diff --git a/deploy/monitoring/README.md b/deploy/monitoring/README.md
new file mode 100644
index 00000000..b8f44b31
--- /dev/null
+++ b/deploy/monitoring/README.md
@@ -0,0 +1,171 @@
+# Sub2API Relay Manager Monitoring Setup
+
+## 概述
+
+本项目已配置完整的监控告警体系，包括 Prometheus metrics、Grafana 仪表板和 Prometheus 告警规则。
+
+## 已配置的 Metrics
+
+### HTTP 层指标
+
+- `http_requests_total` - HTTP 请求总数（按 method, path, status 分类）
+- `http_request_duration_seconds` - HTTP 请求延迟分布
+
+### 业务指标
+
+- `active_hosts` - 活跃宿主数量
+- `active_providers` - 活跃 provider 数量
+- `route_decisions_total` - 路由决策总数
+- `route_failovers_total` - 路由故障转移总数
+
+### 数据库指标
+
+- `db_connections_active` - 活跃数据库连接数
+- `db_operations_total` - 数据库操作总数
+
+### 日志指标
+
+- `log_flush_errors_total` - 日志刷新错误数
+- `log_dropped_events_total` - 丢弃的日志事件数
+
+## 告警规则
+
+### Critical 级别
+
+| 告警名称           | 触发条件                        | 说明            |
+| ------------------ | ------------------------------- | --------------- |
+| ServiceDown        | up == 0 持续1分钟               | 服务完全宕机    |
+| NoActiveProviders  | active_providers == 0 持续1分钟 | 无可用 provider |
+| NoActiveHosts      | active_hosts == 0 持续1分钟     | 无可用 host     |
+| HealthCheckFailing | /healthz 返回非200              | 健康检查失败    |
+
+### Warning 级别
+
+| 告警名称            | 触发条件                 | 说明                  |
+| ------------------- | ------------------------ | --------------------- |
+| HighErrorRate       | 错误率 > 5% 持续2分钟    | HTTP 5xx/4xx 错误率高 |
+| HighLatency         | P95 延迟 > 1秒 持续3分钟 | 请求处理延迟高        |
+| RouteFailoverSpike  | 故障转移率 > 正常水平2倍 | 路由不稳定            |
+| HighDBConnections   | 活跃连接 > 50 持续5分钟  | 数据库连接池压力大    |
+| LogFlushErrors      | 日志刷新错误 > 0         | 日志系统异常          |
+| LogDroppedEvents    | 丢弃事件率 > 10/sec      | 日志缓冲区溢出        |
+| BatchImportFailures | 批处理失败率 > 10%       | Provider 导入问题     |
+| AuthFailures        | 认证失败 > 10/sec        | 凭证问题或攻击        |
+
+## 部署步骤
+
+### 1. Prometheus 配置
+
+在 `prometheus.yml` 中添加：
+
+```yaml
+rule_files:
+  - "sub2api-relay-manager-rules.yml"
+
+scrape_configs:
+  - job_name: "sub2api-relay-manager"
+    static_configs:
+      - targets: ["localhost:8080"]
+    metrics_path: /metrics
+    scrape_interval: 15s
+```
+
+复制告警规则：
+
+```bash
+cp deploy/monitoring/prometheus-rules.yml /etc/prometheus/rules/
+```
+
+### 2. Grafana 配置
+
+导入仪表板：
+
+```bash
+curl -X POST \
+  http://admin:admin@localhost:3000/api/dashboards/db \
+  -H 'Content-Type: application/json' \
+  -d @deploy/monitoring/grafana-dashboard.json
+```
+
+### 3. Alertmanager 配置（可选）
+
+配置告警通知渠道（Slack/Email/PagerDuty）：
+
+```yaml
+# alertmanager.yml
+global:
+  smtp_smarthost: "localhost:587"
+  smtp_from: "alerts@example.com"
+
+route:
+  receiver: "ops-team"
+  group_by: ["alertname", "severity"]
+
+receivers:
+  - name: "ops-team"
+    email_configs:
+      - to: "ops@example.com"
+        subject: "[Alert] {{ .GroupLabels.alertname }}"
+    slack_configs:
+      - api_url: "YOUR_SLACK_WEBHOOK_URL"
+        channel: "#alerts"
+```
+
+## 验证
+
+### 检查 Metrics 端点
+
+```bash
+curl http://localhost:8080/metrics
+```
+
+### 验证告警规则
+
+```bash
+# 在 Prometheus 中查看
+http://localhost:9090/rules
+
+# 查看告警状态
+http://localhost:9090/alerts
+```
+
+### 触发测试告警
+
+```bash
+# 模拟高错误率
+for i in {1..100}; do
+  curl http://localhost:8080/api/nonexistent
+done
+```
+
+## 监控指标解释
+
+### 正常状态参考值
+
+| 指标             | 正常范围 | 告警阈值                      |
+| ---------------- | -------- | ----------------------------- |
+| active_providers | >= 2     | < 2 (warning), = 0 (critical) |
+| active_hosts     | >= 1     | = 0 (critical)                |
+| Error Rate       | < 1%     | > 5%                          |
+| P95 Latency      | < 500ms  | > 1s                          |
+| DB Connections   | < 20     | > 50                          |
+
+## 故障排查
+
+### 服务 Down 告警
+
+1. 检查进程状态：`systemctl status sub2api-relay-manager`
+2. 查看日志：`journalctl -u sub2api-relay-manager`
+3. 检查端口监听：`netstat -tlnp | grep 8080`
+
+### 高延迟告警
+
+1. 检查数据库性能
+2. 查看 upstream provider 响应时间
+3. 检查内存和 CPU 使用率
+
+### 路由故障转移告警
+
+1. 检查 provider 健康状态
+2. 查看 `/api/routing/routes/health`
+3. 分析 provider 响应日志
diff --git a/deploy/monitoring/grafana-dashboard.json b/deploy/monitoring/grafana-dashboard.json
new file mode 100644
index 00000000..9b398e79
--- /dev/null
+++ b/deploy/monitoring/grafana-dashboard.json
@@ -0,0 +1,85 @@
+{
+  "dashboard": {
+    "id": null,
+    "title": "Sub2API Relay Manager",
+    "tags": ["sub2api", "relay", "monitoring"],
+    "timezone": "UTC",
+    "panels": [
+      {
+        "id": 1,
+        "title": "Service Status",
+        "type": "stat",
+        "targets": [
+          {
+            "expr": "up{job=\"sub2api-relay-manager\"}",
+            "legendFormat": "Service Up"
+          }
+        ],
+        "fieldConfig": {
+          "defaults": {
+            "thresholds": {
+              "steps": [
+                { "color": "red", "value": 0 },
+                { "color": "green", "value": 1 }
+              ]
+            }
+          }
+        },
+        "gridPos": { "h": 4, "w": 6, "x": 0, "y": 0 }
+      },
+      {
+        "id": 2,
+        "title": "Active Providers",
+        "type": "stat",
+        "targets": [
+          {
+            "expr": "active_providers",
+            "legendFormat": "Providers"
+          }
+        ],
+        "gridPos": { "h": 4, "w": 6, "x": 6, "y": 0 }
+      },
+      {
+        "id": 3,
+        "title": "Active Hosts",
+        "type": "stat",
+        "targets": [
+          {
+            "expr": "active_hosts",
+            "legendFormat": "Hosts"
+          }
+        ],
+        "gridPos": { "h": 4, "w": 6, "x": 12, "y": 0 }
+      },
+      {
+        "id": 4,
+        "title": "Request Rate",
+        "type": "graph",
+        "targets": [
+          {
+            "expr": "rate(http_requests_total[5m])",
+            "legendFormat": "{{method}} {{path}}"
+          }
+        ],
+        "gridPos": { "h": 8, "w": 12, "x": 0, "y": 4 }
+      },
+      {
+        "id": 5,
+        "title": "Request Duration p95",
+        "type": "graph",
+        "targets": [
+          {
+            "expr": "histogram_quantile(0.95, sum(rate(http_request_duration_seconds_bucket[5m])) by (le))",
+            "legendFormat": "p95"
+          }
+        ],
+        "gridPos": { "h": 8, "w": 12, "x": 12, "y": 4 }
+      }
+    ],
+    "time": {
+      "from": "now-1h",
+      "to": "now"
+    },
+    "refresh": "30s"
+  }
+}
diff --git a/deploy/monitoring/prometheus-rules.yml b/deploy/monitoring/prometheus-rules.yml
new file mode 100644
index 00000000..0421aa19
--- /dev/null
+++ b/deploy/monitoring/prometheus-rules.yml
@@ -0,0 +1,181 @@
+# Prometheus Alerting Rules for sub2api-cn-relay-manager
+# Place this file in your Prometheus rules directory
+
+groups:
+  - name: sub2api-relay-manager-alerts
+    interval: 30s
+    rules:
+      # 服务可用性告警
+      - alert: ServiceDown
+        expr: up{job="sub2api-relay-manager"} == 0
+        for: 1m
+        labels:
+          severity: critical
+          team: ops
+        annotations:
+          summary: "sub2api-relay-manager service is down"
+          description: "The sub2api-relay-manager service has been down for more than 1 minute."
