The MPLS-OPS Archive

Cell Relay Retreat>MPLS-OPS Archive>month:2005-Jan> msg00022



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

Re: Re: Layer 2 MPLS VPN

  • From: Yakov Rekhter <yakov@juniper.net>
  • Date: Mon, 17 Jan 2005 06:42:28 -0800
  • cc: raszuk@cisco.com, sthaug@nethelp.no, Andrew.Frazer@nz.unisys.com, mpls-ops@mplsrc.com
  • Resent-Date: Mon, 17 Jan 2005 10:17:07 -0500
  • X-SpamProbe: GOOD 0.0000000 f8a48bd8bb48a559e50dcd5533e73c63

Irwin,

> The need for IP multicast, which is tough to do over a 2547 service,

Current IP multicast support in VPLS has its own set of problems,
mostly to do with the fact that it uses ingress replication as the
only way to handle IP multicast.

While ingress replication is quite viable in the scenarios where the 
bandwidth of the multicast traffic is low or/and there is a sparse 
distribution of receiving PEs, such that the number of replications 
performed by the ingress PE on each outgoing interface for a particular 
customer multicast data packet is small, in other scenarios viability
of ingress replication is (highly) questionable. And other scenarios
do exist in real life.

This is not to say that the current support for IP multicast in 2547 
does not have its own problems - it does.

The bottom line is that neither current support for IP multicast in VPLS,
nor in 2547 VPNs is good enough.

Yakov.

-------
The MPLS-OPS Mailing List
Subscribe/Unsubscribe:  http://www.mplsrc.com/mplsops.shtml
Archive: http://www.mplsrc.com/mpls-ops_archive.shtml