| bots.d | ||
| conf.d | ||
| .gitignore | ||
| config | ||
| description | ||
| HEAD | ||
| README.md | ||
The Nginx Ultimate Bad Bot, Referer Blocker and Anti DDOS
##The Ultimate Bad Bot and Referer Blocker for Nginx Web Servers including anti DDOS system
Created by: https://github.com/mitchellkrogza
For Nginx Web Server - https://www.nginx.com/
Recommend to saved as: /etc/nginx/conf.d/globalblacklist.conf
Why? .... because all files located in /conf.d/ are automatically loaded by Nginx in the main nginx.conf file.
WHY BLOCK BAD BOTS ?
#####Bad bots are:
- Bad Referers
- Spam Referers
- Spam bots
- Vulnerability scanners
- E-mail harvesters
- Content scrapers
- Aggressive bots that scrape content
- Bots or Servers linked to viruses or malware
- Government surveillance bots
Bots attempt to make themselves look like other software or web sites by disguising their user agent. Their user agent names may look harmless, perfectly legitimate even.
For example, "^Java" but according to Project Honeypot, it's actually one of the most dangerous BUT a lot of legitimate bots out there have "Java" in their user agent string so the approach taken by many to block "Java" is not only ignorant but also blocking out very legitimate crawlers including some of Google's and Bing's and makes it very clear to me that those people writing bot blocking scripts seldom ever test them.
Unfortunately most bot blocker scripts out there are simply copy and pasted from other people's scripts and made to look like their own work. This one was inspired by the one created by https://github.com/mariusv and I contributed to that project but went off into a totally new layout, cleaned it up big time and started from scratch. It is now a completely independent project. It's clean, it works and has been thoroughly tested.
Welcome to the Ultimate Bad Bot and Referer Blocker for Nginx Web Server with and Anti DDOS system
THE BASICS
This bot blocker list is designed to be a global Nginx include file and uses the Nginx map $http_user_agent, map $http_referer and geo $validate_client directives.
This way the .conf file is loaded once into memory by Nginx and is available to all web sites that you operate. You simply need to use an Include statement in an Nginx vhost conf file.
My methods uses no complex regex other than the Name of the Bot. Nginx case matching will do the rest. You can use Regex if you like but it's NOT needed and I proved it by testing with the Chrome extension User-Agent Switcher for Chrome. (handy util and a must for everyone to test these kinds of blocking scripts)
-
The user agent "Aboundex" is found without using "~*Aboundex" ... which means a case insensitive match and is much simpler for anyone to maintain than other lists using complicated and messy Regex patterns.
-
If we have a rule, like "~*Image\ Stripper" and a bot decides to change its User-Agent string to "NOT Image Stripper I Promise" he is picked up regardless and blocked immediately.
I only capitalise bot names in my list for ease of reading and maintenance, remember its not case-sensitive so will catch any combination like "Bot" "bOt" and "bOT".
For those of you who SUCK with Regex this is your saviour !!!
IT'S CENTRALISED:
The beauty of this is that it is one central file used by all your web sites. This means there is only place to make amendments ie. adding new bots that you discover in your log files. Any changes are applied immediately to all sites after a simple "sudo service nginx reload". But of course always do a sudo nginx -t to test any config changes before you reload.
IT IS TINY AND LIGHTWEIGHT
The file is tiny in size. At the time of this writing and the first public commit of this the file size including all the commenting "which nginx ignores" is a mere 58 kb in size. It is so lightweight that Nginx does not even know it's there. It already contains hundreds of entries.
IT IS ACCURATE AND IS FALSE POSITIVE PROOF
Unlike many other bad bot blockers out there for Nginx and Apache where people simply copy and paste lists from others, this list has been built from the ground up and tested thoroughly and I mean thoroughly. It comes from actual server logs that are monitored daily and there are at least 3-10 new additions to this file almost daily.
It has also been throughly tested for false positives using months of constant and regular testing and monitoring of log files.
All web sites listed in the bad referers are checked one by one before they are even added. Simply copying anything that look suspicious in your log file and adding it to a blocker like this without actually seeing what it is first .... well it's foolish to say the least.
DROP THEM AND THAT'S IT
Nginx has a lovely error called 444 which just literally drops the connection. All these rules issue a 444 response so if a rule matches, the requesting IP simply get's no response and it would appear that your server does not exist to them or appears to be offline.
RATE LIMITING FUNCTIONALITY BUILT IN
For bot's or spiders that you still want to allow but want to limit their visitation rate, you can use the built in rate limiting functions I have included. The file is extensively commented throughout so you should figure it out otherwise simply message me if you are having problems.
