All things related to multicast

[Gordon Rogier <>]

Ok Dave and Steve.  Thanks for the clarification and I understand.

*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*
Gordon Rogier                        

            
Network Engineer                     785-864-0381wk 785-550-4468 cell
Great Plains Network                 785-864-9330 FAX

On Thu, 1 Jul 1999, David Meyer wrote:

> 
>       What Steve said. It's private G, not private S that 
>       we need if we want to do this kind of scoping.
> 
>       Dave
> 
> According to Wallace, Steven S:
> > 
> > > 
> > > It would seem to me that if you want to keep something from 
> > > leaking out of
> > > I2, you need to put up a private "source" address range.  Say assign a
> > > subset of this source range to each I2 member via a glop 
> > > agorithm.  Using
> > > this method, mbgp and msdp scoping could be applied to deny 
> > > this private S
> > > information from leaking to commodity I1 service (which is probably
> > > already configured intrinsitically by not having already 
> > > defined these in
> > > a route-map permit type config) and thusly prevent (S,G) 
> > > state from being
> > > built.
> > > 
> > 
> > I would think blocking sources would break RPF, potentially causing 
> > routing
> > loops.  If there's an (*,g), the packet is going to get forwarded, just 
> > not
> > via a source based tree...and if there's not mbgp information preventing 
> > the
> > packet from looping back to the source, not good.
> > 
> > Steven
> > 
> 
>