Randy, do you know if this exploit applies to site-to-site tunnels?  The docs I have read don't specify remote access or site-to-site (or both).  For site-to-site would an attacker have to spoof a valid peer address?  Thanks.