服务器CDN回源502时如何快速切换源站

# 模拟主源站故障（停止Nginx服务）
systemctl stop nginx

# 持续请求CDN地址，观察是否自动切换
curl -I https://cdn.example.com/test.html | grep "X-Served-By"
# 正常应返回备用源站标识，如：X-Served-By: backup-node-123

upstream primary_backend {
    server 192.0.2.1:80 max_fails=3 fail_timeout=30s;
    server 192.0.2.2:80 backup;  # 备用节点（手动切换时启用）
}

upstream backup_backend {
    server 198.51.100.2:80;
}

server {
    location / {
        # 默认使用主源站
        set $backend "primary_backend";

        # 健康检查失败时切换到备用源站
        if ($upstream_addr ~* "192\.0\.2\.1:80.*failed") {
            set $backend "backup_backend";
        }

        proxy_pass http://$backend;
        proxy_set_header Host $host;
    }
}

-- 在Nginx的init_by_lua_block中加载共享字典
init_by_lua_block {
    local health_check = ngx.shared.health_check
    health_check:set("primary_status", "healthy", 10)  -- 10秒过期
}

-- 在content_by_lua_block中动态选择源站
location / {
    access_by_lua_block {
        local health_check = ngx.shared.health_check
        local primary_status = health_check:get("primary_status")

        if primary_status == "unhealthy" then
            ngx.var.backend = "backup_backend"
        else
            ngx.var.backend = "primary_backend"
        end
    }

    proxy_pass http://$backend;
}

-- 通过外部脚本更新健康状态（如每5秒执行一次）
-- curl -X POST "http://nginx-server/health_update?status=unhealthy"

# 1. 检查当前使用的源站
curl -sI https://cdn.example.com/test.html | grep "X-Served-By"

# 2. 模拟备用源站故障（测试切换回主源站）
systemctl stop nginx@backup
curl -s "http://primary-ip/healthz?force_fail=1"  # 触发主源站返回503
sleep 35  # 等待CDN健康检查周期
curl -sI https://cdn.example.com/test.html | grep "X-Served-By"  # 应切换回主源站

# 3. 检查DNS解析记录（DNS切换方案）
dig +short origin.example.com

# 告警规则示例：源站切换频率过高
groups:
- name: cdn-source-switch-alert
  rules:
  - alert: FrequentSourceSwitch
    expr: increase(cdn_source_switch_total[5m]) > 3  # 5分钟内切换超过3次
    for: 1m
    labels:
      severity: critical
    annotations:
      summary: "CDN源站频繁切换"
      description: "域名 {{ $labels.domain }} 在5分钟内切换源站 {{ $value }} 次，可能存在主源站持续故障"

#!/bin/bash
# 分析CDN边缘节点日志，找出502错误与源站切换的关联
LOG_FILE="/var/log/cdn/edge.log"
START_TIME=$(date -d "10 minutes ago" +%s)

awk -v start=$START_TIME '{
    # 解析时间戳（假设日志格式为：2023-10-16T14:30:00Z）
    if ($1 ~ /^[0-9]{4}-[0-9]{2}-[0-9]{2}T/) {
        cmd="date -d \"" $1 "\" +%s 2>/dev/null"
        cmd | getline timestamp
        close(cmd)
        if (timestamp >= start) {
            # 统计502错误与源站切换事件
            if ($9 == "502") {
                print "502 Error:", $0
            } else if ($0 ~ /source_switch/) {
                print "Source Switch:", $0
            }
        }
    }
}' "$LOG_FILE" | sort -k1,1 > /tmp/cdn_analysis.log

# 生成关联报告
echo "=== 502 Error & Source Switch Correlation ==="
grep -A 2 "502 Error" /tmp/cdn_analysis.log | grep -B 2 "Source Switch" | head -20