完善中央与边缘部署、远程写入与监控文档

- 增加中央与边缘完整配置和部署脚本
- 引入 VictoriaMetrics 数据源与 remote_write 故障排查说明
- 新增 edge-agent 配置脚本、ONVIF 自建 exporter 与 ping 监控示例

Co-authored-by: Cursor <cursoragent@cursor.com>

edge-agent/blackbox/config.yml

@@ -0,0 +1,33 @@
+modules:
+  icmp:
+    prober: icmp
+    timeout: 5s
+    icmp:
+      preferred_ip_protocol: "ip4"
+  tcp_connect:
+    prober: tcp
+    timeout: 5s
+  http_2xx:
+    prober: http
+    timeout: 5s
+    http:
+      valid_http_versions: ["HTTP/1.1", "HTTP/2.0"]
+      valid_status_codes: [200]
+      method: GET
+  http_post_2xx:
+    prober: http
+    timeout: 5s
+    http:
+      valid_http_versions: ["HTTP/1.1", "HTTP/2.0"]
+      valid_status_codes: [200]
+      method: POST
+      headers:
+        Content-Type: application/json
+      body: '{}'
+  tcp_connect_tls:
+    prober: tcp
+    timeout: 5s
+    tcp:
+      tls: true
+      tls_config:
+        insecure_skip_verify: false

edge-agent/config/csv-to-json.sh

@@ -0,0 +1,63 @@
+#!/bin/bash
+
+# CSV转JSON脚本 - 将设备CSV表格转换为Prometheus监控JSON配置
+# 使用方法: ./csv-to-json.sh devices.csv > onvif-targets.json
+
+set -e
+
+CSV_FILE=${1:-"devices.csv"}
+OUTPUT_FILE=${2:-"onvif-targets.json"}
+
+# 检查jq是否安装
+if ! command -v jq &> /dev/null; then
+    echo "❌ jq未安装，请先安装jq:"
+    echo "   Ubuntu/Debian: sudo apt-get install jq"
+    echo "   CentOS/RHEL: sudo yum install jq"
+    echo "   Alpine: apk add jq"
+    exit 1
+fi
+
+# 检查CSV文件是否存在
+if [ ! -f "$CSV_FILE" ]; then
+    echo "❌ CSV文件 $CSV_FILE 不存在"
+    exit 1
+fi
+
+echo "🔄 正在转换 $CSV_FILE 到 $OUTPUT_FILE..."
+
+# 使用jq将CSV转换为JSON
+# 1. 读取CSV文件，跳过标题行
+# 2. 将每行转换为JSON对象
+# 3. 构建Prometheus targets格式
+
+tail -n +2 "$CSV_FILE" | while IFS=',' read -r ip device_type model location username password onvif_port; do
+    # 构建labels对象
+    labels="{
+        \"device_type\": \"$device_type\",
+        \"model\": \"$model\",
+        \"location\": \"$location\",
+        \"username\": \"$username\",
+        \"password\": \"$password\""
+    
+    # 如果onvif_port不是默认的80，则添加到labels中
+    if [ "$onvif_port" != "80" ] && [ -n "$onvif_port" ]; then
+        labels="$labels,
+        \"onvif_port\": \"$onvif_port\""
+    fi
+    
+    labels="$labels
+    }"
+    
+    # 输出JSON对象
+    echo "{
+        \"targets\": [\"$ip\"],
+        \"labels\": $labels
+    }"
+done | jq -s '.' > "$OUTPUT_FILE"
+
+echo "✅ 转换完成！"
+echo "📊 生成了 $(jq length "$OUTPUT_FILE") 个设备配置"
+echo "📁 输出文件: $OUTPUT_FILE"
+echo ""
+echo "🔍 预览生成的JSON:"
+jq . "$OUTPUT_FILE"

edge-agent/config/csv-to-ping-json.sh

@@ -0,0 +1,61 @@
+#!/bin/bash
+
+# CSV转Ping JSON脚本 - 将Ping目标CSV表格转换为Prometheus监控JSON配置
+# 使用方法: ./csv-to-ping-json.sh ping-targets.csv > ping-targets.json
+
+set -e
+
+CSV_FILE=${1:-"ping-targets.csv"}
+OUTPUT_FILE=${2:-"ping-targets.json"}
+
+# 检查jq是否安装
+if ! command -v jq &> /dev/null; then
+    echo "❌ jq未安装，请先安装jq:"
+    echo "   Ubuntu/Debian: sudo apt-get install jq"
+    echo "   CentOS/RHEL: sudo yum install jq"
+    echo "   Alpine: apk add jq"
+    exit 1
+fi
+
+# 检查CSV文件是否存在
+if [ ! -f "$CSV_FILE" ]; then
+    echo "❌ CSV文件 $CSV_FILE 不存在"
+    exit 1
+fi
+
+echo "🔄 正在转换 $CSV_FILE 到 $OUTPUT_FILE..."
+
+# 使用jq将CSV转换为JSON
+tail -n +2 "$CSV_FILE" | while IFS=',' read -r ip device group network; do
+    # 构建labels对象
+    labels="{
+        \"device\": \"$device\""
+    
+    # 添加可选的group标签
+    if [ -n "$group" ]; then
+        labels="$labels,
+        \"group\": \"$group\""
+    fi
+    
+    # 添加可选的network标签
+    if [ -n "$network" ]; then
+        labels="$labels,
+        \"network\": \"$network\""
+    fi
+    
+    labels="$labels
+    }"
+    
+    # 输出JSON对象
+    echo "{
+        \"targets\": [\"$ip\"],
+        \"labels\": $labels
+    }"
+done | jq -s '.' > "$OUTPUT_FILE"
+
+echo "✅ 转换完成！"
+echo "📊 生成了 $(jq length "$OUTPUT_FILE") 个Ping目标配置"
+echo "📁 输出文件: $OUTPUT_FILE"
+echo ""
+echo "🔍 预览生成的JSON:"
+jq . "$OUTPUT_FILE"