(Read configuration instructions below thoroughly)
FEATURES:
- Extensive Lists of Bad and Known Bad Bots and Scrapers (updated almost daily)
- Alphabetically ordered for easier maintenance
- Commented sections of certain important bots to be sure of before blocking
- Includes the IP range of Cyveillance who are known to ignore robots.txt rules and snoop around all over the Internet.
- Your own IP Ranges that you want to avoid blocking can be easily added.
- Ability to add other IP ranges and IP blocks that you want to block out.
Usage: recommended to be saved as /etc/nginx/conf.d/globalblacklist.conf
####PLEASE READ: The configuration instructions below !!!!
WARNING:
Please understand why you are using this before you even use this. Please do not simply copy and paste without understanding what this is doing. Do not become a copy and paste Linux "Guru", learn things properly before you use them and always test everything you do one step at a time.
ANOTHER WARNING:
Make sure to add all your own IP addresses the white list section near the bottom of the globalblacklist.conf file !!!!
MONITOR WHAT YOU ARE DOING:
MAKE SURE to monitor your web site logs after implementing this. I suggest you first load this into one site and monitor it for any possible false positives before putting this into production on all your web sites.
Also monitor your logs daily for new bad referers and user-agent strings that you want to block. Your best source of adding to this list is your own server logs, not mine.
Feel free to contribute bad referers from your own logs to this project by sending a PR. You can however rely on this list to keep out 99% of the baddies out there.
HOW TO MONITOR YOUR LOGS DAILY (The Easy Way):
With great thanks and appreciation to https://blog.nexcess.net/2011/01/21/one-liners-for-apache-log-files/
To monitor your top referer's for a web site's log file's on a daily basis use the following simple cron jobs which will email you a list of top referer's / user agents every morning from a particular web site's log files. This is an example for just one cron job for one site. Set up multiple one's for each one you want to monitor. Here is a cron that runs at 8am every morning and emails me the stripped down log of referers. When I say stripped down, the domain of the site and other referers like Google and Bing are stripped from the results. Of course you must change the log file name, domain name and your email address in the examples below. The second cron for collecting User agents does not do any stripping out of any referers but you can add that functionality if you like copying the awk statement !~ from the first example.
Cron for Monitoring Daily Referers on Nginx
00 08 * * * tail -10000 /var/log/nginx/mydomain-access.log | awk '$11 !~ /google|bing|yahoo|yandex|mywebsite.com/' | awk '{print $11}' | tr -d '"' | sort | uniq -c | sort -rn | head -1000 | mail - s "Top 1000 Referers for Mydomain.com" me@mydomain.com
Cron for Monitoring Daily User Agents on Nginx
00 08 * * * tail -50000 /var/log/nginx/mydomain-access.log | awk '{print $12}' | tr -d '"' | sort | uniq -c | sort -rn | head -1000 | mail -s "Top 1000 Agents for Mydomain.com" me@mydomain.com
CONFIGURATION:
####First Step:
sudo mkdir /etc/nginx/bots.d
- copy the blockbots.conf file into that folder
- copy the ddos.conf file into the same folder
####Second Step:
sudo nano /etc/nginx/nginx.conf
#####Add the following rate limiting zones
-
limit_req_zone $ratelimited zone=flood:50m rate=90r/s; -
limit_conn_zone $ratelimited zone=addr:50m;
PLEASE NOTE: The above rate limiting rules are for the DDOS filter, it may seem like high values to you but for wordpress sites with plugins and lots of images, it's not. This will not limit any real visitor to your Wordpress sites but it will immediately rate limit any aggressive bot. Remember that other bots and user agents are rate limited using a different rate limiting rule at the bottom of the globalblacklist.conf file.
####Third Step:
Open a site config file for Nginx (just one for now) and add the following lines
-
include /etc/nginx/bots.d/blockbots.conf; -
include /etc/nginx/bots.d/ddos.conf;
####Fourth Step:
Make sure to edit the globalblacklist.conf file near the bottom there is a section to whitelist your own IP addresses. Please add all your own IP addresses there before putting this into operation.
####Last Step:
sudo nginx -t (make sure it returns no errors and if none then) sudo service nginx reload
IT FORKING WORKS !!!
##WARRANTY OR LICENSE
- This is free to use and modify as you wish.
- No warranties are express or implied.
- You use this entirely at your own Risk.
- Fork your own copy from this repo and feel free to change it to your needs or contribute to it.