The MPLS WG Archive

Cell Relay Retreat>MPLS WG Archive>month:2000-Aug> msg00316



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

Fw: MPLS and CAC

  • From: Curtis Villamizar <curtis@avici.com>
  • Date: Wed, 30 Aug 2000 09:54:35 -0400
  • cc: curtis@avici.com, mpls@UU.NET


In message <005801c01240$f6682080$0300000a@pacbell.net>, "Bora Akyol" writes:
> Of course overbooking is a poor man's substitute for calculating the
> equivalent capacity of the traffic that is passing through the tunnel.

wrt equivalent capacity:

How do you propose calculating the equivalent capacity of best effort
Internet traffic?  What happens to your calculation when a peer
network (ie: some other AS not under your control) loses a link and
prefers a different inter-provider interconnect?  What happens to your
equivalent capacity predictions when a hot web site or other event
appears (sports events, major news events, etc)?

There is a great deal of unpredictability in Internet traffic
patterns.

wrt overbooking:

In addition to the MPLS usage that is analogous to equivalent capacity
in ATM, overbooking is there to handle the transient traffic layout
when a fault occurs.  The layout is then gradually optimized.

Curtis