The MPLS WG Archive

Cell Relay Retreat>MPLS WG Archive>month:2000-Dec> msg00011



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

AW: New MPLS charter

  • From: Hummel Heinrich <Heinrich.Hummel@icn.siemens.de>
  • Date: Fri, 1 Dec 2000 16:07:16 +0100
  • Cc: mpls@UU.NET

>    Shahram,
>  
> Although at the beginning of MPLS development it was believed that longest prefix match is slow and label switching will improve it, but future developments showed otherwise. In fact very good algorithms became available that could do a maximum of 3 lookups to do the longest prefix match. So the benefits of MPLS with regard to fast lookup diminished.
	 
	Good point. And I also believe that the existing label transport within BGP4 is an excellent move in the same direction that  I am thinking.
	 
	Give me some time to think about other good reasons. Routing is a large area, maybe anybody else find new reasons to do so !?  :-)

>  
> > 	I mentioned several protocols where to setup labels for 
> > the control plane: Particularly BGP4 and LDP(unsolicited 
> > downstream mode).
> 
> But you still need IP to transport BGP, LDP, ..., don't you?
	  
	Sure.


> Yours, and have a good weekend
> 
	  Heinrich