I've had it with a certain bot that that ignore my robots.txt and
proceeds to mirror my site, several times a day, burning up my upload
bandwidth. They've moved it to 5 different addresses since midnight.
I want to nail the door shut on the first attempted access by these AH's.
Does anyone have a ready made script that can watch my httpd "other" log,
and if a certain name is at the end of the line, grabs the ipv4 src
address as arg3 of the line, and applies it to iptables DROP rules?
Or do I have to invent a new wheel for this?
Basic rules that simplify it somewhat.
1. this is ipv4 only country and not likely to change in the future
decade.
2. the list of offending bot names will probably never go beyond 50, if
that many. 5 would be realistic.
3. the src address in the log is at a fixed offset, obtainable with the
bash MID$ but the dns return will need some acrobatics involving the
bash RIGHT$ function.
4. it should track the number of hits, and after so many in a /24 block,
autoswitch to a /16 block in order to keep the rules file from
exploding.
Any help will be much appreciated. PM's in this case welcome as I can't
see broadcasting our armament against these MF'ers being broadcast on a
public list.
Thanks all.
Cheers, Gene Heskett