Merge pull request 'feat(monitoring): 监控页三分区仪表 + 上游探针' (#22) from feat-monitoring-enrich into main

Reviewed-on: #22

Dockerfile

@@ -15,7 +15,14 @@ RUN pip install --no-cache-dir -r requirements.txt
 COPY src ./src
 RUN pip install --no-cache-dir .
 
-# 将工作目录所有权移交给非 root 用户
+# 日志目录预创建: compose 运行时把 applogs 命名卷挂到 /app/logs,
+# 卷挂载点默认由 Docker 以 root:root 创建 —— 镜像内不预建的话,
+# 非 root 进程写日志会 PermissionError(Errno 13)。
+# 命名卷为空且首次挂载时,Docker 会复制镜像内该目录的内容与属主,
+# 因此在这里 mkdir + chown 即可让卷目录归 profeto 所有。
+RUN mkdir -p /app/logs
+
+# 将工作目录所有权移交给非 root 用户(含上面的日志目录)
 RUN chown -R profeto:profeto /app
 
 EXPOSE 8000

frontend/src/admin/dal.ts

@@ -363,6 +363,19 @@ export function fetchIngestStatus(): Promise<{ sources: IngestSourceStatus[] }>
   return api.get<{ sources: IngestSourceStatus[] }>(`${API_BASE}/admin/ingest/status`)
 }
 
+/**
+ * 上游数据源(bzzoiro)可达性探针:轻量 GET,不携带 Key、不消耗配额
+ */
+export function fetchUpstreamProbe(): Promise<{
+  ok: boolean
+  status_code?: number
+  latency_ms: number
+  endpoint: string
+  error?: string
+}> {
+  return api.get<never>(`${API_BASE}/admin/monitoring/upstream`)
+}
+
 /**
  * 采集任务状态轮询(单任务)
  */

frontend/src/admin/pages/Collection.tsx

@@ -94,6 +94,10 @@ export default function CollectionPage() {
           setTaskStatus(job.status === 'success' ? 'done' : 'error')
           stopPolling()
           loadRecentJobs() // 终态后刷新历史列表
+          // 广播采集终态:数据完整性等依赖页立即刷新,不必等轮询周期
+          window.dispatchEvent(new CustomEvent('profeto:ingest-done', {
+            detail: { jobId: job.id, status: job.status },
+          }))
         }
       } catch { /* 单次轮询失败不影响后续 */ }
     }

frontend/src/admin/pages/DataCompleteness.tsx

@@ -55,6 +55,9 @@ export default function DataCompletenessPage() {
   const [loading, setLoading] = useState(true)
   const [error, setError] = useState<string | null>(null)
   const [highlightedLeague, setHighlightedLeague] = useState<string | null>(null)
+  // 页面文案承诺「每 5 秒自动刷新」,此前并未实现(仅挂载加载一次),
+  // 采集完成后数字不动 —— 现补齐:5s 轮询 + 采集完成事件即时刷新。
+  const [autoRefresh, setAutoRefresh] = useState(true)
   const leagueRefs = useRef<Record<string, HTMLDivElement | null>>({})
 
   const load = useCallback(async () => {
@@ -72,6 +75,20 @@ export default function DataCompletenessPage() {
 
   useEffect(() => { load() }, [load])
 
+  // 5s 自动轮询(与页面文案一致);已有数据时刷新不闪骨架
+  useEffect(() => {
+    if (!autoRefresh) return
+    const t = setInterval(load, 5_000)
+    return () => clearInterval(t)
+  }, [autoRefresh, load])
+
+  // 采集页任务终态广播 → 立即刷新(不等下一个 5s 周期)
+  useEffect(() => {
+    const onIngestDone = () => load()
+    window.addEventListener('profeto:ingest-done', onIngestDone)
+    return () => window.removeEventListener('profeto:ingest-done', onIngestDone)
+  }, [load])
+
   // 点击问题项 → 滚动到对应联赛卡片并高亮
   const scrollToLeague = useCallback((code: string) => {
     setHighlightedLeague(code)
@@ -92,9 +109,20 @@ export default function DataCompletenessPage() {
         title="数据完整性"
         description="按联赛统计 bzzoiro 数据采集覆盖度。每 5 秒自动刷新,或点击右上角按钮手动刷新。"
         action={
+          <div className="flex items-center gap-2">
+            <label className="flex cursor-pointer items-center gap-1.5 text-2xs text-ink-500">
+              <input
+                type="checkbox"
+                checked={autoRefresh}
+                onChange={e => setAutoRefresh(e.target.checked)}
+                className="accent-current"
+              />
+              5s 自动刷新
+            </label>
             <button onClick={load} disabled={loading} className="btn-sm btn-outline">
               {loading ? <><Spinner /> 刷新中</> : '刷新'}
             </button>
+          </div>
         }
       />
 

frontend/src/admin/pages/Logs.tsx

@@ -53,20 +53,21 @@ export default function LogsPage() {
   }, [load])
 
   const scrollRef = useRef<HTMLDivElement>(null)
-  const userScrolledUp = useRef(false)
+  const userScrolledDown = useRef(false)
 
