mut HashMap<u64, Instant>,
@@ -137,11 +412,16 @@ pub(super) async fn reap_draining_writers(
     let activity = pool.registry.writer_activity_snapshot().await;
     let mut draining_writers = Vec::new();
     let mut empty_writer_ids = Vec::<u64>::new();
+    let mut timeout_expired_writer_ids = Vec::<u64>::new();
     let mut force_close_writer_ids = Vec::<u64>::new();
     for writer in writers {
         if !writer.draining.load(std::sync::atomic::Ordering::Relaxed) {
             continue;
         }
+        if draining_writer_timeout_expired(pool, &writer, now_epoch_secs, drain_ttl_secs) {
+            timeout_expired_writer_ids.push(writer.id);
+            continue;
+        }
         if activity
             .bound_clients_by_writer
             .get(&writer.id)
@@ -207,14 +487,6 @@ pub(super) async fn reap_draining_writers(
                 "ME draining writer remains non-empty past drain TTL"
             );
         }
-        let deadline_epoch_secs = writer
-            .drain_deadline_epoch_secs
-            .load(std::sync::atomic::Ordering::Relaxed);
-        if deadline_epoch_secs != 0 && now_epoch_secs >= deadline_epoch_secs {
-            warn!(writer_id = writer.id, "Drain timeout, force-closing");
-            force_close_writer_ids.push(writer.id);
-            active_draining_writer_ids.remove(&writer.id);
-        }
     }
 
     warn_next_allowed.retain(|writer_id, _| active_draining_writer_ids.contains(writer_id));
@@ -299,11 +571,22 @@ pub(super) async fn reap_draining_writers(
         }
     }
 
-    let close_budget = health_drain_close_budget();
+    let mut closed_writer_ids = HashSet::<u64>::new();
+    for writer_id in timeout_expired_writer_ids {
+        if !closed_writer_ids.insert(writer_id) {
+            continue;
+        }
+        pool.stats.increment_pool_force_close_total();
+        pool.remove_writer_and_close_clients(writer_id, MeWriterTeardownReason::ReapTimeoutExpired)
+            .await;
+        pool.stats
+            .increment_me_draining_writers_reap_progress_total();
+    }
+
     let requested_force_close = force_close_writer_ids.len();
     let requested_empty_close = empty_writer_ids.len();
     let requested_close_total = requested_force_close.saturating_add(requested_empty_close);
-    let mut closed_writer_ids = HashSet::<u64>::new();
+    let close_budget = health_drain_close_budget();
     let mut closed_total = 0usize;
     for writer_id in force_close_writer_ids {
         if closed_total >= close_budget {
@@ -313,7 +596,10 @@ pub(super) async fn reap_draining_writers(
             continue;
         }
         pool.stats.increment_pool_force_close_total();
-        pool.remove_writer_and_close_clients(writer_id).await;
+        pool.remove_writer_and_close_clients(writer_id, MeWriterTeardownReason::ReapThresholdForce)
+            .await;
+        pool.stats
+            .increment_me_draining_writers_reap_progress_total();
         closed_total = closed_total.saturating_add(1);
     }
     for writer_id in empty_writer_ids {
@@ -323,7 +609,10 @@ pub(super) async fn reap_draining_writers(
         if !closed_writer_ids.insert(writer_id) {
             continue;
         }
-        pool.remove_writer_and_close_clients(writer_id).await;
+        pool.remove_writer_and_close_clients(writer_id, MeWriterTeardownReason::ReapEmpty)
+            .await;
+        pool.stats
+            .increment_me_draining_writers_reap_progress_total();
         closed_total = closed_total.saturating_add(1);
     }
 
@@ -392,6 +681,8 @@ async fn check_family(
     adaptive_idle_since: &mut HashMap<(i32, IpFamily), Instant>,
     adaptive_recover_until: &mut HashMap<(i32, IpFamily), Instant>,
     floor_warn_next_allowed: &mut HashMap<(i32, IpFamily), Instant>,
+    drain_warn_next_allowed: &mut HashMap<u64, Instant>,
+    drain_soft_evict_next_allowed: &mut HashMap<u64, Instant>,
 ) -> bool {
     let enabled = match family {
         IpFamily::V4 => pool.decision.ipv4_me,
@@ -472,8 +763,15 @@ async fn check_family(
         floor_plan.active_writers_current,
         floor_plan.warm_writers_current,
     );
+    let mut next_drain_reap_at = Instant::now();
 
     for (dc, endpoints) in dc_endpoints {
+        if Instant::now() >= next_drain_reap_at {
+            reap_draining_writers(pool, drain_warn_next_allowed, drain_soft_evict_next_allowed)
+                .await;
+            next_drain_reap_at = Instant::now()
+                + Duration::from_secs(HEALTH_DRAIN_REAP_OPPORTUNISTIC_INTERVAL_SECS);
+        }
         if endpoints.is_empty() {
             continue;
         }
@@ -617,6 +915,12 @@ async fn check_family(
 
         let mut restored = 0usize;
         for _ in 0..missing {
+            if Instant::now() >= next_drain_reap_at {
+                reap_draining_writers(pool, drain_warn_next_allowed, drain_soft_evict_next_allowed)
+                    .await;
+                next_drain_reap_at = Instant::now()
+                    + Duration::from_secs(HEALTH_DRAIN_REAP_OPPORTUNISTIC_INTERVAL_SECS);
+            }
             if reconnect_budget == 0 {
                 break;
             }
@@ -1468,6 +1772,187 @@ async fn maybe_rotate_single_endpoint_shadow(
     );
 }
 
+/// Last-resort safety net for draining writers stuck past their deadline.
+///
+/// Runs every `TICK_SECS` and force-closes any draining writer whose
+/// `drain_deadline_epoch_secs` has been exceeded by more than a threshold.
+///
+/// Two thresholds:
+///   - `SOFT_THRESHOLD_SECS` (60s): writers with no bound clients
+///   - `HARD_THRESHOLD_SECS` (300s): writers WITH bound clients (unconditional)
+///
+/// Intentionally kept trivial and independent of pool config to minimise
+/// the probability of panicking itself. Uses `SystemTime` directly
+/// as a fallback clock source and timeouts on every lock acquisition
+/// and writer removal so one stuck writer cannot block the rest.
+pub async fn me_zombie_writer_watchdog(pool: Arc<MePool>) {
+    use std::time::{SystemTime, UNIX_EPOCH};
+
+    const TICK_SECS: u64 = 30;
+    const SOFT_THRESHOLD_SECS: u64 = 60;
+    const HARD_THRESHOLD_SECS: u64 = 300;
+    const LOCK_TIMEOUT_SECS: u64 = 5;
+    const REMOVE_TIMEOUT_SECS: u64 = 10;
+    const HARD_DETACH_TIMEOUT_STREAK: u8 = 3;
+
+    let mut removal_timeout_streak = HashMap::<u64, u8>::new();
+
+    loop {
+        tokio::time::sleep(Duration::from_secs(TICK_SECS)).await;
+
+        let now = match SystemTime::now().duration_since(UNIX_EPOCH) {
+            Ok(d) => d.as_secs(),
+            Err(_) => continue,
+        };
+
+        // Phase 1: collect zombie IDs under a short read-lock with timeout.
+        let zombie_ids_with_meta: Vec<(u64, bool)> = {
+            let Ok(ws) = tokio::time::timeout(
+                Duration::from_secs(LOCK_TIMEOUT_SECS),
+                pool.writers.read(),
+            )
+            .await
+            else {
+                warn!("zombie_watchdog: writers read-lock timeout, skipping tick");
+                continue;
+            };
+            ws.iter()
+                .filter(|w| w.draining.load(std::sync::atomic::Ordering::Relaxed))
+                .filter_map(|w| {
+                    let deadline = w
+                        .drain_deadline_epoch_secs
+                        .load(std::sync::atomic::Ordering::Relaxed);
+                    if deadline == 0 {
+                        return None;
+                    }
+                    let overdue = now.saturating_sub(deadline);
+                    if overdue == 0 {
+                        return None;
+                    }
+                    let started = w
+                        .draining_started_at_epoch_secs
+                        .load(std::sync::atomic::Ordering::Relaxed);
+                    let drain_age = now.saturating_sub(started);
+                    if drain_age > HARD_THRESHOLD_SECS {
+                        return Some((w.id, true));
+                    }
+                    if overdue > SOFT_THRESHOLD_SECS {
+                        return Some((w.id, false));
+                    }
+                    None
+                })
+                .collect()
+        };
+        // read lock released here
+
+        if zombie_ids_with_meta.is_empty() {
+            removal_timeout_streak.clear();
+            continue;
+        }
+
+        let mut active_zombie_ids = HashSet::<u64>::with_capacity(zombie_ids_with_meta.len());
+        for (writer_id, _) in &zombie_ids_with_meta {
+            active_zombie_ids.insert(*writer_id);
+        }
+        removal_timeout_streak.retain(|writer_id, _| active_zombie_ids.contains(writer_id));
+
+        warn!(
+            zombie_count = zombie_ids_with_meta.len(),
+            soft_threshold_secs = SOFT_THRESHOLD_SECS,
+            hard_threshold_secs = HARD_THRESHOLD_SECS,
+            "Zombie draining writers detected by watchdog, force-closing"
+        );
+
+        // Phase 2: remove each writer individually with a timeout.
+        // One stuck removal cannot block the rest.
+        for (writer_id, had_clients) in &zombie_ids_with_meta {
+            let result = tokio::time::timeout(
+                Duration::from_secs(REMOVE_TIMEOUT_SECS),
+                pool.remove_writer_and_close_clients(
+                    *writer_id,
+                    MeWriterTeardownReason::WatchdogStuckDraining,
+                ),
+            )
+            .await;
+            match result {
+                Ok(true) => {
+                    removal_timeout_streak.remove(writer_id);
+                    pool.stats.increment_pool_force_close_total();
+                    pool.stats
+                        .increment_me_draining_writers_reap_progress_total();
+                    info!(
+                        writer_id,
+                        had_clients,
+                        "Zombie writer removed by watchdog"
+                    );
+                }
+                Ok(false) => {
+                    removal_timeout_streak.remove(writer_id);
+                    debug!(
+                        writer_id,
+                        had_clients,
+                        "Zombie writer watchdog removal became no-op"
+                    );
+                }
+                Err(_) => {
+                    pool.stats.increment_me_writer_teardown_timeout_total();
+                    let streak = removal_timeout_streak
+                        .entry(*writer_id)
+                        .and_modify(|value| *value = value.saturating_add(1))
+                        .or_insert(1);
+                    warn!(
+                        writer_id,
+                        had_clients,
+                        timeout_streak = *streak,
+                        "Zombie writer removal timed out"
+                    );
+                    if *streak < HARD_DETACH_TIMEOUT_STREAK {
+                        continue;
+                    }
+                    pool.stats.increment_me_writer_teardown_escalation_total();
+
+                    let hard_detach = tokio::time::timeout(
+                        Duration::from_secs(REMOVE_TIMEOUT_SECS),
+                        pool.remove_draining_writer_hard_detach(
+                            *writer_id,
+                            MeWriterTeardownReason::WatchdogStuckDraining,
+                        ),
+                    )
+                    .await;
+                    match hard_detach {
+                        Ok(true) => {
+                            removal_timeout_streak.remove(writer_id);
+                            pool.stats.increment_pool_force_close_total();
+                            pool.stats
+                                .increment_me_draining_writers_reap_progress_total();
+                            info!(
+                                writer_id,
+                                had_clients,
+                                "Zombie writer hard-detached after repeated timeouts"
+                            );
+                        }
+                        Ok(false) => {
+                            removal_timeout_streak.remove(writer_id);
+                            debug!(
+                                writer_id,
+                                had_clients,
+                                "Zombie hard-detach skipped (writer already gone or no longer draining)"
+                            );
+                        }
+                        Err(_) => {
+                            pool.stats.increment_me_writer_teardown_timeout_total();
+                            warn!(
+                                writer_id,
+                                had_clients,
+                                "Zombie hard-detach timed out, will retry next tick"
+                            );
+                        }
+                    }
+                }
+            }
+        }
+    }
+}
 #[cfg(test)]
 mod tests {
     use std::collections::HashMap;
@@ -1479,13 +1964,19 @@ mod tests {
     use tokio::sync::mpsc;
     use tokio_util::sync::CancellationToken;
 
