refactor: config/apps 目录重组、文档重构、架构图收窄

- 中央：config/(prometheus,alertmanager,grafana)、apps/(tile-cache,topology-editor)
- 边缘：config/(vmagent,blackbox,targets)、apps/(onvif-exporter)
- env: TRAEFIK_PROVIDER、prometheus/env.example 详细说明
- 文档：README/doc 重构，EDGE_CACHE 合并到 EDGE_AGENT_CONFIG
- targets.csv 更新流程说明，ARCHITECTURE 图收窄

Made-with: Cursor

central-server/config/prometheus/alert_rules.yml

@@ -0,0 +1,49 @@
+groups:
+- name: onvif_alerts
+  rules:
+  - alert: ONVIFDeviceDown
+    expr: up{job="onvif-devices"} == 0
+    for: 1m
+    labels:
+      severity: critical
+    annotations:
+      summary: "ONVIF设备离线"
+      description: "ONVIF设备 {{ $labels.instance }} 已离线超过1分钟"
+
+  - alert: ONVIFDeviceHighTemperature
+    expr: onvif_device_temperature > 70
+    for: 2m
+    labels:
+      severity: warning
+    annotations:
+      summary: "ONVIF设备温度过高"
+      description: "设备 {{ $labels.instance }} 温度达到 {{ $value }}°C"
+
+  - alert: ONVIFDeviceLowStorage
+    expr: onvif_storage_usage_percent > 90
+    for: 5m
+    labels:
+      severity: warning
+    annotations:
+      summary: "ONVIF设备存储空间不足"
+      description: "设备 {{ $labels.instance }} 存储使用率达到 {{ $value }}%"
+
+- name: network_alerts
+  rules:
+  - alert: NetworkDeviceDown
+    expr: probe_success{job="network-ping"} == 0
+    for: 2m
+    labels:
+      severity: critical
+    annotations:
+      summary: "网络设备离线"
+      description: "网络设备 {{ $labels.instance }} 无法ping通"
+
+  - alert: HighNetworkLatency
+    expr: probe_duration_seconds{job="network-ping"} > 1
+    for: 5m
+    labels:
+      severity: warning
+    annotations:
+      summary: "网络延迟过高"
+      description: "设备 {{ $labels.instance }} 延迟达到 {{ $value }}秒"

central-server/config/prometheus/prometheus.yml

@@ -0,0 +1,67 @@
+# Prometheus 中央服务器配置模板
+# ============================================
+# 说明：
+#   - 本文件为配置模板，包含 ${变量名} 占位符
+#   - 部署时由 deploy.sh 从 .env 读取变量，用 envsubst 生成 prometheus.yml
+#   - 请勿直接编辑 prometheus.yml，修改应在此模板或 .env 中进行
+#
+# 变量来源：central-server/.env（参考 env.example）；本文件位于 config/prometheus/
+# 涉及变量：PROMETHEUS_SCRAPE_INTERVAL, PROMETHEUS_EVALUATION_INTERVAL,
+#          PROMETHEUS_CLUSTER_NAME, VICTORIAMETRICS_PORT,
+#          PROMETHEUS_REMOTE_WRITE_MAX_SAMPLES, PROMETHEUS_REMOTE_WRITE_CAPACITY,
+#          PROMETHEUS_REMOTE_WRITE_MAX_SHARDS
+# ============================================
+
+global:
+  scrape_interval: 15s
+  evaluation_interval: 15s
+  external_labels:
+    cluster: 'central-monitoring'
+
+# 远程写入：将中央 Prometheus 抓取到的本地服务指标推送到 VictoriaMetrics
+# （边缘节点由 vmagent 直接 remote_write 到 VictoriaMetrics）
+remote_write:
+  - url: http://victoria-metrics:8428/api/v1/write
+    queue_config:
+      max_samples_per_send: 10000
+      capacity: 20000
+      max_shards: 10
+
+# 抓取配置：仅抓取中央本机 Docker 容器（Prometheus/VM/Alertmanager/Grafana）
+scrape_configs:
+  # 抓取中央Prometheus自身
+  - job_name: 'prometheus-central'
+    scrape_interval: 15s
+    static_configs:
+      - targets: ['prometheus-central:9090']
+
+  # 抓取VictoriaMetrics (VictoriaMetrics 提供 /metrics 端点)
+  - job_name: 'victoria-metrics'
+    scrape_interval: 15s
+    metrics_path: '/metrics'
+    static_configs:
+      - targets: ['victoria-metrics:8428']
+
+  # 抓取Alertmanager
+  - job_name: 'alertmanager'
+    scrape_interval: 15s
+    static_configs:
+      - targets: ['alertmanager:9093']
+
+  # 抓取Grafana (需要启用 metrics 功能)
+  - job_name: 'grafana'
+    scrape_interval: 15s
+    metrics_path: '/metrics'
+    static_configs:
+      - targets: ['grafana:3000']
+
+# 告警规则：alert_rules.yml 与 prometheus.yml 同目录
+rule_files:
+  - "alert_rules.yml"
+
+# Alertmanager：告警路由与静默
+alerting:
+  alertmanagers:
+    - static_configs:
+        - targets:
+          - alertmanager:9093

central-server/config/prometheus/prometheus.yml.template

@@ -0,0 +1,67 @@
+# Prometheus 中央服务器配置模板
+# ============================================
+# 说明：
+#   - 本文件为配置模板，包含 ${变量名} 占位符
+#   - 部署时由 deploy.sh 从 .env 读取变量，用 envsubst 生成 prometheus.yml
+#   - 请勿直接编辑 prometheus.yml，修改应在此模板或 .env 中进行
+#
+# 变量来源：central-server/.env（参考 env.example）；本文件位于 config/prometheus/
+# 涉及变量：PROMETHEUS_SCRAPE_INTERVAL, PROMETHEUS_EVALUATION_INTERVAL,
+#          PROMETHEUS_CLUSTER_NAME, VICTORIAMETRICS_PORT,
+#          PROMETHEUS_REMOTE_WRITE_MAX_SAMPLES, PROMETHEUS_REMOTE_WRITE_CAPACITY,
+#          PROMETHEUS_REMOTE_WRITE_MAX_SHARDS
+# ============================================
+
+global:
+  scrape_interval: ${PROMETHEUS_SCRAPE_INTERVAL}s
+  evaluation_interval: ${PROMETHEUS_EVALUATION_INTERVAL}s
+  external_labels:
+    cluster: '${PROMETHEUS_CLUSTER_NAME}'
+
+# 远程写入：将中央 Prometheus 抓取到的本地服务指标推送到 VictoriaMetrics
+# （边缘节点由 vmagent 直接 remote_write 到 VictoriaMetrics）
+remote_write:
+  - url: http://victoria-metrics:${VICTORIAMETRICS_PORT}/api/v1/write
+    queue_config:
+      max_samples_per_send: ${PROMETHEUS_REMOTE_WRITE_MAX_SAMPLES}
+      capacity: ${PROMETHEUS_REMOTE_WRITE_CAPACITY}
+      max_shards: ${PROMETHEUS_REMOTE_WRITE_MAX_SHARDS}
+
+# 抓取配置：仅抓取中央本机 Docker 容器（Prometheus/VM/Alertmanager/Grafana）
+scrape_configs:
+  # 抓取中央Prometheus自身
+  - job_name: 'prometheus-central'
+    scrape_interval: ${PROMETHEUS_SCRAPE_INTERVAL}s
+    static_configs:
+      - targets: ['prometheus-central:9090']
+
+  # 抓取VictoriaMetrics (VictoriaMetrics 提供 /metrics 端点)
+  - job_name: 'victoria-metrics'
+    scrape_interval: ${PROMETHEUS_SCRAPE_INTERVAL}s
+    metrics_path: '/metrics'
+    static_configs:
+      - targets: ['victoria-metrics:${VICTORIAMETRICS_PORT}']
+
+  # 抓取Alertmanager
+  - job_name: 'alertmanager'
+    scrape_interval: ${PROMETHEUS_SCRAPE_INTERVAL}s
+    static_configs:
+      - targets: ['alertmanager:9093']
+
+  # 抓取Grafana (需要启用 metrics 功能)
+  - job_name: 'grafana'
+    scrape_interval: ${PROMETHEUS_SCRAPE_INTERVAL}s
+    metrics_path: '/metrics'
+    static_configs:
+      - targets: ['grafana:3000']
+
+# 告警规则：alert_rules.yml 与 prometheus.yml 同目录
+rule_files:
+  - "alert_rules.yml"
+
+# Alertmanager：告警路由与静默
+alerting:
+  alertmanagers:
+    - static_configs:
+        - targets:
+          - alertmanager:9093