-  // 检测用户是否向上滚动过
+  // 检测用户是否向下滚动回看历史(旧日志在下方)
   const handleScroll = () => {
     const el = scrollRef.current
     if (!el) return
-    const atBottom = el.scrollHeight - el.scrollTop - el.clientHeight < 50
+    const atTop = el.scrollTop < 50
-    userScrolledUp.current = !atBottom
+    userScrolledDown.current = !atTop
   }
 
-  // 加载后自动滚动到底部(仅当用户未向上滚动时)
+  // 列表最新在最上面(后端倒序返回);打开页面/自动刷新时定位到顶部=最新。
+  // 用户向下滚动回看历史时暂停定位,拉回顶部即恢复。
   useEffect(() => {
-    if (autoRefresh && !userScrolledUp.current && scrollRef.current) {
+    if (autoRefresh && !userScrolledDown.current && scrollRef.current) {
-      scrollRef.current.scrollTop = scrollRef.current.scrollHeight
+      scrollRef.current.scrollTop = 0
     }
   }, [entries, autoRefresh])
 

frontend/src/admin/pages/Monitoring.tsx

@@ -1,20 +1,49 @@
 /**
- * Admin 后台 - 监控面板(报刊风)
+ * Admin 后台 - 监控页(报刊风·三分区)
  *
- * 功能:
+ * 此前只有存活/DB 两个布尔,与顶栏健康点重复,信息丰度不足。
- * - /health 存活检查(自动:30 秒一轮;可手动刷新)
+ * 现在聚合全站已有信号,回答三个问题:
- * - /health/ready 数据库就绪检查
+ *   1. 基础设施活着吗 —— 服务存活 / DB 就绪 / 版本 / 运行时间 / 上游 bzzoiro 可达性
- * - 服务名 / 版本 / 运行时间 / 检查项(后端返回什么就展示什么)
+ *   2. 数据管线健康吗 —— 最近采集成功率 / 死信 / 数据缺口
+ *   3. LLM 服务正常吗 —— 预测成功率 / 平均延迟
+ * 顶部「需要关注」聚合条:任何一项异常即亮红并直达处理页。
+ * 30s 自动巡检 + 手动巡检。
  */
 
-import { useCallback, useEffect, useState } from 'react'
+import { useCallback, useEffect, useMemo, useState } from 'react'
-import { fetchHealth } from '../dal'
+import { Link } from 'react-router-dom'
+import { fetchHealth, fetchUpstreamProbe, fetchIngestJobs, fetchIngestFailures, fetchDataCompleteness, fetchLLMUsageStats } from '../dal'
+import type { DataCompletenessResponse, IngestFailureItem } from '../dal'
+import type { IngestJob } from '../types'
 import { api } from '../api'
-import { Card, CardBody, CardHeader, SectionHeader, Alert, Spinner, Badge } from '../components'
+import { Alert, SectionHeader, Spinner } from '../components'
+
+type UpstreamProbe = { ok: boolean; status_code?: number; latency_ms: number; endpoint: string; error?: string }
+type LLMStats = { total_predictions: number; avg_latency_ms: number; success_rate: number }
+
+interface TodoItem {
+  key: string
+  label: string
+  to: string
+}
+
+function MetricCard({ label, children }: { label: string; children: React.ReactNode }) {
+  return (
+    <div className="border border-ink-900 bg-paper-50 p-5">
+      <div className="text-2xs tracking-[0.2em] text-ink-400">{label}</div>
+      <div className="mt-2">{children}</div>
+    </div>
+  )
+}
 
 export default function MonitoringPage() {
-  const [health, setHealth] = useState<any>(null)
+  const [health, setHealth] = useState<Record<string, unknown> | null>(null)
   const [ready, setReady] = useState<'ready' | 'not_ready' | null>(null)
+  const [upstream, setUpstream] = useState<UpstreamProbe | null>(null)
+  const [recentJobs, setRecentJobs] = useState<IngestJob[] | null>(null)
+  const [failures, setFailures] = useState<IngestFailureItem[]>([])
+  const [completeness, setCompleteness] = useState<DataCompletenessResponse | null>(null)
+  const [llmStats, setLlmStats] = useState<LLMStats | null>(null)
   const [loading, setLoading] = useState(true)
   const [error, setError] = useState<string | null>(null)
   const [lastCheck, setLastCheck] = useState<string>('')
@@ -22,20 +51,28 @@ export default function MonitoringPage() {
   const refresh = useCallback(async () => {
     setLoading(true)
     setError(null)
-    try {
+    // 各信号独立容错:单项失败只降级对应卡片
-      const [h, r] = await Promise.allSettled([
+    const [h, r, u, j, f, c, l] = await Promise.allSettled([
       fetchHealth(),
       api.get<{ status: string }>('/health/ready'),
+      fetchUpstreamProbe(),
+      fetchIngestJobs({ limit: 10 }),
+      fetchIngestFailures(),
+      fetchDataCompleteness(),
+      fetchLLMUsageStats() as Promise<LLMStats>,
     ])
     setHealth(h.status === 'fulfilled' ? h.value : null)
     setReady(r.status === 'fulfilled' ? (r.value?.status as 'ready' | 'not_ready') : null)
+    setUpstream(u.status === 'fulfilled' ? u.value : null)
+    setRecentJobs(j.status === 'fulfilled' && Array.isArray(j.value) ? j.value : null)
+    setFailures(f.status === 'fulfilled' ? f.value : [])
+    setCompleteness(c.status === 'fulfilled' ? c.value : null)
+    setLlmStats(l.status === 'fulfilled' ? l.value : null)
     if (h.status === 'rejected') {
       setError(h.reason instanceof Error ? h.reason.message : '无法连接到后端')
     }
     setLastCheck(new Date().toLocaleTimeString('zh-CN', { hour12: false }))
-    } finally {
     setLoading(false)
-    }
   }, [])
 
   useEffect(() => {
@@ -45,108 +82,188 @@ export default function MonitoringPage() {
   }, [refresh])
 
   const alive = health?.status === 'healthy' || health?.status === 'ok'
+  const deadLetterCount = failures.filter(f => f.status !== 'resolved').length
+  const missingStatsLeagues = completeness?.leagues.filter(
+    l => l.matches.finished > 0 && l.stats.rows === 0,
+  ).length ?? 0
+  const recentFailedJobs = useMemo(
+    () => (recentJobs ?? []).filter(j => j.status === 'failed').length,
+    [recentJobs],
+  )
+
+  // ── 「需要关注」聚合:任何一项异常即亮红 ──
+  const todos: TodoItem[] = [
+    !alive && { key: 'alive', label: '服务存活异常', to: '/admin/logs' },
+    ready === 'not_ready' && { key: 'db', label: '数据库未就绪', to: '/admin/logs' },
+    upstream && !upstream.ok && { key: 'upstream', label: '上游 bzzoiro 不可达', to: '/admin/logs' },
+    recentFailedJobs > 0 && { key: 'jobs', label: `最近采集失败 ${recentFailedJobs} 次`, to: '/admin/collection' },
+    deadLetterCount > 0 && { key: 'deadletter', label: `死信待处理 ${deadLetterCount} 条`, to: '/admin/data-pipeline' },
+    missingStatsLeagues > 0 && { key: 'missing', label: `${missingStatsLeagues} 个联赛缺统计`, to: '/admin/data-completeness' },
+  ].filter((t): t is TodoItem => t !== false)
+
+  const okJobs = (recentJobs ?? []).filter(j => j.status === 'success').length
+  const runJobs = (recentJobs ?? []).filter(j => j.status === 'success' || j.status === 'failed').length
+  const lastJobTime = recentJobs?.[0]?.created_at
 
   return (
     <div className="space-y-6">
       <SectionHeader
         title="系统监控"
-        description="存活与数据库就绪检查,每 30 秒自动巡检一次。"
+        description="基础设施 / 数据管线 / LLM 三区巡检,每 30 秒自动刷新。"
-      />
+        action={
-
+          <div className="flex items-center gap-2">
-      <div className="flex items-center justify-between">
+            <span className="text-2xs text-ink-400">{lastCheck && `最近巡检 ${lastCheck}`}</span>
-        <span className="text-2xs text-ink-400">
-          {lastCheck && `最近巡检 ${lastCheck}`}
-        </span>
             <button onClick={refresh} disabled={loading} className="btn btn-sm">
               {loading ? (<><Spinner /> 检查中</>) : '立即巡检'}
             </button>
           </div>
+        }
+      />
 
       {error && (
-        <Alert
+        <Alert kind="error" title="无法连接到后端" message={`${error}\n请确认服务是否正常运行,以及登录会话是否已过期。`} />
-          kind="error"
-          title="无法连接到后端"
-          message={`${error}\n请确认服务是否正常运行,以及登录会话是否已过期。`}
-        />
       )}
 
-      <div className="grid gap-4 sm:grid-cols-2 lg:grid-cols-3">
+      {/* ── 需要关注:任何一项异常即亮红并直达处理页 ── */}
-        {/* 存活状态 */}
+      {!loading && todos.length > 0 ? (
-        <div className="border border-ink-900 bg-paper-50 p-5">
+        <div className="grid gap-3 sm:grid-cols-2 lg:grid-cols-3">
-          <div className="text-2xs tracking-[0.2em] text-ink-400">存活状态</div>
+          {todos.map(t => (
-          <div className="mt-2 flex items-center gap-2">
+            <Link
-            <span
+              key={t.key}
-              className={`inline-block h-2 w-2 ${alive ? 'bg-ink-900' : 'bg-press'}`}
+              to={t.to}
-              aria-hidden="true"
+              className="flex items-center gap-3 border border-press bg-press-wash/40 px-4 py-3 transition-colors hover:bg-press-wash"
-            />
+            >
+              <span className="h-2 w-2 shrink-0 animate-pulse rounded-full bg-press" aria-hidden="true" />
+              <span className="text-xs text-ink-800">{t.label}</span>
+              <span className="ml-auto text-press" aria-hidden="true">→</span>
+            </Link>
+          ))}
+        </div>
+      ) : (
+        !loading && (
+          <p className="flex items-center gap-2 border-b border-ink-200 pb-3 text-2xs text-ink-400">
+            <span className="inline-block h-1.5 w-1.5 bg-ink-900" aria-hidden="true" />
+            各项巡检正常:服务、数据库、上游、采集与数据完整性均无异常。
+          </p>
+        )
+      )}
+
+      {/* ── 基础设施 ── */}
+      <section>
+        <h2 className="section-head mb-3">基础设施</h2>
+        <div className="grid gap-4 sm:grid-cols-2 lg:grid-cols-4">
+          <MetricCard label="存活状态">
+            <div className="flex items-center gap-2">
+              <span className={`inline-block h-2 w-2 ${alive ? 'bg-ink-900' : 'bg-press'}`} aria-hidden="true" />
               <span className={`font-serif text-xl font-bold ${alive ? 'text-ink-900' : 'text-press'}`}>
                 {health ? (alive ? '正常' : String(health.status)) : '—'}
               </span>
             </div>
-        </div>
+          </MetricCard>
 
-        {/* 数据库就绪 */}
+          <MetricCard label="数据库就绪">
-        <div className="border border-ink-900 bg-paper-50 p-5">
+            <div className="flex items-center gap-2">
-          <div className="text-2xs tracking-[0.2em] text-ink-400">数据库就绪</div>
+              <span className={`inline-block h-2 w-2 ${ready === 'ready' ? 'bg-ink-900' : ready === null ? 'bg-ink-300' : 'bg-press'}`} aria-hidden="true" />
-          <div className="mt-2 flex items-center gap-2">
+              <span className={`font-serif text-xl font-bold ${ready === 'not_ready' ? 'text-press' : 'text-ink-900'}`}>
-            <span
-              className={`inline-block h-2 w-2 ${ready === 'ready' ? 'bg-ink-900' : ready === null ? 'bg-ink-300' : 'bg-press'}`}
-              aria-hidden="true"
-            />
-            <span
-              className={`font-serif text-xl font-bold ${ready === 'not_ready' ? 'text-press' : 'text-ink-900'}`}
-            >
                 {ready === 'ready' ? '就绪' : ready === 'not_ready' ? '未就绪' : '—'}
               </span>
             </div>
-        </div>
+          </MetricCard>
 
-        {/* 服务名 */}
+          <MetricCard label="版本 / 运行时间">
-        <div className="border border-ink-900 bg-paper-50 p-5">
+            <div className="font-serif text-xl font-bold tabular-nums text-ink-900">
-          <div className="text-2xs tracking-[0.2em] text-ink-400">服务</div>
+              {health?.version ? String(health.version) : '—'}
-          <div className="mt-2 font-serif text-xl font-bold text-ink-900">
-            {health?.service || 'profeto'}
             </div>
+            <div className="mt-0.5 text-2xs tabular-nums text-ink-400">
+              {health?.uptime_seconds != null
+                ? `已运行 ${Math.floor(Number(health.uptime_seconds) / 3600)}h ${Math.floor((Number(health.uptime_seconds) % 3600) / 60)}m`
+                : '—'}
             </div>
+          </MetricCard>
 
-        {/* 版本 */}
+          <MetricCard label="上游 bzzoiro">
-        {health?.version && (
+            {upstream ? (
-          <div className="border border-ink-900 bg-paper-50 p-5">
+              <div>
-            <div className="text-2xs tracking-[0.2em] text-ink-400">版本</div>
+                <div className="flex items-center gap-2">
-            <div className="mt-2 font-serif text-xl font-bold tabular-nums text-ink-900">
+                  <span className={`inline-block h-2 w-2 ${upstream.ok ? 'bg-ok-500' : 'bg-press'}`} aria-hidden="true" />
-              {health.version}
+                  <span className={`font-serif text-xl font-bold ${upstream.ok ? 'text-ink-900' : 'text-press'}`}>
+                    {upstream.ok ? '可达' : '不可达'}
+                  </span>
+                </div>
+                <div className="mt-0.5 text-2xs tabular-nums text-ink-400">
+                  {upstream.ok
+                    ? `HTTP ${upstream.status_code} · ${upstream.latency_ms}ms`
+                    : upstream.error?.slice(0, 40) || `HTTP ${upstream.status_code}`}
                 </div>
               </div>
+            ) : (
+              <span className="font-serif text-xl text-ink-300">—</span>
             )}
+          </MetricCard>
+        </div>
+      </section>
 
-        {/* 运行时间 */}
+      {/* ── 数据管线 ── */}
-        {health?.uptime_seconds != null && (
+      <section>
-          <div className="border border-ink-900 bg-paper-50 p-5">
+        <h2 className="section-head mb-3">数据管线</h2>
-            <div className="text-2xs tracking-[0.2em] text-ink-400">运行时间</div>
+        <div className="grid gap-4 sm:grid-cols-2 lg:grid-cols-4">
-            <div className="mt-2 font-serif text-xl font-bold tabular-nums text-ink-900">
+          <MetricCard label="最近采集">
-              {Math.floor(health.uptime_seconds / 3600)}h{' '}
+            {recentJobs && runJobs > 0 ? (
-              {Math.floor((health.uptime_seconds % 3600) / 60)}m
+              <div className="font-serif text-xl font-bold tabular-nums text-ink-900">
-            </div>
+                {okJobs}/{runJobs} <span className="text-xs font-normal text-ink-500">成功</span>
               </div>
+            ) : (
+              <div className="font-serif text-xl text-ink-300">—</div>
             )}
+            <div className="mt-0.5 text-2xs text-ink-400">
+              {lastJobTime ? `最近 ${new Date(lastJobTime).toLocaleString('zh-CN', { month: '2-digit', day: '2-digit', hour: '2-digit', minute: '2-digit', hour12: false })}` : '暂无记录'}
+            </div>
+          </MetricCard>
 
-        {/* 检查项 */}
+          <MetricCard label="死信待处理">
-        {health?.checks && Object.keys(health.checks).length > 0 && (
+            <div className={`font-serif text-xl font-bold tabular-nums ${deadLetterCount > 0 ? 'text-press' : 'text-ink-900'}`}>
-          <div className="border border-ink-900 bg-paper-50 p-5 sm:col-span-2 lg:col-span-1">
+              {deadLetterCount}
-            <div className="text-2xs tracking-[0.2em] text-ink-400">健康检查</div>
-            <div className="mt-2 space-y-1">
-              {Object.entries(health.checks).map(([key, val]) => (
-                <div key={key} className="flex items-center justify-between text-sm">
-                  <span className="text-2xs text-ink-500">{key}</span>
-                  <Badge status={String(val) === 'pass' ? 'success' : 'error'}>
-                    {String(val)}
-                  </Badge>
             </div>
-              ))}
+            <div className="mt-0.5 text-2xs text-ink-400">失败记录,可重试</div>
+          </MetricCard>
+
+          <MetricCard label="数据缺口">
+            <div className={`font-serif text-xl font-bold tabular-nums ${missingStatsLeagues > 0 ? 'text-press' : 'text-ink-900'}`}>
+              {missingStatsLeagues}
             </div>
+            <div className="mt-0.5 text-2xs text-ink-400">联赛有完赛缺统计</div>
+          </MetricCard>
+
+          <MetricCard label="统计覆盖">
+            <div className="font-serif text-xl font-bold tabular-nums text-ink-900">
+              {completeness ? `${completeness.totals.stats_coverage_pct}%` : '—'}
             </div>
-        )}
+            <div className="mt-0.5 text-2xs text-ink-400">有统计 / 已完赛</div>
+          </MetricCard>
         </div>
+      </section>
+
+      {/* ── LLM ── */}
+      <section>
+        <h2 className="section-head mb-3">LLM 服务</h2>
+        <div className="grid gap-4 sm:grid-cols-3">
+          <MetricCard label="预测总数">
+            <div className="font-serif text-xl font-bold tabular-nums text-ink-900">
+              {llmStats ? llmStats.total_predictions : '—'}
+            </div>
+          </MetricCard>
+          <MetricCard label="平均延迟">
+            <div className="font-serif text-xl font-bold tabular-nums text-ink-900">
+              {llmStats && llmStats.avg_latency_ms > 0 ? `${(llmStats.avg_latency_ms / 1000).toFixed(1)}s` : '—'}
+            </div>
+          </MetricCard>
+          <MetricCard label="有效率">
+            <div className={`font-serif text-xl font-bold tabular-nums ${llmStats && llmStats.success_rate < 80 ? 'text-warn-700' : 'text-ink-900'}`}>
+              {llmStats ? `${llmStats.success_rate.toFixed(0)}%` : '—'}
+            </div>
+          </MetricCard>
+        </div>
+      </section>
     </div>
   )
 }

frontend/src/lib/http.test.ts

@@ -21,8 +21,10 @@ globalThis.window = {
 }
 
 // 捕获每次 fetch 的入参供断言
+let lastUrl = ''
 let lastInit: RequestInit | undefined
-globalThis.fetch = async (_url: string, init?: RequestInit) => {
+globalThis.fetch = async (url: string | URL | Request, init?: RequestInit) => {
+  lastUrl = String(url)
   lastInit = init
   return new Response(JSON.stringify({ ok: true }), { status: 200, headers: { 'content-type': 'application/json' } })
 }
@@ -63,3 +65,19 @@ test('DELETE: method=DELETE, 无 body, 无 Content-Type', async () => {
   assert.equal(lastInit?.body, undefined)
   assert.equal((lastInit?.headers as Record<string, string>)?.['Content-Type'], undefined)
 })
+
+// ── 健康检查豁免: /health* 挂在根路径(后端 app.py 不在 /api/v1 下),
+// 加前缀会 404 → 管理后台右上角永远「系统异常」 ──
+test('build_url: /health 与 /health/ready 不加 /api/v1 前缀', async () => {
+  await http.get('/health')
+  assert.equal(lastUrl, '/health')
+  await http.get('/health/ready')
+  assert.equal(lastUrl, '/health/ready')
+})
+
+test('build_url: 常规 API 路径仍统一加 /api/v1', async () => {
+  await http.get('/admin/stats')
+  assert.equal(lastUrl, '/api/v1/admin/stats')
+  await http.get('/matches')
+  assert.equal(lastUrl, '/api/v1/matches')
+})

frontend/src/lib/http.ts

@@ -17,22 +17,30 @@ export const UNAUTHORIZED_EVENT = 'profeto:unauthorized'
 const API_BASE = '/api/v1'
 const DEFAULT_TIMEOUT = 30_000
 