-    use super::reap_draining_writers;
+    use super::{
+        FamilyCircuitState, apply_family_circuit_result, reap_draining_writers,
+        should_run_family_check,
+    };
     use crate::config::{GeneralConfig, MeRouteNoWriterMode, MeSocksKdfPolicy, MeWriterPickMode};
     use crate::crypto::SecureRandom;
+    use crate::network::IpFamily;
     use crate::network::probe::NetworkDecision;
     use crate::stats::Stats;
     use crate::transport::middle_proxy::codec::WriterCommand;
-    use crate::transport::middle_proxy::pool::{MePool, MeWriter, WriterContour};
+    use crate::transport::middle_proxy::pool::{
+        MeFamilyRuntimeState, MePool, MeWriter, WriterContour,
+    };
     use crate::transport::middle_proxy::registry::ConnMeta;
 
     async fn make_pool(me_pool_drain_threshold: u64) -> Arc<MePool> {
@@ -1544,6 +2035,7 @@ mod tests {
             general.me_adaptive_floor_max_warm_writers_global,
             general.hardswap,
             general.me_pool_drain_ttl_secs,
+            general.me_instadrain,
             general.me_pool_drain_threshold,
             general.me_pool_drain_soft_evict_enabled,
             general.me_pool_drain_soft_evict_grace_secs,
@@ -1662,4 +2154,47 @@ mod tests {
         assert_eq!(pool.registry.get_writer(conn_b).await.unwrap().writer_id, 20);
         assert_eq!(pool.registry.get_writer(conn_c).await.unwrap().writer_id, 30);
     }
+
+    #[tokio::test]
+    async fn suppressed_family_probe_skip_preserves_suppressed_state() {
+        let pool = make_pool(0).await;
+        let now = Instant::now();
+        let now_epoch_secs = MePool::now_epoch_secs();
+        let suppressed_until_epoch_secs = now_epoch_secs.saturating_add(60);
+        pool.set_family_runtime_state(
+            IpFamily::V6,
+            MeFamilyRuntimeState::Suppressed,
+            now_epoch_secs,
+            suppressed_until_epoch_secs,
+            7,
+            0,
+        );
+
+        let mut circuit = FamilyCircuitState {
+            state: MeFamilyRuntimeState::Suppressed,
+            state_since_at: now,
+            suppressed_until: Some(now + Duration::from_secs(60)),
+            next_probe_at: now + Duration::from_secs(5),
+            fail_streak: 7,
+            recover_success_streak: 0,
+        };
+
+        assert!(!should_run_family_check(&mut circuit, now));
+        assert!(!apply_family_circuit_result(
+            &pool,
+            IpFamily::V6,
+            &mut circuit,
+            None,
+            true,
+            now,
+            now_epoch_secs,
+        ));
+        assert_eq!(circuit.state, MeFamilyRuntimeState::Suppressed);
+        assert_eq!(circuit.fail_streak, 7);
+        assert_eq!(circuit.recover_success_streak, 0);
+        assert_eq!(
+            pool.family_runtime_state(IpFamily::V6),
+            MeFamilyRuntimeState::Suppressed,
+        );
+    }
 }

src/transport/middle_proxy/health_adversarial_tests.rs

@@ -81,6 +81,7 @@ async fn make_pool(
         general.me_adaptive_floor_max_warm_writers_global,
         general.hardswap,
         general.me_pool_drain_ttl_secs,
+        general.me_instadrain,
         general.me_pool_drain_threshold,
         general.me_pool_drain_soft_evict_enabled,
         general.me_pool_drain_soft_evict_grace_secs,
@@ -213,7 +214,7 @@ async fn reap_draining_writers_respects_threshold_across_multiple_overflow_cycle
         insert_draining_writer(
             &pool,
             writer_id,
-            now_epoch_secs.saturating_sub(600).saturating_add(writer_id),
+            now_epoch_secs.saturating_sub(20),
             1,
             0,
         )
@@ -230,7 +231,7 @@ async fn reap_draining_writers_respects_threshold_across_multiple_overflow_cycle
     }
 
     assert_eq!(writer_count(&pool).await, threshold as usize);
-    assert_eq!(sorted_writer_ids(&pool).await, vec![58, 59, 60]);
+    assert_eq!(sorted_writer_ids(&pool).await, vec![1, 2, 3]);
 }
 
 #[tokio::test]
@@ -315,7 +316,12 @@ async fn reap_draining_writers_maintains_warn_state_subset_property_under_bulk_c
 
         let ids = sorted_writer_ids(&pool).await;
         for writer_id in ids.into_iter().take(3) {
-            let _ = pool.remove_writer_and_close_clients(writer_id).await;
+            let _ = pool
+                .remove_writer_and_close_clients(
+                    writer_id,
+                    crate::stats::MeWriterTeardownReason::ReapEmpty,
+                )
+                .await;
         }
 
         reap_draining_writers(&pool, &mut warn_next_allowed, &mut soft_evict_next_allowed).await;

src/transport/middle_proxy/health_integration_tests.rs

@@ -80,6 +80,7 @@ async fn make_pool(
         general.me_adaptive_floor_max_warm_writers_global,
         general.hardswap,
         general.me_pool_drain_ttl_secs,
+        general.me_instadrain,
         general.me_pool_drain_threshold,
         general.me_pool_drain_soft_evict_enabled,
         general.me_pool_drain_soft_evict_grace_secs,

src/transport/middle_proxy/health_regression_tests.rs

@@ -4,6 +4,7 @@ use std::sync::Arc;
 use std::sync::atomic::{AtomicBool, AtomicU8, AtomicU32, AtomicU64, Ordering};
 use std::time::{Duration, Instant};
 
+use bytes::Bytes;
 use tokio::sync::mpsc;
 use tokio_util::sync::CancellationToken;
 
@@ -11,7 +12,9 @@ use super::codec::WriterCommand;
 use super::health::{health_drain_close_budget, reap_draining_writers};
 use super::pool::{MePool, MeWriter, WriterContour};
 use super::registry::ConnMeta;
-use crate::config::{GeneralConfig, MeRouteNoWriterMode, MeSocksKdfPolicy, MeWriterPickMode};
+use crate::config::{
+    GeneralConfig, MeBindStaleMode, MeRouteNoWriterMode, MeSocksKdfPolicy, MeWriterPickMode,
+};
 use crate::crypto::SecureRandom;
 use crate::network::probe::NetworkDecision;
 use crate::stats::Stats;
@@ -73,6 +76,7 @@ async fn make_pool(me_pool_drain_threshold: u64) -> Arc<MePool> {
         general.me_adaptive_floor_max_warm_writers_global,
         general.hardswap,
         general.me_pool_drain_ttl_secs,
+        general.me_instadrain,
         general.me_pool_drain_threshold,
         general.me_pool_drain_soft_evict_enabled,
         general.me_pool_drain_soft_evict_grace_secs,
@@ -179,15 +183,23 @@ async fn current_writer_ids(pool: &Arc<MePool>) -> Vec<u64> {
 async fn reap_draining_writers_drops_warn_state_for_removed_writer() {
     let pool = make_pool(128).await;
     let now_epoch_secs = MePool::now_epoch_secs();
-    let conn_ids =
+    let conn_ids = insert_draining_writer(
-        insert_draining_writer(&pool, 7, now_epoch_secs.saturating_sub(180), 1, 0).await;
+        &pool,
+        7,
+        now_epoch_secs.saturating_sub(180),
+        1,
+        now_epoch_secs.saturating_add(3_600),
+    )
+    .await;
     let mut warn_next_allowed = HashMap::new();
     let mut soft_evict_next_allowed = HashMap::new();
 
     reap_draining_writers(&pool, &mut warn_next_allowed, &mut soft_evict_next_allowed).await;
     assert!(warn_next_allowed.contains_key(&7));
 
-    let _ = pool.remove_writer_and_close_clients(7).await;
+    let _ = pool
+        .remove_writer_and_close_clients(7, crate::stats::MeWriterTeardownReason::ReapEmpty)
+        .await;
     assert!(pool.registry.get_writer(conn_ids[0]).await.is_none());
 
     reap_draining_writers(&pool, &mut warn_next_allowed, &mut soft_evict_next_allowed).await;
@@ -209,6 +221,89 @@ async fn reap_draining_writers_removes_empty_draining_writers() {
     assert_eq!(current_writer_ids(&pool).await, vec![3]);
 }
 
