P1-2 nginx 限流 + 扫描器拦截 (20260810-153624)

三档限流 (api 30r/s, admin 5r/s, login 1r/s) + UA / path 黑名单 (return 444) + 自动黑名单脚本 (5min cron, 24h TTL). 实测首跑封禁 222.128.55.233 (4xx=31) 和 120.55.61.72 (UA=zgrab).

nginx/aiflow.conf

@@ -0,0 +1,129 @@
+# 七虎 AI PaaS nginx site — enhanced with rate limiting + scanner defense
+# Owner: 七虎平台运维 (Senior Developer)
+# 维护: edit on host /app/backend/nginx/aiflow.conf, push to container /etc/nginx/conf.d/aiflow.conf
+# 重载: docker exec qihu-aiflow sh -c "nginx -t && nginx -s stop && nginx"
+
+# 三档速率限制
+limit_req_zone $binary_remote_addr zone=api_limit:10m rate=30r/s;
+limit_req_zone $binary_remote_addr zone=admin_limit:10m rate=5r/s;
+limit_req_zone $binary_remote_addr zone=login_limit:1m rate=1r/s;
+
+# 动态黑名单(由 /app/backend/nginx/block_ip.py 写入)
+include /etc/nginx/blocked_ips.conf;
+
+server {
+    listen 8080 default_server;
+    server_name _;
+
+    client_max_body_size 100m;
+    client_body_timeout 30s;
+    keepalive_timeout 75s;
+
+    gzip off;
+
+    # ===== 反扫描器: bad URI 直接 444 =====
+    if ($request_uri ~* "^/(wp-admin|wp-login|xmlrpc\.php|administrator/|sites/default|manager/html|server-status|server-info|elmah\.axd|web\.config|readme|license\.txt|install\.php|shell\.php)") {
+        return 444;
+    }
+    if ($request_uri ~* "^/(phpmyadmin|\.env|\.git/|actuator|actuator/|\.well-known/security)") {
+        return 444;
+    }
+    if ($http_user_agent ~* "(zgrab|nmap|masscan|nikto|sqlmap|acunetix|nessus|wpscan|dirbuster|gobuster|ffuf|hydra|metasploit|MJ12bot|AhrefsBot|SemrushBot|DotBot|PetalBot|Aspera|HTTrack|Scrapy|libwww|Havij|TSWeb)") {
+        return 444;
+    }
+    if ($http_user_agent ~* "(^curl/|^wget/|^python-requests)") {
+        return 444;
+    }
+
+    access_log /var/log/nginx/access.log;
+    error_log /var/log/nginx/error.log warn;
+
+    # 索引 HTML 不缓存
+    location = /index.html {
+        add_header Cache-Control "no-cache";
+        try_files /index.html =404;
+    }
+
+    # 静态资源永久缓存
+    location ~* \.(svg|png|jpg|jpeg|gif|ico|webp|woff2?|ttf|eot|css|js|map)$ {
+        root /app/frontend/dist;
+        try_files $uri =404;
+        access_log off;
+        expires 30d;
+        add_header Cache-Control "public, max-age=2592000";
+    }
+
+    # SPA fallback
+    location / {
+        try_files $uri $uri/ /index.html;
+    }
+
+    # ===== 普通 API: 30r/s =====
+    location /app/ {
+        limit_req zone=api_limit burst=60 nodelay;
+        proxy_pass http://localhost:18080/app/;
+        proxy_set_header Host $host;
+        proxy_set_header X-Real-IP $remote_addr;
+        proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
+        proxy_set_header X-Forwarded-Proto $scheme;
+        proxy_read_timeout 600s;
+        proxy_connect_timeout 10s;
+    }
+
+    # ===== ADMIN API: 5r/s 严格 =====
+    location ~ ^/app/api/(admin/|market/admin/|internal/) {
+        limit_req zone=admin_limit burst=10 nodelay;
+        proxy_pass http://localhost:18080;
+        proxy_set_header Host $host;
+        proxy_set_header X-Real-IP $remote_addr;
+        proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
+        proxy_set_header X-Forwarded-Proto $scheme;
+        proxy_read_timeout 600s;
+        proxy_send_timeout 600s;
+    }
+
+    # ===== AUTH: 1r/s 防爆破 =====
+    location ~ ^/app/api/(auth/login|auth/register|casdoor/) {
+        limit_req zone=login_limit burst=5 nodelay;
+        proxy_pass http://localhost:18080;
+        proxy_set_header Host $host;
+        proxy_set_header X-Real-IP $remote_addr;
+        proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
+        proxy_set_header X-Forwarded-Proto $scheme;
+        proxy_read_timeout 60s;
+        proxy_send_timeout 60s;
+    }
+
+    # ===== SSE STREAMING: 禁缓冲 长连接 =====
+    location ~ ^/app/(api/chat/events|api/pet/) {
+        proxy_pass http://localhost:18080;
+        proxy_set_header Host $host;
+        proxy_set_header X-Real-IP $remote_addr;
+        proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
+        proxy_set_header X-Forwarded-Proto $scheme;
+        proxy_set_header Connection "";
+        proxy_buffering off;
+        proxy_cache off;
+        proxy_request_buffering off;
+        proxy_http_version 1.1;
+        proxy_read_timeout 86400s;
+        proxy_send_timeout 86400s;
+        chunked_transfer_encoding on;
+    }
+
+    # 健康检查
+    location = /health {
+        access_log off;
+        return 200 "ok\n";
+        add_header Content-Type text/plain;
+    }
+
+    # 405/413 nice error
+    error_page 413 /custom_413;
+    location = /custom_413 {
+        internal;
+        return 413 "Payload Too Large\n";
+    }
+}
+
+

