your project looks nice and usefull. What i am making will
utilze graphs & a sql database. I want to graph all packets droped
the last-hour, today, this-week, this year.

Eventully i would like to establish traffic patterns
so i can wright more abstract rules that what iptables
can do. If anyone is intrested I will post here when its done.

Also i figure I can use the logging in iptables, maybe with snmp,
to make netflows that are aware of state, ports, and ips.

jd
[EMAIL PROTECTED]



On Mon, 2003-03-31 at 22:38, Scott R. Godin wrote:
> Jdavis wrote:
> 
> >> 
> >> ($left,$right) = split(/word/, $sentence);
> >> 
> > 
> > I am trying this but its not working. Im lost :)
> > could someone take a look...
> > 
> > This is the beggining of a scrip to make reports
> > based on droped iptable packets
> 
> I've done something like this with my tailfilter project. 
> 
>         http://www.webdragon.net/tf/
> 
> >From the synopsis: 
> 
> -=-
> 
> Tailfilter is something that started off as a perl one-liner, took on
a life 
> of its own, and swiftly grew out of control because I couldn't stop 
> thinking of ways to enhance the original idea.
> 
> Essentially, it's a logfile filter that reformats the output from the
log 
> and 'pretty-prints' a more legible arrangement that lends itself
better to 
> rapid and/or cursory analysis, as well as offering immediate audible 
> notification of new events as they occur (optional). It additionally
caches 
> the results of DNS lookups, as well as the TCP-based services in 
> /etc/services to speed up the info lookups considerably.
> 
> To use it, simply do one of the following (or similar, depending on
where 
> you store the tailfilter script):
> 
>  tail -f /var/log/messages |tailfilter
> 
>  sudo tail -n 50 -f /var/log/messages |./tailfilter -l 40 -c
--iptables
> 
> -=-
> 
> It works with both ipchains and iptables currently, as well as running
under 
> anything above 5.6.0 (I haven't tried it with 5.005_03 but I don't
believe 
> it will work, there. Anyone willing to play with it?) 
> 
> I'm still pondering ways to suppress flood symptoms in the script,
such as 
> when a nmap comes in masked by several hundred other (bogus) IP's, but
this 
> is a rare enough occurrance, that as long as you have the audible 
> notification on, you can stop/pause the script until the flood passes.
A 
> flood from a single IP address is already cached by the existing
script, 
> but when masked behind hundreds of other bogus non-valid IP's, a flood
of 
> hostname lookups inevitably occurs. 
> 
> I'm also planning on replacing the external `host` lookup with perl's
own, 
> now that I've found out how to do it properly. The only misgiving I
have 
> about that is the notification error messages that one gets from
`host` 
> have different qualities depending on the error, whereas I don't
*think* 
> the internal perl way of doing it, does. Still experimenting in my (I
wish) 
> copious spare time. 
> 
> suggestions are welcome. (as are patches :-)
-- 
jd
[EMAIL PROTECTED]

Bad spellers of the world untie!

"I can't tell if I have worked all my life or 
if I have never worked a single day of my life"
                        Miguel de Icaza













-- 
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to