+#[tokio::test]
+async fn reap_draining_writers_does_not_block_on_stuck_writer_close_signal() {
+    let pool = make_pool(128).await;
+    let now_epoch_secs = MePool::now_epoch_secs();
+
+    let (blocked_tx, blocked_rx) = mpsc::channel::<WriterCommand>(1);
+    assert!(
+        blocked_tx
+            .try_send(WriterCommand::Data(Bytes::from_static(b"stuck")))
+            .is_ok()
+    );
+    let blocked_rx_guard = tokio::spawn(async move {
+        let _hold_rx = blocked_rx;
+        tokio::time::sleep(Duration::from_secs(30)).await;
+    });
+
+    let blocked_writer_id = 90u64;
+    let blocked_writer = MeWriter {
+        id: blocked_writer_id,
+        addr: SocketAddr::new(
+            IpAddr::V4(Ipv4Addr::LOCALHOST),
+            4500 + blocked_writer_id as u16,
+        ),
+        source_ip: IpAddr::V4(Ipv4Addr::LOCALHOST),
+        writer_dc: 2,
+        generation: 1,
+        contour: Arc::new(AtomicU8::new(WriterContour::Draining.as_u8())),
+        created_at: Instant::now() - Duration::from_secs(blocked_writer_id),
+        tx: blocked_tx.clone(),
+        cancel: CancellationToken::new(),
+        degraded: Arc::new(AtomicBool::new(false)),
+        rtt_ema_ms_x10: Arc::new(AtomicU32::new(0)),
+        draining: Arc::new(AtomicBool::new(true)),
+        draining_started_at_epoch_secs: Arc::new(AtomicU64::new(
+            now_epoch_secs.saturating_sub(120),
+        )),
+        drain_deadline_epoch_secs: Arc::new(AtomicU64::new(0)),
+        allow_drain_fallback: Arc::new(AtomicBool::new(false)),
+    };
+    pool.writers.write().await.push(blocked_writer);
+    pool.registry
+        .register_writer(blocked_writer_id, blocked_tx)
+        .await;
+    pool.conn_count.fetch_add(1, Ordering::Relaxed);
+
+    insert_draining_writer(&pool, 91, now_epoch_secs.saturating_sub(110), 0, 0).await;
+
+    let mut warn_next_allowed = HashMap::new();
+    let mut soft_evict_next_allowed = HashMap::new();
+
+    let reap_res = tokio::time::timeout(
+        Duration::from_millis(500),
+        reap_draining_writers(&pool, &mut warn_next_allowed, &mut soft_evict_next_allowed),
+    )
+    .await;
+    blocked_rx_guard.abort();
+
+    assert!(reap_res.is_ok(), "reap should not block on close signal");
+    assert!(current_writer_ids(&pool).await.is_empty());
+    assert_eq!(pool.stats.get_me_writer_close_signal_drop_total(), 2);
+    assert_eq!(pool.stats.get_me_writer_close_signal_channel_full_total(), 1);
+    assert_eq!(pool.stats.get_me_draining_writers_reap_progress_total(), 2);
+    let activity = pool.registry.writer_activity_snapshot().await;
+    assert!(!activity.bound_clients_by_writer.contains_key(&blocked_writer_id));
+    assert!(!activity.bound_clients_by_writer.contains_key(&91));
+    let (probe_conn_id, _rx) = pool.registry.register().await;
+    assert!(
+        !pool.registry
+            .bind_writer(
+                probe_conn_id,
+                blocked_writer_id,
+                ConnMeta {
+                    target_dc: 2,
+                    client_addr: SocketAddr::new(IpAddr::V4(Ipv4Addr::LOCALHOST), 6400),
+                    our_addr: SocketAddr::new(IpAddr::V4(Ipv4Addr::LOCALHOST), 443),
+                    proto_flags: 0,
+                },
+            )
+            .await
+    );
+    let _ = pool.registry.unregister(probe_conn_id).await;
+}
+
 #[tokio::test]
 async fn reap_draining_writers_overflow_closes_oldest_non_empty_writers() {
     let pool = make_pool(2).await;
@@ -247,17 +342,17 @@ async fn reap_draining_writers_deadline_force_close_applies_under_threshold() {
 
 #[tokio::test]
 async fn reap_draining_writers_limits_closes_per_health_tick() {
-    let pool = make_pool(128).await;
+    let pool = make_pool(1).await;
     let now_epoch_secs = MePool::now_epoch_secs();
     let close_budget = health_drain_close_budget();
-    let writer_total = close_budget.saturating_add(19);
+    let writer_total = close_budget.saturating_add(20);
     for writer_id in 1..=writer_total as u64 {
         insert_draining_writer(
             &pool,
             writer_id,
             now_epoch_secs.saturating_sub(20),
             1,
-            now_epoch_secs.saturating_sub(1),
+            0,
         )
         .await;
     }
@@ -280,8 +375,8 @@ async fn reap_draining_writers_backlog_drains_across_ticks() {
             &pool,
             writer_id,
             now_epoch_secs.saturating_sub(20),
-            1,
+            0,
-            now_epoch_secs.saturating_sub(1),
+            0,
         )
         .await;
     }
@@ -309,7 +404,7 @@ async fn reap_draining_writers_threshold_backlog_converges_to_threshold() {
         insert_draining_writer(
             &pool,
             writer_id,
-            now_epoch_secs.saturating_sub(200).saturating_add(writer_id),
+            now_epoch_secs.saturating_sub(20),
             1,
             0,
         )
@@ -345,27 +440,27 @@ async fn reap_draining_writers_threshold_zero_preserves_non_expired_non_empty_wr
 
 #[tokio::test]
 async fn reap_draining_writers_prioritizes_force_close_before_empty_cleanup() {
-    let pool = make_pool(128).await;
+    let pool = make_pool(1).await;
     let now_epoch_secs = MePool::now_epoch_secs();
     let close_budget = health_drain_close_budget();
-    for writer_id in 1..=close_budget as u64 {
+    for writer_id in 1..=close_budget.saturating_add(1) as u64 {
         insert_draining_writer(
             &pool,
             writer_id,
             now_epoch_secs.saturating_sub(20),
             1,
-            now_epoch_secs.saturating_sub(1),
+            0,
         )
         .await;
     }
-    let empty_writer_id = close_budget as u64 + 1;
+    let empty_writer_id = close_budget.saturating_add(2) as u64;
     insert_draining_writer(&pool, empty_writer_id, now_epoch_secs.saturating_sub(20), 0, 0).await;
     let mut warn_next_allowed = HashMap::new();
     let mut soft_evict_next_allowed = HashMap::new();
 
     reap_draining_writers(&pool, &mut warn_next_allowed, &mut soft_evict_next_allowed).await;
 
-    assert_eq!(current_writer_ids(&pool).await, vec![empty_writer_id]);
+    assert_eq!(current_writer_ids(&pool).await, vec![1, empty_writer_id]);
 }
 
 #[tokio::test]
@@ -434,7 +529,12 @@ async fn reap_draining_writers_warn_state_never_exceeds_live_draining_population
 
         let existing_writer_ids = current_writer_ids(&pool).await;
         for writer_id in existing_writer_ids.into_iter().take(4) {
-            let _ = pool.remove_writer_and_close_clients(writer_id).await;
+            let _ = pool
+                .remove_writer_and_close_clients(
+                    writer_id,
+                    crate::stats::MeWriterTeardownReason::ReapEmpty,
+                )
+                .await;
         }
         reap_draining_writers(&pool, &mut warn_next_allowed, &mut soft_evict_next_allowed).await;
         assert!(warn_next_allowed.len() <= pool.writers.read().await.len());
@@ -487,7 +587,14 @@ async fn reap_draining_writers_soft_evicts_stuck_writer_with_per_writer_cap() {
         .store(1, Ordering::Relaxed);
 
     let now_epoch_secs = MePool::now_epoch_secs();
-    insert_draining_writer(&pool, 77, now_epoch_secs.saturating_sub(240), 3, 0).await;
+    insert_draining_writer(
+        &pool,
+        77,
+        now_epoch_secs.saturating_sub(240),
+        3,
+        now_epoch_secs.saturating_add(3_600),
+    )
+    .await;
     let mut warn_next_allowed = HashMap::new();
     let mut soft_evict_next_allowed = HashMap::new();
 
@@ -511,7 +618,14 @@ async fn reap_draining_writers_soft_evict_respects_cooldown_per_writer() {
         .store(60_000, Ordering::Relaxed);
 
     let now_epoch_secs = MePool::now_epoch_secs();
-    insert_draining_writer(&pool, 88, now_epoch_secs.saturating_sub(240), 3, 0).await;
+    insert_draining_writer(
+        &pool,
+        88,
+        now_epoch_secs.saturating_sub(240),
+        3,
+        now_epoch_secs.saturating_add(3_600),
+    )
+    .await;
     let mut warn_next_allowed = HashMap::new();
     let mut soft_evict_next_allowed = HashMap::new();
 
@@ -524,12 +638,40 @@ async fn reap_draining_writers_soft_evict_respects_cooldown_per_writer() {
     assert_eq!(pool.stats.get_pool_drain_soft_evict_writer_total(), 1);
 }
 
+#[tokio::test]
+async fn reap_draining_writers_instadrain_removes_non_expired_writers_immediately() {
+    let pool = make_pool(0).await;
+    pool.me_instadrain.store(true, Ordering::Relaxed);
+    let now_epoch_secs = MePool::now_epoch_secs();
+    insert_draining_writer(&pool, 101, now_epoch_secs.saturating_sub(5), 1, 0).await;
+    insert_draining_writer(&pool, 102, now_epoch_secs.saturating_sub(4), 1, 0).await;
+    let mut warn_next_allowed = HashMap::new();
+    let mut soft_evict_next_allowed = HashMap::new();
+
+    reap_draining_writers(&pool, &mut warn_next_allowed, &mut soft_evict_next_allowed).await;
+
+    assert!(current_writer_ids(&pool).await.is_empty());
+}
+
 #[test]
 fn general_config_default_drain_threshold_remains_enabled() {
-    assert_eq!(GeneralConfig::default().me_pool_drain_threshold, 128);
+    assert_eq!(GeneralConfig::default().me_pool_drain_threshold, 32);
     assert!(GeneralConfig::default().me_pool_drain_soft_evict_enabled);
     assert_eq!(
-        GeneralConfig::default().me_pool_drain_soft_evict_per_writer,
+        GeneralConfig::default().me_pool_drain_soft_evict_grace_secs,
-        1
+        10
     );
+    assert_eq!(
+        GeneralConfig::default().me_pool_drain_soft_evict_per_writer,
+        2
+    );
+    assert_eq!(
+        GeneralConfig::default().me_pool_drain_soft_evict_budget_per_core,
+        16
+    );
+    assert_eq!(
+        GeneralConfig::default().me_pool_drain_soft_evict_cooldown_ms,
+        1000
+    );
+    assert_eq!(GeneralConfig::default().me_bind_stale_mode, MeBindStaleMode::Never);
 }

src/transport/middle_proxy/mod.rs

@@ -30,7 +30,7 @@ mod health_adversarial_tests;
 
 use bytes::Bytes;
 