nginx/block_ip.py

@@ -0,0 +1,197 @@
+#!/usr/bin/env python3
+"""qihu-platform auto-block — read nginx access.log, push scanner IPs to blocked_ips.conf.
+
+Usage:  python3 /app/backend/nginx/block_ip.py
+Cron:   */5 * * * *  python3 /app/backend/nginx/block_ip.py >> /var/log/qihu-block.log 2>&1
+"""
+from __future__ import annotations
+import ipaddress
+import os
+import re
+import subprocess
+import sys
+import time
+from collections import defaultdict
+from pathlib import Path
+
+ACCESS_LOG = "/var/log/nginx/access.log"
+ACCESS_LOG_1 = "/var/log/nginx/access.log.1"
+BLOCK_CONF = "/etc/nginx/blocked_ips.conf"
+STATE_FILE = "/var/lib/qihu/block_state.json"
+WINDOW_SEC = 300  # only consider last 5 min
+THRESHOLD_444 = 6
+THRESHOLD_4XX = 20
+BLOCK_TTL_HOURS = 24
+
+LOG_RE = re.compile(
+    r"^(\S+)\s+-\s+\S+\s+\[([^\]]+)\]\s+\"(\S+)\s+(\S+)[^\"]*\"\s+(\d+)\s+\S+\s+\"[^\"]*\"\s+\"([^\"]*)\""
+)
+SCANNER_UA = re.compile(
+    r"(zgrab|nmap|masscan|nikto|sqlmap|acunetix|nessus|wpscan|dirbuster|gobuster|ffuf|hydra|metasploit|Havij|TSWeb)",
+    re.IGNORECASE,
+)
+SCAN_PATH = re.compile(
+    r"^/(wp-admin|wp-login|xmlrpc\.php|administrator|phpmyadmin|\.env|\.git/|actuator|server-status|server-info|elmah)",
+    re.IGNORECASE,
+)
+
+
+def now_ts() -> int:
+    return int(time.time())
+
+
+def parse_log_line(line: str):
+    m = LOG_RE.match(line)
+    if not m:
+        return None
+    ip, ts_str, method, path, status, ua = m.groups()
+    return ip, ts_str, method, path, int(status), ua
+
+
+def nginx_time_to_epoch(s: str) -> int:
+    """10/Aug/2026:14:48:20 +0800 -> epoch"""
+    try:
+        from datetime import datetime
+        return int(datetime.strptime(s.split()[0], "%d/%b/%Y:%H:%M:%S").timestamp())
+    except Exception:
+        return 0
+
+
+def load_state() -> dict:
+    p = Path(STATE_FILE)
+    if p.exists():
+        try:
+            import json
+            return json.loads(p.read_text())
+        except Exception:
+            pass
+    return {"blocks": {}, "last_run": 0}
+
+
+def save_state(state: dict) -> None:
+    Path(STATE_FILE).parent.mkdir(parents=True, exist_ok=True)
+    import json
+    Path(STATE_FILE).write_text(json.dumps(state, indent=2, sort_keys=True))
+
+
+def is_public_ip(ip: str) -> bool:
+    """Skip private / loopback / container IPs."""
+    try:
+        a = ipaddress.ip_address(ip)
+        return not (a.is_private or a.is_loopback or a.is_link_local or a.is_reserved or a.is_multicast)
+    except ValueError:
+        return False
+
+
+def scan() -> dict:
+    """Return {ip: {444: n, 4xx: n, ua_hit: bool, path_hit: bool}}."""
+    stats: dict[str, dict] = defaultdict(lambda: {"444": 0, "4xx": 0, "ua_hit": False, "path_hit": False, "last": 0})
+    cutoff = now_ts() - WINDOW_SEC
+    paths = [ACCESS_LOG]
+    # also include rotated log if recent
+    if os.path.exists(ACCESS_LOG_1):
+        paths.append(ACCESS_LOG_1)
+    for p in paths:
+        if not os.path.exists(p):
+            continue
+        with open(p, "r", errors="replace") as f:
+            for line in f:
+                parsed = parse_log_line(line)
+                if not parsed:
+                    continue
+                ip, ts_str, method, path, status, ua = parsed
+                ts = nginx_time_to_epoch(ts_str)
+                if ts and ts < cutoff and p == ACCESS_LOG:
+                    continue  # only skip old in current log
+                if not is_public_ip(ip):