+/** 健康检查端点挂在根路径(app.py 不在 /api/v1 下,与 compose healthcheck 一致);
+ *  加前缀会 404,导致管理后台健康状态永远显示「系统异常」 */
+const ROOT_ONLY_PREFIXES = ['/health']
+
 /** 所有 API 路径统一走 /api/v1,避免浏览器直接请求 /matches 被 nginx 当 SPA 回退 */
 function build_url(path: string): string {
   if (path.startsWith('http')) return path
+  if (ROOT_ONLY_PREFIXES.some(p => path === p || path.startsWith(`${p}/`))) return path
   if (path.startsWith(API_BASE)) return path
   if (path.startsWith('/')) return `${API_BASE}${path}`
   return `${API_BASE}/${path}`
 }
 
 export class ApiError extends Error {
-  constructor(
+  status: number
-    message: string,
+  data?: unknown
-    public status: number,
+
-    public data?: unknown,
+  constructor(message: string, status: number, data?: unknown) {
-  ) {
     super(message)
     this.name = 'ApiError'
+    // 显式赋值而非构造函数参数属性(public x):strip-types 不支持后者,
+    // 会让 node --experimental-strip-types 跑 lib/http.test.ts 直接失败
+    this.status = status
+    this.data = data
   }
 }
 

src/api/app.py

@@ -3,6 +3,7 @@ from __future__ import annotations
 
 import logging
 import os
+import time
 from collections.abc import AsyncIterator
 from contextlib import asynccontextmanager
 
@@ -13,6 +14,9 @@ from src.core.config import settings
 
 logger = logging.getLogger(__name__)
 
+# 进程启动时间(monotonic,不受系统时钟跳变影响):/health 的 uptime 来源
+_PROCESS_STARTED_MONOTONIC = time.monotonic()
+
 
 async def _fail_stale_ingest_jobs() -> None:
     """P1-E: 启动时将上次遗留的 pending/running ingest_jobs 标 failed。
@@ -156,6 +160,7 @@ def create_app() -> FastAPI:
     from src.api.routes.auth import router as auth_router
     from src.api.routes.admin_settings import router as admin_settings_router
     from src.api.routes.schedules import router as schedules_router
+    from src.api.routes.admin_monitoring import router as admin_monitoring_router
 
     app.include_router(matches_router)
     app.include_router(predict_router)
@@ -165,10 +170,25 @@ def create_app() -> FastAPI:
     app.include_router(auth_router)
     app.include_router(admin_settings_router)
     app.include_router(schedules_router)
+    app.include_router(admin_monitoring_router)
 
     @app.get("/health")
     async def health():
-        return {"status": "healthy", "service": "profeto"}
+        """存活检查。version/uptime_seconds 供管理端监控页展示。"""
+        version = None
+        try:
+            from importlib.metadata import version as _pkg_version
+
+            version = _pkg_version("profeto")
+        except Exception:
+            # 包元数据缺失时返回 None,前端降级隐藏版本卡片
+            version = None
+        return {
+            "status": "healthy",
+            "service": "profeto",
+            "version": version,
+            "uptime_seconds": round(time.monotonic() - _PROCESS_STARTED_MONOTONIC),
+        }
 
 
     @app.get("/health/ready")

src/api/routes/admin_monitoring.py

@@ -0,0 +1,44 @@
+"""后台管理:监控增强探针(上游可达性)。
+
+所有接口需管理员鉴权(require_admin)。路由前缀 /api/v1/admin。
+"""
+from __future__ import annotations
+
+import logging
+import time
+
+import httpx
+from fastapi import APIRouter, Depends
+
+from src.api.deps import require_admin
+from src.core.config import settings
+from src.core.runtime_config import get_runtime_value
+
+logger = logging.getLogger(__name__)
+
+router = APIRouter(prefix="/api/v1/admin", tags=["admin"], dependencies=[Depends(require_admin)])
+
+
+@router.get("/monitoring/upstream")
+async def upstream_probe():
+    """上游数据源(bzzoiro)可达性探针。
+
+    轻量 GET 根端点:不携带 API Key、不触发采集,不消耗配额。
+    HTTP <500 视为可达 —— 404/401 也说明 DNS/网络/TLS 正常,业务语义层的
+    失败(签名、限流)由采集管线自身的死信与日志上报,不在探针职责内。
+    """
+    base = ((await get_runtime_value("BZZOIRO_BASE")) or settings.BZZOIRO_BASE).rstrip("/")
+    started = time.perf_counter()
+    try:
+        async with httpx.AsyncClient(
+            timeout=httpx.Timeout(connect=5.0, read=5.0, write=5.0, pool=5.0),
+        ) as client:
+            resp = await client.get(base)
+        latency_ms = round((time.perf_counter() - started) * 1000)
+        ok = resp.status_code < 500
+        logger.info("上游探针 %s → %s (%sms)", base, resp.status_code, latency_ms)
+        return {"ok": ok, "status_code": resp.status_code, "latency_ms": latency_ms, "endpoint": base}
+    except Exception as e:
+        latency_ms = round((time.perf_counter() - started) * 1000)
+        logger.warning("上游探针失败 %s: %s", base, e)
+        return {"ok": False, "error": str(e)[:200], "latency_ms": latency_ms, "endpoint": base}

src/api/routes/admin_quality.py

@@ -234,13 +234,14 @@ async def run_data_quality_check(db: AsyncSession = Depends(get_db_read)):
     checks = []
 
