The MPLS WG Archive

Cell Relay Retreat>MPLS WG Archive>month:2001-Jan> msg00096



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

Some Queries

  • From: "Cheng-Yin Lee" <leecy@nortelnetworks.com>
  • Date: Tue, 09 Jan 2001 14:04:52 -0500
  • CC: long@pointreyesnet.com, vijay@umbc.edu, mpls@UU.NET
  • Organization: Nortel Networks
  • X-Orig: <leecy@nortelnetworks.com>

Kireeti,

Kireeti Kompella wrote:

> Hi Vijay,
>
> > Then we start to add hierarchy here. One design being the lets mesh the
> > aggregation routers instead of the edge routers. Problem with that design
> > is that now the MPLS domain does not go all the way to the edge, causing
> > the edge to slosh ip traffic flows based on topology changes inside your
> > network, causing your bandwidth settings on LSPS to fail miserably, as
> > traffic may now start traversing LSPs which were previously (thanks to the
> > wonders of SPF) never loaded, thus the automatic TE scripts set their
> > reserved bandwidth to 0.
>
> So you advertise the LSPs into your SPF, and give them a fixed metric.
> Now your SPFs behave: end of slosh.  (Confess!)

Could you elaborate how this scheme works, how this solves the problem
described by Vijay above?

thanks,
cyl



  • References: