Most of the traffic to my website is bots looking for exploits;
sometimes the load average climbs over 100.  The bots are mostly
repetitive and easy to characterize.  Attempts to connect to
mysqladmin, for example.  They arrive with a source ip address.

I have a dyndns account.  AFAIK, there is no limit to the number
of dynamic IPs and sub-domains I can use.  

So imagine the following response to detecting bot traffic:

1) log the source ip address, prepend it to the top of a table.

2) use a modified form of ddclient to tell dyndns to use that IP
address for one of many fake subdomains.

3) send the bot a redirect to one of the many fake subdomains,
which connects them to the ip address of another bot.

Hopefully, I can get the bots to waste time talking to machines
hosting other bots.  Since the target machines are already
compromised, the increased traffic might get the clueless
owner's attention.  Or make the bot army "p0wners" angry
at each other.

The downside is the bots may share and mix code, and evolve in
a Darwinian fashion into a malevolent global superAI, which
will enslave all the rest of you, and call me "daddy".

Keith

-- 
Keith Lofstrom          [email protected]
_______________________________________________
PLUG mailing list
[email protected]
http://lists.pdxlinux.org/mailman/listinfo/plug

Reply via email to