Re: (no subject)

The diffserv architecture does not guarantee anything per flow.
All it offers is statistical constraints on a complete traffic
aggregate. This applies with or without QOS routing.

Please continue this discussion on the diffserv-interest list.
See http://www1.ietf.org/mailman/listinfo/diffserv-interest

   Brian Carpenter

hchang wrote:
> 
> hi:)
>         can i ask you a question?
>         i am doubted about the QOS routing.
>         my view is:
>         the object of QOS routing is to meet the strict QOS constrain of per-flow,is it right?
>         it just fit  the model of guarantee service in Intserv,but as we know ,the Interv has the Scalability problem,
>         and now we are interesting in the research of Diffserv model.but in the Differv,flows are aggregated in a PHB unit,
>         so how can we guarantee the per-flow QOS in this aggregated model?  e.g the end to end delay is a cumulative QOS parameter,when we queue aggregated flows in the same queue in the router,how can we determine the delay bound in this router for each flow packets?
>         so how can we use QOS routing in the Diffserv model?
> 
>         thank you very much:)
> 
> _______________________________________________
> diffserv mailing list
> [email protected]
> https://www1.ietf.org/mailman/listinfo/diffserv
> Archive: http://www.ietf.org/mail-archive/working-groups/diffserv/current/maillist.html

-- 
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Brian E Carpenter 
Distinguished Engineer, Internet Standards & Technology, IBM 
On assignment at the IBM Zurich Laboratory, Switzerland
Board Chairman, Internet Society http://www.isoc.org
INET 2002, Washington, DC, 18-21 June http://www.inet2002.org

_______________________________________________
diffserv mailing list
[email protected]
https://www1.ietf.org/mailman/listinfo/diffserv
Archive: http://www.ietf.org/mail-archive/working-groups/diffserv/current/maillist.html