-pub use health::me_health_monitor;
+pub use health::{me_drain_timeout_enforcer, me_health_monitor, me_zombie_writer_watchdog};
 #[allow(unused_imports)]
 pub use ping::{run_me_ping, format_sample_line, format_me_route, MePingReport, MePingSample, MePingFamily};
 pub use pool::MePool;

src/transport/middle_proxy/pool.rs

@@ -18,6 +18,8 @@ use crate::transport::UpstreamManager;
 use super::ConnRegistry;
 use super::codec::WriterCommand;
 
+const ME_FORCE_CLOSE_SAFETY_FALLBACK_SECS: u64 = 300;
+
 #[derive(Debug, Clone, Copy, PartialEq, Eq, Hash)]
 pub(super) struct RefillDcKey {
     pub dc: i32,
@@ -72,6 +74,64 @@ impl WriterContour {
     }
 }
 
+#[derive(Debug, Clone, Copy, PartialEq, Eq)]
+#[repr(u8)]
+pub(crate) enum MeFamilyRuntimeState {
+    Healthy = 0,
+    Degraded = 1,
+    Suppressed = 2,
+    Recovering = 3,
+}
+
+impl MeFamilyRuntimeState {
+    pub(crate) fn from_u8(value: u8) -> Self {
+        match value {
+            1 => Self::Degraded,
+            2 => Self::Suppressed,
+            3 => Self::Recovering,
+            _ => Self::Healthy,
+        }
+    }
+
+    pub(crate) fn as_str(self) -> &'static str {
+        match self {
+            Self::Healthy => "healthy",
+            Self::Degraded => "degraded",
+            Self::Suppressed => "suppressed",
+            Self::Recovering => "recovering",
+        }
+    }
+}
+
+#[derive(Debug, Clone, Copy, PartialEq, Eq)]
+#[repr(u8)]
+pub(crate) enum MeDrainGateReason {
+    Open = 0,
+    CoverageQuorum = 1,
+    Redundancy = 2,
+    SuppressionActive = 3,
+}
+
+impl MeDrainGateReason {
+    pub(crate) fn from_u8(value: u8) -> Self {
+        match value {
+            1 => Self::CoverageQuorum,
+            2 => Self::Redundancy,
+            3 => Self::SuppressionActive,
+            _ => Self::Open,
+        }
+    }
+
+    pub(crate) fn as_str(self) -> &'static str {
+        match self {
+            Self::Open => "open",
+            Self::CoverageQuorum => "coverage_quorum",
+            Self::Redundancy => "redundancy",
+            Self::SuppressionActive => "suppression_active",
+        }
+    }
+}
+
 #[derive(Debug, Clone)]
 pub struct SecretSnapshot {
     pub epoch: u64,
@@ -171,6 +231,7 @@ pub struct MePool {
     pub(super) endpoint_quarantine: Arc<Mutex<HashMap<SocketAddr, Instant>>>,
     pub(super) kdf_material_fingerprint: Arc<RwLock<HashMap<SocketAddr, (u64, u16)>>>,
     pub(super) me_pool_drain_ttl_secs: AtomicU64,
+    pub(super) me_instadrain: AtomicBool,
     pub(super) me_pool_drain_threshold: AtomicU64,
     pub(super) me_pool_drain_soft_evict_enabled: AtomicBool,
     pub(super) me_pool_drain_soft_evict_grace_secs: AtomicU64,
@@ -200,6 +261,20 @@ pub struct MePool {
     pub(super) me_health_interval_ms_unhealthy: AtomicU64,
     pub(super) me_health_interval_ms_healthy: AtomicU64,
     pub(super) me_warn_rate_limit_ms: AtomicU64,
+    pub(super) me_family_v4_runtime_state: AtomicU8,
+    pub(super) me_family_v6_runtime_state: AtomicU8,
+    pub(super) me_family_v4_state_since_epoch_secs: AtomicU64,
+    pub(super) me_family_v6_state_since_epoch_secs: AtomicU64,
+    pub(super) me_family_v4_suppressed_until_epoch_secs: AtomicU64,
+    pub(super) me_family_v6_suppressed_until_epoch_secs: AtomicU64,
+    pub(super) me_family_v4_fail_streak: AtomicU32,
+    pub(super) me_family_v6_fail_streak: AtomicU32,
+    pub(super) me_family_v4_recover_success_streak: AtomicU32,
+    pub(super) me_family_v6_recover_success_streak: AtomicU32,
+    pub(super) me_last_drain_gate_route_quorum_ok: AtomicBool,
+    pub(super) me_last_drain_gate_redundancy_ok: AtomicBool,
+    pub(super) me_last_drain_gate_block_reason: AtomicU8,
+    pub(super) me_last_drain_gate_updated_at_epoch_secs: AtomicU64,
     pub(super) runtime_ready: AtomicBool,
     pool_size: usize,
     pub(super) preferred_endpoints_by_dc: Arc<RwLock<HashMap<i32, Vec<SocketAddr>>>>,
@@ -228,6 +303,14 @@ impl MePool {
             .as_secs()
     }
 
+    fn normalize_force_close_secs(force_close_secs: u64) -> u64 {
+        if force_close_secs == 0 {
+            ME_FORCE_CLOSE_SAFETY_FALLBACK_SECS
+        } else {
+            force_close_secs
+        }
+    }
+
     pub fn new(
         proxy_tag: Option<Vec<u8>>,
         proxy_secret: Vec<u8>,
@@ -279,6 +362,7 @@ impl MePool {
         me_adaptive_floor_max_warm_writers_global: u32,
         hardswap: bool,
         me_pool_drain_ttl_secs: u64,
+        me_instadrain: bool,
         me_pool_drain_threshold: u64,
         me_pool_drain_soft_evict_enabled: bool,
         me_pool_drain_soft_evict_grace_secs: u64,
@@ -462,6 +546,7 @@ impl MePool {
             endpoint_quarantine: Arc::new(Mutex::new(HashMap::new())),
             kdf_material_fingerprint: Arc::new(RwLock::new(HashMap::new())),
             me_pool_drain_ttl_secs: AtomicU64::new(me_pool_drain_ttl_secs),
+            me_instadrain: AtomicBool::new(me_instadrain),
             me_pool_drain_threshold: AtomicU64::new(me_pool_drain_threshold),
             me_pool_drain_soft_evict_enabled: AtomicBool::new(me_pool_drain_soft_evict_enabled),
             me_pool_drain_soft_evict_grace_secs: AtomicU64::new(me_pool_drain_soft_evict_grace_secs),
@@ -474,7 +559,9 @@ impl MePool {
             me_pool_drain_soft_evict_cooldown_ms: AtomicU64::new(
                 me_pool_drain_soft_evict_cooldown_ms.max(1),
             ),
-            me_pool_force_close_secs: AtomicU64::new(me_pool_force_close_secs),
+            me_pool_force_close_secs: AtomicU64::new(Self::normalize_force_close_secs(
+                me_pool_force_close_secs,
+            )),
             me_pool_min_fresh_ratio_permille: AtomicU32::new(Self::ratio_to_permille(
                 me_pool_min_fresh_ratio,
             )),
@@ -503,6 +590,20 @@ impl MePool {
             me_health_interval_ms_unhealthy: AtomicU64::new(me_health_interval_ms_unhealthy.max(1)),
             me_health_interval_ms_healthy: AtomicU64::new(me_health_interval_ms_healthy.max(1)),
             me_warn_rate_limit_ms: AtomicU64::new(me_warn_rate_limit_ms.max(1)),
+            me_family_v4_runtime_state: AtomicU8::new(MeFamilyRuntimeState::Healthy as u8),
+            me_family_v6_runtime_state: AtomicU8::new(MeFamilyRuntimeState::Healthy as u8),
+            me_family_v4_state_since_epoch_secs: AtomicU64::new(Self::now_epoch_secs()),
+            me_family_v6_state_since_epoch_secs: AtomicU64::new(Self::now_epoch_secs()),
+            me_family_v4_suppressed_until_epoch_secs: AtomicU64::new(0),
+            me_family_v6_suppressed_until_epoch_secs: AtomicU64::new(0),
+            me_family_v4_fail_streak: AtomicU32::new(0),
+            me_family_v6_fail_streak: AtomicU32::new(0),
+            me_family_v4_recover_success_streak: AtomicU32::new(0),
+            me_family_v6_recover_success_streak: AtomicU32::new(0),
+            me_last_drain_gate_route_quorum_ok: AtomicBool::new(false),
+            me_last_drain_gate_redundancy_ok: AtomicBool::new(false),
+            me_last_drain_gate_block_reason: AtomicU8::new(MeDrainGateReason::Open as u8),
+            me_last_drain_gate_updated_at_epoch_secs: AtomicU64::new(Self::now_epoch_secs()),
             runtime_ready: AtomicBool::new(false),
             preferred_endpoints_by_dc: Arc::new(RwLock::new(preferred_endpoints_by_dc)),
         })
@@ -520,10 +621,158 @@ impl MePool {
         self.runtime_ready.load(Ordering::Relaxed)
     }
 
