惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

D
Docker
B
Blog RSS Feed
Microsoft Security Blog
Microsoft Security Blog
Y
Y Combinator Blog
N
Netflix TechBlog - Medium
M
MIT News - Artificial intelligence
奇客Solidot–传递最新科技情报
奇客Solidot–传递最新科技情报
B
Blog
钛媒体:引领未来商业与生活新知
钛媒体:引领未来商业与生活新知
C
Check Point Blog
The GitHub Blog
The GitHub Blog
让小产品的独立变现更简单 - ezindie.com
让小产品的独立变现更简单 - ezindie.com
P
Proofpoint News Feed
Martin Fowler
Martin Fowler
大猫的无限游戏
大猫的无限游戏
GbyAI
GbyAI
博客园_首页
A
About on SuperTechFans
Blog — PlanetScale
Blog — PlanetScale
人人都是产品经理
人人都是产品经理
T
Tailwind CSS Blog
aimingoo的专栏
aimingoo的专栏
T
The Blog of Author Tim Ferriss
The Cloudflare Blog

DEV Community

Authentication Security Deep Dive: From Brute Force to Salted Hashing (With Java Examples) Why AI Systems Don’t Fail — They Drift Spilling beans for how i learn for exam😁"Reinforcement Learning Cheat Sheet" I Replaced Chrome with Safari for AI Browser Automation. Here's What Broke (and What Finally Worked) How Python Borrows Other People's Work The $40 Architecture: Processing 1 Billion API Requests with 99.99% Uptime Vibe Coding: A Workflow Guide (From Zero to SaaS) Most webhook security guides protect the wrong side. The scary part is delivery. Headless CMS for TanStack Start: Build a Blog with Cosmic EU Age Verification App "Hacked in 2 Minutes" — What Actually Happened Comfy Cloud’s delete function does not actually remove files Running AI Models on GPU Cloud Servers: A Beginner Guide Event-driven media intelligence with AWS Step Functions and Bedrock I scored 500 AI prompts across 8 quality dimensions — here's what broke How to Call Google Gemini API from Next.js (Free Tier, No Backend Needed) The Portal Protocol: Reclaiming Human Connection in the Age of AI How to Fix Your Team's Scattered Knowledge Problem With a Self-Hosted Forum Intro to tc Cloud Functors: A Graph-First Mental Model for the Modern Cloud Designing Multi-Tenant Backends With Both Ownership and Team Access I Built a Neumorphic CSS Library with 77+ Components — Here's What I Learned PostgreSQL Performance Optimization: Why Connection Pooling Is Critical at Scale Cómo construí un SaaS multi-rubro para gestionar expensas en Argentina con FastAPI + Vue 3 🚀 I Built an Ethical Hacking Scanner Tool – Open Source Project I Replaced /usage and /context in Claude Code With a Single Statusline A Pythonic Way to Handle Emails (IMAP/SMTP) with Auto-Discovery and AI-Ready Design I Collected 8.9 Million Polymarket Price Points — Here's What I Found About How Markets Really Move EcoTrack AI — Carbon Footprint Tracker & Dashboard Everyone's Using AI. No One Agrees How. 5 self-hosted ebook managers worth trying in 2026 Building Your First AI Agent with LangChain: From Chatbot to Autonomous Assistant
How 77% of a Magento Store's Traffic Turned Out to Be Bot...
Kishan Saval · 2026-05-27 · via DEV Community

A store owner pinged me with a worrying screenshot: Google Analytics showed a sudden spike of "active users," almost all from a single country. Their
first guess was a viral moment. It wasn't. It was a bot flood — and once I pulled the server logs, the numbers were brutal: 77% of all incoming
requests were automated traffic
hammering the site.

Here's exactly how I diagnosed it and shut it down, with the config you can reuse.

## Step 1: Read the actual logs, not the dashboard

Analytics dashboards lie to you about bots, because many bots execute JavaScript and show up as real "users." The truth is in the web server logs. I
pulled the last 15 minutes of nginx access logs and aggregated by client IP and user-agent.

The pattern was instantly obvious:

  • ~77% of requests came from one cloud-hosting IP range (a data-center, not real humans), rotating fake Chrome user-agents — including Chrome version numbers that don't exist yet. That's a dead giveaway.
  • The rest were crawlers stuck in infinite pagination crawl-traps like /blog/tag/x/page/11410 and ?p=6095 — URLs that should never have been generated.

## Step 2: Understand why a firewall won't help (the part people get wrong)

The owner's instinct was "just block the IP with a firewall." But the site sits behind Cloudflare, and that changes everything:

When you're behind a CDN/proxy, your origin server only sees the CDN's IP addresses at the network layer. The real visitor IP lives in the
X-Forwarded-For HTTP header — which iptables (a layer-3/4 firewall) cannot read.

So iptables -s <botIP> -j DROP blocks nothing. The block has to happen where the real IP is visible: at the CDN, or in nginx, which can parse
the header. nginx is also the perfect place because a 403 there is served instantly — before the heavy application (Magento/PHP) ever boots.

## Step 3: The nginx rules

Two map blocks (HTTP context) read the real client IP from X-Forwarded-For and flag bad traffic:

  # Real client IP behind a CDN is the FIRST token of X-Forwarded-For
  map $http_x_forwarded_for $bad_ip {
      default                       0;
  }

  # Non-compliant / spoofed crawlers
  map $http_user_agent $bad_ua {
      default                       0;
      "~*(Bytespider|PetalBot|MJ12bot|DotBot)"  1;
  }   

Enter fullscreen mode Exit fullscreen mode

Then, inside the server block, drop them before PHP and kill the pagination crawl-trap:

  if ($bad_ip) { return 403; }
  if ($bad_ua) { return 403; }

  # No listing has thousands of pages — anything past page 9 is junk.
  location ~ "/page/[1-9][0-9]+/?$" { return 410; }

Enter fullscreen mode Exit fullscreen mode

A few things worth calling out:

  • 410 Gone beats 404 for crawl-traps — it tells well-behaved crawlers (Googlebot, GPTBot) to permanently drop the URL.
  • Don't block the good bots. I left Googlebot and ClaudeBot untouched (the site's robots.txt allows them) and only blocked the spoofed/abusive ones.
  • Validate before reloading. Always nginx -t first, then nginx -s reload (zero downtime). One unquoted {2,} regex in a location will take your whole site down.

## Step 4: The result

After the reload, I sampled live traffic across a 10-minute window:

  • ~95% of requests were now served as cheap 403/410 responses by nginx, never reaching Magento.
  • PHP/database load from the flood dropped to zero — the server breathed again.
  • Real users, Googlebot, and legitimate crawlers were completely unaffected.
  • Within minutes the attacker's volume began to taper (bots back off once they keep hitting walls).

The durable follow-up is to push the same block up to Cloudflare's WAF (a rule on the hosting ASN + Bot Fight Mode), so the junk is dropped at the
edge and never even reaches the origin as a 403.

## The takeaway

If your store's analytics suddenly balloon with traffic from one country or one network and your bounce rate looks weird, check your server logs
before you celebrate.
A surprising amount of "traffic" is bots that inflate your numbers, burn your server resources, and — on Magento — can flood
your customer table with fake registrations. The fix is usually cheap and fast if you block at the right layer.


I'm Kishan Savaliya, an Adobe-Certified Magento & Hyvä developer. I help store owners with exactly this kind of thing — performance, security, and
clean code. If your store feels slow or you're seeing strange traffic, you can find me and what I do at
kishansavaliya.com.