Disclaimer
The Author of this posting offers the information contained within this posting without consideration and with the reader's understanding that there's no implied or expressed suitability or fitness for any purpose. Information provided is for informational purposes only and should not be construed as rendering professional advice of any kind. Usage of this posting's information is solely at reader's own risk.
Liability Disclaimer
In no event shall Author be liable for any damages whatsoever (including, without limitation, damages for loss of use, data or profit) arising out of the use or inability to use the posting's information even if Author has been advised of the possibility of such damage.
Posting
As you ask the questions, neither really.
What happens (assuming we're working with an ISR), an interface FIFO queue "overflows" into the policy-map. If the overflow packets fall into any class queues with WRED defined, then those packets, as they are to be added to the class are subject to the WRED policy. Bandwidth defined for the class has no impact beyond perhaps how many packets are already in the class queue(s), which WRED takes into account when computing average overall class queue depth. Interface congestion's impact is none of the foregoing can happen until the interface congests, but WRED only "cares" about the class packets it applied too. I.e. a particular class could be empty while the interface is filling another class or classes.
