IPS HLBR - Version 1.0 can detect malicious traffic using regular expressions Version 1.0 of Hogwash Light BR, released march 5th 2006, brings two interesting new features. The first one is the ability of using regular expressions to detect intrusion attempts and e-mails with virus or phishing. The second is the use of lists with banned words. HLBR is an IPS (Intrusion Prevention System) that reads network traffic in the layer 2 of the OSI model. Since it works like a bridge, it stays in-line in the network topology and doesn't need an IP address. So, HLBR is invisible to attackers. Traffic filtering (including the packets contents) can be done with simple rules. Version 1.0 can use regular expressions to filter the packets. Below is an example of rule with regular expressions: <rule> ip dst(email) tcp dst(25) tcp regex(filename="[^\n]+\.scr") message=(mailvirus-1-re) .scr attach action=virus </rule> In short, all TCP traffic destined to port 25 of the e-mail server will be filtered. If the text: filename="anything_different_of_line_breaks.scr" is found inside the packet, that means there are an attachment .scr in the e-mail (virus). So this packet will suffer the action named 'virus'. This action logs the event, dumps the malicious traffic in tcpdump format and drops the packet. Below is an example of rule against a type of buffer overflow attempt against DNS servers: <rule> ip dst(dns) udp dst(53) udp nocase(|41cd 80c7 062f 6269 6ec7 4604 2f73 6800 89f0 83c0 0889 4608|) message=(dnsattacks-1) tsl bind attack action=action1 </rule> In this case, due to the use of pipe characters (|), HLBR will check the traffic for the hexadecimal sequence given as an attack signature. HLBR lets you use rules for blocking attacks against network servers. In order to fully understand it please read our documentation at http://hlbr.sourceforge.net/ips-en.html - explanations about the IPS concept including charts. HLBR site is at http://hlbr.sourceforge.net. (Translated from Portuguese by André Bertelli - andre (a) bertelli.name)