According to Wil Cooley:
>
>I'd like to know. My impression was that tagged-VLANs could be
>subverted; I think there was a paper written on it and possibly
>exploit code published. But manually-configured VLANs seem secure,
>unless the switch management software is subverted and the VLAN
>configuration mucked with.
>
Errrrrrrrr.... 802.3q works by certain magic bits in the ether net
packet being set. I would hazard a guess that your reference to
"tagged VLANs" you are talking about a port with vlan tagging turned
on, and "manually-configured VLANs" is a port that is a member of a
vlan but does not have tagging turned on. Normally you turn tagging
on on a "trunk" port that is feeding a downstream switch that is
handling multiple vlans. I suspect (but don't know for sure) that
what a switch port that has tagging turned off does with a packet that
has tagging set on it is "implementation defined"... if that packet
gets to a port that does understand the tagging then the port will
just believe the tag and put it on the appropriate vlan. This is why
attempting to build a secure environment on 802.3q tagging is doomed.
--
===============================================================================
Brett Lymn, Computer Systems Administrator, BAE SYSTEMS
===============================================================================
_______________________________________________
Firewalls mailing list
[EMAIL PROTECTED]
http://lists.gnac.net/mailman/listinfo/firewalls
