The MPLS WG Archive

Cell Relay Retreat>MPLS WG Archive>month:2000-Jul> msg00401



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

mpls-in-ip

  • From: "Loa Andersson" <loa.andersson@nortelnetworks.com>
  • Date: Wed, 26 Jul 2000 15:45:02 +0200
  • CC: tom@ennovatenetworks.com, mpls@UU.NET
  • Organization: Nortel Networks
  • X-Sybari-Space: 00000000 00000000 00000000

Neil and Tom,

we been touching this before, but it is my view that LSPs (in the 
vanilla sense) are CL, ER-LSPs loosely routed are also CL, it takes
certain measures ER routed and pinned down, to make them look and
behave as CO entities.

It might be argued that those CO look alikes are the only LSPs we
want, but that is another story.

/Loa

neil.2.harrison@bt.com wrote:
> 
> Hi Tom,
> 
> Just an observation but......if we allow LSPs (ostensibly a CO entity) to be
> a client layer of an IP server layer (a CNLS entity) will this not present
> some *interesting* problems?  For example:
> -       I guess IP would also be a client of the LSP.....and in theory one
> could recurse this relationship many times;
> -       QoS control would be a issue (esp if there was no control over how
> many times the recursion could exist......this is not just as a single
> instance, but also in a concatenated sense);
> -       ........and so would defect handling and OAM in general.
> 
> Have these factors been considered?
> 
> neil

-- 

Loa Andersson
Director Routing Architecture Lab, EMEA
St Eriksgatan 115A, PO Box 6701
113 85 Stockholm, Sweden
phone: +46 8 50 88 36 34,   mobile + 46 70 522 78 34
e-mail: loa.andersson@nortelnetworks.com


  • References: