Re: remarks- source specific multicast for IP

[Hugh Holbrook <holbrook@cisco.com>]

> Date: Wed, 28 Feb 2001 18:52:13 +0100
> From: Ali Boudani <aboudani@irisa.fr>
> 
> Hello,
> just a few remarks
> 1- Elimination of cross-delivery of traffic when two sources
>     simultaneously use the same source-specific destination address.
>     The simultaneous use of an SSM destination address by multiple
>     sources is explicitly supported.
> 
> You are supposing that the delivery is done usually by shared tree??
> but in all cases the use of ssm will not eliminate the cross-delivery
> because it is not a protocol matter more than topology matter??
> isnt true?

We're referring to the property that subscriptions are maintained
independently per (S,G) -- (S1,G1) and (S2,G2) are routed
independently, even if G1 and G2 are the same.  This has nothing to do
with topology -- it's a property of the service model.  Shared trees
are not used by SSM.

> 2- the adding of SSM upgrades to the existing protocol hasnt any impact
> in scalability in building multicast trees??

Not from any effect of making protocol changes to PIM-SM.  The same
mechanisms are used.  SSM applies most naturally to broadcast-like
sessions where there is a single natural source anyway, and there is
no scalability impact for those sessions.  If anything, the protocol
costs go down a little bit for those apps because the overhead of
going to the shared tree is eliminated.

> 3- is the source address should be in teh range 232/8 also or it is just
> for the group ??

No.  Source addresses for SSM are normal IP host addresses.

-Hugh