+    pub(super) fn set_family_runtime_state(
+        &self,
+        family: IpFamily,
+        state: MeFamilyRuntimeState,
+        state_since_epoch_secs: u64,
+        suppressed_until_epoch_secs: u64,
+        fail_streak: u32,
+        recover_success_streak: u32,
+    ) {
+        match family {
+            IpFamily::V4 => {
+                self.me_family_v4_runtime_state
+                    .store(state as u8, Ordering::Relaxed);
+                self.me_family_v4_state_since_epoch_secs
+                    .store(state_since_epoch_secs, Ordering::Relaxed);
+                self.me_family_v4_suppressed_until_epoch_secs
+                    .store(suppressed_until_epoch_secs, Ordering::Relaxed);
+                self.me_family_v4_fail_streak
+                    .store(fail_streak, Ordering::Relaxed);
+                self.me_family_v4_recover_success_streak
+                    .store(recover_success_streak, Ordering::Relaxed);
+            }
+            IpFamily::V6 => {
+                self.me_family_v6_runtime_state
+                    .store(state as u8, Ordering::Relaxed);
+                self.me_family_v6_state_since_epoch_secs
+                    .store(state_since_epoch_secs, Ordering::Relaxed);
+                self.me_family_v6_suppressed_until_epoch_secs
+                    .store(suppressed_until_epoch_secs, Ordering::Relaxed);
+                self.me_family_v6_fail_streak
+                    .store(fail_streak, Ordering::Relaxed);
+                self.me_family_v6_recover_success_streak
+                    .store(recover_success_streak, Ordering::Relaxed);
+            }
+        }
+    }
+
+    pub(crate) fn family_runtime_state(&self, family: IpFamily) -> MeFamilyRuntimeState {
+        match family {
+            IpFamily::V4 => MeFamilyRuntimeState::from_u8(
+                self.me_family_v4_runtime_state.load(Ordering::Relaxed),
+            ),
+            IpFamily::V6 => MeFamilyRuntimeState::from_u8(
+                self.me_family_v6_runtime_state.load(Ordering::Relaxed),
+            ),
+        }
+    }
+
+    pub(crate) fn family_runtime_state_since_epoch_secs(&self, family: IpFamily) -> u64 {
+        match family {
+            IpFamily::V4 => self
+                .me_family_v4_state_since_epoch_secs
+                .load(Ordering::Relaxed),
+            IpFamily::V6 => self
+                .me_family_v6_state_since_epoch_secs
+                .load(Ordering::Relaxed),
+        }
+    }
+
+    pub(crate) fn family_suppressed_until_epoch_secs(&self, family: IpFamily) -> u64 {
+        match family {
+            IpFamily::V4 => self
+                .me_family_v4_suppressed_until_epoch_secs
+                .load(Ordering::Relaxed),
+            IpFamily::V6 => self
+                .me_family_v6_suppressed_until_epoch_secs
+                .load(Ordering::Relaxed),
+        }
+    }
+
+    pub(crate) fn family_fail_streak(&self, family: IpFamily) -> u32 {
+        match family {
+            IpFamily::V4 => self.me_family_v4_fail_streak.load(Ordering::Relaxed),
+            IpFamily::V6 => self.me_family_v6_fail_streak.load(Ordering::Relaxed),
+        }
+    }
+
+    pub(crate) fn family_recover_success_streak(&self, family: IpFamily) -> u32 {
+        match family {
+            IpFamily::V4 => self
+                .me_family_v4_recover_success_streak
+                .load(Ordering::Relaxed),
+            IpFamily::V6 => self
+                .me_family_v6_recover_success_streak
+                .load(Ordering::Relaxed),
+        }
+    }
+
+    pub(crate) fn is_family_temporarily_suppressed(
+        &self,
+        family: IpFamily,
+        now_epoch_secs: u64,
+    ) -> bool {
+        self.family_suppressed_until_epoch_secs(family) > now_epoch_secs
+    }
+
+    pub(super) fn family_enabled_for_drain_coverage(
+        &self,
+        family: IpFamily,
+        now_epoch_secs: u64,
+    ) -> bool {
+        let configured = match family {
+            IpFamily::V4 => self.decision.ipv4_me,
+            IpFamily::V6 => self.decision.ipv6_me,
+        };
+        configured && !self.is_family_temporarily_suppressed(family, now_epoch_secs)
+    }
+
+    pub(super) fn set_last_drain_gate(
+        &self,
+        route_quorum_ok: bool,
+        redundancy_ok: bool,
+        block_reason: MeDrainGateReason,
+        updated_at_epoch_secs: u64,
+    ) {
+        self.me_last_drain_gate_route_quorum_ok
+            .store(route_quorum_ok, Ordering::Relaxed);
+        self.me_last_drain_gate_redundancy_ok
+            .store(redundancy_ok, Ordering::Relaxed);
+        self.me_last_drain_gate_block_reason
+            .store(block_reason as u8, Ordering::Relaxed);
+        self.me_last_drain_gate_updated_at_epoch_secs
+            .store(updated_at_epoch_secs, Ordering::Relaxed);
+    }
+
+    pub(crate) fn last_drain_gate_route_quorum_ok(&self) -> bool {
+        self.me_last_drain_gate_route_quorum_ok
+            .load(Ordering::Relaxed)
+    }
+
+    pub(crate) fn last_drain_gate_redundancy_ok(&self) -> bool {
+        self.me_last_drain_gate_redundancy_ok
+            .load(Ordering::Relaxed)
+    }
+
+    pub(crate) fn last_drain_gate_block_reason(&self) -> MeDrainGateReason {
+        MeDrainGateReason::from_u8(
+            self.me_last_drain_gate_block_reason
+                .load(Ordering::Relaxed),
+        )
+    }
+
+    pub(crate) fn last_drain_gate_updated_at_epoch_secs(&self) -> u64 {
+        self.me_last_drain_gate_updated_at_epoch_secs
+            .load(Ordering::Relaxed)
+    }
+
     pub fn update_runtime_reinit_policy(
         &self,
         hardswap: bool,
         drain_ttl_secs: u64,
+        instadrain: bool,
         pool_drain_threshold: u64,
         pool_drain_soft_evict_enabled: bool,
         pool_drain_soft_evict_grace_secs: u64,
@@ -568,6 +817,7 @@ impl MePool {
         self.hardswap.store(hardswap, Ordering::Relaxed);
         self.me_pool_drain_ttl_secs
             .store(drain_ttl_secs, Ordering::Relaxed);
+        self.me_instadrain.store(instadrain, Ordering::Relaxed);
         self.me_pool_drain_threshold
             .store(pool_drain_threshold, Ordering::Relaxed);
         self.me_pool_drain_soft_evict_enabled
@@ -582,8 +832,10 @@ impl MePool {
         );
         self.me_pool_drain_soft_evict_cooldown_ms
             .store(pool_drain_soft_evict_cooldown_ms.max(1), Ordering::Relaxed);
-        self.me_pool_force_close_secs
+        self.me_pool_force_close_secs.store(
-            .store(force_close_secs, Ordering::Relaxed);
+            Self::normalize_force_close_secs(force_close_secs),
+            Ordering::Relaxed,
+        );
         self.me_pool_min_fresh_ratio_permille
             .store(Self::ratio_to_permille(min_fresh_ratio), Ordering::Relaxed);
         self.me_hardswap_warmup_delay_min_ms
@@ -728,12 +980,9 @@ impl MePool {
     }
 
     pub(super) fn force_close_timeout(&self) -> Option<Duration> {
-        let secs = self.me_pool_force_close_secs.load(Ordering::Relaxed);
+        let secs =
-        if secs == 0 {
+            Self::normalize_force_close_secs(self.me_pool_force_close_secs.load(Ordering::Relaxed));
-            None
+        Some(Duration::from_secs(secs))
-        } else {
-            Some(Duration::from_secs(secs))
-        }
     }
 
     pub(super) fn drain_soft_evict_enabled(&self) -> bool {
@@ -1005,9 +1254,10 @@ impl MePool {
     }
 
     pub(super) async fn active_coverage_required_total(&self) -> usize {
+        let now_epoch_secs = Self::now_epoch_secs();
         let mut endpoints_by_dc = HashMap::<i32, HashSet<SocketAddr>>::new();
 
-        if self.decision.ipv4_me {
+        if self.family_enabled_for_drain_coverage(IpFamily::V4, now_epoch_secs) {
             let map = self.proxy_map_v4.read().await;
             for (dc, addrs) in map.iter() {
                 let entry = endpoints_by_dc.entry(*dc).or_default();
@@ -1017,7 +1267,7 @@ impl MePool {
             }
         }
 
-        if self.decision.ipv6_me {
+        if self.family_enabled_for_drain_coverage(IpFamily::V6, now_epoch_secs) {
             let map = self.proxy_map_v6.read().await;
             for (dc, addrs) in map.iter() {
                 let entry = endpoints_by_dc.entry(*dc).or_default();

src/transport/middle_proxy/pool_refill.rs

@@ -74,9 +74,8 @@ impl MePool {
             debug!(
                 %addr,
                 wait_ms = expiry.saturating_duration_since(now).as_millis(),
-                "All ME endpoints are quarantined for the DC group; retrying earliest one"
+                "All ME endpoints are quarantined for the DC group; waiting for quarantine expiry"
             );
-            return vec![addr];
         }
 
         Vec::new()
@@ -165,9 +164,10 @@ impl MePool {
     }
 
     async fn endpoints_for_dc(&self, target_dc: i32) -> Vec<SocketAddr> {
+        let now_epoch_secs = Self::now_epoch_secs();
         let mut endpoints = HashSet::<SocketAddr>::new();
 
-        if self.decision.ipv4_me {
+        if self.family_enabled_for_drain_coverage(IpFamily::V4, now_epoch_secs) {
             let map = self.proxy_map_v4.read().await;
             if let Some(addrs) = map.get(&target_dc) {
                 for (ip, port) in addrs {
@@ -176,7 +176,7 @@ impl MePool {
             }
         }
 
-        if self.decision.ipv6_me {
+        if self.family_enabled_for_drain_coverage(IpFamily::V6, now_epoch_secs) {
             let map = self.proxy_map_v6.read().await;
             if let Some(addrs) = map.get(&target_dc) {
                 for (ip, port) in addrs {

src/transport/middle_proxy/pool_reinit.rs

@@ -11,8 +11,9 @@ use tracing::{debug, info, warn};
 use std::collections::hash_map::DefaultHasher;
 
 use crate::crypto::SecureRandom;
+use crate::network::IpFamily;
 
-use super::pool::{MePool, WriterContour};
+use super::pool::{MeDrainGateReason, MePool, WriterContour};
 
 const ME_HARDSWAP_PENDING_TTL_SECS: u64 = 1800;
 
@@ -120,9 +121,10 @@ impl MePool {
     }
 
     async fn desired_dc_endpoints(&self) -> HashMap<i32, HashSet<SocketAddr>> {
+        let now_epoch_secs = Self::now_epoch_secs();
         let mut out: HashMap<i32, HashSet<SocketAddr>> = HashMap::new();
 
-        if self.decision.ipv4_me {
+        if self.family_enabled_for_drain_coverage(IpFamily::V4, now_epoch_secs) {
             let map_v4 = self.proxy_map_v4.read().await.clone();
             for (dc, addrs) in map_v4 {
                 let entry = out.entry(dc).or_default();
@@ -132,7 +134,7 @@ impl MePool {
             }
         }
 
