implementing-ddos-mitigation-with-cloudflare

使用 Cloudflare 实施 DDoS 缓解

概述

Cloudflare 通过其由 300 多个数据中心组成的全球网络提供多层 DDoS 防护，容量达 477+ Tbps。该平台可防御 L3/4 容量型攻击（SYN 洪水、UDP 放大、DNS 反射）、协议攻击（Ping of Death、Smurf）以及 L7 应用层攻击（HTTP 洪水、Slowloris、缓存绕过）。Cloudflare 的自主检测系统利用流量画像、机器学习和自适应规则集，可在约 3 秒内识别并缓解攻击。本技能涵盖配置 Cloudflare 的 DDoS 防护栈，包括托管规则集、WAF 规则、速率限制、Bot Management 以及源站服务器加固。

何时使用

• 在您的环境中部署或配置使用 Cloudflare 实施 DDoS 缓解能力时
• 建立符合合规要求的安全控制时
• 为此领域构建或改进安全架构时
• 执行需要此实施的安全评估时

前提条件

• Cloudflare 账户（WAF 最低需 Pro 套餐，高级 DDoS 需 Enterprise 套餐）
• 已将 DNS 委派给 Cloudflare 名称服务器的域名
• 源站服务器 IP 地址
• 了解正常流量模式和峰值流量
• 用于自动化的 Cloudflare API token

核心概念

DDoS 攻击类别

层	攻击类型	示例	Cloudflare 防护
L3/4	容量型	SYN 洪水、UDP 洪水、DNS 放大	网络层 DDoS 托管规则
L3/4	协议	Ping of Death、Smurf、IP 分片	Advanced TCP Protection
L7	应用	HTTP 洪水、Slowloris、缓存绕过	HTTP DDoS 托管规则、WAF、速率限制
DNS	DNS 专用	DNS 查询洪水、NXDOMAIN 攻击	Advanced DNS Protection

Cloudflare 防护栈

Internet Traffic
     │
     ▼
┌─────────────────────────┐
│  Cloudflare Edge (PoP)  │
│  ┌───────────────────┐  │
│  │ L3/4 DDoS Mgd Rules│  │  ← Volumetric/Protocol mitigation
│  └───────────────────┘  │
│  ┌───────────────────┐  │
│  │ IP Access Rules    │  │  ← Country/ASN/IP blocks
│  └───────────────────┘  │
│  ┌───────────────────┐  │
│  │ Bot Management     │  │  ← Bot score, JS challenge
│  └───────────────────┘  │
│  ┌───────────────────┐  │
│  │ WAF Managed Rules  │  │  ← OWASP, Cloudflare, Custom
│  └───────────────────┘  │
│  ┌───────────────────┐  │
│  │ Rate Limiting      │  │  ← Request rate enforcement
│  └───────────────────┘  │
│  ┌───────────────────┐  │
│  │ HTTP DDoS Mgd Rules│  │  ← L7 flood detection
│  └───────────────────┘  │
└─────────────────────────┘
     │
     ▼
  Origin Server

工作流

步骤 1：将域名接入 Cloudflare

# Add domain via API
curl -X POST "https://api.cloudflare.com/client/v4/zones" \
  -H "Authorization: Bearer $CF_API_TOKEN" \
  -H "Content-Type: application/json" \
  --data '{
    "name": "example.com",
    "type": "full",
    "plan": {"id": "enterprise"}
  }'

# Update DNS records (proxy enabled for DDoS protection)
curl -X POST "https://api.cloudflare.com/client/v4/zones/$ZONE_ID/dns_records" \
  -H "Authorization: Bearer $CF_API_TOKEN" \
  -H "Content-Type: application/json" \
  --data '{
    "type": "A",
    "name": "example.com",
    "content": "203.0.113.50",
    "proxied": true,
    "ttl": 1
  }'

步骤 2：配置 DDoS 托管规则集

HTTP DDoS 攻击防护覆盖：

# List HTTP DDoS managed ruleset
curl -X GET "https://api.cloudflare.com/client/v4/zones/$ZONE_ID/rulesets/phases/ddos_l7/entrypoint" \
  -H "Authorization: Bearer $CF_API_TOKEN"

# Override HTTP DDoS sensitivity and action
curl -X PUT "https://api.cloudflare.com/client/v4/zones/$ZONE_ID/rulesets/phases/ddos_l7/entrypoint" \
  -H "Authorization: Bearer $CF_API_TOKEN" \
  -H "Content-Type: application/json" \
  --data '{
    "rules": [{
      "action": "execute",
      "action_parameters": {
        "id": "4d21379b4f9f4bb088e0729962c8b3cf",
        "overrides": {
          "rules": [{
            "id": "fdfdac75430c4c47a422bdc024aab531",
            "sensitivity_level": "medium",
            "action": "block"
          }],
          "sensitivity_level": "high"
        }
      },
      "expression": "true"
    }]
  }'

