Probe every domain every minute, and keep it out of its own traffic numbers

NPM's access log only describes hosts somebody visited; the host most likely to be quietly broken
is the one nobody does. Probes carry a User-Agent npm_access_stats.sh drops — unmarked, this
monitor would be fifty thousand requests a day in the very logs it reports on.

Plugin/unraid/pages/auth.php

@@ -266,6 +266,15 @@ require_once dirname(__DIR__) . '/include/ai_chat.php';
 .vv-au-stat-s   { font-size:9px;color:#3a3a3a;margin-top:1px; }
 .vv-au-stat-s.warn { color:#8a6a2a; }
 .vv-au-stat-s.bad  { color:#a34; }
+/* Healthy uptime is deliberately quiet — a page where thirty green numbers shout is a page where
+   the two red ones do not. */
+.vv-au-stat-n.ok2  { color:#4a7a4a; }
+
+/* One bar per probe, last hour. The shape is the part a percentage throws away: one long outage
+   and sixty scattered blips are the same 50% and completely different problems. */
+.vv-au-spark    { display:flex;gap:1px;justify-content:flex-end;align-items:flex-end;height:11px;margin-top:3px; }
+.vv-au-spark i  { width:2px;height:100%;background:#2d5a2d;border-radius:1px;flex-shrink:0; }
+.vv-au-spark i.d{ background:#ef5350; }
 /* A disabled host is still listed — it is a thing you might re-enable — but it should not read
    as part of what is currently serving. */
 .vv-au-off td   { opacity:.42; }
@@ -354,7 +363,7 @@ $tabs    = ['proxies' => 'Proxies', 'users' => 'Users & Groups',
     <div class="vv-au-loading" id="vv-au-proxy-loading">Loading…</div>
     <table class="vv-au-tbl" id="vv-au-proxy-tbl" style="display:none">
       <thead><tr>
-        <th>Host</th><th>Flags</th><th style="text-align:right">Requests</th><th style="text-align:right">Errors</th><th>On</th><th></th>
+        <th>Host</th><th>Flags</th><th style="text-align:right">Uptime</th><th style="text-align:right">Requests</th><th style="text-align:right">Errors</th><th>On</th><th></th>
       </tr></thead>
       <tbody id="vv-au-proxy-body"></tbody>
     </table>
@@ -510,6 +519,8 @@ const IS_OWNER = <?= $isOwner ? 'true' : 'false' ?>;
 let _proxies = [], _certs = [];
 // Keyed by proxy-host id, from Tools/npm_access_stats.sh. Empty until it has run once.
 let _proxyStats = {};
+// Keyed by hostname, from Tools/uptime_probe.sh. Empty until the probe has run once.
+let _uptime = {};
 let _users   = [], _groups = [];
 let _rules   = [], _defaultPolicy = 'deny';
 // The trailing comment on the default_policy line, carried so a save puts it back. The block is
@@ -646,7 +657,8 @@ function _loadProxies() {
 
   // Stats are optional decoration — the list must render whether or not the aggregator has run,
   // so this neither blocks _check() nor fails the load.
-  _get('npm_stats', r => { _proxyStats = (r && r.ok && r.hosts) ? r.hosts : {}; if (proxiesLoaded) _renderProxies(); });
+  _get('npm_stats',  r => { _proxyStats = (r && r.ok && r.hosts)   ? r.hosts   : {}; if (proxiesLoaded) _renderProxies(); });
+  _get('npm_uptime', r => { _uptime     = (r && r.ok && r.domains) ? r.domains : {}; if (proxiesLoaded) _renderProxies(); });
   _get('npm_certs', r => { _certs = r.certs || []; certsLoaded = true; _check(); });
   _get('npm_proxies', r => {
     if (!r.ok) { loading.innerHTML = '<span style="color:#ef5350">'+_esc(r.error)+'</span>'; return; }
@@ -716,6 +728,7 @@ function _renderProxies() {
         <div class="vv-au-fwd">→ ${_esc(fwd)}</div>
       </td>
       <td><div class="vv-au-marks">${marks}</div></td>
+      <td class="vv-au-num">${_upCell(p)}</td>
       <td class="vv-au-num">${reqCell}</td>
       <td class="vv-au-num">${errCell}</td>
       <td>${_togHtml('proxy-' + p.id, enabled, enabled ? 'Enabled — click to disable' : 'Disabled — click to enable')}</td>
@@ -729,6 +742,33 @@ function _renderProxies() {
   }).join('');
 }
 
+// Uptime for the host, from Tools/uptime_probe.sh. A proxy host can carry several domains, so the
+// worst of them is what the row reports — a host is only as reachable as its least reachable name,
+// and averaging would hide one dead domain behind three healthy ones.
+function _upCell(p) {
+  const doms = (p.domain_names || []).map(d => String(d).toLowerCase()).filter(d => !d.includes('*'));
+  const recs = doms.map(d => _uptime[d]).filter(Boolean);
+  if (!recs.length) return '<div class="vv-au-stat-n dim">—</div>';
+
+  const worst = recs.reduce((a, b) => ((a.h24 ?? 101) <= (b.h24 ?? 101) ? a : b));
+  const down  = recs.some(r => r.state === 'down');
+  const pct   = worst.h24;
+  // Banded on the same three-state rule as the error column: fine, worth a look, not working.
+  const cls   = down ? ' bad' : (pct === null ? ' dim' : (pct >= 99.5 ? ' ok2' : (pct >= 95 ? ' warn' : ' bad')));
+
+  // Sixty samples is the last hour at a one-minute cadence. Drawn as bars rather than a number
+  // because the shape — one long outage or sixty scattered blips — is the part a percentage loses.
+  const bars = (worst.samples || []).map(s =>
+    `<i class="${s ? '' : 'd'}"></i>`).join('');
+
+  const title = down
+    ? 'DOWN — ' + (worst.last_detail || 'no response')
+    : 'up' + (worst.last_ms ? ' · ' + worst.last_ms + 'ms' : '');
+
+  return `<div class="vv-au-stat-n${cls}" title="${_esc(title)}">${pct === null ? '—' : (pct >= 99.95 ? '100' : pct.toFixed(pct >= 99 ? 2 : 1))}<span class="vv-au-stat-u">% 24h</span></div>
+    <div class="vv-au-spark" title="last hour, one bar per minute">${bars}</div>`;
+}
+
 function _certName(id) {
   const c = _certs.find(x => String(x.id) === String(id));
   return c ? (c.nice_name || (c.domain_names || []).join(', ')) : 'certificate ' + id;

Plugin/unraid/pages/readme/ui-map.md

@@ -1996,6 +1996,18 @@ Saved into `master.conf`, which does not need to be opened by hand.
 | `TRANSCODE_LOG_RETENTION` | a number box, in days | in this section | days before old entries purged |
 | `TRANSCODE_CHECK_EMBY` | a switch | in this section | — |
 
+## Uptime Probe
+
+Route: Settings tab → All settings → *Uptime Probe*
+
+Saved into `master.conf`, which does not need to be opened by hand.
+
+| Setting | Control | Where | What it does |
+|---|---|---|---|
+| `UPTIME_PROBE_ENABLED` | a switch | in this section | master switch for the per-minute probe |
+| `UPTIME_PROBE_TIMEOUT` | a number box, in seconds | in this section | seconds per domain before it counts as down |
+| `UPTIME_PROBE_LIST_TTL` | a number box, in seconds | in this section | seconds to reuse the domain list from NPM before re-reading it |
+
 ## Version Parity
 
 Route: Settings tab → All settings → *Version Parity*