On Friday, 2001/10/12 at 12:25 CET, "Bruno Fernandes" 
<[EMAIL PROTECTED]> wrote:
> -----Original Message-----
> From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]]
> -is Pix able to identify/block IP-spoofing?
> Yes

The only way I know of for a Pix, or any type of box, to identify spoofing 
is by filters that know which source addresses are permissible for 
incoming traffic on an interface.  With some Cisco IOS versions (not 
available on Pix) you can use "ip verify unicast reverse-path" - a very 
nice trick that uses the box's routing table to determine whether to allow 
a source address.  The address, when used as a destination, must be routed 
out the same interface it arrived on; else it gets discarded.  Boxes 
without such a nice control have to have hardcoded access lists which 
statically permit only the source addresses that the admin thinks should 
be arriving on an interface.

But that only works for interfaces which don't have a default route and 
that don't use dynamic routing (which is not, unfortunately, an issue on 
the Pix).  If the Pix is connected to the Internet typically its outside 
interface will be configured with a default route.  There is no way it can 
identify or block spoofed traffic arriving at such an interface (but it 
can, if so configured with access lists, block address ranges that it 
knows should never arrive on that interface, such as rfc1918 addresses and 
its own inside address ranges).

My answer to the original question is that Pix cannot identify spoofing 
(but it can statically filter by address, which may be used to block 
spoofing in some cases).

Tony Rall
_______________________________________________
Firewalls mailing list
[EMAIL PROTECTED]
http://lists.gnac.net/mailman/listinfo/firewalls

Reply via email to