网络层 DDoS 防护覆盖：

curl -X PUT "https://api.cloudflare.com/client/v4/accounts/$ACCOUNT_ID/rulesets/phases/ddos_l4/entrypoint" \
  -H "Authorization: Bearer $CF_API_TOKEN" \
  -H "Content-Type: application/json" \
  --data '{
    "rules": [{
      "action": "execute",
      "action_parameters": {
        "id": "3b64149bfa6e4220bbbc2bd6db7c867e",
        "overrides": {
          "sensitivity_level": "high"
        }
      },
      "expression": "true"
    }]
  }'

步骤 3：配置速率限制规则

# Create rate limiting rule for login endpoint
curl -X POST "https://api.cloudflare.com/client/v4/zones/$ZONE_ID/rulesets/phases/http_ratelimit/entrypoint" \
  -H "Authorization: Bearer $CF_API_TOKEN" \
  -H "Content-Type: application/json" \
  --data '{
    "rules": [
      {
        "description": "Rate limit login attempts",
        "expression": "(http.request.uri.path eq \"/api/login\")",
        "action": "block",
        "ratelimit": {
          "characteristics": ["cf.colo.id", "ip.src"],
          "period": 60,
          "requests_per_period": 10,
          "mitigation_timeout": 600
        }
      },
      {
        "description": "Rate limit API endpoints",
        "expression": "(http.request.uri.path matches \"^/api/\")",
        "action": "managed_challenge",
        "ratelimit": {
          "characteristics": ["cf.colo.id", "ip.src"],
          "period": 60,
          "requests_per_period": 100,
          "mitigation_timeout": 300
        }
      },
      {
        "description": "Global rate limit per IP",
        "expression": "true",
        "action": "managed_challenge",
        "ratelimit": {
          "characteristics": ["ip.src"],
          "period": 10,
          "requests_per_period": 50,
          "mitigation_timeout": 60
        }
      }
    ]
  }'

步骤 4：配置 WAF 自定义规则

# Block known attack patterns
curl -X POST "https://api.cloudflare.com/client/v4/zones/$ZONE_ID/rulesets/phases/http_request_firewall_custom/entrypoint" \
  -H "Authorization: Bearer $CF_API_TOKEN" \
  -H "Content-Type: application/json" \
  --data '{
    "rules": [
      {
        "description": "Block requests from known bad ASNs",
        "expression": "(ip.geoip.asnum in {12345 67890})",
        "action": "block"
      },
      {
        "description": "Challenge requests without User-Agent",
        "expression": "(not http.user_agent ne \"\")",
        "action": "managed_challenge"
      },
      {
        "description": "Block high-risk countries for admin paths",
        "expression": "(http.request.uri.path contains \"/admin\" and not ip.geoip.country in {\"US\" \"CA\" \"GB\"})",
        "action": "block"
      },
      {
        "description": "Block oversized request bodies",
        "expression": "(http.request.body.size gt 10000000)",
        "action": "block"
      }
    ]
  }'

步骤 5：配置源站防护

确保源站服务器仅接受来自 Cloudflare 的流量：

# Get Cloudflare IP ranges
curl https://api.cloudflare.com/client/v4/ips

# Configure origin server firewall (iptables)
# Allow only Cloudflare IPs
for ip in $(curl -s https://www.cloudflare.com/ips-v4); do
    iptables -A INPUT -p tcp --dport 443 -s $ip -j ACCEPT
    iptables -A INPUT -p tcp --dport 80 -s $ip -j ACCEPT
done

# Drop all other HTTP/HTTPS traffic
iptables -A INPUT -p tcp --dport 443 -j DROP
iptables -A INPUT -p tcp --dport 80 -j DROP

# Enable Authenticated Origin Pulls (mutual TLS)
# Download Cloudflare origin CA certificate
curl -o /etc/ssl/cloudflare-origin-pull.pem \
  https://developers.cloudflare.com/ssl/static/authenticated_origin_pull_ca.pem

# Nginx configuration for authenticated origin pulls
# ssl_client_certificate /etc/ssl/cloudflare-origin-pull.pem;
# ssl_verify_client on;

步骤 6：启用 Under Attack Mode 自动化

#!/usr/bin/env python3
"""Auto-enable Cloudflare Under Attack Mode based on traffic anomalies."""

import requests
import time
import sys

CF_API_TOKEN = "your-api-token"
ZONE_ID = "your-zone-id"
HEADERS = {
    "Authorization": f"Bearer {CF_API_TOKEN}",
    "Content-Type": "application/json",
}
BASE_URL = f"https://api.cloudflare.com/client/v4/zones/{ZONE_ID}"

