Date: Tue, 12 Sep 2006 15:51:08 -0400 From: Bart Silverstrim <bsilver@chrononomicon.com> To: freebsd-questions@freebsd.org Subject: forwarding as a gateway, logging certain traffic Message-ID: <7269D41C-C334-44DC-9549-ACB28F79014A@chrononomicon.com>
next in thread | raw e-mail | index | archive | help
This will probably be kind of wordy, but I could use some advice on how to track it. I have a freebsd system acting as a gateway (it's using IP forwarding) so it can act as a web proxy server and filter for the users. It is also filtering incoming email to act as a mail filter between the Internet and our internal Exchange server. The firewall rules used for forwarding information to Squid are rather simple. Ipfw -list gives: ******* 00049 allow tcp from 10.46.255.253 to any 00050 fwd 10.46.255.253,3128 tcp from any to any 80 00100 allow ip from any to any via lo0 00200 deny ip from any to 127.0.0.0/8 00300 deny ip from 127.0.0.0/8 to any 65000 allow ip from any to any 65535 deny ip from any to any ******** The DHCP server then hands out the IP of the FreeBSD server as the gateway address. Something inside our network is infected with a spam-mailing trojan. We now have our PIX firewall set to block all outgoing traffic to port 25 unless it is from our mail server. After setting up a syslog monitor and checking the logs to see if the culprit would appear, what should appear but...the FreeBSD server. Then I smack my forehead; of course it would show up. It's supposed to be the gateway. The trojan computer hits the BSD system and from there hits the PIX...the PIX will be useless to find the culprit. Is there some way to get the FreeBSD system to log machines using port 25 without interfering with the FreeBSD machine's filtering of email function? Or at least make the traffic visible to sniffing with tcpdump or wireshark or ethereal?
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?7269D41C-C334-44DC-9549-ACB28F79014A>