     # 检查1: 已完赛但无统计的比赛
+    # 注意: MatchStats 主键是 match_id(P0-02),不是 id —— 引用 .id 会 AttributeError
     finished_no_stats = (
         await db.execute(
             select(func.count())
             .select_from(Match)
             .outerjoin(MatchStats, Match.id == MatchStats.match_id)
             .where(Match.match_status == "finished")
-            .where(MatchStats.id.is_(None))
+            .where(MatchStats.match_id.is_(None))
         )
     ).scalar() or 0
 
@@ -276,6 +277,13 @@ async def run_data_quality_check(db: AsyncSession = Depends(get_db_read)):
         db.add(c)
     await db.commit()
 
+    # 成功路径留痕:检查何时跑过、各项结果如何(此前 handler 无任何日志,
+    # 加上未捕获异常走 uvicorn.error 不进内存缓冲,线上排障无据可查)
+    logger.info(
+        "数据质量检查完成: %s",
+        "; ".join(f"{c.check_name}={'通过' if c.passed else '未通过'}({c.actual_value:.0f})" for c in checks),
+    )
+
     return {"ok": True, "checks": [{"name": c.check_name, "passed": c.passed} for c in checks]}
 
 

src/core/log_buffer.py

@@ -80,6 +80,14 @@ def setup_logging(level: str = "INFO", log_file: str = "") -> None:
     if root.level == logging.NOTSET or root.level > logging.INFO:
         root.setLevel(getattr(logging, level.upper(), logging.INFO))
 
+    # uvicorn 的 logger 默认 propagate=False:未捕获异常的 traceback 只进
+    # stderr,不经过 root 的任何 handler —— Admin 日志页与文件日志都看不到,
+    # 线上 500 排障无据可查。打开 propagate 让它们进入内存缓冲/滚动文件。
+    for uv_name in ("uvicorn", "uvicorn.error"):
+        uv_logger = logging.getLogger(uv_name)
+        if not uv_logger.propagate:
+            uv_logger.propagate = True
+
     if not any(isinstance(h, MemoryLogHandler) for h in root.handlers):
         handler = MemoryLogHandler()
         handler.setLevel(logging.INFO)
@@ -112,6 +120,15 @@ def setup_logging(level: str = "INFO", log_file: str = "") -> None:
         file_handler.addFilter(_SQLNoiseFilter())
         root.addHandler(file_handler)
         logging.getLogger(__name__).info("文件日志已启用: %s", log_file)
+    except PermissionError:
+        # 容器场景最常见:挂载卷目录属主是 root,进程是非 root 用户。
+        # 修复方向:镜像内预创建目录并 chown(Dockerfile),或重建空卷。
+        logging.getLogger(__name__).warning(
+            "启用文件日志失败(%s):无写权限。容器部署请确认镜像已预创建该目录"
+            "并 chown 给运行用户(compose 卷挂载点默认 root 属主);"
+            "宿主机直跑请检查目录权限。本次仅保留 stdout/内存日志。",
+            log_file,
+        )
     except Exception:
         logging.getLogger(__name__).warning(
             "启用文件日志失败(%s),仅保留 stdout/内存日志", log_file, exc_info=True,

tests/test_data_quality_run.py

@@ -0,0 +1,66 @@
+"""数据质量检查回归测试。
+
+背景(P0-02 遗留): MatchStats 主键改为 match_id 后,质量检查查询仍引用
+MatchStats.id → AttributeError → POST /admin/data-quality/run 必然 500,
+前端显示「运行失败」;且 handler 无日志,未捕获异常走 uvicorn.error
+(propagate=False)不进内存缓冲/文件日志,排障时无据可查。
+"""
+from __future__ import annotations
+
+from unittest.mock import AsyncMock, MagicMock
+
+from src.api.routes import admin_quality
+
+
+def _fake_db(scalars: list[int]) -> MagicMock:
+    """按顺序返回 scalar() 计数的假 AsyncSession。"""
+    db = MagicMock()
+    results = []
+    for v in scalars:
+        r = MagicMock()
+        r.scalar.return_value = v
+        results.append(r)
+    db.execute = AsyncMock(side_effect=results)
+    db.add = MagicMock()
+    db.commit = AsyncMock()
+    return db
+
+
+async def test_run_data_quality_check_no_attribute_error():
+    """检查查询不得引用 MatchStats.id(P0-02 后该属性不存在)。
+
+    修复前: run_data_quality_check 抛 AttributeError → 500。
+    """
+    db = _fake_db([3, 1])
+
+    out = await admin_quality.run_data_quality_check(db)
+
+    assert out["ok"] is True
+    assert out["checks"] == [
+        {"name": "finished_without_stats", "passed": False},
+        {"name": "league_without_standings", "passed": False},
+    ]
+    # 检查结果落库(2 条 DataQualityCheck)
+    assert db.add.call_count == 2
+    db.commit.assert_awaited_once()
+
+
+async def test_run_data_quality_check_all_passed():
+    db = _fake_db([0, 0])
+
+    out = await admin_quality.run_data_quality_check(db)
+
+    assert out["ok"] is True
+    assert all(c["passed"] for c in out["checks"])
+
+
+async def test_run_data_quality_check_logs_summary(caplog):
+    """成功路径必须留日志:否则线上无从得知检查何时跑过、结果如何。"""
+    import logging
+
+    db = _fake_db([0, 0])
+
+    with caplog.at_level(logging.INFO, logger="src.api.routes.admin_quality"):
+        await admin_quality.run_data_quality_check(db)
+
+    assert any("数据质量检查" in r.message for r in caplog.records)

