Here are the rules I use for ftp outbound from my internal network :
IPTABLES="/usr/sbin/iptables"
NET_INT="10.0.0.0/24"
IFACE_INET="eth0"
IFACE_INT="eth2"
IFACE_DMZ="eth1"
IFACE_LOC="lo"
$ip="10.0.0.67" # My workstation ip
$IPTABLES -A FORWARD -i $IFACE_INT -p tcp --dport 20:21 -s $ip -m state
--state NEW,ESTABLISHED,RELATED -j ACCEPT
$IPTABLES -A FORWARD -i $IFACE_INT -p tcp --dport 1024: -s $ip
-m state --state NEW,ESTABLISHED,RELATED -j ACCEPT
$IPTABLES -A FORWARD -i $IFACE_INT -p udp --dport 1024: -s $ip
-m state --state NEW,ESTABLISHED,RELATED -j ACCEPT
$IPTABLES -A FORWARD -i $IFACE_INET -p tcp --sport 20:21 -d $ip -m state
--state ESTABLISHED,RELATED -j ACCEPT
$IPTABLES -A FORWARD -i $IFACE_INET -p tcp --sport 1024: -d $ip
-m state --state ESTABLISHED,RELATED -j ACCEPT
$IPTABLES -A FORWARD -i $IFACE_INET -p udp --sport 1024: -d $ip
-m state --state ESTABLISHED,RELATED -j ACCEPT
$IPTABLES -t nat -A POSTROUTING -o $IFACE_INET -p tcp --dport 20:21 -s
$ip -j SNAT --to-source $IP_INET
$IPTABLES -t nat -A POSTROUTING -o $IFACE_INET -p tcp --dport
1024: -s $ip -j SNAT --to-source $IP_INET
Enjoy!
Ray
On Wed, 2002-12-11 at 08:08, louie miranda wrote:
> Hi, thanks for all your replies, etc. But my problem is not yet solved. I
> dont know why?!
> Can someone give me how did they did it?
>
>
>
> --
> thanks,
> louie miranda
>
>
> ----- Original Message -----
> From: "Jozsef Kadlecsik" <kadlec@blackhole.kfki.hu>
> To: "Raymond Leach" <raymondl@knowledgefactory.co.za>
> Cc: <rsterenborg@xs4all.nl>; "'Paulo Andre'" <pandre@darkstar.nom.za>;
> "'louie miranda'" <lmiranda@chikka.com>; "'netfilter'"
> <netfilter@lists.netfilter.org>
> Sent: Tuesday, December 10, 2002 7:22 PM
> Subject: RE: portfw on iptables 2.4 kernel problem.
>
>
> > On 10 Dec 2002, Raymond Leach wrote:
> >
> > > Yes, you do. Port 20 (and/or any other) connections after the control
> > > connection are not 'RELATED, ESTABLISHED' to the control connection.
> > > They are new connections either from the client to the server or vice
> > > versa. You therefore need seperate rules for them.
> >
> > If we are speaking about the data channels of the supported protocols
> > (FTP, IRC and all the other protocols from p-o-m), then this is absolutely
> > false.
> >
> > > Remember connection tracking happens at a pakcet level, i.e all states
> > > relate to packets of a connection, not per protocol.
> >
> > In the case of the supported protocols with additional channels, again,
> > untrue. Please do no spread false info! Why would then the RELATED state
> > exist?
> >
> > > > However, I'm not sure if it's better to split them up into 2 rules :
> > > > iptables -A FORWARD -i eth0 -o eth1 -p tcp --dport 21 -m state --state
> > > > NEW -j ACCEPT
> > > > iptables -A FORWARD -i eth0 -o eth1 -m state --state
> ESTABLISHED,RELATED
> > > > -j ACCEPT
> >
> > Because the destination port of the data channels cannot be port 21,
> > therefore you must use two rules. And because you specify the
> > incoming/outgoing interfaces, you need a third rule for the reply packets
> > as well.
> >
> > Regards,
> > Jozsef
> > -
> > E-mail : kadlec@blackhole.kfki.hu, kadlec@sunserv.kfki.hu
> > PGP key : http://www.kfki.hu/~kadlec/pgp_public_key.txt
> > Address : KFKI Research Institute for Particle and Nuclear Physics
> > H-1525 Budapest 114, POB. 49, Hungary
> >
> >
> >
--
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
( Raymond Leach )
) Knowledge Factory (
( )
) Tel: +27 11 445 8100 (
( Fax: +27 11 445 8101 )
) (
( http://www.knowledgefactory.co.za/ )
) http://www.saptg.co.za/ (
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
o o
o o
.--. .--.
| o_o| |o_o |
| \_:| |:_/ |
/ / \\ // \ \
( | |) (| | )
/`\_ _/'\ /'\_ _/`\
\___)=(___/ \___)=(___/
Attachment:
signature.asc
Description: This is a digitally signed message part