edge-agent/config/csv-to-targets.sh

@@ -0,0 +1,129 @@
+#!/bin/bash
+
+# 统一目标配置转换脚本
+# 从 targets.csv 生成 onvif-targets.json 和 ping-targets.json
+# 使用方法: ./csv-to-targets.sh targets.csv
+
+set -e
+
+CSV_FILE=${1:-"targets.csv"}
+
+# 检查jq是否安装
+if ! command -v jq &> /dev/null; then
+    echo "❌ jq未安装，请先安装jq:"
+    echo "   Ubuntu/Debian: sudo apt-get install jq"
+    echo "   CentOS/RHEL: sudo yum install jq"
+    echo "   Fedora: sudo dnf install jq"
+    exit 1
+fi
+
+# 检查CSV文件是否存在
+if [ ! -f "$CSV_FILE" ]; then
+    echo "❌ CSV文件 $CSV_FILE 不存在"
+    exit 1
+fi
+
+echo "🔄 正在从 $CSV_FILE 生成配置文件..."
+echo ""
+
+# 临时文件
+ONVIF_TEMP=$(mktemp)
+PING_TEMP=$(mktemp)
+
+# 处理CSV文件（跳过注释行和标题行）
+tail -n +2 "$CSV_FILE" | grep -v '^#' | while IFS=',' read -r type ip device group network device_type model location username password onvif_port; do
+    # 去除空格
+    type=$(echo "$type" | xargs)
+    ip=$(echo "$ip" | xargs)
+    
+    # 跳过空行
+    if [ -z "$type" ] || [ -z "$ip" ]; then
+        continue
+    fi
+    
+    if [ "$type" = "onvif" ]; then
+        # 处理 ONVIF 设备
+        device_type=$(echo "$device_type" | xargs)
+        model=$(echo "$model" | xargs)
+        location=$(echo "$location" | xargs)
+        username=$(echo "$username" | xargs)
+        password=$(echo "$password" | xargs)
+        onvif_port=$(echo "$onvif_port" | xargs)
+        
+        labels="{
+            \"device_type\": \"$device_type\",
+            \"model\": \"$model\",
+            \"location\": \"$location\",
+            \"username\": \"$username\",
+            \"password\": \"$password\""
+        
+        if [ "$onvif_port" != "80" ] && [ -n "$onvif_port" ]; then
+            labels="$labels,
+            \"onvif_port\": \"$onvif_port\""
+        fi
+        
+        labels="$labels
+        }"
+        
+        echo "{
+            \"targets\": [\"$ip\"],
+            \"labels\": $labels
+        }" >> "$ONVIF_TEMP"
+        
+    elif [ "$type" = "ping" ]; then
+        # 处理 Ping 目标
+        device=$(echo "$device" | xargs)
+        group=$(echo "$group" | xargs)
+        network=$(echo "$network" | xargs)
+        
+        labels="{
+            \"device\": \"$device\""
+        
+        if [ -n "$group" ]; then
+            labels="$labels,
+            \"group\": \"$group\""
+        fi
+        
+        if [ -n "$network" ]; then
+            labels="$labels,
+            \"network\": \"$network\""
+        fi
+        
+        labels="$labels
+        }"
+        
+        echo "{
+            \"targets\": [\"$ip\"],
+            \"labels\": $labels
+        }" >> "$PING_TEMP"
+    fi
+done
+
+# 生成 JSON 文件
+if [ -s "$ONVIF_TEMP" ]; then
+    jq -s '.' "$ONVIF_TEMP" > onvif-targets.json
+    ONVIF_COUNT=$(jq length onvif-targets.json)
+    echo "✅ 生成 ONVIF 设备配置: $ONVIF_COUNT 个设备"
+else
+    echo "[]" > onvif-targets.json
+    echo "⚠️  未找到 ONVIF 设备，生成空配置"
+fi
+
+if [ -s "$PING_TEMP" ]; then
+    jq -s '.' "$PING_TEMP" > ping-targets.json
+    PING_COUNT=$(jq length ping-targets.json)
+    echo "✅ 生成 Ping 目标配置: $PING_COUNT 个目标"
+else
+    echo "[]" > ping-targets.json
+    echo "⚠️  未找到 Ping 目标，生成空配置"
+fi
+
+# 清理临时文件
+rm -f "$ONVIF_TEMP" "$PING_TEMP"
+
+echo ""
+echo "✅ 配置文件生成完成！"
+echo "📁 生成的文件："
+echo "   - onvif-targets.json"
+echo "   - ping-targets.json"
+echo ""

edge-agent/config/devices.csv

@@ -0,0 +1,5 @@
+ip,device_type,model,location,username,password,onvif_port
+192.168.1.100,camera,HIKVISION_DS-2CD2342WD-I,front_door,admin,password1,80
+192.168.1.101,camera,DAHUA_IPC-HFW1230S,back_yard,admin,password2,80
+192.168.1.102,camera,UNIVIEW_IPC3120SR,living_room,admin,password3,8080
+192.168.1.50,nvr,HIKVISION_DS-7608NI-I2,server_rack,admin,password4,80

edge-agent/config/ping-targets.csv

@@ -0,0 +1,6 @@
+ip,device,group,network
+192.168.1.1,main_router,network,internal
+192.168.1.100,front_camera,onvif_cameras,internal
+192.168.1.101,back_camera,onvif_cameras,internal
+192.168.1.102,living_camera,onvif_cameras,internal
+8.8.8.8,google_dns,external,external