tests/test_monitoring_endpoints.py

@@ -0,0 +1,114 @@
+"""监控增强端点测试:/health 扩展字段 + 上游探针。
+
+沿用 test_api_critical.py 的模式:直接调用 handler,不启动完整 app
+lifespan(异步 DB 引擎与同步 TestClient 不兼容);探针的外呼用 mock。
+"""
+from __future__ import annotations
+
+from unittest.mock import AsyncMock, MagicMock, patch
+
+import httpx
+import pytest
+
+
+
+
+class TestHealthExtended:
+    async def _get_health(self):
+        """ASGITransport 走真实路由:/health 无鉴权、不依赖 DB,不触发 lifespan。"""
+        from httpx import ASGITransport, AsyncClient
+
+        from src.api.app import app
+
+        transport = ASGITransport(app=app)
+        async with AsyncClient(transport=transport, base_url="http://test") as client:
+            return await client.get("/health")
+
+    @pytest.mark.asyncio
+    async def test_reports_version_and_uptime(self):
+        resp = await self._get_health()
+
+        assert resp.status_code == 200
+        body = resp.json()
+        assert body["status"] == "healthy"
+        assert isinstance(body["uptime_seconds"], int)
+        assert body["uptime_seconds"] >= 0
+        # 包已随 pip install . 安装,元数据可读;异常时为 None(前端降级隐藏)
+        assert body["version"] is None or isinstance(body["version"], str)
+
+    @pytest.mark.asyncio
+    async def test_metadata_failure_degrades_to_none(self):
+        """包元数据不可读时 version=None,不影响存活判定。"""
+        from importlib.metadata import PackageNotFoundError
+
+        with patch("importlib.metadata.version", side_effect=PackageNotFoundError("profeto")):
+            resp = await self._get_health()
+
+        assert resp.status_code == 200
+        body = resp.json()
+        assert body["status"] == "healthy"
+        assert body["version"] is None
+        assert isinstance(body["uptime_seconds"], int)
+
+
+class TestUpstreamProbe:
+    async def _probe_with(self, mock_client_factory):
+        from src.api.routes import admin_monitoring
+
+        with patch.object(admin_monitoring.httpx, "AsyncClient", mock_client_factory), \
+             patch.object(
+                 admin_monitoring, "get_runtime_value",
+                 AsyncMock(return_value="https://sports.bzzoiro.com/api/v2"),
+             ):
+            return await admin_monitoring.upstream_probe()
+
+    @pytest.mark.asyncio
+    async def test_reachable_upstream(self):
+        """HTTP 200 → ok=True 且带延迟与状态码。"""
+        resp = MagicMock()
+        resp.status_code = 200
+        client = MagicMock()
+        client.get = AsyncMock(return_value=resp)
+        client.__aenter__ = AsyncMock(return_value=client)
+        client.__aexit__ = AsyncMock(return_value=False)
+        factory = MagicMock(return_value=client)
+
+        out = await self._probe_with(factory)
+
+        assert out["ok"] is True
+        assert out["status_code"] == 200
+        assert out["latency_ms"] >= 0
+        assert out["endpoint"].endswith("/api/v2")
+        client.get.assert_awaited_once_with("https://sports.bzzoiro.com/api/v2")
+
+    @pytest.mark.asyncio
+    async def test_5xx_counts_as_unreachable(self):
+        resp = MagicMock()
+        resp.status_code = 502
+        client = MagicMock()
+        client.get = AsyncMock(return_value=resp)
+        client.__aenter__ = AsyncMock(return_value=client)
+        client.__aexit__ = AsyncMock(return_value=False)
+        factory = MagicMock(return_value=client)
+
+        out = await self._probe_with(factory)
+
+        assert out["ok"] is False
+        assert out["status_code"] == 502
+
+    @pytest.mark.asyncio
+    async def test_network_error_degrades_gracefully(self):
+        """连接失败不抛 500:返回 ok=False + error 摘要(探针失败不是故障)。"""
+
+        def factory(*_a, **_kw):
+            client = MagicMock()
+            client.get = AsyncMock(side_effect=httpx.ConnectError("connection refused"))
+            client.__aenter__ = AsyncMock(return_value=client)
+            client.__aexit__ = AsyncMock(return_value=False)
+            return client
+
+        out = await self._probe_with(factory)
+
+        assert out["ok"] is False
+        assert "connection refused" in out["error"]
+        assert out["latency_ms"] >= 0