diff --git a/backend/library_cleanup.py b/backend/library_cleanup.py
index 1e1546e..fd8cb44 100644
--- a/backend/library_cleanup.py
+++ b/backend/library_cleanup.py
@@ -3,12 +3,18 @@ Read-only by design: scan() only reports findings; server.py's /library/cleanup/
 performs the actual deletion/reset the admin picked, one category at a time.
 """
 import shutil
+from concurrent.futures import ThreadPoolExecutor
 from datetime import datetime, timezone
 from pathlib import Path
 from typing import Dict, List
 
 import asyncio
 
+# Movie/episode/track files often live on an NFS/CIFS-mounted NAS, where every is_file() stat
+# is a real network round trip. Checking hundreds of them one at a time (even off the event
+# loop) can take minutes; a bounded worker pool checks them concurrently instead.
+_STAT_WORKERS = 24
+
 
 def _file_path(storage_type: str, storage_path: str, videos_dir: Path) -> Path:
     if storage_type in ("radarr", "sonarr", "scan"):
@@ -19,29 +25,40 @@ def _file_path(storage_type: str, storage_path: str, videos_dir: Path) -> Path:
 def _scan_filesystem(movies: List[Dict], episodes: List[Dict], tracks: List[Dict],
                       videos_dir: Path, hls_dir: Path, live_ids: set) -> Dict[str, List[Dict]]:
     """Every filesystem stat call in the whole scan lives here, run off the event loop via
-    asyncio.to_thread — movie/episode/track paths are often on an NFS/CIFS-mounted NAS, and
-    doing hundreds of blocking stats directly in an async function would freeze the entire
-    server (every other request) for as long as the scan takes."""
+    asyncio.to_thread, with the individual stat calls themselves parallelized across a small
+    worker pool — sequential one-at-a-time NAS round trips were taking minutes for a library
+    of any real size."""
     out: Dict[str, List[Dict]] = {"missing_files": [], "orphaned_hls": []}
 
+    candidates = []  # (finding_dict, path_to_check)
     for m in movies:
         if m.get("storage_type") not in ("local", "radarr", "scan") or not m.get("storage_path"):
             continue
-        if not _file_path(m["storage_type"], m["storage_path"], videos_dir).is_file():
-            out["missing_files"].append({"id": m["id"], "content_type": "movie", "label": m.get("title", "Untitled"), "detail": m["storage_path"]})
-
+        candidates.append((
+            {"id": m["id"], "content_type": "movie", "label": m.get("title", "Untitled"), "detail": m["storage_path"]},
+            _file_path(m["storage_type"], m["storage_path"], videos_dir),
+        ))
     for e in episodes:
         if e.get("_orphaned"):
             continue  # already reported as orphaned_episodes; don't also report a missing file
         if e.get("storage_type") not in ("local", "sonarr", "scan") or not e.get("storage_path"):
             continue
         label = f"S{e.get('season_number')}E{e.get('episode_number')} {e.get('title') or ''}".strip()
-        if not _file_path(e["storage_type"], e["storage_path"], videos_dir).is_file():
-            out["missing_files"].append({"id": e["id"], "content_type": "episode", "label": label, "detail": e["storage_path"]})
-
+        candidates.append((
+            {"id": e["id"], "content_type": "episode", "label": label, "detail": e["storage_path"]},
+            _file_path(e["storage_type"], e["storage_path"], videos_dir),
+        ))
     for t in tracks:
-        if not Path(t["storage_path"]).is_file():
-            out["missing_files"].append({"id": t["id"], "content_type": "track", "label": f"{t.get('artist','')} — {t.get('title','Untitled')}", "detail": t["storage_path"]})
+        candidates.append((
+            {"id": t["id"], "content_type": "track", "label": f"{t.get('artist','')} — {t.get('title','Untitled')}", "detail": t["storage_path"]},
+            Path(t["storage_path"]),
+        ))
+
+    with ThreadPoolExecutor(max_workers=_STAT_WORKERS) as pool:
+        exists = pool.map(lambda pair: pair[1].is_file(), candidates)
+        for (finding, _path), is_present in zip(candidates, exists):
+            if not is_present:
+                out["missing_files"].append(finding)
 
     if hls_dir.is_dir():
         for d in hls_dir.iterdir():
diff --git a/frontend/src/components/Navbar.jsx b/frontend/src/components/Navbar.jsx
index 5e20cc6..910e150 100644
--- a/frontend/src/components/Navbar.jsx
+++ b/frontend/src/components/Navbar.jsx
@@ -35,7 +35,6 @@ export const Navbar = () => {
               <NavLink to="/music" className={linkClass} data-testid="nav-music">Music</NavLink>
               <NavLink to="/requests" className={linkClass} data-testid="nav-requests">Wishlist</NavLink>
               {user.is_admin && <NavLink to="/admin" className={linkClass} data-testid="nav-admin">Admin</NavLink>}
-              {user.is_admin && <NavLink to="/admin/users" className={linkClass} data-testid="nav-users">Users</NavLink>}
             </nav>
           )}
         </div>
diff --git a/frontend/src/pages/Admin.jsx b/frontend/src/pages/Admin.jsx
index ad1e1af..737a839 100644
--- a/frontend/src/pages/Admin.jsx
+++ b/frontend/src/pages/Admin.jsx
@@ -94,9 +94,6 @@ export default function Admin() {
           <Link to="/admin/queue" className="flex items-center gap-2 bg-white/10 hover:bg-white/20 text-white px-5 py-2 text-xs uppercase tracking-[0.2em] border border-white/10" data-testid="admin-queue-link">
             <SettingsIcon size={14} strokeWidth={1.5} /> Queue
           </Link>
-          <Link to="/settings" className="flex items-center gap-2 bg-white/10 hover:bg-white/20 text-white px-5 py-2 text-xs uppercase tracking-[0.2em] border border-white/10" data-testid="admin-settings-link">
-            <SettingsIcon size={14} strokeWidth={1.5} /> Settings
-          </Link>
           <Link to="/admin/users" className="flex items-center gap-2 bg-white/10 hover:bg-white/20 text-white px-5 py-2 text-xs uppercase tracking-[0.2em] border border-white/10" data-testid="admin-users-link">
             <SettingsIcon size={14} strokeWidth={1.5} /> Users
           </Link>