-        if self.decision.ipv6_me {
+        if self.family_enabled_for_drain_coverage(IpFamily::V6, now_epoch_secs) {
             let map_v6 = self.proxy_map_v6.read().await.clone();
             for (dc, addrs) in map_v6 {
                 let entry = out.entry(dc).or_default();
@@ -313,13 +315,23 @@ impl MePool {
 
     pub async fn zero_downtime_reinit_after_map_change(self: &Arc<Self>, rng: &SecureRandom) {
         let desired_by_dc = self.desired_dc_endpoints().await;
+        let now_epoch_secs = Self::now_epoch_secs();
+        let v4_suppressed = self.is_family_temporarily_suppressed(IpFamily::V4, now_epoch_secs);
+        let v6_suppressed = self.is_family_temporarily_suppressed(IpFamily::V6, now_epoch_secs);
         if desired_by_dc.is_empty() {
             warn!("ME endpoint map is empty; skipping stale writer drain");
+            let reason = if (self.decision.ipv4_me && v4_suppressed)
+                || (self.decision.ipv6_me && v6_suppressed)
+            {
+                MeDrainGateReason::SuppressionActive
+            } else {
+                MeDrainGateReason::CoverageQuorum
+            };
+            self.set_last_drain_gate(false, false, reason, now_epoch_secs);
             return;
         }
 
         let desired_map_hash = Self::desired_map_hash(&desired_by_dc);
-        let now_epoch_secs = Self::now_epoch_secs();
         let previous_generation = self.current_generation();
         let hardswap = self.hardswap.load(Ordering::Relaxed);
         let generation = if hardswap {
@@ -390,7 +402,17 @@ impl MePool {
                 .load(Ordering::Relaxed),
         );
         let (coverage_ratio, missing_dc) = Self::coverage_ratio(&desired_by_dc, &active_writer_addrs);
+        let mut route_quorum_ok = coverage_ratio >= min_ratio;
+        let mut redundancy_ok = missing_dc.is_empty();
+        let mut redundancy_missing_dc = missing_dc.clone();
+        let mut gate_coverage_ratio = coverage_ratio;
         if !hardswap && coverage_ratio < min_ratio {
+            self.set_last_drain_gate(
+                false,
+                redundancy_ok,
+                MeDrainGateReason::CoverageQuorum,
+                now_epoch_secs,
+            );
             warn!(
                 previous_generation,
                 generation,
@@ -411,7 +433,17 @@ impl MePool {
                 .collect();
             let (fresh_coverage_ratio, fresh_missing_dc) =
                 Self::coverage_ratio(&desired_by_dc, &fresh_writer_addrs);
-            if !fresh_missing_dc.is_empty() {
+            route_quorum_ok = fresh_coverage_ratio >= min_ratio;
+            redundancy_ok = fresh_missing_dc.is_empty();
+            redundancy_missing_dc = fresh_missing_dc.clone();
+            gate_coverage_ratio = fresh_coverage_ratio;
+            if fresh_coverage_ratio < min_ratio {
+                self.set_last_drain_gate(
+                    false,
+                    redundancy_ok,
+                    MeDrainGateReason::CoverageQuorum,
+                    now_epoch_secs,
+                );
                 warn!(
                     previous_generation,
                     generation,
@@ -421,13 +453,16 @@ impl MePool {
                 );
                 return;
             }
-        } else if !missing_dc.is_empty() {
+        }
+
+        self.set_last_drain_gate(route_quorum_ok, redundancy_ok, MeDrainGateReason::Open, now_epoch_secs);
+        if !redundancy_ok {
             warn!(
-                missing_dc = ?missing_dc,
+                missing_dc = ?redundancy_missing_dc,
-                // Keep stale writers alive when fresh coverage is incomplete.
+                coverage_ratio = format_args!("{gate_coverage_ratio:.3}"),
-                "ME reinit coverage incomplete; keeping stale writers"
+                min_ratio = format_args!("{min_ratio:.3}"),
+                "ME reinit proceeds with weighted quorum while some DC groups remain uncovered"
             );
-            return;
         }
 
         if hardswap {

src/transport/middle_proxy/pool_runtime_api.rs

@@ -1,7 +1,7 @@
 use std::collections::HashMap;
 use std::time::Instant;
 
-use super::pool::{MePool, RefillDcKey};
+use super::pool::{MeDrainGateReason, MePool, RefillDcKey};
 use crate::network::IpFamily;
 
 #[derive(Clone, Debug)]
@@ -36,6 +36,24 @@ pub(crate) struct MeApiNatStunSnapshot {
     pub stun_backoff_remaining_ms: Option<u64>,
 }
 
+#[derive(Clone, Debug)]
+pub(crate) struct MeApiFamilyStateSnapshot {
+    pub family: &'static str,
+    pub state: &'static str,
+    pub state_since_epoch_secs: u64,
+    pub suppressed_until_epoch_secs: Option<u64>,
+    pub fail_streak: u32,
+    pub recover_success_streak: u32,
+}
+
+#[derive(Clone, Debug)]
+pub(crate) struct MeApiDrainGateSnapshot {
+    pub route_quorum_ok: bool,
+    pub redundancy_ok: bool,
+    pub block_reason: &'static str,
+    pub updated_at_epoch_secs: u64,
+}
+
 impl MePool {
     pub(crate) async fn api_refill_snapshot(&self) -> MeApiRefillSnapshot {
         let inflight_endpoints_total = self.refill_inflight.lock().await.len();
@@ -125,4 +143,35 @@ impl MePool {
             stun_backoff_remaining_ms,
         }
     }
+
+    pub(crate) fn api_family_state_snapshot(&self) -> Vec<MeApiFamilyStateSnapshot> {
+        [IpFamily::V4, IpFamily::V6]
+            .into_iter()
+            .map(|family| {
+                let state = self.family_runtime_state(family);
+                let suppressed_until = self.family_suppressed_until_epoch_secs(family);
+                MeApiFamilyStateSnapshot {
+                    family: match family {
+                        IpFamily::V4 => "v4",
+                        IpFamily::V6 => "v6",
+                    },
+                    state: state.as_str(),
+                    state_since_epoch_secs: self.family_runtime_state_since_epoch_secs(family),
+                    suppressed_until_epoch_secs: (suppressed_until != 0).then_some(suppressed_until),
+                    fail_streak: self.family_fail_streak(family),
+                    recover_success_streak: self.family_recover_success_streak(family),
+                }
+            })
+            .collect()
+    }
+
+    pub(crate) fn api_drain_gate_snapshot(&self) -> MeApiDrainGateSnapshot {
+        let reason: MeDrainGateReason = self.last_drain_gate_block_reason();
+        MeApiDrainGateSnapshot {
+            route_quorum_ok: self.last_drain_gate_route_quorum_ok(),
+            redundancy_ok: self.last_drain_gate_redundancy_ok(),
+            block_reason: reason.as_str(),
+            updated_at_epoch_secs: self.last_drain_gate_updated_at_epoch_secs(),
+        }
+    }
 }

src/transport/middle_proxy/pool_status.rs

@@ -126,6 +126,7 @@ pub(crate) struct MeApiRuntimeSnapshot {
     pub me_reconnect_backoff_cap_ms: u64,
     pub me_reconnect_fast_retry_count: u32,
     pub me_pool_drain_ttl_secs: u64,
+    pub me_instadrain: bool,
     pub me_pool_drain_soft_evict_enabled: bool,
     pub me_pool_drain_soft_evict_grace_secs: u64,
     pub me_pool_drain_soft_evict_per_writer: u8,
@@ -583,6 +584,7 @@ impl MePool {
             me_reconnect_backoff_cap_ms: self.me_reconnect_backoff_cap.as_millis() as u64,
             me_reconnect_fast_retry_count: self.me_reconnect_fast_retry_count,
             me_pool_drain_ttl_secs: self.me_pool_drain_ttl_secs.load(Ordering::Relaxed),
+            me_instadrain: self.me_instadrain.load(Ordering::Relaxed),
             me_pool_drain_soft_evict_enabled: self
                 .me_pool_drain_soft_evict_enabled
                 .load(Ordering::Relaxed),

src/transport/middle_proxy/pool_writer.rs

@@ -8,6 +8,7 @@ use bytes::Bytes;
 use bytes::BytesMut;
 use rand::Rng;
 use tokio::sync::mpsc;
+use tokio::sync::mpsc::error::TrySendError;
 use tokio_util::sync::CancellationToken;
 use tracing::{debug, info, warn};
 
@@ -15,11 +16,13 @@ use crate::config::MeBindStaleMode;
 use crate::crypto::SecureRandom;
 use crate::error::{ProxyError, Result};
 use crate::protocol::constants::{RPC_CLOSE_EXT_U32, RPC_PING_U32};
+use crate::stats::{
+    MeWriterCleanupSideEffectStep, MeWriterTeardownMode, MeWriterTeardownReason,
+};
 
 use super::codec::{RpcWriter, WriterCommand};
 use super::pool::{MePool, MeWriter, WriterContour};
 use super::reader::reader_loop;
-use super::registry::BoundConn;
 use super::wire::build_proxy_req_payload;
 
 const ME_ACTIVE_PING_SECS: u64 = 25;
@@ -27,6 +30,12 @@ const ME_ACTIVE_PING_JITTER_SECS: i64 = 5;
 const ME_IDLE_KEEPALIVE_MAX_SECS: u64 = 5;
 const ME_RPC_PROXY_REQ_RESPONSE_WAIT_MS: u64 = 700;
 
+#[derive(Clone, Copy)]
+enum WriterRemoveGuardMode {
+    Any,
+    DrainingOnly,
+}
+
 fn is_me_peer_closed_error(error: &ProxyError) -> bool {
     matches!(error, ProxyError::Io(ioe) if ioe.kind() == ErrorKind::UnexpectedEof)
 }
@@ -43,9 +52,16 @@ impl MePool {
 
         for writer_id in closed_writer_ids {
             if self.registry.is_writer_empty(writer_id).await {
-                let _ = self.remove_writer_only(writer_id).await;
+                let _ = self
+                    .remove_writer_only(writer_id, MeWriterTeardownReason::PruneClosedWriter)
+                    .await;
             } else {
-                let _ = self.remove_writer_and_close_clients(writer_id).await;
+                let _ = self
+                    .remove_writer_and_close_clients(
+                        writer_id,
+                        MeWriterTeardownReason::PruneClosedWriter,
+                    )
+                    .await;
             }
         }
     }
@@ -142,6 +158,9 @@ impl MePool {
             crc_mode: hs.crc_mode,
         };
         let cancel_wr = cancel.clone();
+        let cleanup_done = Arc::new(AtomicBool::new(false));
+        let cleanup_for_writer = cleanup_done.clone();
+        let pool_writer_task = Arc::downgrade(self);
         tokio::spawn(async move {
             loop {
                 tokio::select! {
@@ -159,6 +178,20 @@ impl MePool {
                     _ = cancel_wr.cancelled() => break,
                 }
             }
+            if cleanup_for_writer
+                .compare_exchange(false, true, Ordering::AcqRel, Ordering::Relaxed)
+                .is_ok()
+            {
+                if let Some(pool) = pool_writer_task.upgrade() {
+                    pool.remove_writer_and_close_clients(
+                        writer_id,
+                        MeWriterTeardownReason::WriterTaskExit,
+                    )
+                    .await;
+                } else {
+                    cancel_wr.cancel();
+                }
+            }
         });
         let writer = MeWriter {
             id: writer_id,
@@ -195,7 +228,6 @@ impl MePool {
         let cancel_ping = cancel.clone();
         let tx_ping = tx.clone();
         let ping_tracker_ping = ping_tracker.clone();
-        let cleanup_done = Arc::new(AtomicBool::new(false));
         let cleanup_for_reader = cleanup_done.clone();
         let cleanup_for_ping = cleanup_done.clone();
         let keepalive_enabled = self.me_keepalive_enabled;
@@ -241,21 +273,29 @@ impl MePool {
                 stats_reader_close.increment_me_idle_close_by_peer_total();
                 info!(writer_id, "ME socket closed by peer on idle writer");
             }
-            if let Some(pool) = pool.upgrade()
+            if cleanup_for_reader
-                && cleanup_for_reader
+                .compare_exchange(false, true, Ordering::AcqRel, Ordering::Relaxed)
-                    .compare_exchange(false, true, Ordering::AcqRel, Ordering::Relaxed)
+                .is_ok()
-                    .is_ok()
             {
-                pool.remove_writer_and_close_clients(writer_id).await;
+                if let Some(pool) = pool.upgrade() {
+                    pool.remove_writer_and_close_clients(
+                        writer_id,
+                        MeWriterTeardownReason::ReaderExit,
+                    )
+                    .await;
+                } else {
+                    // Fallback for shutdown races: make writer task exit quickly so stale
+                    // channels are observable by periodic prune.
+                    cancel_reader_token.cancel();
+                }
             }
             if let Err(e) = res {
                 if !idle_close_by_peer {
                     warn!(error = %e, "ME reader ended");
                 }
             }
-            let mut ws = writers_arc.write().await;
+            let remaining = writers_arc.read().await.len();
-            ws.retain(|w| w.id != writer_id);
+            debug!(writer_id, remaining, "ME reader task finished");
-            info!(remaining = ws.len(), "Dead ME writer removed from pool");
         });
 
