Re: [SLUG] scans

From: Ian C. Blenke (icblenke@nks.net)
Date: Fri Dec 13 2002 - 10:52:54 EST


On Friday 13 December 2002 10:08, Matt Miller wrote:
> On Fri, 2002-12-13 at 09:05, Todd Robinson wrote:
> > > This would drop most of it:
> > >
> > > iptables -A INPUT -i eth(x) -p tcp --dport 137:139 -j DROP
> >
> > Thanks, need another one for the udp traffic. Also not sure if you need
> > to do another set for source ports. I hadn't thought of consolodating
> > the rules yet, just been adding new ones as I saw them in the logs...
>
> You are absolutely correct. That was a snafu on my part.
> You could use two rules:
> iptables -A INPUT -i eth(x) -p tcp --dport 137:139 -j DROP
> iptables -A INPUT -i eth(x) -p udp --dport 137:139 -j DROP

Or,

iptables -A INPUT -i eth(x) -p tcp --dport 135,137:139,445 -j DROP
iptables -A INPUT -i eth(x) -p udp --dport 135,137:139,445 -j DROP

The proto/port combos you *really* need to worry about are:

        TCP 135 - Microsoft DCE RPC
        UDP 137 - NetBIOS name lookups
        UDP 138 - NetBIOS datagram (\\MAILSLOT browse, network neighborhood)
        TCP 139 - NetBIOS session (SMB over NetBIOS)
        TCP 445 - SMB native (no NetBIOS)

Then there are the other ports, like PPTP, NetMeeting (H.323), and other
miscellaneous listeners that make a Microsoft box a true piece of swiss
cheese.

-- 
- Ian C. Blenke <icblenke@nks.net>

(This message bound by the following: http://www.nks.net/email_disclaimer.html)



This archive was generated by hypermail 2.1.3 : Fri Aug 01 2014 - 19:13:51 EDT