+                    continue
+                s = stats[ip]
+                if status == 444:
+                    s["444"] += 1
+                if 400 <= status < 500:
+                    s["4xx"] += 1
+                if SCANNER_UA.search(ua):
+                    s["ua_hit"] = True
+                if SCAN_PATH.match(path):
+                    s["path_hit"] = True
+                s["last"] = max(s["last"], ts)
+    return stats
+
+
+def add_to_blocked(ip: str) -> None:
+    """Append `deny <ip>;` to blocked_ips.conf (dedup)."""
+    with open(BLOCK_CONF, "r") as f:
+        existing = f.read()
+    rule = f"deny {ip};"
+    if rule in existing:
+        return
+    with open(BLOCK_CONF, "a") as f:
+        f.write(f"\n# auto {time.strftime('%Y-%m-%dT%H:%M:%S')}\n{rule}\n")
+
+
+def reload_nginx() -> bool:
+    try:
+        subprocess.run(["nginx", "-t"], check=True, capture_output=True, timeout=10)
+        subprocess.run(["nginx", "-s", "reload"], check=True, capture_output=True, timeout=10)
+        return True
+    except Exception as e:
+        print(f"reload fail: {e}", file=sys.stderr)
+        return False
+
+
+def main() -> int:
+    state = load_state()
+    blocks = state.get("blocks", {})
+    stats = scan()
+    new_blocks = 0
+    for ip, s in stats.items():
+        should_block = (
+            s["ua_hit"]
+            or s["444"] >= THRESHOLD_444
+            or s["4xx"] >= THRESHOLD_4XX
+        )
+        if not should_block:
+            continue
+        if ip in blocks and (now_ts() - blocks[ip]) < BLOCK_TTL_HOURS * 3600:
+            continue  # already in block list, still in TTL
+        add_to_blocked(ip)
+        blocks[ip] = now_ts()
+        new_blocks += 1
+        print(f"BLOCKED {ip} (444={s['444']} 4xx={s['4xx']} ua={s['ua_hit']} path={s['path_hit']})")
+    # expire old blocks
+    expired = [ip for ip, t in blocks.items() if now_ts() - t > BLOCK_TTL_HOURS * 3600]
+    if expired:
+        print(f"EXPIRED {len(expired)} blocks: {expired[:5]}...")
+        for ip in expired:
+            del blocks[ip]
+        # rewrite conf without expired
+        rewrite_blocked_conf(blocks)
+    if new_blocks:
+        if not reload_nginx():
+            print("RELOAD FAILED, blocks queued", file=sys.stderr)
+            return 1
+    state["blocks"] = blocks
+    state["last_run"] = now_ts()
+    save_state(state)
+    print(f"DONE scanned={len(stats)} new_blocks={new_blocks} expired={len(expired)}")
+    return 0
+
+
+def rewrite_blocked_conf(blocks: dict) -> None:
+    """Rewrite blocked_ips.conf with current active blocks + deny directives."""
+    lines = [
+        "# Auto-managed deny list (see /app/backend/nginx/block_ip.py)",
+        "# Format: deny 1.2.3.4; (one per line)",
+        "# Cleanup: comment out when removing",
+        "",
+    ]
+    for ip, t in sorted(blocks.items()):
+        lines.append(f"# auto {time.strftime('%Y-%m-%dT%H:%M:%S', time.localtime(t))}")
+        lines.append(f"deny {ip};")
+    Path(BLOCK_CONF).write_text("\n".join(lines) + "\n")
+
+
+if __name__ == "__main__":
+    sys.exit(main())
+
+

nginx/blocked_ips.conf

@@ -0,0 +1,11 @@
+# Auto-managed deny list (see /app/backend/nginx/block_ip.py)
+# Format: deny 1.2.3.4; (one per line)
+# Cleanup: comment out when removing
+# cleaned
+
+# auto 2026-08-10T07:31:45
+deny 222.128.55.233;
+
+# auto 2026-08-10T07:31:45
+deny 120.55.61.72;
+