         let pool_ping = Arc::downgrade(self);
@@ -350,7 +390,11 @@ impl MePool {
                             .compare_exchange(false, true, Ordering::AcqRel, Ordering::Relaxed)
                             .is_ok()
                     {
-                        pool.remove_writer_and_close_clients(writer_id).await;
+                        pool.remove_writer_and_close_clients(
+                            writer_id,
+                            MeWriterTeardownReason::PingSendFail,
+                        )
+                        .await;
                     }
                     break;
                 }
@@ -443,7 +487,11 @@ impl MePool {
                         .compare_exchange(false, true, Ordering::AcqRel, Ordering::Relaxed)
                         .is_ok()
                     {
-                        pool.remove_writer_and_close_clients(writer_id).await;
+                        pool.remove_writer_and_close_clients(
+                            writer_id,
+                            MeWriterTeardownReason::SignalSendFail,
+                        )
+                        .await;
                     }
                     break;
                 }
@@ -477,7 +525,11 @@ impl MePool {
                         .compare_exchange(false, true, Ordering::AcqRel, Ordering::Relaxed)
                         .is_ok()
                     {
-                        pool.remove_writer_and_close_clients(writer_id).await;
+                        pool.remove_writer_and_close_clients(
+                            writer_id,
+                            MeWriterTeardownReason::SignalSendFail,
+                        )
+                        .await;
                     }
                     break;
                 }
@@ -490,23 +542,83 @@ impl MePool {
         Ok(())
     }
 
-    pub(crate) async fn remove_writer_and_close_clients(self: &Arc<Self>, writer_id: u64) {
+    pub(crate) async fn remove_writer_and_close_clients(
-        let conns = self.remove_writer_only(writer_id).await;
+        self: &Arc<Self>,
-        for bound in conns {
+        writer_id: u64,
-            let _ = self.registry.route(bound.conn_id, super::MeResponse::Close).await;
+        reason: MeWriterTeardownReason,
-            let _ = self.registry.unregister(bound.conn_id).await;
+    ) -> bool {
-        }
+        // Full client cleanup now happens inside `registry.writer_lost` to keep
+        // writer reap/remove paths strictly non-blocking per connection.
+        self.remove_writer_with_mode(
+            writer_id,
+            reason,
+            MeWriterTeardownMode::Normal,
+            WriterRemoveGuardMode::Any,
+        )
+        .await
     }
 
-    async fn remove_writer_only(self: &Arc<Self>, writer_id: u64) -> Vec<BoundConn> {
+    pub(super) async fn remove_draining_writer_hard_detach(
+        self: &Arc<Self>,
+        writer_id: u64,
+        reason: MeWriterTeardownReason,
+    ) -> bool {
+        self.remove_writer_with_mode(
+            writer_id,
+            reason,
+            MeWriterTeardownMode::HardDetach,
+            WriterRemoveGuardMode::DrainingOnly,
+        )
+        .await
+    }
+
+    async fn remove_writer_only(
+        self: &Arc<Self>,
+        writer_id: u64,
+        reason: MeWriterTeardownReason,
+    ) -> bool {
+        self.remove_writer_with_mode(
+            writer_id,
+            reason,
+            MeWriterTeardownMode::Normal,
+            WriterRemoveGuardMode::Any,
+        )
+        .await
+    }
+
+    // Authoritative teardown primitive shared by normal cleanup and watchdog path.
+    // Lock-order invariant:
+    // 1) mutate `writers` under pool write lock,
+    // 2) release pool lock,
+    // 3) run registry/metrics/refill side effects.
+    // `registry.writer_lost` must never run while `writers` lock is held.
+    async fn remove_writer_with_mode(
+        self: &Arc<Self>,
+        writer_id: u64,
+        reason: MeWriterTeardownReason,
+        mode: MeWriterTeardownMode,
+        guard_mode: WriterRemoveGuardMode,
+    ) -> bool {
+        let started_at = Instant::now();
+        self.stats
+            .increment_me_writer_teardown_attempt_total(reason, mode);
         let mut close_tx: Option<mpsc::Sender<WriterCommand>> = None;
         let mut removed_addr: Option<SocketAddr> = None;
         let mut removed_dc: Option<i32> = None;
         let mut removed_uptime: Option<Duration> = None;
         let mut trigger_refill = false;
+        let mut removed = false;
         {
             let mut ws = self.writers.write().await;
             if let Some(pos) = ws.iter().position(|w| w.id == writer_id) {
+                if matches!(guard_mode, WriterRemoveGuardMode::DrainingOnly)
+                    && !ws[pos].draining.load(Ordering::Relaxed)
+                {
+                    self.stats.increment_me_writer_teardown_noop_total();
+                    self.stats
+                        .observe_me_writer_teardown_duration(mode, started_at.elapsed());
+                    return false;
+                }
                 let w = ws.remove(pos);
                 let was_draining = w.draining.load(Ordering::Relaxed);
                 if was_draining {
@@ -523,27 +635,65 @@ impl MePool {
                 }
                 close_tx = Some(w.tx.clone());
                 self.conn_count.fetch_sub(1, Ordering::Relaxed);
+                removed = true;
             }
         }
-        let conns = self.registry.writer_lost(writer_id).await;
+        // State invariant:
+        // - writer is removed from `self.writers` (pool visibility),
+        // - writer is removed from registry routing/binding maps via `writer_lost`.
+        // The close command below is only a best-effort accelerator for task shutdown.
+        // Cleanup progress must never depend on command-channel availability.
+        let _ = self.registry.writer_lost(writer_id).await;
         {
             let mut tracker = self.ping_tracker.lock().await;
             tracker.retain(|_, (_, wid)| *wid != writer_id);
         }
         self.rtt_stats.lock().await.remove(&writer_id);
         if let Some(tx) = close_tx {
-            let _ = tx.send(WriterCommand::Close).await;
+            match tx.try_send(WriterCommand::Close) {
+                Ok(()) => {}
+                Err(TrySendError::Full(_)) => {
+                    self.stats.increment_me_writer_close_signal_drop_total();
+                    self.stats
+                        .increment_me_writer_close_signal_channel_full_total();
+                    self.stats.increment_me_writer_cleanup_side_effect_failures_total(
+                        MeWriterCleanupSideEffectStep::CloseSignalChannelFull,
+                    );
+                    debug!(
+                        writer_id,
+                        "Skipping close signal for removed writer: command channel is full"
+                    );
+                }
+                Err(TrySendError::Closed(_)) => {
+                    self.stats.increment_me_writer_close_signal_drop_total();
+                    self.stats.increment_me_writer_cleanup_side_effect_failures_total(
+                        MeWriterCleanupSideEffectStep::CloseSignalChannelClosed,
+                    );
+                    debug!(
+                        writer_id,
+                        "Skipping close signal for removed writer: command channel is closed"
+                    );
+                }
+            }
         }
-        if trigger_refill
+        if let Some(addr) = removed_addr {
-            && let Some(addr) = removed_addr
-            && let Some(writer_dc) = removed_dc
-        {
             if let Some(uptime) = removed_uptime {
                 self.maybe_quarantine_flapping_endpoint(addr, uptime).await;
             }
-            self.trigger_immediate_refill_for_dc(addr, writer_dc);
+            if trigger_refill
+                && let Some(writer_dc) = removed_dc
+            {
+                self.trigger_immediate_refill_for_dc(addr, writer_dc);
+            }
         }
-        conns
+        if removed {
+            self.stats.increment_me_writer_teardown_success_total(mode);
+        } else {
+            self.stats.increment_me_writer_teardown_noop_total();
+        }
+        self.stats
+            .observe_me_writer_teardown_duration(mode, started_at.elapsed());
+        removed
     }
 
     pub(crate) async fn mark_writer_draining_with_timeout(

src/transport/middle_proxy/reader.rs

@@ -8,6 +8,7 @@ use bytes::{Bytes, BytesMut};
 use tokio::io::AsyncReadExt;
 use tokio::net::TcpStream;
 use tokio::sync::{Mutex, mpsc};
+use tokio::sync::mpsc::error::TrySendError;
 use tokio_util::sync::CancellationToken;
 use tracing::{debug, trace, warn};
 