+
+      # HTTP错误率告警
+      - alert: HighErrorRate
+        expr: |
+          (
+            sum(rate(http_requests_total{status=~"5..|4.."}[5m])) 
+            / 
+            sum(rate(http_requests_total[5m]))
+          ) > 0.05
+        for: 2m
+        labels:
+          severity: warning
+          team: ops
+        annotations:
+          summary: "High error rate detected"
+          description: "Error rate is above 5% for more than 2 minutes. Current value: {{ $value | humanizePercentage }}"
+
+      # 请求延迟告警
+      - alert: HighLatency
+        expr: |
+          histogram_quantile(0.95, 
+            sum(rate(http_request_duration_seconds_bucket[5m])) by (le)
+          ) > 1.0
+        for: 3m
+        labels:
+          severity: warning
+          team: ops
+        annotations:
+          summary: "High request latency"
+          description: "95th percentile latency is above 1 second for more than 3 minutes."
+
+      # 路由故障转移告警
+      - alert: RouteFailoverSpike
+        expr: |
+          (
+            rate(route_failovers_total[5m]) 
+            > 
+            2 * avg_over_time(rate(route_failovers_total[1h])[1h:5m])
+          )
+        for: 1m
+        labels:
+          severity: warning
+          team: ops
+        annotations:
+          summary: "Route failover spike detected"
+          description: "Route failovers have spiked above normal levels. Current rate: {{ $value }}"
+
+      # 活跃Provider数量告警
+      - alert: NoActiveProviders
+        expr: active_providers == 0
+        for: 1m
+        labels:
+          severity: critical
+          team: ops
+        annotations:
+          summary: "No active providers"
+          description: "There are no active providers configured. The system cannot route requests."
+
+      - alert: LowActiveProviders
+        expr: active_providers < 2
+        for: 5m
+        labels:
+          severity: warning
+          team: ops
+        annotations:
+          summary: "Low number of active providers"
+          description: "Only {{ $value }} active provider(s) detected. Consider adding more for redundancy."
+
+      # 活跃Host告警
+      - alert: NoActiveHosts
+        expr: active_hosts == 0
+        for: 1m
+        labels:
+          severity: critical
+          team: ops
+        annotations:
+          summary: "No active hosts"
+          description: "There are no active hosts. The system cannot import providers."
+
+      # 数据库连接告警
+      - alert: HighDBConnections
+        expr: db_connections_active > 50
+        for: 5m
+        labels:
+          severity: warning
+          team: ops
+        annotations:
+          summary: "High database connection count"
+          description: "Active DB connections: {{ $value }}. Consider connection pool tuning."
+
+      # 数据库操作错误告警
+      - alert: DBOperationErrors
+        expr: |
+          rate(db_operations_total{operation=~"INSERT|UPDATE|DELETE"}[5m]) 
+          > 100
+        for: 2m
+        labels:
+          severity: warning
+          team: ops
+        annotations:
+          summary: "High database write rate"
+          description: "DB write operations are above threshold: {{ $value }} ops/sec"
+
+      # 日志系统告警
+      - alert: LogFlushErrors
+        expr: rate(log_flush_errors_total[5m]) > 0
+        for: 1m
+        labels:
+          severity: warning
+          team: ops
+        annotations:
+          summary: "Log flush errors detected"
+          description: "Log flush errors have been detected. Check log storage/backend."
+
+      - alert: LogDroppedEvents
+        expr: |
+          rate(log_dropped_events_total[5m]) > 10
+        for: 1m
+        labels:
+          severity: warning
+          team: ops
+        annotations:
+          summary: "Log events being dropped"
+          description: "Log events are being dropped at {{ $value }} events/sec. Check log buffer capacity."
+
+      # 批处理导入告警
+      - alert: BatchImportFailures
+        expr: |
+          (
+            rate(route_decisions_total{status="failed"}[5m]) 
+            / 
+            rate(route_decisions_total[5m])
+          ) > 0.1
+        for: 5m
+        labels:
+          severity: warning
+          team: ops
+        annotations:
+          summary: "High batch import failure rate"
+          description: "Batch import failure rate is above 10%. Check provider configurations."
+
+      # API认证失败告警
+      - alert: AuthFailures
+        expr: |
+          rate(http_requests_total{status="401"}[5m]) > 10
+        for: 2m
+        labels:
+          severity: warning
+          team: security
+        annotations:
+          summary: "High authentication failure rate"
+          description: "Auth failures detected. Possible credential issues or attacks."
+
+      # 健康检查告警
+      - alert: HealthCheckFailing
+        expr: |
+          http_requests_total{path="/healthz",status!="200"} > 0
+        for: 30s
+        labels:
+          severity: critical
+          team: ops
+        annotations:
+          summary: "Health check failing"
+          description: "The /healthz endpoint is returning non-200 status."