The MPLS WG Archive

Cell Relay Retreat>MPLS WG Archive>month:2000-Aug> msg00241



[Date Prev][Date Next][Thread Prev][Thread Next]  
  [Date Index][Thread Index][Author Index][Subject Index]

MPLS and CAC

  • From: David Charlap <david.charlap@marconi.com>
  • Date: Mon, 21 Aug 2000 18:28:15 -0400

Erickson Trejo-Reyes wrote:
>
> While reading about MPLS specification (all my previous experience is
> on ATM networks), I have always found statements like: - a path is
> selected and resources are reserved in every node it traverses... Does
> that imply some sort of call admission control?

Any time resources are reserved (on any system, not just MPLS), there
must be admission control.  Otherwise, a router would be unable to
actually guarantee the resources it claims to be reserving.

This is done as a part of the signalling code (either RSVP-TE or
CR-LDP).

Of course, when tunnels are signalled for best-effort traffic, then no
admission control is necessary.

> (Many have told me that CAC is not considered for MPLS, at least for a
> telephony service on top of it). My question lies in the fact that, in
> order to reserve resources, first, some kind of effective bandwidth
> must be computed (am I wrong?).

It is possible to run telephony services over best-effort tunnels
(without reserving any resources).  This is the way internet-based phone
services run today.  It's not ideal, but it seems to be good enough for
many customers.

> Next, if resources are reserved, that means that in some cases the
> requested resources couldn't be granted (because they would be already
> reserved).

Obviously.  I think you may have misunderstood those who you were
talking to.

I don't think they meant that admission control isn't possible under
MPLS.  I think they meant that it is not used for best-effort traffic,
which is what everybody is signalling with MPLS today.

In the future, when people start signalling QoS-bound tunnels, there
will have to be some form of admission control, since it is always
possible for a resource to be exhausted.

> Finally, if a CAC mechanism is not due to be applied (for those who
> say so), how will a network warrant some bounds on performance metrics
> like delay and packet losses.

It won't be able to guarantee them obviously.

Those implementations that don't use any form of admission control will
either be unable to gurantee a reservation, or some form of
administrative action will be required to prevent the resources from
becoming oversubscribed.

-- David


  • Follow-Ups:
  • References:
    • MPLS and CAC
      • From: "Erickson Trejo-Reyes" <eenet@electeng.leeds.ac.uk>