@@ -173,12 +174,12 @@ pub(crate) async fn reader_loop(
             } else if pt == RPC_CLOSE_EXT_U32 && body.len() >= 8 {
                 let cid = u64::from_le_bytes(body[0..8].try_into().unwrap());
                 debug!(cid, "RPC_CLOSE_EXT from ME");
-                reg.route(cid, MeResponse::Close).await;
+                let _ = reg.route_nowait(cid, MeResponse::Close).await;
                 reg.unregister(cid).await;
             } else if pt == RPC_CLOSE_CONN_U32 && body.len() >= 8 {
                 let cid = u64::from_le_bytes(body[0..8].try_into().unwrap());
                 debug!(cid, "RPC_CLOSE_CONN from ME");
-                reg.route(cid, MeResponse::Close).await;
+                let _ = reg.route_nowait(cid, MeResponse::Close).await;
                 reg.unregister(cid).await;
             } else if pt == RPC_PING_U32 && body.len() >= 8 {
                 let ping_id = i64::from_le_bytes(body[0..8].try_into().unwrap());
@@ -186,13 +187,15 @@ pub(crate) async fn reader_loop(
                 let mut pong = Vec::with_capacity(12);
                 pong.extend_from_slice(&RPC_PONG_U32.to_le_bytes());
                 pong.extend_from_slice(&ping_id.to_le_bytes());
-                if tx
+                match tx.try_send(WriterCommand::DataAndFlush(Bytes::from(pong))) {
-                    .send(WriterCommand::DataAndFlush(Bytes::from(pong)))
+                    Ok(()) => {}
-                    .await
+                    Err(TrySendError::Full(_)) => {
-                    .is_err()
+                        debug!(ping_id, "PONG dropped: writer command channel is full");
-                {
+                    }
-                    warn!("PONG send failed");
+                    Err(TrySendError::Closed(_)) => {
-                    break;
+                        warn!("PONG send failed: writer channel closed");
+                        break;
+                    }
                 }
             } else if pt == RPC_PONG_U32 && body.len() >= 8 {
                 let ping_id = i64::from_le_bytes(body[0..8].try_into().unwrap());
@@ -232,6 +235,13 @@ async fn send_close_conn(tx: &mpsc::Sender<WriterCommand>, conn_id: u64) {
     let mut p = Vec::with_capacity(12);
     p.extend_from_slice(&RPC_CLOSE_CONN_U32.to_le_bytes());
     p.extend_from_slice(&conn_id.to_le_bytes());
-
+    match tx.try_send(WriterCommand::DataAndFlush(Bytes::from(p))) {
-    let _ = tx.send(WriterCommand::DataAndFlush(Bytes::from(p))).await;
+        Ok(()) => {}
+        Err(TrySendError::Full(_)) => {
+            debug!(conn_id, "ME close_conn signal skipped: writer command channel is full");
+        }
+        Err(TrySendError::Closed(_)) => {
+            debug!(conn_id, "ME close_conn signal skipped: writer command channel is closed");
+        }
+    }
 }

src/transport/middle_proxy/registry.rs

@@ -169,6 +169,7 @@ impl ConnRegistry {
         None
     }
 
+    #[allow(dead_code)]
     pub async fn route(&self, id: u64, resp: MeResponse) -> RouteResult {
         let tx = {
             let inner = self.inner.read().await;
@@ -445,30 +446,38 @@ impl ConnRegistry {
     }
 
     pub async fn writer_lost(&self, writer_id: u64) -> Vec<BoundConn> {
-        let mut inner = self.inner.write().await;
+        let mut close_txs = Vec::<mpsc::Sender<MeResponse>>::new();
-        inner.writers.remove(&writer_id);
-        inner.last_meta_for_writer.remove(&writer_id);
-        inner.writer_idle_since_epoch_secs.remove(&writer_id);
-        let conns = inner
-            .conns_for_writer
-            .remove(&writer_id)
-            .unwrap_or_default()
-            .into_iter()
-            .collect::<Vec<_>>();
-
         let mut out = Vec::new();
-        for conn_id in conns {
+        {
-            if inner.writer_for_conn.get(&conn_id).copied() != Some(writer_id) {
+            let mut inner = self.inner.write().await;
-                continue;
+            inner.writers.remove(&writer_id);
-            }
+            inner.last_meta_for_writer.remove(&writer_id);
-            inner.writer_for_conn.remove(&conn_id);
+            inner.writer_idle_since_epoch_secs.remove(&writer_id);
-            if let Some(m) = inner.meta.get(&conn_id) {
+            let conns = inner
-                out.push(BoundConn {
+                .conns_for_writer
-                    conn_id,
+                .remove(&writer_id)
-                    meta: m.clone(),
+                .unwrap_or_default()
-                });
+                .into_iter()
+                .collect::<Vec<_>>();
+
+            for conn_id in conns {
+                if inner.writer_for_conn.get(&conn_id).copied() != Some(writer_id) {
+                    continue;
+                }
+                inner.writer_for_conn.remove(&conn_id);
+                if let Some(client_tx) = inner.map.remove(&conn_id) {
+                    close_txs.push(client_tx);
+                }
+                if let Some(meta) = inner.meta.remove(&conn_id) {
+                    out.push(BoundConn { conn_id, meta });
+                }
             }
         }
+
+        for client_tx in close_txs {
+            let _ = client_tx.try_send(MeResponse::Close);
+        }
+
         out
     }
 
@@ -491,6 +500,7 @@ impl ConnRegistry {
 #[cfg(test)]
 mod tests {
     use std::net::{IpAddr, Ipv4Addr, SocketAddr};
+    use std::time::Duration;
 
     use super::ConnMeta;
     use super::ConnRegistry;
@@ -663,6 +673,39 @@ mod tests {
         assert!(registry.is_writer_empty(20).await);
     }
 
+    #[tokio::test]
+    async fn writer_lost_removes_bound_conn_from_registry_and_signals_close() {
+        let registry = ConnRegistry::new();
+        let (conn_id, mut rx) = registry.register().await;
+        let (writer_tx, _writer_rx) = tokio::sync::mpsc::channel(8);
+        registry.register_writer(10, writer_tx).await;
+        let addr = SocketAddr::new(IpAddr::V4(Ipv4Addr::LOCALHOST), 443);
+
+        assert!(
+            registry
+                .bind_writer(
+                    conn_id,
+                    10,
+                    ConnMeta {
+                        target_dc: 2,
+                        client_addr: addr,
+                        our_addr: addr,
+                        proto_flags: 0,
+                    },
+                )
+                .await
+        );
+
+        let lost = registry.writer_lost(10).await;
+        assert_eq!(lost.len(), 1);
+        assert_eq!(lost[0].conn_id, conn_id);
+        assert!(registry.get_writer(conn_id).await.is_none());
+        assert!(registry.get_meta(conn_id).await.is_none());
+        assert_eq!(registry.unregister(conn_id).await, None);
+        let close = tokio::time::timeout(Duration::from_millis(50), rx.recv()).await;
+        assert!(matches!(close, Ok(Some(MeResponse::Close))));
+    }
+
     #[tokio::test]
     async fn bind_writer_rejects_unregistered_writer() {
         let registry = ConnRegistry::new();

src/transport/middle_proxy/send.rs

@@ -14,6 +14,7 @@ use crate::config::{MeRouteNoWriterMode, MeWriterPickMode};
 use crate::error::{ProxyError, Result};
 use crate::network::IpFamily;
 use crate::protocol::constants::{RPC_CLOSE_CONN_U32, RPC_CLOSE_EXT_U32};
+use crate::stats::MeWriterTeardownReason;
 
 use super::MePool;
 use super::codec::WriterCommand;
@@ -134,7 +135,11 @@ impl MePool {
                             Ok(()) => return Ok(()),
                             Err(TimedSendError::Closed(_)) => {
                                 warn!(writer_id = current.writer_id, "ME writer channel closed");
-                                self.remove_writer_and_close_clients(current.writer_id).await;
+                                self.remove_writer_and_close_clients(
+                                    current.writer_id,
+                                    MeWriterTeardownReason::RouteChannelClosed,
+                                )
+                                .await;
                                 continue;
                             }
                             Err(TimedSendError::Timeout(_)) => {
@@ -151,7 +156,11 @@ impl MePool {
                     }
                     Err(TrySendError::Closed(_)) => {
                         warn!(writer_id = current.writer_id, "ME writer channel closed");
-                        self.remove_writer_and_close_clients(current.writer_id).await;
+                        self.remove_writer_and_close_clients(
+                            current.writer_id,
+                            MeWriterTeardownReason::RouteChannelClosed,
+                        )
+                        .await;
                         continue;
                     }
                 }
@@ -458,7 +467,11 @@ impl MePool {
                     Err(TrySendError::Closed(_)) => {
                         self.stats.increment_me_writer_pick_closed_total(pick_mode);
                         warn!(writer_id = w.id, "ME writer channel closed");
-                        self.remove_writer_and_close_clients(w.id).await;
+                        self.remove_writer_and_close_clients(
+                            w.id,
+                            MeWriterTeardownReason::RouteChannelClosed,
+                        )
+                        .await;
                         continue;
                     }
                 }
@@ -503,7 +516,11 @@ impl MePool {
                 Err(TimedSendError::Closed(_)) => {
                     self.stats.increment_me_writer_pick_closed_total(pick_mode);
                     warn!(writer_id = w.id, "ME writer channel closed (blocking)");
-                    self.remove_writer_and_close_clients(w.id).await;
+                    self.remove_writer_and_close_clients(
+                        w.id,
+                        MeWriterTeardownReason::RouteChannelClosed,
+                    )
+                    .await;
                 }
                 Err(TimedSendError::Timeout(_)) => {
                     self.stats.increment_me_writer_pick_full_total(pick_mode);
@@ -643,13 +660,23 @@ impl MePool {
             let mut p = Vec::with_capacity(12);
             p.extend_from_slice(&RPC_CLOSE_EXT_U32.to_le_bytes());
             p.extend_from_slice(&conn_id.to_le_bytes());
-            if w.tx
+            match w.tx.try_send(WriterCommand::DataAndFlush(Bytes::from(p))) {
-                .send(WriterCommand::DataAndFlush(Bytes::from(p)))
+                Ok(()) => {}
-                .await
+                Err(TrySendError::Full(_)) => {
-                .is_err()
+                    debug!(
-            {
+                        conn_id,
-                debug!("ME close write failed");
+                        writer_id = w.writer_id,
-                self.remove_writer_and_close_clients(w.writer_id).await;
+                        "ME close skipped: writer command channel is full"
+                    );
+                }
+                Err(TrySendError::Closed(_)) => {
+                    debug!("ME close write failed");
+                    self.remove_writer_and_close_clients(
+                        w.writer_id,
+                        MeWriterTeardownReason::CloseRpcChannelClosed,
+                    )
+                    .await;
+                }
             }
         } else {
             debug!(conn_id, "ME close skipped (writer missing)");
@@ -666,8 +693,12 @@ impl MePool {
             p.extend_from_slice(&conn_id.to_le_bytes());
             match w.tx.try_send(WriterCommand::DataAndFlush(Bytes::from(p))) {
                 Ok(()) => {}
-                Err(TrySendError::Full(cmd)) => {
+                Err(TrySendError::Full(_)) => {
-                    let _ = tokio::time::timeout(Duration::from_millis(50), w.tx.send(cmd)).await;
+                    debug!(
+                        conn_id,
+                        writer_id = w.writer_id,
+                        "ME close_conn skipped: writer command channel is full"
+                    );
                 }
                 Err(TrySendError::Closed(_)) => {
                     debug!(conn_id, "ME close_conn skipped: writer channel closed");