Re: 3550 qos dscp mutation map question

From: robbie (robbie@packetized.org)
Date: Tue Oct 26 2004 - 21:48:11 GMT-3


 From how I read the explanations, I understand it to mean that if the
switch is configured to trust-cos on that port, that it marks the packet
with an internal-DSCP value based on the CoS, whatever the CoS may be.
The internal-DSCP value is derived from a 'qos map' of some sort,
whether it's 'mls qos dscp-mutation' or some other map.

Having said that, if your question is 'does the cos-to-dscp map operate
on the frame prior to the dscp-mutation map changing the internal dscp
value of the frame?' then I believe that the answer is no, because you
have to have trust 'mls qos trust dscp' configured on the interface
instead of 'mls qos trust cos'.

I think my understanding of it is right - someone confirm/deny?

cheers,
robbie

(ps: 'cisco catalyst qos: quality of service in campus networks' is a
great book on the subject)

Edwards, Andrew M wrote:
> I haven't been able to drum up a clear answer in my mind on this one...
>
> If I set a port to trust cos, I know that the default COS (or 0) is
> applied to unmarked traffic. What I don't know is if the switchport
> marks the IP packet with the DSCP value to the cos-to-dscp map?
>
> Or, is the DSCP mapping just another way the switch could filter/act on
> the packet?
>
> Any help with this is appreciated.
>
> andy
>
> _______________________________________________________________________
> Subscription information may be found at:
> http://www.groupstudy.com/list/CCIELab.html



This archive was generated by hypermail 2.1.4 : Sat Nov 06 2004 - 17:11:53 GMT-3