edge-agent/config/setup-remote-write.sh

@@ -0,0 +1,63 @@
+#!/bin/bash
+
+# 设置远程推送地址脚本
+# 使用方法: ./setup-remote-write.sh <中央服务器地址> [端口]
+# 支持IP地址和域名
+
+set -e
+
+CENTRAL_HOST=${1:-"192.168.1.10"}
+CENTRAL_PORT=${2:-"8428"}
+CONFIG_FILE="../prometheus-edge/prometheus.yml"
+
+echo "=== 设置Prometheus远程推送地址 ==="
+echo ""
+
+if [ -z "$1" ]; then
+    echo "使用方法: $0 <中央服务器地址> [端口]"
+    echo "示例: $0 192.168.1.10 8428"
+    echo "      $0 prometheus.company.com"
+    echo "      $0 prometheus.local 8428"
+    echo ""
+    read -p "请输入中央服务器地址 (IP或域名): " CENTRAL_HOST
+    read -p "请输入端口 (默认8428): " CENTRAL_PORT_INPUT
+    if [ -n "$CENTRAL_PORT_INPUT" ]; then
+        CENTRAL_PORT=$CENTRAL_PORT_INPUT
+    fi
+fi
+
+echo "🔧 配置信息："
+echo "   中央服务器地址: $CENTRAL_HOST"
+echo "   端口: $CENTRAL_PORT"
+echo "   配置文件: $CONFIG_FILE"
+echo ""
+
+# 检查配置文件是否存在
+if [ ! -f "$CONFIG_FILE" ]; then
+    echo "❌ 配置文件 $CONFIG_FILE 不存在"
+    exit 1
+fi
+
+# 备份原配置文件
+cp "$CONFIG_FILE" "${CONFIG_FILE}.backup.$(date +%Y%m%d_%H%M%S)"
+echo "📋 已备份原配置文件"
+
+# 更新配置文件中的远程推送地址
+sed -i "s|http://\${CENTRAL_SERVER_HOST}:\${CENTRAL_SERVER_PORT}|http://$CENTRAL_HOST:$CENTRAL_PORT|g" "$CONFIG_FILE"
+
+echo "✅ 远程推送地址已更新"
+echo ""
+
+# 显示更新后的配置
+echo "🔍 更新后的remote_write配置："
+grep -A 5 "remote_write:" "$CONFIG_FILE"
+
+echo ""
+echo "🔄 重启Prometheus服务以应用新配置："
+echo "   docker-compose restart prometheus-edge"
+echo ""
+echo "📊 检查远程写入状态："
+echo "   curl http://localhost:9090/api/v1/status/config"
+echo ""
+echo "🔗 查看远程写入目标："
+echo "   curl http://localhost:9090/api/v1/status/tsdb"

edge-agent/config/targets.csv

@@ -0,0 +1,13 @@
+# 统一监控目标配置文件
+# 格式: type,ip,device,group,network,device_type,model,location,username,password,onvif_port
+# type: onvif 或 ping
+# 对于 onvif 类型，需要填写: ip,device_type,model,location,username,password,onvif_port
+# 对于 ping 类型，需要填写: ip,device,group,network
+type,ip,device,group,network,device_type,model,location,username,password,onvif_port
+ping,192.168.2.1,main_router,network,internal,,,,,,
+ping,8.8.8.8,google_dns,external,external,,,,,,
+# ONVIF 设备示例（取消注释并填写实际信息）
+# onvif,192.168.1.100,,,front_door,camera,HIKVISION_DS-2CD2342WD-I,front_door,admin,password1,80
+# onvif,192.168.1.101,,,back_yard,camera,DAHUA_IPC-HFW1230S,back_yard,admin,password2,80
+# onvif,192.168.1.102,,,living_room,camera,UNIVIEW_IPC3120SR,living_room,admin,password3,8080
+# onvif,192.168.1.50,,,server_rack,nvr,HIKVISION_DS-7608NI-I2,server_rack,admin,password4,80

edge-agent/config/test-connection.sh

@@ -0,0 +1,90 @@
+#!/bin/bash
+
+# 测试中央服务器连接脚本
+# 使用方法: ./test-connection.sh <中央服务器地址> [端口]
+
+set -e
+
+CENTRAL_HOST=${1:-"192.168.1.10"}
+CENTRAL_PORT=${2:-"8428"}
+
+echo "=== 测试中央服务器连接 ==="
+echo ""
+
+if [ -z "$1" ]; then
+    echo "使用方法: $0 <中央服务器地址> [端口]"
+    echo "示例: $0 192.168.1.10 8428"
+    echo "      $0 prometheus.company.com"
+    echo "      $0 prometheus.local 9090"
+    echo ""
+    read -p "请输入中央服务器地址 (IP或域名): " CENTRAL_HOST
+    read -p "请输入端口 (默认8428): " CENTRAL_PORT_INPUT
+    if [ -n "$CENTRAL_PORT_INPUT" ]; then
+        CENTRAL_PORT=$CENTRAL_PORT_INPUT
+    fi
+fi
+
+echo "🔧 测试配置："
+echo "   中央服务器地址: $CENTRAL_HOST"
+echo "   端口: $CENTRAL_PORT"
+echo ""
+
+# 测试域名解析
+echo "🌐 测试域名解析..."
+if command -v nslookup &> /dev/null; then
+    nslookup $CENTRAL_HOST
+else
+    echo "   nslookup 不可用，跳过DNS测试"
+fi
+
+# 测试网络连通性
+echo ""
+echo "📡 测试网络连通性..."
+if ping -c 3 $CENTRAL_HOST > /dev/null 2>&1; then
+    echo "   ✅ Ping 成功"
+else
+    echo "   ❌ Ping 失败"
+fi
+
+# 测试端口连通性
+echo ""
+echo "🔌 测试端口连通性..."
+if command -v nc &> /dev/null; then
+    if nc -z $CENTRAL_HOST $CENTRAL_PORT 2>/dev/null; then
+        echo "   ✅ 端口 $CENTRAL_PORT 可访问"
+    else
+        echo "   ❌ 端口 $CENTRAL_PORT 不可访问"
+    fi
+else
+    echo "   nc 不可用，跳过端口测试"
+fi
+
+# 测试HTTP连接
+echo ""
+echo "🌐 测试HTTP连接..."
+HTTP_URL="http://$CENTRAL_HOST:$CENTRAL_PORT"
+if command -v curl &> /dev/null; then
+    if curl -s --connect-timeout 5 $HTTP_URL > /dev/null 2>&1; then
+        echo "   ✅ HTTP连接成功: $HTTP_URL"
+        
+        # 测试VictoriaMetrics API
+        if curl -s --connect-timeout 5 "$HTTP_URL/api/v1/status" > /dev/null 2>&1; then
+            echo "   ✅ VictoriaMetrics API 可访问"
+        else
+            echo "   ⚠️  VictoriaMetrics API 不可访问 (可能不是VictoriaMetrics服务)"
+        fi
+    else
+        echo "   ❌ HTTP连接失败: $HTTP_URL"
+    fi
+else
+    echo "   curl 不可用，跳过HTTP测试"
+fi
+
+echo ""
+echo "📋 测试完成！"
+echo ""
+echo "💡 如果连接失败，请检查："
+echo "   1. 网络连接是否正常"
+echo "   2. 防火墙是否开放端口 $CENTRAL_PORT"
+echo "   3. 中央服务器是否正在运行"
+echo "   4. DNS解析是否正确"

edge-agent/config/update-configs.sh

@@ -0,0 +1,68 @@
+#!/bin/bash
+
+# 更新配置文件脚本 - 从CSV生成所有JSON配置文件
+# 使用方法: ./update-configs.sh
+
+set -e
+
+echo "=== 更新Prometheus监控配置文件 ==="
+echo ""
+
+# 检查jq是否安装
+if ! command -v jq &> /dev/null; then
+    echo "❌ jq未安装，请先安装jq:"
+    echo "   Ubuntu/Debian: sudo apt-get install jq"
+    echo "   CentOS/RHEL: sudo yum install jq"
+    echo "   Alpine: apk add jq"
+    exit 1
+fi
+
+# 进入脚本目录
+cd "$(dirname "$0")"
+
+echo "🔄 正在从CSV文件生成JSON配置..."
+
+# 优先使用统一的 targets.csv
+if [ -f "targets.csv" ]; then
+    echo "📋 使用统一配置文件 targets.csv..."
+    chmod +x csv-to-targets.sh 2>/dev/null || true
+    ./csv-to-targets.sh targets.csv
+else
+    echo "⚠️  targets.csv 不存在，使用旧格式配置文件..."
+    echo ""
+    
+    # 兼容旧格式：生成ONVIF设备配置
+    if [ -f "devices.csv" ]; then
+        echo "📱 生成ONVIF设备配置（从 devices.csv）..."
+        chmod +x csv-to-json.sh 2>/dev/null || true
+        ./csv-to-json.sh devices.csv onvif-targets.json
+    else
+        echo "⚠️  devices.csv 不存在，跳过ONVIF设备配置生成"
+        echo "[]" > onvif-targets.json
+    fi
+    
+    # 兼容旧格式：生成Ping目标配置
+    if [ -f "ping-targets.csv" ]; then
+        echo "🌐 生成Ping目标配置（从 ping-targets.csv）..."
+        chmod +x csv-to-ping-json.sh 2>/dev/null || true
+        ./csv-to-ping-json.sh ping-targets.csv ping-targets.json
+    else
+        echo "⚠️  ping-targets.csv 不存在，跳过Ping目标配置生成"
+        echo "[]" > ping-targets.json
+    fi
+fi
+
+echo ""
+echo "✅ 所有配置文件已更新！"
+echo ""
+echo "📋 生成的文件："
+ls -la *.json 2>/dev/null || echo "   (无JSON文件生成)"
+echo ""
+echo "🔄 配置热重载："
+echo "   - Prometheus会在5分钟内自动检测并重载配置"
+echo "   - 无需重启Docker容器！"
+echo ""
+echo "⚡ 强制立即重载 (可选)："
+echo "   docker-compose restart prometheus-edge"
+echo ""
+echo "📝 编辑CSV文件后重新运行此脚本即可更新配置"