NORMAL_RPS_THRESHOLD = 5000  # Requests per second threshold
CHECK_INTERVAL = 30  # Seconds between checks


def get_current_security_level():
    """Get current security level setting."""
    resp = requests.get(
        f"{BASE_URL}/settings/security_level",
        headers=HEADERS
    )
    return resp.json()["result"]["value"]


def set_security_level(level: str):
    """Set security level (off, essentially_off, low, medium, high, under_attack)."""
    resp = requests.patch(
        f"{BASE_URL}/settings/security_level",
        headers=HEADERS,
        json={"value": level}
    )
    result = resp.json()
    if result["success"]:
        print(f"[+] Security level set to: {level}")
    else:
        print(f"[-] Failed to set security level: {result['errors']}")
    return result["success"]


def get_traffic_analytics():
    """Get recent traffic data from Cloudflare analytics."""
    query = """
    query {
      viewer {
        zones(filter: {zoneTag: "%s"}) {
          httpRequests1mGroups(limit: 1, orderBy: [datetime_DESC]) {
            sum {
              requests
              threats
            }
            dimensions {
              datetime
            }
          }
        }
      }
    }
    """ % ZONE_ID

    resp = requests.post(
        "https://api.cloudflare.com/client/v4/graphql",
        headers=HEADERS,
        json={"query": query}
    )
    return resp.json()


def monitor_and_respond():
    """Monitor traffic and auto-enable under attack mode."""
    current_level = get_current_security_level()
    print(f"[*] Current security level: {current_level}")
    print(f"[*] Monitoring traffic (threshold: {NORMAL_RPS_THRESHOLD} RPS)...")

    attack_mode_active = False
    consecutive_normal = 0

    while True:
        try:
            analytics = get_traffic_analytics()
            zones = analytics.get("data", {}).get("viewer", {}).get("zones", [])

            if zones and zones[0].get("httpRequests1mGroups"):
                data = zones[0]["httpRequests1mGroups"][0]["sum"]
                rps = data["requests"] / 60
                threats = data["threats"]

                print(f"[*] Current RPS: {rps:.0f}, Threats: {threats}")

                if rps > NORMAL_RPS_THRESHOLD and not attack_mode_active:
                    print(f"[!] Traffic spike detected: {rps:.0f} RPS")
                    set_security_level("under_attack")
                    attack_mode_active = True
                    consecutive_normal = 0

                elif rps <= NORMAL_RPS_THRESHOLD and attack_mode_active:
                    consecutive_normal += 1
                    if consecutive_normal >= 5:
                        print("[+] Traffic normalized, disabling under attack mode")
                        set_security_level("high")
                        attack_mode_active = False
                        consecutive_normal = 0

        except Exception as e:
            print(f"[-] Error: {e}")

        time.sleep(CHECK_INTERVAL)


if __name__ == "__main__":
    monitor_and_respond()

监控与告警

Cloudflare Dashboard 指标

• 防火墙事件 - 查看已阻止请求、质询请求、速率受限请求
• DDoS 分析 - 攻击规模、持续时间、类型及缓解状态
• 流量分析 - 按时间统计的请求量、带宽、错误率
• Bot 分析 - Bot 分数分布、已验证 Bot 与自动化威胁对比

告警配置

# Create notification policy for DDoS attacks
curl -X POST "https://api.cloudflare.com/client/v4/accounts/$ACCOUNT_ID/alerting/v3/policies" \
  -H "Authorization: Bearer $CF_API_TOKEN" \
  -H "Content-Type: application/json" \
  --data '{
    "name": "DDoS Attack Alert",
    "alert_type": "dos_attack_l7",
    "enabled": true,
    "mechanisms": {
      "email": [{"id": "[email protected]"}],
      "webhooks": [{"id": "webhook-id"}]
    },
    "filters": {
      "zones": ["'$ZONE_ID'"]
    }
  }'

最佳实践

• 代理所有 DNS 记录 - 确保所有指向源站的 A/AAAA/CNAME 记录均已代理（橙色云）
• 隐藏源站 IP - 绝不要暴露源站服务器 IP；使用 Cloudflare Tunnel 或仅限 Cloudflare IP 访问
• 从日志模式开始 - 先用“日志”操作测试 DDoS 规则覆盖，再切换到“阻止”
• 分层防御 - 结合托管规则集、速率限制、WAF 规则和 Bot Management
• 调整敏感度 - 根据流量中的误报率调整 DDoS 规则敏感度
• 缓存策略 - 最大化缓存命中率，以降低攻击期间的源站负载
• 等候室 - 在流量激增期间为关键页面配置 Cloudflare Waiting Room
• 源站认证 - 启用 Authenticated Origin Pulls，防止直接针对源站的攻击

参考

• Cloudflare DDoS 防护文档
• Cloudflare WAF 文档
• Cloudflare 速率限制
• Cloudflare IP 范围