Fix recount stop: kill whole process group, not just shell

AGENTS.md

@@ -43,7 +43,7 @@ PYTHONNOUSERSITE=1 venv/bin/python recounting_dashboard_upload.py
 - **`numpy<2`** is required for `rknn-toolkit-lite2` compatibility.
 - **Install path in service files is `/opt/bytetrack-counter`** (not the `/opt/jetson-counter` mentioned in README/DEPLOY). The `.env.example` also reflects `/opt/bytetrack-counter`.
 - Service user is **`root`**, not `jetson` (despite README saying otherwise).
-- Two systemd units: `bytetrack-counter.service` + `bytetrack-counter-dashboard.service`.
+- Three systemd units: `bytetrack-counter.service`, `bytetrack-counter-dashboard.service`, and `bytetrack-recounting-upload-dashboard.service` (upload recounting dashboard, port 5003).
 - `counter_live.py` (TensorRT) is Jetson-only and won't work on RK3588.
 - **Reset flow**: `POST /api/reset` → deletes state JSON + touches `{SHM_DIR}/.reset`. `batch_store._check_reset_signal()` watches this marker and clears in-memory state on next crossing.
 - **Upload filename contract**: uploads are expected as `batch_XX_YYYYMMDD_HHmmSS.mp4` (XX = batch number, timestamp = date). `recounting_dashboard_upload.py` parses both to query the live API's `/api/day-detail/<YYYY-MM-DD>` for the recorded count.

bytetrack-recounting-upload-dashboard.service

@@ -0,0 +1,34 @@
+[Unit]
+Description=Edge Counter Recounting Upload Dashboard (Flask, port 5003)
+Documentation=file:///opt/bytetrack-counter/DEPLOY.md
+After=network-online.target bytetrack-counter-dashboard.service
+Wants=network-online.target bytetrack-counter-dashboard.service
+
+[Service]
+Type=simple
+User=root
+Group=root
+
+WorkingDirectory=/opt/bytetrack-counter
+
+EnvironmentFile=/opt/bytetrack-counter/.env
+Environment=PATH=/opt/bytetrack-counter/venv/bin:/usr/local/bin:/usr/bin:/bin
+Environment=PYTHONNOUSERSITE=1
+Environment=FLASK_DEBUG=false
+
+ExecStart=/opt/bytetrack-counter/venv/bin/python recounting_dashboard_upload.py
+
+TimeoutStopSec=15
+KillSignal=SIGTERM
+
+Restart=on-failure
+RestartSec=5
+StartLimitInterval=60s
+StartLimitBurst=3
+
+NoNewPrivileges=true
+ProtectHome=true
+PrivateTmp=false
+
+[Install]
+WantedBy=multi-user.target

recounting_dashboard_upload.py

@@ -8,6 +8,7 @@ import logging
 import os
 import re
 import shlex
+import signal
 import subprocess
 import threading
 import time
@@ -74,13 +75,31 @@ def _api_get(base_url, path, default=None):
 
 def _stop_recount_locked():
     global _recount_proc
-    if _recount_proc:
+    proc = _recount_proc
+    _recount_proc = None
+    if proc is None:
+        return
+    try:
+        pgid = os.getpgid(proc.pid)
+    except (ProcessLookupError, PermissionError):
+        return
+    try:
+        os.killpg(pgid, signal.SIGTERM)
+    except (ProcessLookupError, PermissionError):
+        pass
+    try:
+        proc.wait(timeout=5)
+    except subprocess.TimeoutExpired:
         try:
-            _recount_proc.kill()
+            os.killpg(pgid, signal.SIGKILL)
-            _recount_proc.wait(timeout=3)
+        except (ProcessLookupError, PermissionError):
+            pass
+        try:
+            proc.wait(timeout=3)
         except Exception:
             pass
-        _recount_proc = None
+    except Exception:
+        pass
 
 
 def _folder_date(filepath, base):
@@ -179,7 +198,7 @@ def start_recount():
         _current_file = upload_path
         cmd_str = RECOUNT_CMD.replace("{path}", shlex.quote(str(upload_path)))
         try:
-            _recount_proc = subprocess.Popen(cmd_str, shell=True)
+            _recount_proc = subprocess.Popen(cmd_str, shell=True, start_new_session=True)
             print(f"[{datetime.now().strftime('%Y-%m-%d %H:%M:%S')}] Recount started: {cmd_str}")
         except Exception as e:
             _current_file = None