Question: QoS - I Have Trust Issues

Hi all,

We're just rolling out QoS across the network to support our IP Telephony project.

We're happy to trust the diffserv marking on packets that enter our edge switch ports on the Voice vlan, but my question is, 'how do you deal with traffic that may be incorrectly (or maliciously) marked from non-voice edge ports?'

I guess I could re-mark all non-voice vlan traffic with a diffserv of zero or use various other methods of recognising the non-VOIP traffic.

I'm just interested if, in your experience, you have found incorrectly marked traffic to be a problem at all, or if 'down-marking' non-voice user traffic has caused you issues later on?

Many thanks,
Stephen