edge-agent/deploy.sh

@@ -0,0 +1,156 @@
+#!/bin/bash
+
+# 分布式Prometheus边缘代理部署脚本
+# 适用于Linux系统 (玩客云等设备)
+
+set -e
+
+echo "=== 分布式Prometheus边缘代理部署脚本 ==="
+echo ""
+
+# 检查Docker是否安装
+if ! command -v docker &> /dev/null; then
+    echo "❌ Docker未安装，请先安装Docker"
+    exit 1
+fi
+
+# 检查Docker Compose (优先检查V2，然后检查V1)
+DOCKER_COMPOSE_CMD=""
+if docker compose version &> /dev/null; then
+    DOCKER_COMPOSE_CMD="docker compose"
+    echo "✅ 检测到 Docker Compose V2"
+elif command -v docker-compose &> /dev/null; then
+    DOCKER_COMPOSE_CMD="docker-compose"
+    echo "✅ 检测到 Docker Compose V1"
+else
+    echo "❌ Docker Compose未安装，请先安装Docker Compose"
+    exit 1
+fi
+
+echo "✅ Docker环境检查通过"
+echo ""
+
+# 检查jq是否安装
+if ! command -v jq &> /dev/null; then
+    echo "❌ jq未安装，请先安装jq:"
+    echo "   Ubuntu/Debian: sudo apt-get install jq"
+    echo "   CentOS/RHEL: sudo yum install jq"
+    echo "   Alpine: apk add jq"
+    exit 1
+fi
+
+# 检查并生成配置文件
+echo "🔄 检查并生成配置文件..."
+
+if [ -f "config/devices.csv" ]; then
+    echo "📱 从CSV生成ONVIF设备配置..."
+    cd config
+    chmod +x *.sh
+    ./update-configs.sh
+    cd ..
+else
+    echo "⚠️  config/devices.csv 不存在，使用默认JSON配置"
+fi
+
+if [ ! -f "config/onvif-targets.json" ]; then
+    echo "❌ 配置文件 config/onvif-targets.json 不存在"
+    exit 1
+fi
+
+if [ ! -f "config/ping-targets.json" ]; then
+    echo "❌ 配置文件 config/ping-targets.json 不存在"
+    exit 1
+fi
+
+if [ ! -f "prometheus-edge/prometheus.yml" ]; then
+    echo "❌ 配置文件 prometheus-edge/prometheus.yml 不存在"
+    exit 1
+fi
+
+echo "✅ 配置文件检查通过"
+echo ""
+
+# 创建环境变量文件
+if [ ! -f ".env" ]; then
+    if [ -f "env.example" ]; then
+        cp env.example .env
+        echo "📝 已创建 .env 文件，请编辑其中的配置"
+        echo "   特别是 CENTRAL_SERVER_HOST 和 CENTRAL_SERVER_PORT"
+        echo ""
+        read -p "按回车键继续，或 Ctrl+C 取消..."
+    else
+        echo "❌ env.example 文件不存在"
+        exit 1
+    fi
+fi
+
+# 从 .env 生成 prometheus.yml（使 remote_write 指向中央服务器）
+if [ -f ".env" ]; then
+    set -a
+    source .env
+    set +a
+    CENTRAL_SERVER_HOST=${CENTRAL_SERVER_HOST:-192.168.1.10}
+    CENTRAL_SERVER_PORT=${CENTRAL_SERVER_PORT:-8428}
+    if [ -f "prometheus-edge/prometheus.yml.template" ]; then
+        echo "📝 根据 .env 生成 prometheus.yml (中央: ${CENTRAL_SERVER_HOST}:${CENTRAL_SERVER_PORT})..."
+        export CENTRAL_SERVER_HOST CENTRAL_SERVER_PORT
+        envsubst '${CENTRAL_SERVER_HOST} ${CENTRAL_SERVER_PORT}' < prometheus-edge/prometheus.yml.template > prometheus-edge/prometheus.yml
+        echo "✅ prometheus.yml 已生成"
+    fi
+fi
+
+# 创建数据目录
+mkdir -p prometheus-edge/data
+echo "✅ 数据目录创建完成"
+echo ""
+
+# 停止现有服务
+echo "🛑 停止现有服务..."
+$DOCKER_COMPOSE_CMD down 2>/dev/null || true
+
+# 拉取最新镜像
+echo "📥 拉取Docker镜像..."
+if ! $DOCKER_COMPOSE_CMD pull; then
+    echo ""
+    echo "⚠️  镜像拉取失败，尝试继续启动（如果本地已有镜像）..."
+    echo ""
+fi
+
+# 启动服务
+echo "🚀 启动服务..."
+$DOCKER_COMPOSE_CMD up -d
+
+# 等待服务启动
+echo "⏳ 等待服务启动..."
+sleep 10
+
+# 检查服务状态
+echo ""
+echo "📊 服务状态检查："
+$DOCKER_COMPOSE_CMD ps
+
+echo ""
+echo "📋 服务日志："
+$DOCKER_COMPOSE_CMD logs --tail=20
+
+echo ""
+echo "✅ 部署完成！"
+echo ""
+echo "🔗 访问地址："
+echo "   - Prometheus UI: http://localhost:9092"
+echo "   - 目标状态: http://localhost:9092/targets"
+echo ""
+echo "📝 管理命令："
+echo "   - 查看日志: $DOCKER_COMPOSE_CMD logs -f"
+echo "   - 重启服务: $DOCKER_COMPOSE_CMD restart"
+echo "   - 停止服务: $DOCKER_COMPOSE_CMD down"
+echo ""
+echo "🔄 配置更新："
+echo "   - 编辑CSV: nano config/devices.csv"
+echo "   - 生成JSON: cd config && ./update-configs.sh"
+echo "   - 热重载: 等待5分钟自动重载，或重启prometheus-edge"
+echo ""
+echo "⚠️  请确保："
+echo "   1. 已正确配置 .env 文件中的服务器地址"
+echo "   2. 已更新 config/devices.csv 中的设备信息"
+echo "   3. 网络连接正常，可以访问ONVIF设备"

edge-agent/docker-compose.yml

@@ -0,0 +1,70 @@
+services:
+  # ========== 边缘必选 ==========
+  # 1. 边缘 Prometheus：抓取 + remote_write 推到中央 VictoriaMetrics
+  prometheus-edge:
+    image: prom/prometheus:latest
+    container_name: prometheus-edge
+    restart: unless-stopped
+    environment:
+      - CENTRAL_SERVER_HOST=${CENTRAL_SERVER_HOST:-192.168.1.10}
+      - CENTRAL_SERVER_PORT=${CENTRAL_SERVER_PORT:-8428}
+    volumes:
+      - prometheus-edge-data:/prometheus
+      - ./prometheus-edge/prometheus.yml:/etc/prometheus/prometheus.yml:ro
+      - ./config/onvif-targets.json:/etc/prometheus/onvif-targets.json:ro # 挂载静态设备列表
+      - ./config/ping-targets.json:/etc/prometheus/ping-targets.json:ro # 挂载Ping目标列表
+    mem_limit: "256m"
+    cpus: "2.0"
+    ports:
+      - "9092:9090" # 改为9092避免与中央服务器冲突
+    # 本机同机部署时，容器内通过 host.docker.internal 访问宿主机中央服务
+    extra_hosts:
+      - "host.docker.internal:host-gateway"
+    command:
+      - '--config.file=/etc/prometheus/prometheus.yml'
+      - '--storage.tsdb.retention.time=1h'
+      - '--web.enable-lifecycle' # 启用配置重载API
+    networks:
+      - monitoring_net
+
+  # ========== 可选容器（按需启用）==========
+  # 2. ONVIF Exporter（可选，使用本项目自建）
+  # 启用：docker compose --profile onvif up -d --build（会构建 edge-agent/onvif-exporter 并启动）
+  # 配置文件：config/onvif-targets.json（与 targets.csv 中 onvif 行一致，由 update-configs.sh 生成）
+  onvif-exporter:
+    profiles:
+      - onvif
+    image: onvif-exporter:local
+    build:
+      context: ./onvif-exporter
+      dockerfile: Dockerfile
+    container_name: onvif-exporter
+    restart: unless-stopped
+    environment:
+      - EXPORTER_PORT=9600
+      - TARGETS_FILE=/config/targets.json
+    volumes:
+      - ./config/onvif-targets.json:/config/targets.json:ro
+    mem_limit: "128m"
+    cpus: "1.5"
+    networks:
+      - monitoring_net
+
+  # 3. Blackbox Exporter（必选：网络 Ping 探测）
+  blackbox-exporter:
+    image: prom/blackbox-exporter:latest
+    container_name: blackbox-exporter
+    restart: unless-stopped
+    volumes:
+      - ./blackbox/config.yml:/etc/blackbox_exporter/config.yml:ro
+    mem_limit: "64m"
+    cpus: "0.5"
+    networks:
+      - monitoring_net
+
+networks:
+  monitoring_net:
+    driver: bridge
+
+volumes:
+  prometheus-edge-data:

edge-agent/env.example

@@ -0,0 +1,27 @@
+# 中央Prometheus服务器地址 (支持IP或域名)
+# 本机同机部署时使用: host.docker.internal（容器访问宿主机）
+CENTRAL_SERVER_HOST=192.168.1.10
+# 或者使用域名: CENTRAL_SERVER_HOST=prometheus.company.com
+# 本机同机: CENTRAL_SERVER_HOST=host.docker.internal
+
+# 中央服务器端口
+CENTRAL_SERVER_PORT=8428
+# 常用端口说明:
+# 8428 - VictoriaMetrics (推荐)
+# 9090 - Prometheus
+# 8080 - 自定义端口
+
+# 边缘节点标识
+EDGE_NODE_ID=workernode_1
+
+# 远程写入配置
+# 边缘节点会将数据推送到中央服务器的VictoriaMetrics
+# 格式: http://域名或IP:端口/api/v1/write
+# 默认端口: 8428 (VictoriaMetrics)
+
+# 注意：ONVIF设备密码现在在 config/devices.csv 中为每个设备单独配置
+
+# ONVIF Exporter 镜像（仅在使用 --profile onvif 时需要）
+# 公共 registry 无现成镜像，需自建或使用第三方镜像，参见 doc/ONVIF_ALTERNATIVES.md
+# ONVIF_EXPORTER_IMAGE=localhost/onvif-exporter:local
+

edge-agent/onvif-exporter/Dockerfile

@@ -0,0 +1,12 @@
+# 多阶段构建：在镜像内编译，无需本机安装 Go
+FROM golang:1.21-alpine AS builder
+WORKDIR /app
+COPY go.mod ./
+COPY main.go ./
+RUN go mod tidy && CGO_ENABLED=0 GOOS=linux go build -o /onvif-exporter .
+
+FROM alpine:3.19
+RUN apk add --no-cache ca-certificates
+COPY --from=builder /onvif-exporter /onvif-exporter
+EXPOSE 9600
+ENTRYPOINT ["/onvif-exporter"]

edge-agent/onvif-exporter/README.md

@@ -0,0 +1,34 @@
+# ONVIF Exporter（自建）
+
+本目录为自建的 ONVIF 探测容器，供边缘节点可选使用。通过 ONVIF `GetDeviceInformation` 探测设备是否在线，并暴露 Prometheus 指标。
+
+## 指标
+
+- `onvif_device_up`：1=可达，0=不可达（标签：instance, location, model, device_type）
+- `onvif_probe_duration_seconds`：探测耗时（秒）
+
+## 配置
+
+- 从 **config/onvif-targets.json** 读取设备列表（与 `targets.csv` 中 onvif 行一致，由 `config/update-configs.sh` 生成）。
+- 环境变量：`TARGETS_FILE`（默认 `/config/targets.json`）、`EXPORTER_PORT`（默认 9600）。
+
+## 构建与运行
+
+在边缘节点目录下启用 ONVIF 并构建、启动：
+
+```bash
+cd edge-agent
+docker compose --profile onvif up -d --build
+```
+
+或仅构建镜像：
+
+```bash
+docker build -t onvif-exporter:local ./onvif-exporter
+```
+
+## 依赖
+
+- Go 1.21（仅构建时需要，Dockerfile 内已包含）
+- [github.com/use-go/onvif](https://github.com/use-go/onvif)（ONVIF 协议）
+- [prometheus/client_golang](https://github.com/prometheus/client_golang)（指标暴露）

edge-agent/onvif-exporter/go.mod

@@ -0,0 +1,24 @@
+module github.com/distributed-prometheus/onvif-exporter
+
+go 1.21
+
+require (
+	github.com/prometheus/client_golang v1.19.0
+	github.com/use-go/onvif v0.0.9
+)
+
+require (
+	github.com/beorn7/perks v1.0.1 // indirect
+	github.com/beevik/etree v1.1.0 // indirect
+	github.com/cespare/xxhash/v2 v2.2.0 // indirect
+	github.com/elgs/gostrgen v0.0.0-20161222160715-9d61ae07eeae // indirect
+	github.com/gofrs/uuid v3.2.0+incompatible // indirect
+	github.com/juju/errors v0.0.0-20220331221717-b38fca44723b // indirect
+	github.com/prometheus/client_model v0.6.0 // indirect
+	github.com/prometheus/common v0.50.0 // indirect
+	github.com/prometheus/procfs v0.13.0 // indirect
+	github.com/rs/zerolog v1.26.1 // indirect
+	golang.org/x/net v0.21.0 // indirect
+	golang.org/x/sys v0.17.0 // indirect
+	google.golang.org/protobuf v1.32.0 // indirect
+)

edge-agent/onvif-exporter/main.go

@@ -0,0 +1,122 @@
+package main
+
+import (
+	"encoding/json"
+	"log"
+	"net/http"
+	"os"
+	"strings"
+	"time"
+
+	"github.com/prometheus/client_golang/prometheus"
+	"github.com/prometheus/client_golang/prometheus/promhttp"
+	"github.com/use-go/onvif"
+	"github.com/use-go/onvif/device"
+)
+
+const (
+	defaultTargetsFile = "/config/targets.json"
+	defaultPort        = "9600"
+)
+
+type targetGroup struct {
+	Targets []string          `json:"targets"`
+	Labels  map[string]string `json:"labels"`
+}
+
+func main() {
+	targetsFile := os.Getenv("TARGETS_FILE")
+	if targetsFile == "" {
+		targetsFile = defaultTargetsFile
+	}
+	port := os.Getenv("EXPORTER_PORT")
+	if port == "" {
+		port = defaultPort
+	}
+
+	up := prometheus.NewGaugeVec(prometheus.GaugeOpts{
+		Name: "onvif_device_up",
+		Help: "1 if ONVIF device is reachable, 0 otherwise",
+	}, []string{"instance", "location", "model", "device_type"})
+	duration := prometheus.NewGaugeVec(prometheus.GaugeOpts{
+		Name: "onvif_probe_duration_seconds",
+		Help: "ONVIF probe duration in seconds",
+	}, []string{"instance", "location", "model", "device_type"})
+
+	reg := prometheus.NewRegistry()
+	reg.MustRegister(up, duration)
+
+	go func() {
+		ticker := time.NewTicker(60 * time.Second)
+		defer ticker.Stop()
+		for ; true; <-ticker.C {
+			probe(targetsFile, up, duration)
+		}
+	}()
+	// 启动时立即探测一次
+	probe(targetsFile, up, duration)
+
+	http.Handle("/metrics", promhttp.HandlerFor(reg, promhttp.HandlerOpts{EnableOpenMetrics: true}))
+	http.HandleFunc("/health", func(w http.ResponseWriter, _ *http.Request) { w.WriteHeader(http.StatusOK) })
+	log.Printf("ONVIF exporter listening on :%s", port)
+	if err := http.ListenAndServe(":"+port, nil); err != nil {
+		log.Fatal(err)
+	}
+}
+
+func probe(targetsFile string, up, duration *prometheus.GaugeVec) {
+	data, err := os.ReadFile(targetsFile)
+	if err != nil {
+		log.Printf("read targets file: %v", err)
+		return
+	}
+	var groups []targetGroup
+	if err := json.Unmarshal(data, &groups); err != nil {
+		log.Printf("parse targets: %v", err)
+		return
+	}
+	for _, g := range groups {
+		if len(g.Targets) == 0 {
+			continue
+		}
+		ip := strings.TrimSpace(g.Targets[0])
+		port := "80"
+		if p, ok := g.Labels["onvif_port"]; ok && p != "" {
+			port = strings.TrimSpace(p)
+		}
+		user := strings.TrimSpace(g.Labels["username"])
+		pass := strings.TrimSpace(g.Labels["password"])
+		location := strings.TrimSpace(g.Labels["location"])
+		model := strings.TrimSpace(g.Labels["model"])
+		deviceType := strings.TrimSpace(g.Labels["device_type"])
+		instance := ip + ":" + port
+		labels := prometheus.Labels{
+			"instance":    instance,
+			"location":    location,
+			"model":       model,
+			"device_type": deviceType,
+		}
+
+		start := time.Now()
+		err := probeONVIF(ip, port, user, pass)
+		elapsed := time.Since(start).Seconds()
+		duration.With(labels).Set(elapsed)
+		if err != nil {
+			log.Printf("onvif probe %s: %v", instance, err)
+			up.With(labels).Set(0)
+		} else {
+			up.With(labels).Set(1)
+		}
+	}
+}
+
+func probeONVIF(ip, port, username, password string) error {
+	xaddr := ip + ":" + port
+	params := onvif.DeviceParams{Xaddr: xaddr, Username: username, Password: password}
+	dev, err := onvif.NewDevice(params)
+	if err != nil {
+		return err
+	}
+	_, err = dev.CallMethod(device.GetDeviceInformation{})
+	return err
+}

edge-agent/prometheus-edge/prometheus.yml

@@ -0,0 +1,45 @@
+global:
+  scrape_interval: 120s
+  evaluation_interval: 120s
+  external_labels:
+    region: workernode_1 # 设置一个唯一边缘节点名称
+
+remote_write:
+  - url: http://192.168.2.106:8428/api/v1/write
+    queue_config:
+      max_samples_per_send: 5000
+      capacity: 5000
+      max_shards: 5
+
+scrape_configs:
+  - job_name: 'onvif-devices'
+    scrape_interval: 120s
+    file_sd_configs:
+      - files: ['/etc/prometheus/onvif-targets.json']
+        refresh_interval: 5m
+    metrics_path: /metrics
+    static_configs:
+      - targets: ['onvif-exporter:9600']
+
+  - job_name: 'network-ping'
+    scrape_interval: 300s
+    file_sd_configs:
+      - files: ['/etc/prometheus/ping-targets.json']
+        refresh_interval: 5m
+    metrics_path: /probe
+    params:
+      module: [icmp]
+    static_configs:
+      - targets: ['blackbox-exporter:9115']
+    relabel_configs:
+      - source_labels: [__address__]
+        target_label: __param_target
+      - source_labels: [__param_target]
+        target_label: instance
+      - target_label: __address__
+        replacement: blackbox-exporter:9115
+
+  - job_name: 'prometheus-edge'
+    scrape_interval: 60s
+    static_configs:
+      - targets: ['localhost:9090']

edge-agent/prometheus-edge/prometheus.yml.template

@@ -0,0 +1,45 @@
+global:
+  scrape_interval: 120s
+  evaluation_interval: 120s
+  external_labels:
+    region: workernode_1 # 设置一个唯一边缘节点名称
+
+remote_write:
+  - url: http://${CENTRAL_SERVER_HOST}:${CENTRAL_SERVER_PORT}/api/v1/write
+    queue_config:
+      max_samples_per_send: 5000
+      capacity: 5000
+      max_shards: 5
+
+scrape_configs:
+  - job_name: 'onvif-devices'
+    scrape_interval: 120s
+    file_sd_configs:
+      - files: ['/etc/prometheus/onvif-targets.json']
+        refresh_interval: 5m
+    metrics_path: /metrics
+    static_configs:
+      - targets: ['onvif-exporter:9600']
+
+  - job_name: 'network-ping'
+    scrape_interval: 300s
+    file_sd_configs:
+      - files: ['/etc/prometheus/ping-targets.json']
+        refresh_interval: 5m
+    metrics_path: /probe
+    params:
+      module: [icmp]
+    static_configs:
+      - targets: ['blackbox-exporter:9115']
+    relabel_configs:
+      - source_labels: [__address__]
+        target_label: __param_target
+      - source_labels: [__param_target]
+        target_label: instance
+      - target_label: __address__
+        replacement: blackbox-exporter:9115
+
+  - job_name: 'prometheus-edge'
+    scrape_interval: 60s
+    static_configs:
+      - targets: ['localhost:9090']

edge-agent/quick-setup.sh

@@ -0,0 +1,104 @@
+#!/bin/bash
+
+# 边缘节点快速配置脚本
+# 用于在本机快速设置边缘节点
+
+set -e
+
+echo "=== 边缘节点快速配置脚本 ==="
+echo ""
+
+# 获取本机IP
+LOCAL_IP=$(hostname -I | awk '{print $1}')
+
+echo "📋 检测到本机IP: $LOCAL_IP"
+echo ""
+
+# 1. 创建 .env 文件
+if [ ! -f ".env" ]; then
+    echo "📝 创建 .env 配置文件..."
+    cat > .env << EOF
+# 中央服务器地址（本机）
+CENTRAL_SERVER_HOST=${LOCAL_IP}
+CENTRAL_SERVER_PORT=8428
+
+# 边缘节点标识
+EDGE_NODE_ID=workernode_1
+EOF
+    echo "✅ .env 文件已创建"
+else
+    echo "⚠️  .env 文件已存在，跳过创建"
+    echo "   如需修改，请编辑 .env 文件"
+fi
+echo ""
+
+# 2. 配置统一监控目标（最小化测试配置）
+echo "📝 配置统一监控目标 targets.csv..."
+cat > config/targets.csv << 'EOF'
+# 统一监控目标配置文件
+# 格式: type,ip,device,group,network,device_type,model,location,username,password,onvif_port
+# type: onvif 或 ping
+type,ip,device,group,network,device_type,model,location,username,password,onvif_port
+ping,8.8.8.8,google_dns,external,external,,,,,,
+ping,1.1.1.1,cloudflare_dns,external,external,,,,,,
+# ONVIF 设备示例（取消注释并填写实际信息）
+# onvif,192.168.1.100,,,front_door,camera,HIKVISION_DS-2CD2342WD-I,front_door,admin,password1,80
+EOF
+echo "✅ 统一监控目标已配置（使用公共DNS进行测试）"
+echo ""
+
+# 3. 生成配置文件
+echo "🔄 生成配置文件..."
+cd config
+chmod +x *.sh 2>/dev/null || true
+if [ -f "update-configs.sh" ]; then
+    ./update-configs.sh
+    echo "✅ 配置文件已生成"
+else
+    echo "⚠️  update-configs.sh 不存在，跳过"
+fi
+cd ..
+echo ""
+
+# 4. 检查配置文件
+echo "🔍 检查配置文件..."
+if [ ! -f "config/onvif-targets.json" ]; then
+    echo "📝 创建空的 ONVIF 配置文件..."
+    echo "[]" > config/onvif-targets.json
+fi
+
+if [ ! -f "config/ping-targets.json" ]; then
+    echo "📝 创建空的 Ping 配置文件..."
+    echo "[]" > config/ping-targets.json
+fi
+
+echo "✅ 配置文件检查通过"
+echo ""
+
+# 6. 显示配置摘要
+echo "📊 配置摘要："
+echo "   - 中央服务器: ${LOCAL_IP}:8428"
+echo "   - 边缘节点ID: workernode_1"
+echo "   - 监控目标: 已配置（统一 targets.csv）"
+echo "     * Ping 目标: Google DNS, Cloudflare DNS"
+echo "     * ONVIF 设备: 无（用于测试，可在 targets.csv 中添加）"
+echo ""
+
+# 7. 询问是否立即部署
+read -p "是否立即部署边缘节点？(y/N): " -n 1 -r
+echo
+if [[ $REPLY =~ ^[Yy]$ ]]; then
+    echo ""
+    echo "🚀 开始部署..."
+    bash deploy.sh
+else
+    echo ""
+    echo "✅ 配置完成！"
+    echo ""
+    echo "📝 下一步："
+    echo "   1. 检查 .env 文件配置"
+    echo "   2. 编辑 config/targets.csv 添加监控目标（ping 或 onvif）"
+    echo "   3. 运行: cd config && ./update-configs.sh 生成JSON配置"
+    echo "   4. 运行: bash deploy.sh"
+    echo ""
+fi

edge-agent/run-edge-local.sh

@@ -0,0 +1,38 @@
+#!/bin/bash
+# 本机同机部署边缘：中央与边缘在同一台机器时，一键配置并启动边缘
+# 用法: ./run-edge-local.sh
+
+set -e
+
+SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
+cd "$SCRIPT_DIR"
+
+echo "=== 本机同机部署边缘（中央与边缘在同一台机器）==="
+echo ""
+
+# 确保有 .env
+if [ ! -f ".env" ]; then
+    [ -f "env.example" ] && cp env.example .env || { echo "❌ 缺少 env.example"; exit 1; }
+fi
+
+# 指向本机中央 VictoriaMetrics（容器内用 host.docker.internal 访问宿主机）
+if ! grep -q 'CENTRAL_SERVER_HOST=host.docker.internal' .env 2>/dev/null; then
+    echo "📝 设置中央服务器为本机 (host.docker.internal:8428)..."
+    sed -i 's/^CENTRAL_SERVER_HOST=.*/CENTRAL_SERVER_HOST=host.docker.internal/' .env 2>/dev/null || \
+        echo 'CENTRAL_SERVER_HOST=host.docker.internal' >> .env
+    grep -q '^CENTRAL_SERVER_PORT=' .env || echo 'CENTRAL_SERVER_PORT=8428' >> .env
+    sed -i 's/^CENTRAL_SERVER_PORT=.*/CENTRAL_SERVER_PORT=8428/' .env 2>/dev/null || true
+    echo "✅ 已写入 CENTRAL_SERVER_HOST=host.docker.internal, CENTRAL_SERVER_PORT=8428"
+fi
+
+echo ""
+echo "请确保中央服务器已在本机运行（central-server 已 deploy），且 VictoriaMetrics 监听 8428。"
+echo "按回车继续启动边缘，或 Ctrl+C 取消..."
+read -r
+
+bash deploy.sh
+
+echo ""
+echo "💡 本机部署完成后："
+echo "   - 边缘 Prometheus UI: http://localhost:9092"
+echo "   - 在 Grafana 中选择数据源「VictoriaMetrics」